{ "schema_version": 1, "shared_fields": { "healthcare.dental_vision_monthly": { "reference_basis": true, "source": "Basis declaration only; the 12 cities with a per-city citation carry it in entries below (Verbraucherzentrale Zahnzusatzversicherung for Berlin, Moneyland/Comparis for Zurich, Independer for Amsterdam, Folktandvarden for Stockholm, Medi-Cover for Copenhagen, JDA practice rates for Tokyo, Canada Life/Sun Life individual extended health for Vancouver, KFF 2025 EHBS for the US cities, Ljubljana private price lists). The remaining cities were set from local dental-plan or price-list levels without a recorded citation.", "last_updated": "2026-08-08", "confidence": "low", "note": "REFERENCE BASIS (2026-08-08, cluster dental-scaling-0808): routine dental and vision spend for ONE ADULT per month, in the city local currency, whether it is bought as a supplementary policy or paid at the counter. The engine multiplies it by ADULTS (1 or 2, from state.has_partner), never by household size, because child dental is free or heavily subsidised to 18 across most universal systems. USA IS THE ONE EXCEPTION and it is deliberate: US cities store the SINGLE SHARE of the dental+vision premium and the family dental and ortho above that share is folded into annual_out_of_pocket_by_family_size, so the US line is billed flat and the two columns stay complementary. Multiplying it would double count. Before this declaration the field was billed flat everywhere, so a household of six paid one adult's dental, and in the tax-funded Nordics that flat line was the majority of a family's whole healthcare number. Confidence is low because 166 of 178 cities have no citation and the levels disagree within some countries (Oslo 700 vs the other three Norwegian cities 300; Stockholm 350 vs the other three Swedish cities 200), which is an authorship split between research passes rather than a unit split. Re-levelling those is open work and is NOT a basis question. See docs/superpowers/research/2026-08-08-dental-scaling-and-copy.md." }, "housing.utilities_monthly": { "source": "Calibration: Stadt Leipzig KBU 2025 Nebenkosten wedge 2.93 EUR/sqm; DMB Betriebskostenspiegel 2024 (2.67 avg / 3.68 all-in EUR/sqm, heating 1.32); Ofgem TDCV bands + July 2026 price cap (standing charges ~GBP 318/yr); US aggregators (Apartment List 2025-26 electricity by size, flat water/trash).", "last_updated": "2026-08-07", "confidence": "medium", "note": "REFERENCE BASIS (2026-08-07): every city's utilities_monthly is the 2-BEDROOM reference figure (~83 sqm, the Numbeo 85-sqm basket basis most values were sourced on). The engine derives other dwelling sizes via utilitiesScale(bedrooms) in app.js: below the reference a 0.35 standing-charge floor binds (scale 0.64 studio / 0.80 1BR), above it utilities track floor area on the 37/58/83/111 sqm ladder (1.34 at 3BR, +28 sqm per further bedroom). The stored scalar is no longer charged flat to every household size. Internet/telecom is the separate internet_phone_monthly field and stays size-independent. Calibration: docs/superpowers/research/2026-08-07-utilities-nuance.md. Its per-city follow-up flags were RESOLVED 2026-08-08 (Munich 250 -> 330, Sofia 180 -> 150, Dusseldorf 303 confirmed, Leipzig 306 stands; five further German cities flagged low on the 83-sqm basis): docs/superpowers/research/2026-08-08-utilities-values.md" }, "transit.car_dependent": { "source": "US Census ACS 2023 commuting mode share + household vehicle availability by metro area; Statistics Canada Census 2021 commuting; ABS Census 2021 method of travel to work; Stats NZ; national household travel surveys for the Gulf states and South Africa. Cyprus: the 2026 national modal split of 67 percent private car against 5 percent bus, with public transport at about 3 percent of greater-Nicosia journeys, no rail anywhere on the island, and a bus commute of 50 minutes against 23 by car (Cyprus Mail, 2026-08-05).", "last_updated": "2026-07-31", "confidence": "medium", "note": "Boolean picker flag, not a cost. Set to true when a typical household in the metro needs a car for daily life: metro-wide transit commute share under roughly 20 percent AND household vehicle ownership around 85 percent or higher. True makes the calculator default that city to transit.annual_car_cost_avg; absent (the default) means the metro supports car-free daily life, so the default is transit.monthly_pass x 12. Both cities in a comparison read the same rule, and the user can flip either one under Add more details. The flag changes only which of the two existing numbers is picked; it does not change either dollar value. Currently true for 27 US metros (all but New York, the only US metro clearing the transit-share line on an MSA basis), Calgary and Waterloo, Adelaide/Brisbane/Perth, Auckland, the Gulf metros (Abu Dhabi, Doha, Dubai, Kuwait City, Riyadh), Cape Town and Johannesburg, and the three Cyprus cities (Nicosia, Limassol, Larnaca), which are the first EU metros on this list: Cyprus runs no rail at all, its bus share is 5 percent nationally, and it carries one of the highest car-ownership rates in the EU at over 600 cars per 1,000 inhabitants." }, "housing.homeowners_insurance_pct_of_value": { "source": "Per-country research 2026-08-07: regulators and insurer associations (NAIC-class US state tables, Finans Norge, MNB lakasbiztositas index, UNESPA, ANIA, GDV-adjacent bands, DASK official tariff, GIROJ, PIAM, OIC, OJK+MAIPARK, CMF bid data, INS, Quashed/Finity) with reputable aggregators where no association average exists. Full per-country tables with URLs: docs/superpowers/research/2026-08-07-homeowners-insurance.md and the eight 2026-08-07-hoi-*.md region files.", "last_updated": "2026-08-07", "confidence": "medium", "note": "ANNUAL owner premium as a DECIMAL FRACTION of home market value (0.005 = 0.5 pct/yr); the owner-mode housing cost applies it to median_home_price_by_bedroom at the chosen bedroom count. Basis is the cover an owner-occupier customarily buys (building/structure plus contents where the standard local policy bundles it, including mandatory catastrophe schemes: DASK, PAD, Consorcio, NHC/EQC levies, jishin hoken). One national rate per country EXCEPT USA and Canada, which carry per-metro fractions (US premiums track rebuild cost plus peril risk, not market value; Florida/Texas wind-hail metros run 1.6x-2.3x the US national fraction while coastal high-price metros run under it). Conservative bias throughout: the higher end of every sourced range, rounded up, so owner costs are if anything overstated. Double-count check 2026-08-07: the model carries no condo/HOA/strata fee line anywhere, so charging the full standalone owner premium stacks on nothing; in countries where a building master policy customarily rides inside association fees (Nordic BRF/borettslag, condo markets HK/SG/KR, ES comunidad) the standalone detached-house rate was still used as the conservative anchor. Rate floor 0.0002 applied where computed fractions fell below it (China, Taiwan, Singapore, marginally South Korea)." }, "non_cash.public_holidays_substitute_day": { "source": "Per-country statute, primary-sourced 2026-08-08 in docs/superpowers/research/2026-08-08-holiday-substitution-shard-{A,B,C}.md and recorded per profile in data/_holiday_rules.json. Examples: 5 U.S.C. 6103(b) and EO 11582 s.3(a) for the USA; Banking and Financial Dealings Act 1971 s.1 for the UK; Loi du 4 janvier 1974 art. 6 for Belgium; Helligdagsloven LOV-1995-02-24-12 for Norway, which has no such rule.", "last_updated": "2026-08-08", "confidence": "high", "note": "DERIVED FIELD. Copied from the city's holiday profile in data/_holiday_rules.json by scripts/derive-holiday-effective.mjs. Where a substitution reading was arguable the profile takes the less favourable side and carries confidence: medium plus a note saying so." }, "non_cash.public_holidays_effective": { "source": "Computed, not researched: scripts/derive-holiday-effective.mjs brute-forces the 400-year Gregorian cycle (2001-2400) over the holiday rule list in data/_holiday_rules.json and counts landings inside the working week. Each rule list cites the statute it was built from.", "last_updated": "2026-08-08", "confidence": "high", "note": "DERIVED FIELD, do not hand-edit. tests/holiday-effective.test.mjs re-runs the derivation and fails if a committed value differs, which is what stops this number drifting the way a hand-typed one would. Background: docs/superpowers/research/2026-08-08-public-holidays-effective-build.md." }, "non_cash.commute_minutes_avg": { "reference_basis": true, "source": "Basis declaration only; each city carries its own citation in entries below. Anchor definitions for the standard: US Census Bureau, ACS subject definitions for travel time to work (https://www.census.gov/quickfacts/fact/note/US/LFE305221), tables B08013 / B08303 / S0801; Statistics Canada, commuting duration, Census of Population (https://www150.statcan.gc.ca/n1/pub/14-28-0001/2023001/article/00003-eng.htm); Eurostat qoe_ewcs_3c3, \"Mean duration of commuting time one-way between work and home\" (https://ec.europa.eu/eurostat/databrowser/view/qoe_ewcs_3c3/default/map?lang=en).", "last_updated": "2026-08-08", "confidence": "high", "note": "REFERENCE BASIS (2026-08-08, cluster commute-unit-0808): every city commute_minutes_avg is the mean ONE-WAY, DOOR-TO-DOOR travel time to work in whole minutes, averaged over ALL employed commuters regardless of travel mode, for the METRO AREA, with people who work from home excluded from the average. Before this sweep the column mixed one-way, door-to-door and round-trip figures, which differ by roughly 2x, and 166 of 178 cities carried no citation. Full per-dimension reasoning is in data/_national_fields.json _definitions and docs/superpowers/research/2026-08-08-commute-unit-standardization.md. Moovit Public Transport Index figures may corroborate a value but may not be its headline source, because they cover public-transport riders only and run long in car-dominated metros. Traffic-congestion indices (TomTom, INRIX) measure road delay and are a different quantity entirely. Any conversion from a published round-trip or in-vehicle figure is recorded on the individual city entry." }, "healthcare.annual_out_of_pocket_by_family_size": { "reference_basis": true, "source": "Basis declaration only; each city carries its own derivation in entries below. Per-country statutory structure and realized-average sourcing is in docs/superpowers/research/2026-08-08-healthcare-oop-basis-unification.md and the per-country batch files under 2026-08-08-healthcare-oop/. The US rows derive from docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md (KFF 2025 EHBS + Peterson-KFF realized OOP trended to 2026 + Milliman MMI 2025 upper anchor).", "last_updated": "2026-08-08", "confidence": "medium", "note": "REFERENCE BASIS (2026-08-08 unification, amended the same day by cluster dental-scaling-0808). THIS IS THE ONE PLACE THIS FIELD IS DEFINED; the 178 per-city entries below used to each restate it, which is how it came to say two different things at once. The value is the REALIZED AVERAGE annual out-of-pocket health spend of a HOUSEHOLD of n people, in the city local currency, NET of reimbursement, EXCLUDING insurance premiums (premium_monthly_by_family_size) and EXCLUDING routine dental and vision (dental_vision_monthly). It is never a statutory maximum; a country cap is a ceiling CHECK on this number, never the number itself. Row 1 is the single adult and equals annual_out_of_pocket_avg. THE USA IS THE ONE DECLARED EXCEPTION to the dental clause, and it is deliberate: US rows also carry a dental and vision OUT-OF-POCKET margin of about USD 400 at row 1 and USD 400-450 at row 4, because the 2026-07-17 US recalibration chose not to add a dental-by-family-size key and absorbed the family excess here instead (plan lines 72, 129, 131). That is not a double count: the US dental_vision_monthly is the dental and vision PREMIUM single share while this column is out-of-pocket spend, and app.js bills the US dental line flat for exactly this reason while every other country scales it by adults. Before the amendment the exclusion was declared to hold everywhere, which was false for 28 of 178 cities and was the same one-column-two-meanings defect the unification existed to remove. See docs/superpowers/research/2026-08-08-dental-scaling-and-copy.md." }, "tax.retirement_deduction_cap": { "reference_basis": true, "source": "Per-country primary sources are on each city entry below. Full 70-country research, confidence split and predicted impact: docs/superpowers/research/2026-08-11-retirement-deduction-caps.md. Adjudication rules: docs/corrections-ledger.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "REFERENCE BASIS. The statutory ceiling on the PRE-TAX retirement deduction available to ONE EMPLOYEE from their own earned income, per year, in the city's local currency. It answers \"how much of a voluntary retirement contribution may this employee take off the income-tax base\", and nothing else. What it is NOT: not a tax CREDIT (Belgium, Portugal, South Korea, the Philippines, Pakistan, Israel and Hungary all relieve retirement saving through a credit or a state bonus, so their value is 0 here and the benefit is unmodeled); not an EMPLOYER contribution ceiling; not a limit on what may be PAID IN where the two differ; and not a relief against capital income (Finland). Shape and evaluation order are documented at retirementDeductionCap() in app.js. An ABSENT key means UNCAPPED, which is what every city carried until 2026-08-11 and is why the engine deducted the full gross x retirement_contrib_pct in all 70 countries, including ones that grant an employee no deduction at all. That error only ever understated tax, so it flattered every non-US country. Values applied 2026-08-11 from docs/superpowers/research/2026-08-11-retirement-deduction-caps.md (70 countries, primary-sourced, per-country confidence recorded on each city entry below). Where the value had judgment latitude the LESS favourable reading was taken, per the conservative value bias. ONE DECLARED EXCEPTION: INDONESIA carries no key at all and is deliberately uncapped. PMK 168/2023 Pasal 10 lets a Pegawai Tetap deduct iuran pensiun / JHT-THT with no rupiah ceiling in the tax law, while the pension-law track (KMK 343/1998, solving to 7.5% of PhDP) was superseded by POJK 5/2017 and POJK 27/2023 and the current article could not be read. Writing 0 or 7.5% there would invent a ceiling, which is the failure mode this sweep exists to correct (owner decision, 2026-08-11). KNOWN AND SEPARATE, do not read this field as fixing it: 33 of 70 country branches compute social insurance off income that is already net of this deduction, so a larger contribution still shrinks the SI charge. A max of 0 closes that leak as a side effect; every nonzero row inherits it." } }, "entries": { "seattle": { "tax.federal_brackets": { "source": "IRS Rev. Proc. 2024-40 (2026 projections)", "last_updated": "2026-05-12", "note": "Single filer brackets" }, "tax.federal_brackets_married": { "source": "IRS Rev. Proc. 2024-40", "last_updated": "2026-05-12", "note": "Married filing jointly" }, "tax.state_income_tax_rate": { "source": "WA Dept of Revenue", "last_updated": "2026-05-12", "note": "Washington has no state income tax" }, "tax.social_security_rate": { "source": "SSA 2026", "last_updated": "2026-05-12" }, "tax.social_security_cap": { "source": "SSA 2026 wage base limit", "last_updated": "2026-05-12" }, "tax.medicare_rate": { "source": "IRS payroll tax tables", "last_updated": "2026-05-12" }, "tax.state_disability_insurance_rate": { "source": "WA Cares Fund premium rate https://wacaresfund.wa.gov/", "last_updated": "2026-07-31", "confidence": "high", "note": "WA Cares long-term-care premium, 0.58% of gross wages, employee-paid, explicitly NOT subject to the Social Security cap, so it uses the uncapped slot. Added 2026-07-31: Washington employees were being modeled with zero state wage levy while California was charged 1.3% SDI, which flattered Seattle in every Seattle-vs-California comparison" }, "tax.state_payroll_premium_rate": { "source": "WA ESD, Paid Family & Medical Leave premium rate increases to 1.13% in 2026 https://esd.wa.gov/about-us/news-release/2025/paid-family-medical-leave-premium-rate-increases-113-2026", "last_updated": "2026-07-31", "confidence": "high", "note": "WA PFML 2026 total premium 1.13%, of which the employee pays 71.43% = 0.8072%" }, "tax.state_payroll_premium_cap": { "source": "WA ESD / SSA 2026 contribution and benefit base", "last_updated": "2026-07-31", "confidence": "high", "note": "PFML premiums apply only up to the Social Security wage cap ($184,500 in 2026); max employee contribution $1,489" }, "tax.sales_tax_rate": { "source": "Seattle combined state+local 10.25%", "last_updated": "2026-05-12" }, "tax.property_tax_rate_pct": { "source": "King County Assessor avg effective rate", "last_updated": "2026-05-12" }, "housing.median_rent_by_bedroom": { "source": "Zillow Rent Index, Seattle metro", "last_updated": "2026-Q1" }, "housing.rent_percentiles": { "source": "Census ACS 2024 1-Year, table B25063 gross-rent distribution (metro/CBSA), data.census.gov; P25/P75 interpolated from dollar buckets, interpolated P50 matched published B25064 median (research 2026-07-17 §2). Rounded outward to 2dp per conservative-wider rule.", "last_updated": "2026-08-07", "confidence": "high", "note": "All-unit metro basis (pools bedroom sizes), wider than a within-bedroom band by construction. api.census.gov now requires an API key for queries (checked 2026-08-07); table remains on data.census.gov. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "housing.median_home_price_by_bedroom": { "source": "Zillow Home Value Index, Seattle metro", "last_updated": "2026-Q1" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare.premium_monthly_by_family_size": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "note": "Avg employee contribution (worker share of the premium). Corrected 2026-07-10: values previously held the FULL premium (~3.4x every US peer city) while this note already said employee contribution. Now consistent with NYC/SF/Boston/Chicago convention; KFF 2025 worker share of family premium = $6,850/yr." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: King County WA (FIPS 53033), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $2,400." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor.", "confidence": "high" }, "non_cash.parental_leave_weeks": { "source": "WA PFML https://paidleave.wa.gov/how-paid-leave-works/ ; https://paidleave.wa.gov/help-center/individuals-and-families/new-parents/ ; RCW 50A.15.020 https://app.leg.wa.gov/RCW/default.aspx?cite=50A.15.020", "last_updated": "2026-08-06", "confidence": "high", "note": "24 = 12 weeks of WA PFML family (bonding) leave per covered parent, per-parent and non-transferable, summed for a two-earner household (both-parents flag set). The birthing parent's medical leave (16 wk combined family+medical, 18 with pregnancy complications) is medically certified, EXCLUDED, conservative. Was 12/60: the unpaid FMLA window with an estimated effective rate. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "paidleave.wa.gov: 90% of AWW up to half the state AWW, then 50% of the excess; 2026 maximum USD 1,647 https://paidleave.wa.gov/how-paid-leave-works/", "last_updated": "2026-08-06", "confidence": "high", "note": "0 percentage-paid weeks: WA pays 90% of AWW up to half the state AWW (USD 915) plus 50% of the excess, capped at USD 1,647/wk in 2026 (90% of the 1,830 SAWW), binding from about USD 133,000, at or below the salaries we price, so the entitlement lives in non_cash.parental_leave_flat. Two-tier formula, no honest single headline. At USD 150,000 the effective rate is about 57%. REFUTATION recorded: the previous note's ~1,648 cap was wrong by a dollar, the agency page says 1,647." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Seattle: USD 250,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "USD 250,000 / 1 = USD 250,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 1.0000, R5-rounded down to 1. REPLACES 1 (2026-05-13), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. THE ANCHOR, re-verified. levels.fyi Seattle all-levels SWE median TC measured EXACTLY USD 250,000 on 2026-08-11 (n=7,536), and Seattle senior measured EXACTLY 305,000, both matching data/_compensation_anchor.json. Because Seattle measures exactly the anchor value, TC/250,000 and TC/Seattle-TC are the same number for every city, so the absolute and relative bases are not two competing worlds. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Defined as 1.0 here. Other cities' comp_index values are relative to this." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Seattle", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Seattle", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "IRS Child Tax Credit + state EITC", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 87000 USD, Seattle-Tacoma-Bellevue metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "labor_market.tech_salary_distribution": { "source": "https://www.bls.gov/oes/current/oes_42660.htm", "last_updated": "2025-05", "confidence": "rich", "note": "BLS OEWS May 2025, SOC 15-1252 Software Developers, annual wage percentiles (full-market occupation distribution; straight-time gross pay incl. incentive pay, excl. equity and nonproduction bonuses). Public domain; API pull 2026-07-19." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: King County WA (FIPS 53033), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $2,030/mo and the ages 1-2 rate $2,190/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,900. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.income_tested_childcare": { "source": "https://www.seattle.gov/education/for-parents/child-care-and-preschool/child-care-assistance-program", "last_updated": "2026-08-05", "confidence": "high", "note": "ELIGIBILITY-ONLY shape (CC2): ceiling verified at 110% SMI, $153,244 for a family of four, per the City of Seattle. Deliberately carries NO tiers/credit_pct: CCAP is voucher-minus-copay with no honest percentage, and a fabricated 100% tier would claim free childcare. childcareSubsidyNote renders \"may qualify\"; the billed figure stays the market rate. Full size-graded tables in docs/superpowers/research/2026-08-02-seattle-childcare-cliff.md." }, "non_cash.parental_leave_flat": { "source": "paidleave.wa.gov: 2026 maximum weekly benefit USD 1,647 (90% of the state AWW of 1,830) https://paidleave.wa.gov/how-paid-leave-works/", "last_updated": "2026-08-06", "confidence": "high", "note": "USD 1,647/wk, 2026 maximum. 24 x 1,647 = USD 39,528 total." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Seattle-Tacoma-Bellevue WA Metro Area (31000US42660). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US42660", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 30.0 exactly (aggregate 53,404,364 min / 1,780,464 workers; published MOE +/-0.2). Seattle CITY proper is 25.9, so the previous 28 matched neither geography." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "oslo": { "tax.filing_basis": { "source": "Skatteetaten (skatteetaten.no) - spouses assessed individually on employment income (tax classes abolished 2018)", "last_updated": "2026-07-11", "note": "Activates the per-filer split in compute() for two-earner households." }, "tax.inbound_regime": { "source": "PAYE / kildeskatt pa lonn, skatteloven kapittel 20. Flat 25% that INCLUDES trygdeavgift (Skatteetaten rettskilder: 'Denne skattesatsen inkluderer trygdeavgift'); first year of tax residency only; 2026 income ceiling NOK 725,050 (2025: NOK 697,150). Skatteloven s 20-3 makes the ceiling a hard cliff on the WHOLE income, not a band; s 20-2 bans every deduction. https://www.skatteetaten.no/person/utenlandsk/skal-du-arbeide-i-norge/skattekort/kildeskatt-pa-lonn/ ; https://lovdata.no/dokument/NL/lov/1999-03-26-14/KAPITTEL_20", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: max_income refreshed 697,150 (2025) -> 725,050 (2026). Still salary_relief false and no flat_rate, so inboundRegimeStatus returns note_only at every salary and no math moves. The note-only stance is now doubly defensible: professional salaries clear the cliff, and below it the total deduction ban frequently makes the scheme worse than ordinary tax for a household with a Norwegian mortgage. The forskrift under s 20-7 that sets the 2026 ceiling was not retrieved; the figure is published by Skatteetaten in both language versions." }, "tax.base_income_tax_rate": { "source": "Skatteetaten 2026 (allmenn inntekt)", "last_updated": "2026-05-12" }, "tax.trinnskatt_brackets": { "source": "Trinnskatt 2026 thresholds/rates (skatteetaten.no/en/rates/bracket-tax); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "tax.national_insurance_rate": { "source": "Trygdeavgift salary rate 7.6% 2026 (skatteetaten.no/en/rates/national-insurance-contributions); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "tax.sales_tax_rate": { "source": "Skatteetaten merverdiavgift", "last_updated": "2026-05-12", "note": "Standard MVA 25%" }, "tax.property_tax_rate_pct": { "source": "Oslo kommune", "last_updated": "2026-05-12", "note": "Eiendomsskatt abolished in Oslo" }, "housing.median_rent_by_bedroom": { "source": "SSB / Finn.no leiepriser 2025", "last_updated": "2025-Q4" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "housing.median_home_price_by_bedroom": { "source": "Eiendomsverdi Oslo 2025", "last_updated": "2025-Q4" }, "healthcare.premium_monthly_by_family_size": { "source": "Universal coverage via folketrygden: premium = 0", "last_updated": "2026-05-12" }, "healthcare.annual_out_of_pocket_avg": { "source": "Helsedirektoratet and Helfo egenandelstak + NAV, 1,452,000 residents reached the combined tak in 2023", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "childcare.infant_monthly": { "source": "regjeringen.no / Oslo kommune barnehage maksimalpris 2026 (Stortinget vedtatt)", "last_updated": "2026-07-17", "note": "2026 barnehage maksimalpris NOK 1,200/mo charged 11 months/yr + matpenger ~NOK 320/mo: effective (1,200 x 11/12) + 320 ~= NOK 1,420/mo. One age-independent cap covers infants and preschoolers (was 3,000, the pre-2024 level)." }, "non_cash.vacation_days_statutory": { "source": "Ferieloven (LOV-1988-04-29-21) § 5 nr. 1 https://lovdata.no/dokument/NL/lov/1988-04-29-21", "last_updated": "2026-08-08", "confidence": "high", "note": "CORRECTED 2026-08-08, 25 to 21. The Act grants 25 VIRKEDAGER, and the same paragraph defines the unit: 'Som virkedager regnes alle dager som ikke er sondager eller lovbestemte helge- eller hoytidsdager', so Saturday counts and this is a SIX-day week. 25 virkedager is 4 weeks plus 1 day, which is 21 five-day weekdays. The previous note read '25 working days statutory', a mistranslation of virkedager that hid the conversion. The familiar Norwegian 5 weeks is 30 virkedager under TARIFFAVTALE, a collective agreement, not the statute, and this field is the legal floor. Contrast Sweden, where Semesterlagen 3a excludes Saturday and Sunday from semesterdagar, so its 25 is already five-day and correctly stored. Additive to the 12 public holidays." }, "non_cash.parental_leave_weeks": { "source": "NAV https://www.nav.no/foreldrepenger ; folketrygdloven kap. 14", "last_updated": "2026-08-06", "confidence": "high", "note": "49 wk at 100% RE-CONFIRMED 2026-08-06 against NAV: 3 wk reserved for the mother pre-birth + modrekvote 15 + fedrekvote 15 + fellesperiode 16, ONE family entitlement with quotas inside it, counted once. The 80% option is now 61 wk 1 day (extended from the old 59 so the two elections carry near-equal total value); 49 at 100 is the default and conservative pick. REFUTATION recorded: the father's 2 omsorgspermisjon weeks at birth are statutorily unpaid (employer payment is practice, not statute) and add no paid weeks; fathers may start fedrekvote at birth but that draws down the 49. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md." }, "govt_benefits.child_allowance_per_kid_annual": { "source": "NAV barnetrygd 2026", "last_updated": "2026-05-12", "note": "Norwegian barnetrygd: universal, NOT means-tested." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Oslo: NOK 968,721, held in labor_market.tech_salary_distribution (source_slug greater-oslo-region, as_of 2026-07-18, table confidence thin). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "NOK 968,721 / 9.49 = USD 102,078; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.4083, R5-rounded down to 0.4. REPLACES 0.57 (2026-08-11), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Oslo by 42%. The measured p50 was 0.716 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): NOK 1,067,593 / 9.49 = USD 112,497; / USD 196,000 Seattle senior reference = 0.574, R5-rounded down to 0.57. REPLACES 0.78 (2026-05-13), which divided a NOK 1.25M senior-PM anecdote by USD 150,000 rather than by the USD 200-220k Seattle reference its peers use; 0.78 implied NOK 1,450,831, above Oslo's own measured tech p90 of 1,370,000 and 1.29x SSB's national upper quartile. SSB manedslonn covers avtalt lonn + bonus + uregelmessige tillegg, excludes overtime; Norway has no 13th month and feriepenger replaces June salary, so x12 is the correct annualisation. Limit: SSB publishes no Oslo-only cut, and Oslo pays above the national median, so 0.57 may be slightly low. Low is the R5-conservative direction. Working: docs/superpowers/research/2026-08-11-comp-index-sweep-a.md" }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Oslo", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Oslo", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "childcare.preschool_monthly": { "source": "regjeringen.no / Oslo kommune barnehage maksimalpris 2026", "last_updated": "2026-07-17", "note": "Same national maksimalpris as infants: effective ~NOK 1,420/mo (maksimalpris NOK 1,200 x11 + matpenger ~320)." }, "childcare.after_school_monthly": { "source": "Oslo kommune Aktivitetsskolen (AKS) pris og betaling 2026/2027", "last_updated": "2026-07-17", "note": "Oslo AKS full-price heltidsplass capped at NOK 1,242/mo for the 546k-1.5M household income band (part-time core is free for grades 1-3). Was 2,500." }, "tax.personfradrag": { "source": "Personfradrag 2026 NOK 114,540: regjeringen.no statsbudsjettet 2026 (108,550 -> 114,540) / skatteetaten.no/en/rates/personal-allowance", "source_name": "Skatteetaten / statsbudsjettet 2026", "last_updated": "2026-07-23", "confidence": "high", "note": "Applied INSIDE computeNorwayTaxes to the 22% alminnelig-inntekt base ONLY, together with minstefradrag. Per filer (Norway files individually). standard_deduction_* held at 0 so compute() does not subtract a lump across all three components. Correct-first item #1, 2026-07-17 resolutions doc." }, "tax.minstefradrag_rate": { "source": "statsbudsjettet 2026 (regjeringen.no) + skatteetaten.no 2026 rates: minstefradrag wage rate 46% (skatteetaten.no/en/rates/minimum-standard-deduction)", "source_name": "Skatteetaten", "last_updated": "2026-07-23", "confidence": "high", "note": "Minstefradrag = 46% of wage income, applied to the 22% base only. Capped (see minstefradrag_cap)." }, "tax.minstefradrag_cap": { "source": "statsbudsjettet 2026 (regjeringen.no) + skatteetaten.no 2026 rates: minstefradrag wage upper limit NOK 95,700 for 2026 (2025: 92,000)", "source_name": "Skatteetaten / statsbudsjettet 2026", "last_updated": "2026-07-23", "confidence": "high", "note": "Upper limit on the wage minstefradrag. Binds at any ICP salary. Sum with personfradrag = 210,240 on the 22% base." }, "tax.standard_deduction_single": { "source": "statsbudsjettet 2026 (regjeringen.no) + skatteetaten.no 2026 rates", "source_name": "Skatteetaten", "last_updated": "2026-07-23", "confidence": "high", "note": "Held at 0 BY DESIGN. Norway applies minstefradrag + personfradrag inside computeNorwayTaxes, to the 22% base only; a lump here would wrongly shrink trinnskatt and trygdeavgift too (the pre-2026-07-23 bug). See tax.personfradrag." }, "tax.standard_deduction_married": { "source": "statsbudsjettet 2026 (regjeringen.no) + skatteetaten.no 2026 rates", "source_name": "Skatteetaten", "last_updated": "2026-07-23", "confidence": "high", "note": "Held at 0 BY DESIGN. Norway applies minstefradrag + personfradrag inside computeNorwayTaxes, to the 22% base only; a lump here would wrongly shrink trinnskatt and trygdeavgift too (the pre-2026-07-23 bug). See tax.personfradrag." }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/greater-oslo-region", "last_updated": "2026-07-18", "confidence": "thin", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "EF English Proficiency Index 2024 (Very High band)", "last_updated": "2024", "note": "National-level English usability for tech work. Band: very_high." }, "non_cash.parental_leave_paid_pct": { "source": "NAV https://www.nav.no/foreldrepenger : calculation basis capped at 6G; G = NOK 136,549 from 1 May 2026, 6G = NOK 819,294", "last_updated": "2026-08-17", "confidence": "high", "note": "We store 0: the 2026-08-17 caps extension moved all 49 weeks into non_cash.parental_leave_capped (100% of a basis capped at 6G), so no percentage-paid weeks remain on this field. Was 100 with the cap recorded here but unconsumed, which left /rankings/parental-leave-by-country/ printing Norway at a flat 100% at any salary (raised 2026-08-08 and 2026-08-11, fixed cluster leave-caps-0817)." }, "non_cash.parental_leave_capped": { "source": "folketrygdloven para 14-7 (foreldrepenger basis follows the sykepenger rules, capped at 6G per para 8-10); NAV https://www.nav.no/grunnbelopet : G = NOK 136,549 per 1 May 2026, 6G = NOK 819,294/yr (fetched 2026-08-17)", "last_updated": "2026-08-17", "confidence": "high", "note": "All 49 weeks pay 100% of a basis capped at 6G = NOK 819,294/yr. cap_amount is stored per week (819,294 / 52 = NOK 15,755.65) on the engine weekly-gross convention, the same pre-conversion Stockholm uses for its annual SGI ceiling; foreldrepenger dagsats runs Mon-Fri (5-day week, divisor 260), so annual/52 equals dagsats x 5 exactly. Above 6G the payment is flat: effective rate = 819,294 / salary, so the stored Oslo tech p50 of NOK 968,721 yields 84.6% and a $150,000-equivalent (about NOK 1.41M) yields about 58%. The 80% election (61 wk 1 day) is deliberately not stored: 49 wk at 100% is the default and the conservative pick, per the 2026-08-06 batch2 adjudication. UNIT WATCH: G is an annual figure revised every May 1; re-derive cap_amount each trygdeoppgjor. Research: docs/superpowers/research/2026-08-17-parental-leave-caps.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Helsedirektoratet and Helfo egenandelstak + NAV, 1,452,000 residents reached the combined tak in 2023", "last_updated": "2026-08-08", "confidence": "medium", "note": "REPLACES an exact statutory maximum: NOK 3,040 is precisely the 2023 egenandelstak (series 2,460 / 2,921 / 3,040 / 3,165 / 3,278). 26% of the population reaches the cap and ~1m under-16s pay zero, so the realized adult mean is ~NOK 1,650. Children pay no egenandel under 16, extended to all under-18s from 2026-08-01. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "non_cash.commute_minutes_avg": { "source": "Jernbanedirektoratet / Reisevaneundersokelsen, Reisevaner i de 7 storste byregionene med tilleggsutvalg 2022, Figur 23. https://www.jernbanedirektoratet.no/content/uploads/sites/4/2023/11/Reisevaner-i-de-7-storste-byregionene-med-tilleggsutvalg-2022.pdf", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Reiselengde i tid til/fra arbeid (work-trip length in time)", "year": "2022", "note": "Oslo/Akershus byregion, n=4122. Published one way. Mean reconstructed from the published six-band distribution, giving a range of 29.9 to 31.3." }, "non_cash.sick_leave": { "source": "Folketrygdloven (National Insurance Act) kapittel 8 , §8-2 (opptjeningstid), §8-10; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.nav.no/grunnbelopet", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Skatteetaten, Forskuddsutskrivingen 2026 https://www.skatteetaten.no/en/rettskilder/type/uttalelser/uttalelser/forskuddsutskrivingen-2026/", "last_updated": "2026-08-11", "confidence": "high", "note": "IPS (individuell pensjonssparing), NOK 25,000 for 2026, raised from 15,000 under the 2024 pension settlement; old-IPS and new-IPS share one ceiling. Employee-only, and the source states the same figure as both the payment limit and the deduction limit. KNOWN RESIDUAL, not fixed here: the IPS deduction is against alminnelig inntekt (22% flat) ONLY and must not reduce trinnskatt or trygdeavgift, but the engine nets the retirement deduction off personinntekt upstream, so it still reaches too many bases." } }, "stockholm": { "non_cash.parental_leave_flat": { "source": "Socialforsakringsbalken 12 kap. 19 para (lagstanivadagar) https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/socialforsakringsbalk-2010110_sfs-2010-110/ ; Forsakringskassan foraldrapenning", "last_updated": "2026-08-01", "confidence": "high", "note": "90 of the 480 foraldrapenning days pay a flat SEK 180 per calendar day regardless of income. Stored as 12.857 weeks at SEK 1,260 per week (180 x 7). Owner decision 2026-08-01: flat benefits are shown at the effective rate for the salary in front of us rather than folded into a headline percentage." }, "tax.filing_basis": { "source": "Skatteverket (skatteverket.se) - individual taxation of spouses since 1971; no joint filing", "last_updated": "2026-07-11", "note": "Activates the per-filer split in compute() for two-earner households." }, "tax.inbound_regime": { "source": "Expertskatt / forskarskattelattnad, IL 11 kap. 22-23a ss: 25% of remuneration exempt from income tax, max 7 years. Salary route = monthly remuneration that 'overstiger' 1.5x prisbasbelopp. Prisbasbelopp 2026 = SEK 59,200, fixed by the government 2025-09-18 (https://www.regeringen.se/pressmeddelanden/2025/09/prisbasbelopp-for-2026-faststallt/), so Forskarskattenamnden publishes SEK 88,801/mo as the qualifying minimum for work starting in 2026 = SEK 1,065,612/yr (https://forskarskattenamnden.se/andra-sprak/taxation-of-research-workers-board/conditions-for-tax-relief/remuneration-rate). Application within 3 months; no Swedish home or habitual abode in the 5 calendar years before the work starts; non-Swedish citizenship until 2027.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: min_income 1,058,400 -> 1,065,612. The old figure was wrong twice and both errors ran generous: it used the 2025 prisbasbelopp (58,800) and it used 12 x 88,200 rather than 12 x 88,201, ignoring that 'overstiger' is strictly greater. Two known over-generosities REMAIN and are stated in the regime note rather than modeled, because we do not collect the inputs: the threshold locks to the prisbasbelopp of the year the work STARTED and is never re-indexed, and only regular monthly pay and benefits count toward it, so annual bonuses, commissions and one-off payments cannot be used to clear the line. SCHEDULED: a lagradsremiss of June 2026 proposes lifting the exempt share to 30% and the residency bar to 7 years from 2027-01-01. Not law yet. TODO (re-check 2027-01): adopt only if enacted; until then the stored 25% stands." }, "tax.kommunal_rate": { "source": "SCB Kommunalskatterna 2026 / Skatteverket", "last_updated": "2026-05-15", "note": "Stockholm municipal + regional rate 30.55%" }, "tax.statslig_brackets": { "source": "Skatteverket belopp och procent 2026", "last_updated": "2026-05-15", "note": "20% on income above 643 000 SEK threshold" }, "tax.pensionsavgift_rate": { "source": "Skatteverket / Pensionsmyndigheten 2026; 100% skattereduktion offsets the fee (skatteverket.se skattereduktioner); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "7% up to ceiling 673 038 SEK; the engine nets the fee against its matching skattereduktion, so the net cash effect is zero for filers with sufficient liability." }, "tax.pensionsavgift_cap": { "source": "Pensionsmyndigheten 2026 (8.07 x inkomstbasbelopp 83 400)", "last_updated": "2026-05-15" }, "tax.sales_tax_rate": { "source": "Skatteverket moms 2026", "last_updated": "2026-05-15", "note": "Standard VAT 25%" }, "tax.property_tax_rate_pct": { "source": "Skatteverket fastighetsavgift 2026", "last_updated": "2026-05-15", "note": "Fastighetsavgift is a capped annual amount; 0.15% is effective rate estimate for typical Stockholm apartment" }, "tax.standard_deduction_single": { "source": "Grundavdrag floor for higher incomes ~SEK 17,400 (Skatteverket grundavdrag 2026; ekonomifakta.se); replaces the old 40,000 catch-all that tried to fold in the jobbskatteavdrag; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "The earned-income relief is now modeled explicitly as the jobbskatteavdrag_annual credit, so the deduction returns to the true grundavdrag floor.", "confidence": "high" }, "housing.median_rent_by_bedroom": { "source": "Numbeo Stockholm May 2026 / Rentberry 2026", "last_updated": "2026-05-15", "note": "1BR city center ~14 500-17 000 SEK" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "housing.median_home_price_by_bedroom": { "source": "Investropa / Hemnet Group Stockholm 2026", "last_updated": "2026-05-15" }, "housing.utilities_monthly": { "source": "Numbeo Stockholm May 2026 / Expatistan", "last_updated": "2026-05-15" }, "housing.renters_insurance_monthly": { "source": "Hedvig hemforsäkring 2026 average", "last_updated": "2026-05-15" }, "housing.internet_phone_monthly": { "source": "Numbeo Stockholm broadband + estimated mobile bundle", "last_updated": "2026-05-15" }, "healthcare.premium_monthly_by_family_size": { "source": "Universal coverage via landsting: premium = 0", "last_updated": "2026-05-15" }, "healthcare.annual_out_of_pocket_avg": { "source": "1177.se hoegkostnadsskydd for oeppenvaard and for laekemedel + LIF national prescription co-payment aggregate SEK 8.7bn over ~8.6m adults + Region Stockholm SEK 275 visit fee", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "healthcare.dental_vision_monthly": { "source": "Folktandvarden / typical private dental insurance 2026", "last_updated": "2026-05-15" }, "childcare.infant_monthly": { "source": "Skolverket maxtaxa 2026", "last_updated": "2026-05-15" }, "childcare.preschool_monthly": { "source": "Skolverket maxtaxa 2026", "last_updated": "2026-05-15" }, "childcare.after_school_monthly": { "source": "Skolverket maxtaxa fritidshem 2026", "last_updated": "2026-05-15" }, "food.groceries_monthly_per_adult": { "source": "Numbeo Stockholm food basket May 2026", "last_updated": "2026-05-15" }, "food.groceries_monthly_per_kid": { "source": "Derived from adult figure at ~58%", "last_updated": "2026-05-15" }, "food.dining_out_monthly_baseline": { "source": "Numbeo Stockholm restaurants section May 2026", "last_updated": "2026-05-15" }, "transit.monthly_pass": { "source": "SL (sl.se) 2025 fares: 30-day pass 1 070 SEK; 2026 unchanged", "last_updated": "2025-01-08" }, "transit.annual_car_cost_avg": { "source": "Konsumentverket / Statista Sweden car TCO 2024", "last_updated": "2026-05-15" }, "discretionary.entertainment_monthly_baseline": { "source": "Numbeo Stockholm / SATS prices 2026", "last_updated": "2026-05-15" }, "discretionary.local_vacation_annual_baseline": { "source": "Hikersbay Stockholm travel costs / VisitSweden 2026", "last_updated": "2026-05-15" }, "travel_home.avg_round_trip_per_person_to_other_city": { "source": "Skyscanner / Google Flights ARN-SEA May 2026", "last_updated": "2026-05-15" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Försäkringskassan barnbidrag 2026", "last_updated": "2026-05-15", "note": "Swedish barnbidrag: universal, NOT means-tested." }, "non_cash.vacation_days_statutory": { "source": "Semesterlagen (Swedish Annual Leave Act)", "last_updated": "2026-05-15" }, "non_cash.public_holidays": { "source": "Swedish government public holidays 2026", "last_updated": "2026-05-15" }, "non_cash.parental_leave_weeks": { "source": "Socialforsakringsbalken (2010:110) 12 kap. (foraldrapenning, 480 dagar sammanlagt for foraldrarna, 90 reserverade per parent, 90 dagar pa lagstaniva) and 13 kap. 10 para (tillfallig foraldrapenning vid barns fodelse) https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/socialforsakringsbalk-2010110_sfs-2010-110/ ; Forsakringskassan https://www.forsakringskassan.se/privatperson/foralder/foraldrapenning", "last_updated": "2026-08-01", "confidence": "high", "note": "70 wk = the 480-day foraldrapenning pot (68.57 wk) + 10 tio-pappadagar (1.43 wk). Was 68. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. The 480 days are a SHARED pot (sammanlagt for foraldrarna), so they are counted once; the 90+90 reserved days sit INSIDE the 480 and are not summed on top. The 10 days for the non-birthing parent are tillfallig foraldrapenning under SFB 13 kap. 10 para, a legally separate benefit that does not draw down the 480, so they are additive." }, "non_cash.parental_leave_paid_pct": { "source": "Socialforsakringsbalken 12 kap. 19 para (sjukpenningniva och lagstaniva) https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/socialforsakringsbalk-2010110_sfs-2010-110/ ; Forsakringskassan Aktuella belopp 2026", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: since the 2026-08-06 caps rebuild the 400 percentage-paid days live in non_cash.parental_leave_capped (77.6% up to SEK 8,813/week) and the 90 lagstanivadagar stay in non_cash.parental_leave_flat, so no weeks remain on this field. Was 77.6 with the SGI ceiling ignored and flagged as a sourcing gap, both closed by the capped block." }, "non_cash.commute_minutes_avg": { "source": "Trafikanalys, RVU Sverige 2011-2014, Figur 3.4. https://www.trafa.se/globalassets/statistik/resvanor/2009-2015/rvu-sverige-2011-2014.pdf", "last_updated": "2026-08-08", "confidence": "low", "statistic": "Arbetsresornas genomsnittliga restid (average work-trip travel time)", "year": "2011-2014", "note": "NATIONAL figure adopted for want of a city series, so city variation is lost across all four Swedish cities. One way (the huvudresa ends at the workplace) but it includes en-route errands. Trafikanalys states the storstader run ABOVE the national mean, so 32 is a floor for Stockholm. Replaces a citation that was partly Numbeo. Retrieving the RVU Sverige Tabellbilaga or the regional RVUs would move all four Swedish cities from low to high confidence." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Stockholm: SEK 730,138, held in labor_market.tech_salary_distribution (source_slug greater-stockholm, as_of 2026-07-18, table confidence usable). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "SEK 730,138 / 9.51 = USD 76,776; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3071, R5-rounded down to 0.3. REPLACES 0.42 (2026-08-11), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Stockholm by 40%. The measured p50 was 0.731 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): SEK 799,123 / 9.51 = USD 84,030; / USD 196,000 Seattle senior reference = 0.429, R5-rounded down to 0.42. REPLACES 0.55 (2026-05-15), whose note carried no salary figure of any kind, no anchor and no confidence key, and which implied SEK 1,025,178, 0.95x Stockholm's own measured tech p90 of 1,080,000. SCB publishes a full-time-equivalent manadslon and Sweden has no 13th month, so x12 is the correct annualisation; the SCB base and the levels.fyi total-comp figures are consistent once Stockholm's premium over the national median and bonus/equity are allowed for. Working: docs/superpowers/research/2026-08-11-comp-index-sweep-a.md" }, "tax.standard_deduction_married": { "source": "Same grundavdrag floor as single (Sweden taxes spouses individually; Skatteverket); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.jobbskatteavdrag_annual": { "source": "Jobbskatteavdrag plateau ~SEK 52,400/yr 2026 (ekonomifakta.se/sakomraden/skatt/skatt-pa-arbete/jobbskatteavdraget_1212612.html); phase-out abolished from 2025; verified tax audit 2026-07-14", "confidence": "high", "note": "Credit against kommunal + statlig, clamped at the ladder. Plateau value; low incomes are modestly over-credited.", "last_updated": "2026-07-17" }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/greater-stockholm", "last_updated": "2026-07-18", "confidence": "usable", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "EF English Proficiency Index 2024 (Very High band)", "last_updated": "2024", "note": "National-level English usability for tech work. Band: very_high." }, "tax.property_tax_annual_cap": { "source": "https://www.skatteverket.se/privat/fastigheterochbostad/fastighetsavgiftochfastighetsskatt.4.69ef368911e1304a625800013531.html", "last_updated": "2026-08-05", "confidence": "high", "note": "Kommunal fastighetsavgift, capped: SEK 1,784 per apartment in a flerbostadshus (2026). The bostadsrattsforening is the legal taxpayer and embeds the fee in the monthly avgift, so the owner's DIRECT line is 0; we attribute the per-apartment cap through to the household under the conservative rule rather than charging 0, since the cost is real and lands on the owner via the avgift. Urban own-mode stock is dominantly bostadsratt; a detached-house (smahus) owner would instead pay up to SEK 10,425 directly, and a rare agarlagenhet freeholder pays directly at the smahus cap. Previous model was an uncapped 0.15% of market price, which charged Stockholm 3BR SEK 12,000 against the 1,784 cap." }, "non_cash.parental_leave_capped": { "source": "Forsakringskassan, Aktuella belopp 2026: \"Hogsta mojliga SGI ar 2026: 592 000 kronor (10 ganger prisbasbeloppet)\" and maximum foraldrapenning at sjukpenningniva SEK 1,259/day ; regeringen.se 2025-09-18: 2026 prisbasbelopp SEK 59,200 ; SFB 28 kap. 7 para (the 0.97 SGI reduction factor, named-paragraph citation via lagar.se)", "last_updated": "2026-08-06", "confidence": "high", "note": "Payout ceiling SEK 1,259/day x 7 = SEK 8,813/week over the 400 percentage-paid days (57.143 wk; 57.143 + 12.857 flat = 70.0, the stored total exactly). This closes the 2026-08-01 sourcing gap: 592,000 x 0.97 x 0.80 / 365 = 1,258.6 ~ 1,259, which also reproduces the stored 77.6 (0.97 x 0.80), an exact internal-consistency check. Confidence medium-high on the 0.97 factor specifically (named statute paragraph plus arithmetic reproduction, not a verbatim quote), high on the kronor figures (direct agency table). At SEK 1,000,000 the capped slice pays ~46% and the whole blend ~39%. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Sweden) and docs/superpowers/research/2026-08-06-parental-caps-europe-batch-C.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "1177.se hoegkostnadsskydd for oeppenvaard and for laekemedel + LIF national prescription co-payment aggregate SEK 8.7bn over ~8.6m adults + Region Stockholm SEK 275 visit fee", "last_updated": "2026-08-08", "confidence": "medium", "note": "REPLACES an exact statutory maximum, and the wrong one of two: SEK 1,450 is precisely the 2026 oeppenvaard hoegkostnadsskydd and ignores the separate SEK 3,800 medicine ceiling that any adult filling a prescription also pays into. Built instead from ~SEK 1,010 of realized prescription co-payment plus ~SEK 700 of visit fees. Under-18s are free on BOTH ceilings, so the family rows barely move above two adults. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "non_cash.sick_leave": { "source": "Lag (1991:1047) om sjuklön (\"sjuklönelagen\") for the employer phase; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.forsakringskassan.se/privatperson/e-tjanster-blanketter-och-informationsmaterial/aktuella-belopp", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Skatteverket, Avdrag foer pensionssparande https://www.skatteverket.se/privat/skatter/arbeteochinkomst/avdragforprivatpersoner/np.4.5fc8c94513259a4ba1d800042822.html ; 2026 prisbasbelopp https://www.regeringen.se/pressmeddelanden/2025/09/prisbasbelopp-for-2026-faststallt/", "last_updated": "2026-08-11", "confidence": "high", "note": "No deduction for our ICP. Skatteverket restricts the deduction to someone who \"helt saknar pensionsraett i sin anstaellning\", i.e. every employment they held in the income year lacked any right to tjaenstepension. The 35% / 10 prisbasbelopp rule (SEK 592,000 for 2026) reaches only that population; a salaried tech worker with occupational pension gets nothing. General deductibility was abolished in 2016. This is the single largest correction in the sweep." } }, "copenhagen": { "tax.filing_basis": { "source": "Skattestyrelsen (skat.dk) - individual taxation; unused personal allowance transferable between spouses", "last_updated": "2026-07-11", "note": "Activates the per-filer split in compute() for two-earner households." }, "tax.inbound_regime": { "source": "Forskerskatteordningen, KSL ss 48E-F: flat 27% after 8% AM-bidrag = 32.84% all-in, up to 84 months, no upper cap. KSL s 48E stk. 3 nr. 3 sets a grundbelob of DKK 59,500/mo at 2010 level, indexed under PSL s 20, 'hvortil kommer bidrag efter lov om Arbejdsmarkedets Tillaegspension'. Indexed 2026 figure DKK 65,400/mo measured BEFORE AM-bidrag per Den juridiske vejledning C.F.6.1.4 (https://info.skat.dk/data.aspx?oid=1977289), plus the employee ATP share of DKK 99/mo (borger.dk, ATP-satser for privat virksomhed, effective 2026-01-01). Ten-year prior tax-liability bar, KSL s 48E stk. 3 nr. 1 (https://info.skat.dk/data.aspx?oid=2350238). skat.dk tax-scheme-for-researchers.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: min_income 785,000 -> 786,000 = (65,400 + 99) x 12, rounded up. The 2026 monthly figure was LOWERED from DKK 78,000 (Ivaerksaetterpakke), so the direction in the old prose was right. Two prose corrections: (a) this is NOT note-only, the block carries flat_rate 0.27 and the engine applies it; (b) the regime note used to claim the scheme cuts the rate 'from ~52-56% down to about 33%', which compares a marginal rate to an effective one. Our own Denmark integration test pins the ordinary effective rate near 40% and the gap at ~7.1pp, so the note now says 40%. NOT PRIMARILY SOURCED: the transitional provision limiting DKK 65,400 to employments commencing on or after 2026-01-01 (retsinformation.dk and ft.dk both 403); four Danish secondary sources agree, and it does not affect a current-year arrival." }, "tax.am_bidrag_rate": { "source": "skat.dk/saerlige-situationer/arbejdsmarkedsbidrag", "source_name": "SKAT", "last_updated": "2026-05-15", "confidence": "high", "note": "8% gross wage contribution (AM-bidrag), 2026" }, "tax.bundskat_rate": { "source": "skat.dk/borger/beregn-din-skat/skattesatser", "source_name": "SKAT", "last_updated": "2026-05-15", "confidence": "high", "note": "12.01% bundskat 2026" }, "tax.kommunal_rate": { "source": "Copenhagen 2026 kommuneskat 23.39% (thelocal.dk 2025-11-13, Skatteministeriet); verified tax audit 2026-07-14", "source_name": "SKAT", "last_updated": "2026-07-17", "confidence": "high", "note": "Kobenhavn kommunal skat ~23.7% (2026)" }, "tax.topskat_rate": { "source": "skat.dk 2026 schedule: bundskat/mellemskat/topskat/toptopskat https://skat.dk/en-us/help/botton-bracket-middle-bracket-top-bracket-and-additional-top-bracket-tax", "source_name": "SKAT (skat.dk)", "last_updated": "2026-07-23", "confidence": "high", "note": "CUT to 7.5% by the 2026 reform (was 15%). Now the middle of three upper tiers, not the top one." }, "tax.topskat_threshold": { "source": "skat.dk 2026 schedule: bundskat/mellemskat/topskat/toptopskat https://skat.dk/en-us/help/botton-bracket-middle-bracket-top-bracket-and-additional-top-bracket-tax", "source_name": "SKAT (skat.dk)", "last_updated": "2026-07-23", "confidence": "high", "note": "RAISED to DKK 777,900 on personal income after AM-bidrag by the 2026 reform (was 640,108 applied to the wrong base). Gross-basis equivalent 845,543." }, "tax.standard_deduction_single": { "source": "skat.dk 2026 schedule: bundskat/mellemskat/topskat/toptopskat https://skat.dk/en-us/help/botton-bracket-middle-bracket-top-bracket-and-additional-top-bracket-tax", "source_name": "SKAT (skat.dk)", "last_updated": "2026-07-23", "confidence": "high", "note": "Held at 0 BY DESIGN. Denmark applies its personfradrag inside computeDenmarkTaxes against personal income (0.92 x gross); leaving a lump here would double-subtract it. See tax.personfradrag." }, "tax.sales_tax_rate": { "source": "skat.dk/erhverv/moms/momssatser-og-momsfritagelse", "source_name": "SKAT", "last_updated": "2026-05-15", "confidence": "high", "note": "Standard VAT (moms) 25%" }, "tax.property_tax_rate_pct": { "source": "Vurderingsstyrelsen grundskyld 2026", "source_name": "SKAT/Vurderingsstyrelsen", "last_updated": "2026-05-15", "confidence": "medium", "note": "Grundskyld rate ~0.32% for Copenhagen municipality 2026" }, "housing.median_rent_by_bedroom": { "source": "Boligsiden.dk Copenhagen 2025-2026", "source_name": "Boligsiden", "last_updated": "2026-05-15", "confidence": "medium", "note": "1BR city center ~11 000-12 000 DKK; 0BR estimated; 3BR extrapolated. Cross-checked Numbeo." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "housing.median_home_price_by_bedroom": { "source": "Boligsiden.dk / Danmarks Statistik dst.dk 2025", "source_name": "Boligsiden/DST", "last_updated": "2026-05-15", "confidence": "medium", "note": "Copenhagen median sale prices 2025. 1BR ~3.2M DKK; 3BR central Copenhagen can exceed 8M." }, "housing.utilities_monthly": { "source": "Numbeo Copenhagen utilities 2026", "source_name": "Numbeo", "last_updated": "2026-05-15", "confidence": "medium", "note": "Basic utilities 2BR ~1 400 DKK; district heating prevalent" }, "housing.renters_insurance_monthly": { "source": "Tryg/IF indboforsikring Copenhagen estimate 2026", "source_name": "Tryg/IF", "last_updated": "2026-05-15", "confidence": "medium", "note": "Typical renter's insurance ~130 DKK/mo" }, "housing.internet_phone_monthly": { "source": "Yousee / TDC broadband + Telenor mobile 2026", "source_name": "TDC/Telenor", "last_updated": "2026-05-15", "confidence": "medium", "note": "Broadband ~300 DKK + mobile ~250 DKK = ~550 DKK" }, "healthcare.premium_monthly_by_family_size": { "source": "Universal coverage via det offentlige sygehusvaeesen: premium = 0", "source_name": "sundhedsstyrelsen.dk", "last_updated": "2026-05-15", "confidence": "high" }, "healthcare.annual_out_of_pocket_avg": { "source": "Danmarks Statistik health expenditure DKK 262.9bn (2023) at 13.9% household out-of-pocket + sundhed.dk medicintilskud tiers and the DKK 4,850 egenbetalingsloft", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "healthcare.dental_vision_monthly": { "source": "Medi-Cover / typical private tandlaege supplement 2026", "source_name": "Medi-Cover", "last_updated": "2026-05-15", "confidence": "medium", "note": "No universal adult dental in Denmark; private dental plan ~250-450 DKK/mo. 350 DKK used." }, "childcare.infant_monthly": { "source": "Bornenes- og Undervisningsministeriet 2026 (vuggestue)", "source_name": "UVM/Copenhagen municipality", "last_updated": "2026-05-15", "confidence": "high", "note": "Vuggestue (0-2yo) max parental contribution Copenhagen ~3 000 DKK/mo" }, "childcare.preschool_monthly": { "source": "Copenhagen municipality bornehave takst 2026", "source_name": "Kbh.dk", "last_updated": "2026-05-15", "confidence": "high", "note": "Bornehave (3-5yo) ~2 000 DKK/mo after subsidy" }, "childcare.after_school_monthly": { "source": "Copenhagen SFO/fritidsklub 2026", "source_name": "Kbh.dk", "last_updated": "2026-05-15", "confidence": "medium", "note": "Skolefritidsordning (SFO) ~1 800 DKK/mo for school-age kids" }, "food.groceries_monthly_per_adult": { "source": "Numbeo Copenhagen food basket May 2026", "source_name": "Numbeo", "last_updated": "2026-05-15", "confidence": "medium", "note": "~3 500 DKK/mo; cross-checked with DST forbrugerundersogelse" }, "food.groceries_monthly_per_kid": { "source": "Derived from adult figure at ~57%", "source_name": "Derived", "last_updated": "2026-05-15", "confidence": "medium" }, "food.dining_out_monthly_baseline": { "source": "Numbeo Copenhagen restaurant prices May 2026", "source_name": "Numbeo", "last_updated": "2026-05-15", "confidence": "medium", "note": "Mid-range meal ~180 DKK; casual restaurant 2-3x/week baseline" }, "transit.monthly_pass": { "source": "Rejsekort zone-all Copenhagen 2026 (dk.rejsekort.dk)", "source_name": "Rejsekort/DOT", "last_updated": "2026-05-15", "confidence": "high", "note": "Monthly pass equivalent via rejsekort ~480 DKK for typical commute zones 1-4" }, "transit.annual_car_cost_avg": { "source": "FDM Bilkob og bilejerskab 2025 / Statista DK", "source_name": "FDM", "last_updated": "2026-05-15", "confidence": "medium", "note": "Denmark has high vehicle registration tax; ~85 000 DKK/yr all-in" }, "discretionary.entertainment_monthly_baseline": { "source": "Numbeo Copenhagen leisure 2026", "source_name": "Numbeo", "last_updated": "2026-05-15", "confidence": "medium", "note": "Gym + cinema + streaming + misc = ~1 800 DKK" }, "discretionary.local_vacation_annual_baseline": { "source": "Danmarks Statistik turistundersogelse 2024", "source_name": "DST", "last_updated": "2026-05-15", "confidence": "medium", "note": "~18 000 DKK/yr domestic/regional vacations" }, "travel_home.avg_round_trip_per_person_to_other_city": { "source": "Skyscanner / Google Flights CPH-SEA 2026", "source_name": "Skyscanner", "last_updated": "2026-05-15", "confidence": "high", "note": "Economy class CPH-SEA ~6 000 DKK; consistent with ~$870 USD at 6.9 DKK/USD" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Udbetaling Danmark borne- og ungeydelse 2026", "source_name": "Udbetaling Danmark", "last_updated": "2026-05-15", "confidence": "high", "note": "Borneydelse ~17 800 DKK/yr per child (0-14 years old) Danish borne- og ungeydelse: universal below a high-income threshold, means-tested down 2% above the ~DKK 892k topskat base; largely survives, mildly reduced for the very highest earners." }, "non_cash.vacation_days_statutory": { "source": "Ferieloven LBK nr 152 af 20/02/2024 § 5 stk. 1 and § 6 stk. 2", "last_updated": "2026-08-08", "confidence": "high", "note": "RE-VERIFIED 2026-08-08, 25 stands. Accrual is 2.08 days per month of employment, so 24.96 or 25 a year. Denmark writes the unit into the statute: § 6 stk. 2 says 'Ferie holdes med 5 dage om ugen', leave is taken at 5 days per week, so no six-day conversion applies and this is the opposite risk profile to Norway. § 6 stk. 3 bars leave from being scheduled on public holidays, which confirms the two pools are additive." }, "non_cash.public_holidays": { "source": "LOV nr 214 af 06/03/2023 (afskaffelse af store bededag som helligdag) https://www.retsinformation.dk/eli/lta/2023/214", "last_updated": "2026-08-08", "confidence": "high", "note": "CORRECTED 2026-08-08, 11 to 10. Store Bededag was abolished as a helligdag by LOV nr 214 af 06/03/2023, in force 2024-01-01, with affected employees compensated by a 0.45% salary supplement instead. Our 11 still counted it. The current list is 10: Nytarsdag, Skaertorsdag, Langfredag, Paskedag, 2. paskedag, Kristi himmelfartsdag, Pinsedag, 2. pinsedag, Juledag, 2. juledag. Grundlovsdag, Juleaftensdag and Nytarsaftensdag are collective-agreement days, not statutory helligdage, so they stay out." }, "non_cash.parental_leave_weeks": { "source": "borger.dk, Loenmodtager paa barsel (ny orlovsmodel) https://www.borger.dk/familie-og-boern/barsel-oversigt/barsel-loenmodtagere-ny-orlovsmodel ; barselsloven (2022 model, births from 2 Aug 2022)", "last_updated": "2026-08-06", "confidence": "high", "note": "52 wk = mother 4 pre-birth + 24 after + father/co-parent 24 (2 at birth, 9 earmarked non-transferable and 13 transferable inside each 24). Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. The both-parents flag stays off per the Sweden precedent: reserved weeks inside a standard mother-plus-father split, not a summed child-rearing entitlement driving the count." }, "non_cash.parental_leave_paid_pct": { "source": "borger.dk: barselsdagpenge are computed from your hourly wage and capped at the hojeste sats, DKK 5,085/wk before tax in 2026", "last_updated": "2026-08-06", "confidence": "high", "note": "0 percentage-paid weeks: the statute pays your wage up to a flat cap, and the cap binds for any ICP (it is reached at ~DKK 22,041/month gross), so the whole entitlement lives in non_cash.parental_leave_flat. The previous 50 was an admitted guess that blended employer top-ups into a statutory field." }, "non_cash.commute_minutes_avg": { "source": "DTU Transportvaneundersogelsen (Danish National Travel Survey), faktaark om pendling 2016-2019, Region Hovedstaden. https://www.man.dtu.dk/-/media/institutter/management/scientific-advice/transportvaneundersoegelsen/t-u/tu-udgivelser/faktaark/2016-19_faktaark_pendling.pdf", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Gennemsnitligt tidsforbrug paa pendlerrejser (average time spent on commuter journeys)", "year": "2016-2019", "note": "CONVERSION: the published 59 min is a ROUND TRIP, because TU defines a pendlerrejse as home to work to home; 59/2 = 29.5 one way. About 35% of these chains include an errand, so the figure is biased slightly up. Replaces a citation to 'Danmarks Statistik pendlingsundersogelse 2024', a publication that does not exist; DST's actual commuting statistic is register-based and publishes kilometres and flows, never minutes." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Copenhagen: DKK 695,268, held in labor_market.tech_salary_distribution (source_slug copenhagen-dnk, as_of 2026-07-18, table confidence thin). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "DKK 695,268 / 6.47 = USD 107,460; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.4298, R5-rounded down to 0.42. REPLACES 0.62 (2026-05-15), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Copenhagen by 48%. The measured p50 was 0.693 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior tech roles Copenhagen ~700-900k DKK (~$100-130k USD at 6.9 DKK/USD) vs Seattle ~$200k USD. Ratio 0.55-0.65; 0.62 used per task guidance mid-market estimate." }, "tax.personal_allowance": { "confidence": "high", "source": "Personfradrag 2026 DKK 54,100 (skat.dk); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "tax.personfradrag": { "source": "Personfradrag 2026 DKK 54,100: https://www.exploringdenmark.com/danish-tax-rates-2026/ (cross-checked against SKAT)", "source_name": "SKAT / exploringdenmark", "last_updated": "2026-07-23", "confidence": "high", "note": "Applied INSIDE computeDenmarkTaxes against personal income (0.92 x gross), not against gross. standard_deduction_single/married are therefore held at 0 so compute() passes gross through; a non-zero value there would double-subtract the allowance. Correct-first item #3, 2026-07-17 resolutions doc." }, "tax.mellemskat_rate": { "source": "skat.dk 2026 schedule: bundskat/mellemskat/topskat/toptopskat https://skat.dk/en-us/help/botton-bracket-middle-bracket-top-bracket-and-additional-top-bracket-tax", "source_name": "SKAT (skat.dk)", "last_updated": "2026-07-23", "confidence": "high", "note": "2026 three-tier reform: mellemskat 7.5% on personal income above DKK 641,200 (i.e. after the 8% AM-bidrag). Replaced the pre-2026 single 15% topskat together with the two keys below." }, "tax.mellemskat_threshold": { "source": "skat.dk 2026 schedule: bundskat/mellemskat/topskat/toptopskat https://skat.dk/en-us/help/botton-bracket-middle-bracket-top-bracket-and-additional-top-bracket-tax", "source_name": "SKAT (skat.dk)", "last_updated": "2026-07-23", "confidence": "high", "note": "DKK 641,200 measured on personal income AFTER AM-bidrag. Gross-basis equivalent 696,957 (= 641,200 / 0.92), which confirms the basis." }, "tax.toptopskat_rate": { "source": "skat.dk 2026 schedule: bundskat/mellemskat/topskat/toptopskat https://skat.dk/en-us/help/botton-bracket-middle-bracket-top-bracket-and-additional-top-bracket-tax", "source_name": "SKAT (skat.dk)", "last_updated": "2026-07-23", "confidence": "high", "note": "2026 reform: toptopskat 5% on personal income above DKK 2,592,700. Does not bind at ICP salaries but is modeled for completeness." }, "tax.toptopskat_threshold": { "source": "skat.dk 2026 schedule: bundskat/mellemskat/topskat/toptopskat https://skat.dk/en-us/help/botton-bracket-middle-bracket-top-bracket-and-additional-top-bracket-tax", "source_name": "SKAT (skat.dk)", "last_updated": "2026-07-23", "confidence": "high", "note": "DKK 2,592,700 on personal income after AM-bidrag. Gross-basis equivalent 2,818,152." }, "tax.standard_deduction_married": { "source": "skat.dk 2026 schedule: bundskat/mellemskat/topskat/toptopskat https://skat.dk/en-us/help/botton-bracket-middle-bracket-top-bracket-and-additional-top-bracket-tax", "source_name": "SKAT (skat.dk)", "last_updated": "2026-07-23", "confidence": "high", "note": "Held at 0 BY DESIGN. Denmark applies its personfradrag inside computeDenmarkTaxes against personal income (0.92 x gross); leaving a lump here would double-subtract it. See tax.personfradrag." }, "tax.beskaeftigelsesfradrag_rate": { "source": "DK-1: beskæftigelsesfradrag 12.75% of employment income (2026), a ligningsmæssigt fradrag reducing the kommuneskat base ONLY (not personlig indkomst / bundskat / mellemskat / topskat). SKAT employment-and-job-allowances; PwC DK deductions.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.beskaeftigelsesfradrag_cap": { "source": "DK-1: beskæftigelsesfradrag maximum DKK 63,300 (2026). SKAT; SKM faktaark.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.jobfradrag_rate": { "source": "DK-1: jobfradrag 4.5% of employment income over the threshold (2026), municipal base only. SKAT employment-and-job-allowances.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.jobfradrag_threshold": { "source": "DK-1: jobfradrag applies to employment income over DKK 235,200 (2026). SKAT.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.jobfradrag_cap": { "source": "DK-1: jobfradrag maximum DKK 3,100 (2026). SKAT.", "last_updated": "2026-07-26", "confidence": "high" }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/copenhagen-dnk", "last_updated": "2026-07-18", "confidence": "thin", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "EF English Proficiency Index 2024 (Very High band)", "last_updated": "2024", "note": "National-level English usability for tech work. Band: very_high." }, "non_cash.parental_leave_flat": { "source": "borger.dk https://www.borger.dk/familie-og-boern/barsel-oversigt/barsel-loenmodtagere-ny-orlovsmodel : hojeste sats DKK 5,085/wk before tax, 2026 (DKK 137.43/hr x 37)", "last_updated": "2026-08-06", "confidence": "high", "note": "52 x 5,085 = DKK 264,420 total, about a 29% effective rate at a DKK 900,000 gross. Owner call 2026-08-05 under the conservative rule: state dagpenge only. Funktionaerloven para 7 half pay from the employer (salaried employees, 4 wk pre-birth to 14 wk post) is EXCLUDED from this value: counting it is more generous and carries the higher evidence bar, and the para 7 details were not verified this pass. Flag it as a possible refinement, not a defect." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Danmarks Statistik health expenditure DKK 262.9bn (2023) at 13.9% household out-of-pocket + sundhed.dk medicintilskud tiers and the DKK 4,850 egenbetalingsloft", "last_updated": "2026-08-08", "confidence": "medium", "note": "The old DKK 4,250 was 88% of the medicine egenbetalingsloft, which is only reached at DKK 22,596 of gross annual medicine spend, so it was maximum-flavoured rather than an average. Denmark is the only Nordic country where children add material cost: under-18s get a 60% medicine subsidy from the first krone, which is a discount rather than an exemption, while GP and hospital care are free for everyone in gruppe 1. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "non_cash.sick_leave": { "source": "Sygedagpengeloven (Consolidated Act on Sickness Benefits) , §31 (arbejdsgiverperiode; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.detfagligehus.dk/...", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Skat.dk, Fradrag for indbetalinger til pension https://skat.dk/borger/pension-og-efterloen/fradrag-for-indbetalinger-til-pension ; Livsvarig livrente https://skat.dk/borger/pension-og-efterloen/livsvarig-livrente ; Juridisk vejledning C.A.10.2.1.3.3 (PBL s.18) https://info.skat.dk/data.aspx?oid=2048243", "last_updated": "2026-08-11", "confidence": "medium", "note": "JUDGMENT CALL. Ratepension plus ophoerende livrente carry a combined DKK 68,700 ceiling for 2026, shared with employer-administered schemes; overshoot draws a 20% afgift. Livsvarig livrente is genuinely UNCAPPED where the saver signs a 10-year-or-longer agreement, so the strictly-correct value for that saver is no key at all. 68,700 is chosen because our ICP is a mobile tech worker who would not sign a 10-year binding Danish lifetime-annuity commitment, and it is the conservative side. Aldersopsparing gives no deduction. The opfyldningsfradrag DKK 63,200 is for lump-sum engangsindskud, not a general annual ceiling. NEIGHBOURING BUG, out of scope: the Danish branch charges AM-bidrag on the post-retirement-deduction figure; AM-bidrag is levied on full gross." } }, "tokyo": { "childcare.free_tier_start_age": { "source": "Kodomo Katei Cho (Children and Families Agency), free early childhood education and care policy: \"3 sai kara 5 sai made no subete no kodomotachi no riyoryo ga musho ni narimasu\" (usage fees become free for all children aged 3 to 5), in force 2019-10-01 https://www.cfa.go.jp/policies/kokoseido/mushouka/", "confidence": "high", "note": "Free tuition covers ages 3-5. Ages 0-2 stay fee-liable except for municipal-resident-tax-exempt households, with fees set by household tax band x municipality x hours, so the paid rate applies across the whole 0-2 class rather than to infants only. Before this, ages 1-5 billed the free-tier rate. Watch item: Tokyo has been funding wards to drop 0-2 fees (second and later children since FY2023, first children from 2025-09-01 in adopting wards incl. Meguro, Nerima, Kita), so the Tokyo infant rate may now be high for a growing share of the city; ward coverage is unconfirmed and the value is left alone.", "last_updated": "2026-08-01" }, "tax.filing_basis": { "source": "PwC WWTS - Japan (Tax administration): All income tax returns are filed on an individual basis in Japan; joint tax returns are not permitted.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.national_brackets": { "source": "NTA nta.go.jp/english/taxes/individual/incometax/02.htm", "source_name": "National Tax Agency Japan", "last_updated": "2026-05-15", "confidence": "high", "note": "7 brackets 5/10/20/23/33/40/45% — unchanged since 2015" }, "tax.reconstruction_surcharge": { "source": "NTA nta.go.jp/english/taxes/individual/surtax/01.htm", "source_name": "National Tax Agency Japan", "last_updated": "2026-05-15", "confidence": "high", "note": "2.1% of national income tax; fukkō tokubetsu shotoku-zei, ongoing" }, "tax.prefectural_tax_rate": { "source": "Ministry of Internal Affairs soumu.go.jp", "source_name": "MIC Japan", "last_updated": "2026-05-15", "confidence": "high", "note": "4% flat prefectural inhabitant tax (todofuken-min-zei)" }, "tax.municipal_tax_rate": { "source": "Ministry of Internal Affairs soumu.go.jp", "source_name": "MIC Japan", "last_updated": "2026-05-15", "confidence": "high", "note": "6% flat municipal inhabitant tax; combined inhabitant = 10%" }, "tax.social_insurance_rate": { "source": "Japan Pension Service / MHLW 2026", "source_name": "MHLW Japan", "last_updated": "2026-05-15", "confidence": "high", "note": "Employee share: kosei nenkin ~9.15% + kenko hoken ~5% + koyo hoken ~0.3% = ~14.45%" }, "tax.sales_tax_rate": { "source": "NTA / MOF Japan shouhizei", "source_name": "NTA/MOF Japan", "last_updated": "2026-05-15", "confidence": "high", "note": "Consumption tax (shouhizei) 10% since Oct 2019; reduced 8% for food" }, "tax.property_tax_rate_pct": { "source": "Tokyo Metropolitan Government kotei shisanzei 2026", "source_name": "Tokyo Metro", "last_updated": "2026-05-15", "confidence": "high", "note": "Standard kotei shisanzei 1.4% of assessed value" }, "tax.standard_deduction_single": { "source": "NTA kiso kojo 2026 (nta.go.jp)", "source_name": "NTA Japan", "last_updated": "2026-05-15", "confidence": "high", "note": "Basic deduction 480 000 JPY for income under 24M JPY" }, "housing.median_rent_by_bedroom": { "source": "Suumo / LIFULL HOME'S Tokyo 2025-2026", "source_name": "Suumo/LIFULL", "last_updated": "2026-05-15", "confidence": "medium", "note": "Central Tokyo (Minato/Shibuya/Shinjuku): 1LDK ~150k, 2LDK ~220k, 3LDK ~320k JPY/mo" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "housing.median_home_price_by_bedroom": { "source": "Real Estate Japan / Athome Tokyo mansion prices 2025", "source_name": "Real Estate Japan", "last_updated": "2026-05-15", "confidence": "medium", "note": "Central Tokyo mansion prices; 1BR ~45M, 2BR ~70M, 3BR ~100M JPY" }, "housing.utilities_monthly": { "source": "Numbeo Tokyo utilities 2026 / Tokyo Gas estimates", "source_name": "Numbeo/Tokyo Gas", "last_updated": "2026-05-15", "confidence": "medium", "note": "Electric + gas + water ~15 000 JPY/mo for 1-2BR" }, "housing.renters_insurance_monthly": { "source": "Tokio Marine / Sony Sonpo fire insurance 2026", "source_name": "Tokio Marine", "last_updated": "2026-05-15", "confidence": "medium", "note": "Kasai hoken ~1 000 JPY/mo; mandatory for most rental contracts" }, "housing.internet_phone_monthly": { "source": "NTT Hikari / Docomo SIM plan 2026", "source_name": "NTT/Docomo", "last_updated": "2026-05-15", "confidence": "high", "note": "Fiber ~3 000 + SIM ~2 000 JPY/mo = ~5 000 JPY" }, "healthcare.premium_monthly_by_family_size": { "source": "Japan Health Insurance Association (Kyokai Kenpo) 2026", "source_name": "JHIA", "last_updated": "2026-05-15", "confidence": "high", "note": "Shakai Hoken (employer-based) covers premium via social_insurance_rate. OOP model used here; premium set to 0." }, "healthcare.annual_out_of_pocket_avg": { "source": "MHLW coinsurance schedule (30% for ages 6-69, 20% under 6) + the income-banded kougaku ryouyouhi monthly high-cost ceiling + Tokyo ward-level child medical subsidy", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "healthcare.dental_vision_monthly": { "source": "JDA / typical dental practice Tokyo 2026", "source_name": "JDA", "last_updated": "2026-05-15", "confidence": "medium", "note": "Routine cleaning ~3 000-5 000 JPY/quarter; ~3 000 JPY/mo average" }, "childcare.infant_monthly": { "source": "Tokyo Metropolitan Government hoikuen fee 2025", "source_name": "Tokyo Metro", "last_updated": "2026-05-15", "confidence": "medium", "note": "Hoikuen (public) fees income-tied; for median income earner ~30 000 JPY/mo for infant. Waitlists common in central wards." }, "childcare.preschool_monthly": { "source": "MEXT/Cabinet Office 2019 free early-childhood-education reform; Tokyo ward guidance 2025", "last_updated": "2026-07-17", "note": "Ages 3-5 tuition is free nationally (2019 reform, no income cap = NOT means-tested); families pay food/materials/events only ~JPY 6,000/mo. Was 25,000 (full tuition)." }, "childcare.after_school_monthly": { "source": "Tokyo gakudo hoiku (學童保育) fee 2025", "source_name": "Tokyo Metro", "last_updated": "2026-05-15", "confidence": "medium", "note": "School-age after-care (gakudo) ~15 000 JPY/mo typical" }, "food.groceries_monthly_per_adult": { "source": "Numbeo Tokyo food basket May 2026 / SB Statistics", "source_name": "Numbeo", "last_updated": "2026-05-15", "confidence": "medium", "note": "~45 000 JPY/mo for single adult; quality ingredients slightly premium" }, "food.groceries_monthly_per_kid": { "source": "Derived at ~56% of adult figure", "source_name": "Derived", "last_updated": "2026-05-15", "confidence": "medium" }, "food.dining_out_monthly_baseline": { "source": "Numbeo Tokyo restaurants May 2026", "source_name": "Numbeo", "last_updated": "2026-05-15", "confidence": "medium", "note": "Mid-range restaurant 2 500 JPY/meal; typical lunch set 1 000 JPY; ~25 000 JPY/mo baseline" }, "transit.monthly_pass": { "source": "Tokyo Metro / Toei 30-day pass 2026 (tokyometro.jp)", "source_name": "Tokyo Metro", "last_updated": "2026-05-15", "confidence": "high", "note": "Commuter pass (teiki) ~10 000 JPY/mo for typical 2-3 zone commute; company often reimburses" }, "transit.annual_car_cost_avg": { "source": "JAF car ownership costs Japan 2025", "source_name": "JAF", "last_updated": "2026-05-15", "confidence": "medium", "note": "Car ownership rare in central Tokyo; parking ~40 000 JPY/mo, shakken (biennial inspection) ~100 000 JPY. Total ~600 000 JPY/yr" }, "discretionary.entertainment_monthly_baseline": { "source": "Numbeo Tokyo leisure 2026 / Statista Japan recreation", "source_name": "Numbeo", "last_updated": "2026-05-15", "confidence": "medium", "note": "Gym ~10 000 + cinema + streaming + misc ~20 000 JPY/mo" }, "discretionary.local_vacation_annual_baseline": { "source": "JNTO / JTB domestic travel survey 2024", "source_name": "JTB Research", "last_updated": "2026-05-15", "confidence": "medium", "note": "Domestic travel spending ~150 000 JPY/yr for singles; shinkansen + ryokan common" }, "travel_home.avg_round_trip_per_person_to_other_city": { "source": "Skyscanner / Google Flights NRT-SEA May 2026", "source_name": "Skyscanner", "last_updated": "2026-05-15", "confidence": "high", "note": "Economy NRT-SEA round-trip ~120 000 JPY (~$800 USD at 150 JPY/USD)" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Cabinet Office Japan jido-teate 2024 reform", "source_name": "Cabinet Office Japan", "last_updated": "2026-05-15", "confidence": "high", "note": "jido-teate: 0-2yo ¥120k/yr, 3-12yo ¥120k/yr (3rd child ¥180k). ~¥150k used as approximate per child. Japanese jido teate: income cap abolished Oct 2024, now universal, NOT means-tested." }, "non_cash.vacation_days_statutory": { "source": "Labour Standards Act Japan / MHLW survey 2024", "source_name": "MHLW Japan", "last_updated": "2026-05-15", "confidence": "high", "note": "10 days statutory minimum (requires 5+ days taken since 2019 reform); actual usage culturally suppressed" }, "non_cash.public_holidays": { "source": "Cabinet Office Japan shukujitsu 2026", "source_name": "Cabinet Office Japan", "last_updated": "2026-05-15", "confidence": "high", "note": "16 public holidays in 2026" }, "non_cash.parental_leave_weeks": { "source": "Ikuji-Kaigo Kyugyo Ho art. 5 (childcare leave until the child turns 1, PER PARENT, non-transferable); Labor Standards Act art. 65 (maternity 6 wk pre + 8 wk post); Employment Insurance Act (ikuji kyugyo kyufu-kin); 2026 parameter references gyomu-keisan.jp (Reiwa 8), tfp-group.co.jp", "last_updated": "2026-08-06", "confidence": "high", "note": "110 wk = mother 14 (sanzen-sango kyugyo) + mother ikuji kyugyo from the end of the postnatal 8 wk to the first birthday (44 wk) + father ikuji kyugyo from birth to the first birthday (52 wk). The leave is PER PARENT and non-transferable, so both spans are summed, which is why parental_leave_both_parents is true (the Korea 120-wk precedent, an identical construction). The 4-wk shussei-ji ikuji kyugyo (post-birth papa leave) is taken WITHIN the father's span and its benefit days count against his 180-day 67% clock, so it adds nothing. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. Excluded: the 13% top-up to 80% for 28 days and the papa-mama 14-month extension (conditional on both parents' behaviour), the 18/24-month extension when no daycare place is found (conditional tail), and the JPY 500,000 shussan ikuji ichijikin (a lump sum, not leave pay). Was 52/67, the mother's span alone, as the old note itself admitted." }, "non_cash.parental_leave_paid_pct": { "source": "Health Insurance Act art. 102 (shussan teate-kin, 2/3 of standard remuneration); Employment Insurance Act art. 61-7 (ikuji kyugyo kyufu-kin, 67% first 180 days of each parent's leave, 50% after); ceilings per MHLW leaflet LL080731 ho 02 (2026-08-01) via https://roumu.com/pdf/2026080314.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "We store 0: since the 2026-08-07 Japan caps apply, all 110 weeks live in non_cash.parental_leave_capped (60% up to JPY 297,720/month), so no percentage-paid remainder weeks exist and a leftover headline rate would describe zero weeks. The old 60 blend (65.4 wk at 67 + 44.6 wk at 50, maternity at 66.7) is preserved as the capped block's pct. Research: docs/superpowers/research/2026-08-07-japan-parental-caps.md" }, "non_cash.parental_leave_capped": { "source": "MHLW leaflet LL080731 ho 02 (Reiwa 8 revision, in force 2026-08-01): ikuji kyugyo kyufu-kin shikyu jogen-gaku JPY 332,454/30-day unit at 67% and JPY 248,100 at 50%, wage-month base cap (kyugyo kaishi-ji chingin getsugaku jogen) JPY 496,200; fetched 2026-08-07 via https://roumu.com/pdf/2026080314.pdf (announcement https://roumu.com/archives/132696.html; MHLW hub https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000160564_00050.html); prior-year values confirmed against MHLW LL070722 https://www.mhlw.go.jp/content/001520023.pdf. Statutes: Employment Insurance Act art. 61-7 (ikuji kyugyo kyufu-kin), Health Insurance Act art. 102 (shussan teate-kin)", "last_updated": "2026-08-07", "confidence": "high", "note": "All 110 weeks flattened to ONE capped block, elbow-exact at the wage-month base cap: pct 60 (duration-weighted blend, rounded down) x JPY 496,200 = cap 297,720/month. Maternity (14 wk at 2/3, own std-remuneration ceiling JPY 1.39M/month) is FOLDED UNDER the shared ikuji ceiling per the Austria fold + Chile/Kazakhstan flatten precedents; true duration-weighted ceiling is ~298,064, so the stored 297,720 understates Japan (conservative). The two ikuji phases (67% first 180 days per parent, 50% after) share the same 496,200 elbow, so their flatten is lossless. No floor stored: the statutory kagen (96,090 wage-month) implies ~57,654/month, sub-ICP-relevant; omitting it is the less favorable direction. Ceilings revise every Aug 1 (indexed to average wages) - annual refresh row. Research: docs/superpowers/research/2026-08-07-japan-parental-caps.md" }, "non_cash.commute_minutes_avg": { "source": "Statistics Bureau of Japan, Shakai Seikatsu Kihon Chosa 2021, prefecture ranking table rank12. https://www.stat.go.jp/data/shakai/2021/rank/zuhyou/rank/rank12.xlsx", "last_updated": "2026-08-08", "confidence": "high", "statistic": "1-nichi atari no tsukin tsugaku jikan (commuting-to-work-and-school time per day)", "year": "2021", "note": "CONVERSION: Tokyo-to publishes 1h35m = 95 min, and the table's own footnote reads 'per day, persons aged 10 and over, weekday average', which is a ROUND TRIP; 95/2 = 47.5 one way. CAVEATS carried deliberately: the statistic bundles commuting to SCHOOL with commuting to work, and its universe starts at age 10, so schoolchildren sit inside the mean. Survey self-report. Replaces a citation to a 'Recruit Works Institute / Tokyo Commuter Study' that could not be found to exist." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Tokyo: JPY 8,920,499, held in labor_market.tech_salary_distribution (source_slug tokyo-jpn, as_of 2026-07-18, table confidence usable). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "JPY 8,920,499 / 158.02 = USD 56,452; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2258, R5-rounded down to 0.22. REPLACES 0.37 (2026-08-11), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Tokyo by 68%. The measured p50 was 0.610 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): JPY 11,715,704 / 158.02 = USD 74,141; / USD 196,000 Seattle senior reference = 0.378, R5-rounded down to 0.37. REPLACES 0.48 (2026-05-15), whose own note said '0.45 used' (the file never carried that value), quoted a senior PM rather than an SWE, and stated a ratio range of 0.40-0.50 that its own figures (USD 67-87k over USD 200k) do not produce. MHLW publishes a MONTHLY kimatte shikyu suru genkin kyuyogaku and a SEPARATE annual bonus field, and Japanese bonuses run 3-5 months, so monthly x12 alone understates annual pay by 20-30%. levels.fyi Tokyo skews foreign/big-tech and sits far above the domestic SIer market, so 0.37 is the GENEROUS end of the evidence; a domestic reading gives ~0.23. Working: docs/superpowers/research/2026-08-11-comp-index-sweep-a.md" }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/tokyo-jpn", "last_updated": "2026-07-18", "confidence": "usable", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "EF English Proficiency Index 2024 (Low band)", "last_updated": "2024", "note": "National-level English usability for tech work. Band: low." }, "healthcare.statutory_employee": { "source": "kyoukaikenpo.or.jp FY2026 per-prefecture insurance rate tables (Tokyo 9.85% total, employee half 4.925%; Kanagawa/Aichi/Osaka/Fukuoka within 9.85-10.13%); nenkin.go.jp standard remuneration grade table (health grade 50 = JPY 1,390,000/month) and the JPY 5,730,000 fiscal-year bonus ceiling (1 April to 31 March) shared by health, kaigo and the support levy; mhlw.go.jp FY2026 kodomo-kosodate shienkin 0.23% total / 0.115% employee, first charged on the April 2026 premium.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll. Tokyo's 9.85% is used as the national proxy because the block is enforced identical across a country's cities and the spread across our five metros is 0.07pp of employee rate. DELIBERATE EXCLUSIONS: kaigo 0.81% employee is age-gated to 40-64 and computeStatutoryHealthContribution cannot read age_bracket, so charging it would be wrong for the sub-40 half of our ICP (at the ceiling it would add about JPY 181,500/yr); the 0.115% kodomo-kosodate shienkin rides the same grades and bonus ceiling but funds child support, not health (about JPY 25,800/yr at the ceiling). The 2025 pension reform raising the pension ceiling to JPY 750,000 is not yet in force. TODO (phase-ins 2027-09, 2028-09, 2029-09): apply each step only when it takes effect." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "MHLW coinsurance schedule (30% for ages 6-69, 20% under 6) + the income-banded kougaku ryouyouhi monthly high-cost ceiling + Tokyo ward-level child medical subsidy", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 held at 50,000, structurally plausible against the high-cost ceiling, which is mid-revision (rising August 2026) and so usable only as an approximate check. Tokyo's ward child medical subsidy makes children close to free in practice, but ward income limits leave real uncertainty for a high earner, so the child increment is small rather than zero. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "NTA tax answer 1135 https://www.nta.go.jp/taxes/shiraberu/taxanswer/shotoku/1135.htm ; MHLW iDeCo contribution limits https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/nenkin/nenkin/kyoshutsu/taishousha.html", "last_updated": "2026-08-11", "confidence": "medium", "note": "iDeCo, deductible in full under the small-enterprise-mutual-aid category; the binding ceiling lives in the pension law and is stated MONTHLY, converted x12 here. Category-2 insured WITH a corporate DC/DB plan: JPY 20,000/mo = 240,000/yr. Without one it is 23,000/mo = 276,000/yr. The lower figure is the house-rule choice, because a senior engineer at a large employer usually has a corporate pension; that is a BASIS choice, not a magnitude, and it is recorded in docs/corrections-ledger.md. Separate from the already-modeled kosei nenkin. Life- and annuity-insurance premium relief is deliberately not stacked. REFRESH ITEM: from 1 Dec 2026 the limit becomes 62,000/mo less employer DC/DB and the 20,000 sub-cap is abolished." } }, "london": { "non_cash.parental_leave_flat": { "source": "Social Security Benefits Up-rating Order 2026 (SI 2026/148), arts. 9-10: for GBP 187.18 substitute GBP 194.32, applied to SMP, SPP, adoption pay, ShPP and neonatal pay https://www.legislation.gov.uk/uksi/2026/148/contents/made ; gov.uk https://www.gov.uk/maternity-pay-leave/pay", "last_updated": "2026-08-01", "confidence": "high", "note": "GBP 194.32 per week for tax year 2026/27, effective 5 April 2026 for SMP and 6 April 2026 for SPP and ShPP; the prior rate was GBP 187.18. Covers SMP weeks 7 to 39 (33 weeks) plus the 2 weeks of Statutory Paternity Pay. Statute pays the LOWER of the flat rate and 90% of average weekly earnings, so an ICP always receives the flat rate. Note: gov.uk/employers-maternity-pay-leave still showed the stale 187.18 on 2026-08-01; cite the employee-facing pages. The 2026/27 Lower Earnings Limit was not sourced and is not load-bearing, since our user clears it." }, "tax.inbound_regime": { "source": "4-year FIG regime (from 2025-04-06) relieves FOREIGN income and gains only, not a UK salary: HMRC RFIG41000 https://www.gov.uk/hmrc-internal-manuals/residence-and-fig-regime-manual/rfig41000 . Overseas Workday Relief covers only earnings for duties NOT performed in the UK (EIM43565) and is capped from 2025-04-06 at the lower of 30% of relevant qualifying employment income or GBP 300,000 (EIM43600); conditions incl. 10 immediately preceding tax years of non-UK residence (EIM43560). https://www.gov.uk/hmrc-internal-manuals/employment-income-manual/eim43600", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: re-verified against HMRC manuals. No salary relief remains correct for 2026: a resident performing all duties in the UK gets nothing from FIG (foreign income only) and nothing from OWR (non-UK duties only). Note sharpened with the new GBP 300,000 / 30% OWR cap and the 10-year prior non-residence test. No percentage discount added, deliberately: OWR turns on a UK/non-UK workday split we never ask for, and the conservative default for a relocating employee is 100% UK workdays." }, "tax.paye_brackets": { "source": "HMRC income tax rates; additional-rate threshold 125,140 taxable + PA taper over 100k (gov.uk/income-tax-rates); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "England/Wales/NI 2025-26 PAYE bands (20/40/45 above £12,570 personal allowance); frozen through April 2031" }, "tax.national_insurance_rate": { "source": "HMRC https://www.gov.uk/national-insurance/how-much-you-pay", "last_updated": "2026-05-15", "note": "Class 1 employee NI: 8% between primary threshold and UEL" }, "tax.national_insurance_upper_rate": { "source": "HMRC NI rates 2025-26", "last_updated": "2026-05-15", "note": "2% on earnings above £50,270" }, "tax.sales_tax_rate": { "source": "HMRC https://www.gov.uk/vat-rates", "last_updated": "2026-05-15", "note": "UK VAT standard rate 20% since 2011" }, "tax.property_tax_rate_pct": { "source": "Greater London Properties / borough council tax bands", "last_updated": "2026-05-15", "confidence": "medium", "note": "Council tax is band-based on 1991 valuations; ~0.2% effective on current market value for Band D in typical London borough (~£1,500-£2,000/yr)" }, "tax.personal_allowance": { "source": "HMRC Income tax allowances 2025-26", "last_updated": "2026-05-15", "note": "£12,570 frozen through April 2031" }, "housing.median_rent_by_bedroom": { "source": "Investropa London Rents 2026 / Numbeo May 2026", "last_updated": "2026-05-15", "confidence": "medium", "note": "Zones 1-3 median across boroughs; central London zone 1 runs significantly higher" }, "housing.rent_percentiles": { "source": "ONS, Private rental market in London: April 2025 to March 2026, released 27 Apr 2026, https://www.ons.gov.uk/economy/inflationandpriceindices/adhocs/3389privaterentalmarketinlondonapril2025tomarch2026 (real LQ/median/UQ by bedroom). Envelope across studio-3BR: P25 = min(0.82, 0.84, 0.87, 0.85), P75 = max(1.19, 1.19, 1.24, 1.26). Verified live 2026-08-07.", "last_updated": "2026-08-07", "confidence": "high", "note": "Within-bedroom basis, the only observed asking-relevant quartile source in the research. Referenced-tenancy data; stored medians are asking rents, ratios transfer. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "housing.housing_districts": { "source": "ONS, Private rental market in London: April 2025 to March 2026 (same release as rent_percentiles): borough 2BR medians / London 2BR 1,900 (Kensington & Chelsea 3,563 = 1.88, Westminster 3,300 = 1.74, Hackney 2,400 = 1.26, Lambeth 2,250 = 1.18, Barking & Dagenham 1,700 = 0.89, Croydon 1,593 = 0.84).", "last_updated": "2026-08-07", "confidence": "high", "note": "Research tier label \"outer\" mapped to schema tier \"low\". ONS medians are referenced-tenancy; ratios applied to stored asking medians per research §3." }, "housing.median_home_price_by_bedroom": { "source": "Investropa London Housing Prices 2026 / Land Registry", "last_updated": "2026-05-15", "confidence": "medium", "note": "Studio price adjusted to ensure monotonic increase by bedroom count" }, "housing.utilities_monthly": { "source": "Numbeo London May 2026 (Energy Price Cap basis)", "last_updated": "2026-05-15" }, "housing.renters_insurance_monthly": { "source": "NimbleFins contents insurance 2026 average", "last_updated": "2026-05-15", "note": "UK-wide average; London may run 10-20% higher" }, "housing.internet_phone_monthly": { "source": "MoneySuperMarket broadband London May 2026", "last_updated": "2026-05-15", "note": "Includes broadband + mobile combined estimate" }, "healthcare.premium_monthly_by_family_size": { "source": "NHS (universal coverage; funded via National Insurance, not premiums)", "last_updated": "2026-05-15" }, "healthcare.annual_out_of_pocket_avg": { "source": "ONS Healthcare expenditure, UK Health Accounts 2023 and 2024 https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthcaresystem/bulletins/ukhealthaccounts/2023and2024 (household out-of-pocket GBP 46bn in 2024, of which ONS states ~38% is long-term care) + NHS England prescription charge and prepayment certificate", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "healthcare.dental_vision_monthly": { "source": "Private dental + optical insurance averages (WeCoVR, Specsavers)", "last_updated": "2026-05-15", "confidence": "medium" }, "childcare.infant_monthly": { "source": "Coram Family and Childcare Survey 2026 / Hatching Dragons London 2025", "last_updated": "2026-05-15", "note": "Full-time nursery 0-2yo; London highest in UK" }, "childcare.preschool_monthly": { "source": "Coram Family and Childcare Survey 2026", "last_updated": "2026-05-15", "note": "3-4yo with full-time wraparound; 30-hour gov entitlement reduces this" }, "childcare.after_school_monthly": { "source": "Coram + Trust for London school-age childcare research", "last_updated": "2026-05-15", "confidence": "medium", "note": "Highly variable by borough; Ealing most expensive" }, "childcare.subsidy_per_kid_annual": { "source": "https://www.gov.uk/free-childcare-if-working", "last_updated": "2026-08-05", "note": "ZEROED 2026-08-05 (CC5): the free-hours entitlement is a provider-funded fee reduction, never household cash, so under the framework it belongs inside preschool_monthly, same defect class as the 08-02 double-subtraction batch. Working: docs/superpowers/research/2026-08-05-cc5-brazil-vienna-sourcing.md", "confidence": "high" }, "food.groceries_monthly_per_adult": { "source": "Numbeo London + WeCovr UK Food Budget 2026", "last_updated": "2026-05-15" }, "food.groceries_monthly_per_kid": { "source": "ONS Living Costs and Food Survey + NimbleFins family-of-4 figures", "last_updated": "2026-05-15", "confidence": "medium" }, "food.dining_out_monthly_baseline": { "source": "Numbeo London Food Prices May 2026", "last_updated": "2026-05-15" }, "transit.monthly_pass": { "source": "Transport for London https://tfl.gov.uk/fares/find-fares/caps-and-travelcard-prices", "last_updated": "2026-05-15", "note": "Adult Travelcard zones 1-2; 2026 rates" }, "transit.annual_car_cost_avg": { "source": "AA/RAC + ALA Insurance 2026 + TfL Congestion Charge/ULEZ", "last_updated": "2026-05-15", "confidence": "medium", "note": "Includes potential congestion charge (£18/day) and ULEZ (£12.50/day) for non-compliant vehicles; varies widely" }, "discretionary.entertainment_monthly_baseline": { "source": "Numbeo London + Resamania gym + UK streaming calculator", "last_updated": "2026-05-15" }, "discretionary.local_vacation_annual_baseline": { "source": "Weather2Travel UK family holidays + Budget Your Trip 2026", "last_updated": "2026-05-15", "confidence": "medium" }, "travel_home.avg_round_trip_per_person_to_other_city": { "source": "KAYAK / Skyscanner LHR-SEA median 2026", "last_updated": "2026-05-15", "note": "Round-trip economy median across off-peak and peak booking windows" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "gov.uk Child Benefit, what you'll get https://www.gov.uk/child-benefit/what-youll-get ; HMRC rates by tax year https://www.gov.uk/government/publications/rates-and-allowances-tax-credits-child-benefit-and-guardians-allowance/tax-credits-child-benefit-and-guardians-allowance ; High Income Child Benefit Charge https://www.gov.uk/child-benefit-tax-charge (all fetched 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "STATUTORY additional-child rate, 2026-27 tax year: GBP 17.90 a week x 52 = 930.80 a year. The eldest or only child draws govt_benefits.child_allowance_eldest_annual instead. Means-tested by the ENGINE, not in this field: govt_benefits.child_allowance_taper carries the High Income Child Benefit Charge and computeGovtBenefits() withdraws the benefit per household on the higher earner's gross. Until 2026-09-12 this field stored 0 on the assumption that every published UK household's higher earner was over GBP 80,000. That holds for the USD 280,000 family scenario (London primary about GBP 142k, still 0 through the taper) and fails for the USD 150,000 two-earner persona (London higher earner GBP 74,063 keeps 30%, GBP 701.22 a year) and for any calculator user under the ceiling. Ledger decision 2026-09-12; research docs/superpowers/research/2026-09-12-child-benefit-audit-fixes.md." }, "govt_benefits.child_allowance_eldest_annual": { "source": "gov.uk Child Benefit, what you'll get https://www.gov.uk/child-benefit/what-youll-get ; HMRC rates by tax year https://www.gov.uk/government/publications/rates-and-allowances-tax-credits-child-benefit-and-guardians-allowance/tax-credits-child-benefit-and-guardians-allowance ; High Income Child Benefit Charge https://www.gov.uk/child-benefit-tax-charge (all fetched 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "Eldest or only child, 2026-27 tax year: GBP 27.05 a week x 52 = 1,406.60 a year. Unit: ANNUAL, from the statutory WEEKLY rate (HMRC pays every four weeks, 13 payments a year). Drawn by the first eligible child; every further child draws child_allowance_per_kid_annual. Same means-testing as that field: the taper block withdraws it." }, "govt_benefits.child_allowance_taper": { "source": "gov.uk Child Benefit, what you'll get https://www.gov.uk/child-benefit/what-youll-get ; HMRC rates by tax year https://www.gov.uk/government/publications/rates-and-allowances-tax-credits-child-benefit-and-guardians-allowance/tax-credits-child-benefit-and-guardians-allowance ; High Income Child Benefit Charge https://www.gov.uk/child-benefit-tax-charge (all fetched 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "High Income Child Benefit Charge: threshold GBP 60,000 of adjusted net income for tax years from 2024-25, 1% of the benefit for every GBP 200 over it, 100% at GBP 80,000 or more. Tested on the HIGHER earner (\"whoever has the higher income is responsible for paying the tax charge\"), never on the household total, so basis is higher_earner. The engine tests gross salary plus RSU as the conservative proxy for adjusted net income (ANI is net of pension contributions and Gift Aid, so it is lower and keeps more benefit); at the USD 150,000 persona that is a 70% charge on gross against about 48% on ANI. Whole-percent steps rounded down, as HMRC computes it. A deterministic statute, so no conservative-value tiebreak was applied to the BASIS; the gross proxy is the one conservative call inside it." }, "non_cash.vacation_days_statutory": { "source": "GOV.UK holiday entitlement rights https://www.gov.uk/holiday-entitlement-rights (Working Time Regulations 1998 reg. 13 + 13A)", "last_updated": "2026-08-08", "confidence": "high", "note": "CORRECTED 2026-08-08. GOV.UK: 5.6 weeks paid holiday a year, 'Statutory paid holiday entitlement is limited to 28 days', and 'An employer can choose to include bank holidays as part of a worker's statutory annual leave'. So the UK legal floor is 28 days TOTAL and the bank holidays sit INSIDE it, not on top. We split that as 20 here + 8 in public_holidays so both fields stay individually true and the engine's sum lands on the statutory 28. We previously stored 25, a typical-employer figure that summed to 33 and overstated the UK floor by 5 days. Basis is England and Wales (8 bank holidays), which is where London, Birmingham and Manchester sit; Scotland has 9 and Northern Ireland 10, but the 28-day cap is UK-wide and _national_fields.json requires one value per country, so Edinburgh carries the England and Wales split rather than an average." }, "non_cash.public_holidays": { "source": "GOV.UK bank holidays https://www.gov.uk/bank-holidays (England and Wales)", "last_updated": "2026-08-08", "confidence": "high", "note": "8 statutory bank holidays in England and Wales, the basis we model. Scotland has 9 and Northern Ireland 10; we do not average, and Edinburgh carries the England and Wales figure because _national_fields.json enforces one value per country. Read this together with vacation_days_statutory: GOV.UK says bank holidays 'do not have to be given as paid leave' and may be counted inside the 28-day statutory entitlement, so the 20 + 8 split is the floor, not 8 days on top of a separate 28." }, "non_cash.parental_leave_weeks": { "source": "gov.uk Statutory Maternity Pay https://www.gov.uk/maternity-pay-leave/pay ; Statutory Paternity Pay https://www.gov.uk/paternity-pay-leave/pay ; Shared Parental Leave and Pay https://www.gov.uk/shared-parental-leave-and-pay ; unpaid parental leave https://www.gov.uk/parental-leave/entitlement", "last_updated": "2026-08-01", "confidence": "high", "note": "41 PAID wk = SMP 39 (6 at 90% of average weekly earnings plus 33 flat) + Statutory Paternity Pay 2 (flat). Was 52, which INCLUDED the 13 unpaid weeks of the maternity year and violated the field definition. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. Shared Parental Leave adds ZERO weeks: gov.uk requires the mother to take less than the 39 weeks of maternity pay and use the rest as ShPP, so it redistributes her untaken entitlement and leaves the household ceiling at 39. Excluded: the 13-week unpaid SMP tail; 18 unpaid weeks per parent of parental leave; Neonatal Care Pay, which is conditional on the baby needing care; occupational maternity pay, which is contractual rather than statutory." }, "non_cash.parental_leave_paid_pct": { "source": "gov.uk Statutory Maternity Pay: the first 6 weeks are 90% of average weekly earnings before tax, with no cap https://www.gov.uk/maternity-pay-leave/pay", "last_updated": "2026-08-01", "confidence": "high", "note": "90% applies to the SIX percentage-paid weeks only. The other 35 weeks are flat-rate and live in non_cash.parental_leave_flat. The stored 40 was a hand-blend that folded the flat weeks in as though they tracked salary, and it also averaged across 13 unpaid weeks." }, "non_cash.commute_minutes_avg": { "source": "UK Department for Transport, National Travel Survey table NTS9914a, London ITL1 region. https://assets.publishing.service.gov.uk/media/68a4371132d2c63f869343d0/nts9914.ods", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Average trip duration by trip purpose and region of residence, commuting", "year": "2024", "note": "London region 41.1, and Greater London is the correct metro geography with no adjustment needed. NTS trips are one way. Corroborated by the TUC's Labour Force Survey analysis at 80.6 min round trip, which halves to 40.3. Replaces a citation half of which was a relocation agency's blog." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, London: GBP 99,953, held in labor_market.tech_salary_distribution (source_slug london-metro-area, as_of 2026-07-18, table confidence rich). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "GBP 99,953 / 0.742 = USD 134,708; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.5388, R5-rounded down to 0.53. REPLACES 0.75 (2026-05-15), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated London by 42%. The measured p50 was 0.718 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Mid-market median ~0.51 via Levels.fyi; FAANG/fintech in London reach 0.80-0.85; 0.75 reflects typical senior IC midpoint" }, "tax.filing_basis": { "source": "HMRC - independent taxation of spouses since 1990", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/london-metro-area", "last_updated": "2026-07-18", "confidence": "rich", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "ONS Healthcare expenditure, UK Health Accounts 2023 and 2024 https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthcaresystem/bulletins/ukhealthaccounts/2023and2024 (household out-of-pocket GBP 46bn in 2024, of which ONS states ~38% is long-term care) + NHS England prescription charge and prepayment certificate", "last_updated": "2026-08-08", "confidence": "medium", "note": "Almost NONE of this is NHS cost-sharing, and the _meta note has to say so: realized England prescription charges are only about GBP 10-12 per capita because ~89% of items are dispensed free. The figure is overwhelmingly OTC medicines, self-pay private treatment against NHS waiting lists, and private physiotherapy. The old GBP 400 sat ABOVE the entire statutory ceiling at every household size (England prepayment certificate GBP 114.50 per adult, family of four GBP 229 since children are free, and GBP 0 in Scotland, Wales and Northern Ireland). One UK-wide value is retained this pass; Edinburgh's free prescriptions are a real but small (GBP 10-15/yr) difference and the city-level split is queued. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "non_cash.sick_leave": { "source": "Social Security Contributions and Benefits Act 1992, s.151 (employer's duty to pay; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": "https://www.gov.uk/statutory-sick-pay/eligibility", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "HMRC, Pension schemes rates and allowances https://www.gov.uk/government/publications/rates-and-allowances-pension-schemes/pension-schemes-rates ; Pensions Tax Manual PTM057100 https://www.gov.uk/hmrc-internal-manuals/pensions-tax-manual/ptm057100 ; https://www.gov.uk/tax-on-your-private-pension/annual-allowance", "last_updated": "2026-08-11", "confidence": "high", "note": "Annual allowance, 2025-26 and 2026-27 (identical on every figure). A COMBINED input ceiling covering everything paid in by the member or anyone else including the employer, enforced by the annual allowance charge, plus the separate rule that relief cannot exceed 100% of relevant UK earnings. Taper is GBP 1 per GBP 2 of ADJUSTED income above 260,000 down to a 10,000 floor; our gross understates adjusted income (which adds employer contributions) so the taper bites later than reality. No less_employer_pct: auto-enrolment minimum is 3% of qualifying earnings, about GBP 1,321, and anything larger would be an assumption. Never binds at our incomes." } }, "amsterdam": { "tax.filing_basis": { "source": "PwC WWTS - Netherlands (Tax administration): spouses are taxed separately for employment/business income (Box 1); only Box 2/3 and home items are allocable between fiscal partners.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.inbound_regime": { "source": "30%-regeling (expatregeling). Belastingdienst Deskundigheidsvereiste, 2026 norm: a 'belastbaar jaarloon ... meer is dan EUR 48.013 (exclusief de gerichte vrijstelling)', i.e. measured on the salary REMAINING after the 30% carve-out, so about EUR 68,590 of gross: https://www.belastingdienst.nl/wps/wcm/connect/bldcontentnl/belastingdienst/zakelijk/internationaal/personeel/u_bent_niet_in_nederland_gevestigd_loonheffingen_inhouden/als_u_loonheffingen_gaat_inhouden/extraterritoriale_kosten_en_de_30procentregeling/voorwaarden_voor_de_30procentregeling1/deskundigheidsvereiste . Belastingdienst Inhoud van de expatregeling: maximum tax-free allowance EUR 78,600 in 2026, reached at the WNT norm of EUR 262,000. Max 60 months. Rijksoverheid, Belastingvoordeel buitenlandse werknemers.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31 OVER-GRANT FIX + refresh. (1) min_income_basis set to post_relief: the EUR 48,013 floor was being tested against GROSS, so every salary from EUR 48,013 up got the ruling; the statute measures it on the taxable remainder, which needs about EUR 68,590 of gross. Amsterdam single filer at EUR 60,000: tax 7,056 -> 15,906 (ordinary). (2) This entry previously cited the 2025 pair 46,660 / 246,000 while the data already carried the 2026 pair 48,013 / 262,000; the data was right and the prose was stale. (3) The under-30-with-a-master norm of EUR 36,497 is deliberately NOT used: we model neither age nor degree, so we carry the higher general norm. SCHEDULED: allowance drops 30% -> 27% and the norms step up from 2027-01-01 (Belastingplan 2025, second nota van wijziging). The calculator models the current tax year. TODO (2027-01 refresh): apply the 27% and the stepped-up norms once in force for the modeled year." }, "tax.box1_brackets": { "source": "BDO https://www.bdo.global/en-gb/insights/tax/expatriate-tax/netherlands-changes-to-income-tax-brackets-and-rates-for-2026-enter-into-effect", "last_updated": "2026-05-16", "confidence": "high", "note": "NL 2026 Box 1 COMBINED rates (income tax + AOW/ANW/WLZ social insurance): 35.75% to €38,883; 37.56% to €78,426; 49.50% above" }, "tax.social_insurance_portion_rate": { "source": "BDO Global / Deloitte NL Tax Budget 2026", "last_updated": "2026-05-16", "confidence": "high", "note": "AOW 17.90% + ANW 0.10% + WLZ 9.65% = 27.65% of first-bracket combined rate; applies only up to first bracket cap €38,883" }, "tax.general_tax_credit_max": { "source": "Deloitte NL Belastingplan 2026 / Belastingdienst https://www.belastingdienst.nl/wps/wcm/connect/bldcontentnl/belastingdienst/prive/inkomstenbelasting/heffingskortingen_boxen_tarieven/heffingskortingen/algemene_heffingskorting/tabel-algemene-heffingskorting-2026", "last_updated": "2026-05-16", "confidence": "high", "note": "Algemene heffingskorting 2026 max €3,115; phases out 6.398% above €29,736 to €0 at €78,426" }, "tax.labour_tax_credit_max": { "source": "Deloitte NL Belastingplan 2026", "last_updated": "2026-05-16", "confidence": "high", "note": "Arbeidskorting 2026 max €5,685; phases out at ~6.51% above ~€45,592 to €0 at ~€132,920" }, "tax.sales_tax_rate": { "source": "business.gov.nl https://business.gov.nl/regulation/vat/", "last_updated": "2026-05-16", "confidence": "high", "note": "BTW standard rate 21% (reduced 9% applies to food/books/medicines)" }, "tax.property_tax_rate_pct": { "source": "City of Amsterdam OZB https://www.amsterdam.nl/en/municipal-taxes/property-tax-ozb/", "last_updated": "2026-05-16", "confidence": "high", "note": "Amsterdam residential OZB 2026 = 0.0527% of WOZ value (up from 2025's 0.0431%)" }, "housing.median_rent_by_bedroom": { "source": "Pararius / Investropa Q3 2025 https://www.pararius.com/expat-guide/rental-report-q1-2025", "last_updated": "2026-05-16", "confidence": "medium", "note": "Amsterdam free-sector rents Q3 2025; avg €27.75/sqm. Regulated social housing is cheaper but supply-constrained; these reflect open market" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4).", "last_updated": "2026-08-07", "confidence": "medium", "note": "Multiplier band only by decision: citywide Amsterdam sources disagree by up to 50% for the same quarter and no district table survived cross-checking (research §3). Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "housing.median_home_price_by_bedroom": { "source": "Global Property Guide / Funda Q3 2025 https://www.globalpropertyguide.com/europe/netherlands/square-meter-prices", "last_updated": "2026-05-16", "confidence": "medium", "note": "Amsterdam avg transaction Q3 2025 €658k at €8,559/sqm; derived by-bedroom using typical sizes" }, "housing.utilities_monthly": { "source": "Numbeo Amsterdam May 2026", "last_updated": "2026-05-16", "confidence": "high", "note": "Electricity + heating + water + garbage for 85sqm apartment" }, "housing.renters_insurance_monthly": { "source": "IamExpat / Dutchpat https://www.iamexpat.nl/expat-info/dutch-insurance/home-insurance-netherlands", "last_updated": "2026-05-16", "confidence": "medium", "note": "Typical inboedelverzekering (contents) €5-20/month" }, "housing.internet_phone_monthly": { "source": "Numbeo / NLCompass 2026", "last_updated": "2026-05-16", "confidence": "high", "note": "Internet ~€40 + mobile SIM-only ~€20 bundle; premium bundled (KPN/Ziggo) runs higher" }, "healthcare.premium_monthly_by_family_size": { "source": "Zorgwijzer / Zorgverzekeraars Nederland https://www.zorgwijzer.nl/zorgverzekering-2025", "last_updated": "2026-05-16", "confidence": "high", "note": "Basisverzekering nominal premium 2025 ~€150/month per adult; children under 18 are free, so 3- and 4-person families = 2 adults × 150" }, "healthcare.premium_age_multiplier": { "source": "Rijksoverheid https://www.government.nl/topics/health-insurance/standard-health-insurance", "last_updated": "2026-05-16", "confidence": "high", "note": "Dutch Zvw premiums are community-rated; insurers cannot price by age" }, "healthcare.annual_out_of_pocket_avg": { "source": "Zvw statutory eigen risico (EUR 385 per adult, under-18s exempt) + Zorginstituut Nederland and Vektis data on what share of insured actually exhaust it", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "healthcare.dental_vision_monthly": { "source": "Independer https://www.independer.nl/zorgverzekering/info/tandartsverzekering.aspx", "last_updated": "2026-05-16", "confidence": "medium", "note": "Supplementary dental coverage €15-35/month; vision typically bundled" }, "childcare.infant_monthly": { "source": "IJsterk Amsterdam + nlcompass https://ijsterk.nl/en/childcare-costs/", "last_updated": "2026-05-16", "confidence": "high", "note": "Full-time ~140 hrs/month at Amsterdam avg €12.50/hr; gross before subsidy" }, "childcare.preschool_monthly": { "source": "IJsterk Amsterdam (kinderdagverblijf 0-4)", "last_updated": "2026-05-16", "confidence": "high", "note": "NL kinderdagverblijf covers 0-4; same hourly rate as infant care. Gross before subsidy" }, "childcare.after_school_monthly": { "source": "Kindergarden BSO + CompaNanny", "last_updated": "2026-05-16", "confidence": "high", "note": "BSO ~48 hrs/month at €10.75-12.87/hr Amsterdam range" }, "childcare.subsidy_per_kid_annual": { "source": "rijksoverheid.nl kinderopvangtoeslag 2025; Belastingdienst toeslag tables", "last_updated": "2026-07-17", "note": "kinderopvangtoeslag at the 33.3% high-income floor (2025, households >= EUR 226,546): 33.3% x capped rate EUR 10.71/hr x ~1,770 reimbursable h/yr ~= EUR 6,300 per full-time daycare child. The reimbursement % is means-tested, but 33.3% is the statutory MINIMUM, so it survives at the EUR 226k-269k published NL household (was overstated at a mid-income % of ~60-70%). 2026 raises the floor to 36.5%." }, "food.groceries_monthly_per_adult": { "source": "Numbeo Amsterdam markets + Albert Heijn/Jumbo typical weekly spend", "last_updated": "2026-05-16", "confidence": "high" }, "food.groceries_monthly_per_kid": { "source": "CBS Netherlands household expenditure", "last_updated": "2026-05-16", "confidence": "medium", "note": "~53% of adult grocery spend per CBS consumption data" }, "food.dining_out_monthly_baseline": { "source": "Numbeo Amsterdam restaurants", "last_updated": "2026-05-16", "confidence": "high", "note": "Inexpensive meal ~€18; mid-range 3-course for 2 ~€80. Baseline assumes 6-8 casual outings/month" }, "transit.monthly_pass": { "source": "GVB https://www.gvb.nl/en/travelling/season-tickets", "last_updated": "2026-05-16", "confidence": "high", "note": "GVB Amsterdam city zone monthly pass ~€99-105" }, "transit.annual_car_cost_avg": { "source": "Nibud autokosten https://www.nibud.nl/onderwerpen/uitgaven/autokosten/", "last_updated": "2026-05-16", "confidence": "medium", "note": "Mid-size car €5,500-7,500 total annual + Amsterdam parking permit €200-600/yr. Car ownership in central Amsterdam unusual" }, "discretionary.entertainment_monthly_baseline": { "source": "Numbeo Amsterdam + provided context", "last_updated": "2026-05-16", "confidence": "medium", "note": "Composed: gym ~€40 + 2 cinema ~€28 + streaming ~€25 + occasional outings ~€55" }, "discretionary.local_vacation_annual_baseline": { "source": "CBS Netherlands household expenditure + Airbnb reference", "last_updated": "2026-05-16", "confidence": "medium", "note": "3 long weekends in NL/BE/DE (~€500-700 each) + one ~1-week regional trip (~€1200). Excludes long-haul" }, "travel_home.avg_round_trip_per_person_to_other_city": { "source": "Google Flights AMS-SEA economy survey", "last_updated": "2026-05-16", "confidence": "medium", "note": "Mid-range round-trip AMS-SEA booked 4-8 weeks out averages €800-1000; peak €1100-1400" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Algemene Kinderbijslagwet art. 12, consolidated text valid from 01-08-2026, https://wetten.overheid.nl/BWBR0002368/ (fetched 2026-09-12, two independent reads agree; svb.nl returned 403 in English and Dutch and rijksoverheid.nl 404 on 2026-09-10 and 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "Dutch kinderbijslag: universal, NOT means-tested (the means-tested kindgebonden budget is separate and excluded). Statutory amounts per QUARTER per child, band set by the child's age on the first day of the quarter: EUR 298.40 under 6, 362.35 for 6 to 11, 426.29 for 12 to 17 (AKW art. 12 lid 3). The engine has no per-age benefit hook (computeGovtBenefits() reads one rate and only tests age < 18), so ONE rate is stored for every age and the Rule 5 tiebreak picks the LOWEST band: 298.40 x 4 = 1,193.60 a year, stored as 1,194. That understates a 6-to-11-year-old by EUR 255 a year and a 12-to-17-year-old by EUR 511, conservative against the Netherlands. Set 2026-09-12; was 1,368 = 342.06 x 4, which is the 6-to-11 band of the 2025 H1 table mislabelled as 2026. Ledger 2026-09-12; research docs/superpowers/research/2026-09-12-child-benefit-audit-fixes.md F4." }, "non_cash.vacation_days_statutory": { "source": "Netherlands business.gov.nl https://business.gov.nl/regulation/holidays-leave/", "last_updated": "2026-05-16", "confidence": "high", "note": "Statutory minimum 4× weekly hours = 20 days FT; many employers offer 24-25" }, "non_cash.public_holidays": { "source": "Government of the Netherlands https://www.government.nl/topics/working-hours/question-and-answer/public-holidays-in-the-netherlands", "last_updated": "2026-08-08", "confidence": "high", "note": "RE-VERIFIED 2026-08-08, value stands at 11 and the previous note claiming 9 was wrong. The Dutch government's own list names 11: Nieuwjaarsdag, Goede Vrijdag, Eerste and Tweede paasdag, Koningsdag, Bevrijdingsdag, Hemelvaartsdag, Eerste and Tweede pinksterdag, Eerste and Tweede kerstdag. The 9 came from netting off Easter Sunday and Whit Sunday, which always fall on a Sunday, but that is the weekday-capable basis and this field is the nominal statutory count everywhere else. Real caveat, unchanged by the count: the same page says there is no law requiring a day off on any of them, so paid time off depends on the CAO or the contract. Goede Vrijdag is a normal working day under most private CAOs and Bevrijdingsdag is a paid day off mainly in lustrum years." }, "non_cash.parental_leave_weeks": { "source": "Wet arbeid en zorg (WAZO) art. 3:1, 4:2, 4:2a, 6:2, 6:3 https://wetten.overheid.nl/BWBR0013008 (geldig van 01-01-2025) ; Rijksoverheid https://www.rijksoverheid.nl/vraag-en-antwoord/ouderschapsverlof/recht-op-ouderschapsverlof ; business.gov.nl https://business.gov.nl/regulation/leave-schemes/", "last_updated": "2026-08-01", "confidence": "high", "note": "40 wk = zwangerschaps- en bevallingsverlof 16 (mother) + geboorteverlof 1 (partner) + aanvullend geboorteverlof 5 (partner) + betaald ouderschapsverlof 9 per parent x 2. WAZO gives ouderschapsverlof to each parent individually with no transfer mechanism (Beide ouders hebben recht op ouderschapsverlof). Was 9, which held the paid parental weeks for ONE parent and dropped 16 weeks of maternity at 100%. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Excluded: the 17 unpaid weeks per parent of the 26-week ouderschapsverlof, and the 20-week multiples maternity variant." }, "non_cash.parental_leave_paid_pct": { "source": "WAZO art. 4:2 (geboorteverlof met behoud van loon, employer-paid) https://wetten.overheid.nl/BWBR0013008", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the 1-week partner geboorteverlof, employer-paid wage continuation with no dagloon ceiling (the structural twin of Germany Mutterschutz). The other 39 UWV-paid weeks moved to non_cash.parental_leave_capped in the 2026-08-06 caps rebuild; this field was 83, the whole-blend headline with the maximumdagloon ignored." }, "non_cash.universal_healthcare": { "source": "Government of the Netherlands https://www.government.nl/topics/health-insurance", "last_updated": "2026-05-16", "confidence": "high", "note": "NL has MANDATORY private insurance, not tax-funded universal. Set false" }, "non_cash.commute_minutes_avg": { "source": "CBS StatLine, ODiN dataset 85055NED, motive Van en naar het werk, province Noord-Holland. https://opendata.cbs.nl/statline/#/CBS/nl/dataset/85055NED", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Gemiddelde reisduur per verplaatsing (average travel time per trip)", "year": "2024", "note": "31.79 min; the zeer sterk stedelijk cut is 32.76. GEOGRAPHY: province of residence is the finest cut ODiN publishes. Replaces a Numbeo citation. Survey self-report (travel diary)." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Amsterdam: EUR 102,818, held in labor_market.tech_salary_distribution (source_slug greater-amsterdam-area, as_of 2026-07-18, table confidence usable). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 102,818 / 0.867 = USD 118,591; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.4744, R5-rounded down to 0.47. REPLACES 0.72 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Amsterdam by 53%. The measured p50 was 0.659 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE Amsterdam median TC €130-150k (base €80-130k + bonus/equity). Used midpoint €140k = ~$151k. Seattle senior SWE ~$220k all-in median. Ratio 0.69; cross-checked Levels.fyi tech-weighted (Booking/Adyen/Optiver) pulls higher; median ~0.72. 30% ruling for qualifying expats not reflected here (gross comp index); shows up in take-home" }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/greater-amsterdam-area", "last_updated": "2026-07-18", "confidence": "usable", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "EF English Proficiency Index 2024 (Very High band)", "last_updated": "2024", "note": "National-level English usability for tech work. Band: very_high." }, "non_cash.parental_leave_capped": { "source": "UWV maximumdagloon: EUR 309,91 bruto per dag from 1 July 2026 (Wfsv art. 17 lid 1) https://www.uwv.nl/nl/premies-bedragen/maximum-dagloon ; UWV: aanvullend geboorteverlof and betaald ouderschapsverlof pay 70% van het (gemaximeerde) dagloon https://www.uwv.nl/nl/aanvullend-geboorteverlof/hoogte-uitkering-aanvullend-geboorteverlof-betaald-ouderschapsverlof ; WAZO arts. 3:13, 4:2b, 6:3", "last_updated": "2026-08-06", "confidence": "high", "note": "PRE-CONVERTED to a week cap per the calendar-day convention: the dagloon is a 5-WORKDAY wage cap applied BEFORE the percentage, so we store the weekly PAYOUT ceiling at the 70% tier, 309.91 x 0.70 x 5 = EUR 1,084.69/week, over all 39 UWV weeks at the 83 blend (16 wk maternity 100% + 23 wk at 70%). Conservative: the 16 maternity weeks actually cap higher (100% tier, EUR 1,549.55/wk), and pricing the whole block at the lower ceiling reads them LOW. Cap steps semi-annually with the minimum wage (1 Jul 2026 value stored). At EUR 90,000 the blended effective rate is ~64%. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Netherlands)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Zvw statutory eigen risico (EUR 385 per adult, under-18s exempt) + Zorginstituut Nederland and Vektis data on what share of insured actually exhaust it", "last_updated": "2026-08-08", "confidence": "medium", "note": "REPLACES an exact statutory maximum: EUR 385 is precisely the single-adult compulsory deductible, and only about half of insured adults exhaust it, so the realized figure is EUR 229-236 by two converging routes. The deductible is charged PER ADULT and under-18s are wholly exempt, and GP care, maternity care and district nursing sit outside it entirely, so a Dutch family of four is essentially two adults. Aanvullende verzekering is a PREMIUM and is not booked here. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "non_cash.sick_leave": { "source": "Burgerlijk Wetboek (BW) Boek 7, Artikel 629 (loondoorbetaling bij ziekte).; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.uwv.nl/particulieren/bedragen/detail/maximumdagloon", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Belastingdienst Centraal Aanspreekpunt Pensioenen, AOW-franchises https://centraalaanspreekpuntpensioenen.belastingdienst.nl/publicaties/overzicht-aow-inbouwbedragen-en-aow-franchises/ ; V&A 25-008 https://centraalaanspreekpuntpensioenen.belastingdienst.nl/publicaties/va-25-008-v251209/ ; Kennisgroepen KG:070:2024:3 jaarruimte https://kennisgroepen.belastingdienst.nl/publicaties/kg07020243-vragen-en-antwoorden-over-de-jaarruimteberekening-en-de-opbouw-van-pensioenaanspraken/ ; ABP 2026 premie split", "last_updated": "2026-08-11", "confidence": "low", "note": "THIS VALUE IS A MODEL, NOT A STATUTE, and the formula is what is HIGH confidence. Art. 3.127 Wet IB 2001 post-Wtp: jaarruimte = 30% of premiegrondslag MINUS the actual ouderdomspensioen premium already paid (employer + employee). 2026 AOW-franchise EUR 19,172, aftoppingsgrens EUR 137,800, so the statutory maximum is 0.30 x (137,800 - 19,172) = EUR 35,589 (encoded as 13,085 here, see below). The naive statutory mapping is INERT at our incomes and would look like a fix while changing nothing. 13,085 = 11.03% of premiegrondslag, i.e. 30% less ABP's 18.97% employer share; 0.082 is the lowest in-band gross ratio (at EUR 75k), so the mapping sits at or below the derived truth everywhere above EUR 75k. Expressing this via less_employer_pct after the 2026-08-11 engine reorder was considered and rejected: it cannot subtract the AOW-franchise and would hand every Dutch earner about EUR 2,115 of extra room." } }, "berlin": { "tax.inbound_regime": { "source": "Germany has no inbound or expat income-tax regime; ordinary progressive tax applies. Verified 2026-07-31 against the 2025 CDU/CSU-SPD coalition agreement (zero occurrences of a skilled-worker tax rebate; the only measure is an administrative Work-and-stay-Agentur) https://www.cdu.de/app/uploads/2025/04/Koalitionsvertrag-%E2%80%93-barrierefreie-Version.pdf and against the Steueraenderungsgesetz 2025 in force 2026-01-01 https://www.bundesfinanzministerium.de/Content/DE/Pressemitteilungen/Finanzpolitik/2025/09/2025-09-10-steueraenderungsgesetz-2025.html", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: re-verified. The 2024 proposal for a declining 30/20/10% rebate for incoming skilled workers never reached a Gesetzentwurf and died with the traffic-light coalition in November 2024. Nothing of the kind appears in the 2025 coalition agreement or in the Steueraenderungsgesetz 2025. Recommend a standing December re-check against the BMF press-release index, because the idea is revived by German industry most years and would arrive via an annual Jahressteuergesetz." }, "tax.income_tax_tariff_32a": { "source": "§ 32a Abs. 1 EStG, Veranlagungszeitraum 2026 https://www.gesetze-im-internet.de/estg/__32a.html", "last_updated": "2026-07-31", "confidence": "high", "note": "The statutory continuous tariff, verbatim: bis 12 348 EUR = 0; von 12 349 bis 17 799 = (914,51 y + 1 400) y; von 17 800 bis 69 878 = (173,10 z + 2 397) z + 1 034,87; von 69 879 bis 277 825 = 0,42 x - 11 135,63; ab 277 826 = 0,45 x - 19 470,38. Replaces the 5-bracket approximation that flattened the 17,799-69,878 progression zone to 33.5% and overtaxed the bottom two thirds of it (Berlin 60k single: 14,628 charged vs 9,410 on a reader's payslip). Independent check: the legislator holds the zone-3 intercept d = 2 397 fixed across years, so a = 4,985,000 / (zone2_end - Grundfreibetrag) and c = (4200 - d) / (2 z_max); that identity reproduces the published 2023, 2024, 2025 and 2026 coefficients exactly. Excluded: the § 32a Abs. 1 Satz 2 rounding of zvE and the resulting tax down to full euros (worth under EUR 1, and skipping it errs high)." }, "tax.werbungskosten_pauschbetrag": { "source": "§ 9a Satz 1 Nr. 1 Buchst. a EStG https://www.gesetze-im-internet.de/estg/__9a.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Arbeitnehmer-Pauschbetrag EUR 1,230, granted PER EMPLOYED PERSON, so a two-earner household deducts 2,460. Conservative: real filers who itemise Werbungskosten (commuting, home office, union dues) deduct more than the Pauschbetrag, which would lower German tax further." }, "tax.sonderausgaben_pauschbetrag": { "source": "§ 10c EStG https://www.gesetze-im-internet.de/estg/__10c.html", "last_updated": "2026-07-31", "confidence": "high", "note": "EUR 36, doubled to 72 when spouses are assessed together. Immaterial in size, included so the base matches the statute." }, "tax.vorsorge_deductible_rate_rv": { "source": "§ 10 Abs. 1 Nr. 2 und Abs. 3 EStG https://www.gesetze-im-internet.de/estg/__10.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Altersvorsorgeaufwendungen: the employee's 9.3% Rentenversicherung share, 100% deductible since 2023 (Abs. 3 phases 76% in 2013 up 2pp a year to 100% ab 2023), computed against the RV Beitragsbemessungsgrenze (social_insurance_cap_rvalv, 101,400). The Höchstbetrag (the knappschaftliche Höchstbeitrag, roughly EUR 31k) sits far above 9.3% x 101,400 = 9,430, so it never binds at any income this calculator models and is not implemented." }, "tax.vorsorge_deductible_rate_kvpv": { "source": "§ 10 Abs. 1 Nr. 3 und Abs. 4 Satz 4 EStG https://www.gesetze-im-internet.de/estg/__10.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Basisabsicherung Kranken- und Pflegeversicherung, deductible in full above the EUR 1,900 / 2,800 Höchstbetrag (Abs. 4 Satz 4). KV employee share 8.75% reduced by 4% because the contribution carries a Krankengeld entitlement (8.75 x 0.96 = 8.40) plus PV 1.8 = 10.2%, against the KV/PV Beitragsbemessungsgrenze (social_insurance_cap, 69,750). Conservative choices: the 4% Krankengeld haircut is applied to the whole KV contribution including the Zusatzbeitrag, and ALV (1.3%) is NOT deducted at all because the sonstige-Vorsorge cap is already exhausted by KV/PV. Both readings raise our German tax." }, "tax.solidaritaetszuschlag_threshold": { "source": "Techniker Krankenkasse / Finanztip 2026 https://www.tk.de/firmenkunden/service/fachthemen/fachthema-beitraege/solidaritaetszuschlag-2075802", "last_updated": "2026-05-16", "confidence": "high", "note": "Singles 2026 Freigrenze = €20,350 income tax (up from €19,950 in 2025). Married joint = €40,700. Milderungszone above this; simplified to flat 5.5% above threshold" }, "tax.social_insurance_rate": { "source": "lohn-info.de Sozialversicherung 2026 https://www.lohn-info.de/sozialversicherungsbeitraege2026.html", "last_updated": "2026-05-16", "confidence": "high", "note": "Combined employee share 2026: KV 7.3% + Zusatzbeitrag/2 ~1.45% + PV 1.8% + RV 9.3% + ALV 1.3% = 21.15%. Avg Zusatzbeitrag rose to ~2.9% in 2026" }, "tax.social_insurance_cap": { "source": "Bundesregierung Beitragsbemessungsgrenzen 2026 https://www.bundesregierung.de/breg-de/aktuelles/beitragsgemessungsgrenzen-2386514", "last_updated": "2026-07-17", "confidence": "high", "note": "Lower KV/PV ceiling 2026 = EUR 69,750. RV/ALV continue to the higher ceiling (social_insurance_cap_rvalv 101,400), modeled since the 2026-07-14 tax audit." }, "tax.sales_tax_rate": { "source": "UStG §12 / Bundesregierung", "last_updated": "2026-05-16", "confidence": "high", "note": "VAT (Umsatzsteuer) standard rate 19% in 2026; reduced 7% for food/books. Gastronomy food lowered to 7% from 1 Jan 2026" }, "tax.property_tax_rate_pct": { "source": "Berlin Senate Finance: Grundsteuer ab 01.01.2025 https://www.berlin.de/grundsteuer/ablauf/artikel.1419206.php", "last_updated": "2026-05-16", "confidence": "medium", "note": "Berlin post-2025 reform: Hebesatz 470% × Steuermesszahl 0.31‰ for residential = ~0.146% of Grundsteuerwert. Effective on market value ~0.10-0.20%; midpoint 0.15% used" }, "housing.median_rent_by_bedroom": { "source": "IBB Wohnungsmarktbericht 2025, Angebotsmieten netto kalt, VALUE Marktdatenbank / RegioKontext, n=40,014 listings https://www.ibb.de/media/dokumente/publikationen/berliner-wohnungsmarkt/wohnungsmarktbericht/2025/ibb-wohnungsmarktbericht_2025_bf.pdf ; district table https://www.ibb.de/media/dokumente/publikationen/berliner-wohnungsmarkt/wohnungsmarktbericht/2025/ibb-wohnungsmarktbericht-angebotsmieten_2012-2025.xlsx ; by-room and by-size gradient, 2026 cut of the same VALUE database https://mietspiegeltabelle.de/mietspiegel-berlin/ ; dwelling sizes Amt fuer Statistik Berlin-Brandenburg https://www.statistik-berlin-brandenburg.de/f-i-2-4j/", "last_updated": "2026-08-07", "confidence": "high", "note": "CORRECTED 2026-08-01 from 880/1231/1650/2500, which priced every bedroom count above the market and grew worse with each bedroom. IBB puts the 2025 citywide median asking rent at 15.78 EUR/sqm netto kalt (mean 16.25; inner city 19.22, outer 13.01; dearest district Mitte 20.00; building age from 2023 also 20.00); the 2026 cut of the same VALUE database reproduces it at 15.68. The old 1BR of 1231 implied 78 sqm at that median, larger than the average Berlin RENTAL dwelling of 68.2 sqm, and 22.38 EUR/sqm at a normal 55 sqm, above every district median in the city. The small-flat rescue is refuted quantitatively as well as in IBB p.63 verbatim: the per-sqm premium sits under 40 sqm and at 90 sqm and over, and a 50-60 sqm flat prices at 15.20 EUR/sqm, BELOW the citywide average. Likely root cause of the old figures: furnished short-let stock, which IBB prices at 24.14 EUR/sqm All-In, a basis that also double-counts the utilities we carry separately. Rebuilt as sqm x room-specific EUR/sqm on the German Zimmer convention, our 1BR being a 2-Zimmer-Wohnung: 42/60/85/105 sqm at 18.06/16.17/16.00/16.67 EUR/sqm. Those rates are a 16.6 base times the measured room multipliers (1-Zimmer 1.088, 2-Zimmer 0.974, 3-Zimmer 0.964, 4-Zimmer 1.004). The 16.6 base is 5.2% above the citywide median and 2% above its mean, because an arriving professional does not shortlist Marzahn-Hellersdorf at 11.56 and cannot reach the 18% of listings held by municipal and cooperative landlords at 8.00-8.59. Conservative per the value-bias rule at every step: the sizes are the TOP of each plausibility band, so treat these as ceilings that are safe to ship rather than central estimates; a neutral citywide reading would give 652/876/1186/1506. Rents are netto kalt; utilities, insurance and internet are separate fields. Basis is Angebotsmiete (new lease), never the Mietspiegel stock rent of about 7.10-7.21, which asking rent exceeds by 118.9% on IBB own arithmetic. Working: docs/superpowers/research/2026-08-01-berlin-rent-ibb-verification.md and 2026-08-01-berlin-rent-and-transit-household-scaling.md SIZE LADDER CORRECTED 2026-08-07 (small-fixes wave 1, coordinated eight-city pass incl. Berlin): the common dwelling-size ladder moves from the inherited 42/60/85/105 sqm to 37/58/83/111 sqm, Duesseldorf's PUBLISHED offered-flat sizes (Wohnungsmarktbericht 2025 Abb. 54, MFH column, n=10,370 listings; the only published per-Zimmer size table in the German set, and the conservative top of the seven-city evidence range at 3-Zi 83 and 4-Zi 111). Root cause of the old ladder: German official statistics count a closed kitchen over 6 sqm as a room, so census room distributions ran one room high against listing Zimmer counts; the 42 sqm studio was priced at the 1-Zimmer premium rate measured on 33-37 sqm stock (double-counting the small-flat premium) and the 4-Zimmer was undersized at 105 vs the real ~111. Values 760/970/1360/1750 -> 670/940/1330/1850. Per-room rates and the 1.0587 uplift are UNCHANGED; sizes remain COMMON across all eight German cities because the product prices an equivalent standard of living. Working: docs/superpowers/research/2026-08-07-small-fixes.md" }, "housing.rent_percentiles": { "source": "CBRE / Berlin Hyp Wohnmarktreport Berlin 2026 (https://www.berlinhyp.de/downloads/publikationen/publikationen/2026/cbre-berlin-hyp-wohnmarktreport-berlin-2026_ammwt7iv94_m.pdf, 'Mietpreisspanne Neuvermietungsangebote 2025' table, read 2026-08-17), Berlin gesamt, n=40,035 listings of 2025, asking NET-COLD EUR/m2/month (VALUE Marktdatenbank): median 15.80; middle-segment span 7.89-25.19 (endpoints ~P10/P90, cheapest+dearest 10% stripped); bottom/top decile medians 7.06 / 28.57 (~P5/P95). P25/P75 interpolated per side in normal-quantile space: P25/med = 0.694 (P10-based) to 0.719 (P5-based), applied midpoint 0.7062 -> 0.70; P75/med = 1.278 (P90) / 1.275 (P95, mutually consistent) -> 1.28.", "last_updated": "2026-08-17", "confidence": "medium", "note": "REBASED 2026-08-17 (ledger cluster housing-0817) from 0.69/1.27, which was the district ladder's min and max - a pre-smoothed extreme-of-Bezirk-averages, not a percentile of any distribution (the KNOWN LIMIT recorded 2026-08-13). This closes that open item: Berlin now carries a genuine 40k-listing distribution percentile ON OUR OWN basis (asking net cold), like NYC (ACS B25063) and London (ONS). The numeric move is tiny (0.69/1.27 -> 0.70/1.28) - the old extreme-of-averages band was accidentally close to the true quartiles, unlike Madrid's; the gain is that the field finally measures what its name claims. INTERPOLATION JUDGMENT, recorded: the published pairs are ~P10/P90 and ~P5/P95, not quartiles; upper side is near-lognormal (P90- and P95-based estimates agree, 1.278/1.275), lower side is convex in z-ln space (the cheap regulated-adjacent tail flattens: P10-based 0.694 overstates the spread, P5-based 0.719 understates it less), so the midpoint was applied and both raw estimates are named here. WEDGE: per-m2 with ALL unit sizes pooled - part of the spread is size mix, wider than a within-bedroom band, the same priced residual as the ACS metros. Basis asymmetry with Madrid (SERPAVI sitting-stock tax data) named in both entries. Band and ladder now come from different sources (ladder: IBB Bezirk per-m2, display-only); the band==ladder drift-lock test was retired for Berlin in the same pass and replaced by a pinned test. Report is annual (spring, prior-year listings). Band edge = median_rent_by_bedroom x multiplier, computed at display time. Derivation: docs/superpowers/research/2026-08-17-housing-band-fixes.md (B1); source scoping: docs/superpowers/research/2026-08-16-housing-band-scope-madrid-berlin-quartile-sources.md" }, "housing.housing_districts": { "source": "VALUE Marktdatenbank, computed by RegioKontext GmbH for IBB Wohnungsmarktbericht 2025 (Tabellenband Tab. 25), asking cold rent EUR/m2 2025 over citywide 15.78: Mitte 20.00 = 1.27, Friedrichshain-Kreuzberg 19.40 = 1.23, Charlottenburg-Wilmersdorf 19.17 = 1.21, Pankow 17.00 = 1.08, Reinickendorf 12.80 = 0.81, Spandau 12.50 = 0.79, Marzahn-Hellersdorf 11.56/15.78 = 0.7325 x a 0.936 size-at-fixed-rooms factor = 0.69. Workbook: https://www.ibb.de/media/dokumente/publikationen/berliner-wohnungsmarkt/wohnungsmarktbericht/2025/ibb-wohnungsmarktbericht-angebotsmieten_2012-2025.xlsx (sheet Bezirksdaten_2025). Size factors derived from Amt fuer Statistik Berlin-Brandenburg SB F I 1 - j/23 Tab. 2 and 4: https://download.statistik-berlin-brandenburg.de/f918b03661b6743a/307da9a436ed/SB_F01-01-00_2023j01_BE.pdf", "last_updated": "2026-08-13", "confidence": "medium", "note": "MIXED BASIS ON PURPOSE, and the reason is the point. A per-m2 ratio only transfers to a fixed room count if a k-room flat is the same size in every Bezirk, and in Berlin it is not: Marzahn-Hellersdorf's Plattenbau rooms are small (dwelling-weighted OLS of m2 on rooms across the 12 Bezirke puts its 3-Zimmer flats about 6% under the Berlin norm for their room count), so its 0.73 was too HIGH. Corrected 2026-08-13 to 0.69, the IBB-anchored end of a 0.675-0.686 range; the CBRE/Berlin Hyp offer-weighted anchor gives 0.675 and IBB is preferred because the rest of this field already derives from it, and 0.69 is the less favourable direction for the city per the conservative-value rule. The other six are DELIBERATELY LEFT on the raw per-m2 basis. Applying the same size term uniformly would give Charlottenburg-Wilmersdorf 1.31 and Mitte 1.29, which makes C-W Berlin's most expensive Bezirk on a 1.5% margin, well inside the uncertainty of a MEDIUM-confidence derived input, with nothing published supporting the reordering: no by-room district rent source exists for Berlin at all (IBB's Angebotsmieten workbook 2012-2025, the Tabellenband, the full IBB main report where the string 'Zimmer' appears zero times in 8,609 lines, CBRE/Berlin Hyp Wohnmarktreport 2026, Engel & Voelkers, mietspiegeltabelle.de, immoportal.com, miet-check.de and the Amt fuer Statistik were all searched). The spot-check adjudicated the two ladder ends only and named C-W as the next to examine. OPEN ITEM: if a published by-room or by-size Bezirk source appears, apply the size term to all seven at once (C-W 1.21 to ~1.31, Mitte 1.27 to ~1.29). ALSO CORRECTED 2026-08-13: this field credited ImmoScout24; IBB states its own upstream as the VALUE Marktdatenbank, a multi-portal aggregator, computed by RegioKontext. Citation fix, not a value fix. Still 7 of 12 Bezirke; Neukolln, Tempelhof-Schoneberg, Steglitz-Zehlendorf, Lichtenberg and Treptow-Kopenick lacked corroborated asking figures and are left out. Multipliers are on the asking basis; label any rendered figure as asking (Mietspiegel sits >2x lower). Exposure is largest at the STUDIO end, where the citywide sub-40 m2 per-m2 premium is very large and the peripheral market has almost none; flagged, not measured. Working: docs/superpowers/research/2026-08-13-madrid-berlin-ladder-rebase.md; diagnosis: docs/superpowers/research/2026-08-13-berlin-ladder-spotcheck.md" }, "housing.median_home_price_by_bedroom": { "source": "Guthmann Berlin Market Report 2025 / ImmoScout24 WohnBarometer https://guthmann.estate/en/market-report/berlin/", "last_updated": "2026-05-16", "confidence": "medium", "note": "Existing-stock avg ~€5,130/m² in 2025; derived by-bedroom using typical sizes. New-build is ~€8,220/m² but existing-stock is the median" }, "housing.utilities_monthly": { "source": "Berlin Senate Betriebskostenuebersicht 2026 (Abrechnungsjahr 2024): 1,93 EUR/sqm kalt + 1,50 Heizung = 3,43 EUR/sqm, https://www.berlin.de/sen/wohnen/service/betriebskostenuebersicht/", "confidence": "high", "note": "2BR reference (83 sqm per the German dwelling ladder): 3,43 x 83 + 63 electricity = 348, rounded to 350. Numbeo 85 sqm reads 344.88, agreeing to 0.9%. REPLACES the old 250, whose own note read \"2BR ~75 sqm midpoint\": that midpointing onto 75 sqm was superseded when the ladder moved Berlin 2BR to 83 sqm on 2026-08-07 and was never undone.", "last_updated": "2026-08-08" }, "housing.renters_insurance_monthly": { "source": "Settle in Berlin: Hausratversicherung", "last_updated": "2026-05-16", "confidence": "high", "note": "Berlin Hausratversicherung €5-10/mo; midpoint €8" }, "housing.internet_phone_monthly": { "source": "Settle in Germany: Broadband + Mobile", "last_updated": "2026-05-16", "confidence": "high", "note": "Broadband €30-40 (Telekom/Vodafone) + mid-tier mobile €10-20 ≈ €50 bundle" }, "healthcare.premium_monthly_by_family_size": { "source": "GKV-Spitzenverband https://www.gkv-spitzenverband.de/english/statutory_health_insurance/statutory_health_insurance.jsp", "last_updated": "2026-05-16", "confidence": "high", "note": "GKV premium is folded into our tax model's social_insurance_rate to avoid double-counting. Standalone premium line = 0" }, "healthcare.annual_out_of_pocket_avg": { "source": "Destatis Gesundheitsausgabenrechnung 2024 (private households + NPISH EUR 65bn) + Destatis EVS 2023 Konsumausgaben nach Haushaltstyp + GKV Zuzahlungen aggregate EUR 5.18bn (BMG via Aerzteblatt) + SGB V paras 61-62", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "healthcare.dental_vision_monthly": { "source": "Verbraucherzentrale: Zahnzusatzversicherung", "last_updated": "2026-05-16", "confidence": "medium", "note": "Supplementary dental insurance €15-40/mo for 30s; vision usually paid OOP (GKV covers basic eye exam). Median €25" }, "childcare.infant_monthly": { "source": "Berlin Senate, Kostenbeteiligung https://www.berlin.de/sen/jugend/familie-und-kinder/kindertagesbetreuung/kostenbeteiligung/ (fetched 2026-09-12); resident-reported all-in from the 2026-07-22 correction threads", "last_updated": "2026-09-12", "confidence": "high", "note": "FREE-TIER RESIDUAL, not tuition: Berlin Kita has charged no fee since August 2018 for every age from 1 (\"In Berlin ist die Kindertagesbetreuung seit August 2018 kostenfrei\"). The Senate's only mandatory charge is the lunch share, \"Verpflegungsanteil von 23 Euro im Monat\"; Zusatzangebote a Kita may charge for are optional and capped at EUR 100 a month per child from January 2025. The stored 60 is the all-in a Berlin household reports paying (23 lunch plus about 10 Vespergeld plus extras), inside the statutory range 23 to 123: Rule 1 fixes the floor at 23 and Rule 5 keeps the higher all-in (ledger 2026-09-12). Infants under 1 are rarely in Kita (Elternzeit and Elterngeld phase); private Kitas charge EUR 300 to 800 but about 95% use public." }, "childcare.preschool_monthly": { "source": "Berlin Senate, Kostenbeteiligung https://www.berlin.de/sen/jugend/familie-und-kinder/kindertagesbetreuung/kostenbeteiligung/ (fetched 2026-09-12); resident-reported all-in from the 2026-07-22 correction threads", "last_updated": "2026-09-12", "confidence": "high", "note": "FREE-TIER RESIDUAL, not tuition: Kita has charged no fee since August 2018 (Senate). BASIS decided 2026-09-12: the stored 60 is lunch PLUS extras, not lunch only. The Senate's mandatory share is EUR 23 a month for lunch (\"Wenn Ihr Kind ein Mittagessen erhaelt, beteiligen Sie sich als Eltern an den Kosten mit einem Verpflegungsanteil von 23 Euro im Monat\"); Zusatzangebote are optional and capped at EUR 100 a month per child from January 2025 (\"Obergrenze von 100 Euro im Monat pro Kind\"). The 60 is the resident-reported all-in (23 lunch, about 10 Vespergeld, the rest extras) inside the 23 to 123 statutory range. Rule 1 pins the floor at 23; Rule 5 keeps the higher all-in rather than cutting to the floor. Ledger 2026-09-12; research docs/superpowers/research/2026-09-12-child-benefit-audit-fixes.md F2." }, "childcare.after_school_monthly": { "source": "Berlin Senate, Erlaeuterungen zur Festsetzung der Kostenbeteiligung eFoeB, August 2023 https://www.berlin.de/sen/bjf/service/formulare/erlaeuterungen-kostenfestsetzung-efoeb.pdf ; Familienportal Berlin, Ganztagsbetreuung an Grundschulen https://familienportal.berlin.de/artikel/ganztagsbetreuung-fuer-berliner-grundschulkinder ; TKBG Anlage 2 (Gesamtausgabe gueltig ab 01.08.2022) https://www.berlin.de/jugendamt-pankow/_assets/fachdienst-5/dokumente/tagesbetreuungskostenbeteiligungsgesetz-ab-08_2022.pdf (all fetched 2026-09-12; replaces the 2019 press release pressemitteilung.836474.php, which 404s)", "last_updated": "2026-09-12", "confidence": "high", "note": "Set to 0 on 2026-09-12, was 40. STATUTORY for grades 1 to 3, which is every published household's school-age child (ages 6 to 8): the Hort (ergaenzende Foerderung und Betreuung, eFoeB) charges no Kostenbeteiligung through grade 3 and school lunch is free through grade 6 (Senate explainer: \"Fuer Kinder der Jahrgangsstufen 1 bis 3 wird keine Kostenbeteiligung erhoben\" and \"Eine Kostenbeteiligung fuer das Mittagessen wird nicht erhoben\"; Familienportal: \"Das Mittagessen ist fuer alle Berliner Kinder von der ersten bis zur sechsten Klasse kostenfrei\"). The old 40 rested on a \"food fee ~37/mo\" from a 2019 press release that is now a dead link; that fee does not exist. KNOWN GAP, favourable to Berlin for ages 9 to 12 only: from grade 4 the fee is income-scaled under TKBG Anlage 2, and the top band (income from EUR 81,060 a year, which every ICP household clears) pays EUR 134 a month for the 13:30 to 16:00 module and 194 to 18:00, at 80% with two children. The engine bills one after_school_monthly across ages 6 to 12 with no grade boundary, so a 9-to-12-year-old is under-billed by roughly EUR 1,300 to 1,600 a year until a per-city paid-from age exists (the free_tier_start_age pattern). Ledger 2026-09-12." }, "childcare.subsidy_per_kid_annual": { "source": "Land and municipal Kitabeitrag schedules; double-subtraction correction 2026-08-02", "last_updated": "2026-08-02", "confidence": "high", "note": "Set to 0 on 2026-08-02, was EUR 9,600. The old value was the VALUE OF THE FREE OR SUBSIDISED KITA HOURS, but childcare.preschool_monthly already carries the post-free-tier out-of-pocket, so computeChildcareCost subtracted the same relief a second time and clamped the Berlin household bill to zero. German Kita relief is a benefit in kind delivered as a lower or absent fee, never a payment to the parent, so nothing belongs in this field. Berlin now bills EUR 60/month per child of Verpflegung, roughly EUR 23 of meals plus EUR 10 Vespergeld, confirmed by a Berlin resident in the 2026-07-22 correction threads. This field holds CASH THAT REACHES THE HOUSEHOLD. Where the monthly rate it nets against is already the price the household actually pays, the value is 0 and the relief is described here instead. See docs/superpowers/research/2026-08-02-childcare-double-subtraction.md." }, "food.groceries_monthly_per_adult": { "source": "Numbeo Berlin Markets", "last_updated": "2026-05-16", "confidence": "high", "note": "Hard-discounter market (Aldi/Lidl/Penny) keeps grocery costs low. Milk €1.10/L, eggs €3/dozen, bread €2, chicken €10/kg" }, "food.groceries_monthly_per_kid": { "source": "Numbeo Berlin + Destatis EVS", "last_updated": "2026-05-16", "confidence": "medium", "note": "~54% of adult figure" }, "food.dining_out_monthly_baseline": { "source": "Numbeo Berlin Restaurants", "last_updated": "2026-05-16", "confidence": "high", "note": "Mid-range dinner €15-25/pp; inexpensive ~€12. Notably cheaper than Munich/Frankfurt" }, "transit.monthly_pass": { "source": "https://www.bundesregierung.de/breg-de/aktuelles/deutschlandticket-2134074", "last_updated": "2026-08-05", "confidence": "high", "note": "Deutschlandticket EUR 63/mo since 2026-01-01 (VMK decision 2025-09-18). Was 58 through Dec 2025." }, "transit.annual_car_cost_avg": { "source": "ADAC Autokosten", "last_updated": "2026-05-16", "confidence": "medium", "note": "ADAC mid-size all-in ~€450/mo = €5,400/yr. Berlin parking moderate; permit ~€10/yr" }, "discretionary.entertainment_monthly_baseline": { "source": "Numbeo Berlin", "last_updated": "2026-05-16", "confidence": "medium", "note": "Gym ~€30 + 2 cinema ~€26 + streaming ~€20 + outings ~€54. Lower end of European capitals" }, "discretionary.local_vacation_annual_baseline": { "source": "Destatis EVS + regional weekend cost", "last_updated": "2026-05-16", "confidence": "medium", "note": "3 long weekends (Baltic/Saxon Switzerland/Brandenburg) @ €400 + one €600 short trip. German domestic travel relatively affordable" }, "travel_home.avg_round_trip_per_person_to_other_city": { "source": "Google Flights / Skyscanner BER-SEA economy survey", "last_updated": "2026-05-16", "confidence": "medium", "note": "No direct BER-SEA; one-stop via FRA/AMS/LHR. Mid-range RT €800-900 booked 4-8 weeks out. Peak summer >€1100" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Familienkasse https://www.arbeitsagentur.de/familie-und-kinder/kindergeld-anspruch-hoehe-dauer", "last_updated": "2026-05-16", "confidence": "high", "note": "Kindergeld 2026: EUR 259/mo per child x 12 = EUR 3,108, which is the stored value. Universal regardless of income, NOT means-tested. Corrected 2026-08-02: this note previously computed \"EUR 255/mo per child x 12 = EUR 3,060\", which is the 2025 rate and disagreed with the stored 3,108. The Steuerfortentwicklungsgesetz stepped Kindergeld to EUR 255 on 2025-01-01 and EUR 259 on 2026-01-01, so 3,108 is the correct 2026 figure and the arithmetic in the note was what was wrong. The stored value was NOT changed. Corroborated across the repo: /blog/kita-cost-germany/, /blog/cost-of-raising-kids/, /countries/germany/ and /compare-job-offers/ all carry EUR 3,108 / EUR 259." }, "non_cash.vacation_days_statutory": { "source": "BMAS / Bundesurlaubsgesetz https://www.bmas.de/EN/Labour/Working-Hours-and-Leave/working-hours-and-leave.html", "last_updated": "2026-05-16", "confidence": "high", "note": "Statutory minimum 20 days; typical collectively-bargained 25-30; midpoint 28" }, "non_cash.public_holidays": { "source": "Berlin Senate https://www.berlin.de/sen/finanzen/service/feiertage/", "last_updated": "2026-05-16", "confidence": "high", "note": "10 statutory: 9 federal + International Women's Day (March 8) Berlin-specific" }, "non_cash.parental_leave_weeks": { "source": "MuSchG para 3 (Schutzfristen, 6 weeks before and 8 after) https://www.gesetze-im-internet.de/muschg_2018/__3.html ; BEEG para 4 Abs. 3 (12 Monatsbetraege Basiselterngeld plus 2 Partnermonate, gemeinsam) and para 4 Abs. 4 Satz 3 https://www.gesetze-im-internet.de/beeg/__4.html", "last_updated": "2026-08-01", "confidence": "high", "note": "67 wk = 6 additive pre-birth Mutterschutz weeks + the 14-month Elterngeld pot (60.87 wk). Was 61, the Elterngeld pot alone with Mutterschutz omitted. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. OVERLAP RULE APPLIED: the 8 post-birth Mutterschutz weeks are NOT stacked on top, because BEEG para 4 Abs. 4 Satz 3 deems Lebensmonate carrying Mutterschaftsgeld to be consumed Basiselterngeld months (gelten als Monate, fuer die dieser Elternteil Basiselterngeld bezieht). They are counted once, at the higher rate. Germany therefore reaches 67 and not 75. Elterngeld is a SHARED household pot, so it is not doubled. GERMANY HAS NO STATUTORY PAID PATERNITY LEAVE: the Familienstartzeit bill transposing EU Directive 2019/1158 art. 4 was still not law on 2026-08-01, so zero paternity weeks. Excluded: ElterngeldPlus (an election that halves the monthly amount to double the months) and the Partnerschaftsbonus (conditional on both parents working reduced hours)." }, "non_cash.parental_leave_paid_pct": { "source": "SGB V para 24i (Mutterschaftsgeld) https://www.gesetze-im-internet.de/sgb_5/__24i.html ; MuSchG para 20 (Zuschuss zum Mutterschaftsgeld, employer tops up to full net) https://www.gesetze-im-internet.de/muschg_2018/__20.html", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the percentage-paid remainder of the 67 weeks: the 14.1 Mutterschutz weeks (6 pre-birth + 8 post-birth), where para 24i SGB V pays EUR 13/day and para 20 MuSchG obliges the employer to top up to full net pay with no ceiling. Was 72, a blend that averaged the capped Elterngeld slice into this field; since 2026-08-06 the 52.87 Elterngeld weeks live in non_cash.parental_leave_capped (65% up to EUR 1,800/month, nothing above EUR 175,000 household income) and parentalLeaveEffectiveRate() blends the slices at the user's salary, so every reader-facing surface already prices the cap and the cliff." }, "non_cash.parental_leave_capped": { "source": "BEEG para 2 (Elterngeld: 65% of prior income, minimum EUR 300 and maximum EUR 1,800 per month) https://www.gesetze-im-internet.de/beeg/__2.html ; BEEG para 1 Abs. 8 (no Elterngeld above EUR 175,000 combined zu versteuerndes Einkommen) https://www.gesetze-im-internet.de/beeg/__1.html ; MuSchG para 20 (the 14 Mutterschutz weeks stay OUTSIDE this block: the employer net top-up has no ceiling) https://www.gesetze-im-internet.de/muschg_2018/__20.html", "last_updated": "2026-08-06", "confidence": "high", "note": "52.87 of the 67 weeks: the Elterngeld pot (60.87 wk) minus the 8 overlap weeks counted at the Mutterschutz rate. The EUR 1,800/month cap (unindexed since Elterngeld's 2007 introduction, re-confirmed 2026-08-06) binds above ~EUR 33,000 gross, i.e. at every ICP salary: at EUR 90,000 this slice pays ~24% of gross and the blended Berlin rate is ~40%; at EUR 150,000 the slice pays ~14% and the blend is ~32%, versus the 72 this field carried before the 2026-08-06 caps rebuild. BEEG para 1 Abs. 8 zeroes the slice for households above EUR 175,000: the engine compares HOUSEHOLD GROSS against the ceiling as a conservative proxy for zu versteuerndes Einkommen (gross exceeds taxable income, so the cliff bites earlier, the direction less favorable to Germany). Same conservative direction on the rate basis: the statute pays 65% of Elterngeld-Netto while the engine applies 65% to gross, and above ~EUR 33,000 gross the cap binds under both readings, so the distinction is immaterial at any ICP salary. Floor EUR 300/month (BEEG para 2). Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Germany section) and 2026-08-06-caps-engine-build.md." }, "non_cash.universal_healthcare": { "source": "BMG https://www.bundesgesundheitsministerium.de/en/topics/health-insurance/the-german-healthcare-system.html", "last_updated": "2026-05-16", "confidence": "high", "note": "Mandatory universal via GKV (~88% pop) or PKV. Effectively universal" }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, urb_ctran indicator TT1019V, city DE001C. https://ec.europa.eu/eurostat/databrowser/view/urb_ctran/default/table?lang=en", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average time of journey to work (minutes)", "year": "2016", "note": "32.5 min, BASIS YEAR 2016. GEOGRAPHY MISMATCH, stated because it has a direction: Eurostat cuts this indicator at the CITY, which is NARROWER than the metro our unit declares, so the figure omits longer inbound suburban commutes and is biased LOW. Low flatters the destination, which is the opposite of our usual conservative lean, and it is accepted only because no German metro commute series exists. Survey self-report. Rounding is standard half-up, as on all eight German rows. Replaces a Numbeo citation. METRO-LEVEL GERMAN SERIES: ESTABLISHED HARD NEGATIVE, do not re-run. MiD 2023 Tabelle 6 carries RegioStaR7 breakdowns of population, trips, distance and total time but NO duration by purpose; the national banner table W9 Wegedauer is all-purpose and sits on an \"Ohne regelmaessige berufliche Wege\" base; the small-area products regionalise headline indicators only; and Eurostat publishes no functional-urban-area rows for TT1019V. SrV 2023 was rejected because it publishes an ALL-PURPOSE trip duration, and swapping an all-purpose duration for a work-purpose one is a UNIT error, which is worse than the geography error we accepted. SrV is therefore used only as rank-order corroboration, and it agrees with Eurostat against our old stored spread: both put Frankfurt near the BOTTOM of the big five, where we had it second highest." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Berlin: EUR 91,048, held in labor_market.tech_salary_distribution (source_slug berlin-metropolitan-region, as_of 2026-07-18, table confidence usable). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 91,048 / 0.867 = USD 105,015; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.4201, R5-rounded down to 0.42. REPLACES 0.6 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Berlin by 43%. The measured p50 was 0.700 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE Berlin all-in median ~€115k (base ~€85k + bonus/equity). Naive ratio vs Seattle ($220k) ~0.56. Adjusted to 0.60 reflecting FAANG Berlin (Amazon/Google) higher band. Below London (0.75) and Amsterdam (0.72) — Berlin lacks deep FAANG; local scale-ups (Zalando/Delivery Hero/N26) anchor lower. (London reference corrected 0.85 -> 0.75 on 2026-08-11: the repo has always carried 0.75 and this note cited a value that does not exist.)" }, "tax.solidaritaetszuschlag_threshold_married": { "source": "Soli Freigrenze 2026 joint assessment EUR 40,700 (2 x 20,350 single; tk.de Solidaritaetszuschlag); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.social_insurance_rate_rvalv": { "source": "RV 9.3% + ALV 1.3% employee share = 10.6%, runs to the higher Beitragsbemessungsgrenze (lohn-info.de Sozialversicherungsbeitraege 2026); verified tax audit 2026-07-14", "confidence": "high", "note": "Carved out of the combined 21.15% employee share; the KV/PV remainder (10.55%) stops at the lower ceiling.", "last_updated": "2026-07-17" }, "tax.social_insurance_cap_rvalv": { "source": "Beitragsbemessungsgrenze RV/ALV 2026 = EUR 101,400 (bundesregierung.de/breg-de/aktuelles/beitragsgemessungsgrenzen-2386514); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.pv_childless_surcharge_rate": { "source": "PV Kinderlosenzuschlag § 55 Abs. 3 SGB XI: +0.6pp employee-borne for childless members 23+ (dak.de Beitrag zur Pflegeversicherung Zuschlag Kinderlose; vdek.com Aenderungen Gesundheitswesen 2026); found by the 2026-08-06 corrections wave", "confidence": "high", "note": "Applied when the household has no children; the under-23 exemption is not modeled (ICP is older). Deductible with the rest of the Basis-Pflegeversicherung.", "last_updated": "2026-08-06" }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/berlin-metropolitan-region", "last_updated": "2026-07-18", "confidence": "usable", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "EF English Proficiency Index 2024 (High band)", "last_updated": "2024", "note": "National-level English usability for tech work. Band: high." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Destatis Gesundheitsausgabenrechnung 2024 (private households + NPISH EUR 65bn) + Destatis EVS 2023 Konsumausgaben nach Haushaltstyp + GKV Zuzahlungen aggregate EUR 5.18bn (BMG via Aerzteblatt) + SGB V paras 61-62", "last_updated": "2026-08-08", "confidence": "medium", "note": "Level essentially CONFIRMED at EUR 311 per adult (778/capita gross, ~40% in scope after netting long-term care 46%, dental 9%, vision ~3pp). What Germany needed was a curve, not a level. The curve is steep in adults and nearly flat in children because under-18s are exempt from every Zuzahlung except transport AND each child subtracts EUR 9,756 from the income base before the 2%-of-gross Belastungsgrenze, so adding a child LOWERS the household's statutory ceiling in absolute euros. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "non_cash.sick_leave": { "source": "Entgeltfortzahlungsgesetz (EFZG) §3 (employer phase); SGB V §47 (Krankengeld; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "EStG s.10 Abs 3 https://www.gesetze-im-internet.de/estg/__10.html ; SVBezGrV 2026 s.4 Abs 1 https://www.gesetze-im-internet.de/svbezgrv_2026/BJNR1160A0025.html", "last_updated": "2026-08-11", "confidence": "high", "note": "Ruerup/Basisrente Hoechstbetrag, 2026: EUR 30,825.60 single (knappschaftliche RV BBG 124,800 x 24.7%). A DEDUCTION ceiling and a COMBINED one: employee AND employer statutory pension contributions consume the same pot. less_employer_pct 0.186 = 2 x 9.3% and employer_base_cap 101,400 is the allgemeine RV BBG, because computeGermanyTaxes already deducts the employee 9.3% via vorsorge; a naked max would hand a EUR 95k earner about EUR 8.8k of phantom deduction. Riester (EUR 2,100) and bAV Entgeltumwandlung (8% of the BBG, EUR 8,112) are deliberately NOT stacked. Married couples: the statute doubles the ceiling to 61,651.20 even on one income; our per-earner cap under-deducts there, which errs toward higher tax." } }, "zurich": { "tax.federal_brackets": { "source": "ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Exact 2026 Grundtarif (kalte-Progression vintage confirmed against the 2025 circular). 2026-08-07: restored the missing 5.94% band 76,200-82,100 (the stored table jumped 2.97% to 6.6%, overstating by up to CHF 38.94/yr). Applied to taxable income = gross minus the statutory deduction stack (see tax.payroll_deductible)." }, "tax.cantonal_brackets": { "source": "CH-2 (exact, 2026-07-26): Zürich § 35 Abs. 1 StG Grundtarif (single), einfache Staatssteuer x Gesamtsteuerfuss 214% (Kanton 95% + Stadt Zürich 119%, 2026). Replaces the prior approximate single scale that understated ~22% (wrong thresholds, missing the 11%/12% bands). ESTV Kantonsblatt Zürich (Stand Feb 2026, § 35 Abs. 1); zh.ch Steuerfuss 2026.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.social_insurance_rate": { "source": "BSV Beitragssätze 2026 https://www.bsv.admin.ch/en/contributions-overview ; Swissmem, Sozialversicherungsbeiträge 2026 (employee AHV/IV/EO 5.3%, ALV 1.1% to CHF 148,200, no solidarity line) https://www.swissmem.ch/de/wissen/personalwesen/arbeitsrecht/sozialversicherungsbeitraege-2026.html ; TaxPartner, Wegfall des Solidaritätsprozentes per 1. Januar 2023 https://taxpartner.ch/wegfall-des-solidaritaetsprozentes-an-die-schweizer-arbeitslosenversicherung-per-1-januar-2023/", "last_updated": "2026-08-06", "confidence": "high", "note": "Employee share AHV 4.35% + IV 0.7% + EO 0.25% = 5.3% (uncapped) + ALV 1.1% (capped). Combined 6.4% up to the CHF 148,200 cap. Corrected 2026-08-06: this note used to say the above-cap continuation was missing from the engine. It is not. app.js charges min(gross, cap) x 6.4% plus (gross - cap) x 5.3%, so a Zurich single filer on CHF 300,000 pays CHF 17,530.20 of social insurance, of which CHF 8,045.40 is the uncapped AHV/IV/EO slice above the cap. The 0.5% employee ALV solidarity percent this note also mentioned is gone from Swiss law rather than from our model: it was 1.0% total on wage above the cap, levied from 2011 to retire ALV debt, and the right to collect it lapsed automatically on 2023-01-01 once the ALV compensation fund's equity cleared CHF 2.5bn (SECO announcement 2022-10-13). It is absent from every 2026 rate table, so no schema field for tax.alv_solidarity_rate exists. The lapse trigger is statutory and reversible, so re-check it each year. Working: docs/superpowers/research/2026-08-06-tax-value-fix-batch.md" }, "tax.social_insurance_cap": { "source": "BSV ALV-Höchstbetrag 2026", "last_updated": "2026-05-16", "confidence": "high", "note": "ALV ceiling CHF 148,200/yr (unchanged from 2025). v1 simplification: applies the full 6.4% to this cap; understates above-cap AHV/IV/EO" }, "tax.sales_tax_rate": { "source": "ESTV MwSt 2026 https://www.estv.admin.ch/estv/de/home/mehrwertsteuer.html", "last_updated": "2026-05-16", "confidence": "high", "note": "Standard rate 8.1% (raised from 7.7% on 1 Jan 2024 for AHV 21 financing). Planned increase to 8.8% delayed to ~2028" }, "tax.property_tax_rate_pct": { "source": "Kanton Zürich Steueramt: Liegenschaftensteuer", "last_updated": "2026-05-16", "confidence": "medium", "note": "Zurich has NO separate municipal/cantonal recurring property tax. Owner-occupiers are taxed on Eigenmietwert (imputed rental value) via income tax. 0.1% placeholder is directional approximation; precise model would add ~60-70% of market rent to taxable income (out of scope)" }, "housing.median_rent_by_bedroom": { "source": "Investropa Zurich Rents 2026 (Homegate/ZKB index + City of Zurich rent survey) https://investropa.com/blogs/news/zurich-rents", "last_updated": "2026-05-16", "confidence": "high", "note": "Swiss room count = bedrooms + 1; mapped 0BR=1-room, 1BR=2.5-room, 2BR=3.5-room, 3BR=4.5-room. Among most expensive European cities" }, "housing.rent_percentiles": { "source": "Statistik Stadt Zurich Mietpreiserhebung April 2024, open-data table OD5161 https://data.stadt-zuerich.ch/dataset/bau_whg_mpe_mietpreis_raum_zizahl_gn_jahr_od5161/download/BAU516OD5161.csv (re-derived 2026-08-17): within-city qu25/qu75 of net rent per Wohnung, Ganze Stadt, Nicht gemeinnuetzig segment, 3 Zimmer: 1,492/1,875 = 0.7957 and 2,400/1,875 = 1.2800, rounded outward to 0.79/1.28 per the conservative-wider rule (n=9,132).", "last_updated": "2026-08-17", "confidence": "high", "note": "REBASED 2026-08-17 (ledger cluster housing-0817) from the 0.83/1.33 cross-Kreis min/max-of-medians band, which failed the constant-ratio check on its own source (Kreis1/Kreis12 median ratio 1.602 at 2 Zi, 1.841 at 3 Zi, 2.006 at 4 Zi vs the stored constant 1.602) AND was mislabeled: its inputs 2,102/1,312/1,578 are the 2-ZIMMER Nicht-gemeinnuetzig cells, not 3-Zimmer as previously documented (the citywide 2-Zi market median 1,578 coincidentally equals the 3-Zi all-apartments press-release median). The new pair is the quartile the field name claims, from the SAME official source, and is near-constant across sizes (2/3/4 Zi: 0.824/1.262, 0.796/1.280, 0.782/1.282), which kills the constant-ratio defect by construction; 3 Zimmer (= 2BR, Swiss Zimmer counts the living room) is the ICP anchor matching the flagship 2BR convention. Deliberately NOT widened to the 3-Zi cross-Kreis stock ratio 1.84: on the ASKING basis (the stored medians' basis) the cross-Kreis spread is only ~1.52-1.55 (Neho/RealAdvisor 2026), so the stock ratio would overshoot. Quartiles are of STOCK rents (Bestandsmieten) while stored medians are asking; stock quartiles read wider (tenure-discount left tail), the conservative direction for a band. Non-profit segment excluded: a newcomer cannot access co-op/municipal stock. MPE is biennial; next survey April 2026, publication ~Q4 2026. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Derivation: docs/superpowers/research/2026-08-17-housing-band-fixes.md (Z1); scoping evidence: docs/superpowers/research/2026-08-16-housing-band-scope-zurich.md" }, "housing.median_home_price_by_bedroom": { "source": "Homegate / ImmoScout24 / Neho Zurich 2026 (~CHF 17,000/m²) https://www.homegate.ch/en/property-prices-m2/city-zurich", "last_updated": "2026-05-16", "confidence": "medium", "note": "Estimated from ~CHF 17,000/m² average applied to typical sizes per bedroom. Wide dispersion CHF 13,400–20,300/m² by district" }, "housing.utilities_monthly": { "source": "Numbeo Zurich May 2026 + Expatistan", "last_updated": "2026-05-16", "confidence": "high", "note": "~CHF 227/mo for 85m² (electricity, heat, water, garbage)" }, "housing.renters_insurance_monthly": { "source": "Comparis / moneyland.ch Hausratversicherung", "last_updated": "2026-05-16", "confidence": "high", "note": "Annual CHF 150-400 for combined Hausrat + Privathaftpflicht; midpoint ~CHF 260/yr = €22/mo" }, "housing.internet_phone_monthly": { "source": "The Poor Swiss / internet-offer.ch / Salt bundles 2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Home fiber CHF 50-60 + mobile CHF 25-40; bundles via Salt/Swisscom land ~CHF 80-100" }, "healthcare.premium_monthly_by_family_size": { "source": "Expat Savvy / BAG 2026 Zurich canton https://expat-savvy.ch/blog/decoding-swiss-health-insurance-canton-region-premium/", "last_updated": "2026-05-16", "confidence": "high", "note": "Zurich canton adult avg ~CHF 480/mo (national avg CHF 465, ZH above avg). KVG mandatory basic insurance, CHF 300 standard franchise. Family sizes assume 2 adults + reduced child premiums (~CHF 120/mo)" }, "healthcare.premium_age_multiplier": { "source": "Expat Savvy / BAG age tier structure", "last_updated": "2026-05-16", "confidence": "high", "note": "KVG has 3 age tiers: child <19, young adult 19-25 (~0.70× adult), adult 26+. The 18-24 bracket is mostly young-adult priced" }, "healthcare.annual_out_of_pocket_avg": { "source": "BAG Kostenbeteiligung https://www.bag.admin.ch/de/krankenversicherung-kostenbeteiligung-fuer-in-der-schweiz-wohnhafte-versicherte + BAG OKP statutory cost-sharing aggregate CHF 5.7bn (2024) over 8.96m insured, adult/child split from BAG age-group net benefits", "last_updated": "2026-08-08", "confidence": "high", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "healthcare.dental_vision_monthly": { "source": "Moneyland / Comparis dental supplementary", "last_updated": "2026-05-16", "confidence": "high", "note": "Adult dental insurance CHF 25-90/mo; ~CHF 35 mid-range (75% reimbursement up to CHF 1-3k cap)" }, "childcare.infant_monthly": { "source": "Stadt Zürich Kita fees https://www.stadt-zuerich.ch/de/bildung-und-kinderbetreuung/kinderbetreuung/kinderbetreuung-finden/kita.html", "last_updated": "2026-05-16", "confidence": "medium", "note": "Full-time private Kita CHF 2,600-3,800/mo (~CHF 130-180/day × 21 days). Used midpoint ~CHF 2,900 unsubsidized. Betreuungsgutscheine (income-tested vouchers) reduce this substantially" }, "childcare.preschool_monthly": { "source": "Stadt Zürich Kita fees / kibesuisse", "last_updated": "2026-05-16", "confidence": "medium", "note": "Preschool (2-4) typical unsubsidized CHF 2,400-3,200/mo. Kindergarten (4-6) is part of free public school" }, "childcare.after_school_monthly": { "source": "Stadt Zürich Tagesschule/Hort https://www.stadt-zuerich.ch/de/bildung-und-kinderbetreuung/schule/tagesschule.html", "last_updated": "2026-05-16", "confidence": "medium", "note": "Income-tested; mid-income family CHF 500-900/mo for lunch + after-school. Max unsubsidized ~CHF 1,200+" }, "childcare.subsidy_per_kid_annual": { "source": "Stadt Zürich Betreuungsgutscheine", "last_updated": "2026-05-16", "confidence": "high", "note": "Zurich administers subsidies as price reductions on the gross fee, not annual cash subsidy. Set to 0 per v1 scope; gross fees above already represent unsubsidized cost. Tech-worker household will see significant means-tested reduction in practice" }, "food.groceries_monthly_per_adult": { "source": "Rappn 2026 Swiss grocery budget + Numbeo Zurich basket https://rappn.ch/en/blog/real-cost-of-groceries-in-switzerland-honest-budgets-2026", "last_updated": "2026-05-16", "confidence": "high", "note": "Sources converge CHF 400-600/mo; midpoint CHF 500. Zurich most expensive Swiss city; meat ~2-2.5× German prices; cross-border shopping in DE common" }, "food.groceries_monthly_per_kid": { "source": "Rappn 2026 + per-adult scaling", "last_updated": "2026-05-16", "confidence": "medium", "note": "~55% of adult figure" }, "food.dining_out_monthly_baseline": { "source": "Numbeo Zurich Restaurants + Bellevue-au-Lac 2025 guide", "last_updated": "2026-05-16", "confidence": "medium", "note": "Mid-range CHF 30-50/pp; inexpensive CHF 28; pattern ~6 outings/mo = ~CHF 240" }, "transit.monthly_pass": { "source": "ZVV https://www.zvv.ch/zvv/en/travelcards-and-tickets/travelcards/monthly-and-annual-travelcards.html", "last_updated": "2026-05-16", "confidence": "high", "note": "Zone 110 (Stadt Zurich) monthly travelcard, 2nd class. 2-zone passes ~CHF 156 common for commuters" }, "transit.annual_car_cost_avg": { "source": "TCS Kilometerkosten https://www.tcs.ch/en/test-advice/advice/buying-vehicle/operating-costs.php", "last_updated": "2026-05-16", "confidence": "medium", "note": "Mid-size all-in ~CHF 9,000-12,000/yr + Zurich parking permit CHF 300-500/mo pushes urban total to ~CHF 11-13k" }, "discretionary.entertainment_monthly_baseline": { "source": "Numbeo Zurich Sports & Leisure", "last_updated": "2026-05-16", "confidence": "medium", "note": "Gym ~CHF 110/mo + 2 cinema CHF 40 + streaming CHF 30 + occasional outings CHF 120. Among most expensive globally" }, "discretionary.local_vacation_annual_baseline": { "source": "MySwitzerland + Numbeo cross-check", "last_updated": "2026-05-16", "confidence": "medium", "note": "3 long weekends Swiss Alps/lakes (CHF 700-900 each) + 1 short regional trip. CH domestic travel expensive" }, "travel_home.avg_round_trip_per_person_to_other_city": { "source": "SWISS / Kayak / Expedia ZRH-SEA composite", "last_updated": "2026-05-16", "confidence": "medium", "note": "ZRH-SEA no nonstop; one-stop or partner. CHF 800 mid-range 4-8 weeks advance; cheap CHF 440-620, peak summer higher" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Kanton Zürich Familienzulagen https://www.zh.ch/de/soziales/sozialversicherungen/familienzulagen.html", "last_updated": "2026-05-16", "confidence": "high", "note": "ZH Kinderzulage CHF 200/mo (0-15) + Ausbildungszulage CHF 250/mo (16-25). Blended ~CHF 225/mo = CHF 2,700/yr Swiss Kinderzulagen / allocations familiales: cantonal, universal minimum (>= CHF 200-250/mo), NOT means-tested." }, "non_cash.vacation_days_statutory": { "source": "ch.ch Swiss federal portal https://www.ch.ch/en/work/leave-and-holidays/holiday-entitlement/", "last_updated": "2026-05-16", "confidence": "high", "note": "Statutory minimum 4 weeks (20 days) for employees 20+. Many employers offer 25; using statutory minimum per convention" }, "non_cash.public_holidays": { "source": "Canton of Zurich https://www.zh.ch/en/arbeiten-rente/arbeit/feiertage.html", "last_updated": "2026-05-16", "confidence": "high", "note": "9 paid: 1 federal + 8 cantonal (varies slightly by source)" }, "non_cash.parental_leave_weeks": { "source": "BSV, Maternity, paternity and adoption leave and allowances https://www.bsv.admin.ch/en/income-compensation-parents ; LECA/EOG; ahv-iv.ch leaflets 6.02 (maternity) and 6.04 (paternity)", "last_updated": "2026-08-06", "confidence": "high", "note": "16 wk = 14 maternity (EO, continuous from birth) + 2 paternity (10 working days within 6 months, federal statute since 1 Jan 2021), each parent's own entitlement; no federal shared scheme exists. The previously stored value omitted the 2 paternity weeks its own note cited. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md." }, "non_cash.parental_leave_paid_pct": { "source": "BSV https://www.bsv.admin.ch/en/income-compensation-parents : up to 80% of average income, max. CHF 220 per day, for both the maternity and paternity allowances", "last_updated": "2026-08-06", "confidence": "high", "note": "80% capped at CHF 220/day = CHF 1,540/wk, a cap reached at about CHF 99,000/yr, so every ICP salary receives the flat maximum: at CHF 130,000 the effective rate is about 62% and falls as salary rises. Kept as pct 80 because below the threshold it is a true 80%; the capped-percentage shape has no honest flat-block representation, so the cap arithmetic lives here." }, "non_cash.universal_healthcare": { "source": "ch.ch", "last_updated": "2026-05-16", "confidence": "high", "note": "Mandatory private (KVG/LAMal). Universal coverage but NOT tax-funded. Set false to flag the premium line item" }, "non_cash.commute_minutes_avg": { "source": "Swiss Federal Statistical Office (BFS), PxWeb cube px-x-2105000000_206, Agglomeration Zuerich. https://www.pxweb.bfs.admin.ch/pxweb/de/px-x-2105000000_206/px-x-2105000000_206/px-x-2105000000_206.px", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Durchschnittliche Arbeitswegzeit, in Minuten (ein Hinweg)", "year": "2020", "note": "35.6 min. Agglomeration geography. The previous entry cited the BFS topic page while storing 25, a number that page does not give; the agglomeration cube states the number now stored. Swiss cities were the most understated block in the sweep." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Zurich: CHF 137,843, held in labor_market.tech_salary_distribution (source_slug greater-zurich-area, as_of 2026-07-18, table confidence usable). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "CHF 137,843 / 0.807 = USD 170,809; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.6832, R5-rounded down to 0.68. REPLACES 1.05 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Zurich by 54%. The measured p50 was 0.651 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE Zurich all-in median ~CHF 230k (Google Zurich L5 ~CHF 230-250k). CHF near parity USD (~$260k equiv). Raw ratio vs Seattle ~1.18; adjusted to 1.05 reflecting non-FAANG (UBS/Roche/ABB at CHF 130-180k) and avoiding FAANG over-index. ONLY non-US city in dataset above Seattle parity" }, "tax.bvg": { "source": "BVG statutory minimum, 2026 parameters (unchanged from 2025): entry threshold CHF 22,680, coordination deduction CHF 26,460, insured-salary upper limit CHF 90,720, coordinated salary clamped to 3,780-64,260. ASGA Pensionskasse, Merkblatt Renten und Grenzbetraege per 1.1.2026 https://www.asga.ch/wp-content/uploads/downloads/merkblaetter/merkblaetter-d/renten-und-grenzbetraege_2026_d.pdf ; BSV https://www.bsv.admin.ch/de/altersvorsorge-beruflichen-vorsorge (confirms the 22,680 threshold, savings from age 25, and the BVG Art. 66 employer-parity rule); age credits BVG Art. 16: 7/10/15/18% of coordinated salary by age band (Sicherheitsfonds BVG, sfbvg.ch Sparprozess), employee pays half: 3.5/5/7.5/9%.", "last_updated": "2026-08-06", "confidence": "high", "note": "Statutory MINIMUM only, deliberately: real Swiss professional plans insure above the mandatory band (ueberobligatorisch) and add plan-specific risk/admin premiums, so a typical payslip deducts MORE than this charge; everything above the legal floor is excluded under the statutory-minimum convention (the same rule that keeps Swedish ITP and Danish employer pensions out entirely). The charge docks take-home inside taxes.payroll and is carved to the forced-savings own-money row (funded, FZG-vested, portable). It runs on the same wage base as the AHV/ALV slice and never reduces the income-tax base: the CH-1 deduction floor stays CHF 13,500/27,000, which errs against Switzerland. Applied per earner (entry threshold and coordination deduction per person); the 18-24 bracket pays 0 because retirement credits start at 25 (BVG Art. 7). Decision log: docs/superpowers/research/2026-08-06-pension-omission-pair.md" }, "tax.cantonal_brackets_married": { "source": "CH-2 (exact, 2026-07-26): Zürich § 35 Abs. 2 StG Verheiratetentarif (married/one-parent), einfache Staatssteuer x 214% Steuerfuss. A genuine separate Doppeltarif schedule (thresholds are NOT 2x the Grundtarif; e.g. the 7% band starts at 64,100 vs 45,700 single = 1.40x), so it retains a HNW marriage penalty and is modelled bracket-for-bracket, not as a divisor. ESTV Kantonsblatt Zürich (§ 35 Abs. 2, Steuerperiode 2026).", "last_updated": "2026-07-26", "confidence": "high" }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/greater-zurich-area", "last_updated": "2026-07-18", "confidence": "usable", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "EF English Proficiency Index 2024 (High band)", "last_updated": "2024", "note": "National-level English usability for tech work. Band: high." }, "tax.federal_brackets_married": { "source": "ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Exact 2026 Verheiratetentarif (Art. 36 Abs. 2 DBG), marginal form with the 11.5% flat crossover at 941,300. Replaces the divisor-2 full-splitting approximation, which UNDERSTATED federal married tax (exact 7,579.60 vs splitting 3,185.42 at joint taxable 166,920) - the federal tariff is harsher than splitting. All checkpoints re-derived by hand; see docs/superpowers/research/2026-08-08-ch-federal-worksheet.md." }, "tax.payroll_deductible": { "source": "Art. 33 Abs. 1 lit. d + f DBG SR 642.11, consolidation in force 2026-01-01 (fedlex)", "last_updated": "2026-08-07", "confidence": "high", "note": "Employee AHV/IV/EO, ALV and BVG contributions are fully deductible from the income-tax base (salary-certificate net is the taxable base). Gates the per-earner SI+BVG deduction in computeSwitzerlandTaxes; the charged amounts are the deducted amounts, so the FORCED_SAVINGS carve cannot double-count. Replaces the frozen CHF 13,500/27,000 CH-1 floor (2026-07-26), which bundled SI at one income level." }, "tax.professional_forfait": { "source": "Berufskostenverordnung SR 642.118.1 Art. 7 + Anhang Ziff. 1 (fedlex); 2026 min/max confirmed in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Uebrige Berufskosten pauschale: 3% of NET salary (post-SI/BVG), min CHF 2,000 / max CHF 4,000, per earner with income. Commuting and meal deductions excluded (variable/behavioral, conservative)." }, "tax.insurance_deduction_single": { "source": "Art. 33 Abs. 1 lit. g DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amount in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Versicherungsabzug CHF 1,800 (taxpayer WITH pillar-2 contributions, the employee case). Real KVG premiums exceed this cap and some cantons deduct more (BS 4,200, GE full premiums) - federal amount kept for the one-base model, conservative." }, "tax.insurance_deduction_married": { "source": "Art. 33 Abs. 1 lit. g DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amount in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Versicherungsabzug CHF 3,700 married (WITH pillar-2 amounts). Per-child +700 supplement excluded (conservative)." }, "tax.married_deduction": { "source": "Art. 35 Abs. 1 lit. c DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amount in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Verheiratetenabzug CHF 2,800 (2026), automatic for married couples." }, "tax.two_earner_deduction": { "source": "Art. 33 Abs. 2 DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amounts in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Zweitverdienerabzug: 50% of the lower earned income, min CHF 8,600 / max CHF 14,100, capped at the lower income; applies only when both spouses have earned income. Federal amount used for the one-base model (cantonal equivalents are smaller: BS 1,100, ZH ~6,100); partially offset by the larger cantonal insurance deductions we exclude - see the reasoning in docs/superpowers/research/2026-08-08-tax-structural-ch-ie.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "BAG Kostenbeteiligung https://www.bag.admin.ch/de/krankenversicherung-kostenbeteiligung-fuer-in-der-schweiz-wohnhafte-versicherte + BAG OKP statutory cost-sharing aggregate CHF 5.7bn (2024) over 8.96m insured, adult/child split from BAG age-group net benefits", "last_updated": "2026-08-08", "confidence": "high", "note": "Blended CHF 636 per insured person; children ~CHF 140 each (zero franchise, halved CHF 350 Selbstbehalt cap, exempt from the hospital per-diem), leaving CHF 749 per adult, plus a ~7% conservative uplift for non-KVG self-pay that is neither dental nor long-term care. REPLACES a value that was a single adult's statutory MAXIMUM: the old CHF 1,200 was above the CHF 1,000 minimum-franchise ceiling and was billed unchanged to a family of four whose statutory maximum is CHF 2,700. Geneva keeps its existing 1.083 premium, which is a utilisation difference, not a rules difference: cost-sharing law is federal and identical in every canton. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "non_cash.sick_leave": { "source": "Code of Obligations / Obligationenrecht (OR/CO) Art. 324a, paragraphs 1-4.; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Steuerverwaltung Kanton Bern, Taxinfo Saeule 3a https://taxinfo.sv.fin.be.ch/taxinfo/a2830da2-1a00-4928-b58b-74da5d48a598 ; EFD/ESTV confirmation of the 2026 amounts, 17 Nov 2025 https://www.estv.admin.ch/de/newnsb/xgRMirCsezICX4rtof9Lm", "last_updated": "2026-08-11", "confidence": "high", "note": "Saeule 3a kleiner Abzug, CHF 7,258 for 2026 (unchanged from 2025), for anyone who also contributes to a Pensionskasse, which our ICP does (both CH cities carry a bvg block). The grosser Abzug (20% of income to CHF 36,288, no 2nd pillar) is NOT our case. Employee-only; ESTV states the maximum deduction is simultaneously the contribution limit. 2nd-pillar Einkauf is also deductible but is bounded by a personal Einkaufspotenzial we cannot know, so it is excluded." } }, "vancouver": { "tax.filing_basis": { "source": "PwC WWTS - Canada (Individual, Tax administration): married taxpayers file separately; joint returns are not allowed.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "CRA 2026 https://www.canada.ca/en/revenue-agency/services/tax/individuals/frequently-asked-questions-individuals/canadian-income-tax-rates-individuals-current-previous-years.html", "last_updated": "2026-05-16", "confidence": "high", "note": "Federal lowest rate dropped from 15% to 14% effective 2026 (cut announced 2025, full-year 14% for 2026). CRA indexation factor 2.0%" }, "tax.provincial_brackets": { "source": "Gov.bc.ca + TaxTips.ca https://www2.gov.bc.ca/gov/content/taxes/income-taxes/personal/tax-rates", "last_updated": "2026-05-16", "confidence": "high", "note": "BC 2026 indexation 2.2%. Lowest bracket raised from 5.06% to 5.60%. Indexation paused 2027-2030" }, "tax.cpp_rate": { "source": "CRA https://www.canada.ca/en/revenue-agency/services/tax/businesses/topics/payroll/payroll-deductions-contributions/canada-pension-plan-cpp/cpp-contribution-rates-maximums-exemptions.html", "last_updated": "2026-05-16", "confidence": "high", "note": "Employee CPP 5.95% on earnings between $3,500 basic exemption and YMPE. Unchanged from 2025" }, "tax.cpp_cap": { "source": "CRA YMPE 2026", "last_updated": "2026-05-16", "confidence": "high", "note": "YMPE $74,600 (up from $71,300 in 2025, ~4.6%)" }, "tax.cpp2_rate": { "source": "CRA CPP2 enhanced", "last_updated": "2026-05-16", "confidence": "high", "note": "CPP2 4% on earnings between YMPE and YAMPE" }, "tax.cpp2_cap": { "source": "Penad / CRA 2026 https://penad.com/2025/12/09/2026-cpp-ympe-and-contribution-rates-announced/", "last_updated": "2026-05-16", "confidence": "high", "note": "YAMPE 2026 $85,000; max CPP2 employee contribution $416" }, "tax.ei_rate": { "source": "ESDC 2026 https://www.canada.ca/en/employment-social-development/programs/ei/ei-list/ei-employers/premium-reduction-program/2026-maximum-insurable-earnings.html", "last_updated": "2026-05-16", "confidence": "high", "note": "$1.63 per $100 non-Quebec, down from $1.64 in 2025" }, "tax.ei_cap": { "source": "ESDC MIE 2026", "last_updated": "2026-05-16", "confidence": "high", "note": "MIE $68,900 (up from $65,700 in 2025). Max employee premium $1,123.07" }, "tax.personal_allowance": { "source": "Harvest ETFs / CRA BPA 2026", "last_updated": "2026-05-16", "confidence": "high", "note": "Federal BPA $16,452 (from $16,129 in 2025). Phases down to $14,829 above $258,482 net income; full at/below $181,440" }, "tax.provincial_personal_allowance": { "source": "TaxTips.ca BC BPA 2026", "last_updated": "2026-05-16", "confidence": "high", "note": "BC BPA $13,216, indexed 2.2% from 2025; credited at new lowest rate 5.60%" }, "tax.sales_tax_rate": { "source": "Gov.bc.ca PST", "last_updated": "2026-05-16", "confidence": "high", "note": "5% federal GST + 7% BC PST, both on pre-tax price. Unchanged for 2026" }, "tax.property_tax_rate_pct": { "source": "WOWA + Vancouver tax rates https://wowa.ca/taxes/vancouver-property-tax", "last_updated": "2026-05-16", "confidence": "medium", "note": "Vancouver 2025 Class 1 residential effective rate 0.31154% of assessed value (municipal + school + Metro/TransLink/BCA). 2026 not yet published; using 2025 as best available" }, "housing.median_rent_by_bedroom": { "source": "Zumper Vancouver Rent Research https://www.zumper.com/rent-research/vancouver-bc", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "housing.median_home_price_by_bedroom": { "source": "Greater Vancouver REALTORS MLS HPI + WOWA https://www.gvrealtors.ca/market-watch/MLS-HPI-home-price-comparison.html", "last_updated": "2026-05-16", "confidence": "medium", "note": "REBGV publishes by property type, not bedroom. Composite $1.10M; condo $706.7K; townhouse $1.05M; detached $1.85M. Estimates assume studio/1BR=condo, 2BR mix, 3BR mix townhouse/entry detached" }, "housing.utilities_monthly": { "source": "MovingWaldo BC + BC Hydro/FortisBC", "last_updated": "2026-05-16", "confidence": "medium", "note": "BC Hydro ~$100/mo post-2026 + FortisBC if applicable. Many apartments electric-only with heat in rent" }, "housing.renters_insurance_monthly": { "source": "Rates.ca / APOLLO BC tenant", "last_updated": "2026-05-16", "confidence": "high", "note": "$15-25/mo typical; $20 midpoint" }, "housing.internet_phone_monthly": { "source": "PlanHub Vancouver May 2026", "last_updated": "2026-05-16", "confidence": "high", "note": "Internet ~$70 + mobile ~$48 = ~$118-120" }, "healthcare.premium_monthly_by_family_size": { "source": "BC Government MSP https://www2.gov.bc.ca/gov/content/health/health-drug-coverage/msp/bc-residents/premiums", "last_updated": "2026-05-16", "confidence": "high", "note": "BC eliminated MSP premiums effective Jan 1 2020. Universal public coverage via Medicare for all BC residents" }, "healthcare.annual_out_of_pocket_avg": { "source": "CIHI national health expenditure household out-of-pocket + provincial drug plans: Quebec RAMQ covers a registered parent's children free, Ontario Trillium and BC Fair PharmaCare are income-deductible-based", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "healthcare.dental_vision_monthly": { "source": "Canada Life / Sun Life individual extended health", "last_updated": "2026-05-16", "confidence": "medium", "note": "Employer extended health plan contribution + supplemental dental/vision; individual market CAD 50-100/mo for single adult" }, "childcare.infant_monthly": { "source": "BC $10/day Child Care + market rate avg", "last_updated": "2026-05-16", "confidence": "medium", "note": "BC's Child Care BC + federal CWELCC cap fees at participating centers (~$200/mo for 0-5), but demand >> supply. Many families pay market rate CAD 1,400-1,800/mo for infant care. Used CAD 1,500 as blended" }, "childcare.preschool_monthly": { "source": "BC Child Care + market", "last_updated": "2026-05-16", "confidence": "medium", "note": "3-5 preschool typically CAD 1,000-1,400/mo unsubsidized; CAD 200/mo at $10aDay sites with available slot" }, "childcare.after_school_monthly": { "source": "BC Child Care + YMCA fee surveys", "last_updated": "2026-05-16", "confidence": "medium", "note": "School-age after-school CAD 400-600/mo" }, "childcare.subsidy_per_kid_annual": { "source": "CWELCC (Canada-Wide Early Learning and Child Care) provincial fee schedules; double-subtraction correction 2026-08-02", "last_updated": "2026-08-02", "confidence": "medium", "note": "Set to 0 on 2026-08-02, was CAD 3,600. The old value was the CWELCC fee reduction. CWELCC is delivered as a lower posted fee at a participating centre, not as cash to the household, and childcare.preschool_monthly already carries the reduced rate, so the reduction was netted twice. The preschool note already recorded CAD 620/month as a blend between the CAD 1,000-1,400 unsubsidised rate and the CAD 200 ChildCareBC rate, so the reduction was inside the monthly figure before this field touched it. This field holds CASH THAT REACHES THE HOUSEHOLD. Where the monthly rate it nets against is already the price the household actually pays, the value is 0 and the relief is described here instead. See docs/superpowers/research/2026-08-02-childcare-double-subtraction.md. Open follow-up: the Vancouver monthly rate itself is a blend of the capped and uncapped rates and still owes its own sourcing pass." }, "food.groceries_monthly_per_adult": { "source": "Numbeo Vancouver Markets", "last_updated": "2026-05-16", "confidence": "high", "note": "Vancouver ~15-20% above Canadian avg per StatCan" }, "food.groceries_monthly_per_kid": { "source": "Statistics Canada household food", "last_updated": "2026-05-16", "confidence": "medium", "note": "~50% of adult per Canada Food Price Report family-of-four" }, "food.dining_out_monthly_baseline": { "source": "Numbeo Vancouver Restaurants", "last_updated": "2026-05-16", "confidence": "high", "note": "~2 mid-range meals (CAD 30-35/pp) + ~4 casual (CAD 20/pp) per month" }, "transit.monthly_pass": { "source": "TransLink Compass https://www.translink.ca/transit-fares/pricing-and-fare-zones", "last_updated": "2026-05-16", "confidence": "high", "note": "1-zone adult monthly pass. 2-zone (cross-bridge commute) ~CAD 144" }, "transit.annual_car_cost_avg": { "source": "CAA Driving Costs + ICBC averages", "last_updated": "2026-05-16", "confidence": "medium", "note": "Mid-size sedan all-in: depreciation $3,500 + ICBC insurance $1,800-2,200 (BC notably high) + fuel $2,400 + maintenance $1,200 + parking $2,400 + registration $200" }, "discretionary.entertainment_monthly_baseline": { "source": "Numbeo Vancouver", "last_updated": "2026-05-16", "confidence": "medium", "note": "Gym ~$60 + streaming ~$25 + 2 cinema ~$32 + outings ~$80" }, "discretionary.local_vacation_annual_baseline": { "source": "TripAdvisor / Destination BC", "last_updated": "2026-05-16", "confidence": "medium", "note": "2 Whistler ski weekends (~$1000 each) + Tofino long weekend (~$1500) + day trips (~$700). Strong outdoorsy BC regional travel culture" }, "travel_home.avg_round_trip_per_person_to_other_city": { "source": "Google Flights YVR-SEA", "last_updated": "2026-05-16", "confidence": "high", "note": "Vancouver-Seattle short-haul (~30 min, ~230km). RT CAD 200-350 booked 4-8 weeks; many drive instead" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "canada.ca Canada Child Benefit calculation sheet (July 2025-June 2026)", "last_updated": "2026-07-17", "note": "CCB is means-tested on adjusted family net income: a 2-child mixed-age family reaches $0 by ~CAD 236k AFNI. The published CAD 231k-339k households receive ~$0 (residual <$150/kid only at the very low end), and Ontario/Alberta have no high-income floor. Set to 0 (was a flat 3,000)." }, "non_cash.vacation_days_statutory": { "source": "BC Employment Standards Act s.57(1)(a) https://www.bclaws.gov.bc.ca/civix/document/id/complete/statreg/96113_01; Ontario ESA 2000 s.33(1)(a); Quebec LNT art. 68; Alberta Employment Standards Code s.34; Canada Labour Code s.184", "last_updated": "2026-08-08", "confidence": "high", "note": "CORRECTED 2026-08-08, 15 to 10. The year-1 statutory floor is 2 weeks in every jurisdiction we carry a city in: BC s.57(1)(a), Ontario s.33(1)(a), Quebec art. 68, Alberta s.34 and the federal Canada Labour Code s.184 all grant 2 weeks after 12 months. 15 days is the 3-week tier, reached at year 5 in BC, Ontario, Alberta and federally, and year 3 in Quebec. The previous version of this note already carried the correct floor and it was never applied to the stored value. Canadian vacation weeks are natively a 5-day-week unit, so no conversion applies, and vacation is additive to public holidays in every Canadian jurisdiction. Applied to all 6 Canadian cities: vacation is genuinely provincial so it is not enforced in _national_fields.json, but every province we touch agrees on the year-1 floor." }, "non_cash.public_holidays": { "source": "BC Statutory Holidays https://www2.gov.bc.ca/gov/content/employment-business/employment-standards-advice/employment-standards/statutory-holidays", "last_updated": "2026-08-08", "confidence": "high", "note": "RE-VERIFIED 2026-08-08, value stands at 11 and the previous note was wrong on both halves. BC recognises 11 statutory holidays, not 10: New Year's Day, Family Day, Good Friday, Victoria Day, Canada Day, B.C. Day, Labour Day, National Day for Truth and Reconciliation, Thanksgiving, Remembrance Day, Christmas Day. Truth and Reconciliation became a BC statutory holiday in 2023, after this note was first written. Easter Monday is NOT one of them; the BC government page states that Easter Sunday, Easter Monday and Boxing Day are not statutory holidays in B.C." }, "non_cash.parental_leave_weeks": { "source": "Employment Insurance Act ss. 22-23: maternity 15 wk + standard parental up to 35 wk per parent (40 when shared) https://www.canada.ca/en/services/benefits/ei/ei-maternity-parental.html", "last_updated": "2026-08-06", "confidence": "high", "note": "50 = EI maternity 15 (mother) + standard parental 35, a shared pot counted once at the one-parent maximum. The 5 extra weeks that exist only when the second parent also claims are excluded as conditional on both parents' behaviour (same ruling as Portugal's sharing bonus). The extended option (up to 61 wk one parent, at 33%) was rejected: near-identical capped dollars stretched over 26 more weeks would inflate the weeks column. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. Quebec runs QPIP instead, so montreal legitimately diverges (city-varying by design in data/_national_fields.json)." }, "non_cash.parental_leave_paid_pct": { "source": "canada.ca, EI maximum insurable earnings for 2026: CAD 68,900 https://www.canada.ca/en/employment-social-development/programs/ei/ei-list/ei-employers/premium-reduction-program/2026-maximum-insurable-earnings.html", "last_updated": "2026-08-06", "confidence": "high", "note": "55% of insurable earnings for both maternity and standard parental weeks, capped at the 2026 MIE of CAD 68,900, max CAD 729/wk. Against a CAD 130,000 Toronto tech salary the cap makes the effective rate about 29%; the 55 is real only below the MIE. Headline kept per field convention; the capped-percentage shape has no honest flat-block representation, so the arithmetic lives here." }, "non_cash.universal_healthcare": { "source": "BC MSP https://www2.gov.bc.ca/gov/content/health/health-drug-coverage/msp", "last_updated": "2026-05-16", "confidence": "high", "note": "Universal Medicare, BC MSP premiums eliminated Jan 2020" }, "non_cash.commute_minutes_avg": { "source": "Statistics Canada, average commute times for the 15 largest CMAs, May 2025 (The Daily, 2025-08-26, Table 4). https://www150.statcan.gc.ca/n1/daily-quotidien/250826/t004a-eng.htm", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Commuting duration, one way", "year": "2025", "note": "Vancouver CMA 31.1. Replaces an undated 'Statistics Canada Commuting Survey' citation that carried the May 2021 Census value." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Vancouver: CAD 173,703, held in labor_market.tech_salary_distribution (source_slug vancouver-can, as_of 2026-07-18, table confidence usable). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "CAD 173,703 / 1.4 = USD 124,074; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.4963, R5-rounded down to 0.49. REPLACES 0.73 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Vancouver by 49%. The measured p50 was 0.680 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE Vancouver median ~CAD 220k (range CAD 180-260k FAANG-tier; mid-tier CAD 120-160k) ≈ USD 160k at FX 0.73. Seattle baseline $220k. Naive ratio 0.73. Amazon/MSFT Vancouver pay below Seattle counterparts (well-documented Canadian tech wage suppression); US-remote workers anchor top of local market upward" }, "tax.bpa_phaseout": { "source": "Federal BPA supplement phases 16,452 -> 14,829 over net income 181,440-258,482 (CRA 2026); https://www.canada.ca/en/revenue-agency/programs/about-canada-revenue-agency-cra/federal-government-budgets/basic-personal-amount.html; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "BPA modeled as a non-refundable credit at the lowest bracket rate with this federal supplement phase-out." }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/vancouver-can", "last_updated": "2026-07-18", "confidence": "usable", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "housing.homeowners_insurance_pct_of_value": { "source": "Provincial average premium (rates.ca Q2 2026, MyChoice city data, IBC) over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "CIHI national health expenditure household out-of-pocket + provincial drug plans: Quebec RAMQ covers a registered parent's children free, Ontario Trillium and BC Fair PharmaCare are income-deductible-based", "last_updated": "2026-08-08", "confidence": "medium", "note": "Medically necessary physician and hospital care is free everywhere, so the whole of this field is prescription drugs and allied health, which is PROVINCIAL, and the existing per-city spread is kept for that reason. Employer drug plans are a benefit and a premium, not out-of-pocket. Montreal carries a much smaller child increment than the other cities because of the RAMQ free-child rule. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "CRA, MP, DB, RRSP, DPSP, ALDA, TFSA limits and the YMPE https://www.canada.ca/en/revenue-agency/services/tax/registered-plans-administrators/pspa/mp-rrsp-dpsp-tfsa-limits-ympe.html ; https://www.canada.ca/en/revenue-agency/services/tax/individuals/topics/rrsps-related-plans/contributing-a-rrsp-prpp/contributions-affect-your-rrsp-prpp-deduction-limit.html", "last_updated": "2026-08-11", "confidence": "high", "note": "RRSP deduction limit (line 20800), 2026: 18% of prior-year earned income to an annual dollar limit. Pinned at 33,810 via the statutory identity RRSP limit year X = money-purchase limit year X-1 (CRA confirms 2025 MP = 33,810). The PENSION ADJUSTMENT is deliberately NOT modeled, so this is the no-workplace-DB case; a DB-plan member has materially less room. Carry-forward of unused room is also unmodeled. Never binds at our incomes; the 18% governs below CAD 187,833." } }, "san-francisco": { "tax.federal_brackets": { "source": "IRS Rev Proc 2025-32 (2026 inflation, OBBBA) https://www.irs.gov/newsroom/irs-releases-tax-inflation-adjustments-for-tax-year-2026-including-amendments-from-the-one-big-beautiful-bill", "last_updated": "2026-05-16", "confidence": "high", "note": "2026 single-filer brackets matched to Seattle for cross-US parity. IRS published top-bracket threshold at ~$640,600 for 2026; we use the indexed 2025 values for now and will refresh Seattle + SF together" }, "tax.federal_brackets_married": { "source": "IRS Rev Proc 2025-32", "last_updated": "2026-05-16", "confidence": "high", "note": "MFJ brackets matched to Seattle" }, "tax.state_brackets": { "source": "CA FTB, 2025 California Tax Rate Schedules (Form 540 booklet) https://www.ftb.ca.gov/forms/2025/2025-540-tax-rate-schedules.pdf", "last_updated": "2026-08-01", "confidence": "high", "note": "Schedule X, single or married/RDP filing separately, tax year 2025: 1% to 11,079 then 2/4/6/8% to 72,724, 9.3% to 371,479, 10.3% to 445,771, 11.3% to 742,953, 12.3% above. Refreshed 2026-08-01: the file carried the TY2024 thresholds (10,756 / 25,499 / ... / 721,314) while already citing the TY2025 schedule, so the data disagreed with its own citation. TY2025 is FTB's newest published year; the TY2026 schedules are not out. Corrected 2026-08-06: this note used to say the 1% Behavioral Health Services Tax above $1M taxable (renamed from the Mental Health Services Tax for TY2025 by Prop 1) was absent from the engine. It is not absent. app.js computes it as taxes.state_surtax from tax.state_surtax_rate 0.01 and tax.state_surtax_threshold 1000000, both of which this city carries, and a single filer on $1,500,000 gross pays $4,839 of it (identical in all four California cities). The threshold is per return and does not double for joint filers, which is the statute rather than a gap: RTC 17043(c) disapplies section 17045, the income-splitting rule that doubles California's regular MFJ bands, so no schema field for tax.state_surtax_threshold_married exists in any city JSON. https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=RTC§ionNum=17043" }, "tax.state_disability_insurance_rate": { "source": "California EDD, SDI Contribution Rates and Benefit Amounts https://edd.ca.gov/en/disability/Contribution_Rates_and_Benefit_Amounts/", "last_updated": "2026-07-30", "confidence": "high", "note": "Employee State Disability Insurance / Paid Family Leave contribution, 1.3% for 2026 (up from 1.2% in 2025). EDD lists the taxable wage ceiling and the maximum contribution as N/A: SB 951 eliminated both effective 2026-01-01, so this is levied on every dollar of wages with no cap. Added 2026-07-30; before that the engine omitted it entirely, understating California payroll by 1.3% of gross. California is the only modeled state with an employee-side SDI levy, so the field is per-city, not a USA-wide constant." }, "tax.city_income_tax_rate": { "source": "SF Treasurer & Tax Collector", "last_updated": "2026-05-16", "confidence": "high", "note": "SF has no personal income tax. Local taxes apply to businesses only" }, "tax.social_security_cap": { "source": "SSA 2026 COLA Fact Sheet https://www.ssa.gov/news/en/cola/factsheets/2026.html", "last_updated": "2026-05-16", "confidence": "high", "note": "2026 SS wage base $184,500 (up from $176,100 in 2025)" }, "tax.sales_tax_rate": { "source": "CDTFA city/county sales tax https://www.cdtfa.ca.gov/taxes-and-fees/rates.aspx", "last_updated": "2026-05-16", "confidence": "high", "note": "SF combined 8.625%: 6.0% state + 1.25% county + 1.375% district" }, "tax.property_tax_rate_pct": { "source": "SF Treasurer FY2025-26 secured property tax https://sftreasurer.org/property/secured-property-taxes", "last_updated": "2026-05-16", "confidence": "high", "note": "FY2025-26 effective rate 1.18268% of assessed. Prop 13 caps base at 1% + voter-approved bonds. Long-held properties pay much less on market-value basis" }, "tax.standard_deduction_single": { "source": "IRS 2026 (OBBBA)", "last_updated": "2026-05-16", "confidence": "high", "note": "2026 single $16,100 (up from $15,000 in 2025)" }, "tax.standard_deduction_married": { "source": "IRS 2026 (OBBBA)", "last_updated": "2026-05-16", "confidence": "high", "note": "2026 MFJ $32,200 (up from $30,000 in 2025)" }, "housing.median_rent_by_bedroom": { "source": "Zillow Rental Manager / Rent.com / RentHop May 2026", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "housing.median_home_price_by_bedroom": { "source": "Redfin SF + Zillow ZHVI May 2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Redfin median $1.7M, Zillow ZHVI $1.27M. Bedroom breakdown estimated using SF ratios" }, "housing.utilities_monthly": { "source": "Utility-Rates SF / Apartment List", "last_updated": "2026-05-16", "confidence": "medium", "note": "Basic only (electric+gas+water+sewer); excludes internet" }, "housing.renters_insurance_monthly": { "source": "Insure.com / Insurify SF 2026", "last_updated": "2026-05-16", "confidence": "high" }, "housing.internet_phone_monthly": { "source": "CompareInternet SF + AT&T/Verizon bundles 2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Fiber broadband ~$60-70 + unlimited mobile ~$45-50" }, "healthcare.premium_monthly_by_family_size": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "high", "note": "Employee contribution to employer premium. KFF: ~$114/mo single, ~$525/mo family avg, adjusted up 15-20% for SF higher cost-of-care" }, "healthcare.premium_age_multiplier": { "source": "ACA age-rating bands (HHS)", "last_updated": "2026-05-16", "confidence": "high", "note": "Standard 3:1 ACA curve as proxy for individual-market exposure" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "healthcare.dental_vision_monthly": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "medium", "note": "The dental and vision PREMIUM, single share. Billed FLAT for US cities (app.js), unlike every other country, where this field is per adult and scales by adults. Reason: the 2026-07-17 recalibration declined to add a dental-by-family-size key and folded the family dental and ortho excess into annual_out_of_pocket_by_family_size instead (plan line 72), so scaling this line would charge that excess twice. Sharpened 2026-08-08: the earlier wording said only the family increment sits in the OOP column, but the plan puts a dental and vision out-of-pocket margin in EVERY row including row 1." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: San Francisco County CA (FIPS 6075), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $3,000." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: San Francisco County CA (FIPS 6075), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $2,480/mo and the ages 1-2 rate $2,870/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $2,400. Does NOT model free universal pre-K where it exists. PRIOR NOTE: Center-based preschool $2,200-2,800/mo DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.after_school_monthly": { "source": "Care.com", "last_updated": "2026-05-16", "confidence": "medium", "note": "After-school programs $600-1,200/mo; midpoint NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." }, "childcare.subsidy_per_kid_annual": { "source": "IRS CDCC / CA FTB", "last_updated": "2026-05-16", "confidence": "high", "note": "US/CA childcare support delivered via tax credits (out of scope for v1). Income-eligible vouchers don't apply to tech-worker household" }, "childcare.income_tested_childcare": { "source": "San Francisco Department of Early Childhood, FY2026-27 Early Learning Scholarship (ELFA) rate and eligibility sheet, effective 2026-07-01 https://www.sf.gov/departments/department-early-childhood", "source_name": "SF DEC ELFA FY25-26 rate sheet + Children's Council SF", "last_updated": "2026-08-02", "confidence": "high", "note": "UPDATED 2026-08-02 to the FY2026-27 bands, effective 2026-07-01: 150% AMI 233,800 to 243,150 and 200% AMI 311,700 to 324,200, about +4.0% year over year. credit_against_monthly stays 2115: the preschooler reimbursement rate was flat, only the AMI dollars moved. DISPLAY ONLY, never netted from the bill (see computeChildcareCost). The reader-facing sentence is now BUILT FROM THIS BLOCK by childcareSubsidyNote in app.js, so it cannot go stale separately; it used to be hardcoded in index.html and build/build-pages.mjs, both of which kept the FY25-26 dollars after this refresh date. KNOWN SIMPLIFICATION: SFDEC actually runs two tiers under the modeled 100% band, 0-110% AMI genuinely free (family of four 178,300) and 111-150% AMI a credit capped at the DEC rate where a family can still owe the gap; we model the single 150% band. Working: docs/superpowers/research/2026-08-02-us-childcare-ceilings.md" }, "food.groceries_monthly_per_adult": { "source": "Numbeo SF + BLS CEX West Region", "last_updated": "2026-05-16", "confidence": "high", "note": "SF among most expensive US grocery markets (~30% above national avg)" }, "food.groceries_monthly_per_kid": { "source": "BLS CEX + USDA Food Plans", "last_updated": "2026-05-16", "confidence": "high", "note": "~55% of adult per USDA moderate-cost plan" }, "food.dining_out_monthly_baseline": { "source": "Numbeo SF Restaurants", "last_updated": "2026-05-16", "confidence": "high", "note": "Mid-range $30-50/pp; baseline ~2 casual outings/week mix" }, "transit.monthly_pass": { "source": "SFMTA https://www.sfmta.com/getting-around/transit/fares-passes/monthly-passes", "last_updated": "2026-05-16", "confidence": "high", "note": "Muni 'A' pass includes BART within SF. South Bay BART/Caltrain commute substantially more" }, "transit.annual_car_cost_avg": { "source": "AAA Your Driving Costs + SF parking adjustment", "last_updated": "2026-05-16", "confidence": "medium", "note": "AAA national $12,300 + ~$3,000 SF parking premium + CA fuel taxes + elevated insurance" }, "discretionary.entertainment_monthly_baseline": { "source": "Numbeo SF + benchmarks", "last_updated": "2026-05-16", "confidence": "high", "note": "Gym $100 + 2 cinema $35 + 3 streaming $45 + occasional outings $70. SF tech tends to spend liberally" }, "discretionary.local_vacation_annual_baseline": { "source": "TripAdvisor + Airbnb Tahoe/Napa/Sonoma/coastal", "last_updated": "2026-05-16", "confidence": "medium", "note": "3-4 long weekends to Tahoe/Napa/Sonoma/Mendocino/Monterey. Premium nearby destinations" }, "travel_home.avg_round_trip_per_person_to_other_city": { "source": "Google Flights SFO domestic blended", "last_updated": "2026-05-16", "confidence": "medium", "note": "SFO-SEA $200 / SFO-NYC $350 / SFO-DFW $250. Blended $300. Interpretation differs for US-origin scenarios" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "IRS Child Tax Credit https://www.irs.gov/credits-deductions/individuals/child-tax-credit", "last_updated": "2026-05-16", "confidence": "high", "note": "Federal CTC $2,000/yr per child under 17. Phase-out $400k MFJ/$200k single. CA no state child credit US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "non_cash.public_holidays": { "source": "US OPM Federal Holidays", "last_updated": "2026-05-16", "confidence": "high", "note": "11 federal; CA observes same + Cesar Chavez Day for state employees; private follows federal 11" }, "non_cash.parental_leave_weeks": { "source": "EDD Paid Family Leave https://edd.ca.gov/en/disability/paid-family-leave/ ; https://edd.ca.gov/en/disability/Calculating_PFL_Benefit_Payment_Amounts/", "last_updated": "2026-08-06", "confidence": "high", "note": "16 = 8 weeks of CA PFL bonding per parent within a 12-month period, each parent's own SDI-funded non-transferable entitlement, summed (both-parents flag set). The birthing parent's SDI pregnancy disability (usually 4 weeks before plus 6 after, 8 for cesarean) is medically certified, EXCLUDED, conservative. Was 12/100: the unpaid FMLA window with a fabricated headline. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "EDD: SB 951 rates for claims from 2025 (90% low-earner band, 70% above USD 20,931.31 highest-quarter earnings); https://edd.ca.gov/en/disability/Calculating_PFL_Benefit_Payment_Amounts/", "last_updated": "2026-08-06", "confidence": "high", "note": "0 percentage-paid weeks: for claims from 2025 under SB 951, CA pays 90% for highest-quarter earnings of 722.50-16,279.90, a flat 1,127/wk transition band to 20,931.30, and 70% above 20,931.31 per quarter, capped at USD 1,765/wk in 2026 (1,681 in 2025), binding from about USD 131,000. CAVEAT recorded: below that threshold CA pays a true 70% for the ICP band, so the flat store overstates for earners under about 131k (the engine clamps at 100%); at the salaries we price it is exact and less favorable than any headline. At USD 150,000 the effective rate is about 61%." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, San Francisco-Oakland-Fremont CA Metro Area (31000US41860). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US41860", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 32.7. Replaces a citation that read 'ACS B08303 (SF city)', which was the wrong geography by its own wording; SF city proper is 32.2." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, San Francisco: USD 275,216, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "USD 275,216 / 1 = USD 275,216; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 1.1009, R5-rounded down to 1.1. REPLACES 1.38 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE all-in median ~$305k (FAANG L5/E5 $280-380k incl. Google/Meta/Apple/OpenAI/Anthropic/Stripe). Seattle baseline $220k. Ratio 1.39. Bay Area pays HIGHEST tech salaries globally" }, "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 97200 USD, San Francisco-Oakland-Berkeley metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "labor_market.tech_salary_distribution": { "source": "https://www.bls.gov/oes/current/oes_41860.htm", "last_updated": "2025-05", "confidence": "rich", "note": "BLS OEWS May 2025, SOC 15-1252 Software Developers, annual wage percentiles (full-market occupation distribution; straight-time gross pay incl. incentive pay, excl. equity and nonproduction bonuses). Public domain; API pull 2026-07-19." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "tax.state_brackets_married": { "source": "CA FTB, 2025 California Tax Rate Schedules (Form 540 booklet) https://www.ftb.ca.gov/forms/2025/2025-540-tax-rate-schedules.pdf", "last_updated": "2026-08-01", "confidence": "high", "note": "Schedule Y, married/RDP filing jointly or qualifying surviving spouse, tax year 2025. Verified band by band as EXACTLY 2x Schedule X on all 8 finite thresholds, which is what RTC 17041(b) prescribes; the base-tax columns double too. Added 2026-08-01. Before this, couples were run through the single-filer ladder, overstating California tax by a flat $3,457.64 for any household above about $141k of taxable income, and by $4,029 at a $450k household. Head of household is NOT 2x; no schema field for tax.state_brackets_hoh exists." }, "non_cash.parental_leave_flat": { "source": "EDD Contribution Rates and Benefit Amounts: 2026 maximum weekly benefit USD 1,765 https://edd.ca.gov/en/disability/Contribution_Rates_and_Benefit_Amounts/", "last_updated": "2026-08-06", "confidence": "high", "note": "USD 1,765/wk, 2026 maximum. 16 x 1,765 = USD 28,240 total." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." } }, "nyc": { "tax.federal_brackets": { "source": "IRS Rev Proc 2025 / OBBBA 2026 https://www.irs.gov/newsroom/irs-releases-tax-inflation-adjustments-for-tax-year-2026-including-amendments-from-the-one-big-beautiful-bill", "last_updated": "2026-05-16", "confidence": "high", "note": "2026 single-filer brackets. Top 37% at $640,600. Matches Seattle/SF for US parity" }, "tax.state_brackets": { "source": "NYS Dept of Taxation and Finance, IT-201-I (2025) New York State tax rate schedule https://www.tax.ny.gov/forms/html-instructions/2025/it/it201i-2025.htm", "last_updated": "2026-08-01", "confidence": "high", "note": "Single / married filing separately, tax year 2025. Re-verified 2026-08-01 band for band against the IT-201-I schedule, base amounts included, and left unchanged. NY does not inflation-index these: TY2024 is byte-identical and TY2026 is not published. The flat 3.876% NYC resident rate is also confirmed right for this income range, since the NYC schedule tops out at 3.876% above $50,000 single / $90,000 joint." }, "tax.recapture": { "source": "NYS Dept of Taxation and Finance, Form IT-201-I (2025) tax computation worksheets, pp. 34-39 https://www.tax.ny.gov/pdf/2025/inc/it201i_2025.pdf (year-stamped permalink; the /pdf/current_forms/it/it201i.pdf alias rolls to the next tax year and will stop matching these TY2025 constants)", "last_updated": "2026-08-01", "confidence": "high", "note": "Alternative tax table benefit recapture, NY Tax Law 601(d-1). Above $107,650 of NY AGI (IT-201 line 33, BEFORE the deduction) New York claws back the benefit of its lower brackets, phased over $50,000 of AGI at every bracket threshold, so a high earner converges on the top rate on every dollar. Transcribed from worksheets 1-11: single/MFS from pp. 36-37, married filing jointly from pp. 34-35. All ten base and incremental constants were re-derived independently from the rate schedule (within a band, taxable x flat_rate minus the graduated walk is constant and equals base + incremental for that stage) and every one reproduces to the published dollar. Head of household is in the source (worksheets 12-16) but the engine has no HOH filing status, so it is not carried. Band selection is by TAXABLE income, not by AGI, re-verified against the worksheet headers on 2026-08-01: each header joins two conditions with 'and', and the AGI clause's upper bound is always $25,000,000 rather than the next band, so AGI only gates entry and supplies the phase-in numerator. Statutory rounding: divide by $50,000 and round to the fourth decimal BEFORE multiplying. Above $25,000,000 of AGI it is a flat 10.9% of taxable income with no graduated walk. NYC city tax has NO recapture: 'recapture' appears exactly 10 times in IT-201-I and all 10 are state worksheets." }, "tax.city_income_tax_rate": { "source": "NYC Dept of Finance / NY Dept of Taxation, NYC resident tax schedule https://www.tax.ny.gov/pit/file/tax-tables/", "last_updated": "2026-07-31", "confidence": "high", "note": "NYC resident tax is progressive 3.078-3.876% but flattens to the top 3.876% above roughly $50k, so the flat rate is modeled at the top of the ladder. Re-verified 2026-07-31 against a user report that NYC city tax was missing: it has been modeled since launch, and the flat approximation runs slightly high, never low" }, "tax.state_payroll_premium_rate": { "source": "NY State Paid Family Leave 2026 employee contribution rate https://paidfamilyleave.ny.gov/cost", "last_updated": "2026-07-31", "confidence": "high", "note": "NY PFL employee contribution is 0.432% of wages for 2026" }, "tax.state_payroll_premium_cap": { "source": "NY State Paid Family Leave 2026 maximum annual employee contribution ($411.91) https://paidfamilyleave.ny.gov/cost", "last_updated": "2026-07-31", "confidence": "high", "note": "NY caps the CONTRIBUTION, not the wage, at $411.91/yr. Expressed here as the equivalent wage ceiling: 411.91 / 0.00432 = $95,350. The separate NY statutory disability (DBL) deduction of $0.60/week ($31.20/yr) is left out as immaterial" }, "tax.social_security_cap": { "source": "SSA 2026 wage base", "last_updated": "2026-05-16", "confidence": "high", "note": "2026 SS wage base $184,500" }, "tax.sales_tax_rate": { "source": "NY Dept of Taxation Pub 718", "last_updated": "2026-05-16", "confidence": "high", "note": "NYC combined 8.875% = 4% state + 4.5% city + 0.375% MCTD. Stable since 2009" }, "tax.property_tax_rate_pct": { "source": "NYC DoF FY2026 + CBCNY analysis https://cbcny.org/nyc-effective-tax-rates", "last_updated": "2026-05-16", "confidence": "medium", "note": "Class 1 nominal 20.630% on assessed value but assessed capped at 6% of market with growth caps — effective ~0.88% on market value. Highly variable by neighborhood/tenure; condos (Class 2) differ" }, "tax.standard_deduction_single": { "source": "IRS 2026 OBBBA", "last_updated": "2026-05-16", "confidence": "high", "note": "$16,100 (2026)" }, "tax.standard_deduction_married": { "source": "IRS 2026 OBBBA", "last_updated": "2026-05-16", "confidence": "high", "note": "$32,200 (2026)" }, "housing.median_rent_by_bedroom": { "source": "StreetEasy / Brick Underground Feb 2026 https://www.brickunderground.com/rent/nyc-manhattan-brooklyn-queens-rental-market-report-february-2026", "last_updated": "2026-05-16", "confidence": "high", "note": "Citywide median (all 5 boroughs blended). Manhattan substantially higher; outer boroughs lower" }, "housing.rent_percentiles": { "source": "Census ACS 2024 1-Year, table B25063 gross-rent distribution (metro/CBSA), data.census.gov; P25/P75 interpolated from dollar buckets, interpolated P50 matched published B25064 median (research 2026-07-17 §2). Rounded outward to 2dp per conservative-wider rule.", "last_updated": "2026-08-07", "confidence": "high", "note": "All-unit metro basis (pools bedroom sizes), wider than a within-bedroom band by construction. api.census.gov now requires an API key for queries (checked 2026-08-07); table remains on data.census.gov. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "housing.housing_districts": { "source": "StreetEasy median asking rent, TWO-BEDROOM series, borough rows divided by the NYC row in the SAME file, averaged over the 12 months 2025-08..2026-07: Manhattan 1.3087 = 1.31, Brooklyn 0.9019 = 0.90, Queens 0.7826 = 0.78. Machine-readable primary download (streeteasy.com itself blocks fetch, the data bucket does not): https://streeteasy-market-data-download.s3.amazonaws.com/rentals/TwoBd/medianAskingRent_TwoBd.zip (dashboard https://streeteasy.com/blog/data-dashboard/), retrieved 2026-08-13.", "last_updated": "2026-08-13", "confidence": "high", "note": "BASIS IS A RATIO, NEVER A LEVEL, and the ratio is BEDROOM-MATCHED. Both numerator and denominator are the same publisher, same table, same month, same bedroom count, so the ratio is basis-invariant and is applied to the stored asking-rent median_rent_by_bedroom. CORRECTED 2026-08-13 from 1.18/0.93/0.80, which were StreetEasy ALL-UNIT borough medians (June 2026: citywide 4,200, Manhattan 4,965, Brooklyn 3,900, Queens 3,350 -- confirmed digit-for-digit at source, so the transcription was right and only the basis was wrong). A mix-weighted ratio does not transfer to one bedroom count: Manhattan's stock skews to studios and 1BRs so its all-unit median is dragged down, Queens skews larger so its is pulled up, and the borough spread compresses. Found by Skyler reading a 2-bed Manhattan figure that looked low; he was right, it was understated by 9.9%. Cross-BOROUGH ACS ratios were EVALUATED AND REJECTED for New York: ACS B25031 is a contract rent across sitting tenancies, and the asking-to-contract wedge runs 3.00x in Manhattan against 1.85x in Queens because of rent-stabilised and long-tenure stock, so the ratio is NOT invariant across boroughs; ACS 2024 1-yr would have given Manhattan 1.10 (below the value being corrected) and would have priced Queens 2BR above Brooklyn 2BR, inverting a gap that is not in dispute. Its own Manhattan column falls with bedroom count (0BR 2,237 / 1BR 2,382 / 2BR 1,998 / 3BR 1,848). This does NOT impugn housing.rent_percentiles, which is a within-ACS quartile ratio where the regulated stock sits on both sides and cancels. Two-bedroom chosen as the single anchor because the engine applies one multiplier at every bedroom count (districtLadder), 2BR is the ICP's unit, and it matches London's ONS borough-2BR basis; the tradeoff is priced and accepted, the ladder lands within 1.6% of StreetEasy's published borough levels at 2BR and understates Manhattan at 3BR (conservative direction). Per-bedroom multipliers would need an engine + schema change; derived values are in the research file, ready. Bronx and Staten Island still omitted (display scope, not a data gap: the same download does now carry both). 12-month mean rather than a single month removes NYC's strong rent seasonality, removes single-month median noise, and removes the analyst's month-choice degree of freedom; it is also NARROWER than both alternatives considered (Feb-2026 vintage-match would give Manhattan 1.318, the 12-month max 1.352), which satisfies the conservative-value rule. Working, including the full before/after ladder and the ACS refutation: docs/superpowers/research/2026-08-13-nyc-district-acs-rebase.md; original diagnosis: docs/superpowers/research/2026-08-13-nyc-district-multiplier-basis.md" }, "housing.median_home_price_by_bedroom": { "source": "Douglas Elliman / Miller Samuel Q1 2026 + StreetEasy", "last_updated": "2026-05-16", "confidence": "medium", "note": "Manhattan Q1 2026 median sale $1.225M (condo/co-op). Citywide blend estimated using Brooklyn $900K, Queens $700K, Bronx/SI $500-700K" }, "housing.utilities_monthly": { "source": "StreetEasy / Rove Travel 2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Many NYC rentals include heat + hot water; electricity-dominant. Typical 2BR ~$260" }, "housing.renters_insurance_monthly": { "source": "NerdWallet / Insure.com NYC 2026", "last_updated": "2026-05-16", "confidence": "high" }, "housing.internet_phone_monthly": { "source": "Compare Internet NY 2026 + carrier bundles", "last_updated": "2026-05-16", "confidence": "medium", "note": "Broadband $60-75 (Verizon Fios/Spectrum) + mobile $40-50 ≈ $110 bundled" }, "healthcare.premium_monthly_by_family_size": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "high", "note": "Employee contribution; NYC adjusted +~10% over national avg" }, "healthcare.premium_age_multiplier": { "source": "ACA age-rating bands (CMS)", "last_updated": "2026-05-16", "confidence": "high" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "healthcare.dental_vision_monthly": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "medium", "note": "The dental and vision PREMIUM, single share. Billed FLAT for US cities (app.js), unlike every other country, where this field is per adult and scales by adults. Reason: the 2026-07-17 recalibration declined to add a dental-by-family-size key and folded the family dental and ortho excess into annual_out_of_pocket_by_family_size instead (plan line 72), so scaling this line would charge that excess twice. Sharpened 2026-08-08: the earlier wording said only the family increment sits in the OOP column, but the plan puts a dental and vision out-of-pocket margin in EVERY row including row 1." }, "childcare.infant_monthly": { "source": "Care.com NYC market", "last_updated": "2026-05-16", "confidence": "low", "note": "NDCP EVALUATED AND REJECTED 2026-08-13 (cluster childcare-0813). Every other US city moved onto the National Database of Childcare Prices; New York did not. NDCP’s own technical report states New York supplied only REIMBURSEMENT rates, not market rates (\"ICF did not have access to the survey data to compute market rates\"), and the whole NY series is modelled from a single 2019 study. Applying it would have cut this block to $1,398 preschool / $1,562 infant off a subsidy ceiling. No verified replacement median was obtainable: ocfs.ny.gov sits behind a bot defence, and the NYC Comptroller’s OCFS-attributed 2024 figures state no percentile and put toddler above infant, an inversion seen in none of the other 27 cities. VALUE UNCHANGED; this is the one US city off the spine. Open work: read the OCFS 2024 MRS New York City region day-care-centre median directly. See docs/superpowers/research/2026-08-13-us-childcare-ndcp.md F9. PRIOR NOTE: Center-based $2,500-3,200; tech-worker neighborhoods (Manhattan, Brooklyn) ~$2,800 median" }, "childcare.preschool_monthly": { "source": "Care.com NYC market", "last_updated": "2026-05-16", "confidence": "low", "note": "NDCP EVALUATED AND REJECTED 2026-08-13 (cluster childcare-0813). Every other US city moved onto the National Database of Childcare Prices; New York did not. NDCP’s own technical report states New York supplied only REIMBURSEMENT rates, not market rates (\"ICF did not have access to the survey data to compute market rates\"), and the whole NY series is modelled from a single 2019 study. Applying it would have cut this block to $1,398 preschool / $1,562 infant off a subsidy ceiling. No verified replacement median was obtainable: ocfs.ny.gov sits behind a bot defence, and the NYC Comptroller’s OCFS-attributed 2024 figures state no percentile and put toddler above infant, an inversion seen in none of the other 27 cities. VALUE UNCHANGED; this is the one US city off the spine. Open work: read the OCFS 2024 MRS New York City region day-care-centre median directly. See docs/superpowers/research/2026-08-13-us-childcare-ndcp.md F9. PRIOR NOTE: NYC has FREE universal 3-K and Pre-K (ages 3-4); tech-worker families often choose private centers due to waitlists/hours. Value reflects private market DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.after_school_monthly": { "source": "Care.com NYC", "last_updated": "2026-05-16", "confidence": "medium", "note": "Private $600-1,000; DOE COMPASS programs free but limited NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." }, "childcare.subsidy_per_kid_annual": { "source": "NY State + IRS tax credits", "last_updated": "2026-05-16", "confidence": "high", "note": "Federal CDCC + NYS Child/Dependent Care Credit are tax credits — out of scope per v1. Free NYC 3-K/Pre-K is a service benefit, not cash" }, "childcare.income_tested_childcare": { "source": "https://www.nyc.gov/site/acs/early-care/eligibility.page", "last_updated": "2026-08-05", "confidence": "high", "note": "Ceiling at 85% of NYS median income ($116,492 for a family of 4, ACS income table for 1 Jun 2026 to 31 May 2027); one table with no separate exit line; a household under it may qualify for CCAP." }, "food.groceries_monthly_per_adult": { "source": "Numbeo NYC + BLS CEX", "last_updated": "2026-05-16", "confidence": "high", "note": "NYC ~25-30% above US avg" }, "food.groceries_monthly_per_kid": { "source": "USDA Food Plans + NYC adjustment", "last_updated": "2026-05-16", "confidence": "medium" }, "food.dining_out_monthly_baseline": { "source": "Numbeo NYC + BLS food-away-from-home", "last_updated": "2026-05-16", "confidence": "medium", "note": "Inexpensive meal $25, mid-range $50-75/pp. Tech-worker baseline higher than US average" }, "transit.monthly_pass": { "source": "MTA OMNY https://new.mta.info/fares", "last_updated": "2026-05-16", "confidence": "high", "note": "OMNY 30-day cap $132 (2025). Subway + local bus" }, "transit.annual_car_cost_avg": { "source": "AAA + NYC parking/congestion adjustments", "last_updated": "2026-05-16", "confidence": "medium", "note": "NYC insurance ~50-80% above national avg, Manhattan parking $300-600/mo (rare), congestion pricing $9/day Manhattan CBD (2025). Outer-borough-weighted ~$14,500" }, "discretionary.entertainment_monthly_baseline": { "source": "Numbeo NYC", "last_updated": "2026-05-16", "confidence": "high", "note": "Mid-tier gym $100, 2 cinema $40, 2 streaming $30, outings $80" }, "discretionary.local_vacation_annual_baseline": { "source": "TripAdvisor / Airbnb regional", "last_updated": "2026-05-16", "confidence": "medium", "note": "3-4 long weekends (Hamptons/Hudson Valley/Catskills/Jersey Shore) $600-900 each + short regional $800-1000" }, "travel_home.avg_round_trip_per_person_to_other_city": { "source": "Google Flights JFK/LGA/EWR", "last_updated": "2026-05-16", "confidence": "high", "note": "Transcontinental US via 3 major hubs. Mid-range $300-450 booked 4-8 weeks" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "IRS CTC + NY ESCC blended", "last_updated": "2026-05-16", "confidence": "medium", "note": "Federal CTC $2,000 + NY Empire State Child Credit $330-1000 (phases out for tech-worker incomes). Blended $2,200 US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "non_cash.public_holidays": { "source": "US OPM Federal Holidays", "last_updated": "2026-05-16", "confidence": "high" }, "non_cash.parental_leave_weeks": { "source": "NY PFL https://paidfamilyleave.ny.gov/2026 ; https://paidfamilyleave.ny.gov/bonding-leave-fostering-child", "last_updated": "2026-08-06", "confidence": "high", "note": "24 = 12 weeks of NY PFL bonding leave per parent, non-transferable (the agency confirms both parents hold separate entitlements and may even take them simultaneously subject to employer agreement), summed (both-parents flag set). NY DBL maternity disability (50% of AWW capped at USD 170/wk, about 6-8 certified weeks, combined DBL+PFL max 26 wk in any 52) is EXCLUDED: certified duration and a nominal cap. Was 12/67. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "paidfamilyleave.ny.gov/2026: 67% of AWW capped at 67% of the NYSAWW of 1,833.63", "last_updated": "2026-08-06", "confidence": "high", "note": "0 percentage-paid weeks: NY PFL pays a true 67% of the employee AWW but capped at 67% of the NYSAWW (1,833.63), i.e. USD 1,228.53/wk in 2026, binding from USD 95,348.76, below every salary we price, so it lives in non_cash.parental_leave_flat. NY is the one single-rate state in the US set; it is stored on the same flat shape for consistency, and its threshold is among the lowest so the flat store is exact at our evaluation salaries. At USD 150,000 the effective rate is about 43%. The 2025 cap of 1,177.32 is superseded." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, New York-Newark-Jersey City NY-NJ Metro Area (31000US35620). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US35620", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 36.5 (published MOE +/-0.2). The previous 42 was New York CITY proper (40.6 in 2024) and stale with it. Largest single geography correction in the sweep. Still the longest US metro commute." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, New York City: USD 195,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "USD 195,000 / 1 = USD 195,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.7800, R5-rounded down to 0.78. REPLACES 1 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Stored 1.00 claimed parity with Seattle; NYC measures 0.78 of it. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE NYC tech-weighted median ~$255k. Google/Meta/Bloomberg/Datadog $250-380k; Citadel/Two Sigma/HRT $400-600k pull up. Sits between Seattle (1.0) and SF (1.39); finance-tech premium unique to NYC" }, "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 76100 USD, New York-Newark-Jersey City metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "labor_market.tech_salary_distribution": { "source": "https://www.bls.gov/oes/current/oes_35620.htm", "last_updated": "2025-05", "confidence": "rich", "note": "BLS OEWS May 2025, SOC 15-1252 Software Developers, annual wage percentiles (full-market occupation distribution; straight-time gross pay incl. incentive pay, excl. equity and nonproduction bonuses). Public domain; API pull 2026-07-19." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "tax.state_brackets_married": { "source": "NYS Dept of Taxation and Finance, IT-201-I (2025) New York State tax rate schedule https://www.tax.ny.gov/forms/html-instructions/2025/it/it201i-2025.htm", "last_updated": "2026-08-01", "confidence": "high", "note": "Married filing jointly / qualifying surviving spouse, tax year 2025: 17,150 / 23,600 / 27,900 / 161,550 / 323,200 / 2,155,350 / 5,000,000 / 25,000,000. Added 2026-08-01; couples had been run through the single ladder. Deliberately NOT a doubling: the low bands run about 2.00 to 2.02x, the 6% band is only 1.50x (323,200 against 215,400), and the top two thresholds are IDENTICAL for both statuses. Doubling would have been wrong." }, "non_cash.parental_leave_flat": { "source": "paidfamilyleave.ny.gov/2026: 2026 maximum weekly benefit USD 1,228.53 (67% of the NYSAWW of 1,833.63), effective 2026-01-01", "last_updated": "2026-08-06", "confidence": "high", "note": "USD 1,228.53/wk, 2026 maximum. 24 x 1,228.53 = USD 29,485 total." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." } }, "austin": { "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 74300 USD, Austin-Round Rock-Georgetown metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "tax.federal_brackets": { "source": "IRS 2026 OBBBA (same as Seattle/SF/NYC)", "last_updated": "2026-05-16", "confidence": "high", "note": "US cross-city parity" }, "tax.state_brackets": { "source": "TX has no state income tax", "last_updated": "2026-05-16", "confidence": "high", "note": "Empty array — validates new model's backward-compat path for flat-tax states" }, "tax.sales_tax_rate": { "source": "Texas Comptroller + City of Austin", "last_updated": "2026-05-16", "confidence": "high", "note": "Combined 8.25% = 6.25% state + 2% local cap (Austin city + Travis County)" }, "tax.property_tax_rate_pct": { "source": "Travis County Taxpayer Impact Statement FY2026 + AustinTexas.gov https://www.austintexas.gov/page/tax-rates", "last_updated": "2026-05-16", "confidence": "high", "note": "Combined effective ~1.9% after homestead exemption (City + Travis County + AISD + healthcare + ESD). Nominal ~2.07%. TX offsets no income tax with elevated property tax" }, "tax.standard_deduction_single": { "source": "IRS 2026 OBBBA", "last_updated": "2026-05-16", "confidence": "high" }, "tax.standard_deduction_married": { "source": "IRS 2026 OBBBA", "last_updated": "2026-05-16", "confidence": "high" }, "housing.median_rent_by_bedroom": { "source": "Rentometer / Zumper / RentCafe May 2026 https://www.rentometer.com/average-rents-in-austin-tx", "last_updated": "2026-05-16", "confidence": "medium", "note": "Austin rents corrected significantly from 2022 peak" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "housing.median_home_price_by_bedroom": { "source": "Redfin + Zillow Mar-Apr 2026 https://www.redfin.com/city/30818/TX/Austin/housing-market", "last_updated": "2026-05-16", "confidence": "medium", "note": "Austin median sale $530K; prices down 2-7% YoY from 2022 peak" }, "housing.utilities_monthly": { "source": "EnergySage Travis County + Austin Apartment Locators", "last_updated": "2026-05-16", "confidence": "high", "note": "Summer AC bills push annual avg higher" }, "housing.renters_insurance_monthly": { "source": "Insure.com / SafeButler Austin", "last_updated": "2026-05-16", "confidence": "high" }, "housing.internet_phone_monthly": { "source": "Compare Internet Hub Texas + Astound Mobile", "last_updated": "2026-05-16", "confidence": "high", "note": "TX internet $68.84/mo + single-line cell $60-70" }, "healthcare.premium_monthly_by_family_size": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "high", "note": "Austin near national avg; not adjusted up like SF/NYC" }, "healthcare.premium_age_multiplier": { "source": "ACA age-rating", "last_updated": "2026-05-16", "confidence": "high" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "healthcare.dental_vision_monthly": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "medium", "note": "The dental and vision PREMIUM, single share. Billed FLAT for US cities (app.js), unlike every other country, where this field is per adult and scales by adults. Reason: the 2026-07-17 recalibration declined to add a dental-by-family-size key and folded the family dental and ortho excess into annual_out_of_pocket_by_family_size instead (plan line 72), so scaling this line would charge that excess twice. Sharpened 2026-08-08: the earlier wording said only the family increment sits in the OOP column, but the plan puts a dental and vision out-of-pocket margin in EVERY row including row 1." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Travis County TX (FIPS 48453), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Texas caveat: from 2015 onward the state reports the prior year figures, so the 2022 row is 2021 fieldwork and this value is marginally understated after escalation. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,350." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Travis County TX (FIPS 48453), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Texas caveat: from 2015 onward the state reports the prior year figures, so the 2022 row is 2021 fieldwork and this value is marginally understated after escalation. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $1,070/mo and the ages 1-2 rate $1,140/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,050. Does NOT model free universal pre-K where it exists. PRIOR NOTE: Austin full-time preschool ~USD 1,050/mo, at parity with the Houston peer. Was 1,650. DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.after_school_monthly": { "source": "AISD Extended Day + Care.com", "last_updated": "2026-05-16", "confidence": "medium", "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." }, "childcare.subsidy_per_kid_annual": { "source": "IRS CDCC (excluded v1)", "last_updated": "2026-05-16", "confidence": "high", "note": "Tax credits out of scope" }, "childcare.income_tested_childcare": { "source": "https://wfscapitalarea.com/childcare/for-parents/", "last_updated": "2026-08-05", "confidence": "high", "note": "Entry ceiling is the statewide Texas 85% SMI line (TWC BCY 2026 income limits, $92,041 for a family of 4, effective 1 Oct 2025 per WD Letter 11-25, which gives boards No Local Flexibility on the table); Texas publishes no separate exit ceiling; a household under the line may qualify but the board runs a waitlist, and the board's own page still prints the superseded BCY25 figure of $87,731." }, "food.groceries_monthly_per_adult": { "source": "Numbeo Austin + BLS CEX South", "last_updated": "2026-05-16", "confidence": "high" }, "food.groceries_monthly_per_kid": { "source": "BLS CEX + USDA", "last_updated": "2026-05-16", "confidence": "medium" }, "food.dining_out_monthly_baseline": { "source": "Numbeo Austin", "last_updated": "2026-05-16", "confidence": "high", "note": "Mid-range $25/pp — cheaper than coastal" }, "transit.monthly_pass": { "source": "CapMetro https://www.capmetro.org/fares", "last_updated": "2026-05-16", "confidence": "high", "note": "Local pass $41; commuter premium $96" }, "transit.annual_car_cost_avg": { "source": "AAA Driving Costs Austin-adjusted", "last_updated": "2026-05-16", "confidence": "medium", "note": "Lower than coastal: cheaper insurance + gas. Austin is car-dependent" }, "discretionary.entertainment_monthly_baseline": { "source": "Numbeo Austin", "last_updated": "2026-05-16", "confidence": "medium", "note": "Music/festival scene (SXSW/ACL) pushes leisure above median" }, "discretionary.local_vacation_annual_baseline": { "source": "TripAdvisor + Airbnb TX regional", "last_updated": "2026-05-16", "confidence": "medium", "note": "Hill Country / Marfa / Big Bend / Gulf Coast" }, "travel_home.avg_round_trip_per_person_to_other_city": { "source": "Google Flights AUS-SEA", "last_updated": "2026-05-16", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Federal CTC only; TX no state credit", "last_updated": "2026-05-16", "confidence": "high", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "non_cash.public_holidays": { "source": "US OPM", "last_updated": "2026-05-16", "confidence": "high" }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: Texas has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. Texas law only enables voluntary paid-leave insurance products; no entitlement. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Austin-Round Rock-San Marcos TX Metro Area (31000US12420). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US12420", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 28.2. Survey self-report, MSA geography. Austin city proper is 24.2, which is the wrong geography for this field." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Austin: USD 188,400, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "USD 188,400 / 1 = USD 188,400; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.7536, R5-rounded down to 0.75. REPLACES 0.95 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE Austin median ~$200-230k (Apple/Google/Meta/Oracle/Tesla). 'Texas band' discount ~5-10% below coastal; remote-first companies pay coastal rates. No state income tax partially offsets. Tech-weighted ~0.95 vs Seattle baseline" }, "labor_market.tech_salary_distribution": { "source": "https://www.bls.gov/oes/current/oes_12420.htm", "last_updated": "2025-05", "confidence": "rich", "note": "BLS OEWS May 2025, SOC 15-1252 Software Developers, annual wage percentiles (full-market occupation distribution; straight-time gross pay incl. incentive pay, excl. equity and nonproduction bonuses). Public domain; API pull 2026-07-19." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." } }, "boston": { "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 84600 USD, Boston-Cambridge-Newton metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "tax.federal_brackets": { "source": "IRS 2026 OBBBA (US parity)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.state_brackets": { "source": "MA DOR, Massachusetts tax rates https://www.mass.gov/info-details/massachusetts-tax-rates + 2025 Form 1 instructions https://www.mass.gov/doc/2025-form-1-instructions/download + 2025 Schedule 4% Surtax https://www.mass.gov/doc/2025-schedule-4-surtax/download", "last_updated": "2026-08-01", "confidence": "high", "married_variant": "none", "note": "VERIFIED 2026-08-01: Massachusetts has NO married-filing-jointly rate schedule. The rate is a flat 5% on wage income for every filing status, and the 4% Art. XLIV surtax threshold is one number per RETURN rather than a doubled one, with the 2025 Schedule 4% Surtax pre-printing \"Surtax threshold. Enter $1,083,150 in line 5\" and offering no filing-status branch. Couples also cannot split to collect two thresholds, because Ch. 50 of the Acts of 2023 requires a joint Massachusetts return whenever the federal return is joint. Filing status reaches only the exemption ($4,400 / $8,800) and No Tax Status. So the engine falling back to this array for married filers is CORRECT, and adding a widened married array would introduce an error. Separately noted: 1,083,150 is the TY2025 surtax threshold and TY2026 is 1,107,750, left as-is deliberately since it is the conservative direction and has no effect at our income range." }, "tax.state_payroll_premium_rate": { "source": "Massachusetts PFML 2026 contribution rates https://www.mass.gov/info-details/family-and-medical-leave-contribution-rates-for-employers", "last_updated": "2026-07-31", "confidence": "high", "note": "Employee share of the MA Paid Family & Medical Leave contribution, 0.46% of eligible wages for 2026" }, "tax.state_payroll_premium_cap": { "source": "MA PFML / SSA 2026 contribution and benefit base", "last_updated": "2026-07-31", "confidence": "high", "note": "MA PFML contributions are capped by the Social Security wage limit ($184,500 in 2026)" }, "tax.sales_tax_rate": { "source": "MA Dept of Revenue", "last_updated": "2026-05-16", "confidence": "high", "note": "MA 6.25% state, no local in Boston" }, "tax.property_tax_rate_pct": { "source": "Boston Assessing Dept FY2026", "last_updated": "2026-05-16", "confidence": "high", "note": "Class 1 residential ~$10.42/$1000 ≈ 1.04% effective" }, "housing.median_rent_by_bedroom": { "source": "Zillow/RentCafe/Apartments.com Boston 2026", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "housing.median_home_price_by_bedroom": { "source": "Redfin Boston Mar 2026 (median $860K)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Citywide median by property type, breakdown by bedroom estimated" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare.premium_monthly_by_family_size": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "high", "note": "Adjusted +10% over national for academic medical centers" }, "non_cash.parental_leave_paid_pct": { "source": "MA DFML, benefit year 2026: https://www.mass.gov/info-details/how-pfml-weekly-benefit-amounts-are-calculated-andor-changed", "last_updated": "2026-08-06", "confidence": "high", "note": "0 percentage-paid weeks: MA PFML pays 80% of the AWW slice at or below 50% of the state AWW (USD 961.24 in 2026) plus 50% of the slice above, capped at USD 1,230.39/wk, and the cap binds from about USD 98,000 of salary, below every salary we price, so the whole entitlement lives in non_cash.parental_leave_flat (the Denmark/NZ shape). No salary exists at which a single statutory percentage describes MA pay, so any stored headline would be fabricated. At USD 150,000 the effective rate is about 43%. Re-verified against correction flag:01784890637238-10ad763b, which expected 100: REFUTED, nothing in MA law pays 100%." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Boston: USD 168,800, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "USD 168,800 / 1 = USD 168,800; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.6752, R5-rounded down to 0.67. REPLACES 1.05 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Stored 1.05 claimed Boston beats Seattle; it measures 0.68 of it. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Boston tech ($220-260k) + biotech (Vertex/Moderna) ML lifts senior comp ~5% above Seattle" }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Boston + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Boston", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Boston", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "IRS Child Tax Credit + state EITC", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "labor_market.tech_salary_distribution": { "source": "https://www.bls.gov/oes/current/oes_14460.htm", "last_updated": "2025-05", "confidence": "rich", "note": "BLS OEWS May 2025, SOC 15-1252 Software Developers, annual wage percentiles (full-market occupation distribution; straight-time gross pay incl. incentive pay, excl. equity and nonproduction bonuses). Public domain; API pull 2026-07-19." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Suffolk County MA (FIPS 25025), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Massachusetts caveat: the state age bands are entered 2-3 months shifted (state infant 0-15mo entered as 0-18, preschool 34mo-5yr entered as 36-60). Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $1,820/mo and the ages 1-2 rate $2,690/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $2,200. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.income_tested_childcare": { "source": "https://www.mass.gov/child-care-financial-assistance", "last_updated": "2026-08-05", "confidence": "high", "note": "Ceiling at 85% SMI for both initial and continuing eligibility since 1 Jan 2026 (EEC policy CCFA-26-01; $134,028 for a family of 4 from the FY2025 EEC income chart, the newest retrievable agency print); a household under it may qualify, subject to the EEC waitlist." }, "non_cash.parental_leave_weeks": { "source": "M.G.L. c.175M ss.2-3 https://malegislature.gov/Laws/GeneralLaws/PartI/TitleXXII/Chapter175M/Section2 ; MA DFML https://www.mass.gov/info-details/how-pfml-weekly-benefit-amounts-are-calculated-andor-changed", "last_updated": "2026-08-06", "confidence": "high", "note": "24 = 12 weeks of MA PFML family (bonding) leave per covered parent, each parent's own non-transferable entitlement, summed across a two-earner household, which is why the both-parents flag is set. The birthing parent's medical-recovery leave (up to 20 wk medical, 26 wk combined) is medically certified rather than a fixed statutory number, so it is EXCLUDED, conservative. FMLA's 12 unpaid job-protected weeks are excluded by definition. Was 12/80: the unpaid FMLA window wearing the 80% lower-tier rate as a headline. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_flat": { "source": "MA DFML: 2026 maximum weekly benefit USD 1,230.39, 64% of the SAWW of 1,922.48; https://www.mass.gov/info-details/how-pfml-weekly-benefit-amounts-are-calculated-andor-changed", "last_updated": "2026-08-06", "confidence": "high", "note": "USD 1,230.39/wk, the 2026 PFML maximum (0.64 x 1,922.48 = 1,230.39 exactly). 24 x 1,230.39 = USD 29,529 total. mass.gov 403d on the 2026-08-06 re-check; the figure rests on this repo's 2026-08-01 direct mass.gov read, the statute's forced arithmetic, and four convergent secondary confirmations (Seyfarth, Prudential, HR Works, MyBenefitAdvisor)." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Boston-Cambridge-Newton MA-NH Metro Area (31000US14460). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US14460", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 31.9. Survey self-report, MSA geography." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Suffolk County MA (FIPS 25025), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Massachusetts caveat: the state age bands are entered 2-3 months shifted (state infant 0-15mo entered as 0-18, preschool 34mo-5yr entered as 36-60). Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $2,800." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "los-angeles": { "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 65000 USD, Los Angeles-Long Beach-Anaheim metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "tax.federal_brackets": { "source": "IRS 2026 OBBBA (US parity)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.state_brackets": { "source": "CA FTB, 2025 California Tax Rate Schedules (Form 540 booklet) https://www.ftb.ca.gov/forms/2025/2025-540-tax-rate-schedules.pdf", "last_updated": "2026-08-01", "confidence": "high", "note": "Schedule X, single or married/RDP filing separately, tax year 2025: 1% to 11,079 then 2/4/6/8% to 72,724, 9.3% to 371,479, 10.3% to 445,771, 11.3% to 742,953, 12.3% above. Refreshed 2026-08-01: the file carried the TY2024 thresholds (10,756 / 25,499 / ... / 721,314) while already citing the TY2025 schedule, so the data disagreed with its own citation. TY2025 is FTB's newest published year; the TY2026 schedules are not out. Corrected 2026-08-06: this note used to say the 1% Behavioral Health Services Tax above $1M taxable (renamed from the Mental Health Services Tax for TY2025 by Prop 1) was absent from the engine. It is not absent. app.js computes it as taxes.state_surtax from tax.state_surtax_rate 0.01 and tax.state_surtax_threshold 1000000, both of which this city carries, and a single filer on $1,500,000 gross pays $4,839 of it (identical in all four California cities). The threshold is per return and does not double for joint filers, which is the statute rather than a gap: RTC 17043(c) disapplies section 17045, the income-splitting rule that doubles California's regular MFJ bands, so no schema field for tax.state_surtax_threshold_married exists in any city JSON. https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=RTC§ionNum=17043" }, "tax.state_disability_insurance_rate": { "source": "California EDD, SDI Contribution Rates and Benefit Amounts https://edd.ca.gov/en/disability/Contribution_Rates_and_Benefit_Amounts/", "last_updated": "2026-07-30", "confidence": "high", "note": "Employee State Disability Insurance / Paid Family Leave contribution, 1.3% for 2026 (up from 1.2% in 2025). EDD lists the taxable wage ceiling and the maximum contribution as N/A: SB 951 eliminated both effective 2026-01-01, so this is levied on every dollar of wages with no cap. Added 2026-07-30; before that the engine omitted it entirely, understating California payroll by 1.3% of gross. Statewide levy, identical across the CA metros." }, "tax.sales_tax_rate": { "source": "CDTFA LA County", "last_updated": "2026-05-16", "confidence": "high", "note": "LA combined 9.75% = 6% state + 1.25% county + 2.5% city/district" }, "tax.property_tax_rate_pct": { "source": "LA County Assessor", "last_updated": "2026-05-16", "confidence": "high", "note": "LA Prop 13 effective ~1.18% (similar to SF; varies by neighborhood)" }, "housing.median_rent_by_bedroom": { "source": "RentCafe LA 2026", "last_updated": "2026-05-16", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Census ACS 2024 1-Year, table B25063 gross-rent distribution (metro/CBSA), data.census.gov; P25/P75 interpolated from dollar buckets, interpolated P50 matched published B25064 median (research 2026-07-17 §2). Rounded outward to 2dp per conservative-wider rule.", "last_updated": "2026-08-07", "confidence": "high", "note": "All-unit metro basis (pools bedroom sizes), wider than a within-bedroom band by construction. api.census.gov now requires an API key for queries (checked 2026-08-07); table remains on data.census.gov. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "housing.median_home_price_by_bedroom": { "source": "Redfin LA city Mar 2026 (median $1.0M)", "last_updated": "2026-05-16", "confidence": "medium" }, "transit.monthly_pass": { "source": "LA Metro fares + capping", "last_updated": "2026-05-16", "confidence": "high", "note": "LA Metro fare-capped ~$77/wk × 4.3; historical 30-day $100 reference used" }, "non_cash.parental_leave_paid_pct": { "source": "EDD: SB 951 rates for claims from 2025 (90% low-earner band, 70% above USD 20,931.31 highest-quarter earnings); https://edd.ca.gov/en/disability/Calculating_PFL_Benefit_Payment_Amounts/", "last_updated": "2026-08-06", "confidence": "high", "note": "0 percentage-paid weeks: for claims from 2025 under SB 951, CA pays 90% for highest-quarter earnings of 722.50-16,279.90, a flat 1,127/wk transition band to 20,931.30, and 70% above 20,931.31 per quarter, capped at USD 1,765/wk in 2026 (1,681 in 2025), binding from about USD 131,000. CAVEAT recorded: below that threshold CA pays a true 70% for the ICP band, so the flat store overstates for earners under about 131k (the engine clamps at 100%); at the salaries we price it is exact and less favorable than any headline. At USD 150,000 the effective rate is about 61%." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Los Angeles: USD 178,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "USD 178,000 / 1 = USD 178,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.7120, R5-rounded down to 0.71. REPLACES 1.1 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Stored 1.10 claimed LA BEATS Seattle; it measures 0.71 of it. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE $220-280k. Entertainment-tech premium lifts above national tech avg" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare._summary": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "medium", "note": "Group-level citation." }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Los-Angeles + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Los-Angeles", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Los-Angeles", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "IRS Child Tax Credit + state EITC", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "labor_market.tech_salary_distribution": { "source": "https://www.bls.gov/oes/current/oes_31080.htm", "last_updated": "2025-05", "confidence": "rich", "note": "BLS OEWS May 2025, SOC 15-1252 Software Developers, annual wage percentiles (full-market occupation distribution; straight-time gross pay incl. incentive pay, excl. equity and nonproduction bonuses). Public domain; API pull 2026-07-19." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "tax.state_brackets_married": { "source": "CA FTB, 2025 California Tax Rate Schedules (Form 540 booklet) https://www.ftb.ca.gov/forms/2025/2025-540-tax-rate-schedules.pdf", "last_updated": "2026-08-01", "confidence": "high", "note": "Schedule Y, married/RDP filing jointly or qualifying surviving spouse, tax year 2025. Verified band by band as EXACTLY 2x Schedule X on all 8 finite thresholds, which is what RTC 17041(b) prescribes; the base-tax columns double too. Added 2026-08-01. Before this, couples were run through the single-filer ladder, overstating California tax by a flat $3,457.64 for any household above about $141k of taxable income, and by $4,029 at a $450k household. Head of household is NOT 2x; no schema field for tax.state_brackets_hoh exists." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Los Angeles County CA (FIPS 6037), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $1,170/mo and the ages 1-2 rate $1,310/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $2,000. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.income_tested_childcare": { "source": "https://www.cdss.ca.gov/Portals/9/Additional-Resources/Letters-and-Notices/CCBs/2026/CCB_26-17.pdf", "last_updated": "2026-08-05", "confidence": "high", "note": "Eligibility-only (CC2 shape): CDSS ceiling at 85% SMI, $115,637 for a family of four (CCB 26-17, FY 2026-27; same threshold initial and ongoing). Tested on ADJUSTED monthly income per 5 CCR 18078(a), so a gross comparison is approximate; CCDF voucher minus copay, never a percentage. May qualify, never a promise of a spot. Revise each July." }, "non_cash.parental_leave_weeks": { "source": "EDD Paid Family Leave https://edd.ca.gov/en/disability/paid-family-leave/ ; https://edd.ca.gov/en/disability/Calculating_PFL_Benefit_Payment_Amounts/", "last_updated": "2026-08-06", "confidence": "high", "note": "16 = 8 weeks of CA PFL bonding per parent within a 12-month period, each parent's own SDI-funded non-transferable entitlement, summed (both-parents flag set). The birthing parent's SDI pregnancy disability (usually 4 weeks before plus 6 after, 8 for cesarean) is medically certified, EXCLUDED, conservative. Was 12/80: the unpaid FMLA window with a fabricated headline. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_flat": { "source": "EDD Contribution Rates and Benefit Amounts: 2026 maximum weekly benefit USD 1,765 https://edd.ca.gov/en/disability/Contribution_Rates_and_Benefit_Amounts/", "last_updated": "2026-08-06", "confidence": "high", "note": "USD 1,765/wk, 2026 maximum. 16 x 1,765 = USD 28,240 total." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Los Angeles-Long Beach-Anaheim CA Metro Area (31000US31080). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US31080", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 30.7. Survey self-report, MSA geography." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Los Angeles County CA (FIPS 6037), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $2,250." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "chicago": { "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 69700 USD, Chicago-Naperville-Elgin metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "tax.federal_brackets": { "source": "IRS 2026 OBBBA (US parity)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.state_brackets": { "source": "Illinois Dept of Revenue (flat 4.95%)", "last_updated": "2026-05-16", "confidence": "high", "note": "IL flat 4.95% — single-bracket array used to exercise the progressive path uniformly" }, "tax.city_income_tax_rate": { "source": "Chicago Dept of Finance", "last_updated": "2026-05-16", "confidence": "high", "note": "Chicago has NO personal income tax (despite reputation; only business taxes are local)" }, "tax.sales_tax_rate": { "source": "IL Dept of Revenue + City of Chicago", "last_updated": "2026-05-16", "confidence": "high", "note": "Chicago combined 10.25% = 6.25% state + 1.75% county + 1.25% city + 1% RTA — among highest in US" }, "tax.property_tax_rate_pct": { "source": "Cook County Assessor / WOWA", "last_updated": "2026-05-16", "confidence": "medium", "note": "Cook County residential effective ~2.27%; among highest in US. Wide variance across Chicago wards" }, "housing.median_rent_by_bedroom": { "source": "Zillow Chicago 2025", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Census ACS 2024 1-Year, table B25063 gross-rent distribution (metro/CBSA), data.census.gov; P25/P75 interpolated from dollar buckets, interpolated P50 matched published B25064 median (research 2026-07-17 §2). Rounded outward to 2dp per conservative-wider rule.", "last_updated": "2026-08-07", "confidence": "high", "note": "All-unit metro basis (pools bedroom sizes), wider than a within-bedroom band by construction. api.census.gov now requires an API key for queries (checked 2026-08-07); table remains on data.census.gov. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "housing.median_home_price_by_bedroom": { "source": "Zillow Chicago median ~$295K", "last_updated": "2026-05-16", "confidence": "high", "note": "Chicago among most affordable major US cities" }, "transit.monthly_pass": { "source": "CTA Ventra 30-day pass", "last_updated": "2026-05-16", "confidence": "high" }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Chicago: USD 150,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "USD 150,000 / 1 = USD 150,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.6000, R5-rounded down to 0.6. REPLACES 0.97 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE $200-240k. Mix of finance (Citadel), tech regional offices (Google/Meta), and HQs (Boeing/McDonalds/Hyatt). ~3% below Seattle" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare._summary": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "medium", "note": "Group-level citation." }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Chicago + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Chicago", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Chicago", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "IRS Child Tax Credit + state EITC", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "labor_market.tech_salary_distribution": { "source": "https://www.bls.gov/oes/current/oes_16980.htm", "last_updated": "2025-05", "confidence": "rich", "note": "BLS OEWS May 2025, SOC 15-1252 Software Developers, annual wage percentiles (full-market occupation distribution; straight-time gross pay incl. incentive pay, excl. equity and nonproduction bonuses). Public domain; API pull 2026-07-19." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Cook County IL (FIPS 17031), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Illinois caveat: the state calculates weekly rates on a 35-HOUR week, shorter than the full-time week the other cities are priced on, and reports per-slot rather than per-provider, over-weighting large providers. Both push Chicago low. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $980/mo and the ages 1-2 rate $1,300/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,650. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.income_tested_childcare": { "source": "https://www.dhs.state.il.us/page.aspx?item=173299", "last_updated": "2026-08-05", "confidence": "high", "note": "Entry at 225% FPL ($72,336 for a family of 4, IDHS policy 01.02.01 effective 1 Jul 2025) with redetermination at 275% FPL ($88,416 exit) and a graduated phase-out up to 85% SMI; a household under the entry line may qualify." }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: Illinois has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. The IL Paid Leave for All Workers Act and the Chicago and Cook County ordinances are any-purpose paid time off (40 hrs/yr), not parental leave, EXCLUDED. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Chicago-Naperville-Elgin IL-IN Metro Area (31000US16980). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US16980", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 30.9. The previous 33 tracked Chicago CITY (33.5), which is the wrong geography for a metro-based field." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Cook County IL (FIPS 17031), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Illinois caveat: the state calculates weekly rates on a 35-HOUR week, shorter than the full-time week the other cities are priced on, and reports per-slot rather than per-provider, over-weighting large providers. Both push Chicago low. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $2,000." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "munich": { "housing.median_rent_by_bedroom": { "source": "Landeshauptstadt Muenchen, Wohnungsmarktbarometer 2026 HJ1 (asking rents, \"je nettokalt\" printed, ImmoScout24 listings, city area) https://stadt.muenchen.de/dam/jcr:37d1fda8-9941-4579-8974-5a69b20dd3b5/Wohnungsmarktbarometer_2026HJ1.pdf ; Landeshauptstadt Muenchen, Wohnungsmarktdaten Muenchen 2025 (Referat fuer Stadtplanung und Bauordnung, Stand Mai 2026) https://stadt.muenchen.de/dam/jcr:0e48bdf2-0db1-435f-b549-a04f8250fff6/LHM_RFS_Wohnungsmarktflyer2025_web.pdf ; per-room gradient VALUE Marktdatenbank 2026 https://mietspiegeltabelle.de/mietspiegel-muenchen/ ; official Mietspiegel fuer Muenchen 2025 (stock-rent contrast only) https://stadt.muenchen.de/infos/mietspiegel.html", "last_updated": "2026-08-07", "confidence": "high", "note": "CORRECTED 2026-08-02 from 1100/1500/2300/3500, replacing a Numbeo group citation. The defect was SIZE, not the view that Munich is expensive: at Munich real rate the old 3BR of 3500 implies a 158 sqm flat (3500 divided by the 22.17 EUR per sqm 4-Zimmer rate), which is a penthouse and not a family flat, and the error compounded with bedroom count so it was worst exactly where the product cares most. The city own Wohnungsmarktbarometer 2026 HJ1 puts an unfurnished re-let at 22.93 EUR/sqm netto kalt and a new build at 26.20; VALUE 2026 independently gives 23.60, a 2.9% gap in the expected direction, and the city two segments bracket the IBB cross-city anchor of 22.62. Per-room rates 26.84/22.91/21.87/22.17 x 1.0587 = 28.41/24.25/23.15/23.47. Studio 1100 to 1190, 1BR 1500 to 1460, 2BR 2300 to 1970, 3BR 3500 to 2460. Stock contrast: the official Mietspiegel 2025 is 15.38, so asking runs about 49% above stock and shipping the Mietspiegel would understate a new lease by a third. Furnished stock, which many arrivals actually sign first, runs about 30% higher again at 29.80 and is deliberately excluded because it bundles goods we price separately. Per-room rates come from the VALUE Marktdatenbank 2026 cut published at mietspiegeltabelle.de, explicitly nettokalt and explicitly market/asking rather than the qualified municipal Mietspiegel. That is the same database family the Berlin correction leant on. Leipzig proves what it serves: its 10.45 equals VALUE own published MEAN of 10.49 against a median of 10.00, so this basis runs about 5% above median, which is the conservative direction. Method is the shipped Berlin recipe re-run per city, so all eight German cities sit on one footing: rent = size x city per-room EUR/sqm x 1.0587. The 1.0587 is Berlin own uplift over this same source (Berlin shipped a 16.6 base against 15.68 here), and re-running it reproduces Berlin 760/970/1360/1750 to the euro. Sizes are held constant across cities at 42/60/85/105 sqm on the German Zimmer convention (our 1BR = 2-Zimmer-Wohnung), because the product prices an equivalent standard of living rather than how much space local households ration themselves into at local prices. Duesseldorf publishes real offered-flat sizes of 37/58/83/111, so the fixed 42 makes the studio the most conservative cell and the fixed 105 makes the 3BR mildly generous on size; the two roughly cancel. Basis is Angebotsmiete (new lease), netto kalt, never the Mietspiegel stock rent. Rents exclude utilities, insurance and internet, which are separate fields. Values are a CEILING per the conservative value-bias rule, not central estimates. Working: docs/superpowers/research/2026-08-02-german-city-rents.md SIZE LADDER CORRECTED 2026-08-07 (small-fixes wave 1, coordinated eight-city pass incl. Berlin): the common dwelling-size ladder moves from the inherited 42/60/85/105 sqm to 37/58/83/111 sqm, Duesseldorf's PUBLISHED offered-flat sizes (Wohnungsmarktbericht 2025 Abb. 54, MFH column, n=10,370 listings; the only published per-Zimmer size table in the German set, and the conservative top of the seven-city evidence range at 3-Zi 83 and 4-Zi 111). Root cause of the old ladder: German official statistics count a closed kitchen over 6 sqm as a room, so census room distributions ran one room high against listing Zimmer counts; the 42 sqm studio was priced at the 1-Zimmer premium rate measured on 33-37 sqm stock (double-counting the small-flat premium) and the 4-Zimmer was undersized at 105 vs the real ~111. Values 1190/1460/1970/2460 -> 1050/1410/1920/2610. Per-room rates and the 1.0587 uplift are UNCHANGED; sizes remain COMMON across all eight German cities because the product prices an equivalent standard of living. Working: docs/superpowers/research/2026-08-07-small-fixes.md" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "non_cash.parental_leave_paid_pct": { "source": "SGB V para 24i (Mutterschaftsgeld) https://www.gesetze-im-internet.de/sgb_5/__24i.html ; MuSchG para 20 (Zuschuss zum Mutterschaftsgeld, employer tops up to full net) https://www.gesetze-im-internet.de/muschg_2018/__20.html", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the percentage-paid remainder of the 67 weeks: the 14.1 Mutterschutz weeks (6 pre-birth + 8 post-birth), where para 24i SGB V pays EUR 13/day and para 20 MuSchG obliges the employer to top up to full net pay with no ceiling. Was 72, a blend that averaged the capped Elterngeld slice into this field; since 2026-08-06 the 52.87 Elterngeld weeks live in non_cash.parental_leave_capped (65% up to EUR 1,800/month, nothing above EUR 175,000 household income) and parentalLeaveEffectiveRate() blends the slices at the user's salary, so every reader-facing surface already prices the cap and the cliff." }, "non_cash.parental_leave_capped": { "source": "BEEG para 2 (Elterngeld: 65% of prior income, minimum EUR 300 and maximum EUR 1,800 per month) https://www.gesetze-im-internet.de/beeg/__2.html ; BEEG para 1 Abs. 8 (no Elterngeld above EUR 175,000 combined zu versteuerndes Einkommen) https://www.gesetze-im-internet.de/beeg/__1.html ; MuSchG para 20 (the 14 Mutterschutz weeks stay OUTSIDE this block: the employer net top-up has no ceiling) https://www.gesetze-im-internet.de/muschg_2018/__20.html", "last_updated": "2026-08-06", "confidence": "high", "note": "52.87 of the 67 weeks: the Elterngeld pot (60.87 wk) minus the 8 overlap weeks counted at the Mutterschutz rate. The EUR 1,800/month cap (unindexed since Elterngeld's 2007 introduction, re-confirmed 2026-08-06) binds above ~EUR 33,000 gross, i.e. at every ICP salary: at EUR 90,000 this slice pays ~24% of gross and the blended Berlin rate is ~40%; at EUR 150,000 the slice pays ~14% and the blend is ~32%, versus the 72 this field carried before the 2026-08-06 caps rebuild. BEEG para 1 Abs. 8 zeroes the slice for households above EUR 175,000: the engine compares HOUSEHOLD GROSS against the ceiling as a conservative proxy for zu versteuerndes Einkommen (gross exceeds taxable income, so the cliff bites earlier, the direction less favorable to Germany). Same conservative direction on the rate basis: the statute pays 65% of Elterngeld-Netto while the engine applies 65% to gross, and above ~EUR 33,000 gross the cap binds under both readings, so the distinction is immaterial at any ICP salary. Floor EUR 300/month (BEEG para 2). Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Germany section) and 2026-08-06-caps-engine-build.md." }, "non_cash.parental_leave_weeks": { "source": "MuSchG para 3 (Schutzfristen, 6 weeks before and 8 after) https://www.gesetze-im-internet.de/muschg_2018/__3.html ; BEEG para 4 Abs. 3 (12 Monatsbetraege Basiselterngeld plus 2 Partnermonate, gemeinsam) and para 4 Abs. 4 Satz 3 https://www.gesetze-im-internet.de/beeg/__4.html", "last_updated": "2026-08-01", "confidence": "high", "note": "67 wk = 6 additive pre-birth Mutterschutz weeks + the 14-month Elterngeld pot (60.87 wk). Was 61, the Elterngeld pot alone with Mutterschutz omitted. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. OVERLAP RULE APPLIED: the 8 post-birth Mutterschutz weeks are NOT stacked on top, because BEEG para 4 Abs. 4 Satz 3 deems Lebensmonate carrying Mutterschaftsgeld to be consumed Basiselterngeld months (gelten als Monate, fuer die dieser Elternteil Basiselterngeld bezieht). They are counted once, at the higher rate. Germany therefore reaches 67 and not 75. Elterngeld is a SHARED household pot, so it is not doubled. GERMANY HAS NO STATUTORY PAID PATERNITY LEAVE: the Familienstartzeit bill transposing EU Directive 2019/1158 art. 4 was still not law on 2026-08-01, so zero paternity weeks. Excluded: ElterngeldPlus (an election that halves the monthly amount to double the months) and the Partnerschaftsbonus (conditional on both parents working reduced hours)." }, "tax.inbound_regime": { "source": "Germany has no inbound or expat income-tax regime; ordinary progressive tax applies. Verified 2026-07-31 against the 2025 CDU/CSU-SPD coalition agreement (zero occurrences of a skilled-worker tax rebate; the only measure is an administrative Work-and-stay-Agentur) https://www.cdu.de/app/uploads/2025/04/Koalitionsvertrag-%E2%80%93-barrierefreie-Version.pdf and against the Steueraenderungsgesetz 2025 in force 2026-01-01 https://www.bundesfinanzministerium.de/Content/DE/Pressemitteilungen/Finanzpolitik/2025/09/2025-09-10-steueraenderungsgesetz-2025.html", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: re-verified. The 2024 proposal for a declining 30/20/10% rebate for incoming skilled workers never reached a Gesetzentwurf and died with the traffic-light coalition in November 2024. Nothing of the kind appears in the 2025 coalition agreement or in the Steueraenderungsgesetz 2025. Recommend a standing December re-check against the BMF press-release index, because the idea is revived by German industry most years and would arrive via an annual Jahressteuergesetz." }, "tax.federal_tax_model": { "source": "Same DE tax model as Berlin (peer-consistent)", "last_updated": "2026-05-16", "confidence": "high", "note": "All federal fields (income_tax_tariff_32a, the para 9a/10c/10 base deductions, soli, social_insurance_rate/cap) identical to Berlin" }, "tax.property_tax_rate_pct": { "source": "Munich Grundsteuer Hebesatz 535% (2026, raised from 490% post-reform)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Effective on Grundsteuerwert ~0.17%. Higher than Berlin (470% Hebesatz, ~0.15%)" }, "childcare.infant_monthly": { "source": "Bavaria Kitabeiträge (means-tested municipal)", "last_updated": "2026-05-16", "confidence": "high", "note": "Bavaria charges Krippenbeiträge ~€200-300/mo (vs Berlin's free public Kita). Bayerischer Krippengeld €100/mo + Beitragszuschuss €100/mo from age 3" }, "non_cash.public_holidays": { "source": "Bavaria has 13 statutory holidays: most of any German state (incl. Catholic holidays)", "last_updated": "2026-05-16", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Munich: EUR 84,269, fetched live 2026-08-11 and recorded at sweep B F8d of docs/superpowers/research/2026-08-11-comp-index-sweep-b.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 84,269 / 0.867 = USD 97,196; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3888, R5-rounded down to 0.38. REPLACES 0.62 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Munich by 63%. The measured p50 was 0.627 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE €100-130k all-in (slightly above Berlin). Germany's highest-paying tech hub" }, "housing.utilities_monthly": { "source": "DMB LV Bayern Betriebskostenspiegel 2024 (avg 2.78 EUR/sqm/mo, Munich +5% = 2.92, https://www.mieterbund-bayern.org/2026/01/09/neuer-betriebskostenspiegel-fuer-bayern-2024/ + https://www.mein-nebenkostenrechner.de/nebenkosten/bayern/muenchen); SWM Grundversorgung from 2026-02-01 (32.44 ct/kWh gross, typical household ~81 EUR/mo, https://www.swm.de/strom/preisanpassung); Numbeo Munich 85-sqm mean 326.20 (range 240-507, fetched 2026-08-08)", "last_updated": "2026-08-08", "confidence": "medium-high", "note": "2BR (~83 sqm) reference incl. household electricity. Raised 250 -> 330: central evidence band 306-326 (2.92x83 + SWM electricity 64-81; Numbeo mean 326), taken at the top under the conservative-value rule since the avg-paid Betriebskosten basis skews low for Munich full-service stock (Germany's most expensive NK city; all-positions ceiling ~403). Derivation: docs/superpowers/research/2026-08-08-utilities-values.md" }, "housing._summary": { "source": "Numbeo Housing (cost-of-living basket) https://www.numbeo.com/cost-of-living/in/Munich", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation (90/10 quality bar). Per-field values were researched by subagents from local marketplaces (Zillow/Hemnet/Funda/Rightmove/etc.); cite each field individually at next refresh pass." }, "healthcare._summary": { "source": "GKV statutory (folded into social_insurance_rate)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Munich", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Munich", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "Familienkasse Kindergeld", "last_updated": "2026-05-16", "confidence": "medium" }, "tax.income_tax_tariff_32a": { "source": "§ 32a Abs. 1 und Abs. 5 EStG, Veranlagungszeitraum 2026 https://www.gesetze-im-internet.de/estg/__32a.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Statutory continuous tariff, identical nationwide (see berlin for the verbatim zone text and the coefficient cross-check). Replaces the 5-bracket approximation that flattened the 17,799-69,878 progression zone to 33.5%. Ehegattensplitting (Abs. 5) is now an explicit 2 x tariff(joint zvE / 2); doubling bracket edges only reproduced that for a piecewise-linear schedule." }, "tax.werbungskosten_pauschbetrag": { "source": "§ 9a Satz 1 Nr. 1 Buchst. a EStG https://www.gesetze-im-internet.de/estg/__9a.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Arbeitnehmer-Pauschbetrag EUR 1,230 per employed person. National, identical to berlin." }, "tax.sonderausgaben_pauschbetrag": { "source": "§ 10c EStG https://www.gesetze-im-internet.de/estg/__10c.html", "last_updated": "2026-07-31", "confidence": "high", "note": "EUR 36, doubled on Zusammenveranlagung. National, identical to berlin." }, "tax.vorsorge_deductible_rate_rv": { "source": "§ 10 Abs. 1 Nr. 2 und Abs. 3 EStG https://www.gesetze-im-internet.de/estg/__10.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Employee RV share 9.3%, 100% deductible since 2023, against the RV Beitragsbemessungsgrenze. National, identical to berlin." }, "tax.vorsorge_deductible_rate_kvpv": { "source": "§ 10 Abs. 1 Nr. 3 und Abs. 4 Satz 4 EStG https://www.gesetze-im-internet.de/estg/__10.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Basis KV 8.75% x 0.96 (Krankengeld haircut) + PV 1.8% = 10.2%, against the KV/PV Beitragsbemessungsgrenze; ALV not deductible. National, identical to berlin." }, "tax.solidaritaetszuschlag_threshold_married": { "source": "Soli Freigrenze 2026 joint assessment EUR 40,700 (2 x 20,350 single; tk.de Solidaritaetszuschlag); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.social_insurance_rate_rvalv": { "source": "RV 9.3% + ALV 1.3% employee share = 10.6%, runs to the higher Beitragsbemessungsgrenze (lohn-info.de Sozialversicherungsbeitraege 2026); verified tax audit 2026-07-14", "confidence": "high", "note": "Carved out of the combined 21.15% employee share; the KV/PV remainder (10.55%) stops at the lower ceiling.", "last_updated": "2026-07-17" }, "tax.social_insurance_cap_rvalv": { "source": "Beitragsbemessungsgrenze RV/ALV 2026 = EUR 101,400 (bundesregierung.de/breg-de/aktuelles/beitragsgemessungsgrenzen-2386514); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.pv_childless_surcharge_rate": { "source": "PV Kinderlosenzuschlag § 55 Abs. 3 SGB XI: +0.6pp employee-borne for childless members 23+ (dak.de Beitrag zur Pflegeversicherung Zuschlag Kinderlose; vdek.com Aenderungen Gesundheitswesen 2026); found by the 2026-08-06 corrections wave", "confidence": "high", "note": "Applied when the household has no children; the under-23 exemption is not modeled (ICP is older). Deductible with the rest of the Basis-Pflegeversicherung.", "last_updated": "2026-08-06" }, "childcare.subsidy_per_kid_annual": { "source": "Land and municipal Kitabeitrag schedules; double-subtraction correction 2026-08-02", "last_updated": "2026-08-02", "confidence": "high", "note": "Set to 0 on 2026-08-02, was EUR 2,400. The old value was the VALUE OF THE FREE OR SUBSIDISED KITA HOURS, but childcare.preschool_monthly already carries the post-free-tier out-of-pocket, so computeChildcareCost subtracted the same relief a second time and clamped the Munich household bill to zero. German Kita relief is a benefit in kind delivered as a lower or absent fee, never a payment to the parent, so nothing belongs in this field. Munich now bills EUR 200/month per Kindergarten place, net of Bavaria’s Beitragszuschuss. This field holds CASH THAT REACHES THE HOUSEHOLD. Where the monthly rate it nets against is already the price the household actually pays, the value is 0 and the relief is described here instead. See docs/superpowers/research/2026-08-02-childcare-double-subtraction.md." }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "German Kindergeld (EUR 259/mo per child = EUR 3,108/yr, which is the stored value): universal, NOT means-tested, and it does not taper at professional incomes. Corrected 2026-08-02 from a stale \"EUR 255/mo\", which is the 2025 rate; the Steuerfortentwicklungsgesetz stepped Kindergeld to EUR 255 on 2025-01-01 and EUR 259 on 2026-01-01. The stored value was already correct and was NOT changed." }, "transit.monthly_pass": { "source": "https://www.bundesregierung.de/breg-de/aktuelles/deutschlandticket-2134074", "last_updated": "2026-08-05", "confidence": "high", "note": "Deutschlandticket EUR 63/mo since 2026-01-01 (VMK decision 2025-09-18). Was 58 through Dec 2025." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Destatis Gesundheitsausgabenrechnung 2024 (private households + NPISH EUR 65bn) + Destatis EVS 2023 Konsumausgaben nach Haushaltstyp + GKV Zuzahlungen aggregate EUR 5.18bn (BMG via Aerzteblatt) + SGB V paras 61-62", "last_updated": "2026-08-08", "confidence": "medium", "note": "Level essentially CONFIRMED at EUR 311 per adult (778/capita gross, ~40% in scope after netting long-term care 46%, dental 9%, vision ~3pp). What Germany needed was a curve, not a level. The curve is steep in adults and nearly flat in children because under-18s are exempt from every Zuzahlung except transport AND each child subtracts EUR 9,756 from the income base before the 2%-of-gross Belastungsgrenze, so adding a child LOWERS the household's statutory ceiling in absolute euros. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Destatis Gesundheitsausgabenrechnung 2024 (private households + NPISH EUR 65bn) + Destatis EVS 2023 Konsumausgaben nach Haushaltstyp + GKV Zuzahlungen aggregate EUR 5.18bn (BMG via Aerzteblatt) + SGB V paras 61-62", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, urb_ctran indicator TT1019V, city DE003C. https://ec.europa.eu/eurostat/databrowser/view/urb_ctran/default/table?lang=en", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average time of journey to work (minutes)", "year": "2016", "note": "26.9 min, BASIS YEAR 2016. CITY geography, narrower than our metro unit, so biased LOW and flattering. Survey self-report, standard half-up rounding. METRO-LEVEL GERMAN SERIES: ESTABLISHED HARD NEGATIVE, do not re-run. MiD 2023 Tabelle 6 carries RegioStaR7 breakdowns of population, trips, distance and total time but NO duration by purpose; the national banner table W9 Wegedauer is all-purpose and sits on an \"Ohne regelmaessige berufliche Wege\" base; the small-area products regionalise headline indicators only; and Eurostat publishes no functional-urban-area rows for TT1019V. SrV 2023 was rejected because it publishes an ALL-PURPOSE trip duration, and swapping an all-purpose duration for a work-purpose one is a UNIT error, which is worse than the geography error we accepted. SrV is therefore used only as rank-order corroboration, and it agrees with Eurostat against our old stored spread: both put Frankfurt near the BOTTOM of the big five, where we had it second highest." }, "non_cash.sick_leave": { "source": "Entgeltfortzahlungsgesetz (EFZG) §3 (employer phase); SGB V §47 (Krankengeld; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "EStG s.10 Abs 3 https://www.gesetze-im-internet.de/estg/__10.html ; SVBezGrV 2026 s.4 Abs 1 https://www.gesetze-im-internet.de/svbezgrv_2026/BJNR1160A0025.html", "last_updated": "2026-08-11", "confidence": "high", "note": "Ruerup/Basisrente Hoechstbetrag, 2026: EUR 30,825.60 single (knappschaftliche RV BBG 124,800 x 24.7%). A DEDUCTION ceiling and a COMBINED one: employee AND employer statutory pension contributions consume the same pot. less_employer_pct 0.186 = 2 x 9.3% and employer_base_cap 101,400 is the allgemeine RV BBG, because computeGermanyTaxes already deducts the employee 9.3% via vorsorge; a naked max would hand a EUR 95k earner about EUR 8.8k of phantom deduction. Riester (EUR 2,100) and bAV Entgeltumwandlung (8% of the BBG, EUR 8,112) are deliberately NOT stacked. Married couples: the statute doubles the ceiling to 61,651.20 even on one income; our per-earner cap under-deducts there, which errs toward higher tax." } }, "toronto": { "tax.filing_basis": { "source": "PwC WWTS - Canada (Individual, Tax administration): married taxpayers file separately; joint returns are not allowed.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "CRA 2026 (peer-consistent with Vancouver)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.provincial_brackets": { "source": "Ontario 2026 brackets/BPA (taxtips.ca/taxrates/on.htm); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high", "note": "ON progressive 5.05% to 13.16% (top above CAD 220k). Excludes 20%/36% Ontario surtax on provincial tax above CAD 5,710/7,307 — small effect at tech-worker incomes" }, "tax.provincial_personal_allowance": { "source": "ON BPA 2026 ~CAD 12,747", "last_updated": "2026-05-16", "confidence": "high" }, "tax.sales_tax_rate": { "source": "HST 13% (5% GST + 8% ON PST)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Toronto residential 2025 ~0.6618%", "last_updated": "2026-05-16", "confidence": "high", "note": "Among lowest in Canada despite high property values" }, "housing.median_rent_by_bedroom": { "source": "Zumper Toronto / Rentals.ca Q1 2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Below Vancouver. 1BR ~CAD 2,400, 3BR ~CAD 3,500" }, "housing.rent_percentiles": { "source": "Calibrated within-listing default (docs/superpowers/research/2026-07-17-housing-range-research.md §4), Toronto-corroborated 2026-08-17 by a Kijiji listing-count CDF (city of Toronto, long-term apartments-condos, asking CAD/month, fixed bedroom count): 2BR apartment+condo pool 0.82/1.19 (n=2,296), 2BR raw pool 0.79/1.26 (n=7,930), 1BR raw 0.83/1.17 (n=2,477) - the default sits inside the measured bracket.", "last_updated": "2026-08-17", "confidence": "medium", "note": "REBASED 2026-08-17 (ledger cluster housing-0817) from the 0.86/1.36 cross-zone band, which failed the constant-ratio check on its own source: the Toronto Central/Etobicoke South spread is 1.07 (studio), 1.37 (1BR), 1.58 (2BR) on CMHC averages and 0.97/1.35/1.28 on medians, so the stored constant was the single most extreme cell in the grid, and the occupied-purpose-built-to-asking wedge is zone-heterogeneous and direction-flipping (liv.rent asking 1BR Downtown/Etobicoke = 0.89 vs CMHC occupied 1.37 - the NYC/B25031 failure mode). ALSO CORRECTED: the old construction's inputs 1,762/2,045/2,776 were the Oct 2025 zone 2BR AVERAGE cells mislabeled as medians (median cells: 1,958/2,000/2,500; the 2,776 numerator is a b-code reliability cell, and 3BR+ is suppressed at Etobicoke South in both statistics). CMHC publishes NO rent distribution or quartiles for Toronto (HMIP Primary Rental Market menu verified 2026-08-17: average/median/vacancy/availability/universe/summary only), so no genuine local quartile source exists; the calibrated default is retained with direct city corroboration, and where it deviates from the universe-matched measurement (apartment+condo 0.82/1.19, the universe of the stored Rentals.ca-tracking medians) it errs WIDE, the safe direction. Kijiji pool medians (2BR 2,309 raw / 2,128 tagged) vs stored 2,950 is a LEVEL observation, out of scope, recorded not adjudicated (condo-vs-purpose-built universe gap). Band edge = median_rent_by_bedroom x multiplier, computed at display time. Derivation + raw CDFs: docs/superpowers/research/2026-08-17-housing-band-fixes.md (T1); scoping evidence: docs/superpowers/research/2026-08-16-housing-band-scope-toronto.md" }, "transit.monthly_pass": { "source": "TTC Toronto Transit Commission", "last_updated": "2026-05-16", "confidence": "high", "note": "TTC monthly pass CAD 156. Subway + streetcar + bus" }, "non_cash.public_holidays": { "source": "ON Employment Standards Act", "last_updated": "2026-05-16", "confidence": "high", "note": "9 statutory; includes Family Day (3rd Monday Feb)" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Toronto: CAD 142,630, held in labor_market.tech_salary_distribution (source_slug greater-toronto-area, as_of 2026-07-18, table confidence rich). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "CAD 142,630 / 1.4 = USD 101,879; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.4075, R5-rounded down to 0.4. REPLACES 0.75 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Toronto by 88%. The measured p50 was 0.543 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE CAD 200-250k. Above Vancouver (0.73), Canada's largest tech market" }, "healthcare._summary": { "source": "Provincial Medicare (universal)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation." }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Toronto + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Toronto", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Toronto", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "CRA Canada Child Benefit", "last_updated": "2026-05-16", "confidence": "medium" }, "tax.surtax_threshold_1": { "source": "Ontario surtax 20% of basic ON tax over 5,818 (2026); https://www.taxtips.ca/taxrates/on.htm; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "First Ontario surtax threshold (20%)." }, "tax.surtax_threshold_2": { "source": "Ontario surtax +36% of basic ON tax over 7,446 (2026); https://www.taxtips.ca/taxrates/on.htm; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "Second Ontario surtax threshold (36%)." }, "tax.health_premium": { "source": "Ontario Health Premium, stepped income-tested levy capped at 900; https://www.ontario.ca/page/health-premium; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "Gates the stepped OHP (600/750/900 plateaus) in computeCanadaTaxes." }, "tax.bpa_phaseout": { "source": "Federal BPA supplement phases 16,452 -> 14,829 over net income 181,440-258,482 (CRA 2026); https://www.canada.ca/en/revenue-agency/programs/about-canada-revenue-agency-cra/federal-government-budgets/basic-personal-amount.html; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "BPA modeled as a non-refundable credit at the lowest bracket rate with this federal supplement phase-out." }, "govt_benefits.child_allowance_per_kid_annual": { "source": "canada.ca Canada Child Benefit calculation sheet (July 2025-June 2026)", "last_updated": "2026-07-17", "note": "CCB is means-tested on adjusted family net income: a 2-child mixed-age family reaches $0 by ~CAD 236k AFNI. The published CAD 231k-339k households receive ~$0 (residual <$150/kid only at the very low end), and Ontario/Alberta have no high-income floor. Set to 0 (was a flat 3,000)." }, "childcare.subsidy_per_kid_annual": { "source": "CWELCC (Canada-Wide Early Learning and Child Care) provincial fee schedules; double-subtraction correction 2026-08-02", "last_updated": "2026-08-02", "confidence": "medium", "note": "Set to 0 on 2026-08-02, was CAD 3,600. The old value was the CWELCC fee reduction. CWELCC is delivered as a lower posted fee at a participating centre, not as cash to the household, and childcare.preschool_monthly already carries the reduced rate, so the reduction was netted twice. The modelled CAD 580/month already sits at Ontario’s CWELCC cap. This field holds CASH THAT REACHES THE HOUSEHOLD. Where the monthly rate it nets against is already the price the household actually pays, the value is 0 and the relief is described here instead. See docs/superpowers/research/2026-08-02-childcare-double-subtraction.md. Open follow-up: the Toronto monthly rate itself is a blend of the capped and uncapped rates and still owes its own sourcing pass." }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/greater-toronto-area", "last_updated": "2026-07-18", "confidence": "rich", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "non_cash.parental_leave_weeks": { "source": "Employment Insurance Act ss. 22-23: maternity 15 wk + standard parental up to 35 wk per parent (40 when shared) https://www.canada.ca/en/services/benefits/ei/ei-maternity-parental.html", "last_updated": "2026-08-06", "confidence": "high", "note": "50 = EI maternity 15 (mother) + standard parental 35, a shared pot counted once at the one-parent maximum. The 5 extra weeks that exist only when the second parent also claims are excluded as conditional on both parents' behaviour (same ruling as Portugal's sharing bonus). The extended option (up to 61 wk one parent, at 33%) was rejected: near-identical capped dollars stretched over 26 more weeks would inflate the weeks column. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. Quebec runs QPIP instead, so montreal legitimately diverges (city-varying by design in data/_national_fields.json)." }, "non_cash.parental_leave_paid_pct": { "source": "canada.ca, EI maximum insurable earnings for 2026: CAD 68,900 https://www.canada.ca/en/employment-social-development/programs/ei/ei-list/ei-employers/premium-reduction-program/2026-maximum-insurable-earnings.html", "last_updated": "2026-08-06", "confidence": "high", "note": "55% of insurable earnings for both maternity and standard parental weeks, capped at the 2026 MIE of CAD 68,900, max CAD 729/wk. Against a CAD 130,000 Toronto tech salary the cap makes the effective rate about 29%; the 55 is real only below the MIE. Headline kept per field convention; the capped-percentage shape has no honest flat-block representation, so the arithmetic lives here." }, "housing.homeowners_insurance_pct_of_value": { "source": "Provincial average premium (rates.ca Q2 2026, MyChoice city data, IBC) over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "CIHI national health expenditure household out-of-pocket + provincial drug plans: Quebec RAMQ covers a registered parent's children free, Ontario Trillium and BC Fair PharmaCare are income-deductible-based", "last_updated": "2026-08-08", "confidence": "medium", "note": "Medically necessary physician and hospital care is free everywhere, so the whole of this field is prescription drugs and allied health, which is PROVINCIAL, and the existing per-city spread is kept for that reason. Employer drug plans are a benefit and a premium, not out-of-pocket. Montreal carries a much smaller child increment than the other cities because of the RAMQ free-child rule. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "CIHI national health expenditure household out-of-pocket + provincial drug plans: Quebec RAMQ covers a registered parent's children free, Ontario Trillium and BC Fair PharmaCare are income-deductible-based", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Statistics Canada, average commute times for the 15 largest CMAs, May 2025 (The Daily, 2025-08-26, Table 4). https://www150.statcan.gc.ca/n1/daily-quotidien/250826/t004a-eng.htm", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Commuting duration, one way", "year": "2025", "note": "Toronto CMA 34.9, the longest in Canada and up 1.6 min year over year." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "CRA, MP, DB, RRSP, DPSP, ALDA, TFSA limits and the YMPE https://www.canada.ca/en/revenue-agency/services/tax/registered-plans-administrators/pspa/mp-rrsp-dpsp-tfsa-limits-ympe.html ; https://www.canada.ca/en/revenue-agency/services/tax/individuals/topics/rrsps-related-plans/contributing-a-rrsp-prpp/contributions-affect-your-rrsp-prpp-deduction-limit.html", "last_updated": "2026-08-11", "confidence": "high", "note": "RRSP deduction limit (line 20800), 2026: 18% of prior-year earned income to an annual dollar limit. Pinned at 33,810 via the statutory identity RRSP limit year X = money-purchase limit year X-1 (CRA confirms 2025 MP = 33,810). The PENSION ADJUSTMENT is deliberately NOT modeled, so this is the no-workplace-DB case; a DB-plan member has materially less room. Carry-forward of unused room is also unmodeled. Never binds at our incomes; the 18% governs below CAD 187,833." } }, "geneva": { "tax.federal_brackets": { "source": "ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Exact 2026 Grundtarif (kalte-Progression vintage confirmed against the 2025 circular). 2026-08-07: restored the missing 5.94% band 76,200-82,100 (the stored table jumped 2.97% to 6.6%, overstating by up to CHF 38.94/yr). Applied to taxable income = gross minus the statutory deduction stack (see tax.payroll_deductible)." }, "tax.cantonal_brackets": { "source": "RCEPF art. 17, bareme indexe 2026 (rsGE D 3 08.05) https://silgeneve.ch/legis/data/rsg_d3_08p05.htm ; ESTV Feuille cantonale GE etat 1.1.2026 https://www.estv2.admin.ch/stp/kb/ge-fr.pdf ; ArCA-2026 Ville de Geneve https://silgeneve.ch/legis/data/rsg_d3_05p30.htm", "last_updated": "2026-08-07", "confidence": "high", "note": "RECALIBRATED 2026-08-07: exact 2026 bareme (art. 41 al. 1 LIPP as indexed; post-2024-reform scale, top base 18%) x 1.7617 combined multiplier = (1 + 0.475 cantonal centimes + 0.01 aide-a-domicile) x 0.88 LDIRPP reduction + 0.4549 Ville de Geneve communal centimes. REFUTES the prior scale, which used bareme x 0.93 (centimes treated as the whole tax, base impot dropped, no LDIRPP, pre-2024 bareme vintage) and understated Geneva income tax ~36% at the ICP probe (12,305.65 vs official 19,180.39 at taxable 105,000). Married stays divisor-2 splitting (art. 41 al. 2 LIPP, confirmed in force 2026). Simplification kept: domicile-commune centimes on the full base (ESTV-calculator standard)." }, "tax.property_tax_rate_pct": { "source": "GE impôt immobilier complémentaire ~0.1%", "last_updated": "2026-05-16", "confidence": "medium", "note": "Minimal residential property tax. Eigenmietwert-equivalent imputed rent applies similar to Zurich" }, "housing.median_rent_by_bedroom": { "source": "Homegate Geneva 2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Peer to Zurich; among most expensive in Europe" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare.premium_monthly_by_family_size": { "source": "Geneva canton avg KVG premium 2026", "last_updated": "2026-05-16", "confidence": "high", "note": "Geneva canton ~CHF 520/mo adult (higher than Zurich's ~CHF 480). Kids ~CHF 130" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Geneva Caisses d'allocations familiales", "last_updated": "2026-05-16", "confidence": "high", "note": "GE Kinderzulage CHF 311/mo (higher than ZH's CHF 200). Annual ~CHF 3,732 Swiss Kinderzulagen / allocations familiales: cantonal, universal minimum (>= CHF 200-250/mo), NOT means-tested." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Geneva: CHF 98,995, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "CHF 98,995 / CHF 143,634 (Zurich, fetched the SAME DAY) = 0.6892 of Zurich; x Zurich's 0.6832 = 0.4709, R5-rounded down to 0.47. A SAME-DAY RATIO is used here rather than the direct CHF/FX/250,000 division used elsewhere, and only here, because Zurich is the one country anchor whose stored table has drifted materially: stored CHF 137,843 against CHF 143,634 live on 2026-08-11, 4.2% apart. Dividing a fresh satellite figure by 250,000 while the anchor sits on a 4.2%-stale table leaks that gap into every Swiss satellite in the FLATTERING direction. Every other same-day anchor drifted under 3% and in the conservative direction (Oslo 2.8%, Sweden 1.6%, Denmark 1.3%, Germany 1.0%), so none of them needed this. REPLACES 0.85 (2026-08-11), and the interim 0.55 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Resolves the note's own 29% self-contradiction: its quoted senior CHF 180-210k reproduces ZURICH's senior median (CHF 180,717), not Geneva's (CHF 136,933). Caveat: Geneva's p25 of CHF 66,394 is implausibly low and drags the median. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.85 was 0.8095 of Zurich's 1.05; that ratio x Zurich's re-derived 0.68 = 0.5505, R5-rounded down to 0.55. Zurich's own figure: CHF 137,843 / 0.807 = USD 170,809, / USD 250,000 = 0.6832. This is a BASIS move, not a re-measurement: Geneva's rank inside Switzerland is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE CHF 180-210k. Below Zurich (1.05). Smaller tech-employer concentration despite proximity to CERN/UN" }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Geneva + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Geneva", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Geneva", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "non_cash._summary": { "source": "Swiss Code of Obligations + BSV", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Vacation, holidays, sick leave, parental leave per country." }, "tax.bvg": { "source": "BVG statutory minimum, 2026 parameters (unchanged from 2025): entry threshold CHF 22,680, coordination deduction CHF 26,460, insured-salary upper limit CHF 90,720, coordinated salary clamped to 3,780-64,260. ASGA Pensionskasse, Merkblatt Renten und Grenzbetraege per 1.1.2026 https://www.asga.ch/wp-content/uploads/downloads/merkblaetter/merkblaetter-d/renten-und-grenzbetraege_2026_d.pdf ; BSV https://www.bsv.admin.ch/de/altersvorsorge-beruflichen-vorsorge (confirms the 22,680 threshold, savings from age 25, and the BVG Art. 66 employer-parity rule); age credits BVG Art. 16: 7/10/15/18% of coordinated salary by age band (Sicherheitsfonds BVG, sfbvg.ch Sparprozess), employee pays half: 3.5/5/7.5/9%.", "last_updated": "2026-08-06", "confidence": "high", "note": "Statutory MINIMUM only, deliberately: real Swiss professional plans insure above the mandatory band (ueberobligatorisch) and add plan-specific risk/admin premiums, so a typical payslip deducts MORE than this charge; everything above the legal floor is excluded under the statutory-minimum convention (the same rule that keeps Swedish ITP and Danish employer pensions out entirely). The charge docks take-home inside taxes.payroll and is carved to the forced-savings own-money row (funded, FZG-vested, portable). It runs on the same wage base as the AHV/ALV slice and never reduces the income-tax base: the CH-1 deduction floor stays CHF 13,500/27,000, which errs against Switzerland. Applied per earner (entry threshold and coordination deduction per person); the 18-24 bracket pays 0 because retirement credits start at 25 (BVG Art. 7). Decision log: docs/superpowers/research/2026-08-06-pension-omission-pair.md" }, "tax.cantonal_splitting_divisor": { "source": "Geneva LIPP (RS/GE D 3 08) art. 41 al. 2, official cantonal legislation database https://silgeneve.ch/legis/data/rsg_d3_08.htm", "last_updated": "2026-08-01", "confidence": "high", "married_variant": "divisor", "note": "Upgraded 2026-08-01 from an aggregator citation to the statute itself. LIPP art. 41 al. 2: \"Pour les epoux vivant en menage commun, le taux applique a leur revenu est celui qui correspond a 50% de ce dernier\", which is full splitting at a divisor of exactly 2, algebraically 2 x tax(joint/2). Art. 41 al. 1 is a single bareme for everyone; all 15 bareme references in LIPP point back to it and the only other one is art. 59 (wealth tax). Corroborated by art. 24, which counts the exempt first tranche \"sur le double de ce montant\" for art. 41 al. 2/3 taxpayers. This is why geneva carries no tax.cantonal_brackets_married and must not be given one: a married bracket array would send the engine down the Doppeltarif branch and model the same relief twice." }, "non_cash.parental_leave_weeks": { "source": "BSV https://www.bsv.admin.ch/en/income-compensation-parents (federal 14 + 2); canton of Geneva LAMat (loi genevoise sur l'assurance-maternite, 16-wk cantonal maternity)", "last_updated": "2026-08-06", "confidence": "high", "note": "18 wk = 16 maternity (canton LAMat GE extends the federal 14) + 2 paternity (10 working days within 6 months, federal statute since 1 Jan 2021), each parent's own entitlement; no federal shared scheme exists. The previously stored value omitted the 2 paternity weeks its own note cited. Cantonal ceiling for the two extra maternity weeks not re-verified this pass; describing the whole period at the federal CHF 220/day cap errs conservative. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md." }, "non_cash.parental_leave_paid_pct": { "source": "BSV https://www.bsv.admin.ch/en/income-compensation-parents : up to 80% of average income, max. CHF 220 per day, for both the maternity and paternity allowances", "last_updated": "2026-08-06", "confidence": "high", "note": "80% capped at CHF 220/day = CHF 1,540/wk, a cap reached at about CHF 99,000/yr, so every ICP salary receives the flat maximum: at CHF 130,000 the effective rate is about 62% and falls as salary rises. Kept as pct 80 because below the threshold it is a true 80%; the capped-percentage shape has no honest flat-block representation, so the cap arithmetic lives here." }, "tax.federal_brackets_married": { "source": "ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Exact 2026 Verheiratetentarif (Art. 36 Abs. 2 DBG), marginal form with the 11.5% flat crossover at 941,300. Replaces the divisor-2 full-splitting approximation, which UNDERSTATED federal married tax (exact 7,579.60 vs splitting 3,185.42 at joint taxable 166,920) - the federal tariff is harsher than splitting. All checkpoints re-derived by hand; see docs/superpowers/research/2026-08-08-ch-federal-worksheet.md." }, "tax.payroll_deductible": { "source": "Art. 33 Abs. 1 lit. d + f DBG SR 642.11, consolidation in force 2026-01-01 (fedlex)", "last_updated": "2026-08-07", "confidence": "high", "note": "Employee AHV/IV/EO, ALV and BVG contributions are fully deductible from the income-tax base (salary-certificate net is the taxable base). Gates the per-earner SI+BVG deduction in computeSwitzerlandTaxes; the charged amounts are the deducted amounts, so the FORCED_SAVINGS carve cannot double-count. Replaces the frozen CHF 13,500/27,000 CH-1 floor (2026-07-26), which bundled SI at one income level." }, "tax.professional_forfait": { "source": "Berufskostenverordnung SR 642.118.1 Art. 7 + Anhang Ziff. 1 (fedlex); 2026 min/max confirmed in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Uebrige Berufskosten pauschale: 3% of NET salary (post-SI/BVG), min CHF 2,000 / max CHF 4,000, per earner with income. Commuting and meal deductions excluded (variable/behavioral, conservative)." }, "tax.insurance_deduction_single": { "source": "Art. 33 Abs. 1 lit. g DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amount in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Versicherungsabzug CHF 1,800 (taxpayer WITH pillar-2 contributions, the employee case). Real KVG premiums exceed this cap and some cantons deduct more (BS 4,200, GE full premiums) - federal amount kept for the one-base model, conservative." }, "tax.insurance_deduction_married": { "source": "Art. 33 Abs. 1 lit. g DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amount in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Versicherungsabzug CHF 3,700 married (WITH pillar-2 amounts). Per-child +700 supplement excluded (conservative)." }, "tax.married_deduction": { "source": "Art. 35 Abs. 1 lit. c DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amount in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Verheiratetenabzug CHF 2,800 (2026), automatic for married couples." }, "tax.two_earner_deduction": { "source": "Art. 33 Abs. 2 DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amounts in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Zweitverdienerabzug: 50% of the lower earned income, min CHF 8,600 / max CHF 14,100, capped at the lower income; applies only when both spouses have earned income. Federal amount used for the one-base model (cantonal equivalents are smaller: BS 1,100, ZH ~6,100); partially offset by the larger cantonal insurance deductions we exclude - see the reasoning in docs/superpowers/research/2026-08-08-tax-structural-ch-ie.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "BAG Kostenbeteiligung https://www.bag.admin.ch/de/krankenversicherung-kostenbeteiligung-fuer-in-der-schweiz-wohnhafte-versicherte + BAG OKP statutory cost-sharing aggregate CHF 5.7bn (2024) over 8.96m insured, adult/child split from BAG age-group net benefits", "last_updated": "2026-08-08", "confidence": "high", "note": "Blended CHF 636 per insured person; children ~CHF 140 each (zero franchise, halved CHF 350 Selbstbehalt cap, exempt from the hospital per-diem), leaving CHF 749 per adult, plus a ~7% conservative uplift for non-KVG self-pay that is neither dental nor long-term care. REPLACES a value that was a single adult's statutory MAXIMUM: the old CHF 1,200 was above the CHF 1,000 minimum-franchise ceiling and was billed unchanged to a family of four whose statutory maximum is CHF 2,700. Geneva keeps its existing 1.083 premium, which is a utilisation difference, not a rules difference: cost-sharing law is federal and identical in every canton. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "BAG Kostenbeteiligung https://www.bag.admin.ch/de/krankenversicherung-kostenbeteiligung-fuer-in-der-schweiz-wohnhafte-versicherte + BAG OKP statutory cost-sharing aggregate CHF 5.7bn (2024) over 8.96m insured, adult/child split from BAG age-group net benefits", "last_updated": "2026-08-08", "confidence": "high", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Swiss Federal Statistical Office (BFS), PxWeb cube px-x-2105000000_206, nationale Agglomeration Genf. https://www.pxweb.bfs.admin.ch/pxweb/de/px-x-2105000000_206/px-x-2105000000_206/px-x-2105000000_206.px", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Durchschnittliche Arbeitswegzeit, in Minuten (ein Hinweg)", "year": "2020", "note": "34.7 min. NATIONALE Agglomeration, i.e. the Swiss side only, so it excludes the large French cross-border commute and is biased short." }, "non_cash.sick_leave": { "source": "Code of Obligations / Obligationenrecht (OR/CO) Art. 324a, paragraphs 1-4.; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Steuerverwaltung Kanton Bern, Taxinfo Saeule 3a https://taxinfo.sv.fin.be.ch/taxinfo/a2830da2-1a00-4928-b58b-74da5d48a598 ; EFD/ESTV confirmation of the 2026 amounts, 17 Nov 2025 https://www.estv.admin.ch/de/newnsb/xgRMirCsezICX4rtof9Lm", "last_updated": "2026-08-11", "confidence": "high", "note": "Saeule 3a kleiner Abzug, CHF 7,258 for 2026 (unchanged from 2025), for anyone who also contributes to a Pensionskasse, which our ICP does (both CH cities carry a bvg block). The grosser Abzug (20% of income to CHF 36,288, no 2nd pillar) is NOT our case. Employee-only; ESTV states the maximum deduction is simultaneously the contribution limit. 2nd-pillar Einkauf is also deductible but is bounded by a personal Einkaufspotenzial we cannot know, so it is excluded." } }, "denver": { "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 77900 USD, Denver-Aurora-Centennial metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "tax.federal_brackets": { "source": "IRS 2026 OBBBA (US parity)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.state_brackets": { "source": "CO DOR: flat 4.40% (stable 2024-2026)", "last_updated": "2026-05-16", "confidence": "high", "note": "CO flat-tax modeled as single-bracket array. Re-verified 2026-07-31 against Tax Foundation's 2026 state rate table" }, "tax.state_payroll_premium_rate": { "source": "Colorado FAMLI 2026 premium rate https://famli.colorado.gov/employers", "last_updated": "2026-07-31", "confidence": "high", "note": "Colorado FAMLI total premium fell from 0.90% to 0.88% on 2026-01-01, split evenly, so the employee pays 0.44%" }, "tax.state_payroll_premium_cap": { "source": "Colorado FAMLI / SSA 2026 contribution and benefit base", "last_updated": "2026-07-31", "confidence": "high", "note": "FAMLI wages are capped at the Social Security wage base ($184,500 in 2026)" }, "tax.sales_tax_rate": { "source": "Denver combined 8.81%", "last_updated": "2026-05-16", "confidence": "high", "note": "2.9% state + 1% RTD + 0.1% SCFD + 4.81% Denver city" }, "tax.property_tax_rate_pct": { "source": "Denver County assessor (CO Gallagher legacy + 6.7% assessment rate)", "last_updated": "2026-05-16", "confidence": "high", "note": "CO has unusually low property tax via Gallagher Amendment legacy. Denver OPT $5.75/mo head tax omitted (negligible at tech-worker income)" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Denver: USD 164,043, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "USD 164,043 / 1 = USD 164,043; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.6562, R5-rounded down to 0.65. REPLACES 0.93 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Median ~$200-230k. Palantir HQ + Google Boulder + Meta + Lockheed + Schwab HQ. Outdoor-lifestyle pull caps below coastal hubs" }, "housing._summary": { "source": "Numbeo Housing (cost-of-living basket) https://www.numbeo.com/cost-of-living/in/Denver", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation (90/10 quality bar). Per-field values were researched by subagents from local marketplaces (Zillow/Hemnet/Funda/Rightmove/etc.); cite each field individually at next refresh pass." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare._summary": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "medium", "note": "Group-level citation." }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Denver + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Denver", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Denver", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "IRS Child Tax Credit + state EITC", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "labor_market.tech_salary_distribution": { "source": "https://www.bls.gov/oes/current/oes_19740.htm", "last_updated": "2025-05", "confidence": "rich", "note": "BLS OEWS May 2025, SOC 15-1252 Software Developers, annual wage percentiles (full-market occupation distribution; straight-time gross pay incl. incentive pay, excl. equity and nonproduction bonuses). Public domain; API pull 2026-07-19." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Denver County CO (FIPS 8031), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Colorado caveat: the state MRS reports respondents’ MAXIMUM price charged for a regular day, so Denver is biased HIGH against the other cities on this spine. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $1,730/mo and the ages 1-2 rate $2,020/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,500. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.income_tested_childcare": { "source": "https://www.denvergov.org/Government/Agencies-Departments-Offices/Agencies-Departments-Offices-Directory/Denver-Human-Services/Be-Supported/Child-Care", "last_updated": "2026-08-05", "confidence": "high", "note": "Denver County entry at 225% FPL ($72,338 for a family of 4, effective 1 Oct 2025; the line is county-set, not Colorado-wide) with recertification at 85% SMI, an exit ceiling of $117,946; Denver CCCAP enrollment is currently FROZEN for new applicants, so a household under the line may qualify on paper but cannot presently enroll." }, "non_cash.parental_leave_weeks": { "source": "CO FAMLI https://famli.colorado.gov/individuals-and-families/reasons-to-take-famli-leave/parental-bonding-leave ; https://famli.colorado.gov/individuals-and-families/individuals-and-families-faqs", "last_updated": "2026-08-06", "confidence": "high", "note": "24 = 12 weeks of CO FAMLI bonding leave per parent (the agency states both parents are eligible and need not take the weeks concurrently), summed (both-parents flag set). The +4 pregnancy or childbirth complication weeks and the up-to-12 Neonatal Care Leave weeks are condition-dependent, EXCLUDED. Was 12/75. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "famli.colorado.gov: 90% of wages at or below 50% of the state AWW, 50% above, capped at 90% of the state AWW", "last_updated": "2026-08-06", "confidence": "high", "note": "0 percentage-paid weeks: FAMLI pays 90% of wages at or below 50% of the state AWW plus 50% above, capped at 90% of the state AWW: USD 1,381.45/wk Jan-Jun 2026 and USD 1,448 from 2026-07-01 (the cap now steps at the July AWW date), binding from about USD 112,000-117,000, so it lives in non_cash.parental_leave_flat. At USD 150,000 the effective rate is about 50%." }, "non_cash.parental_leave_flat": { "source": "famli.colorado.gov FAQ: benefits are capped at USD 1,448 per week as of July 1, 2026 (USD 1,381.45 Jan-Jun) https://famli.colorado.gov/individuals-and-families/individuals-and-families-faqs", "last_updated": "2026-08-06", "confidence": "high", "note": "USD 1,448/wk, the cap in force at the research date (2026-08-06), stored per the batch-2 NZ in-force-now convention with the mid-year step (1,381.45 Jan-Jun) recorded. 24 x 1,448 = USD 34,752 total." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Denver-Aurora-Centennial CO Metro Area (31000US19740). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US19740", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 27.7. Stored value was already correct; citation added." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Denver County CO (FIPS 8031), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Colorado caveat: the state MRS reports respondents’ MAXIMUM price charged for a regular day, so Denver is biased HIGH against the other cities on this spine. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,800." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "portland": { "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 72600 USD, Portland-Vancouver-Hillsboro metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "tax.federal_brackets": { "source": "IRS 2026 OBBBA (US parity)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.state_brackets": { "source": "Tax Foundation, State Individual Income Tax Rates and Brackets 2026 https://taxfoundation.org/data/all/state/state-income-tax-rates-2026/", "last_updated": "2026-07-31", "confidence": "high", "note": "OR has the highest US top state rate. 2026 single thresholds 4.75% to $4,550, 6.75% to $11,400, 8.75% to $125,000, 9.9% above. Refreshed 2026-07-31: the file still carried the ~2019 thresholds ($3,800 / $9,500)" }, "tax.city_income_tax_rate": { "source": "City of Portland Revenue Division, Personal Income Tax https://www.portland.gov/revenue/personal-tax", "last_updated": "2026-07-31", "confidence": "high", "note": "Zeroed 2026-07-31 in favour of tax.city_brackets. The flat 1% was wrong at both ends: it taxed the first $125k that no local tax touches, and it missed the Preschool for All step to 3% above $250k, understating a $450k filer by about $6,100/yr" }, "tax.city_brackets": { "source": "City of Portland Revenue Division, Personal Income Tax (tax year 2026 rates and thresholds) https://www.portland.gov/revenue/personal-tax", "last_updated": "2026-07-31", "confidence": "high", "note": "Single filer, the two Portland-area local taxes stacked as marginal bands: Multnomah County Preschool for All 1.5% over $125,000 and a further 1.5% (3% total) over $250,000, plus Metro Supportive Housing Services 1% over $128,000 (2026 is the first year the SHS threshold is inflation-indexed). Combined marginal: 0% to $125k, 1.5% to $128k, 2.5% to $250k, 4% above" }, "tax.city_brackets_married": { "source": "City of Portland Revenue Division, Personal Income Tax https://www.portland.gov/revenue/personal-tax", "last_updated": "2026-07-31", "confidence": "high", "note": "Joint filer thresholds: PFA 1.5% over $200,000 and 3% over $400,000; SHS 1% over $205,000. Combined marginal: 0% to $200k, 1.5% to $205k, 2.5% to $400k, 4% above" }, "tax.state_payroll_premium_rate": { "source": "Paid Leave Oregon employee contribution rate 2026 https://paidleave.oregon.gov/", "last_updated": "2026-07-31", "confidence": "high", "note": "Employee share 0.6% of wages, max $1,107/yr at the 2026 wage cap" }, "tax.state_payroll_premium_cap": { "source": "Paid Leave Oregon / SSA 2026 contribution and benefit base", "last_updated": "2026-07-31", "confidence": "high", "note": "Premiums apply up to the Social Security wage cap ($184,500 in 2026)" }, "tax.sales_tax_rate": { "source": "OR has NO sales tax (one of 5 US states)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Multnomah Co effective ~1.08% (Measure 50 caps assessment growth at 3%/yr)", "last_updated": "2026-05-16", "confidence": "high", "note": "Nominal millage higher but applied to suppressed assessed values" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Portland: USD 180,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "USD 180,000 / 1 = USD 180,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.7200, R5-rounded down to 0.72. REPLACES 0.9 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. n=385. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE ~$180-220k. Intel Hillsboro (suburb), Nike, Adidas, Salesforce/Tableau, Vacasa anchor. Lifestyle pull + smaller tech-employer base than Seattle" }, "housing._summary": { "source": "Numbeo Housing (cost-of-living basket) https://www.numbeo.com/cost-of-living/in/Portland", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation (90/10 quality bar). Per-field values were researched by subagents from local marketplaces (Zillow/Hemnet/Funda/Rightmove/etc.); cite each field individually at next refresh pass." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare._summary": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "medium", "note": "Group-level citation." }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Portland + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Portland", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Portland", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "IRS Child Tax Credit + state EITC", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "labor_market.tech_salary_distribution": { "source": "https://www.bls.gov/oes/current/oes_38900.htm", "last_updated": "2025-05", "confidence": "rich", "note": "BLS OEWS May 2025, SOC 15-1252 Software Developers, annual wage percentiles (full-market occupation distribution; straight-time gross pay incl. incentive pay, excl. equity and nonproduction bonuses). Public domain; API pull 2026-07-19." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "tax.state_brackets_married": { "source": "Oregon DOR, 2026 Withholding Tax Formulas 150-206-436 (Rev. 12-31-25) https://www.oregon.gov/dor/forms/FormsPubs/withholding-tax-formulas_206-436_2026.pdf, cross-checked against the 2025 Chart S / Chart J rate charts https://www.oregon.gov/dor/programs/individuals/Documents/Part-year%20and%20nonresident%20tax%20rate%20charts.pdf", "last_updated": "2026-08-01", "confidence": "high", "note": "Oregon Chart J, married filing jointly / head of household / qualifying surviving spouse, tax year 2026: 4.75% to 9,100, 6.75% to 22,800, 8.75% to 250,000, 9.9% above. Exactly 2x the Chart S single thresholds, verified in both TY2025 and TY2026. Added 2026-08-01; couples had been run through the single ladder, overstating Oregon tax by about $1,760 for a $300k household. Oregon indexes only the bottom two thresholds, so the 125,000 / 250,000 top-rate threshold is statutory and does not drift." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Multnomah County OR (FIPS 41051), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $1,550/mo and the ages 1-2 rate $2,050/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,300. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.income_tested_childcare": { "source": "https://www.oregon.gov/delc/programs/pages/erdc.aspx", "last_updated": "2026-08-05", "confidence": "high", "note": "Initial eligibility at 200% FPL ($66,000 for a family of 4, transmittal OEP-PT-26-005 effective 1 Mar 2026) with ongoing eligibility at 250% FPL or 85% SMI whichever is higher, an exit ceiling of $104,580; a household under the initial line may qualify, but ERDC runs a waitlist." }, "non_cash.parental_leave_weeks": { "source": "Paid Leave Oregon https://paidleave.oregon.gov/employees/overview.html ; ORS 657B.050 https://www.oregonlegislature.gov/bills_laws/ors/ors657B.html", "last_updated": "2026-08-06", "confidence": "high", "note": "24 = 12 weeks of Paid Leave Oregon family (bonding) leave per parent, summed (both-parents flag set). The 2 additional pregnancy-related weeks are condition-dependent, EXCLUDED. Was 12/50. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "paidleave.oregon.gov: 100% of AWW up to 65% of the state AWW, then 50% of the excess; ORS 657B.050", "last_updated": "2026-08-06", "confidence": "high", "note": "0 percentage-paid weeks: Oregon pays 100% of AWW up to 65% of the state AWW (USD 916.58) plus 50% of the excess, capped at 120% of the state AWW, USD 1,692.16/wk for benefit years starting on or after 2026-06-28 (1,636.56 before), binding from about USD 128,000, so it lives in non_cash.parental_leave_flat. At USD 150,000 the effective rate is about 59%." }, "non_cash.parental_leave_flat": { "source": "OED, minimum and maximum weekly benefit amounts, 2026-05-29: USD 1,692.16 (120% of the state AWW of 1,410.13) for benefit years starting on or after 2026-06-28 https://www.oregon.gov/employ/NewsAndMedia/Documents/2026-05-29-Minimum-Maximum-Weekly-Benefit-Amounts.pdf", "last_updated": "2026-08-06", "confidence": "high", "note": "USD 1,692.16/wk, the rate in force for claims starting at the research date (2026-08-06); stored per the batch-2 NZ in-force-now convention with the 2026-06-28 step recorded. 24 x 1,692.16 = USD 40,612 total." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Portland-Vancouver-Hillsboro OR-WA Metro Area (31000US38900). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US38900", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 25.4. Survey self-report, MSA geography." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Multnomah County OR (FIPS 41051), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,600." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "washington-dc": { "tax.federal_brackets": { "source": "IRS 2026 OBBBA (US parity)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.state_brackets": { "source": "DC OTR, individual income tax rates https://otr.cfo.dc.gov/page/dc-individual-and-fiduciary-income-tax-rates + D.C. Code 47-1806.03(a)(11) https://code.dccouncil.gov/us/dc/council/code/sections/47-1806.03", "last_updated": "2026-08-01", "confidence": "high", "married_variant": "none", "note": "VERIFIED 2026-08-01: DC applies ONE bracket table to every resident regardless of filing status. 47-1806.03(a)(11) imposes the tax \"on the taxable income of every resident\", and both the OTR rate table and the D-40 Calculation I schedule carry a single rate column with no filing-status split; 47-1806.03(e) treats a separately filing spouse as a single person. The seven bands (10,000 / 40,000 / 60,000 / 250,000 / 500,000 / 1,000,000, 4% up to 10.75%) are in force for tax years beginning after 12/31/2021 with no sunset. So the fallback to this array for married filers is CORRECT, and adding a widened married array would introduce an error." }, "tax.sales_tax_rate": { "source": "DC general sales 6%", "last_updated": "2026-05-16", "confidence": "high", "note": "No local additions to base 6%" }, "tax.property_tax_rate_pct": { "source": "DC Class 1 residential 0.85% statutory", "last_updated": "2026-05-16", "confidence": "high", "note": "Effective before homestead deduction. DC owner-occupied homestead deduction reduces assessed value $87,050" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Washington, DC: USD 159,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "USD 159,000 / 1 = USD 159,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.6360, R5-rounded down to 0.63. REPLACES 0.93 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE ~$190-230k DC proper. Federal contractors (Booz Allen, Leidos, Northrop, SAIC, CACI), Capital One McLean (suburbs), fintech. NoVa suburbs pay higher than DC city proper" }, "housing._summary": { "source": "Numbeo Housing (cost-of-living basket) https://www.numbeo.com/cost-of-living/in/Washington", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation (90/10 quality bar). Per-field values were researched by subagents from local marketplaces (Zillow/Hemnet/Funda/Rightmove/etc.); cite each field individually at next refresh pass." }, "housing.rent_percentiles": { "source": "Census ACS 2024 1-Year, table B25063 gross-rent distribution (metro/CBSA), data.census.gov; P25/P75 interpolated from dollar buckets, interpolated P50 matched published B25064 median (research 2026-07-17 §2). Rounded outward to 2dp per conservative-wider rule.", "last_updated": "2026-08-07", "confidence": "high", "note": "All-unit metro basis (pools bedroom sizes), wider than a within-bedroom band by construction. api.census.gov now requires an API key for queries (checked 2026-08-07); table remains on data.census.gov. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare._summary": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "medium", "note": "Group-level citation." }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Washington + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Washington", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Washington", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "IRS Child Tax Credit + state EITC", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 88700 USD, Washington-Arlington-Alexandria metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "labor_market.tech_salary_distribution": { "source": "https://www.bls.gov/oes/current/oes_47900.htm", "last_updated": "2025-05", "confidence": "rich", "note": "BLS OEWS May 2025, SOC 15-1252 Software Developers, annual wage percentiles (full-market occupation distribution; straight-time gross pay incl. incentive pay, excl. equity and nonproduction bonuses). Public domain; API pull 2026-07-19." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: District of Columbia DC (FIPS 11001), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $2,040/mo and the ages 1-2 rate $2,270/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $2,000. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.income_tested_childcare": { "source": "https://osse.dc.gov/childcaresubsidyfaq", "last_updated": "2026-08-05", "confidence": "high", "note": "Entry at 300% FPL ($93,600 for a family of 4, OSSE sliding fee scale effective 25 Oct 2024) with retention up to 85% of DC median income, an exit ceiling of $134,160; a household under the entry line may qualify, but new applicants have been waitlisted since 12 May 2026." }, "non_cash.parental_leave_weeks": { "source": "DC Paid Family Leave https://dcpaidfamilyleave.dc.gov/ ; DC Code s.32-541.04 https://code.dccouncil.gov/us/dc/council/code/sections/32-541.04", "last_updated": "2026-08-06", "confidence": "high", "note": "24 = 12 weeks of DC PFL parental leave per parent, summed (both-parents flag set). The pregnant parent's 2 prenatal weeks share the 12-week medical cap and are medical-care-conditioned, EXCLUDED, conservative, as is certified medical leave. Was 12/50. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "dcpaidfamilyleave.dc.gov ; DC Code s.32-541.04", "last_updated": "2026-08-06", "confidence": "high", "note": "0 percentage-paid weeks: DC pays 90% of AWW up to 150% of the DC minimum wage x 40 (USD 1,104/wk at the 18.40 minimum) plus 50% of the excess, capped at USD 1,190/wk (effective 2025-10-01, CPI-indexed, next re-index 2026-10-01), binding from about USD 78,000, so it lives in non_cash.parental_leave_flat. At USD 150,000 the effective rate is about 41%. The prior 1,153 figure is superseded." }, "non_cash.parental_leave_flat": { "source": "dcpaidfamilyleave.dc.gov benefits calculator: maximum weekly benefit USD 1,190 effective 2025-10-01 https://dcpaidfamilyleave.dc.gov/benefits-calculator/", "last_updated": "2026-08-06", "confidence": "high", "note": "USD 1,190/wk maximum effective 2025-10-01 (next CPI re-index 2026-10-01). 24 x 1,190 = USD 28,560 total." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Washington-Arlington-Alexandria DC-VA-MD-WV Metro Area (31000US47900). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US47900", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 33.8 (published MOE +/-0.3). Second-longest US metro commute after New York; the previous 30 badly undersold it." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: District of Columbia DC (FIPS 11001), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $2,400." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "raleigh-durham": { "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 73800 USD, Raleigh-Cary metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "tax.state_brackets": { "source": "NCDOR Tax Rate Schedules https://www.ncdor.gov/taxes-forms/individual-income-tax/tax-rate-schedules (Session Law 2023-134): flat 3.99% for tax years after 2025", "last_updated": "2026-07-30", "confidence": "high", "note": "Corrected 2026-07-30 from a stale 4.50%. NCDOR schedule: 2025 = 4.25%, tax years after 2025 = 3.99%. Matches charlotte.json, the other NC metro. Further reductions below 3.99% are trigger-based from 2027." }, "tax.sales_tax_rate": { "source": "Wake/Durham combined 7.25%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Wake County FY2026 + Raleigh-weighted blend ~0.95%", "last_updated": "2026-05-16", "confidence": "high", "note": "Durham County slightly higher ~1.05%" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Raleigh-Durham: USD 150,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "USD 150,000 / 1 = USD 150,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.6000, R5-rounded down to 0.6. REPLACES 0.85 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. n=557. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE $160-200k. Real research-triangle cluster but enterprise tech below coastal" }, "housing._summary": { "source": "Numbeo Housing (cost-of-living basket) https://www.numbeo.com/cost-of-living/in/Raleigh", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation (90/10 quality bar). Per-field values were researched by subagents from local marketplaces (Zillow/Hemnet/Funda/Rightmove/etc.); cite each field individually at next refresh pass." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare._summary": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "medium", "note": "Group-level citation." }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Raleigh + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Raleigh", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Raleigh", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "IRS Child Tax Credit + state EITC", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Wake County NC + Durham County NC (FIPS 37183, 37063), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $980/mo and the ages 1-2 rate $1,360/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,200. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.income_tested_childcare": { "source": "https://ncchildcare.ncdhhs.gov/parent/paying-for-child-care", "last_updated": "2026-08-05", "confidence": "high", "note": "Entry at 200% FPL for children 0-5 ($64,296 for a family of 4, statewide chart effective 1 Jul 2025; ages 6-12 enter at 133% FPL) with recertification capped at 85% SMI, an exit ceiling of $92,976; a household under the entry line may qualify." }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: North Carolina has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. North Carolina grants paid parental leave to state employees only. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Raleigh-Cary NC Metro Area (31000US39580). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US39580", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 27.4 on the Raleigh-Cary MSA. Our slug spans two MSAs; the Raleigh-Durham-Cary CSA gives 26.9 and Durham-Chapel Hill 24.4, so the longest was taken per the conservative-value rule. All three round to 27 or below, so the choice does not change the stored integer." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Wake County NC + Durham County NC (FIPS 37183, 37063), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,400." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "atlanta": { "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 65300 USD, Atlanta-Sandy Springs-Roswell metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "tax.state_brackets": { "source": "Office of the Governor of Georgia, Gov. Kemp Signs Legislation Lowering Taxes (HB 463, 2026-05-11) https://gov.georgia.gov/press-releases/2026-05-11/gov-kemp-signs-legislation-lowering-taxes-and-supporting-economic-growth", "last_updated": "2026-07-31", "confidence": "high", "note": "HB 463 lowered the flat rate from 5.19% to 4.99% beginning 2026-01-01. Re-verified 2026-07-31 against a user report claiming Georgia tax was missing and the rate was 5.19%: it is modeled, and 5.19% is the superseded 2025 rate. Tax Foundation's 2026 table still shows 5.19% because it predates the May signing" }, "tax.sales_tax_rate": { "source": "Atlanta combined 8.9% (4% GA + 4% Fulton + 0.4% MARTA + 0.5% Atlanta MOST)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Fulton County effective ~1.07% within Atlanta city limits", "last_updated": "2026-05-16", "confidence": "high" }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Fulton County GA (FIPS 13121), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $900/mo and the ages 1-2 rate $1,000/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $950. Does NOT model free universal pre-K where it exists. PRIOR NOTE: Near-universal enrollment; figure reflects private-market alternatives DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.income_tested_childcare": { "source": "https://georgia.gov/apply-childcare-and-parent-services-caps-program", "last_updated": "2026-08-05", "confidence": "high", "note": "Initial eligibility at 50% SMI ($55,368 for a family of 4, DECAL CAPS Appendix A effective 2 Mar 2026) with ongoing eligibility at 85% SMI, an exit ceiling of $94,126; a household under the initial line may qualify." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Atlanta: USD 148,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "USD 148,000 / 1 = USD 148,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.5920, R5-rounded down to 0.59. REPLACES 0.9 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE $170-210k. Fintech/enterprise SaaS hub; lacks FAANG HQ density" }, "housing._summary": { "source": "Numbeo Housing (cost-of-living basket) https://www.numbeo.com/cost-of-living/in/Atlanta", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation (90/10 quality bar). Per-field values were researched by subagents from local marketplaces (Zillow/Hemnet/Funda/Rightmove/etc.); cite each field individually at next refresh pass." }, "housing.rent_percentiles": { "source": "Census ACS 2024 1-Year, table B25063 gross-rent distribution (metro/CBSA), data.census.gov; P25/P75 interpolated from dollar buckets, interpolated P50 matched published B25064 median (research 2026-07-17 §2). Rounded outward to 2dp per conservative-wider rule.", "last_updated": "2026-08-07", "confidence": "high", "note": "All-unit metro basis (pools bedroom sizes), wider than a within-bedroom band by construction. api.census.gov now requires an API key for queries (checked 2026-08-07); table remains on data.census.gov. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare._summary": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "medium", "note": "Group-level citation." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Atlanta", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Atlanta", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "IRS Child Tax Credit + state EITC", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "labor_market.tech_salary_distribution": { "source": "https://www.bls.gov/oes/current/oes_12060.htm", "last_updated": "2025-05", "confidence": "rich", "note": "BLS OEWS May 2025, SOC 15-1252 Software Developers, annual wage percentiles (full-market occupation distribution; straight-time gross pay incl. incentive pay, excl. equity and nonproduction bonuses). Public domain; API pull 2026-07-19." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: Georgia has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. Georgia grants paid parental leave to state employees only. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, American Community Survey 2024 1-year, tables B08013/B08012 (= S0801_C01_046E), Atlanta-Sandy Springs-Roswell GA Metro Area (31000US12060). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US12060", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 32.4. Survey self-report, MSA geography, home workers excluded from the denominator." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Fulton County GA (FIPS 13121), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,200." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "miami": { "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 54200 USD, Miami-Fort Lauderdale-Pompano Beach metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "tax.state_brackets": { "source": "FL Constitution Art VII Sec 5: NO state income tax", "last_updated": "2026-05-16", "confidence": "high", "note": "FL is one of 9 US states with no income tax" }, "tax.sales_tax_rate": { "source": "Miami-Dade combined 7% (6% FL + 1% Miami-Dade surtax)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Miami-Dade residential effective ~1.02%", "last_updated": "2026-05-16", "confidence": "high", "note": "FL Save Our Homes amendment caps homesteaded assessment growth at 3%/yr" }, "transit.annual_car_cost_avg": { "source": "AAA + FL auto insurance premium (highest in US ~$2,800)", "last_updated": "2026-05-16", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Miami: USD 140,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "USD 140,000 / 1 = USD 140,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.5600, R5-rounded down to 0.56. REPLACES 0.95 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. n=363. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE $170-220k (rising). Comp bifurcated: local startups pay less, NYC/SF relocators retain coastal comp. Crypto/fintech/Citadel anchor" }, "housing._summary": { "source": "RentCafe avg-by-bedroom 2026-07-02 + Zumper median cross-check", "last_updated": "2026-07-26", "confidence": "high", "note": "2026-07-26 refresh: studio/1BR were stored low; 2BR/3BR held. RentCafe avg-by-bedroom (2026-07-02) with Zumper cross-check; average-tier basis." }, "housing.rent_percentiles": { "source": "Census ACS 2024 1-Year, table B25063 gross-rent distribution (metro/CBSA), data.census.gov; P25/P75 interpolated from dollar buckets, interpolated P50 matched published B25064 median (research 2026-07-17 §2). Rounded outward to 2dp per conservative-wider rule.", "last_updated": "2026-08-07", "confidence": "high", "note": "All-unit metro basis (pools bedroom sizes), wider than a within-bedroom band by construction. api.census.gov now requires an API key for queries (checked 2026-08-07); table remains on data.census.gov. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare._summary": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "medium", "note": "Group-level citation." }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Miami + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Miami", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Miami", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "IRS Child Tax Credit + state EITC", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "labor_market.tech_salary_distribution": { "source": "https://www.bls.gov/oes/current/oes_33100.htm", "last_updated": "2025-05", "confidence": "rich", "note": "BLS OEWS May 2025, SOC 15-1252 Software Developers, annual wage percentiles (full-market occupation distribution; straight-time gross pay incl. incentive pay, excl. equity and nonproduction bonuses). Public domain; API pull 2026-07-19." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "travel_home.avg_round_trip_per_person_to_other_city": { "source": "fare survey, Europe-Miami economy round trips (reporter-corroborated 600-700 EUR); pair-aware pricing is staged in docs/roadmap-staging.md", "last_updated": "2026-08-05", "confidence": "medium", "note": "Was 300, a domestic-fare figure on a field mostly consumed by intercontinental pairs (amsterdam-miami reporters). 700 USD approximates the EUR 600-700 economy RT band. Field stays per-city flat until the pair-aware design lands." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Miami-Dade County FL (FIPS 12086), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $750/mo and the ages 1-2 rate $910/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,300. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.income_tested_childcare": { "source": "https://www.elcmdm.org/o/elmdm/page/school-readiness", "last_updated": "2026-08-05", "confidence": "high", "note": "Entry at 55% SMI, statewide since HB 1255 ($59,467 for a family of 4, FY2027 sliding fee scale effective 1 Jul 2026), with exit at 85% SMI ($91,904); a household under the entry line may qualify, and ELC of Miami-Dade/Monroe reports its School Readiness waitlist was eliminated in FY2025-26." }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: Florida has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. Florida preempts local paid-leave mandates and only enables voluntary insurance products. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Miami-Fort Lauderdale-West Palm Beach FL Metro Area (31000US33100). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US33100", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 30.2. Survey self-report, MSA geography." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Miami-Dade County FL (FIPS 12086), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,600." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "pittsburgh": { "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 65000 USD, Pittsburgh metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "tax.state_brackets": { "source": "PA Dept of Revenue: flat 3.07% (lowest US flat-state rate)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.city_income_tax_rate": { "source": "Pittsburgh resident EIT 1% city + 2% PPS school = 3% combined", "last_updated": "2026-05-16", "confidence": "high", "note": "Exercises the city_income_tax_rate path (after NYC)" }, "tax.property_tax_rate_pct": { "source": "Allegheny County effective ~2.0% (varies 1.6-2.4% by school district)", "last_updated": "2026-05-16", "confidence": "high", "note": "PA notoriously high property tax to fund schools" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Pittsburgh: USD 135,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "USD 135,000 / 1 = USD 135,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.5400, R5-rounded down to 0.54. REPLACES 0.82 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. n=257. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE $160-200k. CMU research talent strong but enterprise pay below coastal" }, "housing._summary": { "source": "RentCafe avg-by-bedroom 2026-07-02 + Zumper median cross-check", "last_updated": "2026-07-26", "confidence": "medium", "note": "2026-07-26 refresh: prior base was too low across the board; magnitude uncertain (RentCafe runs hot here, ~25% above Zumper/Steadily, so weighted to the Zumper/Steadily median cluster). Re-verify 2BR/3BR against Zillow ZORI before treating as high-confidence." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare._summary": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "medium", "note": "Group-level citation." }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Pittsburgh + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Pittsburgh", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Pittsburgh", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "IRS Child Tax Credit + state EITC", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Allegheny County PA (FIPS 42003), NDCP study year 2020.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2020 to 2026 by CUUR0000SEEB03 (x1.27208; 2026 = Jan-Jul mean 396.748 against 2020 annual average 311.888). Tier C: Pennsylvania has no 2021 or 2022 data; the latest observed year is 2020, escalated over the longer window at x1.27208. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $1,060/mo and the ages 1-2 rate $1,140/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,000. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.income_tested_childcare": { "source": "https://www.pa.gov/en/agencies/dhs/resources/early-learning-child-care/child-care-works.html", "last_updated": "2026-08-05", "confidence": "high", "note": "Eligibility-only (CC2 shape): Child Care Works ceiling at 200% of the Federal Poverty Guidelines, $66,000 for a family of four. FPL-based, not SMI. May qualify, never a promise of a spot." }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: Pennsylvania has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. Pennsylvania grants paid parental leave to state employees only. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Pittsburgh PA Metro Area (31000US38300). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US38300", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 26.3. Stored value was already correct; citation added." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Allegheny County PA (FIPS 42003), NDCP study year 2020.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2020 to 2026 by CUUR0000SEEB03 (x1.27208; 2026 = Jan-Jul mean 396.748 against 2020 annual average 311.888). Tier C: Pennsylvania has no 2021 or 2022 data; the latest observed year is 2020, escalated over the longer window at x1.27208. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,200." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "salt-lake-city": { "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 64400 USD, Salt Lake City-Murray metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "tax.state_brackets": { "source": "Tax Foundation, State Individual Income Tax Rates and Brackets 2026 https://taxfoundation.org/data/all/state/state-income-tax-rates-2026/", "last_updated": "2026-07-31", "confidence": "high", "note": "Utah flat rate for tax year 2026 is 4.50%. Corrected 2026-07-31 from the stale 4.55% (a 2024 rate). SB 60 proposes a further cut to 4.45%. TODO (next Utah refresh): adopt only once confirmed enacted, the conservative call" }, "tax.sales_tax_rate": { "source": "SLC combined ~7.75% (4.85% UT + ~2.9% local)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Salt Lake County effective ~0.58% (UT 45% primary-residence exemption)", "last_updated": "2026-05-16", "confidence": "high", "note": "Among lowest US effective property tax" }, "non_cash.public_holidays": { "source": "11 federal + Pioneer Day (UT-specific July 24, widely observed)", "last_updated": "2026-05-16", "confidence": "high" }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Salt Lake City: USD 141,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "USD 141,000 / 1 = USD 141,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.5640, R5-rounded down to 0.56. REPLACES 0.88 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. n=427. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE $170-210k. Mature cluster, below Seattle but real" }, "housing._summary": { "source": "Numbeo Housing (cost-of-living basket) https://www.numbeo.com/cost-of-living/in/Salt-Lake-City", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation (90/10 quality bar). Per-field values were researched by subagents from local marketplaces (Zillow/Hemnet/Funda/Rightmove/etc.); cite each field individually at next refresh pass." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare._summary": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "medium", "note": "Group-level citation." }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Salt-Lake-City + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Salt-Lake-City", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Salt-Lake-City", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "IRS Child Tax Credit + state EITC", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Salt Lake County UT (FIPS 49035), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $870/mo and the ages 1-2 rate $1,070/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,000. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.income_tested_childcare": { "source": "https://jobs.utah.gov/customereducation/services/childcare/index.html", "last_updated": "2026-08-05", "confidence": "high", "note": "Ceiling at 85% SMI ($94,824 for a family of 4, DWS Table 4 effective 1 Oct 2024, the chart the agency posts as current); one line with no separate entry/exit split; a household under it may qualify." }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: Utah has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. Utah's 2020 statute grants paid parental leave to state-government employees only. Re-affirmed against correction flag:01784908255118-2eb6ab40: zero is the honest value on the household-inclusive basis too. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Salt Lake City-Murray UT Metro Area (31000US41620). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US41620", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 23.5. Survey self-report, MSA geography." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Salt Lake County UT (FIPS 49035), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,200." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "gothenburg": { "non_cash.parental_leave_flat": { "source": "Socialforsakringsbalken 12 kap. 19 para (lagstanivadagar) https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/socialforsakringsbalk-2010110_sfs-2010-110/ ; Forsakringskassan foraldrapenning", "last_updated": "2026-08-01", "confidence": "high", "note": "90 of the 480 foraldrapenning days pay a flat SEK 180 per calendar day regardless of income. Stored as 12.857 weeks at SEK 1,260 per week (180 x 7). Owner decision 2026-08-01: flat benefits are shown at the effective rate for the salary in front of us rather than folded into a headline percentage." }, "non_cash.parental_leave_paid_pct": { "source": "Socialforsakringsbalken 12 kap. 19 para (sjukpenningniva och lagstaniva) https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/socialforsakringsbalk-2010110_sfs-2010-110/ ; Forsakringskassan Aktuella belopp 2026", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: since the 2026-08-06 caps rebuild the 400 percentage-paid days live in non_cash.parental_leave_capped (77.6% up to SEK 8,813/week) and the 90 lagstanivadagar stay in non_cash.parental_leave_flat, so no weeks remain on this field. Was 77.6 with the SGI ceiling ignored and flagged as a sourcing gap, both closed by the capped block." }, "non_cash.parental_leave_weeks": { "source": "Socialforsakringsbalken (2010:110) 12 kap. (foraldrapenning, 480 dagar sammanlagt for foraldrarna, 90 reserverade per parent, 90 dagar pa lagstaniva) and 13 kap. 10 para (tillfallig foraldrapenning vid barns fodelse) https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/socialforsakringsbalk-2010110_sfs-2010-110/ ; Forsakringskassan https://www.forsakringskassan.se/privatperson/foralder/foraldrapenning", "last_updated": "2026-08-01", "confidence": "high", "note": "70 wk = the 480-day foraldrapenning pot (68.57 wk) + 10 tio-pappadagar (1.43 wk). Was 68. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. The 480 days are a SHARED pot (sammanlagt for foraldrarna), so they are counted once; the 90+90 reserved days sit INSIDE the 480 and are not summed on top. The 10 days for the non-birthing parent are tillfallig foraldrapenning under SFB 13 kap. 10 para, a legally separate benefit that does not draw down the 480, so they are additive." }, "tax.inbound_regime": { "source": "Expertskatt / forskarskattelattnad, IL 11 kap. 22-23a ss: 25% of remuneration exempt from income tax, max 7 years. Salary route = monthly remuneration that 'overstiger' 1.5x prisbasbelopp. Prisbasbelopp 2026 = SEK 59,200, fixed by the government 2025-09-18 (https://www.regeringen.se/pressmeddelanden/2025/09/prisbasbelopp-for-2026-faststallt/), so Forskarskattenamnden publishes SEK 88,801/mo as the qualifying minimum for work starting in 2026 = SEK 1,065,612/yr (https://forskarskattenamnden.se/andra-sprak/taxation-of-research-workers-board/conditions-for-tax-relief/remuneration-rate). Application within 3 months; no Swedish home or habitual abode in the 5 calendar years before the work starts; non-Swedish citizenship until 2027.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: min_income 1,058,400 -> 1,065,612. The old figure was wrong twice and both errors ran generous: it used the 2025 prisbasbelopp (58,800) and it used 12 x 88,200 rather than 12 x 88,201, ignoring that 'overstiger' is strictly greater. Two known over-generosities REMAIN and are stated in the regime note rather than modeled, because we do not collect the inputs: the threshold locks to the prisbasbelopp of the year the work STARTED and is never re-indexed, and only regular monthly pay and benefits count toward it, so annual bonuses, commissions and one-off payments cannot be used to clear the line. SCHEDULED: a lagradsremiss of June 2026 proposes lifting the exempt share to 30% and the residency bar to 7 years from 2027-01-01. Not law yet. TODO (re-check 2027-01): adopt only if enacted; until then the stored 25% stands." }, "tax.filing_basis": { "source": "Skatteverket (skatteverket.se) - individual taxation of spouses since 1971; no joint filing", "last_updated": "2026-07-11", "note": "Activates the per-filer split in compute() for two-earner households." }, "tax.kommunal_rate": { "source": "Göteborg 21.12% + Västra Götaland 11.48% = 32.60% 2026 (SCB kommunalskatterna); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Swedish fastighetsavgift (national cap)", "last_updated": "2026-05-16", "confidence": "high", "note": "Effective on market value ~0.15%" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Gothenburg: SEK 610,803, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "SEK 610,803 / 9.51 = USD 64,227; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2569, R5-rounded down to 0.25. REPLACES 0.38 (2026-08-11), and the interim 0.27 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.38 was 0.9048 of Stockholm's 0.42; that ratio x Stockholm's re-derived 0.3 = 0.2714, R5-rounded down to 0.27. Stockholm's own figure: SEK 730,138 / 9.51 = USD 76,776, / USD 250,000 = 0.3071. This is a BASIS move, not a re-measurement: Gothenburg's rank inside Sweden is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Sweden's 2nd tech city, ~10% below Stockholm Rebased 2026-08-11 with its country anchor (comp-index sweep A moved Oslo, Stockholm, Tokyo and Singapore); the documented ratio to the anchor is preserved, so this note's relative claim still holds." }, "housing._summary": { "source": "Numbeo Housing (cost-of-living basket) https://www.numbeo.com/cost-of-living/in/Gothenburg", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation (90/10 quality bar). Per-field values were researched by subagents from local marketplaces (Zillow/Hemnet/Funda/Rightmove/etc.); cite each field individually at next refresh pass." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Universal (folded into kommunal_rate)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation." }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Gothenburg + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Gothenburg", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Gothenburg", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "Försäkringskassan barnbidrag", "last_updated": "2026-05-16", "confidence": "medium" }, "non_cash._summary": { "source": "Semesterlagen + Försäkringskassan + universal", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Vacation, holidays, sick leave, parental leave per country." }, "tax.standard_deduction_single": { "source": "Grundavdrag floor for higher incomes ~SEK 17,400 (Skatteverket grundavdrag 2026; ekonomifakta.se); replaces the old 40,000 catch-all that tried to fold in the jobbskatteavdrag; verified tax audit 2026-07-14", "confidence": "high", "note": "The earned-income relief is now modeled explicitly as the jobbskatteavdrag_annual credit, so the deduction returns to the true grundavdrag floor.", "last_updated": "2026-07-17" }, "tax.standard_deduction_married": { "source": "Same grundavdrag floor as single (Sweden taxes spouses individually; Skatteverket); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.jobbskatteavdrag_annual": { "source": "Jobbskatteavdrag plateau ~SEK 52,400/yr 2026 (ekonomifakta.se/sakomraden/skatt/skatt-pa-arbete/jobbskatteavdraget_1212612.html); phase-out abolished from 2025; verified tax audit 2026-07-14", "confidence": "high", "note": "Credit against kommunal + statlig, clamped at the ladder. Plateau value; low incomes are modestly over-credited.", "last_updated": "2026-07-17" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "Swedish barnbidrag: universal, NOT means-tested." }, "tax.property_tax_annual_cap": { "source": "https://www.skatteverket.se/privat/fastigheterochbostad/fastighetsavgiftochfastighetsskatt.4.69ef368911e1304a625800013531.html", "last_updated": "2026-08-05", "confidence": "high", "note": "Kommunal fastighetsavgift, capped: SEK 1,784 per apartment in a flerbostadshus (2026). The bostadsrattsforening is the legal taxpayer and embeds the fee in the monthly avgift, so the owner's DIRECT line is 0; we attribute the per-apartment cap through to the household under the conservative rule rather than charging 0, since the cost is real and lands on the owner via the avgift. Urban own-mode stock is dominantly bostadsratt; a detached-house (smahus) owner would instead pay up to SEK 10,425 directly, and a rare agarlagenhet freeholder pays directly at the smahus cap. Previous model was an uncapped 0.15% of market price, which charged Stockholm 3BR SEK 12,000 against the 1,784 cap." }, "non_cash.parental_leave_capped": { "source": "Forsakringskassan, Aktuella belopp 2026: \"Hogsta mojliga SGI ar 2026: 592 000 kronor (10 ganger prisbasbeloppet)\" and maximum foraldrapenning at sjukpenningniva SEK 1,259/day ; regeringen.se 2025-09-18: 2026 prisbasbelopp SEK 59,200 ; SFB 28 kap. 7 para (the 0.97 SGI reduction factor, named-paragraph citation via lagar.se)", "last_updated": "2026-08-06", "confidence": "high", "note": "Payout ceiling SEK 1,259/day x 7 = SEK 8,813/week over the 400 percentage-paid days (57.143 wk; 57.143 + 12.857 flat = 70.0, the stored total exactly). This closes the 2026-08-01 sourcing gap: 592,000 x 0.97 x 0.80 / 365 = 1,258.6 ~ 1,259, which also reproduces the stored 77.6 (0.97 x 0.80), an exact internal-consistency check. Confidence medium-high on the 0.97 factor specifically (named statute paragraph plus arithmetic reproduction, not a verbatim quote), high on the kronor figures (direct agency table). At SEK 1,000,000 the capped slice pays ~46% and the whole blend ~39%. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Sweden) and docs/superpowers/research/2026-08-06-parental-caps-europe-batch-C.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "1177.se hoegkostnadsskydd for oeppenvaard and for laekemedel + LIF national prescription co-payment aggregate SEK 8.7bn over ~8.6m adults + Region Stockholm SEK 275 visit fee", "last_updated": "2026-08-08", "confidence": "medium", "note": "REPLACES an exact statutory maximum, and the wrong one of two: SEK 1,450 is precisely the 2026 oeppenvaard hoegkostnadsskydd and ignores the separate SEK 3,800 medicine ceiling that any adult filling a prescription also pays into. Built instead from ~SEK 1,010 of realized prescription co-payment plus ~SEK 700 of visit fees. Under-18s are free on BOTH ceilings, so the family rows barely move above two adults. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "1177.se hoegkostnadsskydd for oeppenvaard and for laekemedel + LIF national prescription co-payment aggregate SEK 8.7bn over ~8.6m adults + Region Stockholm SEK 275 visit fee", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Trafikanalys, RVU Sverige 2011-2014, Figur 3.4. https://www.trafa.se/globalassets/statistik/resvanor/2009-2015/rvu-sverige-2011-2014.pdf", "last_updated": "2026-08-08", "confidence": "low", "statistic": "Arbetsresornas genomsnittliga restid", "year": "2011-2014", "note": "NATIONAL figure adopted for want of a city series. Goteborg sits in the same SKL storstad group as Stockholm." }, "non_cash.sick_leave": { "source": "Lag (1991:1047) om sjuklön (\"sjuklönelagen\") for the employer phase; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.forsakringskassan.se/privatperson/e-tjanster-blanketter-och-informationsmaterial/aktuella-belopp", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Skatteverket, Avdrag foer pensionssparande https://www.skatteverket.se/privat/skatter/arbeteochinkomst/avdragforprivatpersoner/np.4.5fc8c94513259a4ba1d800042822.html ; 2026 prisbasbelopp https://www.regeringen.se/pressmeddelanden/2025/09/prisbasbelopp-for-2026-faststallt/", "last_updated": "2026-08-11", "confidence": "high", "note": "No deduction for our ICP. Skatteverket restricts the deduction to someone who \"helt saknar pensionsraett i sin anstaellning\", i.e. every employment they held in the income year lacked any right to tjaenstepension. The 35% / 10 prisbasbelopp rule (SEK 592,000 for 2026) reaches only that population; a salaried tech worker with occupational pension gets nothing. General deductibility was abolished in 2016. This is the single largest correction in the sweep." } }, "malmo": { "non_cash.parental_leave_flat": { "source": "Socialforsakringsbalken 12 kap. 19 para (lagstanivadagar) https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/socialforsakringsbalk-2010110_sfs-2010-110/ ; Forsakringskassan foraldrapenning", "last_updated": "2026-08-01", "confidence": "high", "note": "90 of the 480 foraldrapenning days pay a flat SEK 180 per calendar day regardless of income. Stored as 12.857 weeks at SEK 1,260 per week (180 x 7). Owner decision 2026-08-01: flat benefits are shown at the effective rate for the salary in front of us rather than folded into a headline percentage." }, "non_cash.parental_leave_paid_pct": { "source": "Socialforsakringsbalken 12 kap. 19 para (sjukpenningniva och lagstaniva) https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/socialforsakringsbalk-2010110_sfs-2010-110/ ; Forsakringskassan Aktuella belopp 2026", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: since the 2026-08-06 caps rebuild the 400 percentage-paid days live in non_cash.parental_leave_capped (77.6% up to SEK 8,813/week) and the 90 lagstanivadagar stay in non_cash.parental_leave_flat, so no weeks remain on this field. Was 77.6 with the SGI ceiling ignored and flagged as a sourcing gap, both closed by the capped block." }, "non_cash.parental_leave_weeks": { "source": "Socialforsakringsbalken (2010:110) 12 kap. (foraldrapenning, 480 dagar sammanlagt for foraldrarna, 90 reserverade per parent, 90 dagar pa lagstaniva) and 13 kap. 10 para (tillfallig foraldrapenning vid barns fodelse) https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/socialforsakringsbalk-2010110_sfs-2010-110/ ; Forsakringskassan https://www.forsakringskassan.se/privatperson/foralder/foraldrapenning", "last_updated": "2026-08-01", "confidence": "high", "note": "70 wk = the 480-day foraldrapenning pot (68.57 wk) + 10 tio-pappadagar (1.43 wk). Was 68. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. The 480 days are a SHARED pot (sammanlagt for foraldrarna), so they are counted once; the 90+90 reserved days sit INSIDE the 480 and are not summed on top. The 10 days for the non-birthing parent are tillfallig foraldrapenning under SFB 13 kap. 10 para, a legally separate benefit that does not draw down the 480, so they are additive." }, "tax.inbound_regime": { "source": "Expertskatt / forskarskattelattnad, IL 11 kap. 22-23a ss: 25% of remuneration exempt from income tax, max 7 years. Salary route = monthly remuneration that 'overstiger' 1.5x prisbasbelopp. Prisbasbelopp 2026 = SEK 59,200, fixed by the government 2025-09-18 (https://www.regeringen.se/pressmeddelanden/2025/09/prisbasbelopp-for-2026-faststallt/), so Forskarskattenamnden publishes SEK 88,801/mo as the qualifying minimum for work starting in 2026 = SEK 1,065,612/yr (https://forskarskattenamnden.se/andra-sprak/taxation-of-research-workers-board/conditions-for-tax-relief/remuneration-rate). Application within 3 months; no Swedish home or habitual abode in the 5 calendar years before the work starts; non-Swedish citizenship until 2027.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: min_income 1,058,400 -> 1,065,612. The old figure was wrong twice and both errors ran generous: it used the 2025 prisbasbelopp (58,800) and it used 12 x 88,200 rather than 12 x 88,201, ignoring that 'overstiger' is strictly greater. Two known over-generosities REMAIN and are stated in the regime note rather than modeled, because we do not collect the inputs: the threshold locks to the prisbasbelopp of the year the work STARTED and is never re-indexed, and only regular monthly pay and benefits count toward it, so annual bonuses, commissions and one-off payments cannot be used to clear the line. SCHEDULED: a lagradsremiss of June 2026 proposes lifting the exempt share to 30% and the residency bar to 7 years from 2027-01-01. Not law yet. TODO (re-check 2027-01): adopt only if enacted; until then the stored 25% stands." }, "tax.filing_basis": { "source": "Skatteverket (skatteverket.se) - individual taxation of spouses since 1971; no joint filing", "last_updated": "2026-07-11", "note": "Activates the per-filer split in compute() for two-earner households." }, "tax.kommunal_rate": { "source": "Skatteverket 2026: Malmö 21.24% + Skåne 11.18% = 32.42%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Swedish fastighetsavgift", "last_updated": "2026-05-16", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Malmö. Its basis is converted with the rest of Sweden and its rank inside Sweden is preserved. Prior source: Derived, not observed: no all-levels total-comp measurement exists for Malmö. Basis re-derived with its country anchor Stockholm; Malmö's stored ratio to Stockholm is preserved. Prior source: Gaming hub (King/Massive Entertainment, MSFT-acquired) + Øresund bridge to Copenhagen", "last_updated": "2026-08-11", "confidence": "low", "note": "0.36 x 0.7143 = 0.2571, R5-rounded down to 0.25. 0.7143 is the MEDIAN of (re-derived / previous) across the 2 Sweden cities that DO carry an all-levels measurement (Gothenburg 0.38 to 0.25; Stockholm 0.42 to 0.3), so Malmö moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Malmö's rank inside Sweden is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.36 was 0.8571 of Stockholm's 0.42; that ratio x Stockholm's re-derived 0.3 = 0.2571, R5-rounded down to 0.25. Stockholm's own figure: SEK 730,138 / 9.51 = USD 76,776, / USD 250,000 = 0.3071. This is a BASIS move, not a re-measurement: Malmö's rank inside Sweden is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Smaller tech market than Stockholm/Gothenburg Rebased 2026-08-11 with its country anchor (comp-index sweep A moved Oslo, Stockholm, Tokyo and Singapore); the documented ratio to the anchor is preserved, so this note's relative claim still holds." }, "housing._summary": { "source": "Numbeo Housing (cost-of-living basket) https://www.numbeo.com/cost-of-living/in/Malmoe", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation (90/10 quality bar). Per-field values were researched by subagents from local marketplaces (Zillow/Hemnet/Funda/Rightmove/etc.); cite each field individually at next refresh pass." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Universal (folded into kommunal_rate)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation." }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Malmoe + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Malmoe", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Malmoe", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "Försäkringskassan barnbidrag", "last_updated": "2026-05-16", "confidence": "medium" }, "non_cash._summary": { "source": "Semesterlagen + Försäkringskassan + universal", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Vacation, holidays, sick leave, parental leave per country." }, "tax.standard_deduction_single": { "source": "Grundavdrag floor for higher incomes ~SEK 17,400 (Skatteverket grundavdrag 2026; ekonomifakta.se); replaces the old 40,000 catch-all that tried to fold in the jobbskatteavdrag; verified tax audit 2026-07-14", "confidence": "high", "note": "The earned-income relief is now modeled explicitly as the jobbskatteavdrag_annual credit, so the deduction returns to the true grundavdrag floor.", "last_updated": "2026-07-17" }, "tax.standard_deduction_married": { "source": "Same grundavdrag floor as single (Sweden taxes spouses individually; Skatteverket); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.jobbskatteavdrag_annual": { "source": "Jobbskatteavdrag plateau ~SEK 52,400/yr 2026 (ekonomifakta.se/sakomraden/skatt/skatt-pa-arbete/jobbskatteavdraget_1212612.html); phase-out abolished from 2025; verified tax audit 2026-07-14", "confidence": "high", "note": "Credit against kommunal + statlig, clamped at the ladder. Plateau value; low incomes are modestly over-credited.", "last_updated": "2026-07-17" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "Swedish barnbidrag: universal, NOT means-tested." }, "tax.property_tax_annual_cap": { "source": "https://www.skatteverket.se/privat/fastigheterochbostad/fastighetsavgiftochfastighetsskatt.4.69ef368911e1304a625800013531.html", "last_updated": "2026-08-05", "confidence": "high", "note": "Kommunal fastighetsavgift, capped: SEK 1,784 per apartment in a flerbostadshus (2026). The bostadsrattsforening is the legal taxpayer and embeds the fee in the monthly avgift, so the owner's DIRECT line is 0; we attribute the per-apartment cap through to the household under the conservative rule rather than charging 0, since the cost is real and lands on the owner via the avgift. Urban own-mode stock is dominantly bostadsratt; a detached-house (smahus) owner would instead pay up to SEK 10,425 directly, and a rare agarlagenhet freeholder pays directly at the smahus cap. Previous model was an uncapped 0.15% of market price, which charged Stockholm 3BR SEK 12,000 against the 1,784 cap." }, "non_cash.parental_leave_capped": { "source": "Forsakringskassan, Aktuella belopp 2026: \"Hogsta mojliga SGI ar 2026: 592 000 kronor (10 ganger prisbasbeloppet)\" and maximum foraldrapenning at sjukpenningniva SEK 1,259/day ; regeringen.se 2025-09-18: 2026 prisbasbelopp SEK 59,200 ; SFB 28 kap. 7 para (the 0.97 SGI reduction factor, named-paragraph citation via lagar.se)", "last_updated": "2026-08-06", "confidence": "high", "note": "Payout ceiling SEK 1,259/day x 7 = SEK 8,813/week over the 400 percentage-paid days (57.143 wk; 57.143 + 12.857 flat = 70.0, the stored total exactly). This closes the 2026-08-01 sourcing gap: 592,000 x 0.97 x 0.80 / 365 = 1,258.6 ~ 1,259, which also reproduces the stored 77.6 (0.97 x 0.80), an exact internal-consistency check. Confidence medium-high on the 0.97 factor specifically (named statute paragraph plus arithmetic reproduction, not a verbatim quote), high on the kronor figures (direct agency table). At SEK 1,000,000 the capped slice pays ~46% and the whole blend ~39%. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Sweden) and docs/superpowers/research/2026-08-06-parental-caps-europe-batch-C.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "1177.se hoegkostnadsskydd for oeppenvaard and for laekemedel + LIF national prescription co-payment aggregate SEK 8.7bn over ~8.6m adults + Region Stockholm SEK 275 visit fee", "last_updated": "2026-08-08", "confidence": "medium", "note": "REPLACES an exact statutory maximum, and the wrong one of two: SEK 1,450 is precisely the 2026 oeppenvaard hoegkostnadsskydd and ignores the separate SEK 3,800 medicine ceiling that any adult filling a prescription also pays into. Built instead from ~SEK 1,010 of realized prescription co-payment plus ~SEK 700 of visit fees. Under-18s are free on BOTH ceilings, so the family rows barely move above two adults. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "1177.se hoegkostnadsskydd for oeppenvaard and for laekemedel + LIF national prescription co-payment aggregate SEK 8.7bn over ~8.6m adults + Region Stockholm SEK 275 visit fee", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Trafikanalys, RVU Sverige 2011-2014, Figur 3.4. https://www.trafa.se/globalassets/statistik/resvanor/2009-2015/rvu-sverige-2011-2014.pdf", "last_updated": "2026-08-08", "confidence": "low", "statistic": "Arbetsresornas genomsnittliga restid", "year": "2011-2014", "note": "NATIONAL figure adopted for want of a city series. GEOGRAPHY CALL: measured on the Swedish basis, not the Oresund/Copenhagen labour market, because cross-border commuters are a small minority of Malmo-area commuters." }, "non_cash.sick_leave": { "source": "Lag (1991:1047) om sjuklön (\"sjuklönelagen\") for the employer phase; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.forsakringskassan.se/privatperson/e-tjanster-blanketter-och-informationsmaterial/aktuella-belopp", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Skatteverket, Avdrag foer pensionssparande https://www.skatteverket.se/privat/skatter/arbeteochinkomst/avdragforprivatpersoner/np.4.5fc8c94513259a4ba1d800042822.html ; 2026 prisbasbelopp https://www.regeringen.se/pressmeddelanden/2025/09/prisbasbelopp-for-2026-faststallt/", "last_updated": "2026-08-11", "confidence": "high", "note": "No deduction for our ICP. Skatteverket restricts the deduction to someone who \"helt saknar pensionsraett i sin anstaellning\", i.e. every employment they held in the income year lacked any right to tjaenstepension. The 35% / 10 prisbasbelopp rule (SEK 592,000 for 2026) reaches only that population; a salaried tech worker with occupational pension gets nothing. General deductibility was abolished in 2016. This is the single largest correction in the sweep." } }, "uppsala": { "non_cash.parental_leave_flat": { "source": "Socialforsakringsbalken 12 kap. 19 para (lagstanivadagar) https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/socialforsakringsbalk-2010110_sfs-2010-110/ ; Forsakringskassan foraldrapenning", "last_updated": "2026-08-01", "confidence": "high", "note": "90 of the 480 foraldrapenning days pay a flat SEK 180 per calendar day regardless of income. Stored as 12.857 weeks at SEK 1,260 per week (180 x 7). Owner decision 2026-08-01: flat benefits are shown at the effective rate for the salary in front of us rather than folded into a headline percentage." }, "non_cash.parental_leave_paid_pct": { "source": "Socialforsakringsbalken 12 kap. 19 para (sjukpenningniva och lagstaniva) https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/socialforsakringsbalk-2010110_sfs-2010-110/ ; Forsakringskassan Aktuella belopp 2026", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: since the 2026-08-06 caps rebuild the 400 percentage-paid days live in non_cash.parental_leave_capped (77.6% up to SEK 8,813/week) and the 90 lagstanivadagar stay in non_cash.parental_leave_flat, so no weeks remain on this field. Was 77.6 with the SGI ceiling ignored and flagged as a sourcing gap, both closed by the capped block." }, "non_cash.parental_leave_weeks": { "source": "Socialforsakringsbalken (2010:110) 12 kap. (foraldrapenning, 480 dagar sammanlagt for foraldrarna, 90 reserverade per parent, 90 dagar pa lagstaniva) and 13 kap. 10 para (tillfallig foraldrapenning vid barns fodelse) https://www.riksdagen.se/sv/dokument-och-lagar/dokument/svensk-forfattningssamling/socialforsakringsbalk-2010110_sfs-2010-110/ ; Forsakringskassan https://www.forsakringskassan.se/privatperson/foralder/foraldrapenning", "last_updated": "2026-08-01", "confidence": "high", "note": "70 wk = the 480-day foraldrapenning pot (68.57 wk) + 10 tio-pappadagar (1.43 wk). Was 68. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. The 480 days are a SHARED pot (sammanlagt for foraldrarna), so they are counted once; the 90+90 reserved days sit INSIDE the 480 and are not summed on top. The 10 days for the non-birthing parent are tillfallig foraldrapenning under SFB 13 kap. 10 para, a legally separate benefit that does not draw down the 480, so they are additive." }, "tax.inbound_regime": { "source": "Expertskatt / forskarskattelattnad, IL 11 kap. 22-23a ss: 25% of remuneration exempt from income tax, max 7 years. Salary route = monthly remuneration that 'overstiger' 1.5x prisbasbelopp. Prisbasbelopp 2026 = SEK 59,200, fixed by the government 2025-09-18 (https://www.regeringen.se/pressmeddelanden/2025/09/prisbasbelopp-for-2026-faststallt/), so Forskarskattenamnden publishes SEK 88,801/mo as the qualifying minimum for work starting in 2026 = SEK 1,065,612/yr (https://forskarskattenamnden.se/andra-sprak/taxation-of-research-workers-board/conditions-for-tax-relief/remuneration-rate). Application within 3 months; no Swedish home or habitual abode in the 5 calendar years before the work starts; non-Swedish citizenship until 2027.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: min_income 1,058,400 -> 1,065,612. The old figure was wrong twice and both errors ran generous: it used the 2025 prisbasbelopp (58,800) and it used 12 x 88,200 rather than 12 x 88,201, ignoring that 'overstiger' is strictly greater. Two known over-generosities REMAIN and are stated in the regime note rather than modeled, because we do not collect the inputs: the threshold locks to the prisbasbelopp of the year the work STARTED and is never re-indexed, and only regular monthly pay and benefits count toward it, so annual bonuses, commissions and one-off payments cannot be used to clear the line. SCHEDULED: a lagradsremiss of June 2026 proposes lifting the exempt share to 30% and the residency bar to 7 years from 2027-01-01. Not law yet. TODO (re-check 2027-01): adopt only if enacted; until then the stored 25% stands." }, "tax.filing_basis": { "source": "Skatteverket (skatteverket.se) - individual taxation of spouses since 1971; no joint filing", "last_updated": "2026-07-11", "note": "Activates the per-filer split in compute() for two-earner households." }, "tax.kommunal_rate": { "source": "Skatteverket 2026: Uppsala kommun 21.14% + Uppsala län 11.71% = 32.85%", "last_updated": "2026-05-16", "confidence": "high" }, "transit.monthly_pass": { "source": "UL 30-dagarsbiljett 2026 (raised to SEK 1,150 Jan 8)", "last_updated": "2026-05-16", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Uppsala. Its basis is converted with the rest of Sweden and its rank inside Sweden is preserved. Prior source: Derived, not observed: no all-levels total-comp measurement exists for Uppsala. Basis re-derived with its country anchor Stockholm; Uppsala's stored ratio to Stockholm is preserved. Prior source: Uppsala University + biotech (GE Healthcare, Karolinska adjacency)", "last_updated": "2026-08-11", "confidence": "low", "note": "0.34 x 0.7143 = 0.2429, R5-rounded down to 0.24. 0.7143 is the MEDIAN of (re-derived / previous) across the 2 Sweden cities that DO carry an all-levels measurement (Gothenburg 0.38 to 0.25; Stockholm 0.42 to 0.3), so Uppsala moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Uppsala's rank inside Sweden is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.34 was 0.8095 of Stockholm's 0.42; that ratio x Stockholm's re-derived 0.3 = 0.2429, R5-rounded down to 0.24. Stockholm's own figure: SEK 730,138 / 9.51 = USD 76,776, / USD 250,000 = 0.3071. This is a BASIS move, not a re-measurement: Uppsala's rank inside Sweden is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Academic + biotech heavy, small tech market Rebased 2026-08-11 with its country anchor (comp-index sweep A moved Oslo, Stockholm, Tokyo and Singapore); the documented ratio to the anchor is preserved, so this note's relative claim still holds." }, "housing._summary": { "source": "Numbeo Housing (cost-of-living basket) https://www.numbeo.com/cost-of-living/in/Uppsala", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation (90/10 quality bar). Per-field values were researched by subagents from local marketplaces (Zillow/Hemnet/Funda/Rightmove/etc.); cite each field individually at next refresh pass." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Universal (folded into kommunal_rate)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation." }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Uppsala + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Uppsala", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Uppsala", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "Försäkringskassan barnbidrag", "last_updated": "2026-05-16", "confidence": "medium" }, "non_cash._summary": { "source": "Semesterlagen + Försäkringskassan + universal", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Vacation, holidays, sick leave, parental leave per country." }, "tax.standard_deduction_single": { "source": "Grundavdrag floor for higher incomes ~SEK 17,400 (Skatteverket grundavdrag 2026; ekonomifakta.se); replaces the old 40,000 catch-all that tried to fold in the jobbskatteavdrag; verified tax audit 2026-07-14", "confidence": "high", "note": "The earned-income relief is now modeled explicitly as the jobbskatteavdrag_annual credit, so the deduction returns to the true grundavdrag floor.", "last_updated": "2026-07-17" }, "tax.standard_deduction_married": { "source": "Same grundavdrag floor as single (Sweden taxes spouses individually; Skatteverket); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.jobbskatteavdrag_annual": { "source": "Jobbskatteavdrag plateau ~SEK 52,400/yr 2026 (ekonomifakta.se/sakomraden/skatt/skatt-pa-arbete/jobbskatteavdraget_1212612.html); phase-out abolished from 2025; verified tax audit 2026-07-14", "confidence": "high", "note": "Credit against kommunal + statlig, clamped at the ladder. Plateau value; low incomes are modestly over-credited.", "last_updated": "2026-07-17" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "Swedish barnbidrag: universal, NOT means-tested." }, "tax.property_tax_annual_cap": { "source": "https://www.skatteverket.se/privat/fastigheterochbostad/fastighetsavgiftochfastighetsskatt.4.69ef368911e1304a625800013531.html", "last_updated": "2026-08-05", "confidence": "high", "note": "Kommunal fastighetsavgift, capped: SEK 1,784 per apartment in a flerbostadshus (2026). The bostadsrattsforening is the legal taxpayer and embeds the fee in the monthly avgift, so the owner's DIRECT line is 0; we attribute the per-apartment cap through to the household under the conservative rule rather than charging 0, since the cost is real and lands on the owner via the avgift. Urban own-mode stock is dominantly bostadsratt; a detached-house (smahus) owner would instead pay up to SEK 10,425 directly, and a rare agarlagenhet freeholder pays directly at the smahus cap. Previous model was an uncapped 0.15% of market price, which charged Stockholm 3BR SEK 12,000 against the 1,784 cap." }, "non_cash.parental_leave_capped": { "source": "Forsakringskassan, Aktuella belopp 2026: \"Hogsta mojliga SGI ar 2026: 592 000 kronor (10 ganger prisbasbeloppet)\" and maximum foraldrapenning at sjukpenningniva SEK 1,259/day ; regeringen.se 2025-09-18: 2026 prisbasbelopp SEK 59,200 ; SFB 28 kap. 7 para (the 0.97 SGI reduction factor, named-paragraph citation via lagar.se)", "last_updated": "2026-08-06", "confidence": "high", "note": "Payout ceiling SEK 1,259/day x 7 = SEK 8,813/week over the 400 percentage-paid days (57.143 wk; 57.143 + 12.857 flat = 70.0, the stored total exactly). This closes the 2026-08-01 sourcing gap: 592,000 x 0.97 x 0.80 / 365 = 1,258.6 ~ 1,259, which also reproduces the stored 77.6 (0.97 x 0.80), an exact internal-consistency check. Confidence medium-high on the 0.97 factor specifically (named statute paragraph plus arithmetic reproduction, not a verbatim quote), high on the kronor figures (direct agency table). At SEK 1,000,000 the capped slice pays ~46% and the whole blend ~39%. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Sweden) and docs/superpowers/research/2026-08-06-parental-caps-europe-batch-C.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "1177.se hoegkostnadsskydd for oeppenvaard and for laekemedel + LIF national prescription co-payment aggregate SEK 8.7bn over ~8.6m adults + Region Stockholm SEK 275 visit fee", "last_updated": "2026-08-08", "confidence": "medium", "note": "REPLACES an exact statutory maximum, and the wrong one of two: SEK 1,450 is precisely the 2026 oeppenvaard hoegkostnadsskydd and ignores the separate SEK 3,800 medicine ceiling that any adult filling a prescription also pays into. Built instead from ~SEK 1,010 of realized prescription co-payment plus ~SEK 700 of visit fees. Under-18s are free on BOTH ceilings, so the family rows barely move above two adults. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "1177.se hoegkostnadsskydd for oeppenvaard and for laekemedel + LIF national prescription co-payment aggregate SEK 8.7bn over ~8.6m adults + Region Stockholm SEK 275 visit fee", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Trafikanalys, RVU Sverige 2011-2014, Figur 3.4. https://www.trafa.se/globalassets/statistik/resvanor/2009-2015/rvu-sverige-2011-2014.pdf", "last_updated": "2026-08-08", "confidence": "low", "statistic": "Arbetsresornas genomsnittliga restid", "year": "2011-2014", "note": "NATIONAL figure adopted for want of a city series, and the weakest of the four Swedish rows. GEOGRAPHY CALL: treated as its own city rather than a Stockholm suburb (SKL class storre stad). RVU says storre stader run BELOW the national mean, which argues lower, but Uppsala sends heavy rail commuting to Stockholm, which argues higher." }, "non_cash.sick_leave": { "source": "Lag (1991:1047) om sjuklön (\"sjuklönelagen\") for the employer phase; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.forsakringskassan.se/privatperson/e-tjanster-blanketter-och-informationsmaterial/aktuella-belopp", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Skatteverket, Avdrag foer pensionssparande https://www.skatteverket.se/privat/skatter/arbeteochinkomst/avdragforprivatpersoner/np.4.5fc8c94513259a4ba1d800042822.html ; 2026 prisbasbelopp https://www.regeringen.se/pressmeddelanden/2025/09/prisbasbelopp-for-2026-faststallt/", "last_updated": "2026-08-11", "confidence": "high", "note": "No deduction for our ICP. Skatteverket restricts the deduction to someone who \"helt saknar pensionsraett i sin anstaellning\", i.e. every employment they held in the income year lacked any right to tjaenstepension. The 35% / 10 prisbasbelopp rule (SEK 592,000 for 2026) reaches only that population; a salaried tech worker with occupational pension gets nothing. General deductibility was abolished in 2016. This is the single largest correction in the sweep." } }, "bergen": { "tax.inbound_regime": { "source": "PAYE / kildeskatt pa lonn, skatteloven kapittel 20. Flat 25% that INCLUDES trygdeavgift (Skatteetaten rettskilder: 'Denne skattesatsen inkluderer trygdeavgift'); first year of tax residency only; 2026 income ceiling NOK 725,050 (2025: NOK 697,150). Skatteloven s 20-3 makes the ceiling a hard cliff on the WHOLE income, not a band; s 20-2 bans every deduction. https://www.skatteetaten.no/person/utenlandsk/skal-du-arbeide-i-norge/skattekort/kildeskatt-pa-lonn/ ; https://lovdata.no/dokument/NL/lov/1999-03-26-14/KAPITTEL_20", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: max_income refreshed 697,150 (2025) -> 725,050 (2026). Still salary_relief false and no flat_rate, so inboundRegimeStatus returns note_only at every salary and no math moves. The note-only stance is now doubly defensible: professional salaries clear the cliff, and below it the total deduction ban frequently makes the scheme worse than ordinary tax for a household with a Norwegian mortgage. The forskrift under s 20-7 that sets the 2026 ceiling was not retrieved; the figure is published by Skatteetaten in both language versions." }, "tax.filing_basis": { "source": "Skatteetaten (skatteetaten.no) - spouses assessed individually on employment income (tax classes abolished 2018)", "last_updated": "2026-07-11", "note": "Activates the per-filer split in compute() for two-earner households." }, "tax.property_tax_rate_pct": { "source": "Bergen kommune eiendomsskatt 2025 = 0.27% with bunnfradrag NOK 750k", "last_updated": "2026-05-16", "confidence": "high", "note": "Bergen charges eiendomsskatt unlike Oslo (abolished)" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Bergen: NOK 683,003, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "NOK 683,003 / 9.49 = USD 71,971; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2879, R5-rounded down to 0.28. REPLACES 0.52 (2026-08-11), and the interim 0.36 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.52 was 0.9123 of Oslo's 0.57; that ratio x Oslo's re-derived 0.4 = 0.3649, R5-rounded down to 0.36. Oslo's own figure: NOK 968,721 / 9.49 = USD 102,078, / USD 250,000 = 0.4083. This is a BASIS move, not a re-measurement: Bergen's rank inside Norway is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE ~5-10% below Oslo per Tekna/Kode24 surveys Rebased 2026-08-11 with its country anchor (comp-index sweep A moved Oslo, Stockholm, Tokyo and Singapore); the documented ratio to the anchor is preserved, so this note's relative claim still holds." }, "housing._summary": { "source": "Numbeo Housing (cost-of-living basket) https://www.numbeo.com/cost-of-living/in/Bergen", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation (90/10 quality bar). Per-field values were researched by subagents from local marketplaces (Zillow/Hemnet/Funda/Rightmove/etc.); cite each field individually at next refresh pass." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Universal (folketrygden)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation." }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Bergen + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Bergen", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Bergen", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "NAV barnetrygd", "last_updated": "2026-05-16", "confidence": "medium" }, "non_cash._summary": { "source": "Ferieloven + NAV foreldrepenger/sykepenger", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Vacation, holidays, sick leave, parental leave per country." }, "tax.national_insurance_rate": { "confidence": "high", "source": "Trygdeavgift salary rate 7.6% 2026 (skatteetaten.no/en/rates/national-insurance-contributions); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "tax.trinnskatt_brackets": { "confidence": "high", "source": "Trinnskatt 2026 thresholds/rates (skatteetaten.no/en/rates/bracket-tax); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "childcare.infant_monthly": { "source": "regjeringen.no / Oslo kommune barnehage maksimalpris 2026 (Stortinget vedtatt)", "last_updated": "2026-07-17", "note": "2026 barnehage maksimalpris NOK 1,200/mo charged 11 months/yr + matpenger ~NOK 320/mo: effective (1,200 x 11/12) + 320 ~= NOK 1,420/mo. One age-independent cap covers infants and preschoolers (was 3,000, the pre-2024 level)." }, "childcare.preschool_monthly": { "source": "regjeringen.no / Oslo kommune barnehage maksimalpris 2026", "last_updated": "2026-07-17", "note": "Same national maksimalpris as infants: effective ~NOK 1,420/mo (maksimalpris NOK 1,200 x11 + matpenger ~320)." }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "Norwegian barnetrygd: universal, NOT means-tested." }, "tax.personfradrag": { "source": "Personfradrag 2026 NOK 114,540: regjeringen.no statsbudsjettet 2026 (108,550 -> 114,540) / skatteetaten.no/en/rates/personal-allowance", "source_name": "Skatteetaten / statsbudsjettet 2026", "last_updated": "2026-07-23", "confidence": "high", "note": "Applied INSIDE computeNorwayTaxes to the 22% alminnelig-inntekt base ONLY, together with minstefradrag. Per filer (Norway files individually). standard_deduction_* held at 0 so compute() does not subtract a lump across all three components. Correct-first item #1, 2026-07-17 resolutions doc." }, "tax.minstefradrag_rate": { "source": "statsbudsjettet 2026 (regjeringen.no) + skatteetaten.no 2026 rates: minstefradrag wage rate 46% (skatteetaten.no/en/rates/minimum-standard-deduction)", "source_name": "Skatteetaten", "last_updated": "2026-07-23", "confidence": "high", "note": "Minstefradrag = 46% of wage income, applied to the 22% base only. Capped (see minstefradrag_cap)." }, "tax.minstefradrag_cap": { "source": "statsbudsjettet 2026 (regjeringen.no) + skatteetaten.no 2026 rates: minstefradrag wage upper limit NOK 95,700 for 2026 (2025: 92,000)", "source_name": "Skatteetaten / statsbudsjettet 2026", "last_updated": "2026-07-23", "confidence": "high", "note": "Upper limit on the wage minstefradrag. Binds at any ICP salary. Sum with personfradrag = 210,240 on the 22% base." }, "tax.standard_deduction_single": { "source": "statsbudsjettet 2026 (regjeringen.no) + skatteetaten.no 2026 rates", "source_name": "Skatteetaten", "last_updated": "2026-07-23", "confidence": "high", "note": "Held at 0 BY DESIGN. Norway applies minstefradrag + personfradrag inside computeNorwayTaxes, to the 22% base only; a lump here would wrongly shrink trinnskatt and trygdeavgift too (the pre-2026-07-23 bug). See tax.personfradrag." }, "tax.standard_deduction_married": { "source": "statsbudsjettet 2026 (regjeringen.no) + skatteetaten.no 2026 rates", "source_name": "Skatteetaten", "last_updated": "2026-07-23", "confidence": "high", "note": "Held at 0 BY DESIGN. Norway applies minstefradrag + personfradrag inside computeNorwayTaxes, to the 22% base only; a lump here would wrongly shrink trinnskatt and trygdeavgift too (the pre-2026-07-23 bug). See tax.personfradrag." }, "non_cash.parental_leave_weeks": { "source": "NAV https://www.nav.no/foreldrepenger ; folketrygdloven kap. 14", "last_updated": "2026-08-06", "confidence": "high", "note": "49 wk at 100% RE-CONFIRMED 2026-08-06 against NAV: 3 wk reserved for the mother pre-birth + modrekvote 15 + fedrekvote 15 + fellesperiode 16, ONE family entitlement with quotas inside it, counted once. The 80% option is now 61 wk 1 day (extended from the old 59 so the two elections carry near-equal total value); 49 at 100 is the default and conservative pick. REFUTATION recorded: the father's 2 omsorgspermisjon weeks at birth are statutorily unpaid (employer payment is practice, not statute) and add no paid weeks; fathers may start fedrekvote at birth but that draws down the 49. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md." }, "non_cash.parental_leave_paid_pct": { "source": "NAV https://www.nav.no/foreldrepenger : calculation basis capped at 6G; G = NOK 136,549 from 1 May 2026, 6G = NOK 819,294", "last_updated": "2026-08-17", "confidence": "high", "note": "We store 0: the 2026-08-17 caps extension moved all 49 weeks into non_cash.parental_leave_capped (100% of a basis capped at 6G), so no percentage-paid weeks remain on this field. Was 100 with the cap recorded here but unconsumed, which left /rankings/parental-leave-by-country/ printing Norway at a flat 100% at any salary (raised 2026-08-08 and 2026-08-11, fixed cluster leave-caps-0817)." }, "non_cash.parental_leave_capped": { "source": "folketrygdloven para 14-7 (foreldrepenger basis follows the sykepenger rules, capped at 6G per para 8-10); NAV https://www.nav.no/grunnbelopet : G = NOK 136,549 per 1 May 2026, 6G = NOK 819,294/yr (fetched 2026-08-17)", "last_updated": "2026-08-17", "confidence": "high", "note": "All 49 weeks pay 100% of a basis capped at 6G = NOK 819,294/yr. cap_amount is stored per week (819,294 / 52 = NOK 15,755.65) on the engine weekly-gross convention, the same pre-conversion Stockholm uses for its annual SGI ceiling; foreldrepenger dagsats runs Mon-Fri (5-day week, divisor 260), so annual/52 equals dagsats x 5 exactly. Above 6G the payment is flat: effective rate = 819,294 / salary, so the stored Oslo tech p50 of NOK 968,721 yields 84.6% and a $150,000-equivalent (about NOK 1.41M) yields about 58%. The 80% election (61 wk 1 day) is deliberately not stored: 49 wk at 100% is the default and the conservative pick, per the 2026-08-06 batch2 adjudication. UNIT WATCH: G is an annual figure revised every May 1; re-derive cap_amount each trygdeoppgjor. Research: docs/superpowers/research/2026-08-17-parental-leave-caps.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Helsedirektoratet and Helfo egenandelstak + NAV, 1,452,000 residents reached the combined tak in 2023", "last_updated": "2026-08-08", "confidence": "medium", "note": "REPLACES an exact statutory maximum: NOK 3,040 is precisely the 2023 egenandelstak (series 2,460 / 2,921 / 3,040 / 3,165 / 3,278). 26% of the population reaches the cap and ~1m under-16s pay zero, so the realized adult mean is ~NOK 1,650. Children pay no egenandel under 16, extended to all under-18s from 2026-08-01. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Helsedirektoratet and Helfo egenandelstak + NAV, 1,452,000 residents reached the combined tak in 2023", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Jernbanedirektoratet / RVU, Reisevaner i de 7 storste byregionene 2022, Figur 23, Bergensregionen. https://www.jernbanedirektoratet.no/content/uploads/sites/4/2023/11/Reisevaner-i-de-7-storste-byregionene-med-tilleggsutvalg-2022.pdf", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Reiselengde i tid til/fra arbeid", "year": "2022", "note": "n=2782. Band-to-mean reconstruction gives 27.7." }, "non_cash.sick_leave": { "source": "Folketrygdloven (National Insurance Act) kapittel 8 , §8-2 (opptjeningstid), §8-10; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.nav.no/grunnbelopet", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Skatteetaten, Forskuddsutskrivingen 2026 https://www.skatteetaten.no/en/rettskilder/type/uttalelser/uttalelser/forskuddsutskrivingen-2026/", "last_updated": "2026-08-11", "confidence": "high", "note": "IPS (individuell pensjonssparing), NOK 25,000 for 2026, raised from 15,000 under the 2024 pension settlement; old-IPS and new-IPS share one ceiling. Employee-only, and the source states the same figure as both the payment limit and the deduction limit. KNOWN RESIDUAL, not fixed here: the IPS deduction is against alminnelig inntekt (22% flat) ONLY and must not reduce trinnskatt or trygdeavgift, but the engine nets the retirement deduction off personinntekt upstream, so it still reaches too many bases." } }, "trondheim": { "tax.inbound_regime": { "source": "PAYE / kildeskatt pa lonn, skatteloven kapittel 20. Flat 25% that INCLUDES trygdeavgift (Skatteetaten rettskilder: 'Denne skattesatsen inkluderer trygdeavgift'); first year of tax residency only; 2026 income ceiling NOK 725,050 (2025: NOK 697,150). Skatteloven s 20-3 makes the ceiling a hard cliff on the WHOLE income, not a band; s 20-2 bans every deduction. https://www.skatteetaten.no/person/utenlandsk/skal-du-arbeide-i-norge/skattekort/kildeskatt-pa-lonn/ ; https://lovdata.no/dokument/NL/lov/1999-03-26-14/KAPITTEL_20", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: max_income refreshed 697,150 (2025) -> 725,050 (2026). Still salary_relief false and no flat_rate, so inboundRegimeStatus returns note_only at every salary and no math moves. The note-only stance is now doubly defensible: professional salaries clear the cliff, and below it the total deduction ban frequently makes the scheme worse than ordinary tax for a household with a Norwegian mortgage. The forskrift under s 20-7 that sets the 2026 ceiling was not retrieved; the figure is published by Skatteetaten in both language versions." }, "tax.filing_basis": { "source": "Skatteetaten (skatteetaten.no) - spouses assessed individually on employment income (tax classes abolished 2018)", "last_updated": "2026-07-11", "note": "Activates the per-filer split in compute() for two-earner households." }, "tax.property_tax_rate_pct": { "source": "Trondheim kommune eiendomsskatt ~0.30%", "last_updated": "2026-05-16", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Trondheim: NOK 704,082, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "NOK 704,082 / 9.49 = USD 74,192; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2968, R5-rounded down to 0.29. REPLACES 0.51 (2026-08-11), and the interim 0.35 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. University town, sample may skew junior. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.51 was 0.8947 of Oslo's 0.57; that ratio x Oslo's re-derived 0.4 = 0.3579, R5-rounded down to 0.35. Oslo's own figure: NOK 968,721 / 9.49 = USD 102,078, / USD 250,000 = 0.4083. This is a BASIS move, not a re-measurement: Trondheim's rank inside Norway is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Deep research talent but smaller commercial tech than Oslo Rebased 2026-08-11 with its country anchor (comp-index sweep A moved Oslo, Stockholm, Tokyo and Singapore); the documented ratio to the anchor is preserved, so this note's relative claim still holds." }, "housing._summary": { "source": "Numbeo Housing (cost-of-living basket) https://www.numbeo.com/cost-of-living/in/Trondheim", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation (90/10 quality bar). Per-field values were researched by subagents from local marketplaces (Zillow/Hemnet/Funda/Rightmove/etc.); cite each field individually at next refresh pass." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Universal (folketrygden)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation." }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Trondheim + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Trondheim", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Trondheim", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "NAV barnetrygd", "last_updated": "2026-05-16", "confidence": "medium" }, "non_cash._summary": { "source": "Ferieloven + NAV foreldrepenger/sykepenger", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Vacation, holidays, sick leave, parental leave per country." }, "tax.national_insurance_rate": { "confidence": "high", "source": "Trygdeavgift salary rate 7.6% 2026 (skatteetaten.no/en/rates/national-insurance-contributions); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "tax.trinnskatt_brackets": { "confidence": "high", "source": "Trinnskatt 2026 thresholds/rates (skatteetaten.no/en/rates/bracket-tax); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "childcare.infant_monthly": { "source": "regjeringen.no / Oslo kommune barnehage maksimalpris 2026 (Stortinget vedtatt)", "last_updated": "2026-07-17", "note": "2026 barnehage maksimalpris NOK 1,200/mo charged 11 months/yr + matpenger ~NOK 320/mo: effective (1,200 x 11/12) + 320 ~= NOK 1,420/mo. One age-independent cap covers infants and preschoolers (was 3,000, the pre-2024 level)." }, "childcare.preschool_monthly": { "source": "regjeringen.no / Oslo kommune barnehage maksimalpris 2026", "last_updated": "2026-07-17", "note": "Same national maksimalpris as infants: effective ~NOK 1,420/mo (maksimalpris NOK 1,200 x11 + matpenger ~320)." }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "Norwegian barnetrygd: universal, NOT means-tested." }, "tax.personfradrag": { "source": "Personfradrag 2026 NOK 114,540: regjeringen.no statsbudsjettet 2026 (108,550 -> 114,540) / skatteetaten.no/en/rates/personal-allowance", "source_name": "Skatteetaten / statsbudsjettet 2026", "last_updated": "2026-07-23", "confidence": "high", "note": "Applied INSIDE computeNorwayTaxes to the 22% alminnelig-inntekt base ONLY, together with minstefradrag. Per filer (Norway files individually). standard_deduction_* held at 0 so compute() does not subtract a lump across all three components. Correct-first item #1, 2026-07-17 resolutions doc." }, "tax.minstefradrag_rate": { "source": "statsbudsjettet 2026 (regjeringen.no) + skatteetaten.no 2026 rates: minstefradrag wage rate 46% (skatteetaten.no/en/rates/minimum-standard-deduction)", "source_name": "Skatteetaten", "last_updated": "2026-07-23", "confidence": "high", "note": "Minstefradrag = 46% of wage income, applied to the 22% base only. Capped (see minstefradrag_cap)." }, "tax.minstefradrag_cap": { "source": "statsbudsjettet 2026 (regjeringen.no) + skatteetaten.no 2026 rates: minstefradrag wage upper limit NOK 95,700 for 2026 (2025: 92,000)", "source_name": "Skatteetaten / statsbudsjettet 2026", "last_updated": "2026-07-23", "confidence": "high", "note": "Upper limit on the wage minstefradrag. Binds at any ICP salary. Sum with personfradrag = 210,240 on the 22% base." }, "tax.standard_deduction_single": { "source": "statsbudsjettet 2026 (regjeringen.no) + skatteetaten.no 2026 rates", "source_name": "Skatteetaten", "last_updated": "2026-07-23", "confidence": "high", "note": "Held at 0 BY DESIGN. Norway applies minstefradrag + personfradrag inside computeNorwayTaxes, to the 22% base only; a lump here would wrongly shrink trinnskatt and trygdeavgift too (the pre-2026-07-23 bug). See tax.personfradrag." }, "tax.standard_deduction_married": { "source": "statsbudsjettet 2026 (regjeringen.no) + skatteetaten.no 2026 rates", "source_name": "Skatteetaten", "last_updated": "2026-07-23", "confidence": "high", "note": "Held at 0 BY DESIGN. Norway applies minstefradrag + personfradrag inside computeNorwayTaxes, to the 22% base only; a lump here would wrongly shrink trinnskatt and trygdeavgift too (the pre-2026-07-23 bug). See tax.personfradrag." }, "non_cash.parental_leave_weeks": { "source": "NAV https://www.nav.no/foreldrepenger ; folketrygdloven kap. 14", "last_updated": "2026-08-06", "confidence": "high", "note": "49 wk at 100% RE-CONFIRMED 2026-08-06 against NAV: 3 wk reserved for the mother pre-birth + modrekvote 15 + fedrekvote 15 + fellesperiode 16, ONE family entitlement with quotas inside it, counted once. The 80% option is now 61 wk 1 day (extended from the old 59 so the two elections carry near-equal total value); 49 at 100 is the default and conservative pick. REFUTATION recorded: the father's 2 omsorgspermisjon weeks at birth are statutorily unpaid (employer payment is practice, not statute) and add no paid weeks; fathers may start fedrekvote at birth but that draws down the 49. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md." }, "non_cash.parental_leave_paid_pct": { "source": "NAV https://www.nav.no/foreldrepenger : calculation basis capped at 6G; G = NOK 136,549 from 1 May 2026, 6G = NOK 819,294", "last_updated": "2026-08-17", "confidence": "high", "note": "We store 0: the 2026-08-17 caps extension moved all 49 weeks into non_cash.parental_leave_capped (100% of a basis capped at 6G), so no percentage-paid weeks remain on this field. Was 100 with the cap recorded here but unconsumed, which left /rankings/parental-leave-by-country/ printing Norway at a flat 100% at any salary (raised 2026-08-08 and 2026-08-11, fixed cluster leave-caps-0817)." }, "non_cash.parental_leave_capped": { "source": "folketrygdloven para 14-7 (foreldrepenger basis follows the sykepenger rules, capped at 6G per para 8-10); NAV https://www.nav.no/grunnbelopet : G = NOK 136,549 per 1 May 2026, 6G = NOK 819,294/yr (fetched 2026-08-17)", "last_updated": "2026-08-17", "confidence": "high", "note": "All 49 weeks pay 100% of a basis capped at 6G = NOK 819,294/yr. cap_amount is stored per week (819,294 / 52 = NOK 15,755.65) on the engine weekly-gross convention, the same pre-conversion Stockholm uses for its annual SGI ceiling; foreldrepenger dagsats runs Mon-Fri (5-day week, divisor 260), so annual/52 equals dagsats x 5 exactly. Above 6G the payment is flat: effective rate = 819,294 / salary, so the stored Oslo tech p50 of NOK 968,721 yields 84.6% and a $150,000-equivalent (about NOK 1.41M) yields about 58%. The 80% election (61 wk 1 day) is deliberately not stored: 49 wk at 100% is the default and the conservative pick, per the 2026-08-06 batch2 adjudication. UNIT WATCH: G is an annual figure revised every May 1; re-derive cap_amount each trygdeoppgjor. Research: docs/superpowers/research/2026-08-17-parental-leave-caps.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Helsedirektoratet and Helfo egenandelstak + NAV, 1,452,000 residents reached the combined tak in 2023", "last_updated": "2026-08-08", "confidence": "medium", "note": "REPLACES an exact statutory maximum: NOK 3,040 is precisely the 2023 egenandelstak (series 2,460 / 2,921 / 3,040 / 3,165 / 3,278). 26% of the population reaches the cap and ~1m under-16s pay zero, so the realized adult mean is ~NOK 1,650. Children pay no egenandel under 16, extended to all under-18s from 2026-08-01. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Helsedirektoratet and Helfo egenandelstak + NAV, 1,452,000 residents reached the combined tak in 2023", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Jernbanedirektoratet / RVU, Reisevaner i de 7 storste byregionene 2022, Figur 23, Trondheimsregionen. https://www.jernbanedirektoratet.no/content/uploads/sites/4/2023/11/Reisevaner-i-de-7-storste-byregionene-med-tilleggsutvalg-2022.pdf", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Reiselengde i tid til/fra arbeid", "year": "2022", "note": "n=2751. Band-to-mean gives 25.0. Extraction validated: the report's national row matches the independent RVU 2024 report's 2022 row." }, "non_cash.sick_leave": { "source": "Folketrygdloven (National Insurance Act) kapittel 8 , §8-2 (opptjeningstid), §8-10; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.nav.no/grunnbelopet", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Skatteetaten, Forskuddsutskrivingen 2026 https://www.skatteetaten.no/en/rettskilder/type/uttalelser/uttalelser/forskuddsutskrivingen-2026/", "last_updated": "2026-08-11", "confidence": "high", "note": "IPS (individuell pensjonssparing), NOK 25,000 for 2026, raised from 15,000 under the 2024 pension settlement; old-IPS and new-IPS share one ceiling. Employee-only, and the source states the same figure as both the payment limit and the deduction limit. KNOWN RESIDUAL, not fixed here: the IPS deduction is against alminnelig inntekt (22% flat) ONLY and must not reduce trinnskatt or trygdeavgift, but the engine nets the retirement deduction off personinntekt upstream, so it still reaches too many bases." } }, "dublin": { "non_cash.parental_leave_flat": { "source": "Citizens Information and gov.ie: Maternity Benefit, Paternity Benefit and Parent's Benefit all pay EUR 299 per week in 2026 https://www.gov.ie/en/department-of-social-protection/services/maternity-benefit/ ; https://www.citizensinformation.ie/en/social-welfare/families-and-children/parents-benefit/", "last_updated": "2026-08-01", "confidence": "medium", "note": "EUR 299 per week across all 46 weeks. Confidence is medium on the commencement DATE rather than the amount: both government sources give 2026 or currently but no SI-level commencement day could be pinned. Excluded from the stored amount: increases for a qualified adult or child, which are income-tested on partner earnings and taper to zero above EUR 400 per week, so our ICP fails them." }, "non_cash.parental_leave_paid_pct": { "source": "Citizens Information: Maternity, Paternity and Parent's Benefit are all paid at a flat weekly rate https://www.citizensinformation.ie/en/social-welfare/families-and-children/maternity-benefit/", "last_updated": "2026-08-01", "confidence": "high", "note": "0 because Ireland has NO percentage-paid weeks. Every one of the 46 weeks is a flat cash amount, held in non_cash.parental_leave_flat, and the rate a household actually receives is derived at its own salary. The stored 36 was a fabricated headline percentage for a benefit that has none. Never display this 0 on its own." }, "non_cash.parental_leave_weeks": { "source": "Citizens Information Maternity Benefit https://www.citizensinformation.ie/en/social-welfare/families-and-children/maternity-benefit/ , Paternity Benefit https://www.citizensinformation.ie/en/social-welfare/families-and-children/paternity-benefit/ , Parent's Benefit https://www.citizensinformation.ie/en/social-welfare/families-and-children/parents-benefit/ ; unpaid parental leave https://www.citizensinformation.ie/en/employment/employment-rights-and-conditions/leave-and-holidays/parental-leave/", "last_updated": "2026-08-01", "confidence": "high", "note": "46 PAID wk = Maternity Benefit 26 + Paternity Benefit 2 + Parent's Benefit 9 per parent x 2. Was 61, which INCLUDED unpaid weeks and violated the field definition. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. gov.ie states Parent's Benefit is in addition to the existing Maternity Benefit and Paternity Benefit, so there is no overlap to net off, and Both parents have an equal separate entitlement to parent's leave. You cannot transfer your entitlement to your partner, which is why the 9 weeks are summed. Parent's Benefit remained at 9 weeks after Budget 2026. Excluded: the 16-week unpaid maternity extension and 26 unpaid weeks per parent of parental leave, which are what inflated the stored 61; employer top-ups." }, "tax.inbound_regime": { "source": "SARP, TCA 1997 s825C. Revenue Tax and Duty Manual Part 34-00-10 (current version per eBrief 046/26 of 2026-03-05): for arrivals in 2026 to 2030 the relief disregards 30% of income between a EUR 125,000 lower threshold and a EUR 1,000,000 upper threshold, income tax only, USC and PRSI unaffected. Floor raised from EUR 100,000 and scheme extended to 2030 arrivals by s23 Finance Act 2025, commencing 2026-01-01, with earlier arrivals grandfathered on the older floor. Requires full-time employment with the same employer or group OUTSIDE Ireland for at least 6 months immediately before assignment. https://www.revenue.ie/en/tax-professionals/tdm/income-tax-capital-gains-tax-corporation-tax/part-34/34-00-10.pdf", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: all four carried values verified correct for 2026 (note-only, min 125,000, max 1,000,000, 5 years); the EUR 125,000 floor is confirmed as the Finance Act 2025 figure for arrivals from 2026-01-01, not a projection. Note-only kept, and it is the conservative call: SARP fails the 6-month prior intra-group employment test for our ICP, who takes a new job in a new country rather than an intra-group assignment. Trap now stated in the note: the EUR 125,000 eligibility test runs on BASE salary, excluding bonus, commission, BIK and share-based pay, while the relief formula itself includes them, so a total-comp input over-qualifies people at the door. Also new from 2026-01-01: employer certification on Form SARP 1A after 90 but within 180 days costs the claimant the first year." }, "tax.paye_brackets": { "source": "Revenue.ie tax-relief-charts (SRCOP €44k single 2025/26)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.usc_brackets": { "source": "Citizens Information / Revenue.ie USC 2026 (Budget 2026: 2% ceiling 28,700, rates 0.5/2/3/8 unchanged) https://www.citizensinformation.ie/en/money-and-tax/tax/income-tax/universal-social-charge/", "last_updated": "2026-08-07", "confidence": "high", "note": "Bands verified current for 2026. Since 2026-08-07 the engine charges USC on EACH SPOUSE separately even under joint assessment (USC is individual by statute, never aggregated) and on GROSS wages (USC carries no pension relief). The 11% surcharge applies to self-assessed non-PAYE income only - excluded." }, "tax.prsi_rate": { "source": "Gov.ie PRSI Class A employee 4.2% (in force to 30 Sep 2026; 4.35% from 1 Oct 2026) https://www.gov.ie/en/publication/aa0e46-prsi-pay-related-social-insurance/", "last_updated": "2026-08-07", "confidence": "high", "note": "Kept at 4.2% for the 2026 annual model (statutory for 9 of 12 months; the Oct 1 step rides docs/data-refresh-calendar.md). Uncapped; charged on gross wages (no pension relief), per earner (== joint, no cap)." }, "tax.sales_tax_rate": { "source": "Revenue.ie VAT 23% standard", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Gov.ie LPT 0.0906% (2026 valuation period)", "last_updated": "2026-05-16", "confidence": "high" }, "housing._summary": { "source": "Numbeo Dublin May 2026 + Daft.ie", "last_updated": "2026-05-16", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4).", "last_updated": "2026-08-07", "confidence": "medium", "note": "District tiers skipped by decision (2026-08-07): Daft.ie and RTB publish only hedonic averages, never medians or quartiles, so a Dublin ladder would have to be labeled \"average\" (research §3/§9). Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "HSE universal + VHI/Laya/Irish Life supplemental (community-rated)", "last_updated": "2026-05-16", "confidence": "medium" }, "childcare._summary": { "source": "Pobal CCC 2025 + NCS universal subsidy", "last_updated": "2026-05-16", "confidence": "high" }, "food._summary": { "source": "Numbeo Dublin Markets + Restaurants", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "TFI LEAP monthly + AA Ireland motoring", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Dublin Sports & Leisure", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "Aer Lingus + United DUB-NYC direct + LHR/AMS connections", "last_updated": "2026-05-16", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Ireland Child Benefit Budget 2026", "last_updated": "2026-05-16", "confidence": "high", "note": "Irish Child Benefit (EUR 140/mo): universal, NOT means-tested." }, "non_cash._summary": { "source": "Citizens Information IE + Statutory Sick Pay Act 2024", "last_updated": "2026-05-16", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Dublin: EUR 109,953, held in labor_market.tech_salary_distribution (source_slug greater-dublin-area, as_of 2026-07-18, table confidence usable). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 109,953 / 0.867 = USD 126,820; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.5073, R5-rounded down to 0.5. REPLACES 0.78 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Dublin by 56%. The measured p50 was 0.650 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): FAANG/Stripe €200-280k; non-FAANG €100-160k. Index 0.78" }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "Irish NCS universal subsidy (per-hour, not means-tested) applies to all; the income-assessed top-up dies at high income. Modeled value is the universal NCS." }, "childcare.after_school_monthly": { "source": "childcare accuracy pass 2026-07-21", "last_updated": "2026-07-21", "note": "Dublin parent report 2026-07-21: ~EUR 675/kid/mo gross for primary-school after-school care." }, "childcare.subsidy_after_school_annual": { "source": "childcare accuracy pass 2026-07-21", "last_updated": "2026-07-21", "note": "Irish NCS is hour-scaled: after-school care draws far fewer hours than full-day, so the subsidy is much smaller than the full-day NCS (which over-covered cheap after-school to near-free). Calibrated to a reported ~EUR 500/kid/mo net on ~EUR 675 gross. Not means-tested (universal NCS)." }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/greater-dublin-area", "last_updated": "2026-07-18", "confidence": "usable", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "tax.personal_credit": { "source": "Revenue.ie Personal Tax Credit https://www.revenue.ie/en/personal-tax-credits-reliefs-and-exemptions/personal-tax-credits/index.aspx (Budget 2026 left it unchanged)", "last_updated": "2026-08-07", "confidence": "high", "note": "EUR 2,000 single; DOUBLES to 4,000 for a jointly assessed married couple (engine applies x2 when has_partner). Split 2026-08-07 out of the old personal_tax_credit 4,000 bundle so married credits can be modeled correctly." }, "tax.employee_credit": { "source": "Revenue.ie Employee (PAYE) Tax Credit https://www.revenue.ie/en/personal-tax-credits-reliefs-and-exemptions/personal-tax-credits/employee-tax-credit/index.aspx (Budget 2026 unchanged)", "last_updated": "2026-08-07", "confidence": "high", "note": "EUR 2,000 PER INDIVIDUAL with PAYE income, non-transferable between spouses (a two-earner married couple gets two). The 20%-of-income cap binds only under EUR 10,000 - irrelevant at ICP incomes. Married credit totals: one-earner 6,000 (2x personal + 1 employee), two-earner 8,000. Home Carer credit (1,950) excluded: mutually exclusive with the two-income rate band the married PAYE array embodies (conservative)." }, "non_cash.vacation_days_statutory": { "source": "Organisation of Working Time Act 1997 s.19(1)(a) and s.19(6) https://revisedacts.lawreform.ie/eli/1997/act/20/section/19/revised/en/html", "last_updated": "2026-08-08", "confidence": "high", "note": "CORRECTED 2026-08-08, 22 to 20. s.19(1)(a) sets the entitlement at 4 working weeks for a full-time employee, and s.19(6) supplies the unit rule directly: a working week means 'the number of days that the employee concerned usually works in a week'. For the 5-day week we model that is 20 days. The 8%-of-hours limb in s.19(1)(c) is explicitly capped at the same 4 weeks. The Workplace Relations Commission states the same figure. The stored 22 was a common Irish employer package, not the statutory floor. Additive to the 10 public holidays, which are set separately by ss.21-22 and are current: St Brigid's Day was added as the 10th in 2023." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "HSE Drugs Payment Scheme (EUR 80/month per family) + public in-patient charges abolished 2023-04-17 + the EUR 100 emergency-department charge + published GP visit fees, with free GP care for under-8s only", "last_updated": "2026-08-08", "confidence": "medium", "note": "The one country in its batch where the stored value was too LOW at every household size. GP fees of EUR 50-70 a visit scale with headcount and are free only for under-8s, while the DPS cap is per family so the medicines component barely grows. Ireland has no overall ceiling: the DPS caps medicines and nothing else. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "HSE Drugs Payment Scheme (EUR 80/month per family) + public in-patient charges abolished 2023-04-17 + the EUR 100 emergency-department charge + published GP visit fees, with free GP care for under-8s only", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Central Statistics Office Ireland, Census 2022, table F7068. https://data.cso.ie/table/F7068", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Average journey time to work (minutes)", "year": "2022", "note": "Worker-weighted County Dublin 30.9; the wider Greater Dublin Area is 32.0. Confirmed WORK-ONLY, so no school trips are blended in." }, "non_cash.sick_leave": { "source": "Sick Leave Act 2022 (as amended; day count set by ministerial order per s.4).; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": "https://www.irishlegal.com/articles/government-u-turns-on-extension-of-paid-statutory-sick-leave", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Revenue, Tax relief limits on pension contributions https://www.revenue.ie/en/jobs-and-pensions/pension/relief/tax-relief-limits.aspx", "last_updated": "2026-08-11", "confidence": "high", "note": "Age-related percentage relief limit applied to the lesser of earnings and EUR 115,000, 2026. An explicit RELIEF ceiling and explicitly EMPLOYEE-ONLY: Revenue states employer contributions are not counted against the employee earnings threshold, so no less_employer_pct. Statutory bands are u30 15 / 30-39 20 / 40-49 25 / 50-54 30 / 55-59 35 / 60+ 40; every straddle resolves DOWN (25-34 takes 15, 35-44 takes 20, 45-54 takes 25, 55-64 takes 35) per the conservative value bias. Aggregate over ordinary contributions plus AVCs." } }, "singapore": { "non_cash.parental_leave_paid_pct": { "source": "MOM maternity-leave entitlement page: first 8 weeks at gross rate of pay, employer-funded (Employment Act descent, no reimbursement cap attached) https://www.mom.gov.sg/employment-practices/leave/maternity-leave/eligibility-and-entitlement", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the 8-week remainder, GPML weeks 1-8, the employer-funded Employment-Act block with no statutory ceiling. The other 22 weeks moved to non_cash.parental_leave_capped on 2026-08-06. The expat caveat stands: all three schemes require a Singapore-citizen child; a non-citizen child gets 12 Employment Act weeks of which 8 paid." }, "non_cash.parental_leave_weeks": { "source": "MOM Government-Paid Maternity Leave https://www.mom.gov.sg/employment-practices/leave/maternity-leave/eligibility-and-entitlement ; GPPL https://www.mom.gov.sg/employment-practices/leave/paternity-leave ; Shared Parental Leave https://www.mom.gov.sg/employment-practices/leave/shared-parental-leave and https://www.profamilyleave.msf.gov.sg/", "last_updated": "2026-08-01", "confidence": "high", "note": "30 wk = GPML 16 (mother) + GPPL 4 (father, mandatory for births from 1 Apr 2025) + Shared Parental Leave 10 (births from 1 Apr 2026). Was 16, maternity alone. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. SPL is a SHARED family pot with a default 50/50 split, so it is counted once and NOT doubled. MATERIAL EXPAT CAVEAT: all three government-paid schemes require the child to be a Singapore citizen at birth or within 12 months. A non-citizen child gets 12 Employment Act weeks, of which only the first 8 are paid." }, "tax.income_tax_brackets": { "source": "IRAS resident income tax YA 2024+ (top 24% above SGD 1M)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.cpf_rate": { "source": "CPF Board employee share (age ≤55)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.cpf_cap": { "source": "CPF Ordinary Wage ceiling SGD 8,000/mo from 2026-01-01 (taxsummaries.pwc.com/singapore); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "IRAS GST raised to 9% Jan 2024", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "IRAS owner-occupier progressive on Annual Value", "last_updated": "2026-05-16", "confidence": "medium", "note": "~0.10% effective on market for typical condo" }, "housing._summary": { "source": "Expat-SG.com + 99.co + Relocity (private condo, not HDB)", "last_updated": "2026-05-16", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "MOH/Medisave/Medishield Life + private IP (steeply age-rated)", "last_updated": "2026-05-16", "confidence": "medium" }, "childcare._summary": { "source": "ECDA 2026 + dollarsandsense.sg", "last_updated": "2026-05-16", "confidence": "medium", "note": "Subsidies for SC only; expats reduced/zero" }, "food._summary": { "source": "FairPrice/Cold Storage + Numbeo (~30% imported premium vs US)", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "LTA monthly pass SGD 128; car cost includes COE", "last_updated": "2026-05-16", "confidence": "high", "note": "COE SGD ~125k for 10-yr entitlement drives car cost" }, "discretionary._summary": { "source": "Numbeo Singapore + regional Asia travel", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "SIN-SEA/LAX one-stop (Singapore Airlines/ANA)", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "SG Baby Bonus + CDA + PTR annualized", "last_updated": "2026-05-16", "confidence": "medium", "note": "No recurring monthly allowance" }, "non_cash._summary": { "source": "MOM Employment Act 2024 + GPML 16 weeks 100% capped SGD 30k", "last_updated": "2026-05-16", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Singapore: SGD 113,096, held in labor_market.tech_salary_distribution (source_slug singapore, as_of 2026-07-18, table confidence rich). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "SGD 113,096 / 1.28 = USD 88,356; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3534, R5-rounded down to 0.35. REPLACES 0.69 (2026-08-11), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Singapore by 97%. The measured p50 was 0.512 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): SGD 175,408 / 1.28 = USD 137,038; / USD 196,000 Seattle senior reference = 0.699, R5-rounded down to 0.69. REPLACES 0.85 (2026-05-16), which reproduced the FAANG midpoint of its own note (SGD 240,000 / 1.28 / 220,000 = 0.852) and nothing else: the note's mid-tier band tops out at 0.72 under the most generous anchor. Peers that quoted two bands landed between them or below (Zurich moved DOWN 19% from its FAANG figure). 0.85 implied SGD 213,248, above Singapore's own measured tech p90 of 211,000. MOM Occupational Wage Survey gross monthly wage EXCLUDES employer CPF, bonuses and stock, and Singapore pays a 13th-month AWS, so MOM x12 understates total cash by 8-17% before equity; the MOM software-developer line could not be read from a primary table in this pass and is NOT load-bearing here. Working: docs/superpowers/research/2026-08-11-comp-index-sweep-a.md" }, "tax.filing_basis": { "source": "IRAS (iras.gov.sg) - individual taxation; no joint filing", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/singapore", "last_updated": "2026-07-18", "confidence": "rich", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "non_cash.parental_leave_capped": { "source": "Child Development Co-Savings Act 2001 s.9: \"no woman shall be entitled to any payment in excess of the amount of $20,000\" per 8-week block, CPF-inclusive (statute text via Wikisource mirror + an independent verbatim quotation; sso.agc.gov.sg 403d on every attempt) ; MOM 2026 figures: $10,000 per 4 weeks / $2,500 per week for GPML weeks 9-16, GPPL, SPL ; 3+ independent HR/payroll sources confirm employers are not legally required to top up above the cap. Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (Singapore) and docs/superpowers/research/2026-08-06-pl-caps-east-asia.md", "last_updated": "2026-08-06", "confidence": "medium-high", "note": "REFUTES this file's own 2026-08-01 guess that the SGD ceilings were reimbursement-only: CDCSA s.9 caps the EMPLOYEE's entitlement at the same figure as the employer reimbursement for GPML weeks 9-16, all of GPPL and all of SPL, 22 of the 30 stored weeks, at SGD 2,500/week. The cap binds at or below the LOW end of Singapore tech pay (weekly gross at SGD 130k/yr is ~2,500), so real replacement on the capped weeks runs ~46% at SGD 280k. Confidence medium-high on the statutory mechanism (mirror text, not the live portal), high on the current dollar figures (direct MOM fetch)." }, "non_cash.vacation_days_statutory": { "source": "Employment Act 1968 s.43 https://sso.agc.gov.sg/Act/EmA1968", "last_updated": "2026-08-08", "confidence": "high", "note": "CORRECTED 2026-08-08, 14 to 7. s.43 sets paid annual leave at 7 days for the first 12 months of service, rising one day per further year to a cap of 14. We had stored the CAP as if it were the floor. The declared basis is the year-1 legal floor, so 7. Open question left for the country owner: the Employment Act carves managers and executives out of some Part IV protections, which could matter for the people-manager end of our ICP; not resolved here because it needs a job-title policy call." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "MediShield Life deductibles and co-insurance, measured as cash out-of-pocket after MediSave", "last_updated": "2026-08-08", "confidence": "low", "note": "MediSave is FORCED SAVINGS taken through payroll and is deliberately excluded here, so it is not double-counted against the CPF treatment elsewhere in the engine. LEVEL NOT RE-VERIFIED in the 2026-08-08 sweep. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "MediShield Life deductibles and co-insurance, measured as cash out-of-pocket after MediSave", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. LTA statistics and master plans carry no commute-time series, and the SingStat Census of Population 2020 release is a JavaScript application that 404s on direct URLs." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Ministry of Finance, Supplementary Retirement Scheme https://www.mof.gov.sg/news-resources/supplementary-retirement-scheme/", "last_updated": "2026-08-11", "confidence": "high", "note": "SRS contribution cap, which is also the relief: every dollar contributed reduces assessable income. SGD 15,300 is the Citizen/PR figure; foreigners get SGD 35,700, and our mover usually IS a foreigner, but the field carries one max and cannot branch on sg_cpf_resident, so the house rule takes the lower. Unchanged since 1 Jan 2016. CPF is the separate mandatory scheme. The SGD 80,000 personal income tax relief cap does not bind (max CPF relief 20,400 + earned income relief 1,000 + SRS 15,300 = 36,700). CPF cash top-up relief deliberately not stacked." } }, "helsinki": { "tax.filing_basis": { "source": "Vero (vero.fi) - spouses taxed as individuals on earned income", "last_updated": "2026-07-11", "note": "Activates the per-filer split in compute() for two-earner households." }, "tax.state_tax_brackets": { "source": "Pre-reform-basis national scale coherent with the 18.5%-basis municipal rate (netsalaire.com/en/finland/income-tax; cross-checked countrytaxcalc.com, PwC); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high", "note": "The old post-reform scale (top 41.75/44%) double-counted the pre-reform municipal 18.5%. National tax at EUR 138k now ~27,628 (benchmark)." }, "tax.kunnallisvero_rate": { "source": "Helsinki kunnallisvero 18.5% 2025-2026", "last_updated": "2026-05-16", "confidence": "high", "note": "Church tax (1%) excluded — opt-in" }, "tax.social_insurance_rate": { "source": "Employee 2026: TyEL 7.30% + health insurance 1.98% + unemployment 0.89% = 10.17% (varma.fi social security contributions 2026 PDF); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "VAT 25.5% (raised from 24% Sept 2024)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Helsinki kiinteistövero residential 0.41-0.93%; midpoint 0.6%", "last_updated": "2026-05-16", "confidence": "high" }, "housing._summary": { "source": "Tilastokeskus + Vuokraovi/Oikotie + KTI", "last_updated": "2026-05-16", "confidence": "high", "note": "District heating typically included in rent" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Kela universal + THL public health (maksukatto ~€762/yr)", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "Opetushallitus maxtaxa €295/mo national cap", "last_updated": "2026-05-16", "confidence": "high" }, "food._summary": { "source": "Tilastokeskus + Numbeo Helsinki (K-Ruoka/S-Market)", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "HSL AB-zone monthly €71.30; Autoliitto motoring", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "Tilastokeskus leisure + Finnish vacation patterns (mokki/Lapland)", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "Finnair/KLM HEL-SEA via CPH/AMS", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Kela Lapsilisä (1st child €94.88/mo)", "last_updated": "2026-05-16", "confidence": "high", "note": "Finnish lapsilisa (child benefit): universal, NOT means-tested." }, "non_cash._summary": { "source": "Vuosilomalaki + 2022 family leave reform (320 paid days) + Kela sairauspäiväraha", "last_updated": "2026-05-16", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Helsinki: EUR 73,328, fetched live 2026-08-11 and recorded at sweep B F8d of docs/superpowers/research/2026-08-11-comp-index-sweep-b.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 73,328 / 0.867 = USD 84,577; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3383, R5-rounded down to 0.33. REPLACES 0.45 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Helsinki by 36%. The measured p50 was 0.752 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE €70-100k (Supercell/Wolt/Smartly.io). Index 0.45" }, "healthcare.statutory_employee": { "source": "vero.fi, Vakuutetun sairausvakuutusmaksu (guidance 338472): 2026 employee sairaanhoitomaksu 1.10% of earned income and paivarahamaksu 0.88% above an annual earned-income floor of about EUR 17,255, together the 1.98% employee sairausvakuutusmaksu. Corroborated by stm.fi, 'Sairausvakuutusmaksut nousevat hieman vuonna 2026'. Neither component carries a ceiling; the absence is consistent across both sources, whose purpose is to state the full calculation. Public treatment is delivered by the 21 hyvinvointialueet out of general taxation, not out of this contribution.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll. UNCAPPED rests on the consistent absence of a ceiling clause in two primary sources rather than on an explicit 'there is no cap' sentence. The 0.88% paivarahamaksu is deliberately excluded: it funds cash daily allowances, closer to disability insurance than to health care. THE FRAMING MATTERS MORE THAN THE NUMBER HERE: 1.10% is a genuinely small earmarked contribution, and a reader who takes it for the cost of Finnish health care will badly flatter Finland, so the note says outright where the money actually comes from. Also for 2026: the age-graded employee TyEL contribution is abolished and flat at all ages, and employee unemployment insurance rose from 0.59% to 0.89%." }, "non_cash.parental_leave_weeks": { "source": "Kela https://www.kela.fi/web/selkokieli/vanhempainpaivarahat (raskausraha 40 arkipaivaa; vanhempainraha 320 arkipaivaa per child, 160 per parent, up to 63 transferable); 2022 family-leave reform, births since 4 Sep 2022", "last_updated": "2026-08-06", "confidence": "high", "note": "60 wk = raskausraha 40 arkipaivaa (~6.5 wk, starting pre-birth, a benefit SEPARATE from and additional to vanhempainraha) + vanhempainraha 320 arkipaivaa (a family pot counted once: a single parent receives all 320) = 360 arkipaivaa at Kela's 6 payable days a week = 60. The stored 46 was the pre-2022 system (aitiysraha 105 + vanhempainraha 158 days). REFUTATION recorded: Duunitori's claim that raskausraha sits inside the birthing parent's 160-day quota contradicts Kela's own page, which lists them as distinct benefits; Kela wins. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md." }, "non_cash.parental_leave_paid_pct": { "source": "Kela vanhempainpaivarahat; sairausvakuutuslaki 11 luku. 2026 degressive formula: ~70% of annual income EUR 14,282-43,567, 40% of the slice to 67,113, 25% above, minimum EUR 31.99/day", "last_updated": "2026-08-06", "confidence": "medium-high", "note": "70 is the statutory headline. The formula is degressive rather than flat-capped, so no parental_leave_flat block: at a EUR 90,000 Helsinki salary the daily allowance is ~EUR 152, about 51% effective; the full 70 applies only below ~EUR 43,500. Raskausraha pays at the same daily-allowance level in the post-2022 system (medium confidence on that sub-point)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Kela maksukatto (EUR 815) and laeaekekatto (EUR 636.12) + OECD Finland out-of-pocket per capita, netted of dental (14%) and income-tested long-term-care client fees", "last_updated": "2026-08-08", "confidence": "medium", "note": "The old EUR 700 sat between Finland's two ceilings and modelled one. Held down deliberately by tyoeterveyshuolto, which gives most employed Finns employer-funded primary care free at the point of use, and our ICP is exactly that cohort. Finland's child rule is distinct in Europe: a minor's fees accrue to a PARENT's maksukatto rather than opening a new one, so a child adds real cost that then saturates against the parent's cap. Weakest country in its batch. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Kela maksukatto (EUR 815) and laeaekekatto (EUR 636.12) + OECD Finland out-of-pocket per capita, netted of dental (14%) and income-tested long-term-care client fees", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Traficom, Henkiloliikennetutkimus 2021, seuturaportti Helsingin seutu, weekday table. https://www.traficom.fi/files/media/file/HLT%202021%20seuturaportti%20Helsingin%20seutu.pdf", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Keskimaarainen matka-aika, tyo (average travel time per trip, work)", "year": "2021", "note": "Directly published mean, one way, all modes, 14-municipality metro region, n=3620. Work-only: tyoasia (business travel) and study are separate rows." }, "non_cash.sick_leave": { "source": "Työsopimuslaki (Employment Contracts Act, 55/2001) Chapter 2 Section 11 (employer; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Vero.fi, Pitkaaikaissaastamissopimuksen ja vapaaehtoisen yksiloellisen elaekevakuutuksen verotus https://www.vero.fi/syventavat-vero-ohjeet/ohje-hakusivu/48212/pitkaaikaissaastamissopimuksen_ja_vapaa/ ; VM, Vapaaehtoisen elaekesaastamisen verotuki poistuu vuonna 2027 https://vm.fi/-/vapaaehtoisen-elakesaastamisen-verotuki-poistuu-vuonna-2027", "last_updated": "2026-08-11", "confidence": "high", "note": "No earned-income deduction. The EUR 5,000/yr relief for vapaaehtoinen yksiloellinen elaekevakuutus and PS-sopimus is deducted from PAAOMATULOISTA, capital income; with no capital income it converts to the erityinen alijaeaemaehyvitys, a CREDIT worth about EUR 1,400. It never reduces the earned-income base, which is what this field represents. Abolished entirely from 2027. SEPARATE KNOWN GAP, out of scope: computeFinlandTaxes deducts nothing at all, including the mandatory TyEL and unemployment-insurance employee contributions, which ARE deductible in Finland and run the other way." } }, "sydney": { "non_cash.parental_leave_flat": { "source": "DSS Paid Parental Leave Guide 3.1: the daily NMW amount for PLP is 7.6 times the hourly NMW; from 1 July 2026 the hourly NMW is AUD 26.44, the daily amount is AUD 200.94 and the weekly rate is AUD 1,004.70 https://guides.dss.gov.au/paid-parental-leave-guide/3/1 ; Paid Parental Leave Act 2010 ss 6, 65, 11D; Fair Work Act 2009 Part 2-6", "last_updated": "2026-08-01", "confidence": "high", "note": "AUD 1,004.70 per week (AUD 200.94 per payable day x 5), in force from 1 July 2026 and reindexed each 1 July with the Fair Work Commission national minimum wage order. PLP is taxable. Annualised that is about AUD 52,244, which against a typical Australian tech salary is roughly a 29% replacement rate and falls as pay rises. TWO ITEMS DELIBERATELY NOT IN THIS FIGURE. First, the PPL superannuation contribution for births from 1 July 2025 adds 12% of the PLP amount plus nominal interest, paid by the Commissioner of Taxation into the parent super fund, which raises the cash value but adds no weeks. Second, an income-test CLIFF: a claimant qualifies if EITHER their own adjusted taxable income is at or under AUD 186,487 OR the couple combined is at or under AUD 386,525 (2025-26 reference year, indexed each 1 July). A household on AUD 250,000 qualifies on the family test however it splits, but two earners at AUD 200,000 each fail both tests and receive NOTHING, because PLP is all-or-nothing rather than tapered. ATI includes RSUs and reportable super, which pushes tech households toward that cliff. https://guides.dss.gov.au/paid-parental-leave-guide/2/3" }, "non_cash.parental_leave_paid_pct": { "source": "DSS Paid Parental Leave Guide 3.1, Current rate of PLP https://guides.dss.gov.au/paid-parental-leave-guide/3/1", "last_updated": "2026-08-01", "confidence": "high", "note": "0 because Australia has NO percentage-paid weeks. PLP is paid at the national minimum wage as a flat daily amount, so the stored 100 was a fabricated headline rate for a benefit that never tracked salary. The flat amount lives in non_cash.parental_leave_flat and the rate is derived at the household own salary. Never display this 0 on its own." }, "non_cash.parental_leave_weeks": { "source": "DSS Paid Parental Leave Guide 2.9, Maximum PLP entitlements (Guide v1.99, released 1 July 2026) https://guides.dss.gov.au/paid-parental-leave-guide/2/9 ; Paid Parental Leave Act 2010 ss 31AB, 31ABA", "last_updated": "2026-08-01", "confidence": "high", "note": "26 wk = 130 payable days at 5 payable weekdays per PLP week. Was 22, which is the 110-day rule for births from 1 July 2024, so the repo was two annual steps stale. The statutory ladder is 100 days from 1 Jul 2023, 110 from 1 Jul 2024, 120 from 1 Jul 2025 and 130 from 1 Jul 2026, and a child born today attracts 130. COUNTED ONCE, NOT SUMMED: PLP is a single family entitlement for the child (the maximum number of days that can be received for a child for all claimants for the child), and for a partnered claimant the 20 reserved days are carved OUT of the 130 rather than added on top, so two parents claiming together still receive 130 between them. Excluded: the 12 to 24 months of Fair Work Act unpaid parental leave; reserved-period exemptions for ADF deployment or a deceased partner. Dad and Partner Pay no longer exists separately, having been absorbed into PLP in July 2023." }, "tax.filing_basis": { "source": "PwC Worldwide Tax Summaries - Australia (Individual, Tax administration): no joint assessment or joint filing in Australia; spouses assessed individually.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "FY2026-27 stage: bracket-2 15% from 2026-07-01, drops to 14% 2027-07-01 (ato.gov.au personal income tax cuts); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.medicare_levy_rate": { "source": "ATO Medicare Levy 2% standard rate", "last_updated": "2026-05-16", "confidence": "high" }, "tax.sales_tax_rate": { "source": "ATO GST 10% (national)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "City of Sydney council rates + NSW emergency services levy (blended)", "last_updated": "2026-05-16", "confidence": "medium", "note": "NSW no annual state property tax on owner-occupied" }, "housing._summary": { "source": "Domain & CoreLogic Sydney rental/sales Q1 2026", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Medicare + privatehealth.gov.au PHI Bupa/Medibank 2026", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "Mitchell Institute 2025 + Care for Kids Sydney; CCS net", "last_updated": "2026-05-16", "confidence": "medium" }, "food._summary": { "source": "ABS HES 2024 + Numbeo Sydney 2026", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "Opal card weekly cap AUD 50; NRMA running costs 2024", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "ABS HES discretionary", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "Google Flights/Skyscanner SYD-SEA/LAX 2025", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Services Australia FTB Part A typical band", "last_updated": "2026-05-16", "confidence": "medium" }, "non_cash._summary": { "source": "Fair Work Act NES + Commonwealth PPL 22 weeks; BITRE Sydney commute", "last_updated": "2026-05-16", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Sydney: AUD 161,818, held in labor_market.tech_salary_distribution (source_slug sydney-aus, as_of 2026-07-18, table confidence usable). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "AUD 161,818 / 1.42 = USD 113,956; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.4558, R5-rounded down to 0.45. REPLACES 0.75 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Sydney by 67%. The measured p50 was 0.608 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Sydney is AU tech salary ceiling. Index 0.75" }, "childcare.subsidy_per_kid_annual": { "source": "servicesaustralia.gov.au Child Care Subsidy 2025-26 (90% up to 85,279; 0% at 535,279)", "last_updated": "2026-07-17", "note": "Child Care Subsidy at the solved household: CCS% = 90 - (income - 85,279)/5,000. At the median published sydney household (~491k -> 8.8%) applied to per-kid gross childcare. CCS is means-tested; it survives partially at these incomes (unlike FTB, which is 0). The audit's 36-48% guess assumed a lower household; the engine solves sydney higher. Was a flat over-covering 8,500-10,000." }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/sydney-aus", "last_updated": "2026-07-18", "confidence": "usable", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "healthcare.statutory_employee": { "source": "ATO, Medicare levy: 2% of taxable income with no upper income limit (Medicare Levy Act 1986). ATO Medicare levy surcharge income thresholds and rates, 2026-27: singles AUD 105,000 / 123,000 / 164,000 and families AUD 210,000 / 246,000 / 328,000 at 1% / 1.25% / 1.5%, with the family threshold rising AUD 1,500 per dependent child after the first (2025-26 was 101,000 / 118,000 / 158,000 and 202,000 / 236,000 / 316,000). The base levy is assessed individually; the surcharge tests combined income for couples.", "last_updated": "2026-07-31", "confidence": "medium", "note": "DISPLAY ONLY. NOT capped, and that is the load-bearing fact for Australia: it is one of very few systems in this dataset where the health contribution never flattens. BASE MISMATCH: the levy is charged on taxable income, while computeStatutoryHealthContribution applies the rate to gross, so the displayed figure runs a few percent above the charged one. That is the conservative direction and is the same base issue flagged for every country in this rollout. CAVEAT: ato.gov.au and legislation.gov.au returned 403 to every fetch attempt this session, so the 2% rate and the 2026-27 surcharge tiers were read from indexed snippets of the live ATO pages and cross-checked against two independent professional tax references, hence medium confidence. The 2% rate itself is long-standing and uncontested. 2026-27 Medicare levy LOW-INCOME thresholds were not confirmed (historically legislated retroactively); irrelevant at ICP income." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "AIHW health expenditure 2023-24 individual out-of-pocket, AUD 1,634 per person and ~1,338 excluding dental + PBS co-payment and the family-pooled PBS Safety Net", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 RAISED from 800 on primary-source evidence, the largest upward level move outside Austria. Medicare and the PBS are national with no state benefit variation, and the previous Brisbane-versus-rest spread had no documented basis in _meta or the city JSONs, so it is removed. Private hospital cover is a PREMIUM and stays in the premium field. Children add ~30% of an adult: frequently bulk-billed, lower PBS costs, and the family-pooled Safety Net pulls the whole household toward the concessional co-pay. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "AIHW health expenditure 2023-24 individual out-of-pocket, AUD 1,634 per person and ~1,338 excluding dental + PBS co-payment and the family-pooled PBS Safety Net", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "HILDA Survey (Melbourne Institute), average daily commuting time by capital city, 2017 wave. https://theconversation.com/australian-city-workers-average-commute-has-blown-out-to-66-minutes-a-day-how-does-yours-compare-120598", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average daily commuting time (round trip)", "year": "2017", "note": "CONVERSION: 71 min/day round trip / 2 = 35.5 one way, rounded up. Australia's longest." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "ATO, Contributions caps https://www.ato.gov.au/tax-rates-and-codes/key-superannuation-rates-and-thresholds/contributions-caps ; Maximum contributions base https://www.ato.gov.au/businesses-and-organisations/super-for-employers/paying-super-on-payday/what-payments-are-qualifying-earnings/maximum-contributions-base ; Super guarantee https://www.ato.gov.au/tax-rates-and-codes/key-superannuation-rates-and-thresholds/super-guarantee", "last_updated": "2026-08-11", "confidence": "high", "note": "Concessional contributions cap AUD 32,500 for 2026-27, a CONTRIBUTION ceiling shared with compulsory employer Super Guarantee, so the employee has only what SG leaves: 12% of gross up to the AUD 270,830 maximum contribution base (annual from 1 July 2026 under Payday Super). Room is zero at and above 270,830. The 5-year carry-forward is deliberately ignored. SEPARATE, UNFIXED: concessional contributions are taxed 15% inside the fund (30% under Division 293 above AUD 250,000), so treating the allowed portion as fully pre-tax still flatters Australia." } }, "melbourne": { "non_cash.parental_leave_flat": { "source": "DSS Paid Parental Leave Guide 3.1: the daily NMW amount for PLP is 7.6 times the hourly NMW; from 1 July 2026 the hourly NMW is AUD 26.44, the daily amount is AUD 200.94 and the weekly rate is AUD 1,004.70 https://guides.dss.gov.au/paid-parental-leave-guide/3/1 ; Paid Parental Leave Act 2010 ss 6, 65, 11D; Fair Work Act 2009 Part 2-6", "last_updated": "2026-08-01", "confidence": "high", "note": "AUD 1,004.70 per week (AUD 200.94 per payable day x 5), in force from 1 July 2026 and reindexed each 1 July with the Fair Work Commission national minimum wage order. PLP is taxable. Annualised that is about AUD 52,244, which against a typical Australian tech salary is roughly a 29% replacement rate and falls as pay rises. TWO ITEMS DELIBERATELY NOT IN THIS FIGURE. First, the PPL superannuation contribution for births from 1 July 2025 adds 12% of the PLP amount plus nominal interest, paid by the Commissioner of Taxation into the parent super fund, which raises the cash value but adds no weeks. Second, an income-test CLIFF: a claimant qualifies if EITHER their own adjusted taxable income is at or under AUD 186,487 OR the couple combined is at or under AUD 386,525 (2025-26 reference year, indexed each 1 July). A household on AUD 250,000 qualifies on the family test however it splits, but two earners at AUD 200,000 each fail both tests and receive NOTHING, because PLP is all-or-nothing rather than tapered. ATI includes RSUs and reportable super, which pushes tech households toward that cliff. https://guides.dss.gov.au/paid-parental-leave-guide/2/3" }, "non_cash.parental_leave_paid_pct": { "source": "DSS Paid Parental Leave Guide 3.1, Current rate of PLP https://guides.dss.gov.au/paid-parental-leave-guide/3/1", "last_updated": "2026-08-01", "confidence": "high", "note": "0 because Australia has NO percentage-paid weeks. PLP is paid at the national minimum wage as a flat daily amount, so the stored 100 was a fabricated headline rate for a benefit that never tracked salary. The flat amount lives in non_cash.parental_leave_flat and the rate is derived at the household own salary. Never display this 0 on its own." }, "non_cash.parental_leave_weeks": { "source": "DSS Paid Parental Leave Guide 2.9, Maximum PLP entitlements (Guide v1.99, released 1 July 2026) https://guides.dss.gov.au/paid-parental-leave-guide/2/9 ; Paid Parental Leave Act 2010 ss 31AB, 31ABA", "last_updated": "2026-08-01", "confidence": "high", "note": "26 wk = 130 payable days at 5 payable weekdays per PLP week. Was 22, which is the 110-day rule for births from 1 July 2024, so the repo was two annual steps stale. The statutory ladder is 100 days from 1 Jul 2023, 110 from 1 Jul 2024, 120 from 1 Jul 2025 and 130 from 1 Jul 2026, and a child born today attracts 130. COUNTED ONCE, NOT SUMMED: PLP is a single family entitlement for the child (the maximum number of days that can be received for a child for all claimants for the child), and for a partnered claimant the 20 reserved days are carved OUT of the 130 rather than added on top, so two parents claiming together still receive 130 between them. Excluded: the 12 to 24 months of Fair Work Act unpaid parental leave; reserved-period exemptions for ADF deployment or a deceased partner. Dad and Partner Pay no longer exists separately, having been absorbed into PLP in July 2023." }, "tax.filing_basis": { "source": "PwC Worldwide Tax Summaries - Australia (Individual, Tax administration): no joint assessment or joint filing in Australia; spouses assessed individually.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "FY2026-27 stage: bracket-2 15% from 2026-07-01, drops to 14% 2027-07-01 (ato.gov.au personal income tax cuts); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.medicare_levy_rate": { "source": "ATO Medicare Levy 2% standard rate", "last_updated": "2026-05-16", "confidence": "high" }, "tax.sales_tax_rate": { "source": "ATO GST 10% (national)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "City of Melbourne council rates + VIC fire services levy", "last_updated": "2026-05-16", "confidence": "medium" }, "housing._summary": { "source": "Domain Rental Report Q1 2026 + CoreLogic HVI Apr 2026", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Medicare + privatehealth.gov.au mid-tier 2026", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "ACCC Childcare Inquiry 2024 + VIC sector data; CCS net", "last_updated": "2026-05-16", "confidence": "medium" }, "food._summary": { "source": "Numbeo Melbourne 2026 + ABS SLCI", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "Myki Zone 1+2 daily cap; RACV Driving Your Dollars 2025", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "ABS HES discretionary", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "Skyscanner/Google Flights MEL-SEA/LAX 2026", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Services Australia FTB Part A typical band", "last_updated": "2026-05-16", "confidence": "medium" }, "non_cash._summary": { "source": "Fair Work Act NES + VIC public holidays (Melbourne Cup + AFL Grand Final Eve)", "last_updated": "2026-05-16", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Melbourne: AUD 144,559, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "AUD 144,559 / 1.42 = USD 101,802; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.4072, R5-rounded down to 0.4. REPLACES 0.72 (2026-08-11), and the interim 0.43 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.72 was 0.9600 of Sydney's 0.75; that ratio x Sydney's re-derived 0.45 = 0.4320, R5-rounded down to 0.43. Sydney's own figure: AUD 161,818 / 1.42 = USD 113,956, / USD 250,000 = 0.4558. This is a BASIS move, not a re-measurement: Melbourne's rank inside Australia is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Slightly below Sydney. Index 0.72" }, "childcare.subsidy_per_kid_annual": { "source": "servicesaustralia.gov.au Child Care Subsidy 2025-26 (90% up to 85,279; 0% at 535,279)", "last_updated": "2026-07-17", "note": "Child Care Subsidy at the solved household: CCS% = 90 - (income - 85,279)/5,000. At the median published melbourne household (~445k -> 18%) applied to per-kid gross childcare. CCS is means-tested; it survives partially at these incomes (unlike FTB, which is 0). The audit's 36-48% guess assumed a lower household; the engine solves melbourne higher. Was a flat over-covering 8,500-10,000." }, "healthcare.statutory_employee": { "source": "ATO, Medicare levy: 2% of taxable income with no upper income limit (Medicare Levy Act 1986). ATO Medicare levy surcharge income thresholds and rates, 2026-27: singles AUD 105,000 / 123,000 / 164,000 and families AUD 210,000 / 246,000 / 328,000 at 1% / 1.25% / 1.5%, with the family threshold rising AUD 1,500 per dependent child after the first (2025-26 was 101,000 / 118,000 / 158,000 and 202,000 / 236,000 / 316,000). The base levy is assessed individually; the surcharge tests combined income for couples.", "last_updated": "2026-07-31", "confidence": "medium", "note": "DISPLAY ONLY. NOT capped, and that is the load-bearing fact for Australia: it is one of very few systems in this dataset where the health contribution never flattens. BASE MISMATCH: the levy is charged on taxable income, while computeStatutoryHealthContribution applies the rate to gross, so the displayed figure runs a few percent above the charged one. That is the conservative direction and is the same base issue flagged for every country in this rollout. CAVEAT: ato.gov.au and legislation.gov.au returned 403 to every fetch attempt this session, so the 2% rate and the 2026-27 surcharge tiers were read from indexed snippets of the live ATO pages and cross-checked against two independent professional tax references, hence medium confidence. The 2% rate itself is long-standing and uncontested. 2026-27 Medicare levy LOW-INCOME thresholds were not confirmed (historically legislated retroactively); irrelevant at ICP income." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "AIHW health expenditure 2023-24 individual out-of-pocket, AUD 1,634 per person and ~1,338 excluding dental + PBS co-payment and the family-pooled PBS Safety Net", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 RAISED from 800 on primary-source evidence, the largest upward level move outside Austria. Medicare and the PBS are national with no state benefit variation, and the previous Brisbane-versus-rest spread had no documented basis in _meta or the city JSONs, so it is removed. Private hospital cover is a PREMIUM and stays in the premium field. Children add ~30% of an adult: frequently bulk-billed, lower PBS costs, and the family-pooled Safety Net pulls the whole household toward the concessional co-pay. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "AIHW health expenditure 2023-24 individual out-of-pocket, AUD 1,634 per person and ~1,338 excluding dental + PBS co-payment and the family-pooled PBS Safety Net", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "HILDA Survey (Melbourne Institute), average daily commuting time by capital city, 2017 wave. https://theconversation.com/australian-city-workers-average-commute-has-blown-out-to-66-minutes-a-day-how-does-yours-compare-120598", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average daily commuting time (round trip)", "year": "2017", "note": "CONVERSION: 65 min/day round trip / 2 = 32.5 one way." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "ATO, Contributions caps https://www.ato.gov.au/tax-rates-and-codes/key-superannuation-rates-and-thresholds/contributions-caps ; Maximum contributions base https://www.ato.gov.au/businesses-and-organisations/super-for-employers/paying-super-on-payday/what-payments-are-qualifying-earnings/maximum-contributions-base ; Super guarantee https://www.ato.gov.au/tax-rates-and-codes/key-superannuation-rates-and-thresholds/super-guarantee", "last_updated": "2026-08-11", "confidence": "high", "note": "Concessional contributions cap AUD 32,500 for 2026-27, a CONTRIBUTION ceiling shared with compulsory employer Super Guarantee, so the employee has only what SG leaves: 12% of gross up to the AUD 270,830 maximum contribution base (annual from 1 July 2026 under Payday Super). Room is zero at and above 270,830. The 5-year carry-forward is deliberately ignored. SEPARATE, UNFIXED: concessional contributions are taxed 15% inside the fund (30% under Division 293 above AUD 250,000), so treating the allowed portion as fully pre-tax still flatters Australia." } }, "houston": { "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 59800 USD, Houston-Pasadena-The Woodlands metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "tax.federal_brackets": { "source": "IRS Rev. Proc. 2024-40 inflation-adjusted brackets (2025)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.state_brackets": { "source": "Texas has NO state income tax", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "TX state 6.25% + Houston city 1% + MTA 1% = 8.25%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Harris County effective ~2.0% (TX uses property tax to offset no-income-tax)", "last_updated": "2026-05-17", "confidence": "high" }, "housing._summary": { "source": "RentCafe avg-by-bedroom 2026-07-02 + Zumper median cross-check", "last_updated": "2026-07-26", "confidence": "high", "note": "2026-07-26 refresh: prior Q1-2026 base ran materially high; Houston rents fell ~10% YoY. RentCafe avg-by-bedroom (2026-07-02) with Zumper median cross-check; average-tier basis, consistent with the dataset." }, "housing.rent_percentiles": { "source": "Census ACS 2024 1-Year, table B25063 gross-rent distribution (metro/CBSA), data.census.gov; P25/P75 interpolated from dollar buckets, interpolated P50 matched published B25064 median (research 2026-07-17 §2). Rounded outward to 2dp per conservative-wider rule.", "last_updated": "2026-08-07", "confidence": "high", "note": "All-unit metro basis (pools bedroom sizes), wider than a within-bedroom band by construction. api.census.gov now requires an API key for queries (checked 2026-08-07); table remains on data.census.gov. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare._summary": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "high" }, "childcare._summary": { "source": "Care.com Houston market + CCDF subsidy band", "last_updated": "2026-05-17", "confidence": "medium" }, "food._summary": { "source": "BLS CPI-U Houston-The Woodlands-Sugar Land MSA", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "METRO monthly $46", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "BLS CES Houston metro", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "IAH-SEA/LAX nonstop UA/AA 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Federal CTC $2,000/child (no TX top-up)", "last_updated": "2026-05-17", "confidence": "high", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "non_cash._summary": { "source": "FMLA federal floor only; TX preempts local PSL", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Houston: USD 131,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "USD 131,000 / 1 = USD 131,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.5240, R5-rounded down to 0.52. REPLACES 0.78 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. n=453. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Index 0.78" }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-21", "note": "Zeroed: the DCFSA is a per-HOUSEHOLD cap, not per-kid (the old $2,000/kid double-counted a 2-kid family). Now carried in subsidy_household_annual. Not means-tested." }, "childcare.income_tested_childcare": { "source": "https://www.wrksolutions.com/childcare-financial-assistance", "last_updated": "2026-08-05", "confidence": "high", "note": "Entry ceiling is the statewide Texas 85% SMI line (TWC BCY 2026 income limits, $92,041 for a family of 4, effective 1 Oct 2025; the Gulf Coast board prints the same table as $7,670/mo); Texas publishes no separate exit ceiling; a household under the line may qualify but is not guaranteed a place." }, "labor_market.tech_salary_distribution": { "source": "https://www.bls.gov/oes/current/oes_26420.htm", "last_updated": "2025-05", "confidence": "rich", "note": "BLS OEWS May 2025, SOC 15-1252 Software Developers, annual wage percentiles (full-market occupation distribution; straight-time gross pay incl. incentive pay, excl. equity and nonproduction bonuses). Public domain; API pull 2026-07-19." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Harris County TX (FIPS 48201), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Texas caveat: from 2015 onward the state reports the prior year figures, so the 2022 row is 2021 fieldwork and this value is marginally understated after escalation. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $860/mo and the ages 1-2 rate $910/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,050. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: Texas has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. Texas law only enables voluntary paid-leave insurance products; no entitlement. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Houston-Pasadena-The Woodlands TX Metro Area (31000US26420). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US26420", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 31.1 (published MOE +/-0.3). Houston is among the longest US metro commutes; the previous 27 sat below the national mean." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Harris County TX (FIPS 48201), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Texas caveat: from 2015 onward the state reports the prior year figures, so the 2022 row is 2021 fieldwork and this value is marginally understated after escalation. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,300." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "phoenix": { "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 62100 USD, Phoenix-Mesa-Chandler metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "tax.federal_brackets": { "source": "IRS Rev. Proc. 2024-40 (2025)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.state_brackets": { "source": "AZ flat 2.5% PIT (post-2023 reform)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "AZ 5.6% + Maricopa County 0.7% + Phoenix 2.3% = 8.6%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Maricopa County effective ~0.7%", "last_updated": "2026-05-17", "confidence": "high" }, "housing._summary": { "source": "RentCafe avg-by-bedroom 2026-07-02 + Zumper median cross-check", "last_updated": "2026-07-26", "confidence": "high", "note": "2026-07-26 refresh: prior Q1-2026 base ran high. RentCafe avg-by-bedroom (2026-07-02) with Zumper cross-check; average-tier basis." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare._summary": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "high" }, "childcare._summary": { "source": "Care.com Phoenix + AZ CCDF", "last_updated": "2026-05-17", "confidence": "medium" }, "food._summary": { "source": "BLS CPI-U Phoenix-Mesa-Chandler", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "Valley Metro monthly $64", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "BLS CES Phoenix metro", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "PHX-SEA/LAX nonstop multiple 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Federal CTC $2,000/child", "last_updated": "2026-05-17", "confidence": "high", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "non_cash._summary": { "source": "AZ Fair Wages Act earned sick + FMLA", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Phoenix: USD 135,625, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "USD 135,625 / 1 = USD 135,625; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.5425, R5-rounded down to 0.54. REPLACES 0.75 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. n=438. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Major CA-alternative tech destination. Index 0.75" }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-21", "note": "Zeroed: the DCFSA is a per-HOUSEHOLD cap, not per-kid (the old $2,000/kid double-counted a 2-kid family). Now carried in subsidy_household_annual. Not means-tested." }, "childcare.income_tested_childcare": { "source": "https://des.az.gov/services/child-and-family/child-care", "last_updated": "2026-08-05", "confidence": "high", "note": "Entry at 165% FPL ($53,064 for a family of 4, DES form CCA-0229A FFY 2026, effective 1 Oct 2025); enrolled households are redetermined at 85% SMI, an exit ceiling of $91,788; a household under the entry line may qualify for assistance." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Maricopa County AZ (FIPS 4013), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $1,090/mo and the ages 1-2 rate $1,180/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,000. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: Arizona has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. AZ Prop 206 is sick time, not parental leave. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Phoenix-Mesa-Chandler AZ Metro Area (31000US38060). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US38060", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 27.6. Mode-blind: this is not the transit-rider commute, which runs far longer in Phoenix." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Maricopa County AZ (FIPS 4013), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,250." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "san-diego": { "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 72800 USD, San Diego-Chula Vista-Carlsbad metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "tax.federal_brackets": { "source": "IRS 2025", "last_updated": "2026-05-17", "confidence": "high" }, "tax.state_brackets": { "source": "CA FTB, 2025 California Tax Rate Schedules (Form 540 booklet) https://www.ftb.ca.gov/forms/2025/2025-540-tax-rate-schedules.pdf", "last_updated": "2026-08-01", "confidence": "high", "note": "Schedule X, single or married/RDP filing separately, tax year 2025: 1% to 11,079 then 2/4/6/8% to 72,724, 9.3% to 371,479, 10.3% to 445,771, 11.3% to 742,953, 12.3% above. Refreshed 2026-08-01: the file carried the TY2024 thresholds (10,756 / 25,499 / ... / 721,314) while already citing the TY2025 schedule, so the data disagreed with its own citation. TY2025 is FTB's newest published year; the TY2026 schedules are not out. Corrected 2026-08-06: this note used to say the 1% Behavioral Health Services Tax above $1M taxable (renamed from the Mental Health Services Tax for TY2025 by Prop 1) was absent from the engine. It is not absent. app.js computes it as taxes.state_surtax from tax.state_surtax_rate 0.01 and tax.state_surtax_threshold 1000000, both of which this city carries, and a single filer on $1,500,000 gross pays $4,839 of it (identical in all four California cities). The threshold is per return and does not double for joint filers, which is the statute rather than a gap: RTC 17043(c) disapplies section 17045, the income-splitting rule that doubles California's regular MFJ bands, so no schema field for tax.state_surtax_threshold_married exists in any city JSON. https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=RTC§ionNum=17043" }, "tax.state_disability_insurance_rate": { "source": "California EDD, SDI Contribution Rates and Benefit Amounts https://edd.ca.gov/en/disability/Contribution_Rates_and_Benefit_Amounts/", "last_updated": "2026-07-30", "confidence": "high", "note": "Employee State Disability Insurance / Paid Family Leave contribution, 1.3% for 2026 (up from 1.2% in 2025). EDD lists the taxable wage ceiling and the maximum contribution as N/A: SB 951 eliminated both effective 2026-01-01, so this is levied on every dollar of wages with no cap. Added 2026-07-30; before that the engine omitted it entirely, understating California payroll by 1.3% of gross. Statewide levy, identical across the CA metros." }, "tax.sales_tax_rate": { "source": "CA 7.25% + SD county 0.5% = 7.75%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "SD effective 1.1% (Prop 13 capped)", "last_updated": "2026-05-17", "confidence": "high" }, "housing._summary": { "source": "Zillow + Apartments.com SD Q1 2026", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare._summary": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "high" }, "childcare._summary": { "source": "Care.com SD + CA CCDF", "last_updated": "2026-05-17", "confidence": "medium" }, "food._summary": { "source": "BLS CPI-U San Diego-Chula Vista-Carlsbad", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "MTS monthly pass $72", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "BLS CES San Diego metro", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "SAN-SEA/LAX nonstop AS/UA 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Federal CTC + CA Young Child Tax Credit", "last_updated": "2026-05-17", "confidence": "high", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "non_cash._summary": { "source": "CA SB 616 + SD Earned Sick Leave Ordinance + CA PFL 70%", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, San Diego: USD 185,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "USD 185,000 / 1 = USD 185,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.7400, R5-rounded down to 0.74. REPLACES 0.92 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Index 0.92 (slightly below SF, well above national)" }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-21", "note": "Zeroed: the DCFSA is a per-HOUSEHOLD cap, not per-kid (the old $2,000/kid double-counted a 2-kid family). Now carried in subsidy_household_annual. Not means-tested." }, "labor_market.tech_salary_distribution": { "source": "https://www.bls.gov/oes/current/oes_41740.htm", "last_updated": "2025-05", "confidence": "rich", "note": "BLS OEWS May 2025, SOC 15-1252 Software Developers, annual wage percentiles (full-market occupation distribution; straight-time gross pay incl. incentive pay, excl. equity and nonproduction bonuses). Public domain; API pull 2026-07-19." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "tax.state_brackets_married": { "source": "CA FTB, 2025 California Tax Rate Schedules (Form 540 booklet) https://www.ftb.ca.gov/forms/2025/2025-540-tax-rate-schedules.pdf", "last_updated": "2026-08-01", "confidence": "high", "note": "Schedule Y, married/RDP filing jointly or qualifying surviving spouse, tax year 2025. Verified band by band as EXACTLY 2x Schedule X on all 8 finite thresholds, which is what RTC 17041(b) prescribes; the base-tax columns double too. Added 2026-08-01. Before this, couples were run through the single-filer ladder, overstating California tax by a flat $3,457.64 for any household above about $141k of taxable income, and by $4,029 at a $450k household. Head of household is NOT 2x; no schema field for tax.state_brackets_hoh exists." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: San Diego County CA (FIPS 6073), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $1,260/mo and the ages 1-2 rate $1,690/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,500. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.income_tested_childcare": { "source": "https://www.cdss.ca.gov/Portals/9/Additional-Resources/Letters-and-Notices/CCBs/2026/CCB_26-17.pdf", "last_updated": "2026-08-05", "confidence": "high", "note": "Eligibility-only (CC2 shape): CDSS ceiling at 85% SMI, $115,637 for a family of four (CCB 26-17, FY 2026-27; same threshold initial and ongoing). Tested on ADJUSTED monthly income per 5 CCR 18078(a), so a gross comparison is approximate; CCDF voucher minus copay, never a percentage. May qualify, never a promise of a spot. Revise each July." }, "non_cash.parental_leave_weeks": { "source": "EDD Paid Family Leave https://edd.ca.gov/en/disability/paid-family-leave/ ; https://edd.ca.gov/en/disability/Calculating_PFL_Benefit_Payment_Amounts/", "last_updated": "2026-08-06", "confidence": "high", "note": "16 = 8 weeks of CA PFL bonding per parent within a 12-month period, each parent's own SDI-funded non-transferable entitlement, summed (both-parents flag set). The birthing parent's SDI pregnancy disability (usually 4 weeks before plus 6 after, 8 for cesarean) is medically certified, EXCLUDED, conservative. Was 12/70: the unpaid FMLA window with a fabricated headline. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "EDD: SB 951 rates for claims from 2025 (90% low-earner band, 70% above USD 20,931.31 highest-quarter earnings); https://edd.ca.gov/en/disability/Calculating_PFL_Benefit_Payment_Amounts/", "last_updated": "2026-08-06", "confidence": "high", "note": "0 percentage-paid weeks: for claims from 2025 under SB 951, CA pays 90% for highest-quarter earnings of 722.50-16,279.90, a flat 1,127/wk transition band to 20,931.30, and 70% above 20,931.31 per quarter, capped at USD 1,765/wk in 2026 (1,681 in 2025), binding from about USD 131,000. CAVEAT recorded: below that threshold CA pays a true 70% for the ICP band, so the flat store overstates for earners under about 131k (the engine clamps at 100%); at the salaries we price it is exact and less favorable than any headline. At USD 150,000 the effective rate is about 61%." }, "non_cash.parental_leave_flat": { "source": "EDD Contribution Rates and Benefit Amounts: 2026 maximum weekly benefit USD 1,765 https://edd.ca.gov/en/disability/Contribution_Rates_and_Benefit_Amounts/", "last_updated": "2026-08-06", "confidence": "high", "note": "USD 1,765/wk, 2026 maximum. 16 x 1,765 = USD 28,240 total." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, San Diego-Chula Vista-Carlsbad CA Metro Area (31000US41740). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US41740", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 26.1. Survey self-report, MSA geography." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: San Diego County CA (FIPS 6073), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,900." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "minneapolis": { "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 72700 USD, Minneapolis-St. Paul-Bloomington metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "tax.federal_brackets": { "source": "IRS 2025", "last_updated": "2026-05-17", "confidence": "high" }, "tax.state_brackets": { "source": "MN Dept of Revenue, income tax brackets for tax year 2026 https://www.revenue.state.mn.us/press-release/2025-12-16/minnesota-income-tax-brackets-standard-deduction-and-dependent-exemption", "last_updated": "2026-07-31", "confidence": "high", "note": "Rates unchanged at 5.35/6.80/7.85/9.85. Thresholds refreshed 2026-07-31 to the published 2026 single-filer figures ($33,310 / $109,430 / $203,150); the file carried ~2023 values" }, "tax.state_brackets_married": { "source": "MN Dept of Revenue, income tax brackets for tax year 2026 https://www.revenue.state.mn.us/press-release/2025-12-16/minnesota-income-tax-brackets-standard-deduction-and-dependent-exemption", "last_updated": "2026-07-31", "confidence": "high", "note": "2026 married-filing-jointly thresholds $48,700 / $193,480 / $337,930. Previously this array duplicated the single-filer brackets, which overstated Minnesota tax for every couple" }, "tax.state_payroll_premium_rate": { "source": "Minnesota Paid Leave premium rate and contributions https://pl.mn.gov/resources/calculators/premium-rate-and-contributions", "last_updated": "2026-07-31", "confidence": "high", "note": "Minnesota Paid Leave started 2026-01-01 at a 0.88% total premium; the employee share is 0.44%" }, "tax.state_payroll_premium_cap": { "source": "Minnesota Paid Leave / SSA 2026 contribution and benefit base", "last_updated": "2026-07-31", "confidence": "high", "note": "Premiums apply to wages up to the Social Security withholding limit ($184,500 in 2026)" }, "tax.sales_tax_rate": { "source": "MN 6.875% + Hennepin 0.15% + Minneapolis 1.5% = 8.525%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Hennepin effective ~1.3%", "last_updated": "2026-05-17", "confidence": "high" }, "housing._summary": { "source": "Zillow + Apartments.com MSP Q1 2026", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare._summary": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "high" }, "childcare._summary": { "source": "Care.com MSP + MN CCAP", "last_updated": "2026-05-17", "confidence": "medium" }, "food._summary": { "source": "BLS CPI-U Minneapolis-St. Paul-Bloomington", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "Metro Transit monthly $85", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "BLS CES MSP metro", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "MSP-SEA/LAX nonstop DL/AS 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Federal CTC + MN Child Tax Credit (2024+)", "last_updated": "2026-05-17", "confidence": "high", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "non_cash._summary": { "source": "MN ESST (2024) + Minneapolis SST ordinance + MN PFML (2026)", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Minneapolis: USD 126,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "USD 126,000 / 1 = USD 126,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.5040, R5-rounded down to 0.5. REPLACES 0.85 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. n=386. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Index 0.85" }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-21", "note": "Zeroed: the DCFSA is a per-HOUSEHOLD cap, not per-kid (the old $2,000/kid double-counted a 2-kid family). Now carried in subsidy_household_annual. Not means-tested." }, "childcare.income_tested_childcare": { "source": "https://dcyf.mn.gov/programs-directory/child-care-assistance-program", "last_updated": "2026-08-05", "confidence": "high", "note": "Entrance at 47% SMI ($67,679 for a family of 4, DHS-6413N effective 13 Oct 2025) with exits at 67% SMI at redetermination ($96,479) and 85% SMI within the eligibility period ($122,398); a household under the entrance line may qualify, subject to county Basic Sliding Fee waitlists." }, "labor_market.tech_salary_distribution": { "source": "https://www.bls.gov/oes/current/oes_33460.htm", "last_updated": "2025-05", "confidence": "rich", "note": "BLS OEWS May 2025, SOC 15-1252 Software Developers, annual wage percentiles (full-market occupation distribution; straight-time gross pay incl. incentive pay, excl. equity and nonproduction bonuses). Public domain; API pull 2026-07-19." }, "labor_market.english_band": { "source": "English is the / an official working language", "last_updated": "2024", "note": "National-level English usability for tech work. Band: native." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Hennepin County MN (FIPS 27053), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $1,610/mo and the ages 1-2 rate $1,840/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,400. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "non_cash.parental_leave_weeks": { "source": "MN Paid Leave https://pl.mn.gov/individuals/bonding ; https://pl.mn.gov/individuals/how-paid-leave-works", "last_updated": "2026-08-06", "confidence": "high", "note": "24 = 12 weeks of MN Paid Leave bonding per parent (the agency states each parent can take up to 12 weeks and one parent's leave does not reduce the other's), summed (both-parents flag set). Combined family+medical cap of 20 wk per person per benefit year; the birthing parent's medical portion is certified, EXCLUDED. Benefits payable since 2026-01-01. Was 12/90, which predated the launch. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "pl.mn.gov: 90% of wages up to 50% of the SAWW, 66% to the SAWW, 55% above; maximum = the SAWW", "last_updated": "2026-08-06", "confidence": "high", "note": "0 percentage-paid weeks: MN pays 90% of wages up to 50% of the SAWW (USD 711.50), then 66% of the slice to the SAWW, then 55% above, capped at the SAWW of USD 1,423/wk, binding from about USD 104,000. Three-tier formula, no honest single headline. At USD 150,000 the effective rate is about 49%." }, "non_cash.parental_leave_flat": { "source": "pl.mn.gov estimate-your-payments: maximum weekly benefit = the state AWW, USD 1,423 (launch year 2026) https://pl.mn.gov/resources/calculators/estimate-your-payments", "last_updated": "2026-08-06", "confidence": "high", "note": "USD 1,423/wk, launch-year maximum. 24 x 1,423 = USD 34,152 total." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Minneapolis-St. Paul-Bloomington MN-WI Metro Area (31000US33460). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US33460", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 24.1. Stored value was already correct; citation added." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Hennepin County MN (FIPS 27053), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,750." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "nashville": { "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 62700 USD, Nashville-Davidson-Murfreesboro-Franklin metro area; count-weighted from table B20017 male/female FT-YR medians)", "last_updated": "2024", "confidence": "usable", "note": "All-earner metro override, MSA geographic scope. Reconciled 2026-07-28 from city-proper to the Metropolitan Statistical Area so US metros share one scope; combined FT-YR median derived from ACS 2024 B20017 (leans ~2% high vs the exact combined median, the conservative direction)." }, "tax.federal_brackets": { "source": "IRS 2025", "last_updated": "2026-05-17", "confidence": "high" }, "tax.state_brackets": { "source": "TN has NO state income tax (Hall tax repealed 2021)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "TN 7% + Davidson 2.25% = 9.25%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Davidson effective ~0.75%", "last_updated": "2026-05-17", "confidence": "high" }, "housing._summary": { "source": "Zillow + RentCafe Nashville Q1 2026", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "healthcare._summary": { "source": "KFF 2025 EHBS (worker premium share; single $1,440 / family $6,850; large-firm family $6,227) + Peterson-KFF Health System Tracker 2023 OOP ($869/person, $3,564 family of four) trended ~6%/yr to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-07-17", "confidence": "high" }, "childcare._summary": { "source": "Care.com Nashville + TN childcare subsidy", "last_updated": "2026-05-17", "confidence": "medium" }, "food._summary": { "source": "BLS CPI-U Nashville-Davidson", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "WeGo monthly $65", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "BLS CES Nashville metro", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "BNA-SEA/LAX nonstop AA/DL 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Federal CTC (no TN top-up)", "last_updated": "2026-05-17", "confidence": "high", "note": "US Child Tax Credit (~$2,000/kid plus modeled state top-ups): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at the ~$226k-280k published household." }, "non_cash._summary": { "source": "FMLA federal floor; TN preempts local PSL", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Nashville: USD 155,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "USD 155,000 / 1 = USD 155,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.6200, R5-rounded down to 0.62. REPLACES 0.82 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. n=153, thin. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Fastest-growing US tech metro. Index 0.82" }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-21", "note": "Zeroed: the DCFSA is a per-HOUSEHOLD cap, not per-kid (the old $2,000/kid double-counted a 2-kid family). Now carried in subsidy_household_annual. Not means-tested." }, "childcare.income_tested_childcare": { "source": "https://www.tn.gov/humanservices/for-families/child-care-services/child-care-payment-assistance.html", "last_updated": "2026-08-05", "confidence": "high", "note": "Ceiling at 85% SMI ($76,668 for a family of 4, TDHS Smart Steps flyer of Nov 2024, the newest agency print); a household under it may qualify, subject to availability of funding per TDHS." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Davidson County TN (FIPS 47037), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $760/mo and the ages 1-2 rate $870/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,150. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: Tennessee has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. Tennessee grants paid parental leave to state employees only and otherwise enables voluntary insurance products. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Nashville-Davidson--Murfreesboro--Franklin TN Metro Area (31000US34980). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US34980", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 28.7. Metro sprawl has pushed this well past the previous 25." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Davidson County TN (FIPS 47037), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,400." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "tel-aviv": { "childcare.free_tier_start_age": { "source": "State of Israel, Compulsory Education Law (1949) as extended: the State provides free education to children ages 3-4 in public pre-kindergartens (gan chova), gov.il Ministry of Education", "confidence": "high", "note": "Free/compulsory pre-primary starts at 3. Ages 0-3 (maon) are paid at market rate; Tel Aviv is the most expensive nursery market in Israel (infant about NIS 3,804/month, toddler about NIS 2,945/month, havayot.org.il school year 5786). Before this, ages 1-5 billed the free-tier rate. The 3-5 residual is left at 0 but is not truly zero for a dual-earner household: tzaharon (afternoon extension beyond the early gan finish) is regulated at up to NIS 672/month basic and NIS 935/month with add-ons (kol-zchut.org.il), and Tel Aviv municipal programmes run about NIS 730/month. Left out for now because the city already carries a NIS 5,000/kid subsidy that would partly absorb it. TODO (dedicated pass): decide whether tzaharon joins the 3-5 residual.", "last_updated": "2026-08-01" }, "tax.federal_brackets": { "source": "Israel Tax Authority 7-bracket PIT 2026", "last_updated": "2026-05-17", "confidence": "high" }, "tax.social_security_rate": { "source": "btl.gov.il, health-insurance rates page (mas briut employee 3.23% reduced from 01.01.2025, 5.17% full from 01.02.2025) https://www.btl.gov.il/Insurance/Health_Insurance/Pages/%D7%A9%D7%99%D7%A2%D7%95%D7%A8%D7%99%20%D7%93%D7%9E%D7%99%20%D7%91%D7%99%D7%98%D7%95%D7%97%20%D7%91%D7%A8%D7%99%D7%90%D7%95%D7%AA.aspx ; kol-zchut salaried-employee page for the national-insurance halves (1.04% reduced, 7.00% full), giving published employee totals of 4.27% and 12.17%", "last_updated": "2026-08-06", "confidence": "high", "note": "0.12 -> 0.1217 on 2026-08-06. The combined Bituach Leumi + Mas Briut employee charge runs in two bands: 4.27% on the first ILS 7,703 a month, then 12.17% up to the ILS 51,910/month ceiling. The { rate, cap } shape holds one band, so we carry the FULL 12.17% across the whole base. That overstates by (12.17 - 4.27)% x 7,703 x 12 = about ILS 7,302 a year, which is the direction that errs against Israel. The old 12.00% came from the btl.gov.il form-102 rates page, which still renders the pre-2025 3.50/12.00 pair; ignore that page. Israel's mandatory occupational pension (pensia chova, employee 6%) is charged separately under tax.pension_employee_rate since 2026-08-06 and stays outside this value. Working: docs/superpowers/research/2026-08-06-tax-value-fix-batch.md" }, "tax.social_security_cap": { "source": "btl.gov.il, contribution rates page: maximum income for national-insurance and health contributions ILS 51,910/month for 2026 (x12 = ILS 622,920/yr); reduced-rate step ILS 7,703/month. https://www.btl.gov.il/Insurance/National%20Insurance/type_list/Self_Employed/Pages/rates.aspx", "last_updated": "2026-08-06", "confidence": "high", "note": "570000 -> 622920 on 2026-08-06. The old figure was the ILS 47,465/month ceiling of several years ago. The 2026 ceiling tracks 5x the basic amount rather than 5x the average wage, so it is read off the published table and not derived. This now agrees with healthcare.statutory_employee.annual_cap, which was already on 622,920 from its own 2026-07-31 primary read. The cap binds per earner: a two-professional household clears it twice." }, "tax.pension_employee_rate": { "source": "Extension Order for Comprehensive Pension Insurance in the Economy (tzav harchava, in force since 2008; employee 6% since 1.1.2017). kol-zchut, mandatory pension insurance for employees: employer 6.5% + employee 6% + severance 6% https://www.kolzchut.org.il/he/%D7%97%D7%95%D7%91%D7%AA_%D7%91%D7%99%D7%98%D7%95%D7%97_%D7%A4%D7%A0%D7%A1%D7%99%D7%95%D7%A0%D7%99_%D7%9C%D7%A2%D7%95%D7%91%D7%93%D7%99%D7%9D", "last_updated": "2026-08-06", "confidence": "high", "note": "Employee share of pensia chova, charged on the ORDER's insured wage: the gross base wage or the national average wage, whichever is LOWER (kol-zchut: contributions are computed on the base-wage gross or the average wage in the economy, the lower of the two). Carved to the forced-savings own-money row: the pension-fund balance is the worker's own property, inheritable and portable, and the order covers foreign workers alike. Employment contracts commonly insure the FULL wage (near-universal in tech); that contractual excess is excluded under the statutory-minimum convention, so this value understates a typical Tel Aviv payslip's pension line. The 35% pension tax credit (zikui) stays outside the engine, which overstates Israeli tax slightly, the direction that errs against Israel. Decision log: docs/superpowers/research/2026-08-06-pension-omission-pair.md" }, "tax.pension_wage_cap_annual": { "source": "National-insurance average wage (hasachar hamemutza lefi chok haBituach haLeumi) as of 01.01.2026: ILS 13,769/month, x12 = 165,228/yr. kol-zchut average-wage concept page; btl.gov.il 2026 updates (hadasaidkonkitzva2026). The extension order caps the MANDATORY insured wage at this figure.", "last_updated": "2026-08-06", "confidence": "high", "note": "165,228 = 13,769 x 12. Binds for every ICP earner (all far above the average wage), so the statutory-minimum pension charge is a flat ILS 9,913.68/yr per earner, applied per earner under individual filing. Re-check each January: the average wage resets annually (2025 value was 13,316/month, still quoted by some 2026-dated commercial guides). Distinct from tax.social_security_cap 622,920, which caps the separate Bituach Leumi + Mas Briut charge." }, "tax.sales_tax_rate": { "source": "Ma'am (VAT) 18% (raised from 17% Jan 2025)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Tel Aviv Arnona effective ~0.4% (area-based, ₪/m² scaled)", "last_updated": "2026-05-17", "confidence": "medium" }, "housing._summary": { "source": "Madlan + Yad2 Tel Aviv Q1 2026", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Kupat Holim universal via 3.1-5% health payroll", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "Israeli gan free from age 3; private daycare ILS 5-6k", "last_updated": "2026-05-17", "confidence": "high" }, "food._summary": { "source": "Numbeo Tel Aviv 2026 + Shufersal/Rami Levy", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "Rav-Kav monthly Tel Aviv area ₪225", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "CBS leisure + Numbeo Tel Aviv", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "TLV-SEA/LAX nonstop El Al 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Bituach Leumi Yaldei tikun ₪200/mo", "last_updated": "2026-05-17", "confidence": "high" }, "non_cash._summary": { "source": "Hours of Work and Rest Law + BL parental ~15wk", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Tel Aviv: ILS 464,484, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-europe.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "ILS 464,484 / 3.05 = USD 152,290; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.6092, R5-rounded down to 0.6. REPLACES 0.5 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. The deepest and most coherent sample in its batch. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Major US-tech magnet. Index 0.50" }, "tax.filing_basis": { "source": "PwC Worldwide Tax Summaries: Israel, Individual: Tax administration (separate assessment / chishuv nifrad for spouses with independent-source earned income). https://taxsummaries.pwc.com/israel/individual/tax-administration", "last_updated": "2026-07-26", "note": "IL-2 (closeout 2026-07-26): each spouse taxed on their own salary." }, "tax.tax_credit_annual": { "source": "PwC Worldwide Tax Summaries: Israel, Individual: Other tax credits (credit-point value ILS 242/mo = ILS 2,904/yr for 2026; a resident earner gets 2.25 points = ILS 6,534). https://taxsummaries.pwc.com/israel/individual/other-tax-credits-and-incentives", "last_updated": "2026-07-26", "note": "IL-1 (closeout 2026-07-26): 2.25 tax credit points, flat non-refundable annual credit against tax." }, "healthcare.statutory_employee": { "source": "btl.gov.il health insurance rates page and salaried-employee rate table, 2026: health 3.23% reduced (from 01.01.2025) and 5.17% full (from 01.02.2025), employer column blank; national insurance employee 1.04% reduced / 7.00% full, giving the published employee totals 4.27% and 12.17%. Reduced-rate threshold ILS 7,703/month (CPI-linked under the 2026 Arrangements Law, not 60% of the ILS 13,769 section-2 average wage); ceiling ILS 51,910/month = ILS 622,920/yr, which tracks 5x the basic amount rather than 5x the average wage.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll. Rate is the effective BLEND at the ceiling (7,703 x 3.23% + 44,207 x 5.17% = ILS 2,534.31/month = ILS 30,411.71/yr, over a 622,920 base = 4.8822%), because the { rate, annual_cap } shape cannot express two bands and a flat 5.17% would inflate the ceiling figure by ILS 1,793. The blend overstates slightly for mid-band earners, which is the conservative direction. IGNORE the btl.gov.il form-102 rates page: it still renders the pre-2025 3.50/12.00 pair. The tax-block mismatch this note used to flag is CLOSED as of 2026-08-06: tax.social_security_rate is now 0.1217 and tax.social_security_cap is now 622,920, so the health slice below and the payroll line it sits inside are finally on the same ceiling." }, "childcare.subsidy_per_kid_annual": { "source": "maon dirug tariff tables", "last_updated": "2026-08-05", "confidence": "high", "note": "ZEROED 2026-08-05 (CC5, was 5000): Top income band pays the full tariff; the scheme is reduced tuition at lower incomes, never household cash. Working: docs/superpowers/research/2026-08-05-cc5-four-cities-subsidy.md" }, "non_cash.parental_leave_weeks": { "source": "btl.gov.il maternity/paternity pages (fetched 2026-08-07): 15 weeks dmei leida with 10-of-14 or 15-of-22 months of contributions; 8 weeks with 6-of-14; father days even taken together with the spouse are deducted from those to which the mother is entitled", "last_updated": "2026-08-07", "confidence": "high", "note": "15 wk CONFIRMED on the household-inclusive basis by the batch-5 audit (2026-08-07): the chufshat leida extension to 26 weeks is UNPAID beyond week 15 (excluded); the father's dmei leida is a transfer from the SAME 15-week pot (btl exact wording: deducted from the mother's entitlement) so it adds nothing; the 2016 5-day paternity provision consumes the father's own annual-leave/sick banks and is not parental-leave pay (excluded). Mover pricing: the 15-week tier applies (10-of-14 months of insurance before the determining date; an employed-on-arrival mover who then conceives passes with work-to-term; the 8-week 6-of-14 tier is the floor alternative, recorded not stored). Research: docs/superpowers/research/2026-08-07-pl-b5-austria-israel.md." }, "non_cash.parental_leave_paid_pct": { "source": "National Insurance Law: dmei leida = 100% of the calendar-day average wage (higher of 3-month/90 or 6-month/180), paid by Bituach Leumi", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 15 weeks into non_cash.parental_leave_capped (100% up to the daily ceiling), so no percentage-paid weeks remain on this field. No additive paternity exists: the father's dmei leida is a transfer from the SAME 15-week pot (counted once), and the 5-day birth absence draws the father's own annual/sick leave balances, excluded per the conservative rule." }, "non_cash.parental_leave_capped": { "source": "btl.gov.il ShiureyLeda page (fetched 2026-08-06, re-verified 2026-08-07 in the batch-5 basis audit): maximum dmei leida ILS 1,752.33 per calendar day as of 1 January 2026; corroborated by forty.co.il. Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (Israel), 2026-08-06-pl-caps-israel-turkey.md, 2026-08-07-pl-b5-austria-israel.md", "last_updated": "2026-08-06", "confidence": "high", "note": "Payout ceiling ILS 1,752.33/calendar day (~ILS 52,570/month, ~ILS 630,840/yr equivalent) on all 15 weeks; the Switzerland shape: a true 100% below a threshold near the top of the band, binding only above ~ILS 631k/yr base salary (at ILS 900k the effective rate is ~70%). NOT the same figure as the stored tax.social_security_cap (ILS 51,910/month): the two ceilings run on different statutory formulas since the 2026 amendment; do not conflate. The benefit is itself taxable, a net-vs-gross wrinkle recorded, not a cap." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Health basket copayments with the quarterly visit ceiling and the medicine copayment ceiling; Tipat Halav provides free child preventive care", "last_updated": "2026-08-08", "confidence": "low", "note": "LEVEL NOT RE-VERIFIED in the 2026-08-08 sweep; row 1 is held and only the household curve was added. Free child preventive care partially flattens the curve; illness care does not. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Health basket copayments with the quarterly visit ceiling and the medicine copayment ceiling; Tipat Halav provides free child preventive care", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Israel's CBS publishes commute-time BANDS and no mean. A band-derived reconstruction gives roughly 32, but it is low confidence and it SHORTENS the commute, which flatters the destination and therefore carries the higher evidence bar; 35 sits inside the plausible Gush Dan metro band, so the honest gap was kept instead." }, "non_cash.sick_leave": { "source": "Sick Pay Law, 5776-1976 (חוק דמי מחלה, התשל\"ו-1976).; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-H.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "TheMarker Supermarker, tax benefits for kupot gemel and hishtalmut https://www.supermarker.themarker.com/Gemel/TaxBenefitsForKupatGemelAndHishtalmut.aspx ; https://pensuni.com/?p=827 (Pkudat Mas Hachnasa s.45A and s.47)", "last_updated": "2026-08-11", "confidence": "high", "note": "LEAD REFUTED: on insured salary the employee's 7% buys a 35% CREDIT under s.45A, not a deduction. Eligible contribution 7% x ILS 116,400 = ILS 8,148/yr, credit ILS 2,852/yr. The s.47 DEDUCTION applies only to income the employer does NOT insure. Decisive: an employee who voluntarily contributes above the compulsory amount gets no extra tax benefit. Setting 0 REMOVES an unearned break rather than adding a new one, since the engine runs Israel's ladder on full taxable_income with the 35% credit deliberately unmodeled. 2026 figures confirmed: qualifying income ILS 9,700/mo, average wage ILS 13,769/mo, which matches the engine constant exactly." } }, "sao-paulo": { "tax.filing_basis": { "source": "PwC WWTS - Brazil (Tax administration): spouses may elect to file jointly or separately; individual is the two-earner default.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "RFB IRPF 2026 (5 brackets post-2024 inflation adjustment)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.social_security_rate": { "source": "INSS employee 7.5-14% (v1 flat 11% blended)", "last_updated": "2026-05-17", "confidence": "medium" }, "tax.social_security_cap": { "source": "INSS teto previdenciário BRL 7,786/mo × 12 = ~93k", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Combined ICMS+ISS+PIS/COFINS ~25% blended on services", "last_updated": "2026-05-17", "confidence": "medium" }, "tax.property_tax_rate_pct": { "source": "SP IPTU effective ~1.0%", "last_updated": "2026-05-17", "confidence": "high" }, "housing._summary": { "source": "QuintoAndar + Loft São Paulo Q1 2026 + Numbeo", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Zap Imoveis glue-API listing-count CDF, 2026-08-17 (apartments for rent, Sao Paulo city proper - city verified per listing; priceMax proven to filter BARE aluguel with condominio/IPTU in separate fields; asking BRL/month per unit): 2 quartos P25/med/P75 = 2,781/4,936/7,975 (N=36,181) -> 0.563/1.616, applied 0.56/1.62; 1 quarto 0.749/1.432 (N=49,502, studios folded in). Independent corroboration: QuintoAndar duplicate-free pool 2Q 0.714/~1.67 (n=16,758; its p75 sits in a crude 3,000-5,000 price-slug bin).", "last_updated": "2026-08-17", "confidence": "medium", "note": "REBASED 2026-08-17 (ledger cluster housing-0817) from the calibrated default 0.82/1.24, which fails on BOTH edges in the flattering direction on the band's own basis - all eight measured edges across two portals and two bedroom slices sit outside tolerance. Widest measured pair in the repo: the upper multiplier 1.62 EXCEEDS the global sanity bound (p75 < 1.6, set 2026-08-07 when every pair was <= 1.44), so the guard test carries a documented per-city exception (sao-paulo < 1.7) rather than a silent global loosening - see the ledger row. Lower-edge between-pool gap recorded, not averaged: Zap 0.563 (full broker inventory incl. cheap periphery) vs QuintoAndar 0.714 (owner-direct mid-market); Zap is primary for n, granularity, proven filter semantics and the conservative (wider) read. Portal pools truncate the informal cheap tail, so the true citywide spread is wider still. Dual/heterogeneous-market class, same treatment as Mexico City, Bangkok, Mumbai. 2-quartos anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs + semantics proofs: docs/superpowers/research/2026-08-17-housing-band-fixes-sao-paulo.md; decision: docs/superpowers/research/2026-08-17-housing-band-fixes.md (S4). NOTE the stored 2BR median (6,000 BRL) sits above the Zap 2Q pool median (4,936) - a LEVEL question, flagged, out of scope here." }, "healthcare._summary": { "source": "SUS universal (slow); tech families use planos de saúde BRL 800-2k single", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "Private creche/escolinha SP ~BRL 3-5k", "last_updated": "2026-05-17", "confidence": "medium" }, "food._summary": { "source": "Numbeo São Paulo 2026 + IBGE", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "SPTrans/Metrô bilhete único ~BRL 200/mo", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "IBGE POF + Numbeo SP", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "GRU-SEA/LAX 1-stop AA/UA 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Bolsa Família means-tested; no benefit for tech-income families", "last_updated": "2026-05-17", "confidence": "high" }, "non_cash._summary": { "source": "CLT 30 dias férias + 12 feriados + 17wk licença-maternidade 100%", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, São Paulo: BRL 195,823, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "BRL 195,823 / 5.07 = USD 38,624; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1545, R5-rounded down to 0.15. REPLACES 0.32 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. The 13th salary is already inside. Do NOT multiply by 13. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): LatAm flagship. Index 0.32" }, "healthcare.premium_monthly_by_family_size": { "source": "ANS coletivo empresarial coverage data (derivation + caveats in docs/superpowers/research/2026-08-05-brazil-healthcare.md)", "last_updated": "2026-08-05", "confidence": "high", "note": "Re-based 2026-08-05: BRL 800 was the retail individual price; ~70% of ICP-class professionals get the employer channel, which is not household-zero (avg ~60% employer subsidy, employee-paid dependants, copays). Halved single tier with family scaling; OOP raised 3,000 to 5,000 to carry the copay reality. Closes flag:01785000328601-1787fcc3 + flag:01785698167718-26da2666." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Brazilian private consultation and co-payment pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.parental_leave_capped": { "source": "Lei 8.213/91 art. 72 (salario-maternidade = remuneracao integral, paid by the empresa and compensated against INSS contributions, art. 72 par. 1); CF art. 248 (benefits outside the RGPS teto observe the art. 37 XI ceiling, the STF-minister subsidio); STF ADI 1946 (salario-maternidade exempt from the RGPS teto); subsidio = R$ 46,366.19/month from 1 Feb 2025, Lei 14.520/2023 https://www.planalto.gov.br/ccivil_03/_ato2023-2026/2023/lei/l14520.htm (final step; the Dec-2025 judiciary raise covers servidores, not ministers)", "last_updated": "2026-08-17", "confidence": "high", "note": "The 17.143 maternity weeks pay 100% of remuneration up to the constitutional teto of R$ 46,366.19/month (R$ 556,394.28/yr); the 0.857-wk paternity remainder stays on parental_leave_paid_pct 100 (ADCT art. 10 par. 1, employer-paid at full salary, no ceiling). The ceiling that binds a CLT empregada is CF art. 248's STF-minister subsidio, NOT the RGPS teto (R$ 8,475.55 in 2026), per ADI 1946; the RGPS teto does bind the other claimant categories our ICP is not in. Refutation branch checked and rejected: no statute or binding precedent obliges the employer to top up above the teto, and modelling the cap is the conservative call. At the R$ 782,000/yr base scenario the maternity slice pays 71.2% and the blended 18-wk rate is 72.6%; below R$ 556k/yr it stays an honest 100%. UNIT: the subsidio is a MONTHLY figure, revised only by lei (last step 1 Feb 2025). Research: docs/superpowers/research/2026-08-17-parental-leave-caps.md." }, "non_cash.parental_leave_weeks": { "source": "CF art. 7 XVIII (120 dias); CLT art. 392; Lei 8.213/91 arts. 71-73 (salario-maternidade = remuneracao integral); ADCT art. 10 par. 1 (5 days paternity, in force through 2026); Lei 15.371 phase-in: Senado https://www12.senado.leg.br/noticias/materias/2026/04/01/sancionado-o-aumento-gradual-da-licenca-paternidade", "last_updated": "2026-08-06", "confidence": "high", "note": "18 wk = 120 days maternity at 100% (17.143 wk; capped at the constitutional STF-minister teto, BRL 46,366.19/month, modelled in non_cash.parental_leave_capped as of 2026-08-17; the 2026-08-06 judgment that the teto is irrelevant to any ICP was corrected, it binds at the USD-anchored scenarios this site prices, e.g. the R$ 782,000/yr base scenario) + 5 days paternity at full salary, employer-paid (0.71 wk); 17.86 rounds to 18 (the Germany 66.87-to-67 convention). Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. WATCH ITEM: Lei 15.371 (sanctioned 2026-03-31) phases paternity to 10 days from 2027-01-01 (18.6 wk), 15 from 2028 (19.3), 20 from 2029 (20), with an INSS-reimbursed salario-paternidade at full remuneration; re-touch each January. Excluded: Programa Empresa Cidada 180/20-day extensions (employer opt-in via tax incentive) and collective-agreement top-ups." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Brazilian private consultation and co-payment pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "CHECKED for the premium-as-out-of-pocket error and CLEAR: SUS is free at the point of use and these values are verified far below any real plano de saude premium, so they are co-pays and cash consultations rather than the plan itself. Two city tiers preserved. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "non_cash.commute_minutes_avg": { "source": "IBGE, Censo Demografico 2022, SIDRA table 10331, Regiao Metropolitana de Sao Paulo (39 municipios). https://sidra.ibge.gov.br/tabela/10331", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Tempo habitual de deslocamento de casa para o trabalho", "year": "2022", "note": "IBGE's statistic is home-to-work and therefore ONE WAY by construction. Mean computed from IBGE's published bands by midpoint weighting. The previous 70 was inflated by 21 minutes and is consistent with the widely quoted 'two hours a day' round-trip framing. Six of the seven Brazilian values were wrong by 15 to 21 minutes in the same direction." }, "non_cash.sick_leave": { "source": "Lei 8.213/1991 (Plano de Benefícios da Previdência Social) , Art. 60 §3º (employer; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": "https://www.gov.br/inss/pt-br/direitos-e-deveres/inscricao-e-contribuicao/tabela-de-contribuicao-mensal", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Receita Federal, PGBL e VGBL https://www.gov.br/receitafederal/pt-br/acesso-a-informacao/perguntas-frequentes/imposto-de-renda/dirpf/declaracao/pgvl-vgbl (Lei 9.532/97 art. 11)", "last_updated": "2026-08-11", "confidence": "high", "note": "PGBL contributions are deductible up to 12% of rendimentos tributaveis, i.e. 12% of gross for an employee. No monetary ceiling, so no max key. VGBL is NOT deductible. The deduction requires the COMPLETE declaration and this value assumes it, which is consistent with the branch: Sao Paulo carries no standard_deduction, so the engine already models a complete/itemised basis, and at BRL 300k the 12% (R$36,000) beats the simplificado cap (R$17,640) so choosing complete is genuinely rational. Lei 15.270/2025, in force 1 Jan 2026, leaves art. 11 untouched. Never binds at 6% or 10%: pure correctness insurance." } }, "montreal": { "tax.filing_basis": { "source": "PwC WWTS - Canada (Individual, Tax administration): married taxpayers file separately; joint returns are not allowed.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "CRA 2026 federal: peer-identical with Toronto/Vancouver", "last_updated": "2026-05-17", "confidence": "high" }, "tax.state_brackets": { "source": "Revenu Québec 2026 (14/19/24/25.75)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "GST 5% + QST 9.975% = 14.975%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Ville de Montréal effective ~0.94%", "last_updated": "2026-05-17", "confidence": "high" }, "housing._summary": { "source": "Centris.ca + Numbeo Montréal Q1 2026", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "RAMQ universal", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "Quebec CPE network ~CAD 8.70/day = ~$190/mo for low-income, ~$380 for typical tech-family income", "last_updated": "2026-05-17", "confidence": "high" }, "food._summary": { "source": "Numbeo Montréal 2026 + Statistics Canada", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "STM monthly CAD 100", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "StatCan SHS + Numbeo Montréal", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "YUL-SEA/LAX 1-stop AC/AS 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Retraite Quebec Allocation famille 2026 (montants)", "last_updated": "2026-07-17", "note": "Quebec Allocation famille has a UNIVERSAL minimum (~CAD 1,221/child in 2026) that never phases to zero, unlike the federal CCB in the rest of Canada. Means-tested only above the floor, so the floor survives at any income (was 6,500, the low-income max)." }, "non_cash._summary": { "source": "Quebec Labour Standards Act + RQAP parental leave 18+32 weeks at 75%", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Montreal: CAD 111,315, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "CAD 111,315 / 1.4 = USD 79,511; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3180, R5-rounded down to 0.31. REPLACES 0.62 (2026-08-11), and the interim 0.33 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Answers sweep B: the note's 'below Toronto (0.70)' understated it. The same-day ratio to Toronto is 0.772. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.62 was 0.8267 of Toronto's 0.75; that ratio x Toronto's re-derived 0.4 = 0.3307, R5-rounded down to 0.33. Toronto's own figure: CAD 142,630 / 1.4 = USD 101,879, / USD 250,000 = 0.4075. This is a BASIS move, not a re-measurement: Montreal's rank inside Canada is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Below Toronto (0.70) due to QC tax burden. Index 0.62" }, "tax.provincial_brackets": { "confidence": "high", "source": "Quebec 2026 indexed brackets/BPA (Finances Quebec AUTEN_IncomeTax2026.pdf); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "tax.federal_abatement_rate": { "source": "Quebec residents: 16.5% refundable abatement of basic federal tax (CRA line 44000); https://www.taxtips.ca/qctax/federal-tax-abatement.htm; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "Reduces basic federal tax 16.5% for QC residents." }, "tax.qpip_rate": { "source": "Quebec Parental Insurance Plan employee premium 0.43% (2026); https://www.canada.ca/en/employment-social-development/programs/ei/ei-list/ei-employers/premium-reduction-program/2026-qpip-rates-multiples.html; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "QPIP employee premium to the 103,000 cap." }, "tax.qpip_cap": { "source": "QPIP maximum insurable earnings 103,000 (2026); https://www.canada.ca/en/employment-social-development/programs/ei/ei-list/ei-employers/premium-reduction-program/2026-qpip-rates-multiples.html; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "QPIP annual cap." }, "tax.bpa_phaseout": { "source": "Federal BPA supplement phases 16,452 -> 14,829 over net income 181,440-258,482 (CRA 2026); https://www.canada.ca/en/revenue-agency/programs/about-canada-revenue-agency-cra/federal-government-budgets/basic-personal-amount.html; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "BPA modeled as a non-refundable credit at the lowest bracket rate with this federal supplement phase-out." }, "tax.cpp_rate": { "source": "Quebec QPP employee rate 6.30% (5.3% base plan + 1.0% additional plan, 2026) replaces CPP; Retraite Quebec https://www.retraitequebec.gouv.qc.ca ; cross-checked against PwC's 2026 QPP employee maximum of CAD 4,895.30, which equals 6.30% of (74,600 - 3,500) plus the 4% QPP2 on 10,400 exactly", "last_updated": "2026-08-01", "note": "Quebec workers pay QPP, not CPP." }, "tax.ei_rate": { "source": "Reduced EI rate for Quebec (EI-QC) 1.30% (2026), QPIP covers parental benefits; https://www.canada.ca/en/employment-social-development/programs/ei/ei-list/ei-employers/premium-reduction-program/2026-qpip-rates-multiples.html; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "Quebec EI premium is reduced because QPIP funds parental leave." }, "childcare.subsidy_per_kid_annual": { "source": "CWELCC (Canada-Wide Early Learning and Child Care) provincial fee schedules; double-subtraction correction 2026-08-02", "last_updated": "2026-08-02", "confidence": "medium", "note": "Set to 0 on 2026-08-02, was CAD 5,000. The old value was the CWELCC fee reduction. CWELCC is delivered as a lower posted fee at a participating centre, not as cash to the household, and childcare.preschool_monthly already carries the reduced rate, so the reduction was netted twice. Quebec’s reduced contribution is about CAD 9.35/day, and the modelled CAD 380/month already sits at that level; netting CAD 5,000 on top drove the annual bill negative before the floor caught it. This field holds CASH THAT REACHES THE HOUSEHOLD. Where the monthly rate it nets against is already the price the household actually pays, the value is 0 and the relief is described here instead. See docs/superpowers/research/2026-08-02-childcare-double-subtraction.md. Open follow-up: the Montreal monthly rate itself is a blend of the capped and uncapped rates and still owes its own sourcing pass." }, "non_cash.parental_leave_weeks": { "source": "Act respecting parental insurance (QPIP basic plan): maternity 18 wk + paternity 5 wk + parental 32 wk shared; rqap.gouv.qc.ca (2026 maximum insurable earnings CAD 103,000, confirmed via Revenu Quebec / RQAP news 2026-08-06; the rqap.gouv.qc.ca direct fetch failed on a certificate error)", "last_updated": "2026-08-06", "confidence": "high", "note": "55 = QPIP basic plan: maternity 18 (mother) + paternity 5 (father) + parental 32 (shared pot counted once). Quebec runs QPIP instead of EI parental benefits, so this legitimately diverges from the other Canadian cities (city-varying by design in data/_national_fields.json). The special plan (43 wk all at 75%) is the elective alternative; basic is the default and reference plan, and the FPE gap between plans is under 8%. Was 50/75, an EI week count wearing the QPIP special-plan rate. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md." }, "non_cash.parental_leave_paid_pct": { "source": "QPIP basic plan rates: maternity and paternity 70%, parental 7 wk at 70% then 25 wk at 55%, on insurable earnings up to CAD 103,000 (2026, up from 98,000 in 2025)", "last_updated": "2026-08-06", "confidence": "high", "note": "63 = blend across 55 wk: 30 wk at 70 (maternity 18 + paternity 5 + first 7 parental) + 25 wk at 55: (2100 + 1375) / 55 = 63. REFUTATION recorded: a third-party calculator's ~CAD 94,000 cap for 2026 is the stale 2024 figure; RQAP and Revenu Quebec say 98,000 (2025) then 103,000 (2026). The cap is high enough that the 63 is close to real for much of the Montreal ICP band, unlike the ROC's CAD 68,900." }, "housing.homeowners_insurance_pct_of_value": { "source": "Provincial average premium (rates.ca Q2 2026, MyChoice city data, IBC) over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "CIHI national health expenditure household out-of-pocket + provincial drug plans: Quebec RAMQ covers a registered parent's children free, Ontario Trillium and BC Fair PharmaCare are income-deductible-based", "last_updated": "2026-08-08", "confidence": "medium", "note": "Medically necessary physician and hospital care is free everywhere, so the whole of this field is prescription drugs and allied health, which is PROVINCIAL, and the existing per-city spread is kept for that reason. Employer drug plans are a benefit and a premium, not out-of-pocket. Montreal carries a much smaller child increment than the other cities because of the RAMQ free-child rule. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "CIHI national health expenditure household out-of-pocket + provincial drug plans: Quebec RAMQ covers a registered parent's children free, Ontario Trillium and BC Fair PharmaCare are income-deductible-based", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Statistics Canada, average commute times for the 15 largest CMAs, May 2025 (The Daily, 2025-08-26, Table 4). https://www150.statcan.gc.ca/n1/daily-quotidien/250826/t004a-eng.htm", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Commuting duration, one way", "year": "2025", "note": "Montreal CMA 29.0." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "CRA, MP, DB, RRSP, DPSP, ALDA, TFSA limits and the YMPE https://www.canada.ca/en/revenue-agency/services/tax/registered-plans-administrators/pspa/mp-rrsp-dpsp-tfsa-limits-ympe.html ; https://www.canada.ca/en/revenue-agency/services/tax/individuals/topics/rrsps-related-plans/contributing-a-rrsp-prpp/contributions-affect-your-rrsp-prpp-deduction-limit.html", "last_updated": "2026-08-11", "confidence": "high", "note": "RRSP deduction limit (line 20800), 2026: 18% of prior-year earned income to an annual dollar limit. Pinned at 33,810 via the statutory identity RRSP limit year X = money-purchase limit year X-1 (CRA confirms 2025 MP = 33,810). The PENSION ADJUSTMENT is deliberately NOT modeled, so this is the no-workplace-DB case; a DB-plan member has materially less room. Carry-forward of unused room is also unmodeled. Never binds at our incomes; the 18% governs below CAD 187,833." } }, "frankfurt": { "housing.median_rent_by_bedroom": { "source": "VALUE Marktdatenbank 2026, nettokalt asking rents, new construction excluded https://mietspiegeltabelle.de/mietspiegel-frankfurt-am-main/ ; dwelling sizes and basis reconciliation from Stadt Frankfurt am Main, Buergeramt Statistik und Wahlen: Frankfurter Statistische Berichte 2024 (Mikrozensus Zusatzprogramm Wohnen 2022) https://statistikportal.frankfurt.de/download/FSB/FSB_2024.pdf and STATISTIK.AKTUELL 13/2024, Gebaeude- und Wohnungszaehlung 2022 https://statistikportal.frankfurt.de/statistik_aktuell/2024/FSA_2024_13_GWZ_2022.html", "last_updated": "2026-08-07", "confidence": "high", "note": "CORRECTED 2026-08-02 from 1000/1300/1700/2400, replacing an \"ImmoScout24 + Immowelt Q1 2026\" citation that carried no URL. Five readings of Frankfurt citywide asking rent were adjudicated: ImmoScout24 13.95 to 14.99, VALUE 2025 17.29, VALUE 2026 17.94, Engel und Voelkers 19.36, and the Mietspiegel 12.28 which is stock and not comparable. ImmoScout24 is the low outlier and is contradicted by the same agent own district figures (Westend 21.80, Nordend 20.40, Sachsenhausen 19.10), which cannot coexist with a genuine 14.99 citywide. VALUE 2026 adopted: reproducible across two vintages, the only source publishing a by-Zimmer gradient, mid-range rather than the most aggressive reading, and it excludes Neubau so it understates what an arrival actually pays. Per-room rates 20.41/17.49/17.07/17.30 x 1.0587 = 21.61/18.52/18.07/18.32. Studio 1000 to 910, 1BR 1300 to 1110, 2BR 1700 to 1540, 3BR 2400 to 1920. The city own Mikrozensus corroborates both the direction and the U-shaped size gradient: Frankfurt renter households average 68.0 sqm (Berlin 68.2, so Berlin size assumptions are not absurd here), brutto kalt runs 15.60 per sqm under 40 sqm against 10.70 at 60 to 80 sqm, and households that moved in after 2019 pay 14.10 against 9.30 for pre-1999 tenancies, a 52% spread inside one official dataset that independently confirms stock rent badly understates a new lease. Netto kalt is about 91.2% of brutto kalt in Frankfurt. Per-room rates come from the VALUE Marktdatenbank 2026 cut published at mietspiegeltabelle.de, explicitly nettokalt and explicitly market/asking rather than the qualified municipal Mietspiegel. That is the same database family the Berlin correction leant on. Leipzig proves what it serves: its 10.45 equals VALUE own published MEAN of 10.49 against a median of 10.00, so this basis runs about 5% above median, which is the conservative direction. Method is the shipped Berlin recipe re-run per city, so all eight German cities sit on one footing: rent = size x city per-room EUR/sqm x 1.0587. The 1.0587 is Berlin own uplift over this same source (Berlin shipped a 16.6 base against 15.68 here), and re-running it reproduces Berlin 760/970/1360/1750 to the euro. Sizes are held constant across cities at 42/60/85/105 sqm on the German Zimmer convention (our 1BR = 2-Zimmer-Wohnung), because the product prices an equivalent standard of living rather than how much space local households ration themselves into at local prices. Duesseldorf publishes real offered-flat sizes of 37/58/83/111, so the fixed 42 makes the studio the most conservative cell and the fixed 105 makes the 3BR mildly generous on size; the two roughly cancel. Basis is Angebotsmiete (new lease), netto kalt, never the Mietspiegel stock rent. Rents exclude utilities, insurance and internet, which are separate fields. Values are a CEILING per the conservative value-bias rule, not central estimates. Working: docs/superpowers/research/2026-08-02-german-city-rents.md SIZE LADDER CORRECTED 2026-08-07 (small-fixes wave 1, coordinated eight-city pass incl. Berlin): the common dwelling-size ladder moves from the inherited 42/60/85/105 sqm to 37/58/83/111 sqm, Duesseldorf's PUBLISHED offered-flat sizes (Wohnungsmarktbericht 2025 Abb. 54, MFH column, n=10,370 listings; the only published per-Zimmer size table in the German set, and the conservative top of the seven-city evidence range at 3-Zi 83 and 4-Zi 111). Root cause of the old ladder: German official statistics count a closed kitchen over 6 sqm as a room, so census room distributions ran one room high against listing Zimmer counts; the 42 sqm studio was priced at the 1-Zimmer premium rate measured on 33-37 sqm stock (double-counting the small-flat premium) and the 4-Zimmer was undersized at 105 vs the real ~111. Values 910/1110/1540/1920 -> 800/1070/1500/2030. Per-room rates and the 1.0587 uplift are UNCHANGED; sizes remain COMMON across all eight German cities because the product prices an equivalent standard of living. Working: docs/superpowers/research/2026-08-07-small-fixes.md" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "non_cash.parental_leave_paid_pct": { "source": "SGB V para 24i (Mutterschaftsgeld) https://www.gesetze-im-internet.de/sgb_5/__24i.html ; MuSchG para 20 (Zuschuss zum Mutterschaftsgeld, employer tops up to full net) https://www.gesetze-im-internet.de/muschg_2018/__20.html", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the percentage-paid remainder of the 67 weeks: the 14.1 Mutterschutz weeks (6 pre-birth + 8 post-birth), where para 24i SGB V pays EUR 13/day and para 20 MuSchG obliges the employer to top up to full net pay with no ceiling. Was 72, a blend that averaged the capped Elterngeld slice into this field; since 2026-08-06 the 52.87 Elterngeld weeks live in non_cash.parental_leave_capped (65% up to EUR 1,800/month, nothing above EUR 175,000 household income) and parentalLeaveEffectiveRate() blends the slices at the user's salary, so every reader-facing surface already prices the cap and the cliff." }, "non_cash.parental_leave_capped": { "source": "BEEG para 2 (Elterngeld: 65% of prior income, minimum EUR 300 and maximum EUR 1,800 per month) https://www.gesetze-im-internet.de/beeg/__2.html ; BEEG para 1 Abs. 8 (no Elterngeld above EUR 175,000 combined zu versteuerndes Einkommen) https://www.gesetze-im-internet.de/beeg/__1.html ; MuSchG para 20 (the 14 Mutterschutz weeks stay OUTSIDE this block: the employer net top-up has no ceiling) https://www.gesetze-im-internet.de/muschg_2018/__20.html", "last_updated": "2026-08-06", "confidence": "high", "note": "52.87 of the 67 weeks: the Elterngeld pot (60.87 wk) minus the 8 overlap weeks counted at the Mutterschutz rate. The EUR 1,800/month cap (unindexed since Elterngeld's 2007 introduction, re-confirmed 2026-08-06) binds above ~EUR 33,000 gross, i.e. at every ICP salary: at EUR 90,000 this slice pays ~24% of gross and the blended Berlin rate is ~40%; at EUR 150,000 the slice pays ~14% and the blend is ~32%, versus the 72 this field carried before the 2026-08-06 caps rebuild. BEEG para 1 Abs. 8 zeroes the slice for households above EUR 175,000: the engine compares HOUSEHOLD GROSS against the ceiling as a conservative proxy for zu versteuerndes Einkommen (gross exceeds taxable income, so the cliff bites earlier, the direction less favorable to Germany). Same conservative direction on the rate basis: the statute pays 65% of Elterngeld-Netto while the engine applies 65% to gross, and above ~EUR 33,000 gross the cap binds under both readings, so the distinction is immaterial at any ICP salary. Floor EUR 300/month (BEEG para 2). Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Germany section) and 2026-08-06-caps-engine-build.md." }, "non_cash.parental_leave_weeks": { "source": "MuSchG para 3 (Schutzfristen, 6 weeks before and 8 after) https://www.gesetze-im-internet.de/muschg_2018/__3.html ; BEEG para 4 Abs. 3 (12 Monatsbetraege Basiselterngeld plus 2 Partnermonate, gemeinsam) and para 4 Abs. 4 Satz 3 https://www.gesetze-im-internet.de/beeg/__4.html", "last_updated": "2026-08-01", "confidence": "high", "note": "67 wk = 6 additive pre-birth Mutterschutz weeks + the 14-month Elterngeld pot (60.87 wk). Was 61, the Elterngeld pot alone with Mutterschutz omitted. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. OVERLAP RULE APPLIED: the 8 post-birth Mutterschutz weeks are NOT stacked on top, because BEEG para 4 Abs. 4 Satz 3 deems Lebensmonate carrying Mutterschaftsgeld to be consumed Basiselterngeld months (gelten als Monate, fuer die dieser Elternteil Basiselterngeld bezieht). They are counted once, at the higher rate. Germany therefore reaches 67 and not 75. Elterngeld is a SHARED household pot, so it is not doubled. GERMANY HAS NO STATUTORY PAID PATERNITY LEAVE: the Familienstartzeit bill transposing EU Directive 2019/1158 art. 4 was still not law on 2026-08-01, so zero paternity weeks. Excluded: ElterngeldPlus (an election that halves the monthly amount to double the months) and the Partnerschaftsbonus (conditional on both parents working reduced hours)." }, "tax.federal_brackets": { "source": "BMF Einkommensteuertarif 2026: peer-identical with Berlin/Munich", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Umsatzsteuer 19% (national)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Frankfurt Hebesatz 500% × Steuermesszahl 0.31‰ ≈ 0.155% effective", "last_updated": "2026-05-17", "confidence": "high" }, "housing._summary": { "source": "ImmoScout24 + Immowelt Frankfurt Q1 2026", "last_updated": "2026-05-17", "confidence": "high" }, "healthcare._summary": { "source": "GKV universal via social_insurance", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "Hessen Kita Beiträge means-tested; free 3+", "last_updated": "2026-05-17", "confidence": "high" }, "food._summary": { "source": "Statistisches Bundesamt + Numbeo FFM", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "RMV Frankfurt monthly EUR 100", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "Destatis SOEP + Numbeo FFM", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "FRA-SEA/LAX nonstop Lufthansa 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Kindergeld federal EUR 250/mo = ~3000/yr/kid", "last_updated": "2026-05-17", "confidence": "high", "note": "German Kindergeld (EUR 259/mo per child = EUR 3,108/yr, which is the stored value): universal, NOT means-tested, and it does not taper at professional incomes. Corrected 2026-08-02 from a stale \"EUR 255/mo\", which is the 2025 rate; the Steuerfortentwicklungsgesetz stepped Kindergeld to EUR 255 on 2025-01-01 and EUR 259 on 2026-01-01. The stored value was already correct and was NOT changed." }, "non_cash._summary": { "source": "BUrlG 28 days + Elterngeld + Entgeltfortzahlung", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Frankfurt: EUR 77,102, fetched live 2026-08-11 and recorded at sweep B F8b of docs/superpowers/research/2026-08-11-comp-index-sweep-b.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 77,102 / 0.867 = USD 88,930; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3557, R5-rounded down to 0.35. REPLACES 0.63 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Frankfurt by 80%. The measured p50 was 0.565 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Finance premium pushes above Berlin (0.50). Index 0.55" }, "tax.income_tax_tariff_32a": { "source": "§ 32a Abs. 1 und Abs. 5 EStG, Veranlagungszeitraum 2026 https://www.gesetze-im-internet.de/estg/__32a.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Statutory continuous tariff, identical nationwide (see berlin for the verbatim zone text and the coefficient cross-check). Replaces the 5-bracket approximation that flattened the 17,799-69,878 progression zone to 33.5%. Ehegattensplitting (Abs. 5) is now an explicit 2 x tariff(joint zvE / 2); doubling bracket edges only reproduced that for a piecewise-linear schedule." }, "tax.werbungskosten_pauschbetrag": { "source": "§ 9a Satz 1 Nr. 1 Buchst. a EStG https://www.gesetze-im-internet.de/estg/__9a.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Arbeitnehmer-Pauschbetrag EUR 1,230 per employed person. National, identical to berlin." }, "tax.sonderausgaben_pauschbetrag": { "source": "§ 10c EStG https://www.gesetze-im-internet.de/estg/__10c.html", "last_updated": "2026-07-31", "confidence": "high", "note": "EUR 36, doubled on Zusammenveranlagung. National, identical to berlin." }, "tax.vorsorge_deductible_rate_rv": { "source": "§ 10 Abs. 1 Nr. 2 und Abs. 3 EStG https://www.gesetze-im-internet.de/estg/__10.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Employee RV share 9.3%, 100% deductible since 2023, against the RV Beitragsbemessungsgrenze. National, identical to berlin." }, "tax.vorsorge_deductible_rate_kvpv": { "source": "§ 10 Abs. 1 Nr. 3 und Abs. 4 Satz 4 EStG https://www.gesetze-im-internet.de/estg/__10.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Basis KV 8.75% x 0.96 (Krankengeld haircut) + PV 1.8% = 10.2%, against the KV/PV Beitragsbemessungsgrenze; ALV not deductible. National, identical to berlin." }, "tax.solidaritaetszuschlag_threshold_married": { "source": "Soli Freigrenze 2026 joint assessment EUR 40,700 (2 x 20,350 single; tk.de Solidaritaetszuschlag); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.social_insurance_rate_rvalv": { "source": "RV 9.3% + ALV 1.3% employee share = 10.6%, runs to the higher Beitragsbemessungsgrenze (lohn-info.de Sozialversicherungsbeitraege 2026); verified tax audit 2026-07-14", "confidence": "high", "note": "Carved out of the combined 21.15% employee share; the KV/PV remainder (10.55%) stops at the lower ceiling.", "last_updated": "2026-07-17" }, "tax.social_insurance_cap_rvalv": { "source": "Beitragsbemessungsgrenze RV/ALV 2026 = EUR 101,400 (bundesregierung.de/breg-de/aktuelles/beitragsgemessungsgrenzen-2386514); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.pv_childless_surcharge_rate": { "source": "PV Kinderlosenzuschlag § 55 Abs. 3 SGB XI: +0.6pp employee-borne for childless members 23+ (dak.de Beitrag zur Pflegeversicherung Zuschlag Kinderlose; vdek.com Aenderungen Gesundheitswesen 2026); found by the 2026-08-06 corrections wave", "confidence": "high", "note": "Applied when the household has no children; the under-23 exemption is not modeled (ICP is older). Deductible with the rest of the Basis-Pflegeversicherung.", "last_updated": "2026-08-06" }, "childcare.subsidy_per_kid_annual": { "source": "Land and municipal Kitabeitrag schedules; double-subtraction correction 2026-08-02", "last_updated": "2026-08-02", "confidence": "high", "note": "Set to 0 on 2026-08-02, was EUR 3,600. The old value was the VALUE OF THE FREE OR SUBSIDISED KITA HOURS, but childcare.preschool_monthly already carries the post-free-tier out-of-pocket, so computeChildcareCost subtracted the same relief a second time and clamped the Frankfurt household bill to zero. German Kita relief is a benefit in kind delivered as a lower or absent fee, never a payment to the parent, so nothing belongs in this field. Frankfurt now bills EUR 0/month from age 1, since the city extends Hesse’s fee-free Kindergarten below age 3. This field holds CASH THAT REACHES THE HOUSEHOLD. Where the monthly rate it nets against is already the price the household actually pays, the value is 0 and the relief is described here instead. See docs/superpowers/research/2026-08-02-childcare-double-subtraction.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Destatis Gesundheitsausgabenrechnung 2024 (private households + NPISH EUR 65bn) + Destatis EVS 2023 Konsumausgaben nach Haushaltstyp + GKV Zuzahlungen aggregate EUR 5.18bn (BMG via Aerzteblatt) + SGB V paras 61-62", "last_updated": "2026-08-08", "confidence": "medium", "note": "Level essentially CONFIRMED at EUR 311 per adult (778/capita gross, ~40% in scope after netting long-term care 46%, dental 9%, vision ~3pp). What Germany needed was a curve, not a level. The curve is steep in adults and nearly flat in children because under-18s are exempt from every Zuzahlung except transport AND each child subtracts EUR 9,756 from the income base before the 2%-of-gross Belastungsgrenze, so adding a child LOWERS the household's statutory ceiling in absolute euros. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Destatis Gesundheitsausgabenrechnung 2024 (private households + NPISH EUR 65bn) + Destatis EVS 2023 Konsumausgaben nach Haushaltstyp + GKV Zuzahlungen aggregate EUR 5.18bn (BMG via Aerzteblatt) + SGB V paras 61-62", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, urb_ctran indicator TT1019V, city DE005C. https://ec.europa.eu/eurostat/databrowser/view/urb_ctran/default/table?lang=en", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average time of journey to work (minutes)", "year": "2016", "note": "27.0 min, BASIS YEAR 2016. CITY geography, narrower than our metro unit, so biased LOW and flattering. This matters MORE here than anywhere else in the German block: Frankfurt draws exceptionally heavy inbound commuting that a city cut cannot see, and this is the largest downward move in the block at -8. Treat it as the weakest German row and the first to re-check. Survey self-report, standard half-up rounding. METRO-LEVEL GERMAN SERIES: ESTABLISHED HARD NEGATIVE, do not re-run. MiD 2023 Tabelle 6 carries RegioStaR7 breakdowns of population, trips, distance and total time but NO duration by purpose; the national banner table W9 Wegedauer is all-purpose and sits on an \"Ohne regelmaessige berufliche Wege\" base; the small-area products regionalise headline indicators only; and Eurostat publishes no functional-urban-area rows for TT1019V. SrV 2023 was rejected because it publishes an ALL-PURPOSE trip duration, and swapping an all-purpose duration for a work-purpose one is a UNIT error, which is worse than the geography error we accepted. SrV is therefore used only as rank-order corroboration, and it agrees with Eurostat against our old stored spread: both put Frankfurt near the BOTTOM of the big five, where we had it second highest. PROVISIONAL: this is one of the three large German drops and it is the flattering direction, so re-check it first if a metro-level series ever appears." }, "non_cash.sick_leave": { "source": "Entgeltfortzahlungsgesetz (EFZG) §3 (employer phase); SGB V §47 (Krankengeld; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "housing.utilities_monthly": { "source": "DMB Hessen Betriebskostenspiegel, Frankfurt city figure 3,02 EUR/sqm (\"die mit Abstand teuerste Stadt in Hessen\"), https://www.dmb-hessen.de/service/betriebskostenspiegel-d", "confidence": "medium-high", "note": "2BR reference: 3,02 x 83 + 63 electricity = 314 bottom-up; Numbeo 85 sqm 336.30. Resolved central-up to 325.", "last_updated": "2026-08-08" }, "tax.retirement_deduction_cap": { "source": "EStG s.10 Abs 3 https://www.gesetze-im-internet.de/estg/__10.html ; SVBezGrV 2026 s.4 Abs 1 https://www.gesetze-im-internet.de/svbezgrv_2026/BJNR1160A0025.html", "last_updated": "2026-08-11", "confidence": "high", "note": "Ruerup/Basisrente Hoechstbetrag, 2026: EUR 30,825.60 single (knappschaftliche RV BBG 124,800 x 24.7%). A DEDUCTION ceiling and a COMBINED one: employee AND employer statutory pension contributions consume the same pot. less_employer_pct 0.186 = 2 x 9.3% and employer_base_cap 101,400 is the allgemeine RV BBG, because computeGermanyTaxes already deducts the employee 9.3% via vorsorge; a naked max would hand a EUR 95k earner about EUR 8.8k of phantom deduction. Riester (EUR 2,100) and bAV Entgeltumwandlung (8% of the BBG, EUR 8,112) are deliberately NOT stacked. Married couples: the statute doubles the ceiling to 61,651.20 even on one income; our per-earner cap under-deducts there, which errs toward higher tax." } }, "basel": { "tax.federal_brackets": { "source": "ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Exact 2026 Grundtarif (kalte-Progression vintage confirmed against the 2025 circular). 2026-08-07: restored the missing 5.94% band 76,200-82,100 (the stored table jumped 2.97% to 6.6%, overstating by up to CHF 38.94/yr). Applied to taxable income = gross minus the statutory deduction stack (see tax.payroll_deductible)." }, "tax.cantonal_brackets": { "source": "Kanton BS, Uebersicht Einkommenssteuertarife 2018-2026 (18.12.2025) https://media.bs.ch/original_file/d8cf2817b481307cd7943afa4f80b99a8878297a/einkommenssteuertarife-2019-2026.pdf ; Steuertarife-Booklet 17000.a.01.25 https://media.bs.ch/original_file/47a483fc991f91d59b9a86e4fbbbc0a9de77bd79/17000-steuertarife-2025.pdf ; ESTV Kantonsblatt BS (Stand Feb 2026) https://www.estv2.admin.ch/stp/kb/bs-de.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Exact BS Paragraph-36 Tarif A 2026 (21.00% to 212,500 / 27.25% to 316,300 / 28.25%; marginal, 2025=2026 per the official overview) with the Paragraph-35 Sozialabzug single CHF 19,500 baked in as the zero band + threshold shift (exact deduction semantics). City of Basel levies NO communal income tax (Paragraph 2 Abs. 3 StG), so this is the whole cantonal+communal burden. Replaces the 2026-07-14 calibrated scale, which matched no BS tariff year (audit \"too coarse ~6pp\" confirmed: 12.1k vs 21.0k at 100k taxable). Sanity pin: taxable 100,000 -> CHF 21,000.00 (printed table value)." }, "tax.social_insurance_rate": { "source": "AHV/IV/ALV employee 6.4%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "ESTV MwSt: standard rate 8.1% since 2024-01-01 (was 7.7%) https://taxsummaries.pwc.com/switzerland/individual/other-taxes", "last_updated": "2026-07-14", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "BS exempts primary residence from real estate tax", "last_updated": "2026-05-17", "confidence": "high" }, "housing._summary": { "source": "Homegate + Comparis Basel Q1 2026", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "KVG mandatory private; Basel premiums per BAG monitoring", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "BS subsidized Kita scale CHF 110-145/day", "last_updated": "2026-05-17", "confidence": "high" }, "food._summary": { "source": "BFS + Numbeo Basel", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "BVB Umweltabo monthly CHF 80", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "BFS HABE + Numbeo Basel", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "BSL/EAP-SEA/LAX 1-stop via FRA/ZRH 2025-26", "last_updated": "2026-05-17", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Familienzulagen BS CHF 250/mo", "last_updated": "2026-05-17", "confidence": "high", "note": "Swiss Kinderzulagen / allocations familiales: cantonal, universal minimum (>= CHF 200-250/mo), NOT means-tested." }, "non_cash._summary": { "source": "OR vacation 25 days + Mutterschaftsversicherung 14wk 80%", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Basel: CHF 146,145, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "CHF 146,145 / CHF 143,634 (Zurich, fetched the SAME DAY) = 1.0175 of Zurich; x Zurich's 0.6832 = 0.6952, R5-rounded down to 0.69. A SAME-DAY RATIO is used here rather than the direct CHF/FX/250,000 division used elsewhere, and only here, because Zurich is the one country anchor whose stored table has drifted materially: stored CHF 137,843 against CHF 143,634 live on 2026-08-11, 4.2% apart. Dividing a fresh satellite figure by 250,000 while the anchor sits on a 4.2%-stale table leaks that gap into every Swiss satellite in the FLATTERING direction. Every other same-day anchor drifted under 3% and in the conservative direction (Oslo 2.8%, Sweden 1.6%, Denmark 1.3%, Germany 1.0%), so none of them needed this. REPLACES 0.95 (2026-08-11), and the interim 0.61 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Measures 1.7% ABOVE Zurich at the median, but Zurich's p90 is CHF 329,000 against Basel's 181,000, so read them as tied rather than Basel ahead. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.95 was 0.9048 of Zurich's 1.05; that ratio x Zurich's re-derived 0.68 = 0.6152, R5-rounded down to 0.61. Zurich's own figure: CHF 137,843 / 0.807 = USD 170,809, / USD 250,000 = 0.6832. This is a BASIS move, not a re-measurement: Basel's rank inside Switzerland is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Pharma premium. Index 0.95" }, "childcare.subsidy_per_kid_annual": { "source": "bern.ch Betreuungsgutscheine; bs.ch Betreuungsbeitraege; lausanne.ch Reseau-L 2025", "last_updated": "2026-07-17", "note": "Municipal childcare subsidy is means-tested and 0 for the CHF 200k+ household: Bern Betreuungsgutscheine cut off at CHF 160k massgebendes Einkommen; Lausanne Reseau-L means-tested reduction hits the max tariff; Basel-Stadt subsidized places are income-prioritized/capacity-limited, so a high earner is modeled at the private-Kita market rate (infant_monthly). Was 8,000/4,800/2,400." }, "tax.bvg": { "source": "BVG statutory minimum, 2026 parameters (unchanged from 2025): entry threshold CHF 22,680, coordination deduction CHF 26,460, insured-salary upper limit CHF 90,720, coordinated salary clamped to 3,780-64,260. ASGA Pensionskasse, Merkblatt Renten und Grenzbetraege per 1.1.2026 https://www.asga.ch/wp-content/uploads/downloads/merkblaetter/merkblaetter-d/renten-und-grenzbetraege_2026_d.pdf ; BSV https://www.bsv.admin.ch/de/altersvorsorge-beruflichen-vorsorge (confirms the 22,680 threshold, savings from age 25, and the BVG Art. 66 employer-parity rule); age credits BVG Art. 16: 7/10/15/18% of coordinated salary by age band (Sicherheitsfonds BVG, sfbvg.ch Sparprozess), employee pays half: 3.5/5/7.5/9%.", "last_updated": "2026-08-06", "confidence": "high", "note": "Statutory MINIMUM only, deliberately: real Swiss professional plans insure above the mandatory band (ueberobligatorisch) and add plan-specific risk/admin premiums, so a typical payslip deducts MORE than this charge; everything above the legal floor is excluded under the statutory-minimum convention (the same rule that keeps Swedish ITP and Danish employer pensions out entirely). The charge docks take-home inside taxes.payroll and is carved to the forced-savings own-money row (funded, FZG-vested, portable). It runs on the same wage base as the AHV/ALV slice and never reduces the income-tax base: the CH-1 deduction floor stays CHF 13,500/27,000, which errs against Switzerland. Applied per earner (entry threshold and coordination deduction per person); the 18-24 bracket pays 0 because retirement credits start at 25 (BVG Art. 7). Decision log: docs/superpowers/research/2026-08-06-pension-omission-pair.md" }, "non_cash.parental_leave_weeks": { "source": "BSV, Maternity, paternity and adoption leave and allowances https://www.bsv.admin.ch/en/income-compensation-parents ; LECA/EOG; ahv-iv.ch leaflets 6.02 (maternity) and 6.04 (paternity)", "last_updated": "2026-08-06", "confidence": "high", "note": "16 wk = 14 maternity (EO, continuous from birth) + 2 paternity (10 working days within 6 months, federal statute since 1 Jan 2021), each parent's own entitlement; no federal shared scheme exists. The previously stored value omitted the 2 paternity weeks its own note cited. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md." }, "non_cash.parental_leave_paid_pct": { "source": "BSV https://www.bsv.admin.ch/en/income-compensation-parents : up to 80% of average income, max. CHF 220 per day, for both the maternity and paternity allowances", "last_updated": "2026-08-06", "confidence": "high", "note": "80% capped at CHF 220/day = CHF 1,540/wk, a cap reached at about CHF 99,000/yr, so every ICP salary receives the flat maximum: at CHF 130,000 the effective rate is about 62% and falls as salary rises. Kept as pct 80 because below the threshold it is a true 80%; the capped-percentage shape has no honest flat-block representation, so the cap arithmetic lives here." }, "tax.federal_brackets_married": { "source": "ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Exact 2026 Verheiratetentarif (Art. 36 Abs. 2 DBG), marginal form with the 11.5% flat crossover at 941,300. Replaces the divisor-2 full-splitting approximation, which UNDERSTATED federal married tax (exact 7,579.60 vs splitting 3,185.42 at joint taxable 166,920) - the federal tariff is harsher than splitting. All checkpoints re-derived by hand; see docs/superpowers/research/2026-08-08-ch-federal-worksheet.md." }, "tax.payroll_deductible": { "source": "Art. 33 Abs. 1 lit. d + f DBG SR 642.11, consolidation in force 2026-01-01 (fedlex)", "last_updated": "2026-08-07", "confidence": "high", "note": "Employee AHV/IV/EO, ALV and BVG contributions are fully deductible from the income-tax base (salary-certificate net is the taxable base). Gates the per-earner SI+BVG deduction in computeSwitzerlandTaxes; the charged amounts are the deducted amounts, so the FORCED_SAVINGS carve cannot double-count. Replaces the frozen CHF 13,500/27,000 CH-1 floor (2026-07-26), which bundled SI at one income level." }, "tax.professional_forfait": { "source": "Berufskostenverordnung SR 642.118.1 Art. 7 + Anhang Ziff. 1 (fedlex); 2026 min/max confirmed in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Uebrige Berufskosten pauschale: 3% of NET salary (post-SI/BVG), min CHF 2,000 / max CHF 4,000, per earner with income. Commuting and meal deductions excluded (variable/behavioral, conservative)." }, "tax.insurance_deduction_single": { "source": "Art. 33 Abs. 1 lit. g DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amount in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Versicherungsabzug CHF 1,800 (taxpayer WITH pillar-2 contributions, the employee case). Real KVG premiums exceed this cap and some cantons deduct more (BS 4,200, GE full premiums) - federal amount kept for the one-base model, conservative." }, "tax.insurance_deduction_married": { "source": "Art. 33 Abs. 1 lit. g DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amount in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Versicherungsabzug CHF 3,700 married (WITH pillar-2 amounts). Per-child +700 supplement excluded (conservative)." }, "tax.married_deduction": { "source": "Art. 35 Abs. 1 lit. c DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amount in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Verheiratetenabzug CHF 2,800 (2026), automatic for married couples." }, "tax.two_earner_deduction": { "source": "Art. 33 Abs. 2 DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amounts in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Zweitverdienerabzug: 50% of the lower earned income, min CHF 8,600 / max CHF 14,100, capped at the lower income; applies only when both spouses have earned income. Federal amount used for the one-base model (cantonal equivalents are smaller: BS 1,100, ZH ~6,100); partially offset by the larger cantonal insurance deductions we exclude - see the reasoning in docs/superpowers/research/2026-08-08-tax-structural-ch-ie.md." }, "tax.cantonal_brackets_married": { "source": "Same three BS/ESTV primary sources as tax.cantonal_brackets", "last_updated": "2026-08-07", "confidence": "high", "note": "Exact BS Paragraph-36 Tarif B 2026 (21.00% to 424,900 / 27.25% to 632,500 / 28.25%) with the married Sozialabzug CHF 38,000 baked in as the zero band + shift. A legally distinct schedule whose thresholds are the (separately rounded) doubled single thresholds; stored as the schedule, NOT a divisor, because divisor-2 on the shifted single scale would mis-state the zero band (39,000 vs 38,000). Replaces the pre-2019-vintage 22%/26% @ 400,000 leftover that contradicted the dead divisor also stored here." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "BAG Kostenbeteiligung https://www.bag.admin.ch/de/krankenversicherung-kostenbeteiligung-fuer-in-der-schweiz-wohnhafte-versicherte + BAG OKP statutory cost-sharing aggregate CHF 5.7bn (2024) over 8.96m insured, adult/child split from BAG age-group net benefits", "last_updated": "2026-08-08", "confidence": "high", "note": "Blended CHF 636 per insured person; children ~CHF 140 each (zero franchise, halved CHF 350 Selbstbehalt cap, exempt from the hospital per-diem), leaving CHF 749 per adult, plus a ~7% conservative uplift for non-KVG self-pay that is neither dental nor long-term care. REPLACES a value that was a single adult's statutory MAXIMUM: the old CHF 1,200 was above the CHF 1,000 minimum-franchise ceiling and was billed unchanged to a family of four whose statutory maximum is CHF 2,700. Geneva keeps its existing 1.083 premium, which is a utilisation difference, not a rules difference: cost-sharing law is federal and identical in every canton. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "BAG Kostenbeteiligung https://www.bag.admin.ch/de/krankenversicherung-kostenbeteiligung-fuer-in-der-schweiz-wohnhafte-versicherte + BAG OKP statutory cost-sharing aggregate CHF 5.7bn (2024) over 8.96m insured, adult/child split from BAG age-group net benefits", "last_updated": "2026-08-08", "confidence": "high", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Swiss Federal Statistical Office (BFS), PxWeb cube px-x-2105000000_206, nationale Agglomeration Basel. https://www.pxweb.bfs.admin.ch/pxweb/de/px-x-2105000000_206/px-x-2105000000_206/px-x-2105000000_206.px", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Durchschnittliche Arbeitswegzeit, in Minuten (ein Hinweg)", "year": "2020", "note": "30.9 min. GEOGRAPHY: Agglomeration is the Swiss metro unit and matches our basis. Basel is the NATIONALE Agglomeration, i.e. the Swiss side only, so it EXCLUDES the longer cross-border commutes from France and Germany and is therefore biased short. BFS defines the statistic as one Hinweg (one outbound leg), excludes home workers, and reports the Ausbildungsweg separately." }, "non_cash.sick_leave": { "source": "Code of Obligations / Obligationenrecht (OR/CO) Art. 324a, paragraphs 1-4.; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Steuerverwaltung Kanton Bern, Taxinfo Saeule 3a https://taxinfo.sv.fin.be.ch/taxinfo/a2830da2-1a00-4928-b58b-74da5d48a598 ; EFD/ESTV confirmation of the 2026 amounts, 17 Nov 2025 https://www.estv.admin.ch/de/newnsb/xgRMirCsezICX4rtof9Lm", "last_updated": "2026-08-11", "confidence": "high", "note": "Saeule 3a kleiner Abzug, CHF 7,258 for 2026 (unchanged from 2025), for anyone who also contributes to a Pensionskasse, which our ICP does (both CH cities carry a bvg block). The grosser Abzug (20% of income to CHF 36,288, no 2nd pillar) is NOT our case. Employee-only; ESTV states the maximum deduction is simultaneously the contribution limit. 2nd-pillar Einkauf is also deductible but is bounded by a personal Einkaufspotenzial we cannot know, so it is excluded." } }, "wellington": { "tax.filing_basis": { "source": "PwC WWTS - New Zealand (Tax administration): there is no provision for joint returns for spouses.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "IRD NZ PIT 2026: peer-identical with Auckland", "last_updated": "2026-05-17", "confidence": "high" }, "tax.social_security_rate": { "source": "ACC Earner's Levy 1.6%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "GST 15%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Wellington City Council rates effective ~0.35%", "last_updated": "2026-05-17", "confidence": "medium" }, "housing._summary": { "source": "Trade Me + realestate.co.nz Wellington Q1 2026", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Te Whatu Ora + private supp", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "20hr ECE free + private creche", "last_updated": "2026-05-17", "confidence": "high" }, "food._summary": { "source": "Numbeo Wellington + Stats NZ", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "Metlink monthly NZD 175", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "Stats NZ HES + Numbeo", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "WLG-SEA/LAX 1-stop via AKL/SYD 2025-26", "last_updated": "2026-05-17", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Working for Families tax credits", "last_updated": "2026-05-17", "confidence": "high" }, "non_cash._summary": { "source": "Holidays Act 2003 + 26wk paid parental", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Wellington. Its basis is converted with the rest of New Zealand and its rank inside New Zealand is preserved. Prior source: Levels.fyi WLG + Xero (founded here)/Trade Me/Datacom 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.52 x 0.5273 = 0.2742, R5-rounded down to 0.27. 0.5273 is the MEDIAN of (re-derived / previous) across the 1 New Zealand city that DO carry an all-levels measurement (Auckland 0.55 to 0.29), so Wellington moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Wellington's rank inside New Zealand is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Smaller than AKL. Index 0.52" }, "childcare.subsidy_per_kid_annual": { "source": "workandincome.govt.nz Childcare Subsidy + OSCAR income tables 2025", "last_updated": "2026-07-17", "note": "Childcare Subsidy and OSCAR are means-tested step functions cutting to $0 well below the published household (2-child cutout ~NZD 143k; income-tested FamilyBoost dies ~NZD 180k). The NZD 390k-404k published household receives 0 (was double-counted at 5,000)." }, "childcare.preschool_monthly": { "source": "NZ full-time ECE net of 20 Hours ECE (Auckland/Wellington market, 2025)", "last_updated": "2026-07-17", "note": "Full-time daycare net of the universal (NOT means-tested) 20 Hours ECE, mid-market Auckland/Wellington ~NZD 1,200/mo (was 600, which under-counted full-time care above the 20 free hours)." }, "non_cash.parental_leave_weeks": { "source": "Parental Leave and Employment Protection Act 1987 (primary carer leave, 26 wk); Employment New Zealand https://www.employment.govt.nz/news-and-updates/parental-leave-payments-to-increase-from-1-july", "last_updated": "2026-08-06", "confidence": "high", "note": "26 = primary carer leave, ONE pot per child, transferable to a partner but counted once. The partner's 1-2 wk is job-protected but statutorily UNPAID, so it is excluded per the basis. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md." }, "non_cash.parental_leave_paid_pct": { "source": "Employment New Zealand: the payment is the lower of your usual weekly earnings and the maximum rate https://www.employment.govt.nz/news-and-updates/parental-leave-payments-to-increase-from-1-july", "last_updated": "2026-08-06", "confidence": "high", "note": "0 percentage-paid weeks: the statute pays the LOWER of usual weekly earnings and the cap, and the cap always binds for our ICP, so every week is flat-rate and lives in non_cash.parental_leave_flat (the UK SMP shape). The previous 100 was the below-cap rate only; at NZD 150,000 the real rate is about 28%." }, "non_cash.parental_leave_flat": { "source": "Employment New Zealand, Parental leave payments to increase from 1 July: NZD 788.66 to 811.05/wk before tax from 1 July 2026, paid for up to 26 weeks https://www.employment.govt.nz/news-and-updates/parental-leave-payments-to-increase-from-1-july", "last_updated": "2026-08-06", "confidence": "high", "note": "NZD 811.05/wk, the PPL maximum from 1 July 2026. Against NZD 150,000 (2,884.62/wk) that is a 28% effective rate. 26 x 811.05 = NZD 21,087 total." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "OECD Health at a Glance 2025 New Zealand country note + Te Whatu Ora prescription co-payment and Very Low Cost Access GP fee structure, with free GP visits and prescriptions for under-14s", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at 500. A bottom-up build lands at NZD 300-350 and the OECD per-capita figure lands higher but is contaminated by dental, which New Zealand does not subsidise for adults at all and which is a separate field for us; under genuine latitude the conservative rule takes the higher. New Zealand has no all-in annual out-of-pocket ceiling. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "OECD Health at a Glance 2025 New Zealand country note + Te Whatu Ora prescription co-payment and Very Low Cost Access GP fee structure, with free GP visits and prescriptions for under-14s", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Same New Zealand gap as Auckland. The NZAVS regional commute-times brief is the best next lead." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRD, KiwiSaver deductions from employee pay https://www.ird.govt.nz/kiwisaver/kiwisaver-employers/contributions-and-deductions/kiwisaver-deductions-from-employee-pay ; ESCT https://www.ird.govt.nz/employing-staff/deductions-from-income/employer-superannuation-contriibution-tax-esct", "last_updated": "2026-08-11", "confidence": "high", "note": "No employee deduction exists. KiwiSaver employee contributions are CALCULATED ON gross pay but DEDUCTED FROM after-tax pay, so PAYE is assessed on the full gross. Do not read the IRD phrase \"before tax pay\" as pre-tax treatment: it describes the calculation base only. Employer contributions are taxed separately under ESCT and IRD explicitly excludes employee-directed deductions from that." } }, "rotterdam": { "non_cash.parental_leave_paid_pct": { "source": "WAZO art. 4:2 (geboorteverlof met behoud van loon, employer-paid) https://wetten.overheid.nl/BWBR0013008", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the 1-week partner geboorteverlof, employer-paid wage continuation with no dagloon ceiling (the structural twin of Germany Mutterschutz). The other 39 UWV-paid weeks moved to non_cash.parental_leave_capped in the 2026-08-06 caps rebuild; this field was 83, the whole-blend headline with the maximumdagloon ignored." }, "non_cash.parental_leave_weeks": { "source": "Wet arbeid en zorg (WAZO) art. 3:1, 4:2, 4:2a, 6:2, 6:3 https://wetten.overheid.nl/BWBR0013008 (geldig van 01-01-2025) ; Rijksoverheid https://www.rijksoverheid.nl/vraag-en-antwoord/ouderschapsverlof/recht-op-ouderschapsverlof ; business.gov.nl https://business.gov.nl/regulation/leave-schemes/", "last_updated": "2026-08-01", "confidence": "high", "note": "40 wk = zwangerschaps- en bevallingsverlof 16 (mother) + geboorteverlof 1 (partner) + aanvullend geboorteverlof 5 (partner) + betaald ouderschapsverlof 9 per parent x 2. WAZO gives ouderschapsverlof to each parent individually with no transfer mechanism (Beide ouders hebben recht op ouderschapsverlof). Was 9, which held the paid parental weeks for ONE parent and dropped 16 weeks of maternity at 100%. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Excluded: the 17 unpaid weeks per parent of the 26-week ouderschapsverlof, and the 20-week multiples maternity variant." }, "tax.inbound_regime": { "source": "30%-regeling (expatregeling). Belastingdienst Deskundigheidsvereiste, 2026 norm: a 'belastbaar jaarloon ... meer is dan EUR 48.013 (exclusief de gerichte vrijstelling)', i.e. measured on the salary REMAINING after the 30% carve-out, so about EUR 68,590 of gross: https://www.belastingdienst.nl/wps/wcm/connect/bldcontentnl/belastingdienst/zakelijk/internationaal/personeel/u_bent_niet_in_nederland_gevestigd_loonheffingen_inhouden/als_u_loonheffingen_gaat_inhouden/extraterritoriale_kosten_en_de_30procentregeling/voorwaarden_voor_de_30procentregeling1/deskundigheidsvereiste . Belastingdienst Inhoud van de expatregeling: maximum tax-free allowance EUR 78,600 in 2026, reached at the WNT norm of EUR 262,000. Max 60 months. Rijksoverheid, Belastingvoordeel buitenlandse werknemers.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31 OVER-GRANT FIX + refresh. (1) min_income_basis set to post_relief: the EUR 48,013 floor was being tested against GROSS, so every salary from EUR 48,013 up got the ruling; the statute measures it on the taxable remainder, which needs about EUR 68,590 of gross. Amsterdam single filer at EUR 60,000: tax 7,056 -> 15,906 (ordinary). (2) This entry previously cited the 2025 pair 46,660 / 246,000 while the data already carried the 2026 pair 48,013 / 262,000; the data was right and the prose was stale. (3) The under-30-with-a-master norm of EUR 36,497 is deliberately NOT used: we model neither age nor degree, so we carry the higher general norm. SCHEDULED: allowance drops 30% -> 27% and the norms step up from 2027-01-01 (Belastingplan 2025, second nota van wijziging). The calculator models the current tax year. TODO (2027-01 refresh): apply the 27% and the stepped-up norms once in force for the modeled year." }, "tax.filing_basis": { "source": "PwC WWTS - Netherlands (Tax administration): spouses are taxed separately for employment/business income (Box 1); only Box 2/3 and home items are allocable between fiscal partners.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Belastingdienst Box 1 2026: peer-identical with Amsterdam", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "BTW 21%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Rotterdam OZB ~0.085% effective", "last_updated": "2026-05-17", "confidence": "high" }, "housing._summary": { "source": "Funda + Pararius Rotterdam Q1 2026", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Zorgverzekering mandatory private (Zvw)", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "Kinderopvang per uur EUR ~9; 70% kinderopvangtoeslag for typical tech-family income", "last_updated": "2026-05-17", "confidence": "high" }, "food._summary": { "source": "Numbeo Rotterdam + CBS", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "RET monthly ~EUR 95", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "CBS + Numbeo Rotterdam", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "RTM/AMS-SEA/LAX 1-stop KLM 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Algemene Kinderbijslagwet art. 12, consolidated text valid from 01-08-2026, https://wetten.overheid.nl/BWBR0002368/ (fetched 2026-09-12, two independent reads agree; svb.nl returned 403 in English and Dutch and rijksoverheid.nl 404 on 2026-09-10 and 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "Dutch kinderbijslag: universal, NOT means-tested (the means-tested kindgebonden budget is separate and excluded). Statutory amounts per QUARTER per child, band set by the child's age on the first day of the quarter: EUR 298.40 under 6, 362.35 for 6 to 11, 426.29 for 12 to 17 (AKW art. 12 lid 3). The engine has no per-age benefit hook (computeGovtBenefits() reads one rate and only tests age < 18), so ONE rate is stored for every age and the Rule 5 tiebreak picks the LOWEST band: 298.40 x 4 = 1,193.60 a year, stored as 1,194. That understates a 6-to-11-year-old by EUR 255 a year and a 12-to-17-year-old by EUR 511, conservative against the Netherlands. Set 2026-09-12; was 1,368 = 342.06 x 4, which is the 6-to-11 band of the 2025 H1 table mislabelled as 2026. Ledger 2026-09-12; research docs/superpowers/research/2026-09-12-child-benefit-audit-fixes.md F4." }, "non_cash._summary": { "source": "Dutch labour law 24 days + 26wk parental partly paid via UWV", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Rotterdam: EUR 70,311, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 70,311 / 0.867 = USD 81,097; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3244, R5-rounded down to 0.32. REPLACES 0.55 (2026-08-11), and the interim 0.35 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Closes sweep B's stale-anchor finding: the note derived from 'Below Amsterdam (0.65)' while Amsterdam stored 0.72. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.55 was 0.7639 of Amsterdam's 0.72; that ratio x Amsterdam's re-derived 0.47 = 0.3590, R5-rounded down to 0.35. Amsterdam's own figure: EUR 102,818 / 0.867 = USD 118,591, / USD 250,000 = 0.4744. This is a BASIS move, not a re-measurement: Rotterdam's rank inside Netherlands is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Below Amsterdam (0.65). Index 0.55" }, "childcare.subsidy_per_kid_annual": { "source": "rijksoverheid.nl kinderopvangtoeslag 2025; Belastingdienst toeslag tables", "last_updated": "2026-07-17", "note": "kinderopvangtoeslag at the 33.3% high-income floor (2025, households >= EUR 226,546): 33.3% x capped rate EUR 10.71/hr x ~1,770 reimbursable h/yr ~= EUR 6,300 per full-time daycare child. The reimbursement % is means-tested, but 33.3% is the statutory MINIMUM, so it survives at the EUR 226k-269k published NL household (was overstated at a mid-income % of ~60-70%). 2026 raises the floor to 36.5%." }, "childcare.preschool_monthly": { "source": "NL dagopvang (0-4) market rate; aligned with Amsterdam/Utrecht/The Hague/Eindhoven peers", "last_updated": "2026-07-17", "note": "Rotterdam preschool (dagopvang for 1-3yos) was a 0 hole while its 4 NL peers charge 1,380-1,750/mo. NL kids start basisschool at 4, so ages 1-3 are paid daycare == the infant rate; set to 1,400 to match Rotterdam infant and the peer convention." }, "non_cash.parental_leave_capped": { "source": "UWV maximumdagloon: EUR 309,91 bruto per dag from 1 July 2026 (Wfsv art. 17 lid 1) https://www.uwv.nl/nl/premies-bedragen/maximum-dagloon ; UWV: aanvullend geboorteverlof and betaald ouderschapsverlof pay 70% van het (gemaximeerde) dagloon https://www.uwv.nl/nl/aanvullend-geboorteverlof/hoogte-uitkering-aanvullend-geboorteverlof-betaald-ouderschapsverlof ; WAZO arts. 3:13, 4:2b, 6:3", "last_updated": "2026-08-06", "confidence": "high", "note": "PRE-CONVERTED to a week cap per the calendar-day convention: the dagloon is a 5-WORKDAY wage cap applied BEFORE the percentage, so we store the weekly PAYOUT ceiling at the 70% tier, 309.91 x 0.70 x 5 = EUR 1,084.69/week, over all 39 UWV weeks at the 83 blend (16 wk maternity 100% + 23 wk at 70%). Conservative: the 16 maternity weeks actually cap higher (100% tier, EUR 1,549.55/wk), and pricing the whole block at the lower ceiling reads them LOW. Cap steps semi-annually with the minimum wage (1 Jul 2026 value stored). At EUR 90,000 the blended effective rate is ~64%. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Netherlands)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Zvw statutory eigen risico (EUR 385 per adult, under-18s exempt) + Zorginstituut Nederland and Vektis data on what share of insured actually exhaust it", "last_updated": "2026-08-08", "confidence": "medium", "note": "REPLACES an exact statutory maximum: EUR 385 is precisely the single-adult compulsory deductible, and only about half of insured adults exhaust it, so the realized figure is EUR 229-236 by two converging routes. The deductible is charged PER ADULT and under-18s are wholly exempt, and GP care, maternity care and district nursing sit outside it entirely, so a Dutch family of four is essentially two adults. Aanvullende verzekering is a PREMIUM and is not booked here. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Zvw statutory eigen risico (EUR 385 per adult, under-18s exempt) + Zorginstituut Nederland and Vektis data on what share of insured actually exhaust it", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "CBS StatLine, ODiN dataset 85055NED, motive Van en naar het werk, province Zuid-Holland. https://opendata.cbs.nl/statline/#/CBS/nl/dataset/85055NED", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Gemiddelde reisduur per verplaatsing", "year": "2024", "note": "32.05 min. Province geography. Rotterdam and The Hague necessarily share a value because no finer geography is published." }, "non_cash.sick_leave": { "source": "Burgerlijk Wetboek (BW) Boek 7, Artikel 629 (loondoorbetaling bij ziekte).; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.uwv.nl/particulieren/bedragen/detail/maximumdagloon", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Belastingdienst Centraal Aanspreekpunt Pensioenen, AOW-franchises https://centraalaanspreekpuntpensioenen.belastingdienst.nl/publicaties/overzicht-aow-inbouwbedragen-en-aow-franchises/ ; V&A 25-008 https://centraalaanspreekpuntpensioenen.belastingdienst.nl/publicaties/va-25-008-v251209/ ; Kennisgroepen KG:070:2024:3 jaarruimte https://kennisgroepen.belastingdienst.nl/publicaties/kg07020243-vragen-en-antwoorden-over-de-jaarruimteberekening-en-de-opbouw-van-pensioenaanspraken/ ; ABP 2026 premie split", "last_updated": "2026-08-11", "confidence": "low", "note": "THIS VALUE IS A MODEL, NOT A STATUTE, and the formula is what is HIGH confidence. Art. 3.127 Wet IB 2001 post-Wtp: jaarruimte = 30% of premiegrondslag MINUS the actual ouderdomspensioen premium already paid (employer + employee). 2026 AOW-franchise EUR 19,172, aftoppingsgrens EUR 137,800, so the statutory maximum is 0.30 x (137,800 - 19,172) = EUR 35,589 (encoded as 13,085 here, see below). The naive statutory mapping is INERT at our incomes and would look like a fix while changing nothing. 13,085 = 11.03% of premiegrondslag, i.e. 30% less ABP's 18.97% employer share; 0.082 is the lowest in-band gross ratio (at EUR 75k), so the mapping sits at or below the derived truth everywhere above EUR 75k. Expressing this via less_employer_pct after the 2026-08-11 engine reorder was considered and rejected: it cannot subtract the AOW-franchise and would hand every Dutch earner about EUR 2,115 of extra room." } }, "osaka": { "childcare.free_tier_start_age": { "source": "Kodomo Katei Cho (Children and Families Agency), free early childhood education and care policy: \"3 sai kara 5 sai made no subete no kodomotachi no riyoryo ga musho ni narimasu\" (usage fees become free for all children aged 3 to 5), in force 2019-10-01 https://www.cfa.go.jp/policies/kokoseido/mushouka/", "confidence": "high", "note": "Free tuition covers ages 3-5. Ages 0-2 stay fee-liable except for municipal-resident-tax-exempt households, with fees set by household tax band x municipality x hours, so the paid rate applies across the whole 0-2 class rather than to infants only. Before this, ages 1-5 billed the free-tier rate. Watch item: Tokyo has been funding wards to drop 0-2 fees (second and later children since FY2023, first children from 2025-09-01 in adopting wards incl. Meguro, Nerima, Kita), so the Tokyo infant rate may now be high for a growing share of the city; ward coverage is unconfirmed and the value is left alone.", "last_updated": "2026-08-01" }, "tax.filing_basis": { "source": "PwC WWTS - Japan (Tax administration): All income tax returns are filed on an individual basis in Japan; joint tax returns are not permitted.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "NTA 2026: peer-identical with Tokyo", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Consumption tax 10% (national)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Osaka koteishisanzei + toshikeikakuzei ~1.5% effective", "last_updated": "2026-05-17", "confidence": "medium" }, "housing._summary": { "source": "Suumo + Numbeo Osaka Q1 2026", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Kenkō hoken universal via payroll", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "Free preschool 3-5 (2019+); private hoikuen JPY 60k", "last_updated": "2026-05-17", "confidence": "high" }, "food._summary": { "source": "Numbeo Osaka 2026 + Statistics Bureau", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "Osaka Metro monthly equivalent ~JPY 10,000", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "Statistics Bureau + Numbeo Osaka", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "KIX-SEA/LAX nonstop ANA/JAL/UA 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "National jido teate JPY 12,000/mo = ~144,000/yr", "last_updated": "2026-05-17", "confidence": "high", "note": "Japanese jido teate: income cap abolished Oct 2024, now universal, NOT means-tested." }, "non_cash._summary": { "source": "Labour Standards Act + no statutory sick leave; kenkō hoken injury benefit 67%", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Osaka. Its basis is converted with the rest of Japan and its rank inside Japan is preserved. Prior source: Derived, not observed: no all-levels total-comp measurement exists for Osaka. Basis re-derived with its country anchor Tokyo; Osaka's stored ratio to Tokyo is preserved. Prior source: Levels.fyi Osaka + Panasonic/Rakuten/Sharp/NTT 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.32 x 0.5946 = 0.1903, R5-rounded down to 0.19. 0.5946 is the MEDIAN of (re-derived / previous) across the 1 Japan city that DO carry an all-levels measurement (Tokyo 0.37 to 0.22), so Osaka moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Osaka's rank inside Japan is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.32 was 0.8649 of Tokyo's 0.37; that ratio x Tokyo's re-derived 0.22 = 0.1903, R5-rounded down to 0.19. Tokyo's own figure: JPY 8,920,499 / 158.02 = USD 56,452, / USD 250,000 = 0.2258. This is a BASIS move, not a re-measurement: Osaka's rank inside Japan is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Below Tokyo. Rebased 2026-08-11 to 0.32 when Tokyo moved 0.48 to 0.37 in the comp-index sweep; the stored ratio to Tokyo (0.875) is preserved. The previous note cited Tokyo at 0.70 and Osaka at 0.55, neither of which the file has held for some time" }, "childcare.subsidy_per_kid_annual": { "source": "Kodomo Katei Cho free early childhood education policy and Osaka municipal 0-2 fee schedule; double-subtraction correction 2026-08-02", "last_updated": "2026-08-02", "confidence": "medium", "note": "Set to 0 on 2026-08-02, was JPY 360,000. The old value modelled Osaka's 0-2 fee reduction, but childcare.infant_monthly already carries the fee actually charged after that reduction. All four Japanese peers (tokyo, fukuoka, nagoya, yokohama) carried 0 against comparable infant rates, which is what exposed Osaka as the outlier. This field holds CASH THAT REACHES THE HOUSEHOLD. Where the monthly rate it nets against is already the price the household actually pays, the value is 0 and the relief is described here instead. See docs/superpowers/research/2026-08-02-childcare-double-subtraction.md." }, "healthcare.statutory_employee": { "source": "kyoukaikenpo.or.jp FY2026 per-prefecture insurance rate tables (Tokyo 9.85% total, employee half 4.925%; Kanagawa/Aichi/Osaka/Fukuoka within 9.85-10.13%); nenkin.go.jp standard remuneration grade table (health grade 50 = JPY 1,390,000/month) and the JPY 5,730,000 fiscal-year bonus ceiling (1 April to 31 March) shared by health, kaigo and the support levy; mhlw.go.jp FY2026 kodomo-kosodate shienkin 0.23% total / 0.115% employee, first charged on the April 2026 premium.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll. Tokyo's 9.85% is used as the national proxy because the block is enforced identical across a country's cities and the spread across our five metros is 0.07pp of employee rate. DELIBERATE EXCLUSIONS: kaigo 0.81% employee is age-gated to 40-64 and computeStatutoryHealthContribution cannot read age_bracket, so charging it would be wrong for the sub-40 half of our ICP (at the ceiling it would add about JPY 181,500/yr); the 0.115% kodomo-kosodate shienkin rides the same grades and bonus ceiling but funds child support, not health (about JPY 25,800/yr at the ceiling). The 2025 pension reform raising the pension ceiling to JPY 750,000 is not yet in force. TODO (phase-ins 2027-09, 2028-09, 2029-09): apply each step only when it takes effect." }, "non_cash.parental_leave_weeks": { "source": "Ikuji-Kaigo Kyugyo Ho art. 5 (childcare leave until the child turns 1, PER PARENT, non-transferable); Labor Standards Act art. 65 (maternity 6 wk pre + 8 wk post); Employment Insurance Act (ikuji kyugyo kyufu-kin); 2026 parameter references gyomu-keisan.jp (Reiwa 8), tfp-group.co.jp", "last_updated": "2026-08-06", "confidence": "high", "note": "110 wk = mother 14 (sanzen-sango kyugyo) + mother ikuji kyugyo from the end of the postnatal 8 wk to the first birthday (44 wk) + father ikuji kyugyo from birth to the first birthday (52 wk). The leave is PER PARENT and non-transferable, so both spans are summed, which is why parental_leave_both_parents is true (the Korea 120-wk precedent, an identical construction). The 4-wk shussei-ji ikuji kyugyo (post-birth papa leave) is taken WITHIN the father's span and its benefit days count against his 180-day 67% clock, so it adds nothing. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. Excluded: the 13% top-up to 80% for 28 days and the papa-mama 14-month extension (conditional on both parents' behaviour), the 18/24-month extension when no daycare place is found (conditional tail), and the JPY 500,000 shussan ikuji ichijikin (a lump sum, not leave pay). Was 52/67, the mother's span alone, as the old note itself admitted." }, "non_cash.parental_leave_paid_pct": { "source": "Health Insurance Act art. 102 (shussan teate-kin, 2/3 of standard remuneration); Employment Insurance Act art. 61-7 (ikuji kyugyo kyufu-kin, 67% first 180 days of each parent's leave, 50% after); ceilings per MHLW leaflet LL080731 ho 02 (2026-08-01) via https://roumu.com/pdf/2026080314.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "We store 0: since the 2026-08-07 Japan caps apply, all 110 weeks live in non_cash.parental_leave_capped (60% up to JPY 297,720/month), so no percentage-paid remainder weeks exist and a leftover headline rate would describe zero weeks. The old 60 blend (65.4 wk at 67 + 44.6 wk at 50, maternity at 66.7) is preserved as the capped block's pct. Research: docs/superpowers/research/2026-08-07-japan-parental-caps.md" }, "non_cash.parental_leave_capped": { "source": "MHLW leaflet LL080731 ho 02 (Reiwa 8 revision, in force 2026-08-01): ikuji kyugyo kyufu-kin shikyu jogen-gaku JPY 332,454/30-day unit at 67% and JPY 248,100 at 50%, wage-month base cap (kyugyo kaishi-ji chingin getsugaku jogen) JPY 496,200; fetched 2026-08-07 via https://roumu.com/pdf/2026080314.pdf (announcement https://roumu.com/archives/132696.html; MHLW hub https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000160564_00050.html); prior-year values confirmed against MHLW LL070722 https://www.mhlw.go.jp/content/001520023.pdf. Statutes: Employment Insurance Act art. 61-7 (ikuji kyugyo kyufu-kin), Health Insurance Act art. 102 (shussan teate-kin)", "last_updated": "2026-08-07", "confidence": "high", "note": "All 110 weeks flattened to ONE capped block, elbow-exact at the wage-month base cap: pct 60 (duration-weighted blend, rounded down) x JPY 496,200 = cap 297,720/month. Maternity (14 wk at 2/3, own std-remuneration ceiling JPY 1.39M/month) is FOLDED UNDER the shared ikuji ceiling per the Austria fold + Chile/Kazakhstan flatten precedents; true duration-weighted ceiling is ~298,064, so the stored 297,720 understates Japan (conservative). The two ikuji phases (67% first 180 days per parent, 50% after) share the same 496,200 elbow, so their flatten is lossless. No floor stored: the statutory kagen (96,090 wage-month) implies ~57,654/month, sub-ICP-relevant; omitting it is the less favorable direction. Ceilings revise every Aug 1 (indexed to average wages) - annual refresh row. Research: docs/superpowers/research/2026-08-07-japan-parental-caps.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "MHLW coinsurance schedule (30% for ages 6-69, 20% under 6) + the income-banded kougaku ryouyouhi monthly high-cost ceiling + Tokyo ward-level child medical subsidy", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 held at 50,000, structurally plausible against the high-cost ceiling, which is mid-revision (rising August 2026) and so usable only as an approximate check. Tokyo's ward child medical subsidy makes children close to free in practice, but ward income limits leave real uncertainty for a high earner, so the child increment is small rather than zero. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "MHLW coinsurance schedule (30% for ages 6-69, 20% under 6) + the income-banded kougaku ryouyouhi monthly high-cost ceiling + Tokyo ward-level child medical subsidy", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Statistics Bureau of Japan, Shakai Seikatsu Kihon Chosa 2021, rank12, Osaka-fu. https://www.stat.go.jp/data/shakai/2021/rank/zuhyou/rank/rank12.xlsx", "last_updated": "2026-08-08", "confidence": "high", "statistic": "1-nichi atari no tsukin tsugaku jikan", "year": "2021", "note": "CONVERSION: 1h27m = 87 min ROUND TRIP / 2 = 43.5 one way. Same caveats. Survey self-report." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "NTA tax answer 1135 https://www.nta.go.jp/taxes/shiraberu/taxanswer/shotoku/1135.htm ; MHLW iDeCo contribution limits https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/nenkin/nenkin/kyoshutsu/taishousha.html", "last_updated": "2026-08-11", "confidence": "medium", "note": "iDeCo, deductible in full under the small-enterprise-mutual-aid category; the binding ceiling lives in the pension law and is stated MONTHLY, converted x12 here. Category-2 insured WITH a corporate DC/DB plan: JPY 20,000/mo = 240,000/yr. Without one it is 23,000/mo = 276,000/yr. The lower figure is the house-rule choice, because a senior engineer at a large employer usually has a corporate pension; that is a BASIS choice, not a magnitude, and it is recorded in docs/corrections-ledger.md. Separate from the already-modeled kosei nenkin. Life- and annuity-insurance premium relief is deliberately not stacked. REFRESH ITEM: from 1 Dec 2026 the limit becomes 62,000/mo less employer DC/DB and the 20,000 sub-cap is abolished." } }, "mexico-city": { "tax.federal_brackets": { "source": "ISR tarifa anual 2026 inflation-adjusted limits (taxsummaries.pwc.com/mexico); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.social_security_rate": { "source": "IMSS employee ~3% blended", "last_updated": "2026-05-17", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "IVA general 16%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "CDMX predial ~0.2% effective", "last_updated": "2026-05-17", "confidence": "medium" }, "housing._summary": { "source": "Inmuebles24 + Numbeo Mexico City Q1 2026", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Mercado Libre Inmuebles listing-count CDF, 2026-08-17 (departamentos en renta, distrito-federal, fixed recamaras, _PriceRange URL filter, asking MXN/month per unit, citywide): 2-rec P25/med/P75 = 20,686/32,538/48,182 (N=1,900) -> 0.636/1.481, applied 0.64/1.48; 1-rec 0.712/1.281 (N=1,151). Method and figures reproduced independently across 2026-08-16 and 2026-08-17 runs.", "last_updated": "2026-08-17", "confidence": "medium", "note": "REBASED 2026-08-17 (ledger cluster housing-0817) from the calibrated default 0.82/1.24, which is TOO NARROW for this dual formal/informal market - in the flattering direction: solveEquivalenceBand was understating the uncertainty shown. Same basis as the band's claim (citywide listing pool at fixed bedroom count), so no sitting-stock or all-unit excuse applies. Corroborated on both sides of the basis: Inmuebles24 standardized 2-rec colonia means span 0.57x-1.75x of the city index, and INEGI ENIGH 2024 sitting stock reads 0.58/1.42 at 2-dorm (both in docs/superpowers/research/2026-08-16-housing-band-scope-default-sample-b.md). Caveats, opposite directions per edge and both recorded: the ML pool skews premium with USD-priced luxury fattening the TOP tail (read 1.48 with that discount), while the informal cheap stock that never reaches portals means the true citywide LOW edge is wider than measured. 2-rec anchor per the flagship 2BR convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Raw CDFs: docs/superpowers/research/2026-08-17-housing-band-fixes.md (X1). NOTE the stored median (2BR 28,000) sits between the two portals' universes (Inmuebles24 21,751 / ML 32,538) - a LEVEL question, flagged 2026-08-16, out of scope here, adjudicate separately if ever touched." }, "healthcare._summary": { "source": "IMSS universal (slow) + private health insurance standard for tech families", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "Mexico City private daycare/preschool MXN 6-10k", "last_updated": "2026-05-17", "confidence": "medium" }, "food._summary": { "source": "Numbeo CDMX 2026 + INEGI", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "Metro/Metrobús monthly equivalent MXN ~600", "last_updated": "2026-05-17", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo CDMX entertainment", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "MEX-SEA/LAX nonstop Aeroméxico/Delta 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "MX has no federal cash child allowance for tech-income families", "last_updated": "2026-05-17", "confidence": "high" }, "non_cash._summary": { "source": "LFT 2023 reform: 12 days vacation yr 1; 7 PH; IMSS sick 60% from day 4", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Mexico City: MXN 997,005, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "MXN 997,005 / 17.31 = USD 57,597; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2304, R5-rounded down to 0.23. REPLACES 0.3 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Aguinaldo already inside. The MXN figure comes off the country page's city table because the city page rendered USD 51,940; the two reconcile at 19.2 MXN/USD. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): LatAm tech hub. Index 0.30" }, "tax.filing_basis": { "source": "SAT - ISR assessed per individual; no joint returns", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "non_cash.parental_leave_weeks": { "source": "LSS art. 101 (IMSS maternity subsidy, 100% of last daily contribution salary, 42+42 days); LFT art. 170; LFT art. 132 XXVII bis (5 working days paternity, employer-paid); UMA 2026 = MXN 117.31/day (INEGI comunicado 1/26; DOF 2026-01-09)", "last_updated": "2026-08-06", "confidence": "high", "note": "13 = 12 weeks IMSS maternity (84 days at 100%) + 5 working days of employer-paid paternity (1 week; the 20-day reform was never published in the DOF, so 5 days remains the law in 2026), counted per the Brazil precedent for short paternity. Was 12: maternity only. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "LSS art. 101 (IMSS maternity subsidy, 100% of last daily contribution salary, 42+42 days); LFT art. 170; LFT art. 132 XXVII bis (5 working days paternity, employer-paid); UMA 2026 = MXN 117.31/day (INEGI comunicado 1/26; DOF 2026-01-09)", "last_updated": "2026-08-06", "confidence": "high", "note": "100 is the statutory headline for both blocks. Cap arithmetic for the standing caps-convention item: the IMSS maternity subsidy base is capped at 25 UMA/day (UMA 2026 = MXN 117.31) = MXN 2,932.75/day, about MXN 1.07M/yr, so it binds only at the top of the CDMX tech band (at MXN 1.2M the effective maternity rate is about 89%); paternity is employer-paid at full salary, uncapped." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Mexican private consultation pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "CHECKED for the premium-as-out-of-pocket error and CLEAR: the value sits below any real private premium and above the IMSS contribution, which is consistent with cash-pay private consultations and not with a mis-booked premium. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Mexican private consultation pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "INEGI, Censo de Poblacion y Vivienda 2020, movilidad cotidiana tabulados, variable TIE_TRASLADO_TRAB, Zona Metropolitana del Valle de Mexico (63 municipios). https://www.inegi.org.mx/contenidos/programas/ccpv/2020/tabulados/ampliado/cpv2020_a_cdmx_10_movilidad_cotidiana.xlsx", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Tiempo de traslado al trabajo", "year": "2020", "note": "One way; band-midpoint mean over the metropolitan zone." }, "non_cash.sick_leave": { "source": "Ley del Seguro Social (LSS) Art. 96 (payment start / 52-week + 26-week extension),; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": "https://leyes-mx.com/ley_federal_del_trabajo/42.htm", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "SAT, Aportaciones complementarias de retiro https://www.sat.gob.mx/minisitio/DeduccionesPersonales/aportaciones_complementarias.html (LISR art. 151 fr. V) ; INEGI UMA 2026", "last_updated": "2026-08-11", "confidence": "high", "note": "Aportaciones complementarias de retiro, deductible at 10% of cumulative income up to 5 UMA anual. DENOMINATION TRAP HANDLED: the UMA \"anual\" is NOT diario x 365. INEGI computes mensual = diario x 30.4 and anual = mensual x 12 (= x 364.8), so UMA 2026 diario 117.31 gives anual 42,794.64 and 5 UMA = MXN 213,973.20 (2025 was 206,418.60). LISR art. 151 fr. V sits OUTSIDE the general 15% / 5-UMA personal-deduction cap and stands alone. Binds exactly at a 10% contribution." } }, "auckland": { "tax.filing_basis": { "source": "PwC WWTS - New Zealand (Tax administration): there is no provision for joint returns for spouses.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "IRD NZ PIT 2026 (5 brackets 10.5-39%)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.social_security_rate": { "source": "ACC Earner's Levy 1.6%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.social_security_cap": { "source": "ACC earnings ceiling NZD 152,790/yr 2026", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "GST 15%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Auckland Council rates ~0.3% effective", "last_updated": "2026-05-17", "confidence": "medium" }, "housing._summary": { "source": "Trade Me + realestate.co.nz Auckland Q1 2026 + Numbeo", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Te Whatu Ora universal (slow); 30% have private (Southern Cross)", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "20 hrs/wk free ECE ages 3-5; nursery NZD 1.5-2k", "last_updated": "2026-05-17", "confidence": "high" }, "food._summary": { "source": "Numbeo Auckland 2026", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "AT HOP monthly pass ~NZD 235", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Auckland entertainment", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "AKL-SEA/LAX nonstop Air NZ 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Working for Families tax credits ~NZD 4.5k/yr typical", "last_updated": "2026-05-17", "confidence": "high" }, "non_cash._summary": { "source": "Holidays Act 2003: 20 days annual + 10 sick; 26 weeks paid parental at govt-capped rate", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Auckland: NZD 123,437, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "NZD 123,437 / 1.7 = USD 72,610; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2904, R5-rounded down to 0.29. REPLACES 0.55 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Genuine local sample. Excludes the 3% employer KiwiSaver contribution. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Index 0.55" }, "childcare.subsidy_per_kid_annual": { "source": "workandincome.govt.nz Childcare Subsidy + OSCAR income tables 2025", "last_updated": "2026-07-17", "note": "Childcare Subsidy and OSCAR are means-tested step functions cutting to $0 well below the published household (2-child cutout ~NZD 143k; income-tested FamilyBoost dies ~NZD 180k). The NZD 390k-404k published household receives 0 (was double-counted at 5,000)." }, "childcare.preschool_monthly": { "source": "NZ full-time ECE net of 20 Hours ECE (Auckland/Wellington market, 2025)", "last_updated": "2026-07-17", "note": "Full-time daycare net of the universal (NOT means-tested) 20 Hours ECE, mid-market Auckland/Wellington ~NZD 1,200/mo (was 600, which under-counted full-time care above the 20 free hours)." }, "non_cash.parental_leave_weeks": { "source": "Parental Leave and Employment Protection Act 1987 (primary carer leave, 26 wk); Employment New Zealand https://www.employment.govt.nz/news-and-updates/parental-leave-payments-to-increase-from-1-july", "last_updated": "2026-08-06", "confidence": "high", "note": "26 = primary carer leave, ONE pot per child, transferable to a partner but counted once. The partner's 1-2 wk is job-protected but statutorily UNPAID, so it is excluded per the basis. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md." }, "non_cash.parental_leave_paid_pct": { "source": "Employment New Zealand: the payment is the lower of your usual weekly earnings and the maximum rate https://www.employment.govt.nz/news-and-updates/parental-leave-payments-to-increase-from-1-july", "last_updated": "2026-08-06", "confidence": "high", "note": "0 percentage-paid weeks: the statute pays the LOWER of usual weekly earnings and the cap, and the cap always binds for our ICP, so every week is flat-rate and lives in non_cash.parental_leave_flat (the UK SMP shape). The previous 100 was the below-cap rate only; at NZD 150,000 the real rate is about 28%." }, "non_cash.parental_leave_flat": { "source": "Employment New Zealand, Parental leave payments to increase from 1 July: NZD 788.66 to 811.05/wk before tax from 1 July 2026, paid for up to 26 weeks https://www.employment.govt.nz/news-and-updates/parental-leave-payments-to-increase-from-1-july", "last_updated": "2026-08-06", "confidence": "high", "note": "NZD 811.05/wk, the PPL maximum from 1 July 2026. Against NZD 150,000 (2,884.62/wk) that is a 28% effective rate. 26 x 811.05 = NZD 21,087 total." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "OECD Health at a Glance 2025 New Zealand country note + Te Whatu Ora prescription co-payment and Very Low Cost Access GP fee structure, with free GP visits and prescriptions for under-14s", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at 500. A bottom-up build lands at NZD 300-350 and the OECD per-capita figure lands higher but is contaminated by dental, which New Zealand does not subsidise for adults at all and which is a separate field for us; under genuine latitude the conservative rule takes the higher. New Zealand has no all-in annual out-of-pocket ceiling. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "OECD Health at a Glance 2025 New Zealand country note + Te Whatu Ora prescription co-payment and Very Low Cost Access GP fee structure, with free GP visits and prescriptions for under-14s", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. New Zealand's census carries no travel-time question and the Ministry of Transport Household Travel Survey tables sit behind a JavaScript iframe that could not be read. Recorded so nobody re-runs the lookup." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRD, KiwiSaver deductions from employee pay https://www.ird.govt.nz/kiwisaver/kiwisaver-employers/contributions-and-deductions/kiwisaver-deductions-from-employee-pay ; ESCT https://www.ird.govt.nz/employing-staff/deductions-from-income/employer-superannuation-contriibution-tax-esct", "last_updated": "2026-08-11", "confidence": "high", "note": "No employee deduction exists. KiwiSaver employee contributions are CALCULATED ON gross pay but DEDUCTED FROM after-tax pay, so PAYE is assessed on the full gross. Do not read the IRD phrase \"before tax pay\" as pre-tax treatment: it describes the calculation base only. Employer contributions are taxed separately under ESCT and IRD explicitly excludes employee-directed deductions from that." } }, "athens": { "childcare.free_tier_start_age": { "source": "Eurydice (European Commission), Greece, Early childhood education and care: \"Compulsory attendance starts at age 4\"; the two-year compulsory pre-school period was rolled out nationally from the 2020-21 school year https://eurydice.eacea.ec.europa.eu/national-education-systems/greece/early-childhood-education-and-care", "confidence": "high", "note": "Greece is the reason the boundary is per-city rather than global: its free tier starts at 4, not 3. Ages 0-4 (vrefonipiakoi stathmoi) are paid, with an income-tested ESPA voucher that does not reach a high-income household. Before this, ages 1-5 billed the free-tier rate, so a 3-year-old was charged nothing. CAVEAT: the boundary is well sourced but the Athens paid rate of EUR 400/month it now applies to is NOT - a prior sweep could neither confirm nor refute it - and this change makes that rate matter for four years instead of one. Athens childcare rates need their own sourcing pass.", "last_updated": "2026-08-01" }, "tax.inbound_regime": { "source": "Art. 5C L.4172/2013, inserted by L.4646/2019: 50% exemption of Greek employment income, 7 tax years, no floor and no ceiling ('apallassetai ... gia to peninta tois ekato (50%) tou eisodimatos tou apo misthoti ergasia'; 'ligei meta to peras epta (7) synolika forologikon eton'). Prior non-residency 5 of the last 6 years; transfer from an EU/EEA state or one with a tax administrative-cooperation agreement; Greek employer or Greek PE; 2-year stay declaration; NEW job position per AADE decision A.1087/2021. Deadline: 31 December of the year work starts if it starts by 2 July, otherwise 31 December of the following year with relief starting only then.", "last_updated": "2026-07-31", "confidence": "medium", "note": "2026-07-31: all three carried values verified correct (0.50 / 7 years / no bands). Two errors in the old regime note fixed, both in the reader's favour when wrong: it said the prior-residency bar was '7 of last 8 yrs' (it is 5 of the last 6) and it omitted the origin-country gate and the new-position requirement. Engine check passed: computeGreeceTaxes computes EFKA on gross before resolveInboundRegime, so the 50% correctly does not reduce social contributions. Confidence is medium ONLY because of source access, not substance: aade.gr returned HTTP 403 on every page and elib.aade.gr refused the connection this session, so the statutory text above is quoted from taxheaven.gr's reproduction of the codified article and of A.1087/2021 rather than fetched from AADE. Conservative call taken on the disputed 'new job position' requirement: treated as LIVE, because a secondary claim that L. 5222/2025 abolished it does not survive checking (art. 206 of that law touched paras 3, 6 and 8 only). RE-CHECK: AADE decision A.1096/2026 of 2026-04-29 amends A.1087/2021 and could not be retrieved; procedural decisions can move deadlines." }, "tax.federal_brackets": { "source": "Greece 2026 personal income tax reform (Law, presented ITF Thessaloniki 2025): brackets 9/20/26/34/39/44, most bands cut 2pp with a new 40,000-60,000 band at 39%. GreekCityTimes / Eurofast / PwC Greece. https://taxsummaries.pwc.com/greece/individual/taxes-on-personal-income", "last_updated": "2026-07-26", "note": "GR-1 (closeout 2026-07-26): 2026 six-band scale (was 9/22/28/36/44)." }, "tax.social_security_rate": { "source": "EFKA employee 13.87%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.social_security_cap": { "source": "EFKA wage ceiling EUR 7,572.62/mo × 12 = 90,872/yr", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "FPA standard 24%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "ENFIA ~0.1% effective post-2022 reforms", "last_updated": "2026-05-17", "confidence": "medium" }, "housing._summary": { "source": "Spitogatos + Numbeo Athens Q1 2026", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "ESY universal via EFKA; private complementary common", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "Nipiagogeio free 4+; vrefonipiakos paid", "last_updated": "2026-05-17", "confidence": "medium" }, "food._summary": { "source": "Numbeo Athens 2026 + ELSTAT", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "OASA monthly EUR 30", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Athens entertainment", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "ATH-SEA/LAX 1-stop Aegean/AA 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "OPEKA family allowance", "last_updated": "2026-05-17", "confidence": "medium" }, "non_cash._summary": { "source": "Greek labor code: 20 days/yr + 14 PH + 17 weeks maternity 100%", "last_updated": "2026-05-17", "confidence": "high", "note": "Inbound regime (50% PIT for 7 yrs, Law 4646/2019) modeled via inbound_regime block" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Athens: EUR 31,830, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-europe.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 31,830 / 0.867 = USD 36,713; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1469, R5-rounded down to 0.14. REPLACES 0.32 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Senior/all-levels 1.56, the widest in its batch: a bimodal local-versus-multinational market. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Growing remote-work hub. Index 0.32" }, "tax.filing_basis": { "source": "AADE (aade.gr) - spouses file separate income-tax returns (individual taxation)", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "tax.federal_brackets_married": { "source": "Greece 2026 PIT reform (as tax.federal_brackets). Greece files individually, so the married array mirrors single. https://taxsummaries.pwc.com/greece/individual/taxes-on-personal-income", "last_updated": "2026-07-26", "note": "GR-1 (closeout 2026-07-26): mirrors the 2026 single scale." }, "childcare.subsidy_per_kid_annual": { "source": "EETAA voucher rules (income ceiling ~EUR 35k)", "last_updated": "2026-08-05", "confidence": "high", "note": "ZEROED 2026-08-05 (CC5, was 1200): ESPA/EETAA voucher is income-capped far below the ICP; no cash reaches this household. Working: docs/superpowers/research/2026-08-05-cc5-four-cities-subsidy.md" }, "non_cash.parental_leave_weeks": { "source": "Multi-stream household stack (batch-8 restate 2026-08-08): maternity 119 calendar days = 17 wk (15 capped EFKA+DYPA + 2 employer AK 657-658 leg, carried from the 2026-08-06 rebuild) + paternity 14 WORKING days /5 = 2.8 wk (art. 27 L.4808/2021 black-letter 'dekatesseris (14) ergasimes imeres, me apodoches', e-nomothesia.gr full text; employer-paid, zero tenure) + 9-month DYPA special maternity protection = 39.13 wk (art. 142 L.3655/2008 as am. art. 43 L.4997/2022) + art. 28 L.4808/2021 parental allowance, first 2 months DYPA-paid PER PARENT x2 = 17.39 wk. Total 76.32 -> 76.3. Research: docs/superpowers/research/2026-08-08-pl-b8-europe.md", "last_updated": "2026-08-08", "confidence": "high", "note": "Counting rules: the special benefit is mother-owned with up to 7 of 9 months transferable to the father - ONE pot counted once; each parent's art. 28 paid slice is non-transferable so both count; months 3-4 of each parent's art. 28 leave are UNPAID, excluded. The streams stack sequentially (no anti-cumulation found + explicit sequencing language; med-high). Gates: EFKA maternity 200 insured days in the last 2 years = cumulative-career, ANY employer - EU-aggregation applies, LOUD caveat, no haircut; art. 28 allowance needs 1 year with the SAME employer but the leave is usable to child age 8, so the work-to-term mover vests months after birth (deferred-pot precedent, counted)." }, "non_cash.parental_leave_paid_pct": { "source": "Greek Civil Code arts. 657-658 (employer wage claim during justified absence: one month at >= 1 year tenure, half a month under it, net of insurance benefits received), applied to maternity per e-forologia.gr labour-law analysis and mitos.gov.gr; paternity full pay per art. 27 L.4808/2021", "last_updated": "2026-08-08", "confidence": "high", "note": "100 covers the 4.8 pct weeks: the 2-week AK 657-658 employer maternity leg (for the priced mover, under 1 year tenure, half a month of full salary net of EFKA benefits, rounded down from 2.17) + the 2.8-week working-days paternity, both employer-paid at full wage, uncapped. The 15 capped weeks live in parental_leave_capped, the 56.5 flat DYPA weeks in parental_leave_flat. Tenure gate documented per the price-the-mover rule: at 1+ year tenure the employer maternity leg runs ~4.3 weeks, a small upside not stored." }, "non_cash.parental_leave_capped": { "source": "e-EFKA Circular 25/2025: 2026 insurance-class table, 28th (highest) class tekmarto imerominio EUR 110.49/day, EFKA maternity leg 50% = EUR 386.72/week ; mitos.gov.gr (National Registry of Administrative Procedures), Sympliromatikes paroxes mitrotitas: the DYPA supplement is computed after deducting the AK 657-658 employer amounts and its reference salary caps at the highest EFKA insurance class (PD 776/1977 as amended by PD 221/1997) ; research: docs/superpowers/research/2026-08-06-greece-employer-leg.md", "last_updated": "2026-08-08", "confidence": "medium-high", "note": "15 of the 17 maternity weeks: the EFKA (50% of the capped tekmarto imerominio) plus DYPA supplementary legs, whose COMBINED payout ceiling is the 28th-insurance-class salary, EUR 2,762.25/month = EUR 637.44/week on the monthly convention, the lower of the two published conversions (the daily convention reads EUR 773.43/wk; conservative pick). The 2026-08-06 dedicated pass refuted both the open-ended-employer-top-up reading and the fixed 50/40/10 split: beyond the AK 657-658 window the employer owes nothing by statute. Batch-8 re-verify 2026-08-08: EUR 2,762.25 IS the 2026 in-force value - the +2.40% class revaluation (EFKA Circular 1/2026) is already inside the 110.49 daily figure (the 'cap moved' hypothesis was checked and refuted). The 9-month DYPA follow-on benefit is now STORED in parental_leave_flat (was the flagged follow-up). Cap follows the annual EFKA revaluation - recheck each January." }, "non_cash.parental_leave_flat": { "source": "DYPA flat streams at the statutory minimum wage EUR 920/month (MD 8934/2026, in force 2026-04-01) x 14/12 dora proportion (statutory wording, mitos.gov.gr + prosvasis.dypa.gov.gr): 9-month special maternity protection (art. 142 L.3655/2008 as am. art. 43 L.4997/2022) = 39.13 wk + art. 28 L.4808/2021 parental allowance 2 paid months per parent x2 = 17.39 wk; combined 56.52 -> 56.5 wk at 9,660+4,293.33 EUR / 56.52 wk = EUR 246.9/week. Research: docs/superpowers/research/2026-08-08-pl-b8-europe.md", "last_updated": "2026-08-08", "confidence": "high", "note": "Both streams pay the same min-wage+dora rate, so one flat block is exact. MD 19794/14-07-2026 changed special-benefit eligibility details only, not amount or duration. Min wage steps ~annually (next expected 2027, ~950 target) - the weekly_amount follows it; calendar row added. The father-transfer option on the special benefit does not change household totals (one pot)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "OECD/WHO Greece country health profile 2025, which states explicitly that there is no annual ceiling on total copayments + EOPYY pharmaceutical co-payment structure", "last_updated": "2026-08-08", "confidence": "medium", "note": "Greek household out-of-pocket is among the highest in the EU, driven by pharmaceutical co-payments and the private-pay habit that public waiting lists force on a professional household. Confirmed NEGATIVE finding: Greece has neither an annual ceiling nor even an episode-level cap. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "OECD/WHO Greece country health profile 2025, which states explicitly that there is no annual ceiling on total copayments + EOPYY pharmaceutical co-payment structure", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. ELSTAT publishes no commute-time series and TT1019V carries no Greek city for any year 1990-2024. Eurostat's LFS national mean for Greece is 20, and capital-to-national ratios elsewhere imply roughly 29 to 30, so 35 is probably high, but no city figure exists to replace it and city spread was not invented." }, "non_cash.sick_leave": { "source": "Law 2112/1920 art. 5 (salaried employees, \"υπάλληλοι\") as commonly summarized by; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": "https://greecehandbook.com/who-pays-for-sick-leave-in-greece-employer-or-e-efka/", "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "N. 4172/2013 art. 14 par. 1 (e) and (i) https://www.taxheaven.gr/law/4172/2013/arthro/14 ; N. 5078/2023 art. 105 as amended by art. 35 N. 5178/2025 https://www.taxheaven.gr/news/67664/o-nomos-50782023-gia-ta-tameia-epaggelmatikhs-asfalishs-melh-xronos-asfalishs-kai-eisfores ; AADE circular E.2060/2024", "last_updated": "2026-08-11", "confidence": "medium", "note": "REFUTES the phase-1 brief, which predicted zero. Voluntary contributions to occupational insurance funds (TEA) and premiums under group pension insurance contracts are EXCLUDED from employment income, up to a combined employer-plus-employee ceiling of 20% of gross earnings, annually and cumulatively across all pension products. An exclusion from taxable employment income, i.e. a deduction. CAVEAT, hence medium: the vehicle must be OCCUPATIONAL; a purely individual personal pension or life policy bought by the employee gets nothing. The EUR 1,500/yr in art. 14 par. 1 (ia) is medical/hospital and life-risk cover, not retirement, and is kept out. Inert at 6% and 10%, but encoding zero would have invented about USD 9,900 of Greek tax." } }, "budapest": { "childcare.free_tier_start_age": { "source": "Eurydice (European Commission), Hungary, Early childhood education and care: ovoda attendance is obligatory for children aged 3 for at least 4 hours per weekday and is provided free of charge https://eurydice.eacea.ec.europa.eu/national-education-systems/hungary/early-childhood-education-and-care", "confidence": "high", "note": "Ovoda (kindergarten) is compulsory and free from age 3. Bolcsode (0-3) is non-compulsory and fee-liable; Budapest private bolcsode runs roughly HUF 70,000-130,000/month, municipal roughly HUF 15,000-35,000 plus food. Before this, ages 1-5 billed the free-tier rate, so a 1-year-old in bolcsode was charged nothing. The 3-5 residual stays 0: a meal fee (etkezesi teritesi dij) is standard in Hungarian public kindergartens but no Budapest figure could be sourced.", "last_updated": "2026-08-01" }, "tax.federal_brackets": { "source": "NAV SZJA 2026 flat 15%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.social_security_rate": { "source": "TBJ employee 18.5% (pension 10% + health 7% + UB 1.5%)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "VAT 27% (highest in EU)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Municipal building tax usually 0 for primary residence", "last_updated": "2026-05-17", "confidence": "high" }, "housing._summary": { "source": "ingatlan.com + Numbeo Budapest Q1 2026", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "NEAK universal via 7% health payroll; premium = 0", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "Bölcsöde paid; óvoda 3-6 free", "last_updated": "2026-05-17", "confidence": "high" }, "food._summary": { "source": "Numbeo Budapest 2026 + KSH", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "BKK monthly pass HUF 9,500", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Budapest entertainment", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "BUD-SEA/LAX 1-stop 2025-26", "last_updated": "2026-05-17", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Családi pótlék + family tax benefit blended", "last_updated": "2026-05-17", "confidence": "medium" }, "non_cash._summary": { "source": "Munka Törvénykönyve: 20 days + 11 PH + 24 weeks parental at 70%", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Budapest: HUF 14,747,860, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-europe.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "HUF 14,747,860 / 315.31 = USD 46,773; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1871, R5-rounded down to 0.18. REPLACES 0.3 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Strong tech hub. Index 0.30" }, "healthcare.statutory_employee": { "source": "2019. evi CXXII. torveny (Tbj) section 25(1), net.jogtar.hu: 'A tarsadalombiztositasi jarulek merteke 18,5 szazalek'; section 26(1) sets the fund allocation from which the 37.9% health share (0.379 x 18.5% = 7.0115%) is derived; nav.gov.hu 2026 contribution rates page. The jarulekfizetesi felso hatar applied only to the pension slice and was repealed effective 2013-01-01; the current Tbj contains no ceiling provision.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll. The health slice is a revenue-allocation share of one unified 18.5% levy, not a separately levied rate, so it cannot be read off a payslip. UNCAPPED rests on the documented 2013 repeal plus the absence of any ceiling provision in the current Tbj (negative evidence, no quotable 'there is no cap' sentence). Related, outside this block: from 2026-01-01 mothers under 40 with two children are fully PIT-exempt on wage income (three children since 2025-10-01), phasing to older cohorts through 2029; contributions still apply. TODO: check the Budapest family scenario against this exemption." }, "childcare.subsidy_per_kid_annual": { "source": "Hungarian childcare fee rules", "last_updated": "2026-08-05", "confidence": "high", "note": "ZEROED 2026-08-05 (CC5, was 240000): No Hungarian cash childcare subsidy reaches an ICP household; Budapest excluded from the only fee subsidy. Working: docs/superpowers/research/2026-08-05-cc5-four-cities-subsidy.md" }, "non_cash.parental_leave_weeks": { "source": "Three-layer household stack (batch-8 rework 2026-08-08, resolving the 2026-08-06 scheme-mismatch flag): CSED 168 calendar days = 24 wk (Ebtv. 40-42. par, mother) + GYED from CSED end to the child's 2nd birthday = 80 wk (either-parent pot, counted once) + paternity 10 WORKING days /5 = 2 wk (Mt. 118. par, apasagi szabadsag). Total 106. Research: docs/superpowers/research/2026-08-08-pl-b8-europe.md", "last_updated": "2026-08-08", "confidence": "high", "note": "GYES (age 2-3, HUF 28,500/mo) EXCLUDED as token: 0.72% replacement at the rankings basis, ~8.8% of the minimum wage (Azerbaijan under-3-allowance precedent). The 365-insured-days-in-2-years gate on CSED/GYED is cumulative-career, ANY employer - EU movers aggregate periods under EC 883/2004; LOUD caveat, no haircut. Twins variant (GYED to age 3) not modeled." }, "non_cash.parental_leave_paid_pct": { "source": "1997. evi LXXXIII. torveny (Ebtv.) 40-42. par: CSED 100% of the calendar-daily base since 2021-07-01 with NO ceiling ('nincs felso hatara', unanimous 2026 administration incl. Magyar Allamkincstar); Mt. 118. par paternity 10 working days, days 1-5 at 100% absence pay, days 6-10 at 40%", "last_updated": "2026-08-08", "confidence": "high", "note": "97.7 = the duration-weighted blend over the 26 pct-weeks: (24 CSED wk x 100 + 2 paternity wk x 70)/26 = 97.69 rounded down, where the paternity 70 is itself the (5x100 + 5x40)/10 statutory split. The 2026-08-06 scheme mismatch (the GYED cap applied to these CSED weeks) is RESOLVED: CSED is verified uncapped, the cap belongs to the 80 GYED weeks in parental_leave_capped. Tax note, no metric impact (we store gross): CSED fully PIT/contribution-free since 2025-07-01; GYED PIT-free with a 10% pension contribution - REFUTES the older 'CSED = 15% PIT' note." }, "non_cash.parental_leave_capped": { "source": "1997. evi LXXXIII. torveny 42/D para: GYED cap = 70% of 2x the current minimum wage; 2026 minimum wage HUF 322,800/month (Government Decree 426/2025 (XII. 23.), Nemzeti Jogszabalytar) so 2 x 322,800 x 0.70 = HUF 451,920/month; Magyar Allamkincstar GYED page, cross-checked by officina.hu + financera.hu 2026", "last_updated": "2026-08-08", "confidence": "high", "note": "Payout ceiling HUF 451,920/month GROSS on the 80 GYED weeks at 70% (batch-8 rework 2026-08-08: the cap amount was confirmed current but had been attached to the 24 CSED weeks, which are actually 100% uncapped - the 2026-08-06 conservative pick is superseded by the verified per-scheme structure). GYED runs from CSED's end (week 24) to the child's 2nd birthday = 80 wk, an either-parent pot counted once. The cap drifts with the minimum wage each January, annual refresh." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "OECD Hungary out-of-pocket share, among the highest in the EU + NEAK prescription co-payment structure", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 raised from 80,000 to 140,000 on the batch-D research, the largest proportional level change outside Austria. Hungary has one of the highest out-of-pocket shares in the EU, driven by prescription co-payments and private outpatient use, and no annual ceiling was found. FLAGGED: the batch-D sources disagreed by roughly a factor of two on the level, and the conservative rule takes the higher. This is the value in the sweep most in need of a dedicated follow-up. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "OECD Hungary out-of-pocket share, among the highest in the EU + NEAK prescription co-payment structure", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Hungarian Central Statistical Office, STADAT table 9.9.10. https://www.ksh.hu/docs/hun/xstadat/xstadat_evkozi/e_ingazas9_09_10.html", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Munkahelyre valo eljutashoz szukseges ido (varakozassal egyutt)", "year": "2024", "note": "KSH's own printed average for the fovaros row, one way, INCLUDING waiting time, which is what makes it door-to-door and therefore our unit. The sibling table 9.9.11 excludes waiting and gives 29; the waiting-inclusive figure is the correct one here." }, "non_cash.sick_leave": { "source": "Munka Törvénykönyve (Labor Code, Act I of 2012) Art. 126 (betegszabadság, employer-paid) + Act LXXXIII of 1997 on compulsory health insurance benefits (táppénz, insurer-paid), cash-benefit administration currently run by Magyar Államkincstár (Hungarian State Treasury), not NEAK (NEAK handles in-kind; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-C.md", "source_url": "https://www.allamkincstar.gov.hu/egeszsegbiztositas/Betegseg/tappenz-betegszabadsag", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "NAV, Erdemes kihasznalni az onkentes penztari befizetesek utan jaro adokedvezmenyt https://nav.gov.hu/sajtoszoba/hirek/Erdemes_kihasznalni_az_onkentes_penztari_befizetesek_utan_jaro_adokedvezmenyt", "last_updated": "2026-08-11", "confidence": "high", "note": "Neither a deduction nor an ordinary credit. Rendelkezes az adorol: the taxpayer declares in the SZJA return and NAV TRANSFERS 20% of contributions into the member's pension account. Caps are on the REFUND, not the contribution: HUF 150,000 voluntary pension fund, 100,000 NYESZ, 130,000 pension insurance, HUF 280,000 combined. Nothing reduces the tax base, so 0 is correct for this field, but it leaves Hungary looking worse than reality by up to HUF 280k/yr of government money that belongs in a benefits field." } }, "tallinn": { "tax.filing_basis": { "source": "PwC WWTS - Estonia (Individual - Tax administration); EMTA (Estonian Tax and Customs Board) + Grant Thornton EE: spouses file individual returns (only certain deductions are shareable), no joint assessment of income.", "last_updated": "2026-07-27", "note": "2026-07-26 filing-basis classification (Group A, closeout follow-on): separately-assessed regime; activates the per-filer split in compute() for two-earner households, so each earner gets their own 8,400 zero-band. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0). Conservative-direction fix: lowers the previously overstated two-earner tax to the statutory value (2x105k 23.93% -> 23.05%).", "confidence": "high" }, "tax.payroll_deductible": { "source": "Batch B base-deduction convention (tax audit 2026-07-14, verified): the mandatory employee unemployment (1.6%) + funded pension (2%) contributions are deductible from the PIT base without limitation, so the 22% scale runs on income net of the 3.6% (computeEstoniaTaxes, gated by this flag). Source: PwC EE individual/deductions.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.federal_brackets": { "source": "EMTA (Estonian Tax and Customs Board) + Grant Thornton EE \"Calculating the basic exemption from 2026\": flat basic exemption EUR 8,400/yr (EUR 700/mo) applied at every income; the former income-based phase-out (\"tax hump\", 14,400-25,200) was ABOLISHED. Rate flat 22% (the 2026 hike to 24% was cancelled). https://www.grantthornton.ee/en/insights1/calculating-the-basic-exemption-from-2026/", "last_updated": "2026-07-26", "note": "EE (closeout follow-on): 0% band = 2026 flat exemption 8,400 (was 7,848); no phase-out to model. Old-age pensioners get a higher 9,312, excluded (working-age ICP, the wrong persona)." }, "tax.social_security_rate": { "source": "Employee UI 1.6% + funded pension 2% = 3.6% UNCAPPED", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Estonia VAT 22% (raised from 20% 2024)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Land tax minimal", "last_updated": "2026-05-17", "confidence": "medium" }, "housing._summary": { "source": "KV.ee + Numbeo Tallinn Q1 2026", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Haigekassa universal via 13% employer payroll; premium = 0", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "Lasteaed (preschool) low-cost; lastesõim subsidized", "last_updated": "2026-05-17", "confidence": "medium" }, "food._summary": { "source": "Numbeo Tallinn 2026", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "TLT free for Tallinn registered residents (monthly_pass: 0)", "last_updated": "2026-05-17", "confidence": "high", "note": "Unique: free public transit policy since 2013" }, "discretionary._summary": { "source": "Numbeo Tallinn entertainment", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "TLL-SEA/LAX (1-stop) 2025-26", "last_updated": "2026-05-17", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Lapsetoetus child allowance + universal", "last_updated": "2026-05-17", "confidence": "high" }, "non_cash.parental_leave_weeks": { "source": "Sotsiaalkindlustusamet, family benefits: https://sotsiaalkindlustusamet.ee/en/family-benefits-and-allowances/family-benefits-overview/shared-parental-benefit (ema 100 d incl. up to 70 pre-birth, isa 30 d non-transferable, jagatav 475 d); corroborated by nordenbladet.ee + palgakulu.ee ('Hüvitist makstakse kokku 605 päeva'), fetched 2026-08-07", "last_updated": "2026-08-07", "confidence": "high", "note": "86 wk = 605 paid calendar days: ema vanemahuvitis 100 (up to 70 pre-birth + 30 post; unused pre-birth days roll INTO the shared pot, household total constant) + isa vanemahuvitis 30 (father's own, non-transferable) + jagatav 475 (one shared pot, counted once; 60-day simultaneous use consumes the pot double). 605/7 = 86.43, rounded DOWN. Was 68, the shared pot alone, which the 2026-08-01 note itself flagged as an understatement; batch 5 moved it onto the household-inclusive basis 2026-08-07. both_parents flag NOT set (maternity-style ema days + short isa paternity + one shared pot; Sweden precedent). No eligibility tenure gate (universal, residence-based). Research: docs/superpowers/research/2026-08-07-pl-b5-baltics-balkans.md." }, "non_cash.parental_leave_paid_pct": { "source": "Sotsiaalkindlustusamet, shared parental benefit https://sotsiaalkindlustusamet.ee/en/family-benefits-and-allowances/family-benefits-overview/shared-parental-benefit", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the caps rebuild moved every week into non_cash.parental_leave_capped (100% up to EUR 3,806.10/month; 86 weeks since the batch-5 restatement 2026-08-07), so no percentage-paid weeks remain on this field. Was 100 with the ceiling recorded in prose only." }, "non_cash._summary": { "source": "EE Employment Contracts Act: 28 days vacation; jagatav vanemahüvitis 475 days at 100% (capped)", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Tallinn: EUR 63,776, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-europe.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "EUR 63,776 / 0.867 = USD 73,559; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2942, R5-rounded down to 0.29. REPLACES 0.4 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Senior/all-levels 1.08. The all-levels sample is effectively all senior, which is not a possible shape for a real all-levels population. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): e-Residency / digital govt anchor. Index 0.40" }, "tax.federal_brackets_married": { "source": "As tax.federal_brackets (EMTA 2026). Estonia has no joint assessment, so the married array mirrors single. https://www.grantthornton.ee/en/insights1/calculating-the-basic-exemption-from-2026/", "last_updated": "2026-07-26", "note": "EE (closeout follow-on): mirrors the single 2026 scale (8,400 exemption + 22%)." }, "non_cash.parental_leave_capped": { "source": "Sotsiaalkindlustusamet: \"The upper limit of parental benefit is equal to two times the average taxable income subject to social tax in Estonia for one calendar month from the year preceding the year of the child's birth\", 2026 ceiling EUR 3,806.10/month, re-confirmed 2026-08-07 (the 3x -> 2x rule change applies from 2026-01-01; births through 2025-12-31 keep EUR 5,265.09); floor: worked-parent minimum = alampalk EUR 946/month from 2026-04-01 (mkm.ee). https://sotsiaalkindlustusamet.ee/en/family-benefits-and-allowances/family-benefits-overview/shared-parental-benefit", "last_updated": "2026-08-07", "confidence": "high", "note": "All 86 weeks pay 100% of own average taxable income up to EUR 3,806.10/month (2x the national average taxable income, re-based yearly; the batch-5 staleness suspicion was REFUTED, this IS the 2026 figure), floored at alampalk 946 for a parent who worked. The cap binds above ~EUR 45,700/yr, i.e. across the Tallinn ICP band. cap_amount doubles as the payout ceiling because the rate is 100. LOUD MOVER CAVEAT (F10, recorded not priced per the batch-5 A8 ruling: continuous reference-window dilution is documented, only discrete gates/tiers are mover-priced): the benefit base is Estonian social-tax income of the 12 months ending 9 FULL months before the birth month, divided by a FIXED 12, so a conceive-immediately mover has ~0-3 months of income in the window and receives ~EUR 820-1,806/month (8-17% of the $150k-basis salary), not the cap; a mover who conceives ~a year after arrival gets the full capped 100%. From 2026 the work-while-receiving restriction is abolished (no effect on this block). Research: docs/superpowers/research/2026-08-07-pl-b5-baltics-balkans.md, 2026-08-06-parental-caps-europe.md (Estonia)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Tervisekassa visit fees (EUR 20 specialist for an adult, EUR 5 for a child) and prescription deductible structure + Statistics Estonia household out-of-pocket, netted of dental", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 held at 300, which sits above the bottom-up single-adult estimate of EUR 270 and above the population-wide in-scope average of EUR 231, so it is already the conservative end and this research does not force a level change. A child's per-visit cost is much lower and there is no hospital day fee, but children see doctors more often and get no drug-tier break past age 4, so a child is about half an adult. No annual ceiling. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Tervisekassa visit fees (EUR 20 specialist for an adult, EUR 5 for a child) and prescription deductible structure + Statistics Estonia household out-of-pocket, netted of dental", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Statistics Estonia, table TT233 (Labour Force Survey). https://andmed.stat.ee/en/stat/TT233", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average time of going to work by sex and way of going to work", "year": "2019", "note": "NATIONAL all-modes total of 22.6 adopted for want of a city series; the table breaks down by sex, mode and year only. A mode-weighted urban reconstruction gives about 21.5, so 23 sits at the conservative end." }, "non_cash.sick_leave": { "source": "Ravikindlustuse seadus (Health Insurance Act) , haigushüvitis (sickness benefit), paid by employer days 4-8 and by Tervisekassa (formerly Haigekassa, renamed) from day 9.; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-C.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Estonian Tax and Customs Board (EMTA), Contributions to a supplementary funded pension https://www.emta.ee/en/private-client/taxes-and-payment/tax-incentives/contributions-supplementary-funded-pension", "last_updated": "2026-08-11", "confidence": "high", "note": "III pillar contributions are deductible from taxable income up to 15% of income taxable in Estonia and not more than EUR 6,000/yr; the lower binds. The base is reduced and the 22% income tax overpaid on it is refunded, which is a deduction, unlike Hungary where the state pays a bonus into the fund. EMTA explicitly confirms this is NOT caught by the general EUR 1,200 deductions cap in s.28-2, which covers education and donations only. The 15% reaches EUR 6,000 at about EUR 40k, so the absolute cap is the operative term across our band." } }, "taipei": { "non_cash.parental_leave_paid_pct": { "source": "Laodongjizhunfa art. 50 (maternity 8 wk, wages paid as normal) https://law.moj.gov.tw/LawClass/LawSingle.aspx?pcode=N0030001&flno=50 ; Xingbiepingdenggongzuofa art. 15 (7-day paternity, salary continued)", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the 8.8-week remainder: maternity (8 wk) and paternity (~1 wk), employer wage continuation off ACTUAL pay, confirmed 2026-08-06 to sit outside the insured-salary ceiling entirely. The 52.2 parental-leave-proper weeks moved to non_cash.parental_leave_capped. Was 83, the whole-blend headline with the NT$45,800 ceiling ignored per the old field convention." }, "non_cash.parental_leave_weeks": { "source": "勞動基準法第50條 (產假八星期) https://law.moj.gov.tw/LawClass/LawSingle.aspx?pcode=N0030001&flno=50 ; 性別平等工作法第15條 (陪產檢及陪產假七日) https://law.moj.gov.tw/LawClass/LawAll.aspx?pcode=N0030014 ; 就業保險法第19-2條 and 勞工保險局 育嬰留職停薪津貼 https://www.bli.gov.tw/0015727.html", "last_updated": "2026-08-01", "confidence": "high", "note": "61 wk = 產假 8 (mother, employer-paid) + 陪產檢及陪產假 1 (father) + 育嬰留職停薪津貼 6 months x 2 parents (26.1 x 2). Both parents may claim simultaneously since 2022-01-18 and BLI states 2 人合計最多可領取12個月. Was 8, which held maternity alone. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Conservative: 7 calendar days rather than 7 working days for the paternity week; the 2-year unpaid job-protected extension excluded." }, "tax.federal_brackets": { "source": "National Taxation Bureau of Taipei official rate table, income year 2026: net-taxable bands 610,000 / 1,380,000 / 2,770,000 / 5,190,000 at 5/12/20/30/40% (progressive differences 42,700 / 153,100 / 430,100 / 949,100, reproduced by hand) https://www.ntbt.gov.tw/english/multiplehtml/3f18d2625aea4187b0d90e9b929afe4c ; 2026 deduction stack per the MOF Nov-2025 CPI adjustment (Focus Taiwan / OCAC reprint https://www.ocac.gov.tw/OCAC/Eng/Pages/Detail.aspx?nodeid=329&pid=81752008): personal exemption 101,000 + standard deduction 136,000 single / 272,000 married + salary-income special deduction 227,000.", "last_updated": "2026-08-07", "confidence": "high", "note": "TW-1 convention: the 0% first band bakes in the salaried deduction stack (single 464,000 = 101k + 136k + 227k; married 928,000 = 2x101k + 272k + 2x227k) and the real net-income bands are shifted up by it, so the engine can be fed gross salary. Re-based 2026-08-07: the 2026-07-26 fix used the 2026 deduction stack but the 2024-25 bracket edges (590k/1.33M/2.66M/4.98M); the NTBT primary table shows the edges also moved for income year 2026 (filed May 2027). Both single and married ladders shifted. Brackets + stack re-index on the MOF's ~3% cumulative-CPI trigger, announced late Nov for the following income year - see docs/data-refresh-calendar.md. Research: docs/superpowers/research/2026-08-07-tax-closeout-asia.md (F4)." }, "tax.social_security_rate": { "source": "National Health Insurance Administration: 2026 general premium rate 5.17%, employee (Category-1) share 30% = 1.551%. https://www.nhi.gov.tw/en/cp-19435-fdaba-114-2.html", "last_updated": "2026-07-26", "note": "TW-2 (closeout 2026-07-26): NHI employee rate (was a bundled ~5.5% incl. employer-paid Labor Pension)." }, "tax.social_security_cap": { "source": "NHI 2026 premium classification table (投保金額分級表): top monthly insured amount NT$313,000 x 12 = NT$3,756,000/yr. https://www.nhi.gov.tw/en/cp-19435-fdaba-114-2.html", "last_updated": "2026-07-26", "note": "TW-2 (closeout 2026-07-26): NHI insured-salary ceiling (2026); prior value 2,634,000 was stale." }, "tax.sales_tax_rate": { "source": "Business Tax (VAT) 5%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Taipei City house tax ~0.6% effective on assessed", "last_updated": "2026-05-17", "confidence": "medium" }, "housing._summary": { "source": "591房屋 + Numbeo Taipei 2026", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "NHI universal via payroll; premium = 0", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "Free kindergarten 5yr+; 2-5 subsidized; toddler private TWD 18-22k", "last_updated": "2026-05-17", "confidence": "medium" }, "food._summary": { "source": "Numbeo Taipei 2026", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "TPASS monthly TWD 1,280 (launched 2023)", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Taipei entertainment", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "TPE-SEA/LAX EVA/China Airlines 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Taipei childcare subsidy + Universal child allowance TWD 5k/mo combined", "last_updated": "2026-05-17", "confidence": "medium" }, "non_cash._summary": { "source": "Labor Standards Act: 7 days statutory (1st yr); 30 days sick @ 50%; 8 weeks maternity 60%", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Taipei: TWD 1,409,332, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "TWD 1,409,332 / 32.28 = USD 43,660; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1746, R5-rounded down to 0.17. REPLACES 0.4 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Annual and bonus-inclusive. Do NOT apply Taiwan's 14-month multiplier; that would double-count the year-end bonus by about 17%. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Semi hub. Index 0.40" }, "tax.federal_brackets_married": { "source": "Taiwan 2026 deductions (as tax.federal_brackets): married standard deduction = 2x single (272,000); 0% band = 2x the single deduction stack. https://taxsummaries.pwc.com/taiwan/individual/deductions", "last_updated": "2026-07-26", "note": "TW-1 (closeout 2026-07-26): married 0% band TWD 928,000 (= 2x single); net-income bands unchanged." }, "tax.labor_insurance_rate": { "source": "Bureau of Labor Insurance: 2026 Labor Insurance ordinary rate 11.5% + Employment Insurance 1% = 12.5% combined; insured-employee share 20% (7:2:1 employer:employee:government) = 2.5%. https://www.bli.gov.tw/en/0010789.html", "last_updated": "2026-07-26", "note": "TW-2 (closeout 2026-07-26): Labor Insurance employee rate. Employee LI rate has minor year-to-year drift (2.3-2.5%); immaterial (<0.03pp) at ICP income." }, "tax.labor_insurance_cap": { "source": "Bureau of Labor Insurance: 2026 Labor Insurance top monthly insured salary NT$45,800 x 12 = NT$549,600/yr. https://www.bli.gov.tw/en/0010789.html", "last_updated": "2026-07-26", "note": "TW-2 (closeout 2026-07-26): Labor Insurance insured-salary ceiling. Labor Pension 6% is employer-paid and excluded from the employee deduction." }, "healthcare.statutory_employee": { "source": "Quan min jian kang bao xian fa (National Health Insurance Act) via law.moj.gov.tw pcode L0060001: art. 18 caps the general premium rate at 6% and sets the dependant rule 'juan shu zhi bao xian fei, you bei bao xian ren jiao na; chao guo san kou zhe, yi san kou ji' (dependants are paid by the insured, counted as at most three); art. 31 sets the supplementary premium on the annual cumulative bonus above four times the monthly insured amount. NHIA 2026 tables effective 115.01.01: general rate 5.17% with a 30% Category-1 employee share (nhi.gov.tw/ch/cp-19418-9eefb-2576-1.html) and the premium grade table topping out at NT$313,000/month (nhi.gov.tw/ch/cp-19421-f9533-2569-1.html). Supplementary premium 2.11% with a NT$10,000,000 per-payment cap.", "last_updated": "2026-08-07", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll: tax.social_security_rate and tax.social_security_cap for Taipei ARE the NHI line (0.01551 / 3,756,000), so the shown figure equals the charged one exactly. RESOLVED tax-side 2026-08-07 (roadmap item 6): the per-dependant premium is now CHARGED by the engine (tax.nhi_dependant_max gates a household-level term in compute(): each dependant repeats the insured's share inside taxes.payroll, at most three counted), so a single-earner family of four is charged about 4x what this display field shows. The DISPLAY field itself still surfaces the earner-only share to keep the cross-country statutory-health display contract (per-earner rate x capped wage); the dependant multiple is explained in the city-file note. The 2.11% supplementary premium on large bonuses is also excluded because it is conditional on bonus size, not a flat share of regular pay. CAVEAT: nhi.gov.tw returns 403 to fetchers, so its page titles and effective dates were confirmed but the table bodies were read via cross-citation; the statute text from law.moj.gov.tw is fully primary. A 2026 proposal to raise the supplementary cap to NT$50,000,000 and move to annual settlement is NOT in force, earliest 2027." }, "childcare.subsidy_per_kid_annual": { "source": "Taipei childcare allowance schedules", "last_updated": "2026-08-05", "confidence": "high", "note": "ZEROED 2026-08-05 (CC5): the identical TWD 60,000 sits in govt_benefits.child_allowance_per_kid_annual (60000), a proven double count. The monthly rates still need a model decision (Model P: quasi-public 25,000/16,000 with allowance kept, vs Model Q: subsidised 7,000/3,000 with allowance zeroed); held for owner. Working: 2026-08-05-cc5-four-cities-subsidy.md" }, "childcare.infant_monthly": { "source": "Model P per docs/superpowers/research/2026-08-05-cc5-four-cities-subsidy.md (quasi-public/private market rates; owner decision 2026-08-05)", "last_updated": "2026-08-05", "confidence": "medium", "note": "Model P (owner decision 2026-08-05): quasi-public market rate 25000 TWD/mo, with the TWD 60,000 child allowance KEPT in govt_benefits as real cash. The old rate matched no real regime and its phantom subsidy double-counted the allowance. Model Q (subsidised public 7,000/3,000 with allowance zeroed) was declined: it is honest only if the household reliably wins a public place, which waitlists make doubtful. Conservative under rule 5." }, "childcare.preschool_monthly": { "source": "Model P per docs/superpowers/research/2026-08-05-cc5-four-cities-subsidy.md (quasi-public/private market rates; owner decision 2026-08-05)", "last_updated": "2026-08-05", "confidence": "medium", "note": "Model P (owner decision 2026-08-05): quasi-public market rate 16000 TWD/mo, with the TWD 60,000 child allowance KEPT in govt_benefits as real cash. The old rate matched no real regime and its phantom subsidy double-counted the allowance. Model Q (subsidised public 7,000/3,000 with allowance zeroed) was declined: it is honest only if the household reliably wins a public place, which waitlists make doubtful. Conservative under rule 5." }, "non_cash.parental_leave_capped": { "source": "Jiuyebaoxianfa art. 19-2 (60% of average insured salary) https://law.moj.gov.tw/LawClass/LawSingle.aspx?pcode=N0050021&flno=19-2 ; MOL 20% salary subsidy paid in the same transfer ; 2026 labor-insurance salary grade table tops out at NT$45,800/month (confirmed via 3 independent 2026 payroll sources, consistent with the stored tax.labor_insurance_cap). Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (Taiwan) and docs/superpowers/research/2026-08-06-pl-caps-east-asia.md", "last_updated": "2026-08-06", "confidence": "high", "note": "Payout ceiling NT$36,640/month = (60% + 20%) x the NT$45,800 insured-salary ceiling, across the 6-months-x-2-parents allowance weeks. No employer wage-continuation backstop exists behind this benefit (the leave is formally unpaid with a bolted-on insurance allowance), so the ceiling is the whole story: real replacement ~18-37% at Taipei tech salaries versus the 80% slice headline. The 20% top-up remains an MOL administrative program with no sunset date published, flag unchanged from the 2026-08-01 note." }, "tax.nhi_dependant_max": { "source": "National Health Insurance Act via law.moj.gov.tw pcode L0060001: dependants' premiums are paid by the insured, and when the number of dependants exceeds three the premium is calculated on the basis of three (juan shu ... chao guo san kou zhe, yi san kou ji). Employee formula (Category 1): insured payroll x 5.17% x 30% x (1 + number of dependants). Cross-checked: GoGlobal Taiwan hiring guide (formula), NHIA premium pages.", "last_updated": "2026-08-07", "confidence": "high", "note": "Roadmap item 6 (2026-08-07): gates the household-level dependant charge in compute() - every dependant (non-working spouse, children) repeats the insured's NHI share inside taxes.payroll, at most 3 counted, so a single-earner family of four pays 4x the earner-only share the engine used to charge. Children ride the HIGHER earner's insured salary (the statute allows either parent; higher = conservative, documented judgment call). An employed spouse is self-insured and never a dependant. The 2.11% supplementary premium on bonus slices stays unmodeled. Research: docs/superpowers/research/2026-08-07-tax-closeout-asia.md (F5)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "NHI copayment schedule: outpatient registration fee, capped per-prescription drug copayment, inpatient coinsurance with annual caps", "last_updated": "2026-08-08", "confidence": "low", "note": "LEVEL NOT RE-VERIFIED in the 2026-08-08 sweep; row 1 is held and only the household curve was added, on the default child factor. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "NHI copayment schedule: outpatient registration fee, capped per-prescription drug copayment, inpatient coinsurance with annual caps", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. MOTC's daily transport usage survey is the right series but was not reached. A secondary citation of the 2018 Taipei City transport opinion survey suggests roughly 29 min one way, so 35 may be high." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Labor Pension Act art. 14 https://law.moj.gov.tw/LawClass/LawAll.aspx?pcode=N0030020 ; Bureau of Labor Insurance https://www.bli.gov.tw/0017597.html ; monthly contribution wage grade table https://www.bli.gov.tw/0013083.html", "last_updated": "2026-08-11", "confidence": "medium", "note": "Voluntary employee contribution to the worker's own labour-pension account, up to 6% of monthly wage, EXCLUDED from that year's taxable salary income (art. 14 paras 3 and 5). The base is the monthly contribution wage grade table, whose top grade is TWD 150,000/month, so the true annual ceiling is 6% x 150,000 x 12 = TWD 108,000. Source states a MONTHLY rate; annualised here. Genuinely separate from and additional to the employer 6%, which app.js already excludes as employer money. MEDIUM because the BLI grade-table PDF is not machine-readable: the TWD 150,000 top grade comes from three independent 2026 Taiwanese sources that all derive 9,000/mo and 108,000/yr from it. Worth one manual confirmation." } }, "bangalore": { "tax.federal_brackets": { "source": "Income Tax Department e-filing portal, new regime (s.115BAC, renumbered s.202 by the Income-tax Act 2025): https://www.incometax.gov.in/iec/foportal/help/individual/return-applicable-1 . Standard deduction INR 75,000 confirmed unchanged for FY 2026-27 (KPMG GMS Flash Alert 2026-037, 16 Feb 2026: https://kpmg.com/xx/en/our-insights/gms-flash-alert/2026/flash-alert-2026-037.html ; Finance Bill 2026 cl. 2-3 First Schedule retains the rate structure).", "last_updated": "2026-08-01", "confidence": "high", "note": "New regime, salaried, FY 2026-27. Statutory slabs are 4L/8L/12L/16L/20L/24L; we fold the INR 75,000 salaried standard deduction in by shifting EVERY boundary up by 75,000, so the engine can be fed gross salary. CORRECTED 2026-08-01: the previous data shifted ONLY the first boundary (4,00,000 to 4,75,000), which exempted the deduction at 5% instead of at the taxpayer's marginal 30% and overstated tax by INR 19,500 for every salary in the 30% band, rising to INR 21,450 once the 10% surcharge applied. The surcharge boundaries are shifted by the same 75,000 for consistency, since the statute tests surcharge on total income after the deduction. Post-fix the engine reproduces three independent FY 2026-27 calculators to the rupee: gross 30L gives 4,75,800 (15.86%), 50L gives 10,99,800 (22.00%), 1cr gives 29,25,780 (29.26%). Found while fingerprinting user correction flag:01784996219767-35e6f78c, whose own claim was rejected." }, "tax.cess_rate": { "source": "Health & Education Cess 4% on income tax", "last_updated": "2026-05-17", "confidence": "high" }, "tax.social_security_rate": { "source": "EPF employee 12% on basic salary", "last_updated": "2026-05-17", "confidence": "high" }, "tax.social_security_cap": { "source": "EPFO International Worker framework (EPF Scheme 1952 para 83, 2008 amendment, carried forward into the EPF Scheme, 2026 notified 1 July 2026 under the Code on Social Security), described identically across AscentHR, cgaa.org, gconnect.in's EPFO brochure summary and taxaaram.com; SSA list per Dezan Shira / ExpatOrbit, UK signing 10 Feb 2026 per KPMG Flash Alert; US absence per india-briefing.com (\"pursuing\"). epfindia.gov.in itself was unreachable on two separate attempts (DNS, then connection timeout on re-verification 2026-08-06); the gap is network access from our environment, not thin sourcing. docs/superpowers/research/2026-08-06-forced-savings-foreigner-rules.md", "last_updated": "2026-08-06", "confidence": "medium-high", "note": "The INR 15,000/mo wage ceiling (cap INR 180,000/yr, max employee contribution INR 21,600) is the DOMESTIC basis. A foreign-passport International Worker at an EPF-covered establishment contributes 12% on FULL salary with no ceiling, unless a home-country SSA plus Certificate of Coverage exempts them. Engine seam: computeIndiaTaxes reads state.in_epf_intl_worker, absent or false = this capped basis, true = uncapped 12%. The target side of a comparison defaults to the uncapped IW basis when the SOURCE country holds no SSA with India (INDIA_SSA_COUNTRIES in app.js; the US has none, the UK signed Feb 2026); SSA-source movers keep this capped basis rather than the full COC exemption, which is paperwork-gated (Chile Law 18.156 precedent, conservative rule). At INR 5,000,000 gross the two bases are 21,600 vs 600,000 of employee EPF. Confidence medium-high: the mechanism is uniform across independent sources but the EPFO primary text was not reached; a Karnataka HC challenge to the uncapped rule exists and EPFO has stated the framework stands. TODO (2026-08-06): the EPF Scheme, 2026 applies the INR 15,000 ceiling to international workers in some provisions while others still read \"total wages\"; EPFO called this a typographical and drafting inconsistency, said no proposal exists to change the international-worker framework, and is weighing a corrigendum (Business Standard, 16 July 2026). Re-check this cap once the corrigendum issues. Re-checked 2026-08-07: the corrigendum has still NOT issued (latest reporting remains Business Standard 2026-07-16; EPFO maintains the IW framework is carried forward unchanged, so the uncapped IW default stands), and the DOMESTIC INR 15,000/mo ceiling is also confirmed still in force (the proposed rise to 21k is delayed indefinitely; the SC four-month directive of Jan 2026 has produced no enacted decision). The watch now rides docs/data-refresh-calendar.md rather than this dated note." }, "tax.sales_tax_rate": { "source": "GST effective blended ~18% for services", "last_updated": "2026-05-17", "confidence": "medium" }, "tax.property_tax_rate_pct": { "source": "BBMP property tax effective ~0.3% on guidance value", "last_updated": "2026-05-17", "confidence": "medium" }, "housing._summary": { "source": "MagicBricks + 99acres Bangalore Q1 2026 + Numbeo", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "MagicBricks listing-count CDF, 2026-08-18 (SERVER_PRELOADED_STATE_ resultCount; card price = monthly rent, deposit separate as bookingAmtExact, cap inclusive on the rent field - instrument re-verified on this city's own test page this run; apartments only; asking INR/month): 2 BHK P25/med/P75 -> 0.695/1.347 (n=8,261); 1 BHK 0.663/1.385.", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW on both edges at both slices (2 BHK -0.125/+0.107). Applied 0.70/1.35 (nearest). Caveats: single portal; round-number listing mass at 50,000 INR noted in the CDF; informal cheap tail missing from the pool so the true citywide spread is wider - the fix is conservative. 2 BHK anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-india.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (E). NOTE the stored medians sit near the portal pool's ~p70 - a level observation, flagged, out of scope here." }, "healthcare._summary": { "source": "IN has NO universal healthcare (universal_healthcare: false); private health insurance standard for tech families", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "Bangalore international preschools INR 12-30k; domestic help ~INR 15-25k/mo blended", "last_updated": "2026-05-17", "confidence": "medium" }, "food._summary": { "source": "Numbeo Bangalore 2026", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "BMTC monthly bus pass INR 1,500 (most tech workers use Ola/Uber/private)", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Bangalore entertainment", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "BLR-SEA/LAX 1-stop via DXB/SIN/FRA 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No federal cash child allowance for non-disadvantaged families", "last_updated": "2026-05-17", "confidence": "high" }, "non_cash._summary": { "source": "Shops & Establishments Act (Karnataka) + Maternity Benefit Act 26 weeks 100%", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Bangalore: INR 3,613,778, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "INR 3,613,778 / 95.48 = USD 37,849; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1514, R5-rounded down to 0.15. REPLACES 0.23 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. TC, NOT CTC. An Indian recruiter quoting the same package adds employer PF and gratuity on top, so a CTC figure is not comparable. Closes sweep B's finding that Bangalore existed at 0.20, 0.23 and 0.24 across six records simultaneously. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): USD-equivalent low but per-rupee comp high. Index 0.20" }, "tax.filing_basis": { "source": "Income-tax Act - individual assessee; no joint/spousal returns", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "tax.surcharge_brackets": { "source": "IN-1: new-regime high-earner surcharge on the income tax; 0 (<50L) / 10% (50L-1cr) / 15% (1cr-2cr) / 25% (>2cr, the new-regime cap), applied as a step on total income with marginal relief; 4% cess then runs on tax+surcharge. PwC IN taxes-on-personal-income; Taxmann surcharge/cess.", "last_updated": "2026-07-26", "confidence": "high" }, "non_cash.parental_leave_weeks": { "source": "Maternity Benefit Act 1961 s.5(3) as amended 2017: 26 weeks at 100% of average daily wage, employer-paid; Act text via state-government mirror (wb.gov.in PDF; central gov.in domains unreachable at check, flagged)", "last_updated": "2026-08-06", "confidence": "high", "note": "26 = maternity only, and on the household-inclusive basis this IS the household total: India has no statutory paid paternity or parental leave for private-sector employees as of 2026 (CCS Rule 43-A covers government employees only; the 2017 Paternity Benefit Bill lapsed). Caveats: establishments with 10+ employees; 26 weeks for the first two children, 12 for the third and beyond (we store the first-child case). Confirmed on the basis 2026-08-06. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "Maternity Benefit Act 1961 s.5(3) as amended 2017: 26 weeks at 100% of average daily wage, employer-paid; Act text via state-government mirror (wb.gov.in PDF; central gov.in domains unreachable at check, flagged)", "last_updated": "2026-08-06", "confidence": "high", "note": "100% of the average daily wage (3-month average), employer-paid, no monetary cap. Confirmed 2026-08-06." }, "non_cash.vacation_days_statutory": { "source": "Tamil Nadu Shops and Establishments Act 1947 s.25; Delhi Shops and Establishments Act 1954; Telangana Shops and Establishments Act 1988 s.30; Karnataka Shops and Commercial Establishments Act 1961 s.15; Maharashtra Shops and Establishments Act 2017 s.18(3); Factories Act 1948 s.79", "last_updated": "2026-08-08", "confidence": "medium", "note": "CORRECTED 2026-08-08, 18 to 12. India has NO national annual-leave statute for office workers, so this nationally-enforced field has to pick one number for six cities in five states. The old 18 was sourced to Karnataka, and Karnataka's own Shops and Commercial Establishments Act 1961 s.15 does not say 18: it grants one day per twenty days worked, the same accrual formula as Factories Act 1948 s.79, whose administered output is about 15. So 18 was unsupported by the statute it cited. Verified range across the states we actually model: Tamil Nadu (Chennai) s.25 gives a flat 12, Delhi 15, Telangana (Hyderabad) s.30 a flat 15, and Karnataka (Bangalore) plus Maharashtra (Mumbai, Pune) s.18(3) both run the 1-in-20 accrual that lands near 15. We store 12, the lower end, per rule 5 and because Tamil Nadu's is a clean section-specific floor covering one of our own cities rather than an extrapolation. LIVE ALTERNATIVE: 15 is equally defensible and is the modal figure across five of the six cities; flip this if the owner prefers representativeness to conservatism. Also checked: the OSH Code 2020 s.32 is in force from 2025-11-21 with rules from May 2026, but it is scoped to factory-type establishments and does not yet displace the state Shops Acts for commercial or IT employers. No six-day or calendar conversion applies: the flat states state days directly, and the accrual denominator counts actual attendance days." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series", "last_updated": "2026-08-08", "confidence": "low", "note": "Almost all Indian health spending is out of pocket, and employer group cover is a premium that is not booked here. Two city tiers preserved. FLAGGED for a future look: claims and premium data both hint row 1 may be low for a Bangalore or Mumbai private-hospital household. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. India has no ACS-style journey-to-work time question: Census 2011 collects DISTANCE, not time, and the Time Use Survey does not cut to city level. The only quotable series is MoveInSync corporate-shuttle telematics (49 min 2019, 54 min 2023), which measures employer-cab riders travelling to tech parks and is structurally biased long against an all-commuters all-modes field, so it was deliberately not cited." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-F.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Income Tax Department, New tax regime vs old tax regime FAQs https://www.incometax.gov.in/iec/foportal/help/new-tax-vs-old-tax-regime-faqs", "last_updated": "2026-08-11", "confidence": "high", "note": "No deduction under the regime the engine models. The NEW regime (s.115BAC) is the statutory default from AY 2024-25 and disallows Chapter-VIA deductions except 80CCD(2), 80CCH and 80JJAA, so 80C (where EPF would sit) and the extra 80CCD(1B) NPS deduction are both unavailable; 80CCD(2) is EMPLOYER NPS money, not the employee's own. computeIndiaTaxes models the new regime (slabs shifted by the INR 75,000 standard deduction) and charges EPF in payroll without removing it from the base, which is correct there. If an old-regime path is ever added, the value there is INR 200,000 and it must net off the EPF already flowing through payroll." } }, "dubai": { "tax.federal_brackets": { "source": "UAE has NO personal income tax (federal or emirate)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.social_security_rate": { "source": "GPSSA applies to GCC nationals only; expats (ICP) exempt → 0", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Federal VAT 5% (since Jan 2018)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Dubai has 5% housing fee (rental) but no annual property tax on owner-occupied", "last_updated": "2026-05-17", "confidence": "high" }, "housing._summary": { "source": "Bayut + Property Finder Dubai Q1 2026 + Numbeo", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "DHA decree mandates employer-provided private health insurance", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "Dubai private nursery/preschool AED 35-80k/yr blended", "last_updated": "2026-05-17", "confidence": "high" }, "food._summary": { "source": "Numbeo Dubai 2026", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "RTA Nol Monthly pass AED 350", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Dubai entertainment", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "DXB-SEA/LAX nonstop Emirates 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No cash child allowance for expats", "last_updated": "2026-05-17", "confidence": "high" }, "non_cash._summary": { "source": "UAE Labour Law 2022: 30 days annual + 90 days sick (tiered) + 60 days maternity tiered", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Dubai: AED 324,044, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "AED 324,044 / 3.67 = USD 88,295; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3532, R5-rounded down to 0.35. REPLACES 0.85 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. No income tax, so gross is about net. ALLOWANCE INCLUSION UNKNOWN: UAE packages bundle housing and transport at 40 to 50% of the package and levels.fyi has no allowance field. That is a basis risk, not a precision risk. End-of-service gratuity is not included. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0% tax + AED-USD peg → high USD index. 0.85" }, "non_cash.parental_leave_weeks": { "source": "Federal Decree-Law 33/2021 art. 30 (maternity 60 days: 45 full + 15 half pay) and art. 32 (parental leave 5 working days each parent); u.ae official pages re-fetched 2026-08-07. Research: docs/superpowers/research/2026-08-07-pl-b5-africa-gulf-sea.md, 2026-08-06-parental-caps-asia-mideast.md (UAE), 2026-08-06-pl-caps-gulf.md", "last_updated": "2026-08-07", "confidence": "high", "note": "10.57 wk = 60 calendar days maternity (8.57 wk) + 5 working days parental leave PER PARENT (2 x 1.0 wk): art. 32 grants the 5 days to EACH parent (male or female employee, u.ae official reading re-verified 2026-08-07), so the mother's own 5 days count too, and working days convert /5 per the house convention (the old 9.29 dropped the mother's days AND converted the father's /7). Batch-5 restatement on the household-inclusive basis; research: docs/superpowers/research/2026-08-07-pl-b5-africa-gulf-sea.md. Unpaid extensions (45 illness days) excluded." }, "non_cash.parental_leave_paid_pct": { "source": "Federal Decree-Law 33/2021 art. 30: 45 days full pay + 15 days half pay; art. 32: 5 working days full pay per parent. Employer-paid wage continuation, no social-insurance intermediation for expats (GPSSA is nationals-only pension)", "last_updated": "2026-08-07", "confidence": "high", "note": "89.9 blended = (6.43 wk x 100 + 2.14 wk x 50 + 2.0 wk x 100) / 10.57 (maternity 45d full + 15d half, calendar; parental 5 working days x 2 parents at full pay). Was 88.5 on the 65-calendar-day construction that dropped the mother's art. 32 days. Confirmed genuinely UNCAPPED on the 2026-08-06 cap sweep, re-verified 2026-08-07: direct employer continuation of the actual contractual wage at any salary, no tenure gate. Minor open detail: whether \"full pay\" includes fixed allowances rests on one secondary source; does not affect the cap verdict." }, "non_cash.vacation_days_statutory": { "source": "Federal Decree-Law 33/2021 art. 29; UAE government portal https://u.ae/en/information-and-services/jobs/employment-in-the-private-sector/types-of-leaves-and-entitlements-in-the-private-sector/annual-leave", "last_updated": "2026-08-08", "confidence": "medium-high", "note": "CORRECTED 2026-08-08, 30 to 22. Art. 29 grants 30 days after one year, and the government portal states that public holidays or agreed leave days falling within the annual leave are counted as part of it, which makes the 30 a CALENDAR figure. We had stored it unconverted. Conversion convention documented in docs/superpowers/research/2026-08-08-vacation-days-non-us-sweep.md: a calendar-day entitlement taken as a contiguous block converts to the Mon-Fri days inside its span, so 30 calendar days is 22 weekdays and 21 calendar days is 15. The same rule already governs our Spain, Brazil, Peru, Argentina, Ecuador, Estonia and Azerbaijan values. No longer-service tier exists above it, so this is both the floor and the steady state." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "DHA and HAAD mandatory employer-provided cover with its published co-payment and co-payment-cap structure", "last_updated": "2026-08-08", "confidence": "medium", "note": "Employer-provided cover is legally mandatory for the employee, so the tier-1 premium is 0 and dependant coverage is a PREMIUM; this field carries the co-payment only, verified co-pay-sized rather than premium-sized. Two emirate values preserved. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "DHA and HAAD mandatory employer-provided cover with its published co-payment and co-payment-cap structure", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. The RTA/DGHR 2024 flexible-working release publishes only PERCENTAGE reductions in peak travel time and no absolute baseline in minutes, so it cannot support any stored value." }, "non_cash.sick_leave": { "source": "Federal Decree-Law No. 33 of 2021 (Regulating Labour Relations), Article 31.; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-H.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "PwC Worldwide Tax Summaries, United Arab Emirates (reviewed 12 Mar 2026): \"There is currently no personal income tax in the United Arab Emirates\"", "last_updated": "2026-08-11", "confidence": "high", "note": "Recorded for explicitness only; it changes no output by construction. computeUAETaxes returns { federal: 0, payroll: 0, state: 0, total: 0 } and never reads taxable_income, so no deduction ceiling can move a UAE number. A deduction from a zero-rate base is meaningless." } }, "bangkok": { "tax.federal_brackets": { "source": "Thailand Revenue Department PIT 2026 (8 brackets 0-35%)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.social_security_rate": { "source": "Social Security Office employee 5% capped at THB 15k/mo wage", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "VAT 7%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Land & Buildings Tax 2020: residential primary residence largely exempt; ~0.02-0.1% effective", "last_updated": "2026-05-17", "confidence": "medium" }, "housing._summary": { "source": "DDproperty + Hipflat Bangkok Q1 2026 + Numbeo", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "DDproperty listing-count CDF, 2026-08-17 (condos for rent, Bangkok province TH10, fixed bedroom count, maxPrice sweep with inclusive bounds verified by minprice complement; asking THB/month per unit): 2BR P25/med/P75 = 34,000/50,300/75,000 (n=19,788) -> 0.68/1.49; 1BR 18,100/23,500/32,600 (n=28,581) -> 0.77/1.38.", "last_updated": "2026-08-17", "confidence": "medium", "note": "REBASED 2026-08-17 (ledger cluster housing-0817) from the calibrated default 0.82/1.24, which is TOO NARROW for this dual-market city in the flattering direction (2BR both edges far out: -0.14/+0.25). Near-identical reading to Mexico City (0.64/1.49), the class the 2026-08-16 scoping predicted. Pool choice: CONDO pool, the like-for-like universe of the stored medians (2BR 38,000 THB is a condo-pool level); the Thai-local apartment segment barely overlaps it (RentHub open GraphQL priceHistogram: n=7,599, median ~4,000 THB/month, ratios 0.72/1.44 all-room-types, no bedroom filter), so the condo-only measurement TRUNCATES the citywide pool and the applied pair is conservative - the true citywide spread is wider still. Duplication and serviced-condo stock fatten the top tail (read 1.49 with that discount) but cannot explain the lower-edge miss. 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-17-housing-band-fixes-bangkok.md; decision: docs/superpowers/research/2026-08-17-housing-band-fixes.md (S2). NOTE the stored 2BR median sits ~24% below the condo-pool listing median (50,300) and its pool basis is not written down - a LEVEL question, flagged, out of scope here." }, "healthcare._summary": { "source": "Universal Coverage Scheme (gold card) + Social Security Office; private insurance ~THB 4-6k/single for expats", "last_updated": "2026-05-17", "confidence": "medium" }, "childcare._summary": { "source": "Bangkok international preschools THB 12-20k; local private THB 8-15k", "last_updated": "2026-05-17", "confidence": "medium" }, "food._summary": { "source": "Numbeo Bangkok 2026", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "BTS/MRT monthly equivalent THB 1,000-1,500 typical combo card use", "last_updated": "2026-05-17", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo entertainment Bangkok", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "BKK-SEA/LAX 1-stop 2025-26", "last_updated": "2026-05-17", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Child Support Grant THB 600/mo for low-income; minimal for tech families", "last_updated": "2026-05-17", "confidence": "medium" }, "non_cash._summary": { "source": "Labour Protection Act 6 days statutory + 19 PH + 30 days sick + 90 days maternity", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Bangkok: THB 929,623, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "THB 929,623 / 33.4 = USD 27,833; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1113, R5-rounded down to 0.11. REPLACES 0.25 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Agoda-dominated sample, likely biased high, and senior/all-levels of 2.27 fails the thinness screen. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Lower than other ASEAN tech hubs but rising. Index 0.25" }, "tax.filing_basis": { "source": "Revenue Department (rd.go.th) - spouses may file separately; individual filing modeled", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "non_cash.parental_leave_weeks": { "source": "Labour Protection Act (No. 9) B.E. 2568 (2025), Royal Gazette 7 Nov 2025, in force Dec 2025: maternity 98 -> 120 days (employer-paid first 60), NEW 15-day paternity/spousal leave at 100% (s.41/1). Confirmed via mol.go.th announcement + DFDL, PKF, Tilleke & Gibbins, Nagashima client alerts. Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (Thailand) and docs/superpowers/research/2026-08-06-pl-caps-southeast-asia.md", "last_updated": "2026-08-08", "confidence": "high", "note": "19.29 wk = 120 days maternity + 15 days paternity (135/7). Was 13, the STALE pre-amendment 98-day maternity figure with no paternity (Thailand had none before Dec 2025). Restated by the 2026-08-06 caps rebuild. Batch-7 audit (2026-08-08): paternity 15 days ruled CALENDAR, FINAL - the Senate (gov) page quotes the act's plain-days wording, and the LPA itself contrasts plain-day caps (s.59 maternity) with working-day caps (s.57 sick leave, 30 working days); a '15 working days' secondary reading REFUTED (confusion with the s.59/2 rate phrase). No tenure gate on the employer-paid components. Spouse-leave wording is 'khu somrot' (registered spouse) - unregistered partners unsettled, no figure impact for the married ICP. Conditional +15 d newborn-complications leave at 50% excluded." }, "non_cash.parental_leave_paid_pct": { "source": "Labour Protection Act as amended Dec 2025: employer pays full wages for the first 60 maternity days and the 15 paternity days", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the 10.71-week remainder (60 employer-paid maternity days + 15 paternity days, ordinary uncapped wage continuation). The 60 SSO-paid days live in non_cash.parental_leave_capped. The engine blend reproduces the statutory headline ~77.8% at low salaries and ~58-59% at Bangkok tech pay." }, "non_cash.parental_leave_capped": { "source": "SSO insurable-wage ceiling THB 17,500/month from 1 Jan 2026 (first step of the phased rise from 15,000; 20,000 in 2029, 23,000 from 2032), multi-source corroborated (PKF, HLB, AIM, RLC; sso.go.th unfetchable) ; Social Security Act maternity cash benefit 50% of covered wage, up to 90 days per childbirth", "last_updated": "2026-08-08", "confidence": "medium-high", "note": "Payout ceiling 50% x 17,500 = THB 8,750/month on the 60 maternity days the employer does not pay (8.57 wk). Batch-7 upgrade of the old inference flag (2026-08-08): the SSO 50% x 90-day childbirth benefit is CONFIRMED continuing post-amendment (sso.go.th benefit page unchanged + the Senate (gov) note that the remainder of the 120 days runs on social-security rights), but no SSO circular explicitly mapping benefit days to leave days 61-120 has been located - the fold stays conservative and credits only the 60 non-employer-paid days (the other 30 SSO benefit days would stack over employer-paid days and stay uncredited). Gate: >=5 contribution months within the 15 months before the childbirth month - the mover passes in BOTH conceive scenarios. The 50% cash benefit applies to the first TWO childbirths only. Ceiling THB 17,500/mo corroborated for 2026 (max contribution 875/mo); phased rise 20,000 in 2029, 23,000 from 2032 - tracked in docs/data-refresh-calendar.md. The cap binds immediately at any Bangkok professional salary." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Thailand's Office of Transport and Traffic Policy and Planning does publish the right survey, but otp.go.th sat behind Cloudflare and returned 403 to every attempt. This is the cheapest win on any re-run; two candidate PDF URLs are recorded in the research file." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Revenue Department, Deductions and allowances https://www.rd.go.th/english/6045.html ; https://sherrings.com/rmf-tax-allowances-thailand.html", "last_updated": "2026-08-11", "confidence": "high", "note": "Provident fund contributions, \"not more than 15% of wage, but not exceeding 500,000 baht\". The THB 500,000 is a COMBINED ceiling across PVD / RMF / SSF / NSF / pension life insurance. RMF alone permits 30%; the provident fund's 15% is taken per the conservative house rule. Does not bind at our incomes (6% binds only above THB 8.33m, 10% above 5.0m, against a senior Bangkok professional at THB 1.2-2.5m): a top-end guard. SEPARATE GAP noted, out of scope: the engine omits the Thai social-security contribution's own THB 9,000 deduction." } }, "hong-kong": { "tax.filing_basis": { "source": "PwC WWTS - Hong Kong (Tax administration): spouses are normally taxed separately; joint assessment only by election.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "IRD Salaries Tax 2025/26 progressive + HKD 132k personal allowance", "last_updated": "2026-05-17", "confidence": "high" }, "tax.social_security_rate": { "source": "MFA MPF employee 5% capped at HKD 30k/mo income (max HKD 1,500/mo contribution)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Hong Kong has NO VAT/GST", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Rates 5% of rateable value (well below market); ~0.1% effective on owner-occupied", "last_updated": "2026-05-17", "confidence": "medium" }, "housing._summary": { "source": "Squarefoot/Spacious + Numbeo HK Q1 2026", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Hospital Authority universal (slow); private health insurance ~HKD 600-1,500 single (expat standard)", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "Free Quality Kindergarten Scheme (free 3-6); nursery 0-3 expensive", "last_updated": "2026-05-17", "confidence": "high" }, "food._summary": { "source": "Numbeo HK 2026", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "MTR monthly pass equivalent ~HKD 500", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo HK entertainment", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "HKG-SEA/LAX nonstop Cathay 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "IRD salaries tax allowances: HKD 130,000 child allowance per kid reduces TAXABLE INCOME; no cash child benefit exists in Hong Kong", "last_updated": "2026-05-17", "confidence": "high", "note": "Held at 0: Hong Kong pays no cash child allowance. The real benefit is the HKD 130,000-per-kid taxable-income allowance, which the v1 engine excludes (computeHongKongTaxes runs on standard deductions only; worth ~HKD 22,100/kid at the 17% top slice, a known understatement of Hong Kong)." }, "non_cash._summary": { "source": "EO Cap 57 + 17 GH + 4 days/mo sick after 1yr + 14 weeks maternity 80%", "last_updated": "2026-05-17", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Hong Kong: HKD 604,278, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "HKD 604,278 / 7.84 = USD 77,076; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3083, R5-rounded down to 0.3. REPLACES 0.85 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Thin: the top employer sits only 1.1% above the median. Bank-IT and crypto skew. No domestic cross-check was run for Hong Kong or mainland China. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): HKD pegged to USD floor at high index. 0.85" }, "non_cash.parental_leave_weeks": { "source": "Employment Ordinance Cap. 57 ss.12-15 (14 weeks maternity) ; ss.15A-15L paternity - gazette text of Ord. 21/2014 (labour.gov.hk PDF) read verbatim, 5 days since the 2018 (No. 3) amendment for births from 2019-01-18 ; labour.gov.hk EAO2020 + EAO2025 pages. Research: docs/superpowers/research/2026-08-08-pl-b7-east-sea-gulf.md (batch 7) and docs/superpowers/research/2026-08-06-pl-caps-east-asia.md", "last_updated": "2026-08-08", "confidence": "high", "note": "15.0 wk = maternity 14 (mother) + paternity 5 WORKING days /5 = 1.0 (father). Was 14.7 (5/7): the batch-7 day-ruling flipped on the black-letter test - s.15G(1) makes paternity leave 'in addition to rest days, holidays and annual leave' and s.15G(2) DISPLACES a notified day that falls on a rest day/holiday to the next day, so a paternity day can never absorb a non-working day (distinct from the Philippines plain-days precedent). LOUD mover caveat: PAID maternity/paternity require 40 weeks' continuous contract before the leave (ss.12-15 / s.15H) - the work-to-term mover (~48+ wk) passes, but a household conceiving immediately on arrival takes the leave UNPAID (zero paid weeks). Caveat-only per the Israel/Lithuania precedent, no haircut. The 2025 417/468 continuous-contract amendment (applying from 2026-01-18) relaxes eligibility for part-timers only; durations and the 4/5 rate are unchanged through 2026." }, "non_cash.parental_leave_paid_pct": { "source": "Employment Ordinance: 4/5 (80%) of average daily wages for both maternity and paternity ; info.gov.hk RMLPS release: \"Employers must pay MLP for all of the 14 weeks to the employee\", the HKD 80,000 figure caps only the weeks-11-14 reimbursement TO THE EMPLOYER", "last_updated": "2026-08-06", "confidence": "high", "note": "80% confirmed genuinely UNCAPPED for the employee on the 2026-08-06 cap sweep, a surprise reversal: the widely-quoted HKD 80,000 is a reimbursement ceiling to the employer (RMLPS, maternity weeks 11-14 only) and does not reduce the employee's statutory 4/5-of-actual-wages entitlement at any salary. Recorded so nobody re-imports it as an employee cap." }, "non_cash.vacation_days_statutory": { "source": "Employment Ordinance (Cap. 57) s.41AA https://www.elegislation.gov.hk/hk/cap57", "last_updated": "2026-08-08", "confidence": "high", "note": "CORRECTED 2026-08-08, 10 to 7. s.41AA sets paid annual leave at 7 days for years 1 and 2, rising one day per year to a 14-day maximum at 9+ years of service. The declared basis for this field is the year-1 legal floor, so 7 is the value; the stored 10 was the year-5 rung. Additive to the separate statutory holidays." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Hospital Authority public fee schedule (HKD 50 accident and emergency, HKD 75 per inpatient day) plus private GP and specialist self-pay, which is what a professional household actually uses", "last_updated": "2026-08-08", "confidence": "low", "note": "FLAGGED as the value in its batch most likely to be revised UPWARD: public fees are nominal, but a professional household's real spend is private care, which this figure probably does not capture. LEVEL NOT RE-VERIFIED in the 2026-08-08 sweep. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Hospital Authority public fee schedule (HKD 50 accident and emergency, HKD 75 per inpatient day) plus private GP and specialist self-pay, which is what a professional household actually uses", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Hong Kong Transport Department, Travel Characteristics Survey 2022, home-based work journeys. https://www.td.gov.hk/filemanager/en/content_5349/tcs2022_eng.pdf", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean journey time, home-based work", "year": "2022", "note": "Per-trip and therefore already one way, no conversion. CAVEAT: the TCS figure covers MECHANISED trips only, so short walking commutes are excluded, which biases it long. Survey self-report." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRD, Qualifying Annuity Premiums and TVC FAQ https://www.ird.gov.hk/eng/faq/annuity_qp.htm ; MPF FAQ https://www.ird.gov.hk/eng/faq/mpf.htm", "last_updated": "2026-08-11", "confidence": "high", "note": "Tax-deductible voluntary contributions (TVC) plus qualifying deferred annuity premiums (QDAP), HKD 60,000 AGGREGATE per year of assessment (2025/26 and 2026/27), TVC deducted first. A deduction from assessable income under salaries tax, not a credit, and 60,000 total rather than 60,000 each. Ordinary non-TVC voluntary MPF contributions are expressly NOT deductible, so this is the whole voluntary allowance. The separate HKD 18,000 mandatory-MPF deduction is not modeled by the engine at all (MPF is handled as payroll), so 60,000 does not double-count it." } }, "seoul": { "non_cash.parental_leave_paid_pct": { "source": "Geunrogijunbeop art. 74(4) (employer pays 100% of tongsang imgeum for maternity days 1-60, the EI benefit only discharges the employer up to its amount) via easylaw.go.kr ; Namnyeogoyongpyeongdeungbeop art. 18-2 (paternity 20 days, same employer-pay construction)", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the 11.31-week remainder: maternity days 1-60 (8.57 wk) plus the 20-day paternity leave (2.86 wk), both employer-paid wage continuation that the EI reimbursement cannot reduce, genuinely uncapped at any salary (11.43 wk exact, clipped to the 120-week stored total). The other 108.69 weeks moved to non_cash.parental_leave_capped in the 2026-08-06 caps rebuild. Was 91, the whole-blend headline with every monetary ceiling ignored." }, "non_cash.parental_leave_weeks": { "source": "고용보험법 시행령 제95조 (육아휴직 급여) https://law.go.kr/ ; 근로기준법 제74조 (출산전후휴가 90일) ; 남녀고용평등법 제18조의2 (배우자 출산휴가 20일) and 제19조 (육아휴직 1년 per parent) ; 고용24 육아휴직급여 https://www.work24.go.kr/cm/c/f/1100/selecSystInfo.do?systClId=SC00000251&systId=SI00000402", "last_updated": "2026-08-01", "confidence": "high", "note": "120 wk = 출산전후휴가 12.9 (mother) + 배우자 출산휴가 2.9 (father) + 육아휴직 12 months x 2 parents (52.2 x 2). Was 13, which held maternity alone and dropped 육아휴직 entirely. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Conservative: 12 months per parent, not the 18 available only when BOTH parents each take 3 months or more; 20 working days counted as 2.86 wk rather than the 4 calendar weeks of actual absence; single birth, not multiples." }, "tax.federal_brackets": { "source": "NTS PIT 2026 (8 brackets 6-45%)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.local_surcharge_rate": { "source": "Local Income Tax 10% of national PIT (Jiban Sodeukse)", "last_updated": "2026-05-17", "confidence": "high" }, "tax.social_security_rate": { "source": "NP 4.5% + NHIS 3.55% + LTC 0.4% × NHIS + EI 0.9% ≈ 9.4% combined employee", "last_updated": "2026-05-17", "confidence": "medium", "note": "LEGACY fallback only. Superseded 2026-07-26 by the split np_rate + nhi_ltc_rate + ei_rate (KR-4); the engine ignores this when np_rate is present." }, "tax.social_security_cap": { "source": "NP+NHIS combined cap ~KRW 71M/yr (cap varies by component; v1 blended)", "last_updated": "2026-05-17", "confidence": "medium", "note": "LEGACY fallback only. Superseded 2026-07-26 by np_cap (National Pension) with NHI+LTC and EI uncapped (KR-4)." }, "tax.sales_tax_rate": { "source": "VAT 10%", "last_updated": "2026-05-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Seoul property tax 0.1-0.4% on official price (gongsi)", "last_updated": "2026-05-17", "confidence": "medium" }, "housing._summary": { "source": "Naver/Daum 부동산 Seoul Q1 2026 + Numbeo", "last_updated": "2026-05-17", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "NHIS universal via 3.55% payroll; premium = 0", "last_updated": "2026-05-17", "confidence": "high" }, "childcare._summary": { "source": "Free Nuri Curriculum 3-5; 부모급여 KRW 1M/mo under-1", "last_updated": "2026-05-17", "confidence": "high" }, "food._summary": { "source": "Numbeo Seoul 2026", "last_updated": "2026-05-17", "confidence": "high" }, "transit._summary": { "source": "Climate Card (기후동행카드) KRW 62-65k monthly (launched 2024)", "last_updated": "2026-05-17", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Seoul entertainment", "last_updated": "2026-05-17", "confidence": "medium" }, "travel_home._summary": { "source": "ICN-SEA/LAX nonstop Korean Air/Asiana 2025-26", "last_updated": "2026-05-17", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "KR child allowance KRW 100k/mo to age 8", "last_updated": "2026-05-17", "confidence": "high" }, "non_cash._summary": { "source": "LSA 15 days/yr after 1yr; 15 PH; NO statutory sick leave; maternity 90 days 100% + leave 1yr 80% capped", "last_updated": "2026-05-17", "confidence": "high", "note": "South Korea is famously a no-statutory-sick-leave country" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Seoul: KRW 95,948,809, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "KRW 95,948,809 / 1440.53 = USD 66,607; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2664, R5-rounded down to 0.26. REPLACES 0.55 (2026-05-17), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Annual, not monthly. But 2.24x the KRW 42.88M domestic developer average, so this is a big-tech and multinational sample rather than the Korean market. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Top AP tech market ex-Japan. Index 0.55" }, "tax.filing_basis": { "source": "KR-5: no joint filing; spouses file separate returns. PwC KR tax administration: https://taxsummaries.pwc.com/republic-of-korea/individual/tax-administration", "last_updated": "2026-07-26", "confidence": "high" }, "tax.earned_income_deduction_tiers": { "source": "KR-1: earned-income deduction (근로소득공제), tiered 70/40/15/5/2% by KRW band. PwC KR deductions: https://taxsummaries.pwc.com/republic-of-korea/individual/deductions", "last_updated": "2026-07-26", "confidence": "high" }, "tax.earned_income_deduction_cap": { "source": "KR-1: earned-income deduction total capped at KRW 20M. PwC KR deductions (same URL).", "last_updated": "2026-07-26", "confidence": "high" }, "tax.personal_deduction_per_head": { "source": "KR-2: basic personal deduction KRW 1.5M per head. PwC KR deductions.", "last_updated": "2026-07-26", "confidence": "high", "note": "Only the taxpayer head is modelled; dependent (spouse/child) heads are omitted — conservative (overstates tax) and avoids double-counting kids in the individual-filing clone." }, "tax.np_rate": { "source": "KR-4: National Pension employee rate 4.75% for 2026 (pension reform: 9.0%->9.5% total from 2026-01-01). PwC KR other-taxes; koreantaxexpert 2026; Korea Herald/Lockton pension reform.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.np_cap": { "source": "KR-4: National Pension standard-income cap KRW 6.59M/mo = 79.08M/yr wage base (from Jul 2026). PwC KR other-taxes; koreantaxexpert.", "last_updated": "2026-07-26", "confidence": "high", "note": "Value rule: the higher (from-Jul-2026) cap is the less-favorable choice for the country; annual model uses the 12x monthly cap." }, "tax.nhi_ltc_rate": { "source": "KR-4: National Health Insurance incl. Long-term Care, employee ~4.0674% for 2026 (uncapped at ICP income). PwC KR other-taxes (4.0674%); koreantaxexpert (NHI 3.595% + LTC 0.4656%).", "last_updated": "2026-07-26", "confidence": "medium", "note": "The exact 2026 NHI rate is \"tentative\"; the conservative (higher, PwC) combined figure is used per the value rule." }, "tax.ei_rate": { "source": "KR-4: Employment Insurance employee 0.9% (uncapped). PwC KR other-taxes; koreantaxexpert.", "last_updated": "2026-07-26", "confidence": "high" }, "healthcare.statutory_employee": { "source": "MOHW press release '2026 NHI premium rate set at 7.19%' (mohw.go.kr board.es list_no=1487279), which states the employee and employer shares are 3.595% each; MOHW press release '2026 long-term care insurance rate 0.9448%' (list_no=1487817), which sets LTC at 13.14% of the health premium and 0.9448% of income across both parties, so 0.4724% employee (3.595% x 13.14% = 0.4724%, matching 0.9448% / 2). Ceiling: MOHW notice 2025-222 amending the monthly upper and lower health premium limits, issued 2025-12-24 and effective 2026-01-01; the monthly premium ceiling of KRW 9,183,480 implies a wage base of 9,183,480 / 0.0719 = KRW 127,725,730/month = KRW 1,532,708,760/year.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll (tax.nhi_ltc_rate 0.040674, charged uncapped by computeSouthKoreaTaxes; the display cap is far above any modeled salary so the two agree in practice). LTC is included because the regime definition covers long-term care and it is compulsory with no opt-out. CAVEAT ON THE CAP: the rates are primary from MOHW, but the KRW 9,183,480 premium ceiling reached us through five independent Korean outlets citing MOHW notice 2025-222, whose own figures sit in an unopenable attachment, and nhis.or.kr was unreachable. The derived annual wage base is therefore secondary-sourced and approximate. It does not bind at any income this calculator models, so the imprecision costs nothing." }, "childcare.subsidy_per_kid_annual": { "source": "Korean boyukryo / aihaengbok childcare voucher (Ministry of Health and Welfare); double-subtraction correction 2026-08-02", "last_updated": "2026-08-02", "confidence": "medium", "note": "Set to 0 on 2026-08-02, was KRW 6,000,000 (new meta entry; the value carried no provenance before). The old value was the childcare voucher, which the government pays direct to the centre rather than to the household. The monthly rates already carry only the residual parents actually pay, the teukbyeol-hwaldongbi extra-activity and meal charges. Seoul preschool_monthly is already 0 for the covered ages. This field holds CASH THAT REACHES THE HOUSEHOLD. Where the monthly rate it nets against is already the price the household actually pays, the value is 0 and the relief is described here instead. See docs/superpowers/research/2026-08-02-childcare-double-subtraction.md." }, "non_cash.parental_leave_capped": { "source": "Goyongbohumbeop art. 70(4) + Enforcement Decree art. 95(1) via easylaw.go.kr (fetched 2026-08-06): yugahyujik ceilings KRW 2.5M/mo months 1-3, 2.0M months 4-6, 1.6M months 7-12 (in force since 2025-01-01, current 2026); maternity days 61-90 ceiling KRW 2.2M/mo (2026-01-01). Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (South Korea) and docs/superpowers/research/2026-08-06-pl-caps-east-asia.md", "last_updated": "2026-08-06", "confidence": "high", "note": "Multi-phase caps flattened to ONE duration-weighted record because the engine holds a single capped block: (4.29 wk x 2.2M + 26.1 x 2.5M + 26.1 x 2.0M + 52.2 x 1.6M) / 108.69 = KRW 1,935,845/month, stored rounded down to 1,935,000; pct 90.4 is the same weighting of the 100/100/100/80 phase rates. At ICP salaries every phase cap binds, so the weighted form reproduces the exact statutory total payout; between the lowest and highest cap it can overstate by up to ~4% at sub-ICP salaries, the flatten cost recorded here. Excluded per the basis convention: the 6+6 bumo-yugahyujik enhanced ceilings (conditional on both parents' behaviour). At KRW 100M the blended effective rate is ~30%, versus the 91 headline this replaces." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "NHIS coinsurance by care setting (20% inpatient, 30-60% outpatient by facility level) + the income-decile-banded annual out-of-pocket ceiling", "last_updated": "2026-08-08", "confidence": "low", "note": "LEVEL NOT RE-VERIFIED in the 2026-08-08 sweep; row 1 is held and only the household curve was added. Korea's child discount is narrow, mainly under-6 inpatient at 5%, so its child increment is one of the steepest in the Asia-Pacific batch. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "NHIS coinsurance by care setting (20% inpatient, 30-60% outpatient by facility level) + the income-decile-banded annual out-of-pocket ceiling", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Statistics Korea, Population and Housing Census 2020, commuting time by province, reported in KIRI weekly trend 2021-12-06. https://kiri.or.kr/PDF/weeklytrend/20211206/trend20211206_8.pdf", "last_updated": "2026-08-08", "confidence": "high", "statistic": "시도별 통근·통학 소요시간 (commuting-to-work-and-school time by province)", "year": "2020", "note": "37.2 min, published ONE WAY with no conversion. Established as one-way by triangulation against the 2023 Seoul Survey, which states 34.5 min 편도 (one way) explicitly; a round-trip reading would put Seoul at 18.6, contradicting that twofold. GEOGRAPHY: Seoul metro CITY proper. CAVEAT: the statistic blends commuting to work with commuting to school, the same bundling as Japan. A work-only correction was derived and then DROPPED because the Japanese and Korean ratios point in opposite directions, so both countries are carried blended and symmetrically. Survey self-report (census). The widely quoted 1h26m headline is press restating this population round-trip." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "National Tax Service, pension account tax credit https://www.nts.go.kr/nts/cm/cntnts/cntntsView.do?cntntsId=7875", "last_updated": "2026-08-11", "confidence": "high", "note": "No base deduction. Pension-account relief is a 세액공제, deducted from the computed global income tax: contribution base KRW 6,000,000 (pension savings) or 9,000,000 combined with retirement-pension DC/IRP, at 15% below the income thresholds and 12% at our ICP's income. Setting 0 costs the user a real benefit worth at most about KRW 1,080,000 of tax; modeling it needs a tax-credit key, not this one. computeSouthKoreaTaxes already subtracts the full mandatory payroll, so nothing double-counts." } }, "warsaw": { "tax.inbound_regime": { "source": "Ustawa o zryczaltowanym podatku dochodowym, consolidated text Dz.U. 2025 poz. 843 (https://api.sejm.gov.pl/eli/acts/DU/2025/843/text.pdf). Art. 12 ust. 1 pkt 2b sets 12% for software-related services (PKWiU ex 62.01.1, 62.01.2, ex 62.02, 62.03.1, ex 62.09.20.0, 62.02.10.0 and the 58.29 software-publishing groupings); art. 6 ust. 4 pkt 1 lit. a caps prior-year revenue at EUR 2,000,000, converted at the NBP rate of the first business day of October (table 190/A/NBP/2025, EUR 4.2586), so the 2026 limit is PLN 8,517,200. Art. 8 ust. 2 voids the regime for the whole year if the same services are supplied to a current or former employer. Ulga na powrot: art. 21 ust. 1 pkt 152 PIT, Dz.U. 2026 poz. 592.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: confidence upgraded medium -> high, and the source moved from a PwC summary to the Sejm ELI statute text. FRAMING FLAG RAISED, NOT RESOLVED HERE: the 12% ryczalt is NOT an inbound regime. It is a general self-employment option available to a Polish local on day one and unavailable to anyone taxed as an employee, and it taxes REVENUE with no cost deduction, no PLN 30,000 allowance and no joint filing. Poland has no inbound income-tax relief for a newly arrived foreign employee at all; the nearest thing is ulga na powrot (PLN 85,528/yr for 4 years, conditions in art. 21 ust. 43; no schema field for tax.ulga_na_powrot exists). Carrying it under an inbound-regime toggle therefore offers our ICP, a relocating EMPLOYEE, a discount the statute does not give them. Recommend either reclassifying the toggle as a B2B-status scenario or dropping it; not done here because it is a product decision, and the brief for this pass said 'upgrade or flag'. The regime note now says all of this in the reader's own words. The old note's claim that the model UNDERSTATES the benefit (real ryczalt gets Maly ZUS Plus and revenue-bracket NFZ) is true only for someone already self-employed; for an employee it overstates by the whole amount. 2026 changes recorded: revenue limit PLN 8,517,200 (down on a weaker EUR); NFZ base PLN 9,228.64 (M.P. 2026 poz. 117), so a 300k+ contractor pays PLN 1,495.04/mo; Maly ZUS Plus capped at 36 months in any 60 from 2026-01-01." }, "tax.payroll_deductible": { "source": "Batch B base-deduction convention (tax audit 2026-07-14, verified): employee ZUS is deductible from the PIT base, so the brackets run on gross minus ZUS. The NFZ health contribution (9%) is NOT deductible (financed from net income since the 2022 reform). computePolandTaxes, gated by this flag (bracket path only; the ryczalt flat regime forgoes it). Source: PwC PL individual/other-taxes.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.federal_brackets": { "source": "podatki.gov.pl PIT 2026 (12%/32% post-Polski Ład 2.0; PLN 30k kwota wolna)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_rate": { "source": "ZUS / getsix: Social Insurance (ZUS) contributions in Poland 2026 (employee emerytalne 9.76% + rentowe 1.50% = 11.26%, capped at the annual 30x limit). https://getsix.eu/human-resources-payroll-in-poland/social-insurance-institution-zus-contributions-in-poland-in-2026-current-bases-limits-and-contribution-amounts-for-entrepreneurs/", "last_updated": "2026-07-26", "note": "PL-1 (closeout 2026-07-26): capped ZUS stream (pension + disability), split from the former blended 13.71%." }, "tax.social_security_cap": { "source": "ZUS: annual 30x limit (trzydziestokrotność) for pension+disability contributions, 2026 = PLN 282,600. https://getsix.eu/human-resources-payroll-in-poland/social-insurance-institution-zus-contributions-in-poland-in-2026-current-bases-limits-and-contribution-amounts-for-entrepreneurs/", "last_updated": "2026-07-26", "note": "PL-1 (closeout 2026-07-26): 2026 pension+disability cap, up from PLN 260,190 (2025)." }, "tax.health_insurance_rate": { "source": "NFZ składka zdrowotna 9% UNCAPPED (non-deductible since 2022 reform)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.sales_tax_rate": { "source": "VAT podstawowa stawka 23%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Podatek od nieruchomości ~0.1% effective", "last_updated": "2026-05-16", "confidence": "medium" }, "housing._summary": { "source": "Otodom.pl + Numbeo Warsaw Q1 2026", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "NFZ universal via 9% health payroll; premium = 0", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "Żłobki (0-3) paid; przedszkole 3-5 free; bezpłatne zerowka", "last_updated": "2026-05-16", "confidence": "medium" }, "food._summary": { "source": "Numbeo Warsaw 2026 + GUS", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "ZTM Warsaw bilet miesięczny PLN 180", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "GUS + Numbeo entertainment", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "WAW-SEA/LAX (1-stop) LOT/Lufthansa 2025-26", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Rodzina 800+ unconditional PLN 800/mo/kid = PLN 9,600/yr/kid (post-2024 reform; was 500+)", "last_updated": "2026-05-16", "confidence": "high", "note": "Top childcare cash transfer of any country we model Polish 800+ (Rodzina 800 plus, PLN 800/mo): universal, NOT means-tested." }, "non_cash.parental_leave_weeks": { "source": "Ministerstwo Rodziny, Pracy i Polityki Spolecznej, urlop rodzicielski: https://www.gov.pl/web/rodzina/urlop-rodzicielski ; ZUS zasilek macierzynski: https://www.zus.pl/swiadczenia/zasilki/zasilek-macierzynski/wysokosc", "last_updated": "2026-08-06", "confidence": "high", "note": "20 weeks urlop macierzynski plus 41 weeks urlop rodzicielski for a single birth ('41 tygodni w przypadku urodzenia jednego dziecka przy jednym porodzie'), 61 weeks for the household. 9 of the 41 are non-transferable to each parent under Poland's 2023 implementation of the EU work-life balance directive ('Kazdemu z pracownikow rodzicow dziecka przysluguje wylaczne prawo do 9 tygodni urlopu rodzicielskiego'). Corrected 2026-08-01 from 20, which captured only the maternity block and dropped the parental benefit entirely, the same defect the Vienna reporter found in Austria. Found by the cross-country sweep that followed that correction. Re-confirmed on the household-inclusive basis 2026-08-06 (the 20+41 construction already sums the per-parent 9-week reserves inside the shared 41; both-parents flag stays off per the Sweden precedent). See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "ZUS: 'zasilek macierzynski wynosi 70% podstawy wymiaru zasilku' for the urlop rodzicielski period; gov.pl records the 81,5% election. https://www.zus.pl/swiadczenia/zasilki/zasilek-macierzynski/wysokosc", "last_updated": "2026-08-06", "confidence": "high", "note": "Blended across the 61-week window: 20 weeks at 100% plus 41 weeks at 70% gives 79.8%, stored as 80, on the same blended basis as Sweden's 78. Conservative choice: electing the flat 81,5% rate within 21 days of the birth pays more, and we do not assume the household elects it. The father's 9 non-transferable weeks are always at 70%. Cap check closed 2026-08-06: zasilek macierzynski is UNCAPPED for employees (art. 36 basis is the plain 12-month average wage; the art. 46 100%-of-average-wage cap applies only after employment ends; the 250% figure is the voluntary contribution-base cap for the self-employed), so the stored blend is real for an ICP, one of the few uncapped schemes." }, "non_cash._summary": { "source": "Kodeks pracy 20 dni + 13 świąt + urlop macierzyński 20 tygodni 100% then urlop rodzicielski 41 tygodni 70% (post-2023)", "last_updated": "2026-05-16", "confidence": "high", "note": "B2B/ryczałt 12% flat regime for IT contractors is the dominant tax setup among PL tech workers. Modeled in v2 inbound_regime toggle as flat 12% PIT, but UNDERSTATES the real benefit: actual ryczałt also gets Mały ZUS Plus (revenue-scaled, ~50-70% less than regular ZUS) + revenue-bracket NFZ (lower than 9% employee rate). True effective tax on a ryczałt contractor at PLN 300k revenue is closer to 17-20% vs our model's ~24%." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Warsaw: PLN 281,628, fetched live 2026-08-11 and recorded at sweep B F8d of docs/superpowers/research/2026-08-11-comp-index-sweep-b.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "PLN 281,628 / 3.74 = USD 75,302; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3012, R5-rounded down to 0.3. REPLACES 0.36 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Warsaw by 20%. The measured p50 was 0.837 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Top PL tech market. Index 0.36" }, "tax.social_security_rate_uncapped": { "source": "ZUS / getsix: Poland 2026 employee ZUS: chorobowe (sickness) 2.45% is NOT subject to the 30x annual limit. https://getsix.eu/human-resources-payroll-in-poland/social-insurance-institution-zus-contributions-in-poland-in-2026-current-bases-limits-and-contribution-amounts-for-entrepreneurs/", "last_updated": "2026-07-26", "note": "PL-1 (closeout 2026-07-26): chorobowe 2.45%, UNCAPPED — continues on income above the 30x cap." }, "tax.filing_basis": { "source": "PwC Worldwide Tax Summaries: Poland, Individual: Tax administration (\"Married taxpayers ... may, under some conditions, choose ... jointly or separately\"); separate assessment is the default and joint (wspólne rozliczenie) is an optional election. https://taxsummaries.pwc.com/poland/individual/tax-administration", "last_updated": "2026-07-26", "note": "Closeout follow-on: separate spousal assessment is Poland’s default; joint filing is an election. Fixes the two-earner overstatement the PL-1 ZUS fix left (it had stayed joint-stacked)." }, "healthcare.statutory_employee": { "source": "NFZ / ustawa o swiadczeniach opieki zdrowotnej finansowanych ze srodkow publicznych art. 79 (9% rate) and art. 81 ust. 5 (30x ceiling does not apply to health). ZUS 30x annual limit for pension+disability 2026 = PLN 282,600.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll. Rate applied to gross rather than the statutory gross-minus-ZUS base because the field shape cannot express a net-of-contributions base; overstates by roughly 1.2pp of gross, the conservative direction, and mirrors computePolandTaxes. The 2025 bill cutting entrepreneurs health contributions was VETOED in May 2025 and is not in force." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "GUS National Health Account household out-of-pocket + gov.pl free-drug programme for under-18s, which is list- and indication-specific rather than unconditional + published Polish private clinic visit pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "NFZ care is free at the point of use, so this is prescription co-payments plus the private outpatient care a professional household actually buys. An employer medical package (Medicover, LuxMed) is a PREMIUM and is not booked here. Poland has no statutory annual ceiling. The previously unexplained Krakow discount is removed; the Warsaw tier is kept and is consistent with Warsaw private pricing running above other Polish metros. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "GUS National Health Account household out-of-pocket + gov.pl free-drug programme for under-18s, which is list- and indication-specific rather than unconditional + published Polish private clinic visit pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Warszawskie Badanie Ruchu 2015, Etap III raport, section 4.5. https://transport.um.warszawa.pl/documents/62470/8560273/WBR_2015._Etap_III._Raport._Wersja_06_2016.pdf", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Motywacja dom-praca (home to work trip motivation)", "year": "2015", "note": "One way, mode-blind, work-only, Warsaw city. The home-to-work leg is 38; the evening praca-dom return is 42. The home-to-work leg is used throughout this field, matching the ACS convention." }, "non_cash.sick_leave": { "source": "Art. 92 Kodeks pracy (Labor Code, employer-paid `wynagrodzenie chorobowe`) + Ustawa z dnia 25.06.1999 o świadczeniach pieniężnych z ubezpieczenia społecznego w razie choroby i macierzyństwa (\"ustawa zasiłkowa\" , ZUS-paid `zasiłek chorobowy`), Art. 4 (qualifying period), Art. 8 (182-day benefit-perio; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-C.md", "source_url": "https://poradnikprzedsiebiorcy.pl/-okres-wyczekiwania-na-prawo-do-zasilku-chorobowego", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Ministerstwo Rodziny, Pracy i Polityki Spolecznej, IKZE limit wplat https://www.gov.pl/web/rodzina/ikze-limit-wplat (Obwieszczenie 10 Nov 2025, M.P. 2025 poz. 1156)", "last_updated": "2026-08-11", "confidence": "high", "note": "IKZE contributions reduce dochod, the PIT base. The deduction limit equals the annual payment limit, 1.2 x the forecast average monthly wage (1.2 x 9,420 = PLN 11,304 for 2026). Employee-only. IKE is not deductible and PPK employee contributions come out of net pay with neither deduction nor credit. PLN 16,956 is the SELF-EMPLOYED limit and must not be used here. The 2025 figure was PLN 10,407.60." } }, "krakow": { "tax.inbound_regime": { "source": "Ustawa o zryczaltowanym podatku dochodowym, consolidated text Dz.U. 2025 poz. 843 (https://api.sejm.gov.pl/eli/acts/DU/2025/843/text.pdf). Art. 12 ust. 1 pkt 2b sets 12% for software-related services (PKWiU ex 62.01.1, 62.01.2, ex 62.02, 62.03.1, ex 62.09.20.0, 62.02.10.0 and the 58.29 software-publishing groupings); art. 6 ust. 4 pkt 1 lit. a caps prior-year revenue at EUR 2,000,000, converted at the NBP rate of the first business day of October (table 190/A/NBP/2025, EUR 4.2586), so the 2026 limit is PLN 8,517,200. Art. 8 ust. 2 voids the regime for the whole year if the same services are supplied to a current or former employer. Ulga na powrot: art. 21 ust. 1 pkt 152 PIT, Dz.U. 2026 poz. 592.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: confidence upgraded medium -> high, and the source moved from a PwC summary to the Sejm ELI statute text. FRAMING FLAG RAISED, NOT RESOLVED HERE: the 12% ryczalt is NOT an inbound regime. It is a general self-employment option available to a Polish local on day one and unavailable to anyone taxed as an employee, and it taxes REVENUE with no cost deduction, no PLN 30,000 allowance and no joint filing. Poland has no inbound income-tax relief for a newly arrived foreign employee at all; the nearest thing is ulga na powrot (PLN 85,528/yr for 4 years, conditions in art. 21 ust. 43; no schema field for tax.ulga_na_powrot exists). Carrying it under an inbound-regime toggle therefore offers our ICP, a relocating EMPLOYEE, a discount the statute does not give them. Recommend either reclassifying the toggle as a B2B-status scenario or dropping it; not done here because it is a product decision, and the brief for this pass said 'upgrade or flag'. The regime note now says all of this in the reader's own words. The old note's claim that the model UNDERSTATES the benefit (real ryczalt gets Maly ZUS Plus and revenue-bracket NFZ) is true only for someone already self-employed; for an employee it overstates by the whole amount. 2026 changes recorded: revenue limit PLN 8,517,200 (down on a weaker EUR); NFZ base PLN 9,228.64 (M.P. 2026 poz. 117), so a 300k+ contractor pays PLN 1,495.04/mo; Maly ZUS Plus capped at 36 months in any 60 from 2026-01-01." }, "tax.payroll_deductible": { "source": "Batch B base-deduction convention (tax audit 2026-07-14, verified): employee ZUS is deductible from the PIT base, so the brackets run on gross minus ZUS. The NFZ health contribution (9%) is NOT deductible (financed from net income since the 2022 reform). computePolandTaxes, gated by this flag (bracket path only; the ryczalt flat regime forgoes it). Source: PwC PL individual/other-taxes.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.federal_brackets": { "source": "podatki.gov.pl PIT 2026 (same as Warsaw)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_rate": { "source": "ZUS / getsix: Social Insurance (ZUS) contributions in Poland 2026 (employee emerytalne 9.76% + rentowe 1.50% = 11.26%, capped at the annual 30x limit). https://getsix.eu/human-resources-payroll-in-poland/social-insurance-institution-zus-contributions-in-poland-in-2026-current-bases-limits-and-contribution-amounts-for-entrepreneurs/", "last_updated": "2026-07-26", "note": "PL-1 (closeout 2026-07-26): capped ZUS stream (pension + disability), split from the former blended 13.71%." }, "tax.social_security_cap": { "source": "ZUS: annual 30x limit (trzydziestokrotność) for pension+disability contributions, 2026 = PLN 282,600. https://getsix.eu/human-resources-payroll-in-poland/social-insurance-institution-zus-contributions-in-poland-in-2026-current-bases-limits-and-contribution-amounts-for-entrepreneurs/", "last_updated": "2026-07-26", "note": "PL-1 (closeout 2026-07-26): 2026 pension+disability cap, up from PLN 260,190 (2025)." }, "tax.health_insurance_rate": { "source": "NFZ 9% uncapped", "last_updated": "2026-05-16", "confidence": "high" }, "tax.sales_tax_rate": { "source": "VAT 23%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Podatek od nieruchomości ~0.1%", "last_updated": "2026-05-16", "confidence": "medium" }, "housing._summary": { "source": "Otodom.pl Kraków + Numbeo Q1 2026", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "NFZ universal", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "Żłobki + przedszkole same federal framework as Warsaw", "last_updated": "2026-05-16", "confidence": "medium" }, "food._summary": { "source": "Numbeo Kraków 2026 + GUS", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "MPK Kraków bilet miesięczny PLN 138", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "GUS + Numbeo Kraków", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "KRK-SEA/LAX (1-stop) 2025-26", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Rodzina 800+ same federal as Warsaw", "last_updated": "2026-05-16", "confidence": "high", "note": "Polish 800+ (Rodzina 800 plus, PLN 800/mo): universal, NOT means-tested." }, "non_cash.parental_leave_weeks": { "source": "Ministerstwo Rodziny, gov.pl urlop rodzicielski: https://www.gov.pl/web/rodzina/urlop-rodzicielski ; ZUS: https://www.zus.pl/swiadczenia/zasilki/zasilek-macierzynski/wysokosc . National statute, peer-identical with Warsaw.", "last_updated": "2026-08-06", "confidence": "high", "note": "20 weeks urlop macierzynski plus 41 weeks urlop rodzicielski = 61 weeks for the household. Corrected 2026-08-01 from 20, which was the maternity block alone. See the Warsaw entry. Re-confirmed on the household-inclusive basis 2026-08-06 (the 20+41 construction already sums the per-parent 9-week reserves inside the shared 41; both-parents flag stays off per the Sweden precedent). See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "ZUS zasilek macierzynski, 70% for the urlop rodzicielski period: https://www.zus.pl/swiadczenia/zasilki/zasilek-macierzynski/wysokosc . National statute, peer-identical with Warsaw.", "last_updated": "2026-08-06", "confidence": "high", "note": "Blended 20 weeks at 100% plus 41 at 70% = 79.8%, stored as 80. See the Warsaw entry. Cap check closed 2026-08-06: zasilek macierzynski is UNCAPPED for employees (art. 36 basis is the plain 12-month average wage; the art. 46 100%-of-average-wage cap applies only after employment ends; the 250% figure is the voluntary contribution-base cap for the self-employed), so the stored blend is real for an ICP, one of the few uncapped schemes." }, "non_cash._summary": { "source": "Kodeks pracy same federal as Warsaw", "last_updated": "2026-05-16", "confidence": "high", "note": "Same B2B/ryczałt flag as Warsaw — v1 toggle approximation understates real benefit by ~5-7pp because Mały ZUS Plus + revenue-bracket NFZ aren't modeled" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Kraków: PLN 260,107, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "PLN 260,107 / 3.74 = USD 69,547; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2782, R5-rounded down to 0.27. REPLACES 0.32 (2026-08-11), and the interim 0.26 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.32 was 0.8889 of Warsaw's 0.36; that ratio x Warsaw's re-derived 0.3 = 0.2667, R5-rounded down to 0.26. Warsaw's own figure: PLN 281,628 / 3.74 = USD 75,302, / USD 250,000 = 0.3012. This is a BASIS move, not a re-measurement: Kraków's rank inside Poland is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Massive outsourcing hub. Index 0.32" }, "tax.social_security_rate_uncapped": { "source": "ZUS / getsix: Poland 2026 employee ZUS: chorobowe (sickness) 2.45% is NOT subject to the 30x annual limit. https://getsix.eu/human-resources-payroll-in-poland/social-insurance-institution-zus-contributions-in-poland-in-2026-current-bases-limits-and-contribution-amounts-for-entrepreneurs/", "last_updated": "2026-07-26", "note": "PL-1 (closeout 2026-07-26): chorobowe 2.45%, UNCAPPED — continues on income above the 30x cap." }, "tax.filing_basis": { "source": "PwC Worldwide Tax Summaries: Poland, Individual: Tax administration (\"Married taxpayers ... may, under some conditions, choose ... jointly or separately\"); separate assessment is the default and joint (wspólne rozliczenie) is an optional election. https://taxsummaries.pwc.com/poland/individual/tax-administration", "last_updated": "2026-07-26", "note": "Closeout follow-on: separate spousal assessment is Poland’s default; joint filing is an election. Fixes the two-earner overstatement the PL-1 ZUS fix left (it had stayed joint-stacked)." }, "healthcare.statutory_employee": { "source": "NFZ / ustawa o swiadczeniach opieki zdrowotnej finansowanych ze srodkow publicznych art. 79 (9% rate) and art. 81 ust. 5 (30x ceiling does not apply to health). ZUS 30x annual limit for pension+disability 2026 = PLN 282,600.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll. Rate applied to gross rather than the statutory gross-minus-ZUS base because the field shape cannot express a net-of-contributions base; overstates by roughly 1.2pp of gross, the conservative direction, and mirrors computePolandTaxes. The 2025 bill cutting entrepreneurs health contributions was VETOED in May 2025 and is not in force." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "GUS National Health Account household out-of-pocket + gov.pl free-drug programme for under-18s, which is list- and indication-specific rather than unconditional + published Polish private clinic visit pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "NFZ care is free at the point of use, so this is prescription co-payments plus the private outpatient care a professional household actually buys. An employer medical package (Medicover, LuxMed) is a PREMIUM and is not booked here. Poland has no statutory annual ceiling. The previously unexplained Krakow discount is removed; the Warsaw tier is kept and is consistent with Warsaw private pricing running above other Polish metros. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "GUS National Health Account household out-of-pocket + gov.pl free-drug programme for under-18s, which is list- and indication-specific rather than unconditional + published Polish private clinic visit pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Kompleksowe Badania Ruchu Krakow 2023. https://www.bip.krakow.pl/plik.php?zid=457396&wer=0&new=t&mode=shw", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Sredni czas trwania podrozy z domu do pracy", "year": "2023", "note": "One way, mode-blind, work-only, Krakow Metropolitan Area (city plus 35 gminas). Largest upward correction in the sweep." }, "non_cash.sick_leave": { "source": "Art. 92 Kodeks pracy (Labor Code, employer-paid `wynagrodzenie chorobowe`) + Ustawa z dnia 25.06.1999 o świadczeniach pieniężnych z ubezpieczenia społecznego w razie choroby i macierzyństwa (\"ustawa zasiłkowa\" , ZUS-paid `zasiłek chorobowy`), Art. 4 (qualifying period), Art. 8 (182-day benefit-perio; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-C.md", "source_url": "https://poradnikprzedsiebiorcy.pl/-okres-wyczekiwania-na-prawo-do-zasilku-chorobowego", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Ministerstwo Rodziny, Pracy i Polityki Spolecznej, IKZE limit wplat https://www.gov.pl/web/rodzina/ikze-limit-wplat (Obwieszczenie 10 Nov 2025, M.P. 2025 poz. 1156)", "last_updated": "2026-08-11", "confidence": "high", "note": "IKZE contributions reduce dochod, the PIT base. The deduction limit equals the annual payment limit, 1.2 x the forecast average monthly wage (1.2 x 9,420 = PLN 11,304 for 2026). Employee-only. IKE is not deductible and PPK employee contributions come out of net pay with neither deduction nor credit. PLN 16,956 is the SELF-EMPLOYED limit and must not be used here. The 2025 figure was PLN 10,407.60." } }, "prague": { "non_cash.parental_leave_flat": { "source": "Urad prace CR, Rodicovsky prispevek, rules stated as platne od ledna 2026 https://www.uradprace.cz/rodicovsky-prispevek ; redukcni hranice 2026 https://www.cssz.gov.cz/web/cz/podrobne-informace-o-nemocenskem", "last_updated": "2026-08-01", "confidence": "high", "note": "A fixed cash POT of CZK 350,000 per child born from 1 January 2024, drawable until the child turns 3 at up to max(CZK 15,000, 0.7 x 30 x DVZ) per month, with the rate changeable every 3 months. For an ICP the ceiling is 0.7 x 30 x the maximum reduced DVZ of CZK 2,857 = CZK 59,997 a month, so the pot is exhausted in 5.83 months, which is 25.37 weeks at CZK 13,798 a week. The CZK 15,000 floor applies only to parents with no or low DVZ and is not our case. Drawdown speed trades weeks against the weekly amount at constant total cash, so the full-pay-equivalent value is invariant; the maximum drawdown is stored because it yields the fewest weeks. EXPRESSLY NOT MEANS-TESTED (tato davka neni vyplacena v zavislosti na prijmu rodiny), so our high earner receives the full pot, and drawing it is compatible with working." }, "non_cash.parental_leave_paid_pct": { "source": "CSSZ: PPM and otcovska paid at 70% redukovaneho denniho vymerovaciho zakladu https://www.cssz.gov.cz/web/cz/podrobne-informace-o-penezite-pomoci-v-materstvi", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved the 70%-paid PPM + otcovska weeks into non_cash.parental_leave_capped and the rodicovsky prispevek weeks stay in non_cash.parental_leave_flat, so no weeks remain on this field. Was 70 with the redukcni hranice ignored per the old field convention." }, "non_cash.parental_leave_weeks": { "source": "CSSZ Penezita pomoc v materstvi https://www.cssz.gov.cz/penezita-pomoc-v-materstvi and https://www.cssz.gov.cz/web/cz/podrobne-informace-o-penezite-pomoci-v-materstvi ; CSSZ Otcovska https://www.cssz.gov.cz/otcovska ; Urad prace CR Rodicovsky prispevek https://www.uradprace.cz/rodicovsky-prispevek", "last_updated": "2026-08-01", "confidence": "high", "note": "55 wk = penezita pomoc v materstvi 28 (mother) + otcovska 2 (14 days, father) + rodicovsky prispevek 25.4. Was 28, maternity alone, omitting otcovska and the entire CZK 350,000 parental allowance. The first two are non-transferable and held by different parents, so they sum. The rodicovsky prispevek is a SHARED family pot attaching to the youngest child, so it is counted ONCE, and it cannot run while PPM is being paid, so the three components are sequential with no overlap to net off. Its 25.4 weeks are the pot divided by the maximum monthly drawdown, which is the fewest-weeks reading and what a high earner would actually do. Excluded: the 37-week multiples variant. HOST WARNING: cssz.cz redirects to cssz.gov.cz, and mpsv.gov.cz/rodicovsky-prispevek is stale (last updated 7 November 2019, still quoting CZK 300,000 and age 4), so cite uradprace.cz." }, "tax.filing_basis": { "source": "Joint taxation of spouses abolished 2008 (flat-tax reform); spouses file individually. Dostupny advokat; Wikipedia Taxation in the Czech Republic.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Finanční správa ČR PIT 2026 (15% to 36× avg wage, 23% above)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_rate": { "source": "ČSSZ sociální pojištění zaměstnanec 7.10%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_cap": { "source": "ČSSZ maximální vyměřovací základ 48× průměrná mzda 2026 ~CZK 2,234,000", "last_updated": "2026-05-16", "confidence": "high" }, "tax.health_insurance_rate": { "source": "VZP zdravotní pojištění zaměstnanec 4.5% UNCAPPED", "last_updated": "2026-05-16", "confidence": "high" }, "tax.sales_tax_rate": { "source": "DPH základní sazba 21%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Daň z nemovitých věcí ~0.1% effective", "last_updated": "2026-05-16", "confidence": "medium" }, "housing._summary": { "source": "Sreality.cz + Bezrealitky Praha Q1 2026 + Numbeo", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "VZP/ZP universal via 4.5% health payroll; private supp negligible", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "Mateřská školka free 3-6 (povinný předškolní rok); jesle 0-3 paid", "last_updated": "2026-05-16", "confidence": "medium" }, "food._summary": { "source": "Numbeo Praha 2026 + ČSÚ", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "DPP měsíční kupon CZK 550 (Lítačka)", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "ČSÚ + Numbeo Praha entertainment", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "PRG-SEA/LAX (1-stop via FRA/AMS/LHR) 2025-26", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Přídavek na dítě means-tested; typical tech family rarely qualifies but daňové zvýhodnění na dítě (~CZK 15k/yr first kid) acts as benefit", "last_updated": "2026-05-16", "confidence": "medium" }, "non_cash._summary": { "source": "Zákoník práce 20 dní + 13 svátků + mateřská 28 týdnů PPM ~70%", "last_updated": "2026-05-16", "confidence": "high", "note": "V1 omits sleva na poplatníka (~CZK 30,840), daňové zvýhodnění na dítě, sleva na manžela → overstates tax by ~CZK 30-60k/yr for typical filer" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Prague: CZK 1,472,806, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-europe.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "CZK 1,472,806 / 20.99 = USD 70,167; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2807, R5-rounded down to 0.28. REPLACES 0.38 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Strong expat tech scene. Index 0.38" }, "healthcare.statutory_employee": { "source": "VZP (vzp.cz) - zdravotni pojisteni, zamestnanec 4.5%, no maximum assessment base since 2013; the 48x average-wage maximum applies to socialni pojisteni under CSSZ only (zakon 592/1992 Sb.).", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll. Uncapped is the load-bearing fact: it is why Czech health does not flatten for our ICP. Cap-abolition history is secondary-sourced; the live 2026 uncapped fact is primary." }, "non_cash.parental_leave_capped": { "source": "CSSZ, redukcni hranice platne od 1.1.2026: 1 633 / 2 449 / 4 897 Kc per day https://www.cssz.gov.cz/web/cz/podrobne-informace-o-nemocenskem ; CSSZ PPM page (70% of the reduced DVZ, first band counted at 100% for PPM and otcovska) https://www.cssz.gov.cz/web/cz/podrobne-informace-o-penezite-pomoci-v-materstvi", "last_updated": "2026-08-06", "confidence": "high", "note": "Payout ceiling 70% x the maximum reduced DVZ, where the reduced DVZ tops out at 1,633 + 0.6 x (2,449 - 1,633) + 0.3 x (4,897 - 2,449) = CZK 2,857/day, so 0.70 x 2,857 = CZK 1,999.90, stored as CZK 2,000/day (CZK 14,000/wk). weeks is 29.63, the stored 55-week total minus the 25.37 flat weeks: the statutory 28 PPM + 2 otcovska = 30 wk are clipped by 0.37 wk to fit the stored total, a rounding remainder in the conservative direction. Confidence high on the redukcni hranice and the 70% rate, medium-high on the derived day-to-week ceiling (first-pass derivation 2026-08-06). Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Czechia)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Regulatory fee CZK 90 for emergency only + the prescription co-payment annual protective limit (ochranny limit), which is set lower for under-18s and over-65s", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 raised from 5,000 to 6,000 on the batch-D research. The regulatory fee is CZK 90 and applies to emergency only; the real burden is prescription and device co-payments, which run against the ochranny limit of CZK 5,000 for a working-age adult, CZK 1,000 for under-18s and 65-69s and CZK 500 from age 70. Children are therefore partially protected rather than exempt, and the curve flattens without going flat. The ochranny limit is recorded for the queued ceiling field; note it is scoped to drugs and devices, not to all cost-sharing. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Regulatory fee CZK 90 for emergency only + the prescription co-payment annual protective limit (ochranny limit), which is set lower for under-18s and over-65s", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Czech Statistical Office, Scitani lidu domu a bytu 2011, table 702. https://csu.gov.cz/docs/107508/cd3222b8-9f22-18a9-18f9-af981aa5a178/2200013702.pdf", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Doba cesty do zamestnani (travel time to employment), employed persons block", "year": "2011", "note": "Band-midpoint weighted mean: Praha 31.04, Praha-vychod 35.32, Praha-zapad 36.55, giving a metro figure of 32.00. One way as published. Treated as a FLOOR because the band-midpoint estimator understates by 2 to 3 minutes. The 2021 census DROPPED the travel-time question, so 2011 is the most recent Czech data in existence." }, "non_cash.sick_leave": { "source": "Zákoník práce (Labor Code, Zákon č. 262/2006 Sb.) §192 (náhrada mzdy, employer-paid, first 14 calendar days) + Zákon č. 187/2006 Sb. o nemocenském pojištění (Sickness Insurance Act; nemocenské, ČSSZ-paid from day 15; podpůrčí doba 380-day ceiling); reduction thresholds (redukční hranice) for 2026 se; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-C.md", "source_url": "https://www.jenprace.cz/uzitecne-informace/vypocet-nemocenske", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Financni sprava, Informace k povinnemu prispevku na produkty na stari https://financnisprava.gov.cz/cs/dane/dane/dan-z-prijmu/zamestnanci-zamestnavatele/informace-stanoviska-sdeleni/2026/informace-k-povinnemu-prispevku-na-produkty-na-stari (ZDP 586/1992 s.15a/s.15b)", "last_updated": "2026-08-11", "confidence": "medium", "note": "CZK 48,000/yr combined across doplnkove penzijni sporeni, penzijni pripojisteni, DIP and qualifying life insurance, deducted from the base. For DPS / penzijni pripojisteni only the slice ABOVE CZK 1,700/month (20,400/yr) is deductible because that slice attracts the statni prispevek instead, but DIP and life insurance carry no such threshold so the full 48,000 is reachable. The 20,400 threshold is deliberately NOT modeled with less_employer_pct: that would lower the CAP rather than the CONTRIBUTION, overstating the deduction at low contributions and understating it at high ones. The statni prispevek and the separate CZK 50,000 employer-contribution exemption are different rules." } }, "vienna": { "tax.payroll_deductible": { "source": "Batch B base-deduction convention (tax audit 2026-07-14, verified): Austrian Sozialversicherung is a pre-tax deduction, so Einkommensteuer runs on gross minus SV (computeAustriaTaxes, gated by this flag). Source: PwC AT individual/deductions.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.federal_brackets": { "source": "BMF Einkommensteuertarif 2026 (7 brackets, indexed kalte-Progression law)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_rate": { "source": "ÖGK SV Beitragssätze ~18.07% (pension 10.25% + health 3.87% + UV 2.95% + chamber/housing ~1%)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_cap": { "source": "WKO Voraussichtliche Werte in der Sozialversicherung 2026: Höchstbeitragsgrundlage EUR 6,930/mo x 12 + EUR 13,860 special-payment base = EUR 97,020/yr. https://www.wko.at/sozialversicherung/voraussichtliche-werte-in-der-sozialversicherung-2026", "last_updated": "2026-07-26", "note": "AT-3 (closeout 2026-07-26): 2026 annual maximum contribution basis, up from EUR 90,300 (2025)." }, "tax.sales_tax_rate": { "source": "USt Normalsatz 20%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Grundsteuer ~0.1% effective (trivial in AT vs other EU)", "last_updated": "2026-05-16", "confidence": "high" }, "housing._summary": { "source": "ImmoScout24 + Statistik Austria Wien Q1 2026 + Numbeo", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "ÖGK universal funded via SV employee 18.07%; premium = 0 (already in payroll)", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "Wien Gratiskindergarten 0+; only meals contribution. Tagesmutter for under-2 mixed.", "last_updated": "2026-05-16", "confidence": "high" }, "food._summary": { "source": "Numbeo Wien 2026 + Statistik Austria", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "Wiener Linien Jahreskarte €467/yr card (€461 digital) = €38.92/mo, price effective 2026-01-01 (wienerlinien.at/jahreskarte); conservative card price per the value-bias rule. Was €365. Flagged by reporter 816f, 2026-08-06 corrections wave (key 86a0552b)", "last_updated": "2026-08-06", "confidence": "high" }, "discretionary._summary": { "source": "Statistik Austria + Numbeo Wien", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "VIE-SEA/LAX (1-stop via FRA/MUC/LHR) Austrian/Lufthansa 2025-26", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Familienbeihilfe + Familienbonus Plus combined ~€2,500-3,500/yr/kid for typical tech-worker household", "last_updated": "2026-05-16", "confidence": "high", "note": "Austrian Familienbeihilfe: universal, NOT means-tested." }, "non_cash.parental_leave_weeks": { "source": "Bundeskanzleramt / oesterreich.gv.at, einkommensabhaengiges Kinderbetreuungsgeld: https://www.bundeskanzleramt.gv.at/agenda/familie/kinderbetreuungsgeld/basisinformationen-kinderbetreuungsgeld/einkommensabhaengiges-kinderbetreuungsgeld.html; Wochengeld (MSchG s3, ASVG s162) via oesterreich.gv.at + OeGK; Familienzeitbonus (FamZeitbG) via oesterreich.gv.at + WKO/AK", "last_updated": "2026-08-07", "confidence": "high", "note": "73 wk = 8 pre-birth Wochengeld weeks (MSchG: 8 weeks before the due date at ~100% of average net, uncapped for employees; counted per the house basis, consistent with Denmark 4 / Norway 3 / Finland ~6.5 pre-birth weeks; the 2026-08-01 exclusion of these weeks was an inconsistency, corrected in batch 5) + 426 days (60.9 -> 61 wk) of the shared eaKBG window from the birth when both parents claim (365 one parent alone, 61 days reserved each; INCLUSIVE of post-birth Wochengeld via KBGG s6 Ruhen) + 28 days (4.0 wk) Familienzeitbonus, the father's own additive entitlement: its Anrechnung against his later KBG was ABOLISHED for births from 2023-01-01 (parliamentary record + WKO + AK convergent, RIS unreachable 2026-08-07), so the prior exclusion ground was outdated law; 28-day floor of the 28-31 day range is the conservative pick. The flat-rate KBG-Konto variant (fixed sum over 365-851 days) stays excluded: we carry the income-dependent 80% variant, the professional-household default. both_parents flag NOT set: shared pot + short papamonat (Sweden precedent). Household-inclusive basis (2026-08-01); batch-5 restatement 2026-08-07, research: docs/superpowers/research/2026-08-07-pl-b5-austria-israel.md. Was 61 (pre-birth Wochengeld and FZB dropped). User correction flag:01784883616070-0a448b8e." }, "non_cash.parental_leave_paid_pct": { "source": "oesterreich.gv.at Wochengeld (MSchG s3, ASVG s162): average net of the last 3 months plus special-payments surcharge, no ceiling for employees", "last_updated": "2026-08-07", "confidence": "high", "note": "100 covers ONLY the 8 pre-birth Wochengeld weeks (uncapped ~100% of net for employees). The 61 eaKBG weeks live in non_cash.parental_leave_capped (80% to EUR 80.12/day) and the 4 Familienzeitbonus weeks in non_cash.parental_leave_flat. Was 0 when all stored weeks sat in the capped block (2026-08-06); batch 5 added the pre-birth weeks 2026-08-07." }, "non_cash.parental_leave_flat": { "source": "Familienzeitbonusgesetz (FamZeitbG): EUR 54.87/day for 2026 (valorization suspended 2026-27), 28-31 days within the first 91 days of the child's life, via oesterreich.gv.at + WKO/AK (RIS 503'd 2026-08-07); Anrechnung against the father's later KBG abolished for births from 2023-01-01", "last_updated": "2026-08-07", "confidence": "high", "note": "Papamonat: the father's own statutory paid month, additive since the 2023 abolition of the KBG offset. Stored at the 28-day conservative floor = 4.0 wk at EUR 54.87 x 7 = EUR 384.09/wk. At EUR 130,050 ($150k basis) that is ~15.4% of the father's weekly wage. Research: docs/superpowers/research/2026-08-07-pl-b5-austria-israel.md." }, "non_cash._summary": { "source": "ArbVG 5 Wochen Urlaub + 13 Feiertage; Wochengeld 16 Wochen 100% then einkommensabhaengiges Kinderbetreuungsgeld to day 426", "last_updated": "2026-05-16", "confidence": "high", "note": "13./14. Gehalt (sonstige Bezüge) IS modeled, via tax.special_payment_brackets (a real progressive §67 scale topping at 50%, not a flat 6% as this note used to claim) inside computeAustriaTaxes. Corrected 2026-08-06: the note predated the AT-1 closeout fix and was left stale after tax.special_payment_brackets's own entry was updated to describe the shipped behaviour. Verified by running compute() at EUR 150,000 gross with and without special_payment_brackets: tax moves from EUR 59,400 to EUR 67,764, an EUR 8,364 swing. The only remaining unmodeled piece is the EUR 2,615 Freigrenze low-earner edge case, already documented on tax.special_payment_brackets's own entry. Found by tests/meta-claims.check.mjs's category-3 informational list, spot-checked by hand, 2026-08-06." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Vienna: EUR 70,495, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-europe.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 70,495 / 0.867 = USD 81,309; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3252, R5-rounded down to 0.32. REPLACES 0.48 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Vienna tops global livability rankings. Index 0.48" }, "childcare.subsidy_per_kid_annual": { "source": "Stadt Wien beitragsfreier Kindergarten; double-subtraction correction 2026-08-02", "last_updated": "2026-08-02", "confidence": "high", "note": "Set to 0 on 2026-08-02, was EUR 2,000. The old value was the value of Vienna's free municipal Kindergarten place, which is a free place rather than a transfer, and childcare.preschool_monthly is already 0 for the covered ages. This field holds CASH THAT REACHES THE HOUSEHOLD. Where the monthly rate it nets against is already the price the household actually pays, the value is 0 and the relief is described here instead. See docs/superpowers/research/2026-08-02-childcare-double-subtraction.md. Open follow-up: Vienna's preschool_monthly of 0 is itself too low, because the Gratiskindergarten still charges an Essensbeitrag of about EUR 72.33/month." }, "tax.filing_basis": { "source": "PwC Worldwide Tax Summaries: Austria, Individual: Tax administration (no joint assessment; spouses assessed individually). https://taxsummaries.pwc.com/austria/individual/tax-administration", "last_updated": "2026-07-26", "note": "AT-2 (closeout 2026-07-26): Austria has no joint assessment; spouses always assessed individually." }, "tax.special_payment_fraction": { "source": "Austrian 14-payment salary convention (12 monthly + Urlaubsgeld + Weihnachtsgeld); the two special payments are 2/14 of annual pay and exactly fill the Jahressechstel. WKO Lohnverrechnung / BMF Steuerbuch 2026.", "last_updated": "2026-07-26", "note": "AT-1 (closeout follow-on): 2/14 of pay is taxed as sonstige Bezüge, not at the progressive tariff." }, "tax.special_payment_brackets": { "source": "WKO \"Sonstige Bezüge - steuerliche Behandlung\" (§67 EStG): within the Jahressechstel, the first EUR 620 is tax-free, then 6% (to 25,000), 27% (to 50,000), 35.75% (to 83,333), and per-tariff (~50%) above. https://www.wko.at/lohnverrechnung/sonstige-bezuege-steuerliche-behandlung", "last_updated": "2026-07-26", "note": "AT-1 (closeout follow-on): §67 preferential scale for the 13th/14th salary. The EUR 2,615 Freigrenze (full exemption for tiny Jahressechstel) is a low-earner edge case the engine skips; an ICP Jahressechstel sits far above it." }, "healthcare.statutory_employee": { "source": "OEGK contribution tables (Krankenversicherung, employee 3.87%); sozialversicherung.gv.at 2026 Hoechstbeitragsgrundlage EUR 6,930/month for laufende Bezuege plus a separate EUR 13,860 annual cap on Sonderzahlungen = EUR 97,020; ASVG section 45 confirms the ceiling covers health; wko.at contribution table dated 28.01.2026. Vienna Wohnbaufoerderungsbeitrag increase to 1.50% split evenly: Wiener Landesgesetzblatt Nr. 73/2025.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll. The cap is legally two sub-caps (monthly on laufende Bezuege, annual on Sonderzahlungen) whose sum we carry as one annual_cap. WBF and AK are not levied on Sonderzahlungen, so the blended employee total at the ceiling is 17.93% nationally and 18.14% in Vienna rather than a flat 18.07%; that affects the payroll total, not this health rate. The 3.40% Zusatzbeitrag for a co-insured adult is deliberately excluded: it is waived while that person raises a child, which is our ICP." }, "childcare.preschool_monthly": { "source": "https://www.wien.gv.at/bildung/kindergarten-staedtisch-kosten-oeffnungszeiten", "last_updated": "2026-08-05", "confidence": "high", "note": "Gratiskindergarten charges a flat EUR 94/mo Essensbeitrag (effective 2026-01-01; was 84.85 from 2024-09, 72.33 in 2023-24). 12-month year, no income scaling at ICP incomes (Befreiung only below ~EUR 2,019 net/mo household). Same meal-fee class as Berlin 60." }, "non_cash.parental_leave_capped": { "source": "Bundeskanzleramt: einkommensabhaengiges Kinderbetreuungsgeld, \"80 Prozent der Letzteinkuenfte\", \"maximal 80,12 Euro taeglich (rund 2.400 Euro monatlich)\", re-confirmed 2026-08-06 https://www.bundeskanzleramt.gv.at/agenda/familie/kinderbetreuungsgeld/basisinformationen-kinderbetreuungsgeld/einkommensabhaengiges-kinderbetreuungsgeld.html", "last_updated": "2026-08-06", "confidence": "high", "note": "Payout ceiling EUR 80.12/day (EUR 560.84/wk) on the 61-week eaKBG window, re-confirmed for 2026 on 2026-08-07 (LexisNexis records the same value 2025-2027). Single conservative blend within the window: the ~8 POST-birth Wochengeld weeks inside it are genuinely uncapped employer-topped net pay, and folding them under the ceiling reads them LOW, the direction less favorable to Austria. The 8 PRE-birth Wochengeld weeks are additive calendar time and live on parental_leave_paid_pct=100 since batch 5 (2026-08-07). No income-eligibility cliff exists on the income-dependent variant (a Zuverdienstgrenze limits earnings WHILE drawing, not eligibility). Floor: Sonderleistung EUR 41.14/day, immaterial at ICP salaries and not stored. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md + 2026-08-07-pl-b5-austria-israel.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "STATISTIK AUSTRIA Gesundheitsausgaben 2024, EUR 8,980m direct household OOP on the SHA 2011 basis https://www.statistik.at/statistiken/bevoelkerung-und-soziales/gesundheit/gesundheitsversorgung-und-ausgaben/gesundheitsausgaben + Konsumerhebung 2024/25 + OeGK Rezeptgebuehr", "last_updated": "2026-08-08", "confidence": "medium", "note": "The largest single level correction in the sweep. Two independent routes agree on ~EUR 980-1,000 per capita gross of dental and long-term care; the in-scope share is ~53%, giving EUR 520-710, and the conservative pick is 550. The old EUR 250 looks like a statutory-Rezeptgebuehr-scale figure used as total household out-of-pocket, in the one country where statutory cost-sharing is the SMALLEST part of what households pay: indirect cost-sharing is 13.5% of system financing against 7.6% for direct. Austria also exempts under-18s from the hospital per-diem but NOT from the Rezeptgebuehr, so its children cost more than elsewhere in Europe. Weakest input is the in-scope share. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "STATISTIK AUSTRIA Gesundheitsausgaben 2024, EUR 8,980m direct household OOP on the SHA 2011 basis https://www.statistik.at/statistiken/bevoelkerung-und-soziales/gesundheit/gesundheitsversorgung-und-ausgaben/gesundheitsausgaben + Konsumerhebung 2024/25 + OeGK Rezeptgebuehr", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "Value RETAINED at 30 and CORROBORATED rather than sourced. A midpoint reconstruction of Vienna-resident one-way bands from Statistik Austria Volkszaehlung 2001, published by Arbeiterkammer Wien (https://www.arbeiterkammer.at/infopool/wien/Verkehr_und_Infrastruktur_25.pdf), gives 30.3, which agrees with the stored 30. That is treated as corroboration and NOT as the source of record, because the basis year is 2001 and a quarter-century-old figure carrying a live URL would read as verified while saying nothing about commuting today. Eurostat urb_ctran TT1019V has ZERO Austrian rows in any year. The Statistik Austria headline of 24 min was REJECTED as off-unit: it counts only gemeindeuberschreitende (cross-municipality) commuters, excluding nearly everyone who lives and works inside Vienna, and it models the time from the register rather than measuring it." }, "non_cash.sick_leave": { "source": "Angestelltengesetz (AngG) §8 for Angestellte / Austria's own Entgeltfortzahlungsgesetz; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "BMF, Sonderausgaben im Einzelnen https://www.bmf.gv.at/themen/steuern/arbeitnehmerveranlagung/was-kann-ich-geltend-machen/sonderausgaben/sonderausgaben-im-einzelnen.html ; WKO, praemienbeguenstigte Zukunftsvorsorge https://www.wko.at/steuern/praemienbeguenstigte-zukunftsvorsorge", "last_updated": "2026-08-11", "confidence": "high", "note": "No deduction exists. Topf-Sonderausgaben (private Pensionsversicherung, Pensionskassenbeitraege) ran out with the pre-2016-contract transition window on 31.12.2020. Praemienbeguenstigte Zukunftsvorsorge is a PREMIUM SUBSIDY (4.25% on max EUR 3,817.04 = EUR 162.22 for 2026), and the BMF is explicit it is not a Sonderausgabe. The one uncapped carve-out, freiwillige Weiterversicherung and Nachkauf von Versicherungszeiten, buys STATE pension credit for gap years rather than being a recurring private contribution, so it is not this field." } }, "brussels": { "childcare.free_tier_start_age": { "source": "Eurydice (European Commission), Belgium - Flemish Community, Early childhood education and care: \"From age 2-and-a-half, children are legally entitled to pre-primary education (ISCED 020)\" https://eurydice.eacea.ec.europa.eu/national-education-systems/belgium-flemish-community/early-childhood-education-and-care", "confidence": "high", "note": "Statutory entitlement starts at 2 years 6 months. Kid ages in the model are whole numbers, so the boundary is set to 3, the conservative rounding: the whole of age 2 keeps billing the paid kinderopvang rate rather than the free kleuteronderwijs rate. Before this, ages 1-5 all billed the free-tier rate, so a 1-year-old in income-tariff kinderopvang was charged nothing. Rates themselves are unchanged and still owe a review: the 2025 inkomenstarief ceiling is EUR 35.89/day (about EUR 718-790/month) for a top-income family, above all three modelled infant rates, and the free-tier residual is not zero either (statutory maximumfactuur EUR 55/year plus school meals at roughly EUR 76-95/month). See docs/superpowers/research/2026-08-01-childcare-free-tier-boundary.md.", "last_updated": "2026-08-01" }, "tax.payroll_deductible": { "source": "Batch B base-deduction convention (tax audit 2026-07-14, verified): the IPP brackets apply to income net of the deductible employee RSZ (13.07%), so the ladder runs on gross minus RSZ (computeBelgiumTaxes, gated by this flag); RSZ itself stays on full gross. Source: PwC BE individual/deductions.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.federal_brackets": { "source": "SPF Finances IPP 2026 (4 brackets indexed)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.communal_surcharge_rate": { "source": "Brussels-Capital communes additionnels 7.5% (Brussels-City; range 0-9%)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.inbound_regime": { "source": "Bijzonder belastingstelsel voor ingekomen belastingplichtigen (BBIB / RSII). Wet-loi van 18 december 2025 houdende diverse bepalingen, arts. 15-18, Belgisch Staatsblad 30-12-2025, applicable to remuneration paid or allocated from 1 January 2025: expense exclusion raised 30% -> 35% of gross remuneration, the ~EUR 90k tax cap abolished (it survives for social security), and the minimum gross remuneration cut from EUR 75,000 to more than EUR 70,000. Circulaire 2026/C/51 of 1 April 2026 (Administration generale de la Fiscalite) on the retroactive application, contract-adjustment deadline 30 June 2026. Cross-checked Claeys & Engels, EY Belgium and BDO.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31 OVER-GRANT FIX. This entry already said the floor had been lowered to ~EUR 70k, but no min_income was declared, so a 35% exclusion applied at ANY salary. Wired min_income 70000, the exact statutory figure. Brussels single filer at EUR 69,999: tax 16,808 -> 29,015 (ordinary). OPEN QUESTION, not resolved here: the allowance is legally paid bovenop de bezoldiging (on top of gross remuneration), so reduction_pct 0.35 sits between the contract-gross reading (allowance is extra tax-free cash, we understate the break) and the total-package reading (taxable share is 1/1.35 = 74.07%, true carve-out 25.93%, we overstate it). Left at 0.35 pending a modeling decision." }, "tax.social_security_rate": { "source": "ONSS/RSZ employee general scheme 13.07% (uncapped)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_cap": { "source": "BE general scheme has NO employee cap", "last_updated": "2026-05-16", "confidence": "high" }, "tax.sales_tax_rate": { "source": "BTW/TVA standard 21%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Bruxelles Fiscalité précompte immobilier (~1.2% market-value effective via cadastral income)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Real précompte is % of cadastral income; 1.2% is market-value approximation" }, "housing._summary": { "source": "Immoweb + Statbel Brussels-Capital Q1 2026 + Numbeo", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "INAMI universal via RSZ + mutuelle Mutualité Chrétienne/Solidaris", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "ONE crèche tariff income-based; FAMIRIS supplements", "last_updated": "2026-05-16", "confidence": "medium" }, "food._summary": { "source": "Numbeo Brussels 2026 + Statbel", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "STIB-MIVB abonnement mensuel €49", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "Statbel + Numbeo Brussels", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "BRU-SEA/LAX (1-stop via AMS/LHR) Brussels Airlines/Lufthansa 2025-26", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "FAMIRIS Brussels-Capital ~€175-220/mo/child", "last_updated": "2026-05-16", "confidence": "high", "note": "Belgian Groeipakket / allocations familiales: universal (regional), NOT means-tested (social supplements are extra)." }, "non_cash._summary": { "source": "BE labor code 20 jours + 10 jours fériés + maternité 15 sem 82% + paternité 20 jours", "last_updated": "2026-05-16", "confidence": "high", "note": "Régime des cadres étrangers (expat regime) IS modeled, via tax.inbound_regime (engine toggle): 35% expense exclusion, 5 years, hard EUR 70,000 gross floor below which it does not apply. Corrected 2026-08-06 from \"NOT modeled in v1\", which was stale and predated the toggle being wired; same defect class as Rome's pre-2026-08-02 note. Verified by running compute() at EUR 105,000 gross with apply_inbound_regime off and on: Brussels tax moved by EUR 19,753 in the check's reference run (EUR 28,219 at a flat EUR 150,000 gross). Found by tests/meta-claims.check.mjs's toggle test, 2026-08-06." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Brussels: EUR 53,895, fetched live 2026-08-11 and recorded at sweep B F8c of docs/superpowers/research/2026-08-11-comp-index-sweep-b.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "EUR 53,895 / 0.867 = USD 62,163; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2487, R5-rounded down to 0.24. REPLACES 0.5 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Brussels by 108%. The measured p50 was 0.497 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): EU-funded comp adds floor. Index 0.50 CONFIDENCE LOW, and the reason is recorded rather than smoothed over: levels.fyi ranks Brussels (EUR 53,895) BELOW Antwerp (66,830), Ghent (62,057) and Leuven (58,586), which sweep B F8c calls implausible for the capital and EU-institution hub and attributes to sample composition (Brussels senior p25-p75 spread 2.5x against Helsinki 1.6x). All three values come from that one source on one basis, so the country is internally consistent, but its ORDERING is the weakest in the 2026-08-11 re-derivation." }, "childcare.subsidy_per_kid_annual": { "source": "Kind en Gezin / ONE income-related crEche tariff (inkomenstarief); double-subtraction correction 2026-08-02", "last_updated": "2026-08-02", "confidence": "medium", "note": "Set to 0 on 2026-08-02, was EUR 1,200. The old note described it as \"the subsidy vs the uncapped rate\", but childcare.infant_monthly already carries the CAPPED income-related fee, so the cap was applied twice. The Flemish peers antwerp and ghent already carried 0, which is the pattern Brussels should have matched. This field holds CASH THAT REACHES THE HOUSEHOLD. Where the monthly rate it nets against is already the price the household actually pays, the value is 0 and the relief is described here instead. See docs/superpowers/research/2026-08-02-childcare-double-subtraction.md." }, "tax.filing_basis": { "source": "PwC Worldwide Tax Summaries: Belgium, Individual: Sample personal income tax calculation (spouses taxed separately on all earned income / décumul). https://taxsummaries.pwc.com/belgium/individual/sample-personal-income-tax-calculation", "last_updated": "2026-07-26", "note": "BE-3 (closeout 2026-07-26): spouses taxed separately (décumul) though filing one joint return." }, "tax.professional_expense_rate": { "source": "PwC Worldwide Tax Summaries: Belgium, Individual: Deductions (lump-sum professional expenses = 30% of gross taxable income for employees). https://taxsummaries.pwc.com/belgium/individual/deductions", "last_updated": "2026-07-26", "note": "BE-1 (closeout 2026-07-26): 30% professional-expense forfait rate." }, "tax.professional_expense_cap": { "source": "PwC Worldwide Tax Summaries: Belgium, Individual: Deductions (30% forfait capped at EUR 5,930, AY2026/IY2025). https://taxsummaries.pwc.com/belgium/individual/deductions", "last_updated": "2026-07-26", "note": "BE-1 (closeout 2026-07-26): forfait cap EUR 5,930 (confirmed current cap; IY2026 indexation not yet published. TODO (when it publishes): apply the small uplift; the cap stays EUR 5,930 until then)." }, "tax.tax_free_allowance": { "source": "PwC BE: Belgian tax reform 2026 (quotité exemptée / belastingvrije som = EUR 11,180 for income year 2026, up from EUR 10,910). https://news.pwc.be/belgian-tax-reform-2026-tax-and-social-measures-of-the-programme-law/", "last_updated": "2026-07-26", "note": "BE-2 (closeout 2026-07-26): tax-free allowance for income year 2026." }, "tax.tax_free_allowance_credit_rate": { "source": "PwC BE: quotité exemptée credited at the bottom-bracket 25% rate (the allowance falls entirely within the first IPP band). https://news.pwc.be/belgian-tax-reform-2026-tax-and-social-measures-of-the-programme-law/", "last_updated": "2026-07-26", "note": "BE-2 (closeout 2026-07-26): credited at 25% → reduction ~EUR 2,795." }, "non_cash.parental_leave_weeks": { "source": "belgium.be (conge de maternite, 15 semaines); INAMI maternity amounts https://www.inami.fgov.be/fr/themes/grossesse-et-naissance/montants-et-plafonds-des-indemnites/salaries-chomeurs/montant-maximal-de-votre-indemnite-pendant-un-repos-de-maternite-salariee ; ONEM conges thematiques montants https://www.onem.be/documentation/montants/interruption-de-carriere-credit-temps/conges-thematiques", "last_updated": "2026-08-06", "confidence": "high", "note": "54 wk = repos de maternite 15 (6 prenatal of which 1 mandatory, 9 postnatal mandatory) + conge de naissance 20 days (4 wk, the other parent's own right) + conge parental 4 months PER PARENT via ONEM until the child is 12 (34.78 wk summed across both parents, which is why parental_leave_both_parents is true; Netherlands/Ireland precedent). Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. Was 15/82, maternity alone with the first-30-days rate applied to everything." }, "non_cash.parental_leave_paid_pct": { "source": "INAMI: 82% of uncapped gross for the first 30 days, then 75% of salary capped at EUR 186.79/day from 1 Mar 2026 (max EUR 140.09/day, 6 payable days/wk); conge de naissance 3 days at 100% employer-paid then 17 days at 82% capped", "last_updated": "2026-08-06", "confidence": "high", "note": "79 = blend across the 19.2 percentage-paid weeks: maternity 4.29 wk at 82 (uncapped) + 10.71 wk at 75 (capped, max EUR 840.54/wk) + naissance 0.6 wk at 100 + 3.4 wk at 82 (capped): (352 + 803 + 60 + 279) / 19.0 = 79. The EUR 186.79/day ceiling sits near EUR 58,000/yr, so the capped phases pay their daily maxima for an ICP." }, "non_cash.parental_leave_flat": { "source": "ONEM conges thematiques montants, index 01.03.2026: EUR 1,058.86 gross a month (EUR 951.60 net after the 11.11% withholding) for full-time interruption https://www.onem.be/documentation/montants/interruption-de-carriere-credit-temps/conges-thematiques", "last_updated": "2026-08-06", "confidence": "high", "note": "EUR 243.53/wk = 1,058.86 / 4.348. Four months per parent, so 8 months = 34.78 wk for the household. A flat sum regardless of salary, roughly 15-19% of an ICP wage, which is exactly what the flat block exists to price honestly." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "INAMI/RIZIV maximum a facturer 2026 income bands (top band EUR 2,194.20 per household, child sub-cap EUR 792.35) + Statbel household health expenditure", "last_updated": "2026-08-08", "confidence": "medium", "note": "The old EUR 450 was neither a maximum nor a sourced average, just a round number that landed near the right single-adult figure and was then applied flat to households. Belgium has no child exemption, so the curve scales with headcount, shaded down slightly because the maximum a facturer is a HOUSEHOLD ceiling that does not grow with family size. Note it caps tickets moderateurs only, not fee or room supplements, which are the larger half of Belgian out-of-pocket spend. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "INAMI/RIZIV maximum a facturer 2026 income bands (top band EUR 2,194.20 per household, child sub-cap EUR 792.35) + Statbel household health expenditure", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Same Belgian gap as Antwerp. The Eurostat Urban Audit value for BE001C is 33.0 on a 1991 basis and was deliberately rejected for the same reason: adopting it would have moved this city by -5 on thirty-five-year-old evidence. OVG Vlaanderen 7 (2023-24) was CHECKED and REJECTED: its duration figures are DAY-LEVEL and ALL-PURPOSE (22.3 min, Flanders), which is the same unit error as SrV, so adopting it would have swapped one defect for another. FORWARD POINTER: OVG 8 is the source to watch, if it publishes duration by motive." }, "non_cash.sick_leave": { "source": "Arbeidsovereenkomstenwet (Employment Contracts Act, 3 July 1978) art. 52; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.ejustice.just.fgov.be", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "SPF Finances, Epargne-pension https://fin.belgium.be/fr/particuliers/avantages-fiscaux/epargne-pension", "last_updated": "2026-08-11", "confidence": "high", "note": "No base deduction. Epargne-pension is a REDUCTION D'IMPOT applied after the tax is computed and never touches the taxable base: 30% on up to EUR 1,050 (max EUR 315) or 25% on up to EUR 1,350 (max EUR 337.50) for income year 2026. Epargne a long terme is also a credit (30%, ceiling 6% of net professional income plus EUR 183.6, absolute max EUR 2,450). Occupational 2nd-pillar (groepsverzekering) employee contributions DO reduce taxable pay but are scheme-specific and we model no employer scheme. computeBelgiumTaxes already nets the mandatory RSZ 13.07% separately, so nothing here double-counts." } }, "lisbon": { "tax.inbound_regime": { "source": "IFICI, art. 58-A EBF (2024 successor to the NHR), current redacao Lei n.o 82/2023 de 29 de dezembro: 'tributado, em sede de IRS, a taxa especial de 20 % sobre os rendimentos liquidos das categorias A e B ... durante um prazo de 10 anos'. Qualifying activities set by Portaria n.o 352/2024/1 (2024-12-23) as amended by Portaria n.o 52-A/2025/1. 5-year prior non-residency; registration by 15 January of the year after becoming resident; former NHR and art. 12-A CIRS users excluded. https://info.portaldasfinancas.gov.pt/pt/informacao_fiscal/codigos_tributarios/bf_rep/Pages/EBF58A.aspx ; https://info.portaldasfinancas.gov.pt/pt/apoio_contribuinte/questoes_frequentes/pages/faqs-01018.aspx", "last_updated": "2026-07-31", "confidence": "medium", "note": "2026-07-31: the 20% and the 10 years are verified primary and unchanged. Confidence is medium for ELIGIBILITY, not for the rate. OPEN DECISION RAISED, NOT TAKEN HERE: every one of the seven qualifying limbs in art. 58-A n.o 1 gates on the EMPLOYER being certified or recognised (FCT, AICEP, an exporter of 50%+ of turnover, SIFIDE II via ANI, or a Startup Portugal certified startup), so a software engineer at an ordinary Portuguese company does not qualify. That is the same shape as Ireland's SARP, which we carry as note-only. Applying 20% by default flatters Portugal by roughly the gap between 20% and a 35-45% effective IRS rate. Recommend demoting to note-only or gating behind an explicit opt-in; not done in this pass because it changes calculator math and would break tests/integration/tax-correct-first-portugal.test.mjs, which is outside this pass's file ownership. The note now states the employer gate plainly. NOT SOURCED PRIMARILY: the Portaria's detailed profession list (diariodarepublica.pt rendered empty HTML and unparseable PDF); no OE2026 amendment found, but its absence could not be affirmatively ruled out." }, "tax.filing_basis": { "source": "PwC WWTS - Portugal (Tax administration): separate assessment is the default since the 2015 reform; joint filing is elective, so individual is the two-earner default. Existing doubled married brackets remain for single-earner (elective-joint) households.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "IRS 2026 bracket scale (taxsummaries.pwc.com/portugal); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.social_security_rate": { "source": "Segurança Social TSU trabalhador 11% (uncapped)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_cap": { "source": "PT TSU has NO cap on employee contribution", "last_updated": "2026-05-16", "confidence": "high" }, "tax.sales_tax_rate": { "source": "IVA taxa normal 23% continente (DRE)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Câmara Municipal de Lisboa IMI urbano 0.30% (2026)", "last_updated": "2026-05-16", "confidence": "high" }, "housing._summary": { "source": "Idealista PT + Imovirtual Lisboa Q1 2026 + Numbeo", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Imovirtual citywide listing harvest 2026-08-17 (all search pages, __NEXT_DATA__ JSON, apartments for rent, reverseGeocoding council = lisboa/lisboa, deduped by ad id): within-listing P25/P75 of asking EUR/month per unit at fixed tipologia. T2 anchor: 1,497.5/1,875 = 0.799 and 2,300/1,875 = 1.227 (n=956), rounded outward to 0.79/1.23. T1: 0.833/1.233 (n=814); T3: 0.808/1.310 (n=475).", "last_updated": "2026-08-17", "confidence": "medium", "note": "REBASED 2026-08-17 (ledger cluster housing-0817) from the 0.80/1.33 INE per-m2 cross-freguesia min/max band, which failed the constant-ratio check with Madrid's exact signature: the per-unit Santo Antonio/Santa Clara asking ratio is 1.76 at T1, 2.36 at T2, 2.50 at T3 against the stored constant 1.66, i.e. a pooled per-m2 freguesia ratio silently encodes the ~1BR ratio (INE's own release confirms per-m2 medians are highest for T0/T1 in every large municipality). ALSO RESOLVED: the old derivation's citywide denominator 16.0 EUR/m2 appears in NO cited INE release (Lisboa municipio 12-month median is 17.22, quarterly 17.42; same-release arithmetic gives 0.75/1.24) - most plausibly a pre-revision-series value, uncitable since INE restated the whole series 2026-06-26; moot because the basis is replaced, not re-derived. The new pair is a genuine within-distribution percentile at fixed bedroom count - the semantics the field name claims - and renders on the listing basis (rentBandBasis infers listing). NOT chosen: an honest area-basis T2 band (Santo Antonio 1.38 / Santa Clara 0.59 vs citywide median) - min/max of freguesia medians is the pre-smoothed extreme-of-averages construction the 2026-08-13 ledger rejects, and its cheap anchor is an n=5 cell. Known residual: single-pair schema anchored at T2; measured T3 top is 1.31, so 3BR households see a ~+0.08 understated top edge (conservative direction noted, not hidden). Studios untestable (invalid t0 filter). Level observation, out of scope: measured citywide T2 asking median 1,875 vs stored median 1,750 (+7% asking drift), T1 1,500 vs 1,300. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Derivation + per-tipologia table: docs/superpowers/research/2026-08-17-housing-band-fixes.md (L1); scoping evidence: docs/superpowers/research/2026-08-16-housing-band-scope-lisbon.md" }, "healthcare._summary": { "source": "SNS universal + Multicare/Médis private complementary", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "Pré-escolar gratuito 3+; creche/infantário tariff varies; Vale Infância PT", "last_updated": "2026-05-16", "confidence": "medium" }, "food._summary": { "source": "Numbeo Lisboa 2026 + INE", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "Carris/Metro Navegante Mensal €40 (subsidized 2023+)", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "INE + Numbeo Lisboa", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "LIS-SEA/LAX (1-stop via LHR/AMS/CDG) TAP/AA 2025-26", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Abono de família para crianças e jovens (Segurança Social, means-tested)", "last_updated": "2026-05-16", "confidence": "medium" }, "non_cash._summary": { "source": "Código do Trabalho 22 dias + 13 feriados + licença parental 120-150 dias 100%", "last_updated": "2026-05-16", "confidence": "high", "note": "IFICI regime: qualifying inbound R&D/tech workers get 20% flat IRS for 10 yrs (replaced NHR in 2024). CARRIED but DELIBERATELY NOT APPLIED to take-home: tax.inbound_regime holds type portugal_ifici with salary_relief false, so the engine surfaces the regime as a note and the tax number does not move. Verified 2026-08-02: EUR 70,530 at EUR 150,000 gross with the toggle both off and on. The reason is in the regime block’s own note, that every qualifying route gates on your EMPLOYER rather than on you and we cannot test that. Wording tightened 2026-08-02 from \"NOT modeled in v1 engine\", which read as an unfinished task rather than a decision; do not wire a flat rate here without revisiting it. Major draw for US-tech ICP." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Lisbon: EUR 49,547, held in labor_market.tech_salary_distribution (source_slug lisbon-metro-area, as_of 2026-07-18, table confidence thin). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 49,547 / 0.867 = USD 57,148; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2286, R5-rounded down to 0.22. REPLACES 0.34 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Lisbon by 55%. The measured p50 was 0.672 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Top of PT market. Index 0.34" }, "tax.federal_brackets_married": { "source": "Conjugal quotient (coefficient 2): married IRS = 2 x tax(income/2), reproduced by doubling every single-bracket edge; PwC Tax Summaries Portugal; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "Doubled-edge married brackets = the Portuguese conjugal quotient for a two-earner couple." }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-21", "note": "Zeroed: the Portuguese abono de familia is means-tested and is 0 for a high-earner household; there is no universal 3-6 cash subsidy." }, "childcare.preschool_monthly": { "source": "childcare accuracy pass 2026-07-21", "last_updated": "2026-07-21", "note": "Pre-escolar (3-6) tuition is free (public + IPSS), but Lisbon/Porto capacity is scarce (waitlists), so a relocating family typically pays private (~EUR 500/mo). Research 2026-07-21; softest number in the set." }, "tax.specific_deduction_min": { "source": "Autoridade Tributaria / CIRS art. 25: deducao especifica da Categoria A: a maior de 8,54 x IAS ou as contribuicoes obrigatorias para a Seguranca Social. IAS 2026 = 537,13 EUR, logo o piso = 4.587,09 EUR.", "source_name": "CIRS art. 25 (AT)", "last_updated": "2026-07-23", "confidence": "high", "note": "Floor only. The deduction actually applied is max(this floor, employee SS) and is UNCAPPED, so above ~41,700 EUR gross the 11% SS figure binds instead. Applied in computePortugalTaxes (correct-first item #2 of the 2026-07-17 resolutions doc). Re-verify the IAS each January." }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/lisbon-metro-area", "last_updated": "2026-07-18", "confidence": "thin", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "EF English Proficiency Index 2024 (Very High band)", "last_updated": "2024", "note": "National-level English usability for tech work. Band: very_high." }, "non_cash.parental_leave_weeks": { "source": "gov.pt, Having a child: Parental leave and allowance https://www2.gov.pt/en/guias/ter-uma-crianca/licenca-parental ; Codigo do Trabalho arts. 39-65 as amended by Lei 13/2023; DL 91/2009", "last_updated": "2026-08-06", "confidence": "medium-high", "note": "48 wk = licenca parental inicial 120 days at 100% (17.14 wk, a SHARED pot counted once; the mother 42 mandatory post-birth days sit inside it) + father exclusive 28 mandatory + 7 optional days at 100% (5 wk) + licenca parental alargada 3 months PER PARENT at 30% (26.09 wk, summed across both parents, which is why parental_leave_both_parents is true; Italy precedent). Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. Excluded: the 30-day sharing bonus (conditional on both parents each taking 30 days) and the 150-day/80% election (equal total cash, more weeks). Owner call 2026-08-05: primary representation is 48/62; the alternative WITHOUT the alargada tail is 22 wk at 100, recorded here for the next audit rather than stored. Was 17/100, the initial pot alone. Confidence medium-high on the alargada subsidy conditions (gov.pt says usable anytime until the child turns six; historical rules required it to follow the initial leave)." }, "non_cash.parental_leave_paid_pct": { "source": "gov.pt https://www2.gov.pt/en/guias/ter-uma-crianca/licenca-parental (initial leave 120 days at 100%; father days always paid at 100%; alargada at 30% of reference salary; subsidy floor EUR 13.93/day, no ceiling identified)", "last_updated": "2026-08-06", "confidence": "medium-high", "note": "62 = duration-weighted blend of 22.14 wk at 100 (initial pot + father days; no ceiling identified on the subsidy, floor only, so the 100 is real even for an ICP, one of the few uncapped schemes) and 26.09 wk at 30 (alargada; the older 25 is superseded, and the 40 variant requires both parents to take the full three months so 30 is used): (2214 + 783) / 48.2 = 62." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Taxas moderadoras largely abolished for primary and hospital care in 2022, leaving unreferred emergency capped at EUR 40 per episode (Portaria 64-C/2016) + INE household budget survey and the OECD Portugal country profile", "last_updated": "2026-08-08", "confidence": "medium", "note": "The value rises despite the fee abolition, because the abolished category was already a minor share and the population-wide realized average from two converging sources sits above the old 250/230 even before adjusting for a working-age household that pays privately to skip queues. No annual ceiling exists. The existing ~6% Lisbon-over-Porto tier is preserved. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Taxas moderadoras largely abolished for primary and hospital care in 2022, leaving unreferred emergency capped at EUR 40 per episode (Portaria 64-C/2016) + INE household budget survey and the OECD Portugal country profile", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, indicator TT1019V, greater city. https://ec.europa.eu/eurostat/api/dissemination/statistics/1.0/data/urb_ctran?indic_ur=TT1019V&format=JSON&lang=EN", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average time of journey to work", "year": "2011", "note": "Urban Audit 33.24 for the greater city. INE Portugal's Censos 2021 all-modes figure for the wider Area Metropolitana de Lisboa is lower, about 25." }, "non_cash.sick_leave": { "source": "Subsídio de doença, general regime , Decreto-Lei n.º 28/2004 (regime jurídico de; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Ordem dos Contabilistas Certificados, IRS: PPR e deducao a coleta https://occ.pt/pt-pt/noticias/irs-ppr-e-deducao-coleta (quoting EBF art. 21 n.2) ; EBF https://info.portaldasfinancas.gov.pt/pt/informacao_fiscal/codigos_tributarios/Cod_download/Documents/EBF.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "No base deduction. PPR relief is a DEDUCAO A COLETA (EBF art. 21 n.2), a credit of 20% of the amount paid in, capped by age on 1 January at EUR 400 / 350 / 300, and it competes inside the global limit on deducoes a coleta (CIRS art. 78 n.7) shared with health and education. Nothing reduces rendimento coletavel: the Category A deducao especifica (CIRS art. 25) is max(8.54 x IAS, actual SS) plus union dues and nothing else. Lisbon's entire current relief was spurious." } }, "porto": { "tax.inbound_regime": { "source": "IFICI, art. 58-A EBF (2024 successor to the NHR), current redacao Lei n.o 82/2023 de 29 de dezembro: 'tributado, em sede de IRS, a taxa especial de 20 % sobre os rendimentos liquidos das categorias A e B ... durante um prazo de 10 anos'. Qualifying activities set by Portaria n.o 352/2024/1 (2024-12-23) as amended by Portaria n.o 52-A/2025/1. 5-year prior non-residency; registration by 15 January of the year after becoming resident; former NHR and art. 12-A CIRS users excluded. https://info.portaldasfinancas.gov.pt/pt/informacao_fiscal/codigos_tributarios/bf_rep/Pages/EBF58A.aspx ; https://info.portaldasfinancas.gov.pt/pt/apoio_contribuinte/questoes_frequentes/pages/faqs-01018.aspx", "last_updated": "2026-07-31", "confidence": "medium", "note": "2026-07-31: the 20% and the 10 years are verified primary and unchanged. Confidence is medium for ELIGIBILITY, not for the rate. OPEN DECISION RAISED, NOT TAKEN HERE: every one of the seven qualifying limbs in art. 58-A n.o 1 gates on the EMPLOYER being certified or recognised (FCT, AICEP, an exporter of 50%+ of turnover, SIFIDE II via ANI, or a Startup Portugal certified startup), so a software engineer at an ordinary Portuguese company does not qualify. That is the same shape as Ireland's SARP, which we carry as note-only. Applying 20% by default flatters Portugal by roughly the gap between 20% and a 35-45% effective IRS rate. Recommend demoting to note-only or gating behind an explicit opt-in; not done in this pass because it changes calculator math and would break tests/integration/tax-correct-first-portugal.test.mjs, which is outside this pass's file ownership. The note now states the employer gate plainly. NOT SOURCED PRIMARILY: the Portaria's detailed profession list (diariodarepublica.pt rendered empty HTML and unparseable PDF); no OE2026 amendment found, but its absence could not be affirmatively ruled out." }, "tax.filing_basis": { "source": "PwC WWTS - Portugal (Tax administration): separate assessment is the default since the 2015 reform; joint filing is elective, so individual is the two-earner default. Existing doubled married brackets remain for single-earner (elective-joint) households.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "IRS 2026 bracket scale (taxsummaries.pwc.com/portugal); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.social_security_rate": { "source": "Segurança Social TSU trabalhador 11%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_cap": { "source": "PT TSU uncapped", "last_updated": "2026-05-16", "confidence": "high" }, "tax.sales_tax_rate": { "source": "IVA 23% continente", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Câmara Municipal do Porto IMI urbano 0.45% (top of band)", "last_updated": "2026-05-16", "confidence": "high" }, "housing._summary": { "source": "Idealista PT + Imovirtual Porto Q1 2026 + Numbeo", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "SNS universal + Multicare/Médis", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "Pré-escolar gratuito 3+; creche tariff varies", "last_updated": "2026-05-16", "confidence": "medium" }, "food._summary": { "source": "Numbeo Porto 2026 + INE", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "STCP Andante mensal €40", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "INE + Numbeo Porto", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "OPO-SEA/LAX 1-stop via LIS/LHR 2025-26", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Abono de família Segurança Social", "last_updated": "2026-05-16", "confidence": "medium" }, "non_cash._summary": { "source": "Código do Trabalho + same federal labor law as Lisbon", "last_updated": "2026-05-16", "confidence": "high", "note": "IFICI 20% flat IRS regime, same as Lisbon. CARRIED but DELIBERATELY NOT APPLIED to take-home: tax.inbound_regime holds type portugal_ifici with salary_relief false, so the regime is surfaced as a note and the tax number does not move. Verified 2026-08-02: EUR 70,530 at EUR 150,000 gross with the toggle both off and on. Wording tightened 2026-08-02 from \"not modeled\", which read as an unfinished task rather than a decision. See the lisbon entry and the regime block’s own note." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Porto: EUR 42,890, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 42,890 / 0.867 = USD 49,469; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1979, R5-rounded down to 0.19. REPLACES 0.3 (2026-08-11), and the interim 0.19 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.3 was 0.8824 of Lisbon's 0.34; that ratio x Lisbon's re-derived 0.22 = 0.1941, R5-rounded down to 0.19. Lisbon's own figure: EUR 49,547 / 0.867 = USD 57,148, / USD 250,000 = 0.2286. This is a BASIS move, not a re-measurement: Porto's rank inside Portugal is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Below Lisbon. Index 0.30" }, "tax.federal_brackets_married": { "source": "Conjugal quotient (coefficient 2): married IRS = 2 x tax(income/2), reproduced by doubling every single-bracket edge; PwC Tax Summaries Portugal; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "Doubled-edge married brackets = the Portuguese conjugal quotient for a two-earner couple." }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-21", "note": "Zeroed: the Portuguese abono de familia is means-tested and is 0 for a high-earner household; there is no universal 3-6 cash subsidy." }, "childcare.preschool_monthly": { "source": "childcare accuracy pass 2026-07-21", "last_updated": "2026-07-21", "note": "Pre-escolar (3-6) tuition is free (public + IPSS), but Lisbon/Porto capacity is scarce (waitlists), so a relocating family typically pays private (~EUR 450/mo). Research 2026-07-21; softest number in the set." }, "tax.specific_deduction_min": { "source": "Autoridade Tributaria / CIRS art. 25: deducao especifica da Categoria A: a maior de 8,54 x IAS ou as contribuicoes obrigatorias para a Seguranca Social. IAS 2026 = 537,13 EUR, logo o piso = 4.587,09 EUR.", "source_name": "CIRS art. 25 (AT)", "last_updated": "2026-07-23", "confidence": "high", "note": "Floor only. The deduction actually applied is max(this floor, employee SS) and is UNCAPPED, so above ~41,700 EUR gross the 11% SS figure binds instead. Applied in computePortugalTaxes (correct-first item #2 of the 2026-07-17 resolutions doc). Re-verify the IAS each January." }, "non_cash.parental_leave_weeks": { "source": "gov.pt, Having a child: Parental leave and allowance https://www2.gov.pt/en/guias/ter-uma-crianca/licenca-parental ; Codigo do Trabalho arts. 39-65 as amended by Lei 13/2023; DL 91/2009", "last_updated": "2026-08-06", "confidence": "medium-high", "note": "48 wk = licenca parental inicial 120 days at 100% (17.14 wk, a SHARED pot counted once; the mother 42 mandatory post-birth days sit inside it) + father exclusive 28 mandatory + 7 optional days at 100% (5 wk) + licenca parental alargada 3 months PER PARENT at 30% (26.09 wk, summed across both parents, which is why parental_leave_both_parents is true; Italy precedent). Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. Excluded: the 30-day sharing bonus (conditional on both parents each taking 30 days) and the 150-day/80% election (equal total cash, more weeks). Owner call 2026-08-05: primary representation is 48/62; the alternative WITHOUT the alargada tail is 22 wk at 100, recorded here for the next audit rather than stored. Was 17/100, the initial pot alone. Confidence medium-high on the alargada subsidy conditions (gov.pt says usable anytime until the child turns six; historical rules required it to follow the initial leave)." }, "non_cash.parental_leave_paid_pct": { "source": "gov.pt https://www2.gov.pt/en/guias/ter-uma-crianca/licenca-parental (initial leave 120 days at 100%; father days always paid at 100%; alargada at 30% of reference salary; subsidy floor EUR 13.93/day, no ceiling identified)", "last_updated": "2026-08-06", "confidence": "medium-high", "note": "62 = duration-weighted blend of 22.14 wk at 100 (initial pot + father days; no ceiling identified on the subsidy, floor only, so the 100 is real even for an ICP, one of the few uncapped schemes) and 26.09 wk at 30 (alargada; the older 25 is superseded, and the 40 variant requires both parents to take the full three months so 30 is used): (2214 + 783) / 48.2 = 62." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Taxas moderadoras largely abolished for primary and hospital care in 2022, leaving unreferred emergency capped at EUR 40 per episode (Portaria 64-C/2016) + INE household budget survey and the OECD Portugal country profile", "last_updated": "2026-08-08", "confidence": "medium", "note": "The value rises despite the fee abolition, because the abolished category was already a minor share and the population-wide realized average from two converging sources sits above the old 250/230 even before adjusting for a working-age household that pays privately to skip queues. No annual ceiling exists. The existing ~6% Lisbon-over-Porto tier is preserved. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Taxas moderadoras largely abolished for primary and hospital care in 2022, leaving unreferred emergency capped at EUR 40 per episode (Portaria 64-C/2016) + INE household budget survey and the OECD Portugal country profile", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, indicator TT1019V, greater city. https://ec.europa.eu/eurostat/api/dissemination/statistics/1.0/data/urb_ctran?indic_ur=TT1019V&format=JSON&lang=EN", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average time of journey to work", "year": "2011", "note": "Urban Audit 25.21." }, "non_cash.sick_leave": { "source": "Subsídio de doença, general regime , Decreto-Lei n.º 28/2004 (regime jurídico de; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Ordem dos Contabilistas Certificados, IRS: PPR e deducao a coleta https://occ.pt/pt-pt/noticias/irs-ppr-e-deducao-coleta (quoting EBF art. 21 n.2) ; EBF https://info.portaldasfinancas.gov.pt/pt/informacao_fiscal/codigos_tributarios/Cod_download/Documents/EBF.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "No base deduction. PPR relief is a DEDUCAO A COLETA (EBF art. 21 n.2), a credit of 20% of the amount paid in, capped by age on 1 January at EUR 400 / 350 / 300, and it competes inside the global limit on deducoes a coleta (CIRS art. 78 n.7) shared with health and education. Nothing reduces rendimento coletavel: the Category A deducao especifica (CIRS art. 25) is max(8.54 x IAS, actual SS) plus union dues and nothing else. Lisbon's entire current relief was spurious." } }, "milan": { "childcare.free_tier_start_age": { "source": "ISTAT / Dipartimento per le politiche della famiglia / Universita Ca Foscari, I servizi educativi per l'infanzia in Italia, a.e. 2023/2024, p.39: \"La scuola d'infanzia ... se a titolarita pubblica e sostanzialmente gratuita (fatta eccezione per la quota della mensa)\" https://www.istat.it/wp-content/uploads/2025/05/report_infanzia_2023_2024.pdf", "last_updated": "2026-08-01", "confidence": "high", "note": "Scuola dell'infanzia runs 3-6 and is free in state schools; asilo nido (0-3) is paid. Participation from age 3 to compulsory school is 94.7% (2023, Eurostat via the same report). Before this, ages 1-5 all billed one rate, so the bucket straddled paid nido and free infanzia." }, "non_cash.parental_leave_paid_pct": { "source": "D.Lgs 151/2001 art. 22 c.1 (indennita pari all 80 per cento for congedo di maternita, testo in vigore dal 28-5-2003), art. 29 c.1 (100 per cento for congedo di paternita), art. 34 c.1 (3 months at 80 per cento in alternativa tra i genitori, the remainder at 30 per cento) https://www.normattiva.it/uri-res/N2Ls?urn:nir:stato:decreto.legislativo:2001-03-26;151 ; Legge 207/2024 art. 1 commi 217-218 (the third month at 80%) ; D.Lgs 151/2001 art. 23 (no massimale for dependent employees) https://www.normattiva.it/uri-res/N2Ls?urn:nir:stato:decreto.legislativo:2001-03-26;151", "last_updated": "2026-08-06", "confidence": "high", "note": "60% blended (exact 59.9): 34.7 wk at 80% (maternity plus the 3 uplifted parental months) + 2 wk at 100% (paternity) + 26 wk at 30% (the parental tail). This LOWERS the stored 80, because that figure was the maternity rate applied to maternity weeks only. Congedo di maternita was NOT raised to 100%: art. 22 is untouched since 2003, and the 100% figure in circulation comes from collective agreements, which are excluded. Cap research 2026-08-06: genuinely UNCAPPED for standard dependent employees at every tier. D.Lgs 151/2001 art. 23 defines the indennita off the actual retribuzione media globale giornaliera with no massimale language, and the Legge 335/1995 massimale contributivo (EUR 122,295 for 2026, INPS circolare 47/2026) applies ONLY to Gestione Separata workers, not lavoratori dipendenti, so Italy keeps a plain percentage representation, a positive citable finding rather than a research gap. Separate income-eligibility cliff (EUR 19,885.13/yr, 2.5x minimum pension) on part of the 30% tail is flagged for its own pass; months 10-11 are already excluded from the stored 62. See docs/superpowers/research/2026-08-06-parental-caps-europe.md (Italy)." }, "non_cash.parental_leave_weeks": { "source": "D.Lgs 151/2001 Testo Unico consolidated on Normattiva, testo in vigore dal 1-1-2026 https://www.normattiva.it/uri-res/N2Ls?urn:nir:stato:decreto.legislativo:2001-03-26;151; art. 16 (congedo di maternita, 5 mesi), art. 27-bis (congedo di paternita obbligatorio, 10 giorni lavorativi), art. 32 and 34 (congedo parentale)", "last_updated": "2026-08-01", "confidence": "high", "note": "62 wk = congedo di maternita 21.7 + congedo di paternita obbligatorio 2 + congedo parentale 9 paid months (3 non-transferable per parent plus 3 in alternativa tra loro). Was 22, which was congedo di maternita alone and dropped the mandatory 10 paternity days and all 9 paid parental months. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Excluded: months 10-11 of congedo parentale, which art. 34 c.3 means-tests at 2.5x the minimum pension and our ICP fails; CCNL top-ups." }, "tax.filing_basis": { "source": "PwC WWTS - Italy (Tax administration): Modello Redditi PF cannot be filed jointly; each spouse taxed on own income (the joint Modello 730 does not aggregate income).", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Agenzia delle Entrate IRPEF 2026 (3 scaglioni 23/35/43)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.addizionale_regionale_rate": { "source": "Regione Lombardia addizionale IRPEF 2026 (1.73%)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.addizionale_comunale_rate": { "source": "Comune di Milano addizionale comunale 0.8%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_rate": { "source": "INPS aliquota dipendente privato 9.19%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_cap": { "source": "INPS massimale contributivo 2026 ~€119,650", "last_updated": "2026-05-16", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "IVA ordinaria 22%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "IMU prima casa NON di lusso esente (Legge Bilancio); symbolic 0.1%", "last_updated": "2026-05-16", "confidence": "high", "note": "Real cost ~€0 for primary residents; TARI (waste) not modeled" }, "housing._summary": { "source": "Immobiliare.it + Idealista Milano Q1 2026", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "SSN/Lombardia Welfare universale + Unisalute/Generali private supp.", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "Comune di Milano nido tariffa ISEE-based + Bonus Asilo Nido INPS €3,000", "last_updated": "2026-05-16", "confidence": "medium" }, "food._summary": { "source": "Numbeo Milano 2026 + ISTAT", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "ATM abbonamento mensile €39 (urbano 2026)", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "ISTAT + Numbeo entertainment", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "MXP-SEA/LAX (1-stop via AMS/LHR/CDG) 2025-26", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Assegno Unico Universale INPS (2022+); ~€1,800/yr typical tech-worker household", "last_updated": "2026-05-16", "confidence": "high", "note": "Italian Assegno Unico Universale: means-tested by ISEE with a guaranteed minimum (~EUR 57/mo/kid) paid at any income; modeled value is the high-ISEE floor (survives)." }, "non_cash._summary": { "source": "CCNL + Codice Civile 26 gg ferie + congedo maternità 5 mesi 80% + paternità 10 gg", "last_updated": "2026-05-16", "confidence": "high", "note": "Regime Impatriati is modeled via tax.inbound_regime (engine toggle). Big draw for US-tech ICP." }, "tax.inbound_regime": { "source": "Regime impatriati, art. 5 D.Lgs. 209/2023 (transfers of residence from 2024). Agenzia delle Entrate: 'i redditi ... prodotti in Italia entro il limite annuo di 600.000 euro concorrono alla formazione del reddito complessivo limitatamente al 50%', so the EUR 600,000 is a cap on the exempted base, not a cliff. Five tax periods; 3 prior tax periods of non-Italian residency (6 or 7 if the same employer or group); elevata qualificazione o specializzazione; 4-year residency commitment with clawback plus interest. https://www.agenziaentrate.gov.it/portale/lavoratori-impatriati-209-2023/infogen-lavoratori-impatriati-209-2023 ; normattiva urn:nir:stato:decreto.legislativo:2023-12-27;209~art5", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: values verified UNCHANGED for 2026 (0.50 / 600,000 / 5 years). The only 2026 amendment is art. 22 L. 132/2025 (in force 2025-10-10), which merely widened the qualification test to include applied AI research. Notes harmonised across all five Italian cities: Bologna, Naples and Turin still carried the PRE-2024 condition '2+ yrs prior non-IT residency', which is wrong (the current statute is 3 tax periods, 6 or 7 within the same employer group). Caught by the new tax.inbound_regime entry in data/_national_fields.json. The 60% exclusion for a worker with a minor child is DELIBERATELY left out: it requires the child to reside in Italy for the whole benefit period, which we cannot test, and the conservative tiebreaker takes the lower exclusion. WATCH 2027: normattiva shows art. 5 abrogated by D.Lgs. 117/2026 (Testo unico imposte sui redditi, GU n. 152 of 2026-07-03) with effect 2027-01-01; the regime moves to art. 225. The art. 225 text could not be retrieved, so treat the consolidation as presumptively substance-neutral. TODO (before 2027-01): re-verify the art. 225 text." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Milan: EUR 37,601, held in labor_market.tech_salary_distribution (source_slug milan-metro-area, as_of 2026-07-18, table confidence thin). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 37,601 / 0.867 = USD 43,369; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1735, R5-rounded down to 0.17. REPLACES 0.3 (2026-08-10), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Milan by 76%. The measured p50 was 0.578 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.30. Was 0.42, set 2026-05-16 from the narrative 'top of Italian market' with no cross-country figure recorded. Re-derived on the same basis the peer indices use (levels.fyi senior SWE median total comp, USD, over the USD 220,000 Seattle senior reference): Milan senior median EUR 59,027 = USD 68,081 / 220,000 = 0.3095, rounded DOWN per conservative bias. That basis reproduces Madrid (0.392 vs stored 0.40), Barcelona (0.403 vs 0.38) and Berlin (0.603 vs 0.60) to within 6%; the old 0.42 was 26% out. Milan is EUR 59,027 senior against Madrid's EUR 74,792, so it cannot sit above Madrid's 0.40. All figures are RAL, gross annual, 13th/14th mensilita already inside. Still top of the Italian market. Research: docs/superpowers/research/2026-08-10-milan-comp-index.md" }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-21", "note": "Zeroed for 3-6: the Italian \"bonus asilo nido\" applies to 0-3 nido and is means-tested (only a small floor survives for a high earner); it does not apply to free 3-6 infanzia." }, "childcare.preschool_monthly": { "source": "Comune di Milano, Un posto a tavola 2026/2027 (ristorazione scolastica tariffe), band \"Oltre 27.000,01 o in caso di assenza dati ISEE\" = EUR 734,40 per year https://www.comune.milano.it", "note": "EUR 734.40/YEAR flat, charged a prescindere dalle effettive presenze (regardless of actual attendance), so meal days do not affect it. 734.40/12 = 61.20. Was 550 (paritaria/bilingual tier). Applies the owner decision of 2026-07-18 that Spain and Italy follow the statutory free convention, now that the age boundary makes it applicable to the free ages only. REPLACES the 2026-07-21 modelling, which priced Italy at the paritaria/bilingual private tier on the argument that January-only enrolment and Italian-only instruction push international families private. That argument holds equally for Catalan-medium Barcelona and for France, Belgium and Germany, where the repo already models the free public path, so Italy was the only country priced private. Stored value is the ANNUAL out-of-pocket divided by 12, because computeChildcareCost multiplies by 12; the invoice figure is a school-year figure over ~10 months. Sourcing: docs/superpowers/research/2026-08-01-italy-mensa-fees.md.", "last_updated": "2026-08-01", "confidence": "high" }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/milan-metro-area", "last_updated": "2026-07-18", "confidence": "thin", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "EF English Proficiency Index 2024 (Moderate band)", "last_updated": "2024", "note": "National-level English usability for tech work. Band: moderate." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "National ticket framework on specialist visits and diagnostics plus the pharmaceutical co-payment, both regionally set; esenzione per reddito thresholds sit far below our ICP household", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 raised from 400 to 480 on the batch-D research. The ticket on specialist visits and diagnostics plus the pharmaceutical co-payment are both REGIONALLY set, and the esenzione per reddito income thresholds sit far below our ICP household, so neither the adults nor the children are exempt and the curve follows the no-child-discount shape. Italy has no annual out-of-pocket ceiling. Naples is the fifth Italian city, not Florence. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "National ticket framework on specialist visits and diagnostics plus the pharmaceutical co-payment, both regionally set; esenzione per reddito thresholds sit far below our ICP household", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, indicator TT1019V, greater city. https://ec.europa.eu/eurostat/api/dissemination/statistics/1.0/data/urb_ctran?indic_ur=TT1019V&format=JSON&lang=EN", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average time of journey to work", "year": "2011", "note": "Urban Audit 34.8. Same ISTAT work-or-study caveat as Rome." }, "non_cash.sick_leave": { "source": "INPS indennità di malattia (general private-sector \"industria\" regime); Codice; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Agenzia delle Entrate, Risposta n. 76/2024 https://www.agenziaentrate.gov.it/portale/documents/20143/5946162/Risposta+n.+76_2024.pdf/f0933165-e52e-a519-8ca9-85d3f5162f23 ; Risoluzione n. 25/2025 https://www.agenziaentrate.gov.it/portale/documents/20143/8874453/Risoluzione+n.+25_2025/58473993-686e-9f56-723a-034c8a1dd3f6 (D.Lgs. 252/2005 art. 8 c.4)", "last_updated": "2026-08-11", "confidence": "high", "note": "Previdenza complementare deduction from reddito complessivo, EUR 5,164.57/yr, fixed in statute since 2007 (it is 10 million lire) and not indexed. A COMBINED ceiling: employer contributions to the fondo pensione count against the same figure. TFR conferred to the fund does NOT count and is not deductible, having never been taxed as income. less_employer_pct is deliberately left unset, which is the GENEROUS reading, rather than inventing an employer rate; flagged for a later pass. The \"lavoratori di prima occupazione\" extra platform in years 6-25 is deliberately unmodeled." } }, "rome": { "childcare.free_tier_start_age": { "source": "ISTAT / Dipartimento per le politiche della famiglia / Universita Ca Foscari, I servizi educativi per l'infanzia in Italia, a.e. 2023/2024, p.39: \"La scuola d'infanzia ... se a titolarita pubblica e sostanzialmente gratuita (fatta eccezione per la quota della mensa)\" https://www.istat.it/wp-content/uploads/2025/05/report_infanzia_2023_2024.pdf", "last_updated": "2026-08-01", "confidence": "high", "note": "Scuola dell'infanzia runs 3-6 and is free in state schools; asilo nido (0-3) is paid. Participation from age 3 to compulsory school is 94.7% (2023, Eurostat via the same report). Before this, ages 1-5 all billed one rate, so the bucket straddled paid nido and free infanzia." }, "non_cash.parental_leave_paid_pct": { "source": "D.Lgs 151/2001 art. 22 c.1 (indennita pari all 80 per cento for congedo di maternita, testo in vigore dal 28-5-2003), art. 29 c.1 (100 per cento for congedo di paternita), art. 34 c.1 (3 months at 80 per cento in alternativa tra i genitori, the remainder at 30 per cento) https://www.normattiva.it/uri-res/N2Ls?urn:nir:stato:decreto.legislativo:2001-03-26;151 ; Legge 207/2024 art. 1 commi 217-218 (the third month at 80%) ; D.Lgs 151/2001 art. 23 (no massimale for dependent employees) https://www.normattiva.it/uri-res/N2Ls?urn:nir:stato:decreto.legislativo:2001-03-26;151", "last_updated": "2026-08-06", "confidence": "high", "note": "60% blended (exact 59.9): 34.7 wk at 80% (maternity plus the 3 uplifted parental months) + 2 wk at 100% (paternity) + 26 wk at 30% (the parental tail). This LOWERS the stored 80, because that figure was the maternity rate applied to maternity weeks only. Congedo di maternita was NOT raised to 100%: art. 22 is untouched since 2003, and the 100% figure in circulation comes from collective agreements, which are excluded. Cap research 2026-08-06: genuinely UNCAPPED for standard dependent employees at every tier. D.Lgs 151/2001 art. 23 defines the indennita off the actual retribuzione media globale giornaliera with no massimale language, and the Legge 335/1995 massimale contributivo (EUR 122,295 for 2026, INPS circolare 47/2026) applies ONLY to Gestione Separata workers, not lavoratori dipendenti, so Italy keeps a plain percentage representation, a positive citable finding rather than a research gap. Separate income-eligibility cliff (EUR 19,885.13/yr, 2.5x minimum pension) on part of the 30% tail is flagged for its own pass; months 10-11 are already excluded from the stored 62. See docs/superpowers/research/2026-08-06-parental-caps-europe.md (Italy)." }, "non_cash.parental_leave_weeks": { "source": "D.Lgs 151/2001 Testo Unico consolidated on Normattiva, testo in vigore dal 1-1-2026 https://www.normattiva.it/uri-res/N2Ls?urn:nir:stato:decreto.legislativo:2001-03-26;151; art. 16 (congedo di maternita, 5 mesi), art. 27-bis (congedo di paternita obbligatorio, 10 giorni lavorativi), art. 32 and 34 (congedo parentale)", "last_updated": "2026-08-01", "confidence": "high", "note": "62 wk = congedo di maternita 21.7 + congedo di paternita obbligatorio 2 + congedo parentale 9 paid months (3 non-transferable per parent plus 3 in alternativa tra loro). Was 22, which was congedo di maternita alone and dropped the mandatory 10 paternity days and all 9 paid parental months. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Excluded: months 10-11 of congedo parentale, which art. 34 c.3 means-tests at 2.5x the minimum pension and our ICP fails; CCNL top-ups." }, "tax.filing_basis": { "source": "PwC WWTS - Italy (Tax administration): Modello Redditi PF cannot be filed jointly; each spouse taxed on own income (the joint Modello 730 does not aggregate income).", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Agenzia delle Entrate IRPEF 2026", "last_updated": "2026-05-16", "confidence": "high" }, "tax.addizionale_regionale_rate": { "source": "Regione Lazio addizionale IRPEF 2026 (3.33% top: highest among major IT regions)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.addizionale_comunale_rate": { "source": "Roma Capitale addizionale comunale 0.9%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_rate": { "source": "INPS dipendente 9.19%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_cap": { "source": "INPS massimale 2026 ~€119,650", "last_updated": "2026-05-16", "confidence": "medium" }, "tax.inbound_regime": { "source": "Regime impatriati, art. 5 D.Lgs. 209/2023 (transfers of residence from 2024). Agenzia delle Entrate: 'i redditi ... prodotti in Italia entro il limite annuo di 600.000 euro concorrono alla formazione del reddito complessivo limitatamente al 50%', so the EUR 600,000 is a cap on the exempted base, not a cliff. Five tax periods; 3 prior tax periods of non-Italian residency (6 or 7 if the same employer or group); elevata qualificazione o specializzazione; 4-year residency commitment with clawback plus interest. https://www.agenziaentrate.gov.it/portale/lavoratori-impatriati-209-2023/infogen-lavoratori-impatriati-209-2023 ; normattiva urn:nir:stato:decreto.legislativo:2023-12-27;209~art5", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: values verified UNCHANGED for 2026 (0.50 / 600,000 / 5 years). The only 2026 amendment is art. 22 L. 132/2025 (in force 2025-10-10), which merely widened the qualification test to include applied AI research. Notes harmonised across all five Italian cities: Bologna, Naples and Turin still carried the PRE-2024 condition '2+ yrs prior non-IT residency', which is wrong (the current statute is 3 tax periods, 6 or 7 within the same employer group). Caught by the new tax.inbound_regime entry in data/_national_fields.json. The 60% exclusion for a worker with a minor child is DELIBERATELY left out: it requires the child to reside in Italy for the whole benefit period, which we cannot test, and the conservative tiebreaker takes the lower exclusion. WATCH 2027: normattiva shows art. 5 abrogated by D.Lgs. 117/2026 (Testo unico imposte sui redditi, GU n. 152 of 2026-07-03) with effect 2027-01-01; the regime moves to art. 225. The art. 225 text could not be retrieved, so treat the consolidation as presumptively substance-neutral. TODO (before 2027-01): re-verify the art. 225 text." }, "tax.sales_tax_rate": { "source": "IVA ordinaria 22%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "IMU prima casa esente; symbolic 0.1%", "last_updated": "2026-05-16", "confidence": "high" }, "housing._summary": { "source": "Immobiliare.it + Idealista Roma Q1 2026", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "SSN/Regione Lazio Sanità universale", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "Roma Capitale nido tariffa ISEE + Bonus Asilo Nido INPS", "last_updated": "2026-05-16", "confidence": "medium" }, "food._summary": { "source": "Numbeo Roma 2026 + ISTAT", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "ATAC Metrebus mensile €35", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "ISTAT + Numbeo Roma", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "FCO-SEA/LAX (1-stop) 2025-26", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Assegno Unico Universale INPS", "last_updated": "2026-05-16", "confidence": "high", "note": "Italian Assegno Unico Universale: means-tested by ISEE with a guaranteed minimum (~EUR 57/mo/kid) paid at any income; modeled value is the high-ISEE floor (survives)." }, "non_cash._summary": { "source": "CCNL + Codice Civile (same federal labor law as Milan); ATAC commute data", "last_updated": "2026-05-16", "confidence": "high", "note": "Regime Impatriati IS modeled, via tax.inbound_regime (engine toggle): 5 years, 50% IRPEF exemption, capped at EUR 600,000 of income. Corrected 2026-08-02 from \"not modeled\", which was stale and contradicted milan’s note on a byte-identical regime block. Verified by running compute() at EUR 150,000 gross with apply_inbound_regime off and on: Rome tax moves from EUR 69,288 to EUR 36,462. Big draw for US-tech ICP." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Rome: EUR 35,323, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 35,323 / 0.867 = USD 40,742; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1630, R5-rounded down to 0.16. REPLACES 0.25 (2026-08-11), and the interim 0.14 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.25 was 0.8333 of Milan's 0.3; that ratio x Milan's re-derived 0.17 = 0.1417, R5-rounded down to 0.14. Milan's own figure: EUR 37,601 / 0.867 = USD 43,369, / USD 250,000 = 0.1735. This is a BASIS move, not a re-measurement: Rome's rank inside Italy is unchanged and still rests on its prior evidence, so confidence is medium. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.25. Was 0.37, set 2026-05-16 from the narrative 'Below Milan' with no figure recorded. Measured on the peer basis (levels.fyi senior SWE median total comp / USD 220,000 Seattle senior reference): Rome senior median EUR 48,016 = USD 55,382 / 220,000 = 0.2517. RAL, gross annual, mensilita included. Still below Milan (0.30), so that ordering claim survives the correction. Research: docs/superpowers/research/2026-08-10-milan-comp-index.md" }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-21", "note": "Zeroed for 3-6: the Italian \"bonus asilo nido\" applies to 0-3 nido and is means-tested (only a small floor survives for a high earner); it does not apply to free 3-6 infanzia." }, "childcare.preschool_monthly": { "source": "Roma Capitale, tariffe refezione scolastica a.s. 2025/2026 set by Deliberazione di Giunta Capitolina n. 421 del 15/12/2023; top ISEE band (above EUR 45,000) = EUR 94,40/month https://www.comune.roma.it", "note": "EUR 94.40/month over the ~10-month school year = EUR 944/yr; 944/12 = 78.67. Was 450 (paritaria/bilingual tier). Applies the owner decision of 2026-07-18 that Spain and Italy follow the statutory free convention, now that the age boundary makes it applicable to the free ages only. REPLACES the 2026-07-21 modelling, which priced Italy at the paritaria/bilingual private tier on the argument that January-only enrolment and Italian-only instruction push international families private. That argument holds equally for Catalan-medium Barcelona and for France, Belgium and Germany, where the repo already models the free public path, so Italy was the only country priced private. Stored value is the ANNUAL out-of-pocket divided by 12, because computeChildcareCost multiplies by 12; the invoice figure is a school-year figure over ~10 months. Sourcing: docs/superpowers/research/2026-08-01-italy-mensa-fees.md.", "last_updated": "2026-08-01", "confidence": "high" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "National ticket framework on specialist visits and diagnostics plus the pharmaceutical co-payment, both regionally set; esenzione per reddito thresholds sit far below our ICP household", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 raised from 400 to 480 on the batch-D research. The ticket on specialist visits and diagnostics plus the pharmaceutical co-payment are both REGIONALLY set, and the esenzione per reddito income thresholds sit far below our ICP household, so neither the adults nor the children are exempt and the curve follows the no-child-discount shape. Italy has no annual out-of-pocket ceiling. Naples is the fifth Italian city, not Florence. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "National ticket framework on specialist visits and diagnostics plus the pharmaceutical co-payment, both regionally set; esenzione per reddito thresholds sit far below our ICP household", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, indicator TT1019V, greater city. https://ec.europa.eu/eurostat/api/dissemination/statistics/1.0/data/urb_ctran?indic_ur=TT1019V&format=JSON&lang=EN", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average time of journey to work", "year": "2011", "note": "Urban Audit 44.74. CAVEAT: the underlying ISTAT series measures travel to the place of work OR STUDY, so school trips are blended in; no work-only Italian city mean exists. ISTAT's 2021 census commuting matrix carries no travel-time variable at all." }, "non_cash.sick_leave": { "source": "INPS indennità di malattia (general private-sector \"industria\" regime); Codice; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Agenzia delle Entrate, Risposta n. 76/2024 https://www.agenziaentrate.gov.it/portale/documents/20143/5946162/Risposta+n.+76_2024.pdf/f0933165-e52e-a519-8ca9-85d3f5162f23 ; Risoluzione n. 25/2025 https://www.agenziaentrate.gov.it/portale/documents/20143/8874453/Risoluzione+n.+25_2025/58473993-686e-9f56-723a-034c8a1dd3f6 (D.Lgs. 252/2005 art. 8 c.4)", "last_updated": "2026-08-11", "confidence": "high", "note": "Previdenza complementare deduction from reddito complessivo, EUR 5,164.57/yr, fixed in statute since 2007 (it is 10 million lire) and not indexed. A COMBINED ceiling: employer contributions to the fondo pensione count against the same figure. TFR conferred to the fund does NOT count and is not deductible, having never been taxed as income. less_employer_pct is deliberately left unset, which is the GENEROUS reading, rather than inventing an employer rate; flagged for a later pass. The \"lavoratori di prima occupazione\" extra platform in years 6-25 is deliberately unmodeled." } }, "madrid": { "childcare.free_tier_start_age": { "last_updated": "2026-08-01", "confidence": "high", "source": "Ley Organica 2/2006 de Educacion (LOE), consolidated text, art. 15.2: \"El segundo ciclo de la educacion infantil sera gratuito\"; art. 14.1: \"El primero comprende hasta los tres anos, y el segundo, desde los tres a los seis anos de edad\" https://www.boe.es/buscar/doc.php?id=BOE-A-2006-7899 . Enrolment: Ministerio de Educacion FP y Deportes, 2025-09-09, \"La escolarizacion en el segundo ciclo de educacion infantil (3 a 5 anos) es en Espana practicamente universal, con un 98% de los menores matriculados\" https://www.educacionfpydeportes.gob.es/prensa/actualidad/2025/09/20250909-panoramaeducacion2025.html", "note": "Segundo ciclo (3-6) is free by statute in public and concertado schools and 98% enrolled; primer ciclo (0-3) is neither free nor universal. Before this, ages 1-5 all billed one rate, so the bucket straddled both: a 1-2 year old in paid guarderia was billed the free-tier meal fee. Entry to P3 is by BIRTH-YEAR COHORT (children born in year N start the September of year N+3), so a December-born child starts free schooling at 2 years 9 months; whole-number ages make 3 the correct modelling boundary." }, "non_cash.parental_leave_paid_pct": { "source": "Real Decreto-ley 9/2025 preamble (cien por cien replacement) https://www.boe.es/eli/es/rdl/2025/07/29/9 ; LGSS art. 179 (base reguladora) and art. 148 (tope maximo de la base de cotizacion)", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 38 weeks into non_cash.parental_leave_capped (100% of a base capped at the base maxima de cotizacion), so no percentage-paid weeks remain on this field. Was 100 with the cap ignored per the old field convention. Watch item unchanged: the RDL preamble records a government commitment to extend to 20 weeks per parent, which is not law and is not stored." }, "non_cash.parental_leave_weeks": { "source": "Estatuto de los Trabajadores art. 48.4 as amended by Real Decreto-ley 9/2025 de 29 de julio, in force 31/07/2025 https://www.boe.es/buscar/act.php?id=BOE-A-2025-15741 ; consolidated ET https://www.boe.es/buscar/act.php?id=BOE-A-2015-11430", "last_updated": "2026-08-01", "confidence": "high", "note": "38 wk = 19 per parent x 2 (6 obligatory post-partum + 11 usable to age 12 months + 2 usable to age 8). ET art. 48.4 states este derecho es individual de la persona trabajadora sin que pueda transferirse su ejercicio al otro progenitor. Spain has no separate maternity leave: both parents hold the identical permiso por nacimiento y cuidado. Was 16, one parent on the pre-2025 duration. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Excluded: the 8-week permiso parental of art. 48 bis, which RDL 9/2025 confirms remains unpaid; the 32-week monoparental variant; multiple-birth uplifts." }, "tax.inbound_regime": { "source": "Beckham Law, art. 93.2.e) LIRPF (Ley 35/2006, as amended by Ley 28/2022 Ley de Startups and Ley 7/2024): marginal two-band scale on Spanish employment income, 24% up to EUR 600,000 and 47% on the excess. Duration = the tax period of the move plus the five following (6 years); 5 prior tax periods of non-residency; election on modelo 149. AEAT Manual practico IRPF: https://sede.agenciatributaria.gob.es/Sede/ayuda/manuales-videos-folletos/manuales-practicos/irpf-2024/c15-calculo-impuesto-determinacion-cuotas-integras/regimen-fiscal-especial-aplicable-trabajadores.html ; AEAT Regimen especial impatriados art. 93: https://sede.agenciatributaria.gob.es/Sede/ayuda/manuales-videos-folletos/manuales-practicos/manual-tributacion-no-residentes/regimenes-opcionales/regimen-especial-impatriados.html", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31 OVER-GRANT FIX. The block declared flat_rate 0.24 with no max_income and no flat_rate_above, so 24% ran to infinity even though this very entry described the 47% band. Wired max_income 600000 + flat_rate_above 0.47; the engine blends them into one effective rate. Barcelona single filer at EUR 1,000,000: tax 243,887 -> 335,887. Nothing at or below EUR 600,000 moves." }, "tax.federal_brackets": { "source": "Art. 63.1 LIRPF, escala estatal 2026 (via guiafiscal.es/irpf); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.state_brackets": { "source": "Comunidad de Madrid IRPF autonómica 2026 (Ley 19/2010 + reformas)", "last_updated": "2026-05-16", "confidence": "high", "note": "Lowest autonómica in Spain — Madrid keeps top rate at 22.5%" }, "tax.social_security_rate": { "source": "Seguridad Social employee 6.35% (4.7% CC + 1.55% UE + 0.10% FP)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_cap": { "source": "Seguridad Social base maxima 2026 EUR 5,101.20/mo (taxsummaries.pwc.com/spain); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "medium", "note": "Provisional, finalized Q1 2026" }, "tax.sales_tax_rate": { "source": "IVA general 21% (BOE)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Ayuntamiento de Madrid IBI urbano 0.456%", "last_updated": "2026-05-16", "confidence": "high" }, "housing._summary": { "source": "Idealista + Fotocasa Madrid Q1 2026 + Numbeo Madrid", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "SERPAVI 2026 (MIVAU, Sistema Estatal de Referencia del Precio del Alquiler de Vivienda; publication PDF https://publicaciones.transportes.gob.es/downloadcustom/sample/4078, municipios table p.13, read 2026-08-17): Madrid municipio, tax year 2024, vivienda colectiva, n=300,447 - per-dwelling Cuantia P25/Mediana/P75 = 700/869/1,166 EUR/month, multipliers 700/869 = 0.8055 and 1,166/869 = 1.3418, rounded outward to 0.80/1.35. Methodology (percentile identity, IRPF sitting-stock basis, 30-150 m2 filter): https://cdn.mivau.gob.es/portal-web-mivau/vivienda/alquila/2025-09-10_Metodolog%C3%ADa_SERPAVI.pdf", "last_updated": "2026-08-17", "confidence": "high", "note": "REBASED 2026-08-17 (ledger cluster housing-0817) from 0.67/1.44, which was the district ladder's min and max - a pre-smoothed extreme-of-averages, not a percentile of any rent distribution (the KNOWN LIMIT recorded 2026-08-13). This is the genuine quartile source Madrid never had, closing that open item; NYC (ACS B25063) and London (ONS) precedent. The band and the ladder now deliberately come from DIFFERENT sources measuring different quantities: the ladder stays Brains Q2-2026 2-hab district asking ratios (display), the band is the quartile pair of all municipio tenancies (solver input). The band==ladder-min/max drift-lock test was retired for Madrid in the same pass, replaced by a pin of this pair. WEDGES, named not averaged away: (1) sitting-stock tax-declared rents (landlord IRPF, ongoing habitual-residence tenancies) around ASKING by-bedroom medians - long cheap tenancies drag P25 down, so the pair likely reads WIDER than a same-day asking distribution (conservative direction); the within-ONE-distribution ratio class is what the 2026-08-13 ledger kept for NYC B25063 (regulated/long-tenure stock sits on both sides and largely cancels; the refutation was specific to between-geography ratios). (2) All-unit 30-150 m2 pooling (superficie IQR 52-84 m2), like the ACS metros: part of the spread is size mix, wider than a within-bedroom band, priced residual. (3) Vintage tax year 2024; SERPAVI is annual, next edition ~March 2027. Note asymmetry with Berlin, whose quartile source is ASKING listings - opposite bases, each named. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Derivation: docs/superpowers/research/2026-08-17-housing-band-fixes.md (M1); source scoping: docs/superpowers/research/2026-08-16-housing-band-scope-madrid-berlin-quartile-sources.md" }, "housing.housing_districts": { "source": "Brains Real Estate, 'Mercado Renta Bruta por tipologia' Q2-2026, TWO-BEDROOM (2 hab) district asking rent over a 2-hab citywide denominator of EUR 1,757.7: Salamanca 2,529.67 = 1.44, Centro 2,304.73 = 1.31, Chamberi 2,186.06 = 1.24, Usera 1,285.01 = 0.73, Villaverde 1,249.86 = 0.71, Puente de Vallecas 1,181.62 = 0.67. All 21 Madrid-capital districts and five series (todos/1/2/3/4 hab) are a JS object literal in the served HTML of https://brainsre.news/alquiler-pisos-madrid/ (published 2026-08-11), retrieved 2026-08-13.", "last_updated": "2026-08-13", "confidence": "medium", "note": "BASIS IS A RATIO, NEVER A LEVEL, and the ratio is BEDROOM-MATCHED. CORRECTED 2026-08-13 from 1.22/1.16/1.11/0.84/0.84/0.74, which were Idealista per-m2 district ratios over citywide 23.7 EUR/m2. A EUR/m2 average is dominated by small units, which carry the highest EUR/m2, so a per-m2 ratio silently encodes the ONE-BEDROOM ratio and then gets applied at every bedroom count. Measured against this source, the stored ladder captured 99.5% of the true Salamanca:Villaverde spread at 1 hab, 81.5% at 2 hab, 55.4% at 3 hab and 42.6% at 4 hab. Two-bedroom chosen as the single anchor because the engine applies one multiplier at every bedroom count (districtLadder), 2BR is the ICP's unit, and it matches London's ONS-2BR and NYC's StreetEasy-2BR bases; the tradeoff is priced and accepted. Denominator: Brains publishes no Madrid-capital citywide by-bedroom row, so it was set by holding the geometric mean of the six multipliers unchanged, the level being the one thing NOT in dispute (this source's own per-m2 column reproduces all three tested stored values within 0.02). An independent route (the citywide 1-hab of EUR 1,325 implied by the old ladder itself, stepped by this source's bedroom step) lands at EUR 1,755, a 0.2% difference. A third route using the published listing-concentration factor 1.1967 was REJECTED: it predicts a citywide 1-hab of EUR 1,473 against the EUR 1,325 that five districts agree on, so the listing mix does not weight the same at a fixed bedroom count. Bottom of the ladder moves Villaverde to Puente de Vallecas, which is this source's ordering at 2 hab and corrects a stored 13.5% gap between the two that held at no bedroom count; they invert again at 1 hab, 1.4% apart. Rebasing every district onto one Q2-2026 table also retires a vintage defect: Puente de Vallecas' old per-m2 input was a January 2026 figure over a June 2026 denominator. An unsourced 'Fotocasa June 2026 by-bedroom corroborates (Salamanca 1BR 1,468 / Vallecas 1BR 1,059)' was REMOVED from this field: no URL for it exists anywhere in the repo, Fotocasa publishes no rooms breakdown, its implied 1.386 ratio disagreed with the 1.452 it was cited as supporting, and it corroborated at the one bedroom count where the per-m2 ladder was never wrong. Still only the top-3/bottom-3 of 21 districts; no mid tier sourced, none invented. Per-bedroom multipliers would need an engine + schema change; the full 21-district by-bedroom table is in the research file. Working: docs/superpowers/research/2026-08-13-madrid-berlin-ladder-rebase.md; diagnosis: docs/superpowers/research/2026-08-13-madrid-ladder-spotcheck.md" }, "healthcare._summary": { "source": "SNS (universal) + Sanitas/Adeslas private complementary", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "Private escuela infantil market rate EUR 500-800/mo (elmundodemozart.com, domoescuelainfantil.es) + matrícula/materials annualized; Cheque Guardería / Becas de Educación Infantil (sede.comunidad.madrid) confirmed capped at EUR 35,913 renta per cápita: zero for a high-earning household, doubly so under the Beckham inbound regime which files IRNR (Modelo 151) rather than the resident IRPF return the beca application requires", "last_updated": "2026-07-17", "confidence": "medium", "note": "Ground-truthed against a Madrid-resident staff engineer (3 kids, Beckham regime) who reported paying full private-market rates with zero subsidy; prior EUR 3,113/yr subsidy value was the program's poorest-tier maximum, wrongly applied as a universal ICP benefit. Comunidad de Madrid convenio-center discounts (~EUR 180-250/mo off) reported by the same source were left out of the value. TODO: confirm whether they are city-wide or Salamanca-district-specific before pricing them in" }, "food._summary": { "source": "Numbeo Madrid 2026 + INE", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "CRTM Abono Transporte Zona A €32.70 (2026, joven €20)", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "INE + Numbeo entertainment", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "MAD-SEA/LAX (1-stop via AMS/LHR/CDG) 2025-26", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Prestación por hijo a cargo (means-tested, INSS)", "last_updated": "2026-05-16", "confidence": "medium" }, "non_cash._summary": { "source": "Estatuto de Trabajadores 22 días + 14 festivos + permiso parental biparental 19 sem 100% (2026 reform)", "last_updated": "2026-05-16", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Madrid: EUR 53,912, held in labor_market.tech_salary_distribution (source_slug madrid-esp, as_of 2026-07-18, table confidence thin). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 53,912 / 0.867 = USD 62,182; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2487, R5-rounded down to 0.24. REPLACES 0.4 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Madrid by 67%. The measured p50 was 0.622 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Senior SWE €55-85k; FAANG €90-130k. Index 0.40" }, "tax.filing_basis": { "source": "Agencia Tributaria - individual IRPF (the engine files spouses individually; the elective declaracion conjunta is excluded)", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "childcare.preschool_monthly": { "last_updated": "2026-08-01", "confidence": "medium", "source": "Comedor: ORDEN 1977/2023, de 5 de junio, de la Vicepresidencia, Consejeria de Educacion y Universidades (Comunidad de Madrid), BOCM 2023-06-07: \"Se fija en 5,50 euros el precio diario del menu escolar a aplicar en los centros docentes publicos no universitarios de la Comunidad de Madrid\" https://www.bocm.es/boletin/CM_Orden_BOCM/2023/06/07/BOCM-20230607-22.xml . Aula matinal (\"Los Primeros del Cole\"): priced per school by the consejo escolar / AMPA, not by any BOCM order; published 2025-26 non-socio rates EUR 28-58/month depending on start time and whether breakfast is included (e.g. CEIP La Encina https://ampalaencina.es/primeros-del-cole/ ).", "note": "EUR 5.50/day x 15 = EUR 82.50 comedor, plus aula matinal at EUR 48/month over 10 months = EUR 40.00, total EUR 122.50. Was 200 (an unsourced ~EUR 200 blend). The order carries no expiry and BOCM full-text search finds no later menu-price order, so 5.50 is still operative for 2025-26. AULA MATINAL CHOICE: the highest verified full-schedule rate is EUR 58/month but it BUNDLES BREAKFAST, which the food block already charges via groceries_monthly_per_kid; charging it twice is not conservatism, so the no-breakfast column (EUR 28-48) is used at its top, EUR 48. There is no central Madrid rate to cite, so this leg is medium confidence. EUR 5.50/day x 15 = EUR 82.50 comedor + EUR 40.00 aula matinal = EUR 122.50. Stored value is the ANNUAL total divided by 12 (computeChildcareCost multiplies by 12), NOT the school-year invoice, which runs ~10 months. Comedor annualised as daily rate x 180 dias lectivos / 12 = daily x 15; 180 is the conservative end of the 175-181 range. Applies the owner decision of 2026-07-18 that Spain follows the statutory free convention, now that the age boundary makes it applicable to the free ages only. Sourcing: docs/superpowers/research/2026-08-01-childcare-free-tier-boundary.md and 2026-08-01-spain-wraparound-fees.md." }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/madrid-esp", "last_updated": "2026-07-18", "confidence": "thin", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "EF English Proficiency Index 2024 (Moderate band)", "last_updated": "2024", "note": "National-level English usability for tech work. Band: moderate." }, "non_cash.parental_leave_capped": { "source": "Orden PJC/297/2026, de 30 de marzo, art. 2.1: tope maximo de la base de cotizacion 5.101,20 euros mensuales desde el 1 de enero de 2026 https://www.boe.es/diario_boe/txt.php?id=BOE-A-2026-7296 ; LGSS art. 179 (the base reguladora is the contingencias comunes contribution base) and art. 148 (that base is capped at the tope maximo)", "last_updated": "2026-08-06", "confidence": "high", "note": "All 38 weeks pay 100% of a base capped at EUR 5,101.20/month (EUR 61,214.40/yr), one clean segment with no phase split: above that gross the benefit is the cap, not the salary. At EUR 90,000 the effective rate is ~68%, at EUR 150,000 ~41%, versus the naked 100% this replaces. cap_amount doubles as the payout ceiling because the rate is 100. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Spain)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "RD 16/2012 pharmaceutical co-payment tramos, with our ICP household in the uncapped 50-60% band + Ministerio de Sanidad SCS household out-of-pocket with dental at ~32%", "last_updated": "2026-08-08", "confidence": "medium", "note": "Spain is the European exception on the child rule: a dependent minor registered on a working parent's card pays the SAME co-payment percentage as the parent, so the sub-linearity here is a volume argument (free well-child checks and the vaccination schedule never touch the pharmacy co-pay), not a statutory discount. Spain has NO statutory annual ceiling for the household we model; the monthly caps introduced in 2026 apply only below EUR 35,000 of income. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "RD 16/2012 pharmaceutical co-payment tramos, with our ICP household in the uncapped 50-60% band + Ministerio de Sanidad SCS household out-of-pocket with dental at ~32%", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "INE Spain, Censo de Poblacion y Viviendas, Urban Audit indicator 6.4, Area Urbana Funcional de Madrid. https://www.ine.es/jaxiT3/Tabla.htm?t=69327", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Duracion media del desplazamiento al trabajo (solo ida)", "year": "2011", "note": "AUF value 31.8, the Spanish maximum. INE's definition is verbatim one-way: the average time in minutes to travel between the place of residence and the place of work, solo ida. Published at Functional Urban Area level, which matches our metro basis. The AUF breakdown exists only for 2011. Spain's 2021 national figure of 41.4 is a DIFFERENT quantity (ECEPOV's daily work/study question, i.e. round trip) and was rejected." }, "non_cash.sick_leave": { "source": "Ley General de la Seguridad Social, current consolidated text Real Decreto; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Agencia Tributaria, Manual Renta 2025 s.8.2.2.6 https://sede.agenciatributaria.gob.es/Sede/ayuda/manuales-videos-folletos/manuales-ayuda-presentacion/irpf-2025/8-cumplimentacion-irpf/8_2-base-liquidable-general-base-ahorro/8_2_2-reducciones-aportaciones-prevision-social/8_2_2_6-aportaciones-anuales-maximas-limite-reduccion.html", "last_updated": "2026-08-11", "confidence": "high", "note": "Reduccion de la base imponible general for aportaciones a sistemas de prevision social (Ley 35/2006 arts. 51-52), EUR 1,500/yr. A base deduction, not a credit, and individual. NOT 8,500: the +8,500 increment requires EMPLOYER contributions to a plan de empleo, which we do not model, and worker contributions only ride along under a matching coefficient table. pct_of_earnings deliberately omitted: the outer 30%-of-net-earned-income bound never binds at EUR 75-95k and 0.30 of GROSS would be both inert and wrong-based. SEPARATE pre-existing understatement, out of scope: computeSpainTaxes does not treat cotizaciones as a gasto deducible." } }, "barcelona": { "childcare.infant_monthly": { "source": "Ajuntament de Barcelona, \"Text refos de la regulacio dels preus publics pels serveis d'educacio per al curs 2026-2027 i seguents, per a les escoles bressol municipals\", Comissio de Govern 18-12-2025, published BOPB 23-12-2025, CVE 202510200797, art. 3.1.1 tarificacio social: tram 10 (top band) quota mensual EUR 406,85 = escolaritat EUR 245,96 + alimentacio EUR 160,89, billed 10 months September-June https://ajuntament.barcelona.cat/escolesbressol/sites/default/files/normativa_preus_publics_26_27_publicada_al_bopb.pdf . Private-market context: El Periodico 2026-05-19, \"Las guarderias privadas ya superan de media los 500 euros al mes en Espana\".", "last_updated": "2026-08-01", "confidence": "medium", "note": "VALUE UNCHANGED AT 450, but no longer unsourced. This closes the ROADMAP flag that Barcelona was out of family; the actual defect was preschool_monthly at 500, which is fixed separately. Two anchors now bracket the 0-3 rate. MUNICIPAL FLOOR: tram 10 is the automatic band for any household that renounces the social tariff or does not file AEAT-verifiable income data, and a EUR 140k household clears the 4.5x IRSC threshold (EUR 64,538/yr for a 3-member family) regardless, so tram 10 is unambiguous. EUR 406.85 x 10 months = EUR 4,068.50/yr = EUR 339/month on the 12-month basis this field uses. PRIVATE CEILING: only 4,679 municipal places exist across 105 escoles bressol for the whole city, so most families are private; the national private average is above EUR 500/month and Barcelona-specific private point estimates could not be sourced to any primary or high-quality secondary source. 450 (EUR 5,400/yr) sits inside that band, above the municipal floor, and is retained rather than lowered to the municipal figure: dropping to 339 would flatter Spain on the strength of a place most families cannot get. EXCLUDED: the I2 carve-out. Under the Generalitat Acord Marc (Llei 2/2021, disposicio addicional dissetena), 2-3 year olds in MUNICIPAL escoles bressol pay EUR 0 escolaritat and only EUR 160.89 alimentacio. Barcelona therefore has three tiers, not two, and age 2 is billed the full paid rate here. That is the conservative reading and it is also the realistic one for the private majority, where no I2 gratuity applies." }, "childcare.free_tier_start_age": { "last_updated": "2026-08-01", "confidence": "high", "source": "Ley Organica 2/2006 de Educacion (LOE), consolidated text, art. 15.2: \"El segundo ciclo de la educacion infantil sera gratuito\"; art. 14.1: \"El primero comprende hasta los tres anos, y el segundo, desde los tres a los seis anos de edad\" https://www.boe.es/buscar/doc.php?id=BOE-A-2006-7899 . Enrolment: Ministerio de Educacion FP y Deportes, 2025-09-09, \"La escolarizacion en el segundo ciclo de educacion infantil (3 a 5 anos) es en Espana practicamente universal, con un 98% de los menores matriculados\" https://www.educacionfpydeportes.gob.es/prensa/actualidad/2025/09/20250909-panoramaeducacion2025.html", "note": "Segundo ciclo (3-6) is free by statute in public and concertado schools and 98% enrolled; primer ciclo (0-3) is neither free nor universal. Before this, ages 1-5 all billed one rate, so the bucket straddled both: a 1-2 year old in paid guarderia was billed the free-tier meal fee. Entry to P3 is by BIRTH-YEAR COHORT (children born in year N start the September of year N+3), so a December-born child starts free schooling at 2 years 9 months; whole-number ages make 3 the correct modelling boundary." }, "non_cash.parental_leave_paid_pct": { "source": "Real Decreto-ley 9/2025 preamble (cien por cien replacement) https://www.boe.es/eli/es/rdl/2025/07/29/9 ; LGSS art. 179 (base reguladora) and art. 148 (tope maximo de la base de cotizacion)", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 38 weeks into non_cash.parental_leave_capped (100% of a base capped at the base maxima de cotizacion), so no percentage-paid weeks remain on this field. Was 100 with the cap ignored per the old field convention. Watch item unchanged: the RDL preamble records a government commitment to extend to 20 weeks per parent, which is not law and is not stored." }, "non_cash.parental_leave_weeks": { "source": "Estatuto de los Trabajadores art. 48.4 as amended by Real Decreto-ley 9/2025 de 29 de julio, in force 31/07/2025 https://www.boe.es/buscar/act.php?id=BOE-A-2025-15741 ; consolidated ET https://www.boe.es/buscar/act.php?id=BOE-A-2015-11430", "last_updated": "2026-08-01", "confidence": "high", "note": "38 wk = 19 per parent x 2 (6 obligatory post-partum + 11 usable to age 12 months + 2 usable to age 8). ET art. 48.4 states este derecho es individual de la persona trabajadora sin que pueda transferirse su ejercicio al otro progenitor. Spain has no separate maternity leave: both parents hold the identical permiso por nacimiento y cuidado. Was 16, one parent on the pre-2025 duration. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Excluded: the 8-week permiso parental of art. 48 bis, which RDL 9/2025 confirms remains unpaid; the 32-week monoparental variant; multiple-birth uplifts." }, "tax.inbound_regime": { "source": "Beckham Law, art. 93.2.e) LIRPF (Ley 35/2006, as amended by Ley 28/2022 Ley de Startups and Ley 7/2024): marginal two-band scale on Spanish employment income, 24% up to EUR 600,000 and 47% on the excess. Duration = the tax period of the move plus the five following (6 years); 5 prior tax periods of non-residency; election on modelo 149. AEAT Manual practico IRPF: https://sede.agenciatributaria.gob.es/Sede/ayuda/manuales-videos-folletos/manuales-practicos/irpf-2024/c15-calculo-impuesto-determinacion-cuotas-integras/regimen-fiscal-especial-aplicable-trabajadores.html ; AEAT Regimen especial impatriados art. 93: https://sede.agenciatributaria.gob.es/Sede/ayuda/manuales-videos-folletos/manuales-practicos/manual-tributacion-no-residentes/regimenes-opcionales/regimen-especial-impatriados.html", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31 OVER-GRANT FIX. The block declared flat_rate 0.24 with no max_income and no flat_rate_above, so 24% ran to infinity even though this very entry described the 47% band. Wired max_income 600000 + flat_rate_above 0.47; the engine blends them into one effective rate. Barcelona single filer at EUR 1,000,000: tax 243,887 -> 335,887. Nothing at or below EUR 600,000 moves." }, "tax.federal_brackets": { "source": "Art. 63.1 LIRPF, escala estatal 2026 (via guiafiscal.es/irpf); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.state_brackets": { "source": "Catalonia IRPF autonomica 2026 scale (guiafiscal.es/irpf/cataluna); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high", "note": "8-tranche schedule; top 25.5%" }, "tax.social_security_rate": { "source": "Seguridad Social employee 6.35%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_cap": { "source": "Seguridad Social base maxima 2026 EUR 5,101.20/mo (taxsummaries.pwc.com/spain); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "IVA general 21% (BOE)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Ajuntament de Barcelona IBI 0.66%", "last_updated": "2026-05-16", "confidence": "high" }, "housing._summary": { "source": "Idealista + Habitaclia Barcelona Q1 2026; 2024 rent caps in force", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "CatSalut (universal) + Sanitas/Adeslas private complementary", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "Generalitat ajuts escolaritat 0-3; subvenció per plaça I2 (2-3yr) a llars d'infants privades EUR 800/yr 2025-26 (tramits.gencat.cat, lhdigital.cat)", "last_updated": "2026-07-17", "confidence": "low", "note": "Subsidy value corrected from EUR 1,600 (unsourced) to the confirmed EUR 800/yr Generalitat per-place subsidy; that program only covers I2 (age 2-3) at adhered private centers, so applying it flatly across the whole 1-5 preschool bucket is a simplification shared with the pre-existing engine design. Monthly base fees (infant EUR 450 / preschool EUR 300) NOT re-verified this pass — general market data suggests Barcelona private nursery fees run toward the upper end of the EUR 300-1,000+/mo range, so these may be underpriced; flagged as follow-up pending a Barcelona-specific source or ground-truth commenter" }, "food._summary": { "source": "Numbeo Barcelona 2026 + INE", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "TMB T-mes €23 (subsidized 2024-26)", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "INE + Numbeo entertainment", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "BCN-SEA/LAX (1-stop via AMS/CDG/LHR) 2025-26", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Prestación INSS + Catalunya complemento", "last_updated": "2026-05-16", "confidence": "medium" }, "non_cash._summary": { "source": "Estatuto Trabajadores + permiso parental biparental 19 sem 100%", "last_updated": "2026-05-16", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Barcelona: EUR 65,010, held in labor_market.tech_salary_distribution (source_slug greater-barcelona-area, as_of 2026-07-18, table confidence usable). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 65,010 / 0.867 = USD 74,983; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2999, R5-rounded down to 0.29. REPLACES 0.38 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Barcelona by 31%. The measured p50 was 0.789 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Slightly below Madrid. Index 0.38" }, "tax.filing_basis": { "source": "Agencia Tributaria - individual IRPF (the engine files spouses individually; the elective declaracion conjunta is excluded)", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-21", "note": "Zeroed: Spain has no universal cash childcare subsidy for free 3-6 public infantil; the prior value was a means-tested benefit that is 0 for a high earner." }, "childcare.preschool_monthly": { "source": "Menjador: Resolucio EDF/1222/2025, de 2 d'abril, Departament d'Educacio i Formacio Professional, DOGC num. 9389 de 9.4.2025 (CVE-DOGC-A-25094063-2025), operative text: \"Fixar el preu maxim de la prestacio del servei escolar de menjador [...] per al curs escolar 2025-2026 [...] a) Amb caracter general [...] en la quantitat de 7,54 euros/alumne/dia, IVA inclos\" (mirror: https://www.vallesoriental.cat/media/repository/ambits/politiques-socials/Ajuts_de_Menjador/25-26/RESOLUCIO_PREU_MENJADOR_25_26.pdf ); enabling norm Decret 160/1996, art. 14. Dies lectius: Catalan law caps the school year at \"entre cent setanta-cinc i cent setanta-vuit dies lectius per curs\"; curs 2025-2026 runs 177. Acollida matinal: priced per school by the AFA, not centrally; Barcelona-city AFA published 2025-26 non-member rates for the 7:30-9:00 slot, e.g. https://afaitaca.cat and https://afa.escolalesacacies.cat .", "last_updated": "2026-08-01", "confidence": "medium", "note": "EUR 7.54/day x 177 dies lectius = EUR 1,334.58/yr = EUR 111.21/month, plus acollida matinal EUR 46/month over 10 months = EUR 38.33, total EUR 149.55. Was 500, modelled at the concertado/private bilingual tier, which is the ROADMAP out-of-family flag: it exceeded Barcelona's own infant_monthly of 450 even though Catalonia's 3-6 tier is free. TWO CORRECTIONS TO CARRIED-OVER INPUTS, both found by reading the operative text rather than a summary. (1) The rate is 7,54, not the 7,83 an earlier sweep used; 7,83 appears nowhere in the resolution. The schedule is 7,54 (2h30 midday block) / 7,13 (2h) for a comensal fix, and 8,29 / 7,85 for comensals esporadics, defined in clause 3 as pupils attending under three days a week, which a dual-earner family is not. (2) Catalonia runs 177 dies lectius, not the 180 used elsewhere in this pass. Together those two took the annual comedor from EUR 1,409.40 to EUR 1,334.58, a 5.6% overstatement removed. ACOLLIDA: the non-member rate is used (AFA membership is a discount, and the ICP is unlikely to be a member in year one) and it excludes breakfast, which the food block already charges via groceries_monthly_per_kid. WATCH: Resolucio EDF/483/2026 (20 Feb 2026) exists and most likely prices curs 2026-2027; if it raises the rate this value should follow. Stored value is the ANNUAL total divided by 12 (computeChildcareCost multiplies by 12), NOT the school-year invoice, which runs ~10 months. Applies the owner decision of 2026-07-18 that Spain follows the statutory free convention, now that the age boundary makes it applicable to the free ages only. Sourcing: docs/superpowers/research/2026-08-01-childcare-free-tier-boundary.md and 2026-08-01-spain-wraparound-fees.md." }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/greater-barcelona-area", "last_updated": "2026-07-18", "confidence": "usable", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "EF English Proficiency Index 2024 (Moderate band)", "last_updated": "2024", "note": "National-level English usability for tech work. Band: moderate." }, "non_cash.parental_leave_capped": { "source": "Orden PJC/297/2026, de 30 de marzo, art. 2.1: tope maximo de la base de cotizacion 5.101,20 euros mensuales desde el 1 de enero de 2026 https://www.boe.es/diario_boe/txt.php?id=BOE-A-2026-7296 ; LGSS art. 179 (the base reguladora is the contingencias comunes contribution base) and art. 148 (that base is capped at the tope maximo)", "last_updated": "2026-08-06", "confidence": "high", "note": "All 38 weeks pay 100% of a base capped at EUR 5,101.20/month (EUR 61,214.40/yr), one clean segment with no phase split: above that gross the benefit is the cap, not the salary. At EUR 90,000 the effective rate is ~68%, at EUR 150,000 ~41%, versus the naked 100% this replaces. cap_amount doubles as the payout ceiling because the rate is 100. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Spain)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "RD 16/2012 pharmaceutical co-payment tramos, with our ICP household in the uncapped 50-60% band + Ministerio de Sanidad SCS household out-of-pocket with dental at ~32%", "last_updated": "2026-08-08", "confidence": "medium", "note": "Spain is the European exception on the child rule: a dependent minor registered on a working parent's card pays the SAME co-payment percentage as the parent, so the sub-linearity here is a volume argument (free well-child checks and the vaccination schedule never touch the pharmacy co-pay), not a statutory discount. Spain has NO statutory annual ceiling for the household we model; the monthly caps introduced in 2026 apply only below EUR 35,000 of income. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "RD 16/2012 pharmaceutical co-payment tramos, with our ICP household in the uncapped 50-60% band + Ministerio de Sanidad SCS household out-of-pocket with dental at ~32%", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "INE Spain, Urban Audit indicator 6.4, Area Urbana Funcional de Barcelona. https://www.ine.es/jaxiT3/Tabla.htm?t=69327", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Duracion media del desplazamiento al trabajo (solo ida)", "year": "2011", "note": "AUF value 26.7." }, "non_cash.sick_leave": { "source": "Ley General de la Seguridad Social, current consolidated text Real Decreto; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Agencia Tributaria, Manual Renta 2025 s.8.2.2.6 https://sede.agenciatributaria.gob.es/Sede/ayuda/manuales-videos-folletos/manuales-ayuda-presentacion/irpf-2025/8-cumplimentacion-irpf/8_2-base-liquidable-general-base-ahorro/8_2_2-reducciones-aportaciones-prevision-social/8_2_2_6-aportaciones-anuales-maximas-limite-reduccion.html", "last_updated": "2026-08-11", "confidence": "high", "note": "Reduccion de la base imponible general for aportaciones a sistemas de prevision social (Ley 35/2006 arts. 51-52), EUR 1,500/yr. A base deduction, not a credit, and individual. NOT 8,500: the +8,500 increment requires EMPLOYER contributions to a plan de empleo, which we do not model, and worker contributions only ride along under a matching coefficient table. pct_of_earnings deliberately omitted: the outer 30%-of-net-earned-income bound never binds at EUR 75-95k and 0.30 of GROSS would be both inert and wrong-based. SEPARATE pre-existing understatement, out of scope: computeSpainTaxes does not treat cotizaciones as a gasto deducible." } }, "valencia": { "childcare.free_tier_start_age": { "last_updated": "2026-08-01", "confidence": "high", "source": "Ley Organica 2/2006 de Educacion (LOE), consolidated text, art. 15.2: \"El segundo ciclo de la educacion infantil sera gratuito\"; art. 14.1: \"El primero comprende hasta los tres anos, y el segundo, desde los tres a los seis anos de edad\" https://www.boe.es/buscar/doc.php?id=BOE-A-2006-7899 . Enrolment: Ministerio de Educacion FP y Deportes, 2025-09-09, \"La escolarizacion en el segundo ciclo de educacion infantil (3 a 5 anos) es en Espana practicamente universal, con un 98% de los menores matriculados\" https://www.educacionfpydeportes.gob.es/prensa/actualidad/2025/09/20250909-panoramaeducacion2025.html", "note": "Segundo ciclo (3-6) is free by statute in public and concertado schools and 98% enrolled; primer ciclo (0-3) is neither free nor universal. Before this, ages 1-5 all billed one rate, so the bucket straddled both: a 1-2 year old in paid guarderia was billed the free-tier meal fee. Entry to P3 is by BIRTH-YEAR COHORT (children born in year N start the September of year N+3), so a December-born child starts free schooling at 2 years 9 months; whole-number ages make 3 the correct modelling boundary." }, "non_cash.parental_leave_paid_pct": { "source": "Real Decreto-ley 9/2025 preamble (cien por cien replacement) https://www.boe.es/eli/es/rdl/2025/07/29/9 ; LGSS art. 179 (base reguladora) and art. 148 (tope maximo de la base de cotizacion)", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 38 weeks into non_cash.parental_leave_capped (100% of a base capped at the base maxima de cotizacion), so no percentage-paid weeks remain on this field. Was 100 with the cap ignored per the old field convention. Watch item unchanged: the RDL preamble records a government commitment to extend to 20 weeks per parent, which is not law and is not stored." }, "non_cash.parental_leave_weeks": { "source": "Estatuto de los Trabajadores art. 48.4 as amended by Real Decreto-ley 9/2025 de 29 de julio, in force 31/07/2025 https://www.boe.es/buscar/act.php?id=BOE-A-2025-15741 ; consolidated ET https://www.boe.es/buscar/act.php?id=BOE-A-2015-11430", "last_updated": "2026-08-01", "confidence": "high", "note": "38 wk = 19 per parent x 2 (6 obligatory post-partum + 11 usable to age 12 months + 2 usable to age 8). ET art. 48.4 states este derecho es individual de la persona trabajadora sin que pueda transferirse su ejercicio al otro progenitor. Spain has no separate maternity leave: both parents hold the identical permiso por nacimiento y cuidado. Was 16, one parent on the pre-2025 duration. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Excluded: the 8-week permiso parental of art. 48 bis, which RDL 9/2025 confirms remains unpaid; the 32-week monoparental variant; multiple-birth uplifts." }, "tax.inbound_regime": { "source": "Beckham Law, art. 93.2.e) LIRPF (Ley 35/2006, as amended by Ley 28/2022 Ley de Startups and Ley 7/2024): marginal two-band scale on Spanish employment income, 24% up to EUR 600,000 and 47% on the excess. Duration = the tax period of the move plus the five following (6 years); 5 prior tax periods of non-residency; election on modelo 149. AEAT Manual practico IRPF: https://sede.agenciatributaria.gob.es/Sede/ayuda/manuales-videos-folletos/manuales-practicos/irpf-2024/c15-calculo-impuesto-determinacion-cuotas-integras/regimen-fiscal-especial-aplicable-trabajadores.html ; AEAT Regimen especial impatriados art. 93: https://sede.agenciatributaria.gob.es/Sede/ayuda/manuales-videos-folletos/manuales-practicos/manual-tributacion-no-residentes/regimenes-opcionales/regimen-especial-impatriados.html", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31 OVER-GRANT FIX. The block declared flat_rate 0.24 with no max_income and no flat_rate_above, so 24% ran to infinity even though this very entry described the 47% band. Wired max_income 600000 + flat_rate_above 0.47; the engine blends them into one effective rate. Barcelona single filer at EUR 1,000,000: tax 243,887 -> 335,887. Nothing at or below EUR 600,000 moves." }, "tax.federal_brackets": { "source": "Art. 63.1 LIRPF, escala estatal 2026 (via guiafiscal.es/irpf); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.state_brackets": { "source": "Generalitat Valenciana IRPF autonómica 2026 (Ley 13/1997 + 2023 reform)", "last_updated": "2026-05-16", "confidence": "high", "note": "Top 29.5% — highest autonómica in Spain" }, "tax.social_security_rate": { "source": "Seguridad Social employee 6.35%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_cap": { "source": "Seguridad Social base maxima 2026 EUR 5,101.20/mo (taxsummaries.pwc.com/spain); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "IVA general 21% (BOE)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Ajuntament de València IBI 1.071%", "last_updated": "2026-05-16", "confidence": "high", "note": "Highest of 3 SP cities — VLC keeps higher IBI to offset lower COL" }, "housing._summary": { "source": "Idealista + Habitaclia Valencia Q1 2026", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Conselleria de Sanitat Universal + private complementary", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "Bono Infantil Generalitat Valenciana (sede.gva.es, tus-ayudas.es): 0-1/1-2 tramos income-tested up to EUR 72,460-78,822 household income depending on family size, above which the aid is EUR 0; 2-3 tramo is a universal EUR 280/mo paid direct to the center regardless of income", "last_updated": "2026-07-17", "confidence": "low", "note": "subsidy_per_kid_annual zeroed (was EUR 1,600, unsourced) since a well-paid tech household exceeds the income-tested 0-2 caps; the universal 2-3 tramo isn't cleanly representable as a flat per-kid subsidy in this schema (same limitation as Barcelona's I2 subsidy) so it's folded qualitatively into preschool_monthly instead. preschool_monthly raised from EUR 0 (was implicitly 'fully free', wrong for the 0-2 portion of the age-1-5 bucket and for extras/hours in the free 2-6 portion) to a rough blended EUR 200/mo — precision here is a follow-up item, not independently ground-truthed like Madrid" }, "food._summary": { "source": "Numbeo Valencia 2026 + INE", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "EMT València + Metrovalencia bono mensual ~€40", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "INE + Numbeo Valencia", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "VLC-SEA/LAX (multi-stop) 2025-26", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Prestación INSS", "last_updated": "2026-05-16", "confidence": "medium" }, "non_cash._summary": { "source": "Estatuto Trabajadores + permiso parental biparental 19 sem 100%", "last_updated": "2026-05-16", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Valencia. Its basis is converted with the rest of Spain and its rank inside Spain is preserved. Prior source: Derived, not observed: no all-levels total-comp measurement exists for Valencia. Basis re-derived with its country anchor Madrid; Valencia's stored ratio to Madrid is preserved. Prior source: Levels.fyi Valencia + remote-work salary surveys", "last_updated": "2026-08-11", "confidence": "low", "note": "0.34 x 0.6000 = 0.2040, R5-rounded down to 0.2. 0.6000 is the MEDIAN of (re-derived / previous) across the 5 Spain cities that DO carry an all-levels measurement (Barcelona 0.38 to 0.29; Bilbao 0.4 to 0.18; Madrid 0.4 to 0.24; Málaga 0.35 to 0.21; Seville 0.33 to 0.15), so Valencia moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Valencia's rank inside Spain is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.34 was 0.8500 of Madrid's 0.4; that ratio x Madrid's re-derived 0.24 = 0.2040, R5-rounded down to 0.2. Madrid's own figure: EUR 53,912 / 0.867 = USD 62,182, / USD 250,000 = 0.2487. This is a BASIS move, not a re-measurement: Valencia's rank inside Spain is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Lower than MAD/BCN; growing remote hub. Index 0.34" }, "tax.filing_basis": { "source": "Agencia Tributaria - individual IRPF (the engine files spouses individually; the elective declaracion conjunta is excluded)", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "childcare.preschool_monthly": { "source": "Comedor: Conselleria d'Educacio, Generalitat Valenciana, precio maximo autorizado curso 2025/26 EUR 4,35/dia para Educacion Infantil, Primaria y ESO https://portal.edu.gva.es/03004442/wp-content/uploads/sites/1619/2025/09/Precios-comedor-2025-2026.pdf (corroborated by the official procedure page https://sede.gva.es/es/detall-tramit?id_proc=G1245 , which carries the same ceiling into 2026/2027). Escola matinera: priced per centre by the AMPA/AFA, not centrally; published 2025-26 tariffs on the Generalitat school portal: CEIP San Blas 7:30-8:15 EUR 40/mes (the published price IS the non-member price) https://portal.edu.gva.es/sanblas/servici-escola-matinera-25-26/ ; CEIP Ausias March Mislata 07:30-09:00 EUR 37/mes non-member https://portal.edu.gva.es/ausiasmarchmislata/wp-content/uploads/sites/1236/2025/09/ESCOLA-MATINERA-CURS-25-26-CABRIOLES.pdf ; El Campello municipal protocol, the same service priced both ways, EUR 35/mes without breakfast vs EUR 48 with https://portal.edu.gva.es/pladebarraques/wp-content/uploads/sites/365/2025/09/Protocolo-Escola-Matinera-castellano.pdf", "last_updated": "2026-08-01", "confidence": "medium", "note": "EUR 4.35/day x 15 = EUR 65.25 comedor + escola matinera at EUR 40/month over 10 months = EUR 33.33, total EUR 98.58. Was 180 (an unsourced blend). MATINERA CHOICE: EUR 40 is the top of the without-breakfast range (EUR 35-40 for a 7:00-7:30 start); the with-breakfast figure is EUR 48, and the food block already charges breakfast via groceries_monthly_per_kid, so using it would double-count. The El Campello protocol is what makes that split clean: it prices the identical service at 35 without and 48 with. No central rate exists (verified: the Conselleria fixes only the comedor module; the Ajuntament de Valencia's only instrument is a means-tested ayuda requiring an existing cheque escolar or comedor grant plus both parents working, which a EUR 140k household does not get). CAVEAT: no Valencia-CITY CEIP tariff could be pinned; the sourced schools are Mislata, El Campello and Alicante, so treat EUR 35-40 as a Comunitat-wide figure. EXCLUDED: the Bono Infantil 2-3 tramo (~EUR 280/month) is UNIVERSAL rather than income-tested and does reach a high-income household; the schema has no universal-cash-benefit slot and a flat per-kid subsidy would leak into the free 3-6 years. Stored value is the ANNUAL total divided by 12 (computeChildcareCost multiplies by 12), NOT the school-year invoice, which runs ~10 months. Comedor annualised as daily rate x 180 dias lectivos / 12 = daily x 15. Applies the owner decision of 2026-07-18 that Spain follows the statutory free convention, now that the age boundary makes it applicable to the free ages only. Sourcing: docs/superpowers/research/2026-08-01-childcare-free-tier-boundary.md and 2026-08-01-spain-wraparound-fees.md." }, "non_cash.parental_leave_capped": { "source": "Orden PJC/297/2026, de 30 de marzo, art. 2.1: tope maximo de la base de cotizacion 5.101,20 euros mensuales desde el 1 de enero de 2026 https://www.boe.es/diario_boe/txt.php?id=BOE-A-2026-7296 ; LGSS art. 179 (the base reguladora is the contingencias comunes contribution base) and art. 148 (that base is capped at the tope maximo)", "last_updated": "2026-08-06", "confidence": "high", "note": "All 38 weeks pay 100% of a base capped at EUR 5,101.20/month (EUR 61,214.40/yr), one clean segment with no phase split: above that gross the benefit is the cap, not the salary. At EUR 90,000 the effective rate is ~68%, at EUR 150,000 ~41%, versus the naked 100% this replaces. cap_amount doubles as the payout ceiling because the rate is 100. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Spain)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "RD 16/2012 pharmaceutical co-payment tramos, with our ICP household in the uncapped 50-60% band + Ministerio de Sanidad SCS household out-of-pocket with dental at ~32%", "last_updated": "2026-08-08", "confidence": "medium", "note": "Spain is the European exception on the child rule: a dependent minor registered on a working parent's card pays the SAME co-payment percentage as the parent, so the sub-linearity here is a volume argument (free well-child checks and the vaccination schedule never touch the pharmacy co-pay), not a statutory discount. Spain has NO statutory annual ceiling for the household we model; the monthly caps introduced in 2026 apply only below EUR 35,000 of income. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "RD 16/2012 pharmaceutical co-payment tramos, with our ICP household in the uncapped 50-60% band + Ministerio de Sanidad SCS household out-of-pocket with dental at ~32%", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "INE Spain, Urban Audit indicator 6.4, Area Urbana Funcional de Valencia. https://www.ine.es/jaxiT3/Tabla.htm?t=69327", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Duracion media del desplazamiento al trabajo (solo ida)", "year": "2011", "note": "AUF value 21.3." }, "non_cash.sick_leave": { "source": "Ley General de la Seguridad Social, current consolidated text Real Decreto; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Agencia Tributaria, Manual Renta 2025 s.8.2.2.6 https://sede.agenciatributaria.gob.es/Sede/ayuda/manuales-videos-folletos/manuales-ayuda-presentacion/irpf-2025/8-cumplimentacion-irpf/8_2-base-liquidable-general-base-ahorro/8_2_2-reducciones-aportaciones-prevision-social/8_2_2_6-aportaciones-anuales-maximas-limite-reduccion.html", "last_updated": "2026-08-11", "confidence": "high", "note": "Reduccion de la base imponible general for aportaciones a sistemas de prevision social (Ley 35/2006 arts. 51-52), EUR 1,500/yr. A base deduction, not a credit, and individual. NOT 8,500: the +8,500 increment requires EMPLOYER contributions to a plan de empleo, which we do not model, and worker contributions only ride along under a matching coefficient table. pct_of_earnings deliberately omitted: the outer 30%-of-net-earned-income bound never binds at EUR 75-95k and 0.30 of GROSS would be both inert and wrong-based. SEPARATE pre-existing understatement, out of scope: computeSpainTaxes does not treat cotizaciones as a gasto deducible." } }, "paris": { "non_cash.parental_leave_paid_pct": { "source": "service-public.gouv.fr F2265 (indemnites journalieres de maternite) https://www.service-public.gouv.fr/particuliers/vosdroits/F2265 ; F39685 (conge supplementaire de naissance) https://www.service-public.gouv.fr/particuliers/vosdroits/F39685", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 37 weeks into non_cash.parental_leave_capped (the 83 blend up to the IJ daily ceiling), so no percentage-paid weeks remain on this field. Was 83 with three caps ignored per the old field convention." }, "non_cash.parental_leave_weeks": { "source": "Conge supplementaire de naissance, service-public.gouv.fr https://www.service-public.gouv.fr/particuliers/vosdroits/F39685 (verifie le 03 juin 2026) ; conge de maternite https://www.service-public.gouv.fr/particuliers/vosdroits/F2265 ; conge de paternite et d'accueil de l'enfant, 25 jours calendaires (4 employer-paid + 21 CPAM) https://www.service-public.gouv.fr/particuliers/vosdroits/F3156 ; conge parental d'education https://www.service-public.gouv.fr/particuliers/vosdroits/F2280", "last_updated": "2026-08-01", "confidence": "high", "note": "37 wk = conge de maternite 16 (mother) + conge de paternite et d'accueil 3.6 (25 calendar days, father) + conge supplementaire de naissance 17.4 (2 months per parent x 2). Was 16, maternity alone. MAJOR CHANGE IN FORCE: the conge supplementaire de naissance (CSN) applies from 1 July 2026 to children born on or after 1 January 2026, giving each parent 1 or 2 months as a personal non-transferable right, so it SUMS across both parents. It must be taken AFTER maternity and paternity leave, so those weeks stack rather than offset. PreParE is EXCLUDED from the headline under the overlap rule: service-public states the CSN cannot be cumulated with PreParE for the same period, and the CSN pays roughly six times more, so the weeks are counted once at the higher rate. Excluded: the 3 jours ouvrables of the OLD employer-paid conge de naissance (0.5 wk of noise, and omitting it is the less favourable reading); the conge parental d'education, which service-public confirms is unpaid absent a collective agreement; multiple-birth uplifts." }, "childcare.free_tier_start_age": { "source": "service-public.gouv.fr F1898 (Direction de l'information legale et administrative): \"L'instruction est obligatoire pour tous les enfants, francais et etrangers, a partir de 3 ans et jusqu'a l'age de 16 ans revolus\" (loi du 26 juillet 2019 pour une ecole de la confiance) https://www.service-public.fr/particuliers/vosdroits/F1898 ; education.gouv.fr https://www.education.gouv.fr/l-inscription-l-ecole-maternelle-8651", "last_updated": "2026-08-01", "confidence": "high", "note": "Ecole maternelle runs 3-6, is free, and has been COMPULSORY since 2019, so participation at 3-5 is at or above 96% (Eurostat). Creche (0-3) is paid. Before this, ages 1-5 all billed one rate, so a 1-year-old in creche was billed the maternelle wraparound rate." }, "tax.federal_brackets": { "source": "impots.gouv.fr barème IR 2026 (indexed +1.8%)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.csg_crds_rate": { "source": "service-public.fr CSG 9.2% + CRDS 0.5%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_rate": { "source": "URSSAF employee-side contributions excl. CSG/CRDS: vieillesse 7.30% + retraite complémentaire T1 ~4.01% ≈ 11.3%", "last_updated": "2026-07-11", "confidence": "high", "note": "Corrected 2026-07-11: previous 0.22 blended figure double-counted CSG/CRDS (which the engine adds separately via csg_crds_rate). IR barème now also runs on the net imposable base in computeFranceTaxes." }, "tax.sales_tax_rate": { "source": "TVA standard 20% (economie.gouv.fr)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Ville de Paris taxe foncière 13.5% × 50% valeur locative ≈ effective 0.5%", "last_updated": "2026-05-16", "confidence": "medium" }, "housing._summary": { "source": "SeLoger + MeilleursAgents Paris Q1 2026 + Numbeo Paris", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4), KEPT over a Paris-specific derivation: the encadrement quartier pair (0.88/1.18 vs a Belleville mid anchor) comes from 3 inner quartiers the research says understate the spread, and the statutory -30%/+20% band is a legal rule, not percentiles. Default is the wider, conservative choice.", "last_updated": "2026-08-07", "confidence": "medium", "note": "District tiers deferred: a clean Paris ladder needs the full ~80-quartier encadrement pull plus a citywide anchor choice (research §3). Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Assurance Maladie (Sécu base 70%) + mutuelle complémentaire (Harmonie Mutuelle/MGEN)", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "CAF crèche tarif + école maternelle gratuite age 3+", "last_updated": "2026-05-16", "confidence": "high" }, "food._summary": { "source": "Numbeo Paris 2026 + INSEE", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "RATP Navigo monthly €88.80 (Jan 2026)", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "INSEE discretionary + Numbeo entertainment", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "Air France CDG-SEA/LAX nonstop or 1-stop 2025-26", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "CAF allocations familiales (typical means-tested band)", "last_updated": "2026-05-16", "confidence": "medium", "note": "French allocations familiales: means-tested/income-modulated (from the 2nd child), retaining a reduced high-income tier; modeled value is the high-income modulated rate." }, "non_cash._summary": { "source": "Code du travail 5 weeks CP + 11 jours fériés + Sécu IJSS + congé maternité 16 weeks", "last_updated": "2026-05-16", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Paris: EUR 64,372, held in labor_market.tech_salary_distribution (source_slug paris-fra, as_of 2026-07-18, table confidence usable). That table was re-fetched live 2026-08-11 and matched within 3.0% (sweep A F2).", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 64,372 / 0.867 = USD 74,247; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2970, R5-rounded down to 0.29. REPLACES 0.55 (2026-05-16), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Paris by 90%. The measured p50 was 0.540 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): FAANG/Datadog/Mistral SWE €110-160k; mid €70-90k. Index 0.55" }, "childcare.subsidy_per_kid_annual": { "source": "paris.fr quotient familial / cantine 2025; CAF PSU crèche", "last_updated": "2026-07-17", "note": "French childcare (crèche PSU / périscolaire + cantine) is means-tested by income/quotient familial; top-tranche households pay the full/top rate, so the modeled prices are net top-tranche costs and the means-tested subsidy is 0. Was 2,200-2,400 (a mid-income CMG/PSU level that a top earner does not receive)." }, "childcare.after_school_monthly": { "source": "paris.fr cantine (EUR 7/meal top tranche) + périscolaire étude (top tranche) 2025", "last_updated": "2026-07-17", "note": "Paris périscolaire + cantine at the top QF tranche ~= EUR 1,800/yr (cantine ~EUR 1,000 + étude/centre de loisirs ~EUR 800) = EUR 150/mo net. Was EUR 220/mo grossed then wiped by a 2,400 subsidy to a ~EUR 240 net, understating the top-tranche cost." }, "childcare.preschool_monthly": { "source": "Ville de Paris, tranche T10 (quotient familial > EUR 5,000, the top tranche): cantine EUR 7.00/meal https://www.paris.fr/pages/la-cantine-2074 ; https://www.paris.fr/pages/quotient-familial-parisien-et-tranche-tarifaire-27928 ; periscolaire schedule: gouter recreatif EUR 6.25/month per weekly session and Wednesday centre de loisirs EUR 33.60/month", "last_updated": "2026-08-01", "confidence": "medium", "note": "VALUE UNCHANGED AT 150, now sourced. Paris is a 4.5-day week = 162 school days. Sourced components: cantine 7.00 x 162 = 1,134, gouter recreatif 4 afternoons x 6.25 x 10 months = 250, Wednesday centre de loisirs 33.60 x 10 = 336, total EUR 1,720/yr = EUR 143.33/month. That is a FLOOR: the accueil du matin slot is not priced in the retrieved schedule, and the periscolaire euro amounts come from the 2023/24 edition (tranche thresholds still match). 150 sits just above the sourced floor, so it stands under the conservative value bias. Free tuition, so the modelled cost is the working-parent wraparound at the TOP quotient-familial tranche: cantine plus periscolaire plus, on a 4-day-week commune, Wednesday care. Stored value is the ANNUAL total divided by 12 (computeChildcareCost multiplies by 12); French communes bill over the ~10-month school year. School days: the annee scolaire is 36 weeks (Code de l'education art. L521-1) = 144 days on a 4-day week, 162 on a 4.5-day week. Sourcing and the full per-slot grids: docs/superpowers/research/2026-08-01-france-cantine-fees.md." }, "labor_market.tech_salary_distribution": { "source": "https://www.levels.fyi/t/software-engineer/locations/paris-fra", "last_updated": "2026-07-18", "confidence": "usable", "note": "levels.fyi total comp (salary + stock + bonus), all-levels Software Engineer, snapshot. Self-reported; skews senior / big-tech, so it reads as senior-leaning tech pay, not the full local market. Sample sizes not published." }, "labor_market.english_band": { "source": "EF English Proficiency Index 2024 (Moderate band)", "last_updated": "2024", "note": "National-level English usability for tech work. Band: moderate." }, "transit.statutory_employer_share_pct": { "source": "https://entreprendre.service-public.gouv.fr/vosdroits/F37900 (Code du travail L3261-2)", "last_updated": "2026-08-05", "confidence": "high", "note": "Every French employer must reimburse 50% of a public-transit subscription (prise en charge obligatoire). The engine nets this share on the pass branch and the row label discloses it. Statutory-share convention, owner decision 2026-08-04; Belgium checked and does NOT qualify (sectoral agreements, not one statute)." }, "transit.monthly_pass": { "source": "https://www.iledefrance-mobilites.fr/titres-et-tarifs (Navigo mensuel 2026)", "last_updated": "2026-08-05", "confidence": "high", "note": "Navigo mensuel EUR 90.80 (2026), sticker price before the statutory 50% employer share the engine nets. Was 88." }, "non_cash.parental_leave_capped": { "source": "service-public.gouv.fr F2265: IJ maternite/paternite plafonnee a 104,02 EUR par jour (2026), salaire journalier de base capped at the PASS of 4 005 EUR/month https://www.service-public.gouv.fr/particuliers/vosdroits/F2265 ; F39685: CSN paid 70% month 1 / 60% month 2 of a base capped at the same PASS https://www.service-public.gouv.fr/particuliers/vosdroits/F39685", "last_updated": "2026-08-06", "confidence": "high", "note": "Conservative single blend of TWO statutory cap mechanisms: the 19.6 IJ weeks (maternity 16 + paternity 3.6) cap the PAYOUT at EUR 104.02/day, and the 17.4 CSN weeks cap the BASE at EUR 4,005/month before the 70%/60% rate (a weekly ceiling around EUR 555 to 647). We store one block, 83% up to EUR 104.02/day (EUR 728.14/wk), the best-known ceiling, which reads the CSN weeks slightly LOW, the direction less favorable to France. At EUR 90,000 the effective rate is ~42%, matching the 40-45% the 2026-08-01 note already estimated. If the engine ever supports two capped blocks, France is a candidate to split. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (France)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "DREES Fiche 21, reste a charge des menages measured after BOTH Assurance Maladie and the complementaire (EUR 292 per inhabitant, 2024), netted of dental and optical + ameli participation forfaitaire and franchises medicales, EUR 100/yr cap each", "last_updated": "2026-08-08", "confidence": "medium", "note": "DREES measures exactly our basis, which matters in France more than anywhere: the mutuelle has been employer-mandated since ANI 2016 and is a PREMIUM, so the honest French out-of-pocket figure is the residual after it. Under-18s are statutorily exempt from both participations and are covered by 100% Sante, so a child adds ~40% of an adult. All French cities are normalised onto one curve: the previous Lyon-300 against Paris/Toulouse/Nantes/Marseille-350 split had no stated basis. A Paris depassements uplift is queued, not shipped. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "DREES Fiche 21, reste a charge des menages measured after BOTH Assurance Maladie and the complementaire (EUR 292 per inhabitant, 2024), netted of dental and optical + ameli participation forfaitaire and franchises medicales, EUR 100/yr cap each", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Institut Paris Region, Enquete regionale sur la mobilite des Franciliens (EMG 2023). https://www.institutparisregion.fr/mobilite-et-transports/deplacements/enquete-regionale-sur-la-mobilite-des-franciliens/", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Duree moyenne d'un deplacement domicile-travail", "year": "2023", "note": "Ile-de-France, one way, all modes. The older INSEE round-trip figure of 68 min halves to 34, corroborating the direction." }, "non_cash.sick_leave": { "source": "Indemnités journalières de sécurité sociale (IJSS): Code de la sécurité sociale; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": "https://www.echangesassurances.org/actualites/indemnites-journalieres-maladie-plafond-juillet-2026", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "service-public.gouv.fr, Plafond epargne retraite (PER) https://www.service-public.gouv.fr/particuliers/vosdroits/F14709?lang=fr ; impots.gouv.fr https://www.impots.gouv.fr/particulier/questions/que-signifie-la-somme-imprimee-dans-la-rubrique-plafond-epargne-retraite-de", "last_updated": "2026-08-11", "confidence": "high", "note": "PER individuel deduction from revenu net global, CGI art. 163 quatervicies, for versements paid in 2026 (plafond computed on 2025 income, PASS 2025 = EUR 47,100): the greater of 10% of prior-year revenus d'activite capped at 8 x PASS (EUR 37,680) or 10% of PASS (EUR 4,710). BASIS TRAP HANDLED: the statutory 10% runs on activity income NET of cotisations sociales and after the 10% abattement, which is 0.10 x (1 - 0.113 - 0.068) x 0.90 = 7.37% of GROSS, so 0.0737 is used. It shipped as 0.074 on 2026-08-11 and was corrected the same day: rounding UP was the one value in the 69-country set that went the generous way, and the house rule is to err less favorable to the country; applying 0.10 to gross would overstate the ceiling by about a quarter. earnings_cap omitted as redundant (max binds first, around EUR 511k of gross). The EUR 4,710 floor has no engine key and omitting it is the conservative direction. 5-year carry-forward is unmodelable." } }, "lyon": { "non_cash.parental_leave_paid_pct": { "source": "service-public.gouv.fr F2265 (indemnites journalieres de maternite) https://www.service-public.gouv.fr/particuliers/vosdroits/F2265 ; F39685 (conge supplementaire de naissance) https://www.service-public.gouv.fr/particuliers/vosdroits/F39685", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 37 weeks into non_cash.parental_leave_capped (the 83 blend up to the IJ daily ceiling), so no percentage-paid weeks remain on this field. Was 83 with three caps ignored per the old field convention." }, "non_cash.parental_leave_weeks": { "source": "Conge supplementaire de naissance, service-public.gouv.fr https://www.service-public.gouv.fr/particuliers/vosdroits/F39685 (verifie le 03 juin 2026) ; conge de maternite https://www.service-public.gouv.fr/particuliers/vosdroits/F2265 ; conge de paternite et d'accueil de l'enfant, 25 jours calendaires (4 employer-paid + 21 CPAM) https://www.service-public.gouv.fr/particuliers/vosdroits/F3156 ; conge parental d'education https://www.service-public.gouv.fr/particuliers/vosdroits/F2280", "last_updated": "2026-08-01", "confidence": "high", "note": "37 wk = conge de maternite 16 (mother) + conge de paternite et d'accueil 3.6 (25 calendar days, father) + conge supplementaire de naissance 17.4 (2 months per parent x 2). Was 16, maternity alone. MAJOR CHANGE IN FORCE: the conge supplementaire de naissance (CSN) applies from 1 July 2026 to children born on or after 1 January 2026, giving each parent 1 or 2 months as a personal non-transferable right, so it SUMS across both parents. It must be taken AFTER maternity and paternity leave, so those weeks stack rather than offset. PreParE is EXCLUDED from the headline under the overlap rule: service-public states the CSN cannot be cumulated with PreParE for the same period, and the CSN pays roughly six times more, so the weeks are counted once at the higher rate. Excluded: the 3 jours ouvrables of the OLD employer-paid conge de naissance (0.5 wk of noise, and omitting it is the less favourable reading); the conge parental d'education, which service-public confirms is unpaid absent a collective agreement; multiple-birth uplifts." }, "childcare.after_school_monthly": { "source": "Ville de Lyon, Tableau tarifs periscolaires 2026 (tranche 7, QFM > 2,100 EUR or non-Lyonnais, the top tranche): pause du midi incl. meal EUR 7.30/day; accueil du matin free at every tranche; PEPS du soir Forfait Max EUR 420/year; ateliers du mercredi EUR 212/year https://www.lyon.fr/sites/lyonfr/files/content/documents/2026-04/Tableau%20tarifs%20p%C3%A9riscolaires%202026.pdf", "last_updated": "2026-08-01", "confidence": "high", "note": "French communes price periscolaire and the cantine off ONE tariff schedule that covers maternelle and elementaire alike, so the school-age rate is set equal to the maternelle rate rather than left at a separate legacy value. Aligned 2026-08-01 with the same sourced annual total. Was 200." }, "childcare.free_tier_start_age": { "source": "service-public.gouv.fr F1898 (Direction de l'information legale et administrative): \"L'instruction est obligatoire pour tous les enfants, francais et etrangers, a partir de 3 ans et jusqu'a l'age de 16 ans revolus\" (loi du 26 juillet 2019 pour une ecole de la confiance) https://www.service-public.fr/particuliers/vosdroits/F1898 ; education.gouv.fr https://www.education.gouv.fr/l-inscription-l-ecole-maternelle-8651", "last_updated": "2026-08-01", "confidence": "high", "note": "Ecole maternelle runs 3-6, is free, and has been COMPULSORY since 2019, so participation at 3-5 is at or above 96% (Eurostat). Creche (0-3) is paid. Before this, ages 1-5 all billed one rate, so a 1-year-old in creche was billed the maternelle wraparound rate." }, "tax.federal_brackets": { "source": "impots.gouv.fr barème IR 2026 (indexed +1.8%)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.csg_crds_rate": { "source": "service-public.fr CSG 9.2% + CRDS 0.5%", "last_updated": "2026-05-16", "confidence": "high" }, "tax.social_security_rate": { "source": "URSSAF employee-side contributions excl. CSG/CRDS: vieillesse 7.30% + retraite complémentaire T1 ~4.01% ≈ 11.3%", "last_updated": "2026-07-11", "confidence": "high", "note": "Corrected 2026-07-11: previous 0.22 blended figure double-counted CSG/CRDS (which the engine adds separately via csg_crds_rate). IR barème now also runs on the net imposable base in computeFranceTaxes." }, "tax.sales_tax_rate": { "source": "TVA standard 20% (economie.gouv.fr)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Métropole de Lyon taxe foncière effective ~0.7% (higher than Paris due to Lyon foncière increase 2023)", "last_updated": "2026-05-16", "confidence": "medium" }, "housing._summary": { "source": "SeLoger + MeilleursAgents Lyon Q1 2026 + Numbeo Lyon", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Assurance Maladie + Lyon mutuelle averages", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "CAF crèche tarif + école maternelle gratuite age 3+", "last_updated": "2026-05-16", "confidence": "high" }, "food._summary": { "source": "Numbeo Lyon 2026 + INSEE", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "TCL Lyon abonnement mensuel €76 (Jan 2026)", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "INSEE + Numbeo Lyon entertainment", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "LYS-SEA/LAX 1-stop via CDG/AMS 2025-26", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "CAF allocations familiales", "last_updated": "2026-05-16", "confidence": "medium", "note": "French allocations familiales: means-tested/income-modulated (from the 2nd child), retaining a reduced high-income tier; modeled value is the high-income modulated rate." }, "non_cash._summary": { "source": "Code du travail (same as Paris) + Lyon INSEE commute", "last_updated": "2026-05-16", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Lyon: EUR 44,293, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 44,293 / 0.867 = USD 51,088; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2044, R5-rounded down to 0.2. REPLACES 0.48 (2026-08-11), and the interim 0.25 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.48 was 0.8727 of Paris's 0.55; that ratio x Paris's re-derived 0.29 = 0.2531, R5-rounded down to 0.25. Paris's own figure: EUR 64,372 / 0.867 = USD 74,247, / USD 250,000 = 0.2970. This is a BASIS move, not a re-measurement: Lyon's rank inside France is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): ~15% below Paris. Index 0.48" }, "childcare.subsidy_per_kid_annual": { "source": "paris.fr quotient familial / cantine 2025; CAF PSU crèche", "last_updated": "2026-07-17", "note": "French childcare (crèche PSU / périscolaire + cantine) is means-tested by income/quotient familial; top-tranche households pay the full/top rate, so the modeled prices are net top-tranche costs and the means-tested subsidy is 0. Was 2,200-2,400 (a mid-income CMG/PSU level that a top earner does not receive)." }, "childcare.preschool_monthly": { "source": "Ville de Lyon, Tableau tarifs periscolaires 2026 (tranche 7, QFM > 2,100 EUR or non-Lyonnais, the top tranche): pause du midi incl. meal EUR 7.30/day; accueil du matin free at every tranche; PEPS du soir Forfait Max EUR 420/year; ateliers du mercredi EUR 212/year https://www.lyon.fr/sites/lyonfr/files/content/documents/2026-04/Tableau%20tarifs%20p%C3%A9riscolaires%202026.pdf", "note": "Lyon runs a 4-DAY week, so 144 school days. 7.30 x 144 = 1,051 + 420 soir + 212 mercredi = EUR 1,683/yr; 1683/12 = 140.25. Morning care is free in Lyon, which is why it lands below Nantes. Was 200/mo, unsourced. Free tuition, so the modelled cost is the working-parent wraparound at the TOP quotient-familial tranche: cantine plus periscolaire plus, on a 4-day-week commune, Wednesday care. Stored value is the ANNUAL total divided by 12 (computeChildcareCost multiplies by 12); French communes bill over the ~10-month school year. School days: the annee scolaire is 36 weeks (Code de l'education art. L521-1) = 144 days on a 4-day week, 162 on a 4.5-day week. Sourcing and the full per-slot grids: docs/superpowers/research/2026-08-01-france-cantine-fees.md.", "last_updated": "2026-08-01", "confidence": "high" }, "transit.statutory_employer_share_pct": { "source": "https://entreprendre.service-public.gouv.fr/vosdroits/F37900 (Code du travail L3261-2)", "last_updated": "2026-08-05", "confidence": "high", "note": "Every French employer must reimburse 50% of a public-transit subscription (prise en charge obligatoire). The engine nets this share on the pass branch and the row label discloses it. Statutory-share convention, owner decision 2026-08-04; Belgium checked and does NOT qualify (sectoral agreements, not one statute)." }, "non_cash.parental_leave_capped": { "source": "service-public.gouv.fr F2265: IJ maternite/paternite plafonnee a 104,02 EUR par jour (2026), salaire journalier de base capped at the PASS of 4 005 EUR/month https://www.service-public.gouv.fr/particuliers/vosdroits/F2265 ; F39685: CSN paid 70% month 1 / 60% month 2 of a base capped at the same PASS https://www.service-public.gouv.fr/particuliers/vosdroits/F39685", "last_updated": "2026-08-06", "confidence": "high", "note": "Conservative single blend of TWO statutory cap mechanisms: the 19.6 IJ weeks (maternity 16 + paternity 3.6) cap the PAYOUT at EUR 104.02/day, and the 17.4 CSN weeks cap the BASE at EUR 4,005/month before the 70%/60% rate (a weekly ceiling around EUR 555 to 647). We store one block, 83% up to EUR 104.02/day (EUR 728.14/wk), the best-known ceiling, which reads the CSN weeks slightly LOW, the direction less favorable to France. At EUR 90,000 the effective rate is ~42%, matching the 40-45% the 2026-08-01 note already estimated. If the engine ever supports two capped blocks, France is a candidate to split. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (France)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "DREES Fiche 21, reste a charge des menages measured after BOTH Assurance Maladie and the complementaire (EUR 292 per inhabitant, 2024), netted of dental and optical + ameli participation forfaitaire and franchises medicales, EUR 100/yr cap each", "last_updated": "2026-08-08", "confidence": "medium", "note": "DREES measures exactly our basis, which matters in France more than anywhere: the mutuelle has been employer-mandated since ANI 2016 and is a PREMIUM, so the honest French out-of-pocket figure is the residual after it. Under-18s are statutorily exempt from both participations and are covered by 100% Sante, so a child adds ~40% of an adult. All French cities are normalised onto one curve: the previous Lyon-300 against Paris/Toulouse/Nantes/Marseille-350 split had no stated basis. A Paris depassements uplift is queued, not shipped. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "DREES Fiche 21, reste a charge des menages measured after BOTH Assurance Maladie and the complementaire (EUR 292 per inhabitant, 2024), netted of dental and optical + ameli participation forfaitaire and franchises medicales, EUR 100/yr cap each", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, indicator TT1019V, greater city. https://ec.europa.eu/eurostat/api/dissemination/statistics/1.0/data/urb_ctran?indic_ur=TT1019V&format=JSON&lang=EN", "last_updated": "2026-08-08", "confidence": "low", "statistic": "Average time of journey to work", "year": "2012-2013", "note": "Urban Audit greater-city values 24.3 and 24.6. TT1019V is validated as one-way because Spain's Urban Audit values independently reproduce INE's own one-way series. CAVEAT: the FRENCH waves of this series are unreliable in places (2011 shows Nimes at 70.38 and Marseille at 65.63, bracketed by far lower values either side, which looks like a round-trip basis break, and every 2014 French value is a round integer). Geography is the core city, which understates a metro. Eurostat's LFS 2019 national mean for France is 26." }, "non_cash.sick_leave": { "source": "Indemnités journalières de sécurité sociale (IJSS): Code de la sécurité sociale; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": "https://www.echangesassurances.org/actualites/indemnites-journalieres-maladie-plafond-juillet-2026", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "service-public.gouv.fr, Plafond epargne retraite (PER) https://www.service-public.gouv.fr/particuliers/vosdroits/F14709?lang=fr ; impots.gouv.fr https://www.impots.gouv.fr/particulier/questions/que-signifie-la-somme-imprimee-dans-la-rubrique-plafond-epargne-retraite-de", "last_updated": "2026-08-11", "confidence": "high", "note": "PER individuel deduction from revenu net global, CGI art. 163 quatervicies, for versements paid in 2026 (plafond computed on 2025 income, PASS 2025 = EUR 47,100): the greater of 10% of prior-year revenus d'activite capped at 8 x PASS (EUR 37,680) or 10% of PASS (EUR 4,710). BASIS TRAP HANDLED: the statutory 10% runs on activity income NET of cotisations sociales and after the 10% abattement, which is 0.10 x (1 - 0.113 - 0.068) x 0.90 = 7.37% of GROSS, so 0.0737 is used. It shipped as 0.074 on 2026-08-11 and was corrected the same day: rounding UP was the one value in the 69-country set that went the generous way, and the house rule is to err less favorable to the country; applying 0.10 to gross would overstate the ceiling by about a quarter. earnings_cap omitted as redundant (max binds first, around EUR 511k of gross). The EUR 4,710 floor has no engine key and omitting it is the conservative direction. 5-year carry-forward is unmodelable." } }, "brisbane": { "non_cash.parental_leave_flat": { "source": "DSS Paid Parental Leave Guide 3.1: the daily NMW amount for PLP is 7.6 times the hourly NMW; from 1 July 2026 the hourly NMW is AUD 26.44, the daily amount is AUD 200.94 and the weekly rate is AUD 1,004.70 https://guides.dss.gov.au/paid-parental-leave-guide/3/1 ; Paid Parental Leave Act 2010 ss 6, 65, 11D; Fair Work Act 2009 Part 2-6", "last_updated": "2026-08-01", "confidence": "high", "note": "AUD 1,004.70 per week (AUD 200.94 per payable day x 5), in force from 1 July 2026 and reindexed each 1 July with the Fair Work Commission national minimum wage order. PLP is taxable. Annualised that is about AUD 52,244, which against a typical Australian tech salary is roughly a 29% replacement rate and falls as pay rises. TWO ITEMS DELIBERATELY NOT IN THIS FIGURE. First, the PPL superannuation contribution for births from 1 July 2025 adds 12% of the PLP amount plus nominal interest, paid by the Commissioner of Taxation into the parent super fund, which raises the cash value but adds no weeks. Second, an income-test CLIFF: a claimant qualifies if EITHER their own adjusted taxable income is at or under AUD 186,487 OR the couple combined is at or under AUD 386,525 (2025-26 reference year, indexed each 1 July). A household on AUD 250,000 qualifies on the family test however it splits, but two earners at AUD 200,000 each fail both tests and receive NOTHING, because PLP is all-or-nothing rather than tapered. ATI includes RSUs and reportable super, which pushes tech households toward that cliff. https://guides.dss.gov.au/paid-parental-leave-guide/2/3" }, "non_cash.parental_leave_paid_pct": { "source": "DSS Paid Parental Leave Guide 3.1, Current rate of PLP https://guides.dss.gov.au/paid-parental-leave-guide/3/1", "last_updated": "2026-08-01", "confidence": "high", "note": "0 because Australia has NO percentage-paid weeks. PLP is paid at the national minimum wage as a flat daily amount, so the stored 100 was a fabricated headline rate for a benefit that never tracked salary. The flat amount lives in non_cash.parental_leave_flat and the rate is derived at the household own salary. Never display this 0 on its own." }, "non_cash.parental_leave_weeks": { "source": "DSS Paid Parental Leave Guide 2.9, Maximum PLP entitlements (Guide v1.99, released 1 July 2026) https://guides.dss.gov.au/paid-parental-leave-guide/2/9 ; Paid Parental Leave Act 2010 ss 31AB, 31ABA", "last_updated": "2026-08-01", "confidence": "high", "note": "26 wk = 130 payable days at 5 payable weekdays per PLP week. Was 22, which is the 110-day rule for births from 1 July 2024, so the repo was two annual steps stale. The statutory ladder is 100 days from 1 Jul 2023, 110 from 1 Jul 2024, 120 from 1 Jul 2025 and 130 from 1 Jul 2026, and a child born today attracts 130. COUNTED ONCE, NOT SUMMED: PLP is a single family entitlement for the child (the maximum number of days that can be received for a child for all claimants for the child), and for a partnered claimant the 20 reserved days are carved OUT of the 130 rather than added on top, so two parents claiming together still receive 130 between them. Excluded: the 12 to 24 months of Fair Work Act unpaid parental leave; reserved-period exemptions for ADF deployment or a deceased partner. Dad and Partner Pay no longer exists separately, having been absorbed into PLP in July 2023." }, "tax.filing_basis": { "source": "PwC Worldwide Tax Summaries - Australia (Individual, Tax administration): no joint assessment or joint filing in Australia; spouses assessed individually.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "FY2026-27 stage: bracket-2 15% from 2026-07-01, drops to 14% 2027-07-01 (ato.gov.au personal income tax cuts); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.medicare_levy_rate": { "source": "ATO Medicare Levy 2% standard rate", "last_updated": "2026-05-16", "confidence": "high" }, "tax.sales_tax_rate": { "source": "ATO GST 10% (national)", "last_updated": "2026-05-16", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Brisbane City Council rates schedule 2025-26", "last_updated": "2026-05-16", "confidence": "medium" }, "housing._summary": { "source": "Domain & realestate.com.au Brisbane Q1 2026 + CoreLogic", "last_updated": "2026-05-16", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Medicare + privatehealth.gov.au QLD tier 2026", "last_updated": "2026-05-16", "confidence": "high" }, "childcare._summary": { "source": "ACCC Childcare Inquiry + Care for Kids Brisbane; CCS net", "last_updated": "2026-05-16", "confidence": "medium" }, "food._summary": { "source": "Numbeo Brisbane 2026 + ABS HES adjusted", "last_updated": "2026-05-16", "confidence": "high" }, "transit._summary": { "source": "TransLink Go Card Zone 1 daily cap AUD 11.40; RACQ ownership cost", "last_updated": "2026-05-16", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Brisbane entertainment + Tourism Research Australia", "last_updated": "2026-05-16", "confidence": "medium" }, "travel_home._summary": { "source": "Skyscanner/Google Flights BNE-SEA/LAX (one-stop via SYD)", "last_updated": "2026-05-16", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Services Australia FTB Part A typical band", "last_updated": "2026-05-16", "confidence": "medium" }, "non_cash._summary": { "source": "Fair Work Act NES + QLD holidays incl. Brisbane Show Day", "last_updated": "2026-05-16", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Brisbane: AUD 122,187, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "AUD 122,187 / 1.42 = USD 86,047; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3442, R5-rounded down to 0.34. REPLACES 0.65 (2026-08-11), and the interim 0.39 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.65 was 0.8667 of Sydney's 0.75; that ratio x Sydney's re-derived 0.45 = 0.3900, R5-rounded down to 0.39. Sydney's own figure: AUD 161,818 / 1.42 = USD 113,956, / USD 250,000 = 0.4558. This is a BASIS move, not a re-measurement: Brisbane's rank inside Australia is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 30% cheaper than Sydney. Index 0.65" }, "childcare.subsidy_per_kid_annual": { "source": "servicesaustralia.gov.au Child Care Subsidy 2025-26 (90% up to 85,279; 0% at 535,279)", "last_updated": "2026-07-17", "note": "Child Care Subsidy at the solved household: CCS% = 90 - (income - 85,279)/5,000. At the median published brisbane household (~424k -> 22%) applied to per-kid gross childcare. CCS is means-tested; it survives partially at these incomes (unlike FTB, which is 0). The audit's 36-48% guess assumed a lower household; the engine solves brisbane higher. Was a flat over-covering 8,500-10,000." }, "healthcare.statutory_employee": { "source": "ATO, Medicare levy: 2% of taxable income with no upper income limit (Medicare Levy Act 1986). ATO Medicare levy surcharge income thresholds and rates, 2026-27: singles AUD 105,000 / 123,000 / 164,000 and families AUD 210,000 / 246,000 / 328,000 at 1% / 1.25% / 1.5%, with the family threshold rising AUD 1,500 per dependent child after the first (2025-26 was 101,000 / 118,000 / 158,000 and 202,000 / 236,000 / 316,000). The base levy is assessed individually; the surcharge tests combined income for couples.", "last_updated": "2026-07-31", "confidence": "medium", "note": "DISPLAY ONLY. NOT capped, and that is the load-bearing fact for Australia: it is one of very few systems in this dataset where the health contribution never flattens. BASE MISMATCH: the levy is charged on taxable income, while computeStatutoryHealthContribution applies the rate to gross, so the displayed figure runs a few percent above the charged one. That is the conservative direction and is the same base issue flagged for every country in this rollout. CAVEAT: ato.gov.au and legislation.gov.au returned 403 to every fetch attempt this session, so the 2% rate and the 2026-27 surcharge tiers were read from indexed snippets of the live ATO pages and cross-checked against two independent professional tax references, hence medium confidence. The 2% rate itself is long-standing and uncontested. 2026-27 Medicare levy LOW-INCOME thresholds were not confirmed (historically legislated retroactively); irrelevant at ICP income." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "AIHW health expenditure 2023-24 individual out-of-pocket, AUD 1,634 per person and ~1,338 excluding dental + PBS co-payment and the family-pooled PBS Safety Net", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 RAISED from 800 on primary-source evidence, the largest upward level move outside Austria. Medicare and the PBS are national with no state benefit variation, and the previous Brisbane-versus-rest spread had no documented basis in _meta or the city JSONs, so it is removed. Private hospital cover is a PREMIUM and stays in the premium field. Children add ~30% of an adult: frequently bulk-billed, lower PBS costs, and the family-pooled Safety Net pulls the whole household toward the concessional co-pay. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "AIHW health expenditure 2023-24 individual out-of-pocket, AUD 1,634 per person and ~1,338 excluding dental + PBS co-payment and the family-pooled PBS Safety Net", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "HILDA Survey (Melbourne Institute), average daily commuting time by capital city, 2017 wave. https://theconversation.com/australian-city-workers-average-commute-has-blown-out-to-66-minutes-a-day-how-does-yours-compare-120598", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average daily commuting time (round trip)", "year": "2017", "note": "CONVERSION: 67 min/day round trip / 2 = 33.5 one way. The previous 28 was an error against this same table." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "ATO, Contributions caps https://www.ato.gov.au/tax-rates-and-codes/key-superannuation-rates-and-thresholds/contributions-caps ; Maximum contributions base https://www.ato.gov.au/businesses-and-organisations/super-for-employers/paying-super-on-payday/what-payments-are-qualifying-earnings/maximum-contributions-base ; Super guarantee https://www.ato.gov.au/tax-rates-and-codes/key-superannuation-rates-and-thresholds/super-guarantee", "last_updated": "2026-08-11", "confidence": "high", "note": "Concessional contributions cap AUD 32,500 for 2026-27, a CONTRIBUTION ceiling shared with compulsory employer Super Guarantee, so the employee has only what SG leaves: 12% of gross up to the AUD 270,830 maximum contribution base (annual from 1 July 2026 under Payday Super). Room is zero at and above 270,830. The 5-year carry-forward is deliberately ignored. SEPARATE, UNFIXED: concessional contributions are taxed 15% inside the fund (30% under Division 293 above AUD 250,000), so treating the allowed portion as fully pre-tax still flatters Australia." } }, "stavanger": { "tax.inbound_regime": { "source": "PAYE / kildeskatt pa lonn, skatteloven kapittel 20. Flat 25% that INCLUDES trygdeavgift (Skatteetaten rettskilder: 'Denne skattesatsen inkluderer trygdeavgift'); first year of tax residency only; 2026 income ceiling NOK 725,050 (2025: NOK 697,150). Skatteloven s 20-3 makes the ceiling a hard cliff on the WHOLE income, not a band; s 20-2 bans every deduction. https://www.skatteetaten.no/person/utenlandsk/skal-du-arbeide-i-norge/skattekort/kildeskatt-pa-lonn/ ; https://lovdata.no/dokument/NL/lov/1999-03-26-14/KAPITTEL_20", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: max_income refreshed 697,150 (2025) -> 725,050 (2026). Still salary_relief false and no flat_rate, so inboundRegimeStatus returns note_only at every salary and no math moves. The note-only stance is now doubly defensible: professional salaries clear the cliff, and below it the total deduction ban frequently makes the scheme worse than ordinary tax for a household with a Norwegian mortgage. The forskrift under s 20-7 that sets the 2026 ceiling was not retrieved; the figure is published by Skatteetaten in both language versions." }, "tax.filing_basis": { "source": "Skatteetaten (skatteetaten.no) - spouses assessed individually on employment income (tax classes abolished 2018)", "last_updated": "2026-07-11", "note": "Activates the per-filer split in compute() for two-earner households." }, "tax.property_tax_rate_pct": { "source": "Stavanger kommune eiendomsskatt", "last_updated": "2026-05-16", "confidence": "high", "note": "Effective 0.20 after bunnfradrag NOK 1.2M" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Stavanger: NOK 713,543, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "NOK 713,543 / 9.49 = USD 75,189; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3008, R5-rounded down to 0.3. REPLACES 0.54 (2026-08-11), and the interim 0.37 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Sample may lean petroleum-sector IT. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.54 was 0.9474 of Oslo's 0.57; that ratio x Oslo's re-derived 0.4 = 0.3789, R5-rounded down to 0.37. Oslo's own figure: NOK 968,721 / 9.49 = USD 102,078, / USD 250,000 = 0.4083. This is a BASIS move, not a re-measurement: Stavanger's rank inside Norway is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Oil-capital premium pushes above Bergen/Trondheim Rebased 2026-08-11 with its country anchor (comp-index sweep A moved Oslo, Stockholm, Tokyo and Singapore); the documented ratio to the anchor is preserved, so this note's relative claim still holds." }, "housing._summary": { "source": "Numbeo Housing (cost-of-living basket) https://www.numbeo.com/cost-of-living/in/Stavanger", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation (90/10 quality bar). Per-field values were researched by subagents from local marketplaces (Zillow/Hemnet/Funda/Rightmove/etc.); cite each field individually at next refresh pass." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Universal (folketrygden)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation." }, "childcare._summary": { "source": "Numbeo (preschool) https://www.numbeo.com/cost-of-living/in/Stavanger + government social services for subsidies", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Subsidy fields require country-specific authority per Phase 4 conventions." }, "food._summary": { "source": "Numbeo Markets + Restaurants https://www.numbeo.com/cost-of-living/in/Stavanger", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Numbeo runs 12-28% high vs official statistics; acceptable for golden test ±30% tolerance." }, "transit._summary": { "source": "Local transit authority (per-city schedule)", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Monthly pass + car cost researched per city." }, "discretionary._summary": { "source": "Numbeo Sports & Leisure + dining https://www.numbeo.com/cost-of-living/in/Stavanger", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Gym, cinema, restaurants from Numbeo basket." }, "travel_home._summary": { "source": "Google Flights / Kayak sampled fares 2025-2026", "last_updated": "2026-05-16", "confidence": "medium", "note": "Round-trip economy mid-range estimate." }, "govt_benefits._summary": { "source": "NAV barnetrygd", "last_updated": "2026-05-16", "confidence": "medium" }, "non_cash._summary": { "source": "Ferieloven + NAV foreldrepenger/sykepenger", "last_updated": "2026-05-16", "confidence": "medium", "note": "Group-level citation. Vacation, holidays, sick leave, parental leave per country." }, "tax.national_insurance_rate": { "confidence": "high", "source": "Trygdeavgift salary rate 7.6% 2026 (skatteetaten.no/en/rates/national-insurance-contributions); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "tax.trinnskatt_brackets": { "confidence": "high", "source": "Trinnskatt 2026 thresholds/rates (skatteetaten.no/en/rates/bracket-tax); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "childcare.infant_monthly": { "source": "regjeringen.no / Oslo kommune barnehage maksimalpris 2026 (Stortinget vedtatt)", "last_updated": "2026-07-17", "note": "2026 barnehage maksimalpris NOK 1,200/mo charged 11 months/yr + matpenger ~NOK 320/mo: effective (1,200 x 11/12) + 320 ~= NOK 1,420/mo. One age-independent cap covers infants and preschoolers (was 3,000, the pre-2024 level)." }, "childcare.preschool_monthly": { "source": "regjeringen.no / Oslo kommune barnehage maksimalpris 2026", "last_updated": "2026-07-17", "note": "Same national maksimalpris as infants: effective ~NOK 1,420/mo (maksimalpris NOK 1,200 x11 + matpenger ~320)." }, "govt_benefits.child_allowance_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-17", "note": "Norwegian barnetrygd: universal, NOT means-tested." }, "tax.personfradrag": { "source": "Personfradrag 2026 NOK 114,540: regjeringen.no statsbudsjettet 2026 (108,550 -> 114,540) / skatteetaten.no/en/rates/personal-allowance", "source_name": "Skatteetaten / statsbudsjettet 2026", "last_updated": "2026-07-23", "confidence": "high", "note": "Applied INSIDE computeNorwayTaxes to the 22% alminnelig-inntekt base ONLY, together with minstefradrag. Per filer (Norway files individually). standard_deduction_* held at 0 so compute() does not subtract a lump across all three components. Correct-first item #1, 2026-07-17 resolutions doc." }, "tax.minstefradrag_rate": { "source": "statsbudsjettet 2026 (regjeringen.no) + skatteetaten.no 2026 rates: minstefradrag wage rate 46% (skatteetaten.no/en/rates/minimum-standard-deduction)", "source_name": "Skatteetaten", "last_updated": "2026-07-23", "confidence": "high", "note": "Minstefradrag = 46% of wage income, applied to the 22% base only. Capped (see minstefradrag_cap)." }, "tax.minstefradrag_cap": { "source": "statsbudsjettet 2026 (regjeringen.no) + skatteetaten.no 2026 rates: minstefradrag wage upper limit NOK 95,700 for 2026 (2025: 92,000)", "source_name": "Skatteetaten / statsbudsjettet 2026", "last_updated": "2026-07-23", "confidence": "high", "note": "Upper limit on the wage minstefradrag. Binds at any ICP salary. Sum with personfradrag = 210,240 on the 22% base." }, "tax.standard_deduction_single": { "source": "statsbudsjettet 2026 (regjeringen.no) + skatteetaten.no 2026 rates", "source_name": "Skatteetaten", "last_updated": "2026-07-23", "confidence": "high", "note": "Held at 0 BY DESIGN. Norway applies minstefradrag + personfradrag inside computeNorwayTaxes, to the 22% base only; a lump here would wrongly shrink trinnskatt and trygdeavgift too (the pre-2026-07-23 bug). See tax.personfradrag." }, "tax.standard_deduction_married": { "source": "statsbudsjettet 2026 (regjeringen.no) + skatteetaten.no 2026 rates", "source_name": "Skatteetaten", "last_updated": "2026-07-23", "confidence": "high", "note": "Held at 0 BY DESIGN. Norway applies minstefradrag + personfradrag inside computeNorwayTaxes, to the 22% base only; a lump here would wrongly shrink trinnskatt and trygdeavgift too (the pre-2026-07-23 bug). See tax.personfradrag." }, "non_cash.parental_leave_weeks": { "source": "NAV https://www.nav.no/foreldrepenger ; folketrygdloven kap. 14", "last_updated": "2026-08-06", "confidence": "high", "note": "49 wk at 100% RE-CONFIRMED 2026-08-06 against NAV: 3 wk reserved for the mother pre-birth + modrekvote 15 + fedrekvote 15 + fellesperiode 16, ONE family entitlement with quotas inside it, counted once. The 80% option is now 61 wk 1 day (extended from the old 59 so the two elections carry near-equal total value); 49 at 100 is the default and conservative pick. REFUTATION recorded: the father's 2 omsorgspermisjon weeks at birth are statutorily unpaid (employer payment is practice, not statute) and add no paid weeks; fathers may start fedrekvote at birth but that draws down the 49. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md." }, "non_cash.parental_leave_paid_pct": { "source": "NAV https://www.nav.no/foreldrepenger : calculation basis capped at 6G; G = NOK 136,549 from 1 May 2026, 6G = NOK 819,294", "last_updated": "2026-08-17", "confidence": "high", "note": "We store 0: the 2026-08-17 caps extension moved all 49 weeks into non_cash.parental_leave_capped (100% of a basis capped at 6G), so no percentage-paid weeks remain on this field. Was 100 with the cap recorded here but unconsumed, which left /rankings/parental-leave-by-country/ printing Norway at a flat 100% at any salary (raised 2026-08-08 and 2026-08-11, fixed cluster leave-caps-0817)." }, "non_cash.parental_leave_capped": { "source": "folketrygdloven para 14-7 (foreldrepenger basis follows the sykepenger rules, capped at 6G per para 8-10); NAV https://www.nav.no/grunnbelopet : G = NOK 136,549 per 1 May 2026, 6G = NOK 819,294/yr (fetched 2026-08-17)", "last_updated": "2026-08-17", "confidence": "high", "note": "All 49 weeks pay 100% of a basis capped at 6G = NOK 819,294/yr. cap_amount is stored per week (819,294 / 52 = NOK 15,755.65) on the engine weekly-gross convention, the same pre-conversion Stockholm uses for its annual SGI ceiling; foreldrepenger dagsats runs Mon-Fri (5-day week, divisor 260), so annual/52 equals dagsats x 5 exactly. Above 6G the payment is flat: effective rate = 819,294 / salary, so the stored Oslo tech p50 of NOK 968,721 yields 84.6% and a $150,000-equivalent (about NOK 1.41M) yields about 58%. The 80% election (61 wk 1 day) is deliberately not stored: 49 wk at 100% is the default and the conservative pick, per the 2026-08-06 batch2 adjudication. UNIT WATCH: G is an annual figure revised every May 1; re-derive cap_amount each trygdeoppgjor. Research: docs/superpowers/research/2026-08-17-parental-leave-caps.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Helsedirektoratet and Helfo egenandelstak + NAV, 1,452,000 residents reached the combined tak in 2023", "last_updated": "2026-08-08", "confidence": "medium", "note": "REPLACES an exact statutory maximum: NOK 3,040 is precisely the 2023 egenandelstak (series 2,460 / 2,921 / 3,040 / 3,165 / 3,278). 26% of the population reaches the cap and ~1m under-16s pay zero, so the realized adult mean is ~NOK 1,650. Children pay no egenandel under 16, extended to all under-18s from 2026-08-01. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Helsedirektoratet and Helfo egenandelstak + NAV, 1,452,000 residents reached the combined tak in 2023", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Jernbanedirektoratet / RVU, Reisevaner i de 7 storste byregionene 2022, Figur 23, Nord-Jaeren. https://www.jernbanedirektoratet.no/content/uploads/sites/4/2023/11/Reisevaner-i-de-7-storste-byregionene-med-tilleggsutvalg-2022.pdf", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Reiselengde i tid til/fra arbeid", "year": "2022", "note": "Nord-Jaeren covers Stavanger, Sandnes, Sola and Randaberg, n=2755. Band-to-mean gives 25.1." }, "non_cash.sick_leave": { "source": "Folketrygdloven (National Insurance Act) kapittel 8 , §8-2 (opptjeningstid), §8-10; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.nav.no/grunnbelopet", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Skatteetaten, Forskuddsutskrivingen 2026 https://www.skatteetaten.no/en/rettskilder/type/uttalelser/uttalelser/forskuddsutskrivingen-2026/", "last_updated": "2026-08-11", "confidence": "high", "note": "IPS (individuell pensjonssparing), NOK 25,000 for 2026, raised from 15,000 under the 2024 pension settlement; old-IPS and new-IPS share one ceiling. Employee-only, and the source states the same figure as both the payment limit and the deduction limit. KNOWN RESIDUAL, not fixed here: the IPS deduction is against alminnelig inntekt (22% flat) ONLY and must not reduce trinnskatt or trygdeavgift, but the engine nets the retirement deduction off personinntekt upstream, so it still reaches too many bases." } }, "utrecht": { "non_cash.parental_leave_paid_pct": { "source": "WAZO art. 4:2 (geboorteverlof met behoud van loon, employer-paid) https://wetten.overheid.nl/BWBR0013008", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the 1-week partner geboorteverlof, employer-paid wage continuation with no dagloon ceiling (the structural twin of Germany Mutterschutz). The other 39 UWV-paid weeks moved to non_cash.parental_leave_capped in the 2026-08-06 caps rebuild; this field was 83, the whole-blend headline with the maximumdagloon ignored." }, "non_cash.parental_leave_weeks": { "source": "Wet arbeid en zorg (WAZO) art. 3:1, 4:2, 4:2a, 6:2, 6:3 https://wetten.overheid.nl/BWBR0013008 (geldig van 01-01-2025) ; Rijksoverheid https://www.rijksoverheid.nl/vraag-en-antwoord/ouderschapsverlof/recht-op-ouderschapsverlof ; business.gov.nl https://business.gov.nl/regulation/leave-schemes/", "last_updated": "2026-08-01", "confidence": "high", "note": "40 wk = zwangerschaps- en bevallingsverlof 16 (mother) + geboorteverlof 1 (partner) + aanvullend geboorteverlof 5 (partner) + betaald ouderschapsverlof 9 per parent x 2. WAZO gives ouderschapsverlof to each parent individually with no transfer mechanism (Beide ouders hebben recht op ouderschapsverlof). Was 9, which held the paid parental weeks for ONE parent and dropped 16 weeks of maternity at 100%. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Excluded: the 17 unpaid weeks per parent of the 26-week ouderschapsverlof, and the 20-week multiples maternity variant." }, "tax.inbound_regime": { "source": "30%-regeling (expatregeling). Belastingdienst Deskundigheidsvereiste, 2026 norm: a 'belastbaar jaarloon ... meer is dan EUR 48.013 (exclusief de gerichte vrijstelling)', i.e. measured on the salary REMAINING after the 30% carve-out, so about EUR 68,590 of gross: https://www.belastingdienst.nl/wps/wcm/connect/bldcontentnl/belastingdienst/zakelijk/internationaal/personeel/u_bent_niet_in_nederland_gevestigd_loonheffingen_inhouden/als_u_loonheffingen_gaat_inhouden/extraterritoriale_kosten_en_de_30procentregeling/voorwaarden_voor_de_30procentregeling1/deskundigheidsvereiste . Belastingdienst Inhoud van de expatregeling: maximum tax-free allowance EUR 78,600 in 2026, reached at the WNT norm of EUR 262,000. Max 60 months. Rijksoverheid, Belastingvoordeel buitenlandse werknemers.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31 OVER-GRANT FIX + refresh. (1) min_income_basis set to post_relief: the EUR 48,013 floor was being tested against GROSS, so every salary from EUR 48,013 up got the ruling; the statute measures it on the taxable remainder, which needs about EUR 68,590 of gross. Amsterdam single filer at EUR 60,000: tax 7,056 -> 15,906 (ordinary). (2) This entry previously cited the 2025 pair 46,660 / 246,000 while the data already carried the 2026 pair 48,013 / 262,000; the data was right and the prose was stale. (3) The under-30-with-a-master norm of EUR 36,497 is deliberately NOT used: we model neither age nor degree, so we carry the higher general norm. SCHEDULED: allowance drops 30% -> 27% and the norms step up from 2027-01-01 (Belastingplan 2025, second nota van wijziging). The calculator models the current tax year. TODO (2027-01 refresh): apply the 27% and the stepped-up norms once in force for the modeled year." }, "tax.filing_basis": { "source": "PwC WWTS - Netherlands (Tax administration): spouses are taxed separately for employment/business income (Box 1); only Box 2/3 and home items are allocable between fiscal partners.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.box1_brackets": { "source": "Belastingdienst Box 1 2026: peer-identical with Amsterdam", "last_updated": "2026-05-25", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Gemeente Utrecht OZB 2026 ~0.0895% of WOZ value", "last_updated": "2026-05-25", "confidence": "high" }, "tax.sales_tax_rate": { "source": "BTW standard rate 21%", "last_updated": "2026-05-25", "confidence": "high" }, "housing._summary": { "source": "Pararius / Funda Utrecht Q1 2026", "last_updated": "2026-05-25", "confidence": "high", "note": "Randstad university city; rents above national average, below Amsterdam." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Zorgverzekering basisverzekering (Zvw) ~EUR 150/mo; kids <18 free", "last_updated": "2026-05-25", "confidence": "high" }, "childcare._summary": { "source": "Kinderopvang ~EUR 10.71/hr 2026 cap; kinderopvangtoeslag nets ~EUR 9,500/kid for mid tech income", "last_updated": "2026-05-25", "confidence": "medium" }, "food._summary": { "source": "Numbeo Utrecht + CBS 2026", "last_updated": "2026-05-25", "confidence": "medium" }, "transit._summary": { "source": "U-OV / GVU regional pass ~EUR 100/mo", "last_updated": "2026-05-25", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo Utrecht + CBS", "last_updated": "2026-05-25", "confidence": "medium" }, "travel_home._summary": { "source": "AMS->SEA/LAX 1-stop KLM/Delta transatlantic 2025-26", "last_updated": "2026-05-25", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Algemene Kinderbijslagwet art. 12, consolidated text valid from 01-08-2026, https://wetten.overheid.nl/BWBR0002368/ (fetched 2026-09-12, two independent reads agree; svb.nl returned 403 in English and Dutch and rijksoverheid.nl 404 on 2026-09-10 and 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "Dutch kinderbijslag: universal, NOT means-tested (the means-tested kindgebonden budget is separate and excluded). Statutory amounts per QUARTER per child, band set by the child's age on the first day of the quarter: EUR 298.40 under 6, 362.35 for 6 to 11, 426.29 for 12 to 17 (AKW art. 12 lid 3). The engine has no per-age benefit hook (computeGovtBenefits() reads one rate and only tests age < 18), so ONE rate is stored for every age and the Rule 5 tiebreak picks the LOWEST band: 298.40 x 4 = 1,193.60 a year, stored as 1,194. That understates a 6-to-11-year-old by EUR 255 a year and a 12-to-17-year-old by EUR 511, conservative against the Netherlands. Set 2026-09-12; was 1,368 = 342.06 x 4, which is the 6-to-11 band of the 2025 H1 table mislabelled as 2026. Ledger 2026-09-12; research docs/superpowers/research/2026-09-12-child-benefit-audit-fixes.md F4." }, "non_cash._summary": { "source": "Dutch labour law: 20 statutory vacation days, 26wk parental leave (9 paid ~70% via UWV)", "last_updated": "2026-05-25", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Utrecht: EUR 78,040, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 78,040 / 0.867 = USD 90,012; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3600, R5-rounded down to 0.36. REPLACES 0.64 (2026-08-11), and the interim 0.41 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.64 was 0.8889 of Amsterdam's 0.72; that ratio x Amsterdam's re-derived 0.47 = 0.4178, R5-rounded down to 0.41. Amsterdam's own figure: EUR 102,818 / 0.867 = USD 118,591, / USD 250,000 = 0.4744. This is a BASIS move, not a re-measurement: Utrecht's rank inside Netherlands is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.64 vs Amsterdam 0.72." }, "childcare.subsidy_per_kid_annual": { "source": "rijksoverheid.nl kinderopvangtoeslag 2025; Belastingdienst toeslag tables", "last_updated": "2026-07-17", "note": "kinderopvangtoeslag at the 33.3% high-income floor (2025, households >= EUR 226,546): 33.3% x capped rate EUR 10.71/hr x ~1,770 reimbursable h/yr ~= EUR 6,300 per full-time daycare child. The reimbursement % is means-tested, but 33.3% is the statutory MINIMUM, so it survives at the EUR 226k-269k published NL household (was overstated at a mid-income % of ~60-70%). 2026 raises the floor to 36.5%." }, "non_cash.parental_leave_capped": { "source": "UWV maximumdagloon: EUR 309,91 bruto per dag from 1 July 2026 (Wfsv art. 17 lid 1) https://www.uwv.nl/nl/premies-bedragen/maximum-dagloon ; UWV: aanvullend geboorteverlof and betaald ouderschapsverlof pay 70% van het (gemaximeerde) dagloon https://www.uwv.nl/nl/aanvullend-geboorteverlof/hoogte-uitkering-aanvullend-geboorteverlof-betaald-ouderschapsverlof ; WAZO arts. 3:13, 4:2b, 6:3", "last_updated": "2026-08-06", "confidence": "high", "note": "PRE-CONVERTED to a week cap per the calendar-day convention: the dagloon is a 5-WORKDAY wage cap applied BEFORE the percentage, so we store the weekly PAYOUT ceiling at the 70% tier, 309.91 x 0.70 x 5 = EUR 1,084.69/week, over all 39 UWV weeks at the 83 blend (16 wk maternity 100% + 23 wk at 70%). Conservative: the 16 maternity weeks actually cap higher (100% tier, EUR 1,549.55/wk), and pricing the whole block at the lower ceiling reads them LOW. Cap steps semi-annually with the minimum wage (1 Jul 2026 value stored). At EUR 90,000 the blended effective rate is ~64%. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Netherlands)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Zvw statutory eigen risico (EUR 385 per adult, under-18s exempt) + Zorginstituut Nederland and Vektis data on what share of insured actually exhaust it", "last_updated": "2026-08-08", "confidence": "medium", "note": "REPLACES an exact statutory maximum: EUR 385 is precisely the single-adult compulsory deductible, and only about half of insured adults exhaust it, so the realized figure is EUR 229-236 by two converging routes. The deductible is charged PER ADULT and under-18s are wholly exempt, and GP care, maternity care and district nursing sit outside it entirely, so a Dutch family of four is essentially two adults. Aanvullende verzekering is a PREMIUM and is not booked here. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Zvw statutory eigen risico (EUR 385 per adult, under-18s exempt) + Zorginstituut Nederland and Vektis data on what share of insured actually exhaust it", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "CBS StatLine, ODiN dataset 85055NED, motive Van en naar het werk, province Utrecht. https://opendata.cbs.nl/statline/#/CBS/nl/dataset/85055NED", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Gemiddelde reisduur per verplaatsing", "year": "2024", "note": "32.48 min. Province geography. Largest Dutch correction; the previous 25 was 7 minutes short." }, "non_cash.sick_leave": { "source": "Burgerlijk Wetboek (BW) Boek 7, Artikel 629 (loondoorbetaling bij ziekte).; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.uwv.nl/particulieren/bedragen/detail/maximumdagloon", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Belastingdienst Centraal Aanspreekpunt Pensioenen, AOW-franchises https://centraalaanspreekpuntpensioenen.belastingdienst.nl/publicaties/overzicht-aow-inbouwbedragen-en-aow-franchises/ ; V&A 25-008 https://centraalaanspreekpuntpensioenen.belastingdienst.nl/publicaties/va-25-008-v251209/ ; Kennisgroepen KG:070:2024:3 jaarruimte https://kennisgroepen.belastingdienst.nl/publicaties/kg07020243-vragen-en-antwoorden-over-de-jaarruimteberekening-en-de-opbouw-van-pensioenaanspraken/ ; ABP 2026 premie split", "last_updated": "2026-08-11", "confidence": "low", "note": "THIS VALUE IS A MODEL, NOT A STATUTE, and the formula is what is HIGH confidence. Art. 3.127 Wet IB 2001 post-Wtp: jaarruimte = 30% of premiegrondslag MINUS the actual ouderdomspensioen premium already paid (employer + employee). 2026 AOW-franchise EUR 19,172, aftoppingsgrens EUR 137,800, so the statutory maximum is 0.30 x (137,800 - 19,172) = EUR 35,589 (encoded as 13,085 here, see below). The naive statutory mapping is INERT at our incomes and would look like a fix while changing nothing. 13,085 = 11.03% of premiegrondslag, i.e. 30% less ABP's 18.97% employer share; 0.082 is the lowest in-band gross ratio (at EUR 75k), so the mapping sits at or below the derived truth everywhere above EUR 75k. Expressing this via less_employer_pct after the 2026-08-11 engine reorder was considered and rejected: it cannot subtract the AOW-franchise and would hand every Dutch earner about EUR 2,115 of extra room." } }, "eindhoven": { "non_cash.parental_leave_paid_pct": { "source": "WAZO art. 4:2 (geboorteverlof met behoud van loon, employer-paid) https://wetten.overheid.nl/BWBR0013008", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the 1-week partner geboorteverlof, employer-paid wage continuation with no dagloon ceiling (the structural twin of Germany Mutterschutz). The other 39 UWV-paid weeks moved to non_cash.parental_leave_capped in the 2026-08-06 caps rebuild; this field was 83, the whole-blend headline with the maximumdagloon ignored." }, "non_cash.parental_leave_weeks": { "source": "Wet arbeid en zorg (WAZO) art. 3:1, 4:2, 4:2a, 6:2, 6:3 https://wetten.overheid.nl/BWBR0013008 (geldig van 01-01-2025) ; Rijksoverheid https://www.rijksoverheid.nl/vraag-en-antwoord/ouderschapsverlof/recht-op-ouderschapsverlof ; business.gov.nl https://business.gov.nl/regulation/leave-schemes/", "last_updated": "2026-08-01", "confidence": "high", "note": "40 wk = zwangerschaps- en bevallingsverlof 16 (mother) + geboorteverlof 1 (partner) + aanvullend geboorteverlof 5 (partner) + betaald ouderschapsverlof 9 per parent x 2. WAZO gives ouderschapsverlof to each parent individually with no transfer mechanism (Beide ouders hebben recht op ouderschapsverlof). Was 9, which held the paid parental weeks for ONE parent and dropped 16 weeks of maternity at 100%. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Excluded: the 17 unpaid weeks per parent of the 26-week ouderschapsverlof, and the 20-week multiples maternity variant." }, "tax.inbound_regime": { "source": "30%-regeling (expatregeling). Belastingdienst Deskundigheidsvereiste, 2026 norm: a 'belastbaar jaarloon ... meer is dan EUR 48.013 (exclusief de gerichte vrijstelling)', i.e. measured on the salary REMAINING after the 30% carve-out, so about EUR 68,590 of gross: https://www.belastingdienst.nl/wps/wcm/connect/bldcontentnl/belastingdienst/zakelijk/internationaal/personeel/u_bent_niet_in_nederland_gevestigd_loonheffingen_inhouden/als_u_loonheffingen_gaat_inhouden/extraterritoriale_kosten_en_de_30procentregeling/voorwaarden_voor_de_30procentregeling1/deskundigheidsvereiste . Belastingdienst Inhoud van de expatregeling: maximum tax-free allowance EUR 78,600 in 2026, reached at the WNT norm of EUR 262,000. Max 60 months. Rijksoverheid, Belastingvoordeel buitenlandse werknemers.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31 OVER-GRANT FIX + refresh. (1) min_income_basis set to post_relief: the EUR 48,013 floor was being tested against GROSS, so every salary from EUR 48,013 up got the ruling; the statute measures it on the taxable remainder, which needs about EUR 68,590 of gross. Amsterdam single filer at EUR 60,000: tax 7,056 -> 15,906 (ordinary). (2) This entry previously cited the 2025 pair 46,660 / 246,000 while the data already carried the 2026 pair 48,013 / 262,000; the data was right and the prose was stale. (3) The under-30-with-a-master norm of EUR 36,497 is deliberately NOT used: we model neither age nor degree, so we carry the higher general norm. SCHEDULED: allowance drops 30% -> 27% and the norms step up from 2027-01-01 (Belastingplan 2025, second nota van wijziging). The calculator models the current tax year. TODO (2027-01 refresh): apply the 27% and the stepped-up norms once in force for the modeled year." }, "tax.filing_basis": { "source": "PwC WWTS - Netherlands (Tax administration): spouses are taxed separately for employment/business income (Box 1); only Box 2/3 and home items are allocable between fiscal partners.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.box1_brackets": { "source": "Belastingdienst Box 1 2026: peer-identical with Amsterdam", "last_updated": "2026-05-25", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Gemeente Eindhoven OZB 2026 ~0.0973% of WOZ value", "last_updated": "2026-05-25", "confidence": "high" }, "tax.sales_tax_rate": { "source": "BTW standard rate 21%", "last_updated": "2026-05-25", "confidence": "high" }, "housing._summary": { "source": "Pararius / Funda Eindhoven Q1 2026", "last_updated": "2026-05-25", "confidence": "high", "note": "Brainport tech demand; below Randstad core on rent." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Zorgverzekering basisverzekering (Zvw) ~EUR 150/mo; kids <18 free", "last_updated": "2026-05-25", "confidence": "high" }, "childcare._summary": { "source": "Kinderopvang ~EUR 10.71/hr 2026 cap; kinderopvangtoeslag nets ~EUR 8,500/kid for mid tech income", "last_updated": "2026-05-25", "confidence": "medium" }, "food._summary": { "source": "Numbeo Eindhoven + CBS 2026", "last_updated": "2026-05-25", "confidence": "medium" }, "transit._summary": { "source": "Bravo / Hermes regional pass ~EUR 95/mo", "last_updated": "2026-05-25", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo Eindhoven + CBS", "last_updated": "2026-05-25", "confidence": "medium" }, "travel_home._summary": { "source": "EIN/AMS->SEA/LAX 1-stop 2025-26 transatlantic", "last_updated": "2026-05-25", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Algemene Kinderbijslagwet art. 12, consolidated text valid from 01-08-2026, https://wetten.overheid.nl/BWBR0002368/ (fetched 2026-09-12, two independent reads agree; svb.nl returned 403 in English and Dutch and rijksoverheid.nl 404 on 2026-09-10 and 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "Dutch kinderbijslag: universal, NOT means-tested (the means-tested kindgebonden budget is separate and excluded). Statutory amounts per QUARTER per child, band set by the child's age on the first day of the quarter: EUR 298.40 under 6, 362.35 for 6 to 11, 426.29 for 12 to 17 (AKW art. 12 lid 3). The engine has no per-age benefit hook (computeGovtBenefits() reads one rate and only tests age < 18), so ONE rate is stored for every age and the Rule 5 tiebreak picks the LOWEST band: 298.40 x 4 = 1,193.60 a year, stored as 1,194. That understates a 6-to-11-year-old by EUR 255 a year and a 12-to-17-year-old by EUR 511, conservative against the Netherlands. Set 2026-09-12; was 1,368 = 342.06 x 4, which is the 6-to-11 band of the 2025 H1 table mislabelled as 2026. Ledger 2026-09-12; research docs/superpowers/research/2026-09-12-child-benefit-audit-fixes.md F4." }, "non_cash._summary": { "source": "Dutch labour law: 20 statutory vacation days, 26wk parental leave (9 paid ~70% via UWV)", "last_updated": "2026-05-25", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Eindhoven: EUR 70,079, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 70,079 / 0.867 = USD 80,829; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3233, R5-rounded down to 0.32. REPLACES 0.62 (2026-08-11), and the interim 0.4 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. The ASML/Philips metro does NOT measure above Rotterdam, contrary to the stored ordering. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.62 was 0.8611 of Amsterdam's 0.72; that ratio x Amsterdam's re-derived 0.47 = 0.4047, R5-rounded down to 0.4. Amsterdam's own figure: EUR 102,818 / 0.867 = USD 118,591, / USD 250,000 = 0.4744. This is a BASIS move, not a re-measurement: Eindhoven's rank inside Netherlands is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.62; strong hardware/semiconductor cluster." }, "childcare.subsidy_per_kid_annual": { "source": "rijksoverheid.nl kinderopvangtoeslag 2025; Belastingdienst toeslag tables", "last_updated": "2026-07-17", "note": "kinderopvangtoeslag at the 33.3% high-income floor (2025, households >= EUR 226,546): 33.3% x capped rate EUR 10.71/hr x ~1,770 reimbursable h/yr ~= EUR 6,300 per full-time daycare child. The reimbursement % is means-tested, but 33.3% is the statutory MINIMUM, so it survives at the EUR 226k-269k published NL household (was overstated at a mid-income % of ~60-70%). 2026 raises the floor to 36.5%." }, "non_cash.parental_leave_capped": { "source": "UWV maximumdagloon: EUR 309,91 bruto per dag from 1 July 2026 (Wfsv art. 17 lid 1) https://www.uwv.nl/nl/premies-bedragen/maximum-dagloon ; UWV: aanvullend geboorteverlof and betaald ouderschapsverlof pay 70% van het (gemaximeerde) dagloon https://www.uwv.nl/nl/aanvullend-geboorteverlof/hoogte-uitkering-aanvullend-geboorteverlof-betaald-ouderschapsverlof ; WAZO arts. 3:13, 4:2b, 6:3", "last_updated": "2026-08-06", "confidence": "high", "note": "PRE-CONVERTED to a week cap per the calendar-day convention: the dagloon is a 5-WORKDAY wage cap applied BEFORE the percentage, so we store the weekly PAYOUT ceiling at the 70% tier, 309.91 x 0.70 x 5 = EUR 1,084.69/week, over all 39 UWV weeks at the 83 blend (16 wk maternity 100% + 23 wk at 70%). Conservative: the 16 maternity weeks actually cap higher (100% tier, EUR 1,549.55/wk), and pricing the whole block at the lower ceiling reads them LOW. Cap steps semi-annually with the minimum wage (1 Jul 2026 value stored). At EUR 90,000 the blended effective rate is ~64%. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Netherlands)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Zvw statutory eigen risico (EUR 385 per adult, under-18s exempt) + Zorginstituut Nederland and Vektis data on what share of insured actually exhaust it", "last_updated": "2026-08-08", "confidence": "medium", "note": "REPLACES an exact statutory maximum: EUR 385 is precisely the single-adult compulsory deductible, and only about half of insured adults exhaust it, so the realized figure is EUR 229-236 by two converging routes. The deductible is charged PER ADULT and under-18s are wholly exempt, and GP care, maternity care and district nursing sit outside it entirely, so a Dutch family of four is essentially two adults. Aanvullende verzekering is a PREMIUM and is not booked here. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Zvw statutory eigen risico (EUR 385 per adult, under-18s exempt) + Zorginstituut Nederland and Vektis data on what share of insured actually exhaust it", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "CBS StatLine, ODiN dataset 85055NED, motive Van en naar het werk, province Noord-Brabant. https://opendata.cbs.nl/statline/#/CBS/nl/dataset/85055NED", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Gemiddelde reisduur per verplaatsing", "year": "2024", "note": "28.03 min. Province geography, broader than the metro." }, "non_cash.sick_leave": { "source": "Burgerlijk Wetboek (BW) Boek 7, Artikel 629 (loondoorbetaling bij ziekte).; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.uwv.nl/particulieren/bedragen/detail/maximumdagloon", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Belastingdienst Centraal Aanspreekpunt Pensioenen, AOW-franchises https://centraalaanspreekpuntpensioenen.belastingdienst.nl/publicaties/overzicht-aow-inbouwbedragen-en-aow-franchises/ ; V&A 25-008 https://centraalaanspreekpuntpensioenen.belastingdienst.nl/publicaties/va-25-008-v251209/ ; Kennisgroepen KG:070:2024:3 jaarruimte https://kennisgroepen.belastingdienst.nl/publicaties/kg07020243-vragen-en-antwoorden-over-de-jaarruimteberekening-en-de-opbouw-van-pensioenaanspraken/ ; ABP 2026 premie split", "last_updated": "2026-08-11", "confidence": "low", "note": "THIS VALUE IS A MODEL, NOT A STATUTE, and the formula is what is HIGH confidence. Art. 3.127 Wet IB 2001 post-Wtp: jaarruimte = 30% of premiegrondslag MINUS the actual ouderdomspensioen premium already paid (employer + employee). 2026 AOW-franchise EUR 19,172, aftoppingsgrens EUR 137,800, so the statutory maximum is 0.30 x (137,800 - 19,172) = EUR 35,589 (encoded as 13,085 here, see below). The naive statutory mapping is INERT at our incomes and would look like a fix while changing nothing. 13,085 = 11.03% of premiegrondslag, i.e. 30% less ABP's 18.97% employer share; 0.082 is the lowest in-band gross ratio (at EUR 75k), so the mapping sits at or below the derived truth everywhere above EUR 75k. Expressing this via less_employer_pct after the 2026-08-11 engine reorder was considered and rejected: it cannot subtract the AOW-franchise and would hand every Dutch earner about EUR 2,115 of extra room." } }, "the-hague": { "non_cash.parental_leave_paid_pct": { "source": "WAZO art. 4:2 (geboorteverlof met behoud van loon, employer-paid) https://wetten.overheid.nl/BWBR0013008", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the 1-week partner geboorteverlof, employer-paid wage continuation with no dagloon ceiling (the structural twin of Germany Mutterschutz). The other 39 UWV-paid weeks moved to non_cash.parental_leave_capped in the 2026-08-06 caps rebuild; this field was 83, the whole-blend headline with the maximumdagloon ignored." }, "non_cash.parental_leave_weeks": { "source": "Wet arbeid en zorg (WAZO) art. 3:1, 4:2, 4:2a, 6:2, 6:3 https://wetten.overheid.nl/BWBR0013008 (geldig van 01-01-2025) ; Rijksoverheid https://www.rijksoverheid.nl/vraag-en-antwoord/ouderschapsverlof/recht-op-ouderschapsverlof ; business.gov.nl https://business.gov.nl/regulation/leave-schemes/", "last_updated": "2026-08-01", "confidence": "high", "note": "40 wk = zwangerschaps- en bevallingsverlof 16 (mother) + geboorteverlof 1 (partner) + aanvullend geboorteverlof 5 (partner) + betaald ouderschapsverlof 9 per parent x 2. WAZO gives ouderschapsverlof to each parent individually with no transfer mechanism (Beide ouders hebben recht op ouderschapsverlof). Was 9, which held the paid parental weeks for ONE parent and dropped 16 weeks of maternity at 100%. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Excluded: the 17 unpaid weeks per parent of the 26-week ouderschapsverlof, and the 20-week multiples maternity variant." }, "tax.inbound_regime": { "source": "30%-regeling (expatregeling). Belastingdienst Deskundigheidsvereiste, 2026 norm: a 'belastbaar jaarloon ... meer is dan EUR 48.013 (exclusief de gerichte vrijstelling)', i.e. measured on the salary REMAINING after the 30% carve-out, so about EUR 68,590 of gross: https://www.belastingdienst.nl/wps/wcm/connect/bldcontentnl/belastingdienst/zakelijk/internationaal/personeel/u_bent_niet_in_nederland_gevestigd_loonheffingen_inhouden/als_u_loonheffingen_gaat_inhouden/extraterritoriale_kosten_en_de_30procentregeling/voorwaarden_voor_de_30procentregeling1/deskundigheidsvereiste . Belastingdienst Inhoud van de expatregeling: maximum tax-free allowance EUR 78,600 in 2026, reached at the WNT norm of EUR 262,000. Max 60 months. Rijksoverheid, Belastingvoordeel buitenlandse werknemers.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31 OVER-GRANT FIX + refresh. (1) min_income_basis set to post_relief: the EUR 48,013 floor was being tested against GROSS, so every salary from EUR 48,013 up got the ruling; the statute measures it on the taxable remainder, which needs about EUR 68,590 of gross. Amsterdam single filer at EUR 60,000: tax 7,056 -> 15,906 (ordinary). (2) This entry previously cited the 2025 pair 46,660 / 246,000 while the data already carried the 2026 pair 48,013 / 262,000; the data was right and the prose was stale. (3) The under-30-with-a-master norm of EUR 36,497 is deliberately NOT used: we model neither age nor degree, so we carry the higher general norm. SCHEDULED: allowance drops 30% -> 27% and the norms step up from 2027-01-01 (Belastingplan 2025, second nota van wijziging). The calculator models the current tax year. TODO (2027-01 refresh): apply the 27% and the stepped-up norms once in force for the modeled year." }, "tax.filing_basis": { "source": "PwC WWTS - Netherlands (Tax administration): spouses are taxed separately for employment/business income (Box 1); only Box 2/3 and home items are allocable between fiscal partners.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.box1_brackets": { "source": "Belastingdienst Box 1 2026: peer-identical with Amsterdam", "last_updated": "2026-05-25", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Gemeente Den Haag OZB 2026 ~0.1270% of WOZ value", "last_updated": "2026-05-25", "confidence": "high" }, "tax.sales_tax_rate": { "source": "BTW standard rate 21%", "last_updated": "2026-05-25", "confidence": "high" }, "housing._summary": { "source": "Pararius / Funda The Hague Q1 2026", "last_updated": "2026-05-25", "confidence": "high", "note": "Government + international-org/expat demand; Randstad pricing." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Zorgverzekering basisverzekering (Zvw) ~EUR 150/mo; kids <18 free", "last_updated": "2026-05-25", "confidence": "high" }, "childcare._summary": { "source": "Kinderopvang ~EUR 10.71/hr 2026 cap; kinderopvangtoeslag nets ~EUR 8,400/kid for mid tech income", "last_updated": "2026-05-25", "confidence": "medium" }, "food._summary": { "source": "Numbeo The Hague + CBS 2026", "last_updated": "2026-05-25", "confidence": "medium" }, "transit._summary": { "source": "HTM regional pass ~EUR 100/mo", "last_updated": "2026-05-25", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo The Hague + CBS", "last_updated": "2026-05-25", "confidence": "medium" }, "travel_home._summary": { "source": "AMS/RTM->SEA/LAX 1-stop 2025-26 transatlantic", "last_updated": "2026-05-25", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Algemene Kinderbijslagwet art. 12, consolidated text valid from 01-08-2026, https://wetten.overheid.nl/BWBR0002368/ (fetched 2026-09-12, two independent reads agree; svb.nl returned 403 in English and Dutch and rijksoverheid.nl 404 on 2026-09-10 and 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "Dutch kinderbijslag: universal, NOT means-tested (the means-tested kindgebonden budget is separate and excluded). Statutory amounts per QUARTER per child, band set by the child's age on the first day of the quarter: EUR 298.40 under 6, 362.35 for 6 to 11, 426.29 for 12 to 17 (AKW art. 12 lid 3). The engine has no per-age benefit hook (computeGovtBenefits() reads one rate and only tests age < 18), so ONE rate is stored for every age and the Rule 5 tiebreak picks the LOWEST band: 298.40 x 4 = 1,193.60 a year, stored as 1,194. That understates a 6-to-11-year-old by EUR 255 a year and a 12-to-17-year-old by EUR 511, conservative against the Netherlands. Set 2026-09-12; was 1,368 = 342.06 x 4, which is the 6-to-11 band of the 2025 H1 table mislabelled as 2026. Ledger 2026-09-12; research docs/superpowers/research/2026-09-12-child-benefit-audit-fixes.md F4." }, "non_cash._summary": { "source": "Dutch labour law: 20 statutory vacation days, 26wk parental leave (9 paid ~70% via UWV)", "last_updated": "2026-05-25", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, The Hague: EUR 72,165, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 72,165 / 0.867 = USD 83,235; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3329, R5-rounded down to 0.33. REPLACES 0.6 (2026-08-11), and the interim 0.39 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.6 was 0.8333 of Amsterdam's 0.72; that ratio x Amsterdam's re-derived 0.47 = 0.3917, R5-rounded down to 0.39. Amsterdam's own figure: EUR 102,818 / 0.867 = USD 118,591, / USD 250,000 = 0.4744. This is a BASIS move, not a re-measurement: The Hague's rank inside Netherlands is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.60; less private-tech-dense than Amsterdam/Utrecht." }, "childcare.subsidy_per_kid_annual": { "source": "rijksoverheid.nl kinderopvangtoeslag 2025; Belastingdienst toeslag tables", "last_updated": "2026-07-17", "note": "kinderopvangtoeslag at the 33.3% high-income floor (2025, households >= EUR 226,546): 33.3% x capped rate EUR 10.71/hr x ~1,770 reimbursable h/yr ~= EUR 6,300 per full-time daycare child. The reimbursement % is means-tested, but 33.3% is the statutory MINIMUM, so it survives at the EUR 226k-269k published NL household (was overstated at a mid-income % of ~60-70%). 2026 raises the floor to 36.5%." }, "non_cash.parental_leave_capped": { "source": "UWV maximumdagloon: EUR 309,91 bruto per dag from 1 July 2026 (Wfsv art. 17 lid 1) https://www.uwv.nl/nl/premies-bedragen/maximum-dagloon ; UWV: aanvullend geboorteverlof and betaald ouderschapsverlof pay 70% van het (gemaximeerde) dagloon https://www.uwv.nl/nl/aanvullend-geboorteverlof/hoogte-uitkering-aanvullend-geboorteverlof-betaald-ouderschapsverlof ; WAZO arts. 3:13, 4:2b, 6:3", "last_updated": "2026-08-06", "confidence": "high", "note": "PRE-CONVERTED to a week cap per the calendar-day convention: the dagloon is a 5-WORKDAY wage cap applied BEFORE the percentage, so we store the weekly PAYOUT ceiling at the 70% tier, 309.91 x 0.70 x 5 = EUR 1,084.69/week, over all 39 UWV weeks at the 83 blend (16 wk maternity 100% + 23 wk at 70%). Conservative: the 16 maternity weeks actually cap higher (100% tier, EUR 1,549.55/wk), and pricing the whole block at the lower ceiling reads them LOW. Cap steps semi-annually with the minimum wage (1 Jul 2026 value stored). At EUR 90,000 the blended effective rate is ~64%. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Netherlands)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Zvw statutory eigen risico (EUR 385 per adult, under-18s exempt) + Zorginstituut Nederland and Vektis data on what share of insured actually exhaust it", "last_updated": "2026-08-08", "confidence": "medium", "note": "REPLACES an exact statutory maximum: EUR 385 is precisely the single-adult compulsory deductible, and only about half of insured adults exhaust it, so the realized figure is EUR 229-236 by two converging routes. The deductible is charged PER ADULT and under-18s are wholly exempt, and GP care, maternity care and district nursing sit outside it entirely, so a Dutch family of four is essentially two adults. Aanvullende verzekering is a PREMIUM and is not booked here. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Zvw statutory eigen risico (EUR 385 per adult, under-18s exempt) + Zorginstituut Nederland and Vektis data on what share of insured actually exhaust it", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "CBS StatLine, ODiN dataset 85055NED, motive Van en naar het werk, province Zuid-Holland. https://opendata.cbs.nl/statline/#/CBS/nl/dataset/85055NED", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Gemiddelde reisduur per verplaatsing", "year": "2024", "note": "32.05 min, the same Zuid-Holland figure as Rotterdam, because province is the finest geography ODiN publishes." }, "non_cash.sick_leave": { "source": "Burgerlijk Wetboek (BW) Boek 7, Artikel 629 (loondoorbetaling bij ziekte).; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.uwv.nl/particulieren/bedragen/detail/maximumdagloon", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Belastingdienst Centraal Aanspreekpunt Pensioenen, AOW-franchises https://centraalaanspreekpuntpensioenen.belastingdienst.nl/publicaties/overzicht-aow-inbouwbedragen-en-aow-franchises/ ; V&A 25-008 https://centraalaanspreekpuntpensioenen.belastingdienst.nl/publicaties/va-25-008-v251209/ ; Kennisgroepen KG:070:2024:3 jaarruimte https://kennisgroepen.belastingdienst.nl/publicaties/kg07020243-vragen-en-antwoorden-over-de-jaarruimteberekening-en-de-opbouw-van-pensioenaanspraken/ ; ABP 2026 premie split", "last_updated": "2026-08-11", "confidence": "low", "note": "THIS VALUE IS A MODEL, NOT A STATUTE, and the formula is what is HIGH confidence. Art. 3.127 Wet IB 2001 post-Wtp: jaarruimte = 30% of premiegrondslag MINUS the actual ouderdomspensioen premium already paid (employer + employee). 2026 AOW-franchise EUR 19,172, aftoppingsgrens EUR 137,800, so the statutory maximum is 0.30 x (137,800 - 19,172) = EUR 35,589 (encoded as 13,085 here, see below). The naive statutory mapping is INERT at our incomes and would look like a fix while changing nothing. 13,085 = 11.03% of premiegrondslag, i.e. 30% less ABP's 18.97% employer share; 0.082 is the lowest in-band gross ratio (at EUR 75k), so the mapping sits at or below the derived truth everywhere above EUR 75k. Expressing this via less_employer_pct after the 2026-08-11 engine reorder was considered and rejected: it cannot subtract the AOW-franchise and would hand every Dutch earner about EUR 2,115 of extra room." } }, "hamburg": { "housing.median_rent_by_bedroom": { "source": "Wohnungsmarktstudie des Gymnasiums Ohmoor 2026 (40th annual edition, n about 5,300 listings Jan-Mar 2026, Immowelt data, Hamburg city separated from Umland), via Mieterverein zu Hamburg https://www.mieterverein-hamburg.de/wp-content/uploads/2026/05/PM_Mieterverein_Ohmoor_Mietenstudie.pdf ; per-room gradient VALUE Marktdatenbank 2026 https://mietspiegeltabelle.de/mietspiegel-hamburg/ ; dwelling sizes Statistikamt Nord, Hochbautaetigkeit und Wohnungsbestand in Hamburg 2023 (F II 1, 2, 4 - j 23 HH) https://www.statistik-nord.de/fileadmin/Dokumente/F_II_1_2_4_j23_HH.xlsx", "last_updated": "2026-08-07", "confidence": "high", "note": "CORRECTED 2026-08-02 from 850/1100/1550/2100, replacing an \"Investropa + Numbeo\" citation. The Ohmoor study, running annually since 1986, puts Hamburg new-lease Kaltmiete at 15.88 EUR/sqm for Q1 2026 against 12.65 in the Umland, and VALUE 2026 gives 16.24; ageing the 2025 Ohmoor figure of 15.62 forward by 4% lands on 16.24 exactly, so two unrelated sources reconcile. Per-room rates 17.76/15.72/15.96/16.94 x 1.0587 = 18.80/16.64/16.90/17.93. Studio 850 to 790, 1BR 1100 to 1000, 2BR 1550 to 1440, 3BR 2100 to 1880. Hamburg was the mildest of the seven: every move is under 10% and the ladder shape was already at the dataset median, so this is a re-basing onto a real source rather than a rescue. The Ohmoor figure is an unweighted arithmetic mean across listings, which sits above a size-weighted rate because small flats carry a per-sqm premium, so it must not be applied flat to every size band; the per-room gradient does that work. Statistikamt Nord puts the average Mehrfamilienhaus dwelling at 66.29 sqm and new-build flats at 73.2 sqm, and notes Wohnraeume exclude kitchen and bathroom so the count is directly comparable to listing Zimmer. Per-room rates come from the VALUE Marktdatenbank 2026 cut published at mietspiegeltabelle.de, explicitly nettokalt and explicitly market/asking rather than the qualified municipal Mietspiegel. That is the same database family the Berlin correction leant on. Leipzig proves what it serves: its 10.45 equals VALUE own published MEAN of 10.49 against a median of 10.00, so this basis runs about 5% above median, which is the conservative direction. Method is the shipped Berlin recipe re-run per city, so all eight German cities sit on one footing: rent = size x city per-room EUR/sqm x 1.0587. The 1.0587 is Berlin own uplift over this same source (Berlin shipped a 16.6 base against 15.68 here), and re-running it reproduces Berlin 760/970/1360/1750 to the euro. Sizes are held constant across cities at 42/60/85/105 sqm on the German Zimmer convention (our 1BR = 2-Zimmer-Wohnung), because the product prices an equivalent standard of living rather than how much space local households ration themselves into at local prices. Duesseldorf publishes real offered-flat sizes of 37/58/83/111, so the fixed 42 makes the studio the most conservative cell and the fixed 105 makes the 3BR mildly generous on size; the two roughly cancel. Basis is Angebotsmiete (new lease), netto kalt, never the Mietspiegel stock rent. Rents exclude utilities, insurance and internet, which are separate fields. Values are a CEILING per the conservative value-bias rule, not central estimates. Working: docs/superpowers/research/2026-08-02-german-city-rents.md SIZE LADDER CORRECTED 2026-08-07 (small-fixes wave 1, coordinated eight-city pass incl. Berlin): the common dwelling-size ladder moves from the inherited 42/60/85/105 sqm to 37/58/83/111 sqm, Duesseldorf's PUBLISHED offered-flat sizes (Wohnungsmarktbericht 2025 Abb. 54, MFH column, n=10,370 listings; the only published per-Zimmer size table in the German set, and the conservative top of the seven-city evidence range at 3-Zi 83 and 4-Zi 111). Root cause of the old ladder: German official statistics count a closed kitchen over 6 sqm as a room, so census room distributions ran one room high against listing Zimmer counts; the 42 sqm studio was priced at the 1-Zimmer premium rate measured on 33-37 sqm stock (double-counting the small-flat premium) and the 4-Zimmer was undersized at 105 vs the real ~111. Values 790/1000/1440/1880 -> 700/970/1400/1990. Per-room rates and the 1.0587 uplift are UNCHANGED; sizes remain COMMON across all eight German cities because the product prices an equivalent standard of living. Working: docs/superpowers/research/2026-08-07-small-fixes.md" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "non_cash.parental_leave_paid_pct": { "source": "SGB V para 24i (Mutterschaftsgeld) https://www.gesetze-im-internet.de/sgb_5/__24i.html ; MuSchG para 20 (Zuschuss zum Mutterschaftsgeld, employer tops up to full net) https://www.gesetze-im-internet.de/muschg_2018/__20.html", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the percentage-paid remainder of the 67 weeks: the 14.1 Mutterschutz weeks (6 pre-birth + 8 post-birth), where para 24i SGB V pays EUR 13/day and para 20 MuSchG obliges the employer to top up to full net pay with no ceiling. Was 72, a blend that averaged the capped Elterngeld slice into this field; since 2026-08-06 the 52.87 Elterngeld weeks live in non_cash.parental_leave_capped (65% up to EUR 1,800/month, nothing above EUR 175,000 household income) and parentalLeaveEffectiveRate() blends the slices at the user's salary, so every reader-facing surface already prices the cap and the cliff." }, "non_cash.parental_leave_capped": { "source": "BEEG para 2 (Elterngeld: 65% of prior income, minimum EUR 300 and maximum EUR 1,800 per month) https://www.gesetze-im-internet.de/beeg/__2.html ; BEEG para 1 Abs. 8 (no Elterngeld above EUR 175,000 combined zu versteuerndes Einkommen) https://www.gesetze-im-internet.de/beeg/__1.html ; MuSchG para 20 (the 14 Mutterschutz weeks stay OUTSIDE this block: the employer net top-up has no ceiling) https://www.gesetze-im-internet.de/muschg_2018/__20.html", "last_updated": "2026-08-06", "confidence": "high", "note": "52.87 of the 67 weeks: the Elterngeld pot (60.87 wk) minus the 8 overlap weeks counted at the Mutterschutz rate. The EUR 1,800/month cap (unindexed since Elterngeld's 2007 introduction, re-confirmed 2026-08-06) binds above ~EUR 33,000 gross, i.e. at every ICP salary: at EUR 90,000 this slice pays ~24% of gross and the blended Berlin rate is ~40%; at EUR 150,000 the slice pays ~14% and the blend is ~32%, versus the 72 this field carried before the 2026-08-06 caps rebuild. BEEG para 1 Abs. 8 zeroes the slice for households above EUR 175,000: the engine compares HOUSEHOLD GROSS against the ceiling as a conservative proxy for zu versteuerndes Einkommen (gross exceeds taxable income, so the cliff bites earlier, the direction less favorable to Germany). Same conservative direction on the rate basis: the statute pays 65% of Elterngeld-Netto while the engine applies 65% to gross, and above ~EUR 33,000 gross the cap binds under both readings, so the distinction is immaterial at any ICP salary. Floor EUR 300/month (BEEG para 2). Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Germany section) and 2026-08-06-caps-engine-build.md." }, "non_cash.parental_leave_weeks": { "source": "MuSchG para 3 (Schutzfristen, 6 weeks before and 8 after) https://www.gesetze-im-internet.de/muschg_2018/__3.html ; BEEG para 4 Abs. 3 (12 Monatsbetraege Basiselterngeld plus 2 Partnermonate, gemeinsam) and para 4 Abs. 4 Satz 3 https://www.gesetze-im-internet.de/beeg/__4.html", "last_updated": "2026-08-01", "confidence": "high", "note": "67 wk = 6 additive pre-birth Mutterschutz weeks + the 14-month Elterngeld pot (60.87 wk). Was 61, the Elterngeld pot alone with Mutterschutz omitted. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. OVERLAP RULE APPLIED: the 8 post-birth Mutterschutz weeks are NOT stacked on top, because BEEG para 4 Abs. 4 Satz 3 deems Lebensmonate carrying Mutterschaftsgeld to be consumed Basiselterngeld months (gelten als Monate, fuer die dieser Elternteil Basiselterngeld bezieht). They are counted once, at the higher rate. Germany therefore reaches 67 and not 75. Elterngeld is a SHARED household pot, so it is not doubled. GERMANY HAS NO STATUTORY PAID PATERNITY LEAVE: the Familienstartzeit bill transposing EU Directive 2019/1158 art. 4 was still not law on 2026-08-01, so zero paternity weeks. Excluded: ElterngeldPlus (an election that halves the monthly amount to double the months) and the Partnerschaftsbonus (conditional on both parents working reduced hours)." }, "tax.income_tax_tariff_32a": { "source": "§ 32a Abs. 1 EStG, Veranlagungszeitraum 2026 https://www.gesetze-im-internet.de/estg/__32a.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Statutory continuous tariff, identical nationwide (see berlin for the verbatim zone text and the coefficient cross-check). Replaces the 5-bracket approximation that flattened the 17,799-69,878 progression zone to 33.5%." }, "tax.werbungskosten_pauschbetrag": { "source": "§ 9a Satz 1 Nr. 1 Buchst. a EStG https://www.gesetze-im-internet.de/estg/__9a.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Arbeitnehmer-Pauschbetrag EUR 1,230 per employed person. National, identical to berlin." }, "tax.sonderausgaben_pauschbetrag": { "source": "§ 10c EStG https://www.gesetze-im-internet.de/estg/__10c.html", "last_updated": "2026-07-31", "confidence": "high", "note": "EUR 36, doubled on Zusammenveranlagung. National, identical to berlin." }, "tax.vorsorge_deductible_rate_rv": { "source": "§ 10 Abs. 1 Nr. 2 und Abs. 3 EStG https://www.gesetze-im-internet.de/estg/__10.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Employee RV share 9.3%, 100% deductible since 2023, against the RV Beitragsbemessungsgrenze. National, identical to berlin." }, "tax.vorsorge_deductible_rate_kvpv": { "source": "§ 10 Abs. 1 Nr. 3 und Abs. 4 Satz 4 EStG https://www.gesetze-im-internet.de/estg/__10.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Basis KV 8.75% x 0.96 (Krankengeld haircut) + PV 1.8% = 10.2%, against the KV/PV Beitragsbemessungsgrenze; ALV not deductible. National, identical to berlin." }, "tax.property_tax_rate_pct": { "source": "Hamburg Flächenmodell Grundsteuer B 975% (low Messzahl 0.70)", "last_updated": "2026-05-26", "confidence": "medium", "note": "Flächenmodell, not Bundesmodell — set to German-norm ~0.15% effective (revenue-neutral reform)." }, "tax.sales_tax_rate": { "source": "German VAT 19%", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Investropa + Numbeo Hamburg 2026", "last_updated": "2026-05-26", "confidence": "high" }, "healthcare._summary": { "source": "GKV statutory insurance (income-based, taken in payroll => premium 0); Eigenanteil ~€300/yr", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "hamburg.de Elternbeitrag: 5h/day free + income-capped fees", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Hamburg 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Deutschlandticket €58/mo (nationwide)", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Hamburg", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "HAM->US transatlantic economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Kindergeld €259/mo = €3,108/yr (national, inherited)", "last_updated": "2026-05-26", "confidence": "high", "note": "German Kindergeld (EUR 259/mo per child = EUR 3,108/yr, which is the stored value): universal, NOT means-tested, and it does not taper at professional incomes. Corrected 2026-08-02 from a stale \"EUR 255/mo\", which is the 2025 rate; the Steuerfortentwicklungsgesetz stepped Kindergeld to EUR 255 on 2025-01-01 and EUR 259 on 2026-01-01. The stored value was already correct and was NOT changed." }, "non_cash._summary": { "source": "German labour law (national leave/parental, inherited); Hamburg public holidays 10 (Land-specific)", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Hamburg: EUR 73,603, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 73,603 / 0.867 = USD 84,894; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3396, R5-rounded down to 0.33. REPLACES 0.6 (2026-08-11), and the interim 0.42 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.6 was 1.0000 of Berlin's 0.6; that ratio x Berlin's re-derived 0.42 = 0.4200, R5-rounded down to 0.42. Berlin's own figure: EUR 91,048 / 0.867 = USD 105,015, / USD 250,000 = 0.4201. This is a BASIS move, not a re-measurement: Hamburg's rank inside Germany is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.60, tracks Berlin." }, "tax.solidaritaetszuschlag_threshold_married": { "source": "Soli Freigrenze 2026 joint assessment EUR 40,700 (2 x 20,350 single; tk.de Solidaritaetszuschlag); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.social_insurance_rate_rvalv": { "source": "RV 9.3% + ALV 1.3% employee share = 10.6%, runs to the higher Beitragsbemessungsgrenze (lohn-info.de Sozialversicherungsbeitraege 2026); verified tax audit 2026-07-14", "confidence": "high", "note": "Carved out of the combined 21.15% employee share; the KV/PV remainder (10.55%) stops at the lower ceiling.", "last_updated": "2026-07-17" }, "tax.social_insurance_cap_rvalv": { "source": "Beitragsbemessungsgrenze RV/ALV 2026 = EUR 101,400 (bundesregierung.de/breg-de/aktuelles/beitragsgemessungsgrenzen-2386514); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.pv_childless_surcharge_rate": { "source": "PV Kinderlosenzuschlag § 55 Abs. 3 SGB XI: +0.6pp employee-borne for childless members 23+ (dak.de Beitrag zur Pflegeversicherung Zuschlag Kinderlose; vdek.com Aenderungen Gesundheitswesen 2026); found by the 2026-08-06 corrections wave", "confidence": "high", "note": "Applied when the household has no children; the under-23 exemption is not modeled (ICP is older). Deductible with the rest of the Basis-Pflegeversicherung.", "last_updated": "2026-08-06" }, "childcare.subsidy_per_kid_annual": { "source": "Land and municipal Kitabeitrag schedules; double-subtraction correction 2026-08-02", "last_updated": "2026-08-02", "confidence": "high", "note": "Set to 0 on 2026-08-02, was EUR 1,800. The old value was the VALUE OF THE FREE OR SUBSIDISED KITA HOURS, but childcare.preschool_monthly already carries the post-free-tier out-of-pocket, so computeChildcareCost subtracted the same relief a second time and clamped the Hamburg household bill to zero. German Kita relief is a benefit in kind delivered as a lower or absent fee, never a payment to the parent, so nothing belongs in this field. Hamburg now bills EUR 100/month per child, the food contribution on the fee-free five-hour Basisgutschein. This field holds CASH THAT REACHES THE HOUSEHOLD. Where the monthly rate it nets against is already the price the household actually pays, the value is 0 and the relief is described here instead. See docs/superpowers/research/2026-08-02-childcare-double-subtraction.md." }, "transit.monthly_pass": { "source": "https://www.bundesregierung.de/breg-de/aktuelles/deutschlandticket-2134074", "last_updated": "2026-08-05", "confidence": "high", "note": "Deutschlandticket EUR 63/mo since 2026-01-01 (VMK decision 2025-09-18). Was 58 through Dec 2025." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Destatis Gesundheitsausgabenrechnung 2024 (private households + NPISH EUR 65bn) + Destatis EVS 2023 Konsumausgaben nach Haushaltstyp + GKV Zuzahlungen aggregate EUR 5.18bn (BMG via Aerzteblatt) + SGB V paras 61-62", "last_updated": "2026-08-08", "confidence": "medium", "note": "Level essentially CONFIRMED at EUR 311 per adult (778/capita gross, ~40% in scope after netting long-term care 46%, dental 9%, vision ~3pp). What Germany needed was a curve, not a level. The curve is steep in adults and nearly flat in children because under-18s are exempt from every Zuzahlung except transport AND each child subtracts EUR 9,756 from the income base before the 2%-of-gross Belastungsgrenze, so adding a child LOWERS the household's statutory ceiling in absolute euros. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Destatis Gesundheitsausgabenrechnung 2024 (private households + NPISH EUR 65bn) + Destatis EVS 2023 Konsumausgaben nach Haushaltstyp + GKV Zuzahlungen aggregate EUR 5.18bn (BMG via Aerzteblatt) + SGB V paras 61-62", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, urb_ctran indicator TT1019V, city DE002C. https://ec.europa.eu/eurostat/databrowser/view/urb_ctran/default/table?lang=en", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average time of journey to work (minutes)", "year": "2016", "note": "30.1 min, BASIS YEAR 2016. CITY geography, narrower than our metro unit, so biased LOW and flattering; this row moves -7. Hamburg did not participate in SrV 2023, so no city travel-survey cross-check exists either. Survey self-report, standard half-up rounding. METRO-LEVEL GERMAN SERIES: ESTABLISHED HARD NEGATIVE, do not re-run. MiD 2023 Tabelle 6 carries RegioStaR7 breakdowns of population, trips, distance and total time but NO duration by purpose; the national banner table W9 Wegedauer is all-purpose and sits on an \"Ohne regelmaessige berufliche Wege\" base; the small-area products regionalise headline indicators only; and Eurostat publishes no functional-urban-area rows for TT1019V. SrV 2023 was rejected because it publishes an ALL-PURPOSE trip duration, and swapping an all-purpose duration for a work-purpose one is a UNIT error, which is worse than the geography error we accepted. SrV is therefore used only as rank-order corroboration, and it agrees with Eurostat against our old stored spread: both put Frankfurt near the BOTTOM of the big five, where we had it second highest. PROVISIONAL: this is one of the three large German drops and it is the flattering direction, so re-check it first if a metro-level series ever appears." }, "non_cash.sick_leave": { "source": "Entgeltfortzahlungsgesetz (EFZG) §3 (employer phase); SGB V §47 (Krankengeld; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "housing.utilities_monthly": { "source": "Mieterverein Hamburg Betriebskostenspiegel (city premium +8.8% over the DMB national) + Numbeo 85 sqm 341.56, https://www.mieterverein-hamburg.de/mediathek/pressemitteilungen/neuer-betriebskostenspiegel-fuer-hamburg-mieter-zahlen-im-schnitt-346-euro-pro-quadratmeter/", "confidence": "medium-high", "note": "2BR reference. Band 305 to 342; 335 sits below both Numbeo and the ratio transfer because Hamburg has no current city-measured LEVEL, only a published premium. The 3,46 EUR/sqm in the press release is NOT an average-paid figure: it is 0,28 above the DMB all-positions 3,18 for the same year, and reading it as average-paid would overstate Hamburg by 59%.", "last_updated": "2026-08-08" }, "tax.retirement_deduction_cap": { "source": "EStG s.10 Abs 3 https://www.gesetze-im-internet.de/estg/__10.html ; SVBezGrV 2026 s.4 Abs 1 https://www.gesetze-im-internet.de/svbezgrv_2026/BJNR1160A0025.html", "last_updated": "2026-08-11", "confidence": "high", "note": "Ruerup/Basisrente Hoechstbetrag, 2026: EUR 30,825.60 single (knappschaftliche RV BBG 124,800 x 24.7%). A DEDUCTION ceiling and a COMBINED one: employee AND employer statutory pension contributions consume the same pot. less_employer_pct 0.186 = 2 x 9.3% and employer_base_cap 101,400 is the allgemeine RV BBG, because computeGermanyTaxes already deducts the employee 9.3% via vorsorge; a naked max would hand a EUR 95k earner about EUR 8.8k of phantom deduction. Riester (EUR 2,100) and bAV Entgeltumwandlung (8% of the BBG, EUR 8,112) are deliberately NOT stacked. Married couples: the statute doubles the ceiling to 61,651.20 even on one income; our per-earner cap under-deducts there, which errs toward higher tax." } }, "cologne": { "housing.median_rent_by_bedroom": { "source": "Stadt Koeln, Amt fuer Stadtentwicklung und Statistik, \"Das Mietwohnungsangebot in Koeln 2025\", Koelner Statistische Nachrichten 9/2026 (median Angebotsmiete, Nettokaltmiete, FUB IGES Mietdatenbank, 12,205 offers of which 11,659 usable, city area) https://www.stadt-koeln.de/mediaasset/content/pdf15/statistik-bauen-und-wohnen/ksn_9_2026_das_mietwohnungsangebot_in_koeln_2025.pdf ; Stadt Koeln, Wohnungsmarktbericht Koeln 2024, KSN 2/2024 (Bestandsmiete contrast) https://www.stadt-koeln.de/mediaasset/content/pdf15/statistik-bauen-und-wohnen/ksn_2_2024_wohnungsmarktbericht_k%C3%B6ln_2024.pdf ; per-room gradient VALUE Marktdatenbank 2026 https://mietspiegeltabelle.de/mietspiegel-koeln/", "last_updated": "2026-08-07", "confidence": "high", "note": "CORRECTED 2026-08-02 from 700/900/1400/1600, replacing an \"Investropa + Numbeo\" citation that also carried an unjustified high confidence. The city own report gives a 2025 median Angebotsmiete of 14.70 EUR/sqm nettokalt, up 4.8% from 14.00, and states outright that all mittlere values are medians; VALUE 2026 gives 15.93, the gap being one year of growth plus a mean-versus-median effect. Per-room rates 18.01/15.72/14.98/15.39 x 1.0587 = 19.07/16.64/15.86/16.29. Studio 700 to 800, 1BR 900 to 1000, 2BR 1400 to 1350, 3BR 1600 to 1710. The real defect here was SHAPE, not level: the stored steps ran +56% then +14%, so the 2BR sat at the 90th percentile of the whole dataset while the 3BR step fell BELOW its 10th percentile. The old 1400 at 85 sqm implies 16.47 per sqm, above every Cologne size band from 40 sqm up (the city puts them all between 13.30 and 14.60 on a 2025 median basis) and above every Stadtteil median but the top three. Corrected values are the ceiling reading and are close to the independently derived inner-ring uplift of 700/1010/1290/1680. Stock contrast: the city puts the 2023 Bestandsmiete median at 10.00, so asking runs about 47% above sitting-tenant rent. Two caveats the city itself raises and we carry: 15% of offers are now furnished at 19.30 against 14.30 unfurnished, and empirica finds Koeln asking rents run about 15% above rents on leases that actually close, so these are a ceiling on what is asked. Per-room rates come from the VALUE Marktdatenbank 2026 cut published at mietspiegeltabelle.de, explicitly nettokalt and explicitly market/asking rather than the qualified municipal Mietspiegel. That is the same database family the Berlin correction leant on. Leipzig proves what it serves: its 10.45 equals VALUE own published MEAN of 10.49 against a median of 10.00, so this basis runs about 5% above median, which is the conservative direction. Method is the shipped Berlin recipe re-run per city, so all eight German cities sit on one footing: rent = size x city per-room EUR/sqm x 1.0587. The 1.0587 is Berlin own uplift over this same source (Berlin shipped a 16.6 base against 15.68 here), and re-running it reproduces Berlin 760/970/1360/1750 to the euro. Sizes are held constant across cities at 42/60/85/105 sqm on the German Zimmer convention (our 1BR = 2-Zimmer-Wohnung), because the product prices an equivalent standard of living rather than how much space local households ration themselves into at local prices. Duesseldorf publishes real offered-flat sizes of 37/58/83/111, so the fixed 42 makes the studio the most conservative cell and the fixed 105 makes the 3BR mildly generous on size; the two roughly cancel. Basis is Angebotsmiete (new lease), netto kalt, never the Mietspiegel stock rent. Rents exclude utilities, insurance and internet, which are separate fields. Values are a CEILING per the conservative value-bias rule, not central estimates. Working: docs/superpowers/research/2026-08-02-german-city-rents.md SIZE LADDER CORRECTED 2026-08-07 (small-fixes wave 1, coordinated eight-city pass incl. Berlin): the common dwelling-size ladder moves from the inherited 42/60/85/105 sqm to 37/58/83/111 sqm, Duesseldorf's PUBLISHED offered-flat sizes (Wohnungsmarktbericht 2025 Abb. 54, MFH column, n=10,370 listings; the only published per-Zimmer size table in the German set, and the conservative top of the seven-city evidence range at 3-Zi 83 and 4-Zi 111). Root cause of the old ladder: German official statistics count a closed kitchen over 6 sqm as a room, so census room distributions ran one room high against listing Zimmer counts; the 42 sqm studio was priced at the 1-Zimmer premium rate measured on 33-37 sqm stock (double-counting the small-flat premium) and the 4-Zimmer was undersized at 105 vs the real ~111. Values 800/1000/1350/1710 -> 710/970/1320/1810. Per-room rates and the 1.0587 uplift are UNCHANGED; sizes remain COMMON across all eight German cities because the product prices an equivalent standard of living. Working: docs/superpowers/research/2026-08-07-small-fixes.md" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "non_cash.parental_leave_paid_pct": { "source": "SGB V para 24i (Mutterschaftsgeld) https://www.gesetze-im-internet.de/sgb_5/__24i.html ; MuSchG para 20 (Zuschuss zum Mutterschaftsgeld, employer tops up to full net) https://www.gesetze-im-internet.de/muschg_2018/__20.html", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the percentage-paid remainder of the 67 weeks: the 14.1 Mutterschutz weeks (6 pre-birth + 8 post-birth), where para 24i SGB V pays EUR 13/day and para 20 MuSchG obliges the employer to top up to full net pay with no ceiling. Was 72, a blend that averaged the capped Elterngeld slice into this field; since 2026-08-06 the 52.87 Elterngeld weeks live in non_cash.parental_leave_capped (65% up to EUR 1,800/month, nothing above EUR 175,000 household income) and parentalLeaveEffectiveRate() blends the slices at the user's salary, so every reader-facing surface already prices the cap and the cliff." }, "non_cash.parental_leave_capped": { "source": "BEEG para 2 (Elterngeld: 65% of prior income, minimum EUR 300 and maximum EUR 1,800 per month) https://www.gesetze-im-internet.de/beeg/__2.html ; BEEG para 1 Abs. 8 (no Elterngeld above EUR 175,000 combined zu versteuerndes Einkommen) https://www.gesetze-im-internet.de/beeg/__1.html ; MuSchG para 20 (the 14 Mutterschutz weeks stay OUTSIDE this block: the employer net top-up has no ceiling) https://www.gesetze-im-internet.de/muschg_2018/__20.html", "last_updated": "2026-08-06", "confidence": "high", "note": "52.87 of the 67 weeks: the Elterngeld pot (60.87 wk) minus the 8 overlap weeks counted at the Mutterschutz rate. The EUR 1,800/month cap (unindexed since Elterngeld's 2007 introduction, re-confirmed 2026-08-06) binds above ~EUR 33,000 gross, i.e. at every ICP salary: at EUR 90,000 this slice pays ~24% of gross and the blended Berlin rate is ~40%; at EUR 150,000 the slice pays ~14% and the blend is ~32%, versus the 72 this field carried before the 2026-08-06 caps rebuild. BEEG para 1 Abs. 8 zeroes the slice for households above EUR 175,000: the engine compares HOUSEHOLD GROSS against the ceiling as a conservative proxy for zu versteuerndes Einkommen (gross exceeds taxable income, so the cliff bites earlier, the direction less favorable to Germany). Same conservative direction on the rate basis: the statute pays 65% of Elterngeld-Netto while the engine applies 65% to gross, and above ~EUR 33,000 gross the cap binds under both readings, so the distinction is immaterial at any ICP salary. Floor EUR 300/month (BEEG para 2). Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Germany section) and 2026-08-06-caps-engine-build.md." }, "non_cash.parental_leave_weeks": { "source": "MuSchG para 3 (Schutzfristen, 6 weeks before and 8 after) https://www.gesetze-im-internet.de/muschg_2018/__3.html ; BEEG para 4 Abs. 3 (12 Monatsbetraege Basiselterngeld plus 2 Partnermonate, gemeinsam) and para 4 Abs. 4 Satz 3 https://www.gesetze-im-internet.de/beeg/__4.html", "last_updated": "2026-08-01", "confidence": "high", "note": "67 wk = 6 additive pre-birth Mutterschutz weeks + the 14-month Elterngeld pot (60.87 wk). Was 61, the Elterngeld pot alone with Mutterschutz omitted. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. OVERLAP RULE APPLIED: the 8 post-birth Mutterschutz weeks are NOT stacked on top, because BEEG para 4 Abs. 4 Satz 3 deems Lebensmonate carrying Mutterschaftsgeld to be consumed Basiselterngeld months (gelten als Monate, fuer die dieser Elternteil Basiselterngeld bezieht). They are counted once, at the higher rate. Germany therefore reaches 67 and not 75. Elterngeld is a SHARED household pot, so it is not doubled. GERMANY HAS NO STATUTORY PAID PATERNITY LEAVE: the Familienstartzeit bill transposing EU Directive 2019/1158 art. 4 was still not law on 2026-08-01, so zero paternity weeks. Excluded: ElterngeldPlus (an election that halves the monthly amount to double the months) and the Partnerschaftsbonus (conditional on both parents working reduced hours)." }, "tax.income_tax_tariff_32a": { "source": "§ 32a Abs. 1 EStG, Veranlagungszeitraum 2026 https://www.gesetze-im-internet.de/estg/__32a.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Statutory continuous tariff, identical nationwide (see berlin for the verbatim zone text and the coefficient cross-check). Replaces the 5-bracket approximation that flattened the 17,799-69,878 progression zone to 33.5%." }, "tax.werbungskosten_pauschbetrag": { "source": "§ 9a Satz 1 Nr. 1 Buchst. a EStG https://www.gesetze-im-internet.de/estg/__9a.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Arbeitnehmer-Pauschbetrag EUR 1,230 per employed person. National, identical to berlin." }, "tax.sonderausgaben_pauschbetrag": { "source": "§ 10c EStG https://www.gesetze-im-internet.de/estg/__10c.html", "last_updated": "2026-07-31", "confidence": "high", "note": "EUR 36, doubled on Zusammenveranlagung. National, identical to berlin." }, "tax.vorsorge_deductible_rate_rv": { "source": "§ 10 Abs. 1 Nr. 2 und Abs. 3 EStG https://www.gesetze-im-internet.de/estg/__10.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Employee RV share 9.3%, 100% deductible since 2023, against the RV Beitragsbemessungsgrenze. National, identical to berlin." }, "tax.vorsorge_deductible_rate_kvpv": { "source": "§ 10 Abs. 1 Nr. 3 und Abs. 4 Satz 4 EStG https://www.gesetze-im-internet.de/estg/__10.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Basis KV 8.75% x 0.96 (Krankengeld haircut) + PV 1.8% = 10.2%, against the KV/PV Beitragsbemessungsgrenze; ALV not deductible. National, identical to berlin." }, "tax.property_tax_rate_pct": { "source": "Cologne Grundsteuer B Hebesatz 475% (NRW Bundesmodell) × 0.31‰ ≈ 0.15%", "last_updated": "2026-05-26", "confidence": "high" }, "tax.sales_tax_rate": { "source": "German VAT 19%", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Investropa + Numbeo Cologne 2026", "last_updated": "2026-05-26", "confidence": "high" }, "healthcare._summary": { "source": "GKV statutory insurance (premium in payroll); Eigenanteil ~€300/yr", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Stadt Köln Elternbeitragssatzung 2025 (income-scaled; last 2 preschool yrs free)", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Cologne 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Deutschlandticket €58/mo (nationwide)", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Cologne", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "CGN->US transatlantic economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Kindergeld €259/mo = €3,108/yr (national, inherited)", "last_updated": "2026-05-26", "confidence": "high", "note": "German Kindergeld (EUR 259/mo per child = EUR 3,108/yr, which is the stored value): universal, NOT means-tested, and it does not taper at professional incomes. Corrected 2026-08-02 from a stale \"EUR 255/mo\", which is the 2025 rate; the Steuerfortentwicklungsgesetz stepped Kindergeld to EUR 255 on 2025-01-01 and EUR 259 on 2026-01-01. The stored value was already correct and was NOT changed." }, "non_cash._summary": { "source": "German labour law (national, inherited); NRW public holidays 11 (Land-specific)", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Cologne: EUR 76,054, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 76,054 / 0.867 = USD 87,721; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3509, R5-rounded down to 0.35. REPLACES 0.57 (2026-08-11), and the interim 0.39 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.57 was 0.9500 of Berlin's 0.6; that ratio x Berlin's re-derived 0.42 = 0.3990, R5-rounded down to 0.39. Berlin's own figure: EUR 91,048 / 0.867 = USD 105,015, / USD 250,000 = 0.4201. This is a BASIS move, not a re-measurement: Cologne's rank inside Germany is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.57, below Berlin (smaller tech ecosystem)." }, "tax.solidaritaetszuschlag_threshold_married": { "source": "Soli Freigrenze 2026 joint assessment EUR 40,700 (2 x 20,350 single; tk.de Solidaritaetszuschlag); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.social_insurance_rate_rvalv": { "source": "RV 9.3% + ALV 1.3% employee share = 10.6%, runs to the higher Beitragsbemessungsgrenze (lohn-info.de Sozialversicherungsbeitraege 2026); verified tax audit 2026-07-14", "confidence": "high", "note": "Carved out of the combined 21.15% employee share; the KV/PV remainder (10.55%) stops at the lower ceiling.", "last_updated": "2026-07-17" }, "tax.social_insurance_cap_rvalv": { "source": "Beitragsbemessungsgrenze RV/ALV 2026 = EUR 101,400 (bundesregierung.de/breg-de/aktuelles/beitragsgemessungsgrenzen-2386514); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.pv_childless_surcharge_rate": { "source": "PV Kinderlosenzuschlag § 55 Abs. 3 SGB XI: +0.6pp employee-borne for childless members 23+ (dak.de Beitrag zur Pflegeversicherung Zuschlag Kinderlose; vdek.com Aenderungen Gesundheitswesen 2026); found by the 2026-08-06 corrections wave", "confidence": "high", "note": "Applied when the household has no children; the under-23 exemption is not modeled (ICP is older). Deductible with the rest of the Basis-Pflegeversicherung.", "last_updated": "2026-08-06" }, "transit.monthly_pass": { "source": "https://www.bundesregierung.de/breg-de/aktuelles/deutschlandticket-2134074", "last_updated": "2026-08-05", "confidence": "high", "note": "Deutschlandticket EUR 63/mo since 2026-01-01 (VMK decision 2025-09-18). Was 58 through Dec 2025." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Destatis Gesundheitsausgabenrechnung 2024 (private households + NPISH EUR 65bn) + Destatis EVS 2023 Konsumausgaben nach Haushaltstyp + GKV Zuzahlungen aggregate EUR 5.18bn (BMG via Aerzteblatt) + SGB V paras 61-62", "last_updated": "2026-08-08", "confidence": "medium", "note": "Level essentially CONFIRMED at EUR 311 per adult (778/capita gross, ~40% in scope after netting long-term care 46%, dental 9%, vision ~3pp). What Germany needed was a curve, not a level. The curve is steep in adults and nearly flat in children because under-18s are exempt from every Zuzahlung except transport AND each child subtracts EUR 9,756 from the income base before the 2%-of-gross Belastungsgrenze, so adding a child LOWERS the household's statutory ceiling in absolute euros. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Destatis Gesundheitsausgabenrechnung 2024 (private households + NPISH EUR 65bn) + Destatis EVS 2023 Konsumausgaben nach Haushaltstyp + GKV Zuzahlungen aggregate EUR 5.18bn (BMG via Aerzteblatt) + SGB V paras 61-62", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, urb_ctran indicator TT1019V, city DE004C. https://ec.europa.eu/eurostat/databrowser/view/urb_ctran/default/table?lang=en", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average time of journey to work (minutes)", "year": "2016", "note": "28.0 min, BASIS YEAR 2016. Same geography mismatch as Berlin: CITY geography, narrower than our metro unit, so biased LOW, which flatters the destination. Survey self-report, standard half-up rounding. for a metro-level German series. METRO-LEVEL GERMAN SERIES: ESTABLISHED HARD NEGATIVE, do not re-run. MiD 2023 Tabelle 6 carries RegioStaR7 breakdowns of population, trips, distance and total time but NO duration by purpose; the national banner table W9 Wegedauer is all-purpose and sits on an \"Ohne regelmaessige berufliche Wege\" base; the small-area products regionalise headline indicators only; and Eurostat publishes no functional-urban-area rows for TT1019V. SrV 2023 was rejected because it publishes an ALL-PURPOSE trip duration, and swapping an all-purpose duration for a work-purpose one is a UNIT error, which is worse than the geography error we accepted. SrV is therefore used only as rank-order corroboration, and it agrees with Eurostat against our old stored spread: both put Frankfurt near the BOTTOM of the big five, where we had it second highest. PROVISIONAL: this is one of the three large German drops and it is the flattering direction, so re-check it first if a metro-level series ever appears." }, "non_cash.sick_leave": { "source": "Entgeltfortzahlungsgesetz (EFZG) §3 (employer phase); SGB V §47 (Krankengeld; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "housing.utilities_monthly": { "source": "DMB NRW Betriebskostenspiegel 2,71 EUR/sqm (state-level) + the municipal-fee model, https://mieterbund.de/aktuelles/meldungen/aktueller-betriebskostenspiegel-fuer-deutschland/", "confidence": "medium", "note": "2BR reference: state 2,71 gives a 288 floor; the municipal-fee model puts Koeln 1.4% above Duesseldorf, whose 303 was independently confirmed 2026-08-08, so 305 is the Duesseldorf-consistent value. Numbeo discarded here (196 to 744 spread). Cologne 2026 fees rise further (Grundsteuer 475% to 550%, waste +6 to +15.6%), so 305 is not aggressive.", "last_updated": "2026-08-08" }, "tax.retirement_deduction_cap": { "source": "EStG s.10 Abs 3 https://www.gesetze-im-internet.de/estg/__10.html ; SVBezGrV 2026 s.4 Abs 1 https://www.gesetze-im-internet.de/svbezgrv_2026/BJNR1160A0025.html", "last_updated": "2026-08-11", "confidence": "high", "note": "Ruerup/Basisrente Hoechstbetrag, 2026: EUR 30,825.60 single (knappschaftliche RV BBG 124,800 x 24.7%). A DEDUCTION ceiling and a COMBINED one: employee AND employer statutory pension contributions consume the same pot. less_employer_pct 0.186 = 2 x 9.3% and employer_base_cap 101,400 is the allgemeine RV BBG, because computeGermanyTaxes already deducts the employee 9.3% via vorsorge; a naked max would hand a EUR 95k earner about EUR 8.8k of phantom deduction. Riester (EUR 2,100) and bAV Entgeltumwandlung (8% of the BBG, EUR 8,112) are deliberately NOT stacked. Married couples: the statute doubles the ceiling to 61,651.20 even on one income; our per-earner cap under-deducts there, which errs toward higher tax." } }, "dusseldorf": { "housing.median_rent_by_bedroom": { "source": "Landeshauptstadt Duesseldorf, Wohnungsmarktbericht 2025, chapter 6.2 (published 2026-05-12; \"Bei den dargestellten Mietwerten handelt es sich um Angebotspreise (Nettokaltmiete)\"; median; n = 10,370 Mietangebote; calendar 2024; VALUE AG Marktdatenbank; city area) http://ris-oparl.itk-rheinland.de/Oparl/bodies/0015/downloadfiles/00586901.pdf ; landing page https://www.duesseldorf-radar.de/informationsvorlage/wohnungsmarktbericht-2025 ; 2026 uprating and per-room gradient VALUE Marktdatenbank 2026 https://mietspiegeltabelle.de/mietspiegel-duesseldorf/", "last_updated": "2026-08-07", "confidence": "high", "note": "CORRECTED 2026-08-02 from 650/800/1100/1500, replacing an \"Investropa + Numbeo\" citation. This is the only one of the seven where the stored values were too LOW at every bedroom count. The city own report gives a 2024 median asking rent of 13.10 EUR/sqm netto kalt (13.05 in Mehrfamilienhaeuser, which are 98% of listings) rising 6.1% in 2024 alone, and VALUE 2026 gives 15.00. Per-room rates 15.93/14.53/14.89/15.93 x 1.0587 = 16.86/15.38/15.76/16.86. Studio 650 to 710, 1BR 800 to 920, 2BR 1100 to 1340, 3BR 1500 to 1770. Duesseldorf is the best-documented city in the German set and its report is the source of the size evidence used across all seven: published average sizes of offered flats by Zimmer count in Mehrfamilienhaeusern are 1-Zi 37, 2-Zi 58, 3-Zi 83, 4-Zi 111 sqm. Its per-room medians show the classic U (1-Zi 14.09, trough at 2-Zi 12.64, 3-Zi 13.00, 4-Zi 13.89, 5+ 15.11), and re-weighting the four 2026 rates by the report own listing mix returns 14.97 against an independently stated 15.00. REFUTED here: the assumption that Duesseldorf outranks Cologne. Two independent 2026 series say the opposite (VALUE Cologne 15.93 against Duesseldorf 15.00; ImmoScout24 WohnBarometer Q1 2026 Cologne 13.12 against Duesseldorf 12.85), and Duesseldorf leads only on 4-Zimmer and new build, which is why its 3BR ends above Cologne while its studio and 1BR end below. Stock contrast: Zensus 2022 median Bestandsmiete 8.88, so asking runs about 47% above stock. Per-room rates come from the VALUE Marktdatenbank 2026 cut published at mietspiegeltabelle.de, explicitly nettokalt and explicitly market/asking rather than the qualified municipal Mietspiegel. That is the same database family the Berlin correction leant on. Leipzig proves what it serves: its 10.45 equals VALUE own published MEAN of 10.49 against a median of 10.00, so this basis runs about 5% above median, which is the conservative direction. Method is the shipped Berlin recipe re-run per city, so all eight German cities sit on one footing: rent = size x city per-room EUR/sqm x 1.0587. The 1.0587 is Berlin own uplift over this same source (Berlin shipped a 16.6 base against 15.68 here), and re-running it reproduces Berlin 760/970/1360/1750 to the euro. Sizes are held constant across cities at 42/60/85/105 sqm on the German Zimmer convention (our 1BR = 2-Zimmer-Wohnung), because the product prices an equivalent standard of living rather than how much space local households ration themselves into at local prices. Duesseldorf publishes real offered-flat sizes of 37/58/83/111, so the fixed 42 makes the studio the most conservative cell and the fixed 105 makes the 3BR mildly generous on size; the two roughly cancel. Basis is Angebotsmiete (new lease), netto kalt, never the Mietspiegel stock rent. Rents exclude utilities, insurance and internet, which are separate fields. Values are a CEILING per the conservative value-bias rule, not central estimates. Working: docs/superpowers/research/2026-08-02-german-city-rents.md SIZE LADDER CORRECTED 2026-08-07 (small-fixes wave 1, coordinated eight-city pass incl. Berlin): the common dwelling-size ladder moves from the inherited 42/60/85/105 sqm to 37/58/83/111 sqm, Duesseldorf's PUBLISHED offered-flat sizes (Wohnungsmarktbericht 2025 Abb. 54, MFH column, n=10,370 listings; the only published per-Zimmer size table in the German set, and the conservative top of the seven-city evidence range at 3-Zi 83 and 4-Zi 111). Root cause of the old ladder: German official statistics count a closed kitchen over 6 sqm as a room, so census room distributions ran one room high against listing Zimmer counts; the 42 sqm studio was priced at the 1-Zimmer premium rate measured on 33-37 sqm stock (double-counting the small-flat premium) and the 4-Zimmer was undersized at 105 vs the real ~111. Values 710/920/1340/1770 -> 620/890/1310/1870. Per-room rates and the 1.0587 uplift are UNCHANGED; sizes remain COMMON across all eight German cities because the product prices an equivalent standard of living. Working: docs/superpowers/research/2026-08-07-small-fixes.md" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "non_cash.parental_leave_paid_pct": { "source": "SGB V para 24i (Mutterschaftsgeld) https://www.gesetze-im-internet.de/sgb_5/__24i.html ; MuSchG para 20 (Zuschuss zum Mutterschaftsgeld, employer tops up to full net) https://www.gesetze-im-internet.de/muschg_2018/__20.html", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the percentage-paid remainder of the 67 weeks: the 14.1 Mutterschutz weeks (6 pre-birth + 8 post-birth), where para 24i SGB V pays EUR 13/day and para 20 MuSchG obliges the employer to top up to full net pay with no ceiling. Was 72, a blend that averaged the capped Elterngeld slice into this field; since 2026-08-06 the 52.87 Elterngeld weeks live in non_cash.parental_leave_capped (65% up to EUR 1,800/month, nothing above EUR 175,000 household income) and parentalLeaveEffectiveRate() blends the slices at the user's salary, so every reader-facing surface already prices the cap and the cliff." }, "non_cash.parental_leave_capped": { "source": "BEEG para 2 (Elterngeld: 65% of prior income, minimum EUR 300 and maximum EUR 1,800 per month) https://www.gesetze-im-internet.de/beeg/__2.html ; BEEG para 1 Abs. 8 (no Elterngeld above EUR 175,000 combined zu versteuerndes Einkommen) https://www.gesetze-im-internet.de/beeg/__1.html ; MuSchG para 20 (the 14 Mutterschutz weeks stay OUTSIDE this block: the employer net top-up has no ceiling) https://www.gesetze-im-internet.de/muschg_2018/__20.html", "last_updated": "2026-08-06", "confidence": "high", "note": "52.87 of the 67 weeks: the Elterngeld pot (60.87 wk) minus the 8 overlap weeks counted at the Mutterschutz rate. The EUR 1,800/month cap (unindexed since Elterngeld's 2007 introduction, re-confirmed 2026-08-06) binds above ~EUR 33,000 gross, i.e. at every ICP salary: at EUR 90,000 this slice pays ~24% of gross and the blended Berlin rate is ~40%; at EUR 150,000 the slice pays ~14% and the blend is ~32%, versus the 72 this field carried before the 2026-08-06 caps rebuild. BEEG para 1 Abs. 8 zeroes the slice for households above EUR 175,000: the engine compares HOUSEHOLD GROSS against the ceiling as a conservative proxy for zu versteuerndes Einkommen (gross exceeds taxable income, so the cliff bites earlier, the direction less favorable to Germany). Same conservative direction on the rate basis: the statute pays 65% of Elterngeld-Netto while the engine applies 65% to gross, and above ~EUR 33,000 gross the cap binds under both readings, so the distinction is immaterial at any ICP salary. Floor EUR 300/month (BEEG para 2). Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Germany section) and 2026-08-06-caps-engine-build.md." }, "non_cash.parental_leave_weeks": { "source": "MuSchG para 3 (Schutzfristen, 6 weeks before and 8 after) https://www.gesetze-im-internet.de/muschg_2018/__3.html ; BEEG para 4 Abs. 3 (12 Monatsbetraege Basiselterngeld plus 2 Partnermonate, gemeinsam) and para 4 Abs. 4 Satz 3 https://www.gesetze-im-internet.de/beeg/__4.html", "last_updated": "2026-08-01", "confidence": "high", "note": "67 wk = 6 additive pre-birth Mutterschutz weeks + the 14-month Elterngeld pot (60.87 wk). Was 61, the Elterngeld pot alone with Mutterschutz omitted. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. OVERLAP RULE APPLIED: the 8 post-birth Mutterschutz weeks are NOT stacked on top, because BEEG para 4 Abs. 4 Satz 3 deems Lebensmonate carrying Mutterschaftsgeld to be consumed Basiselterngeld months (gelten als Monate, fuer die dieser Elternteil Basiselterngeld bezieht). They are counted once, at the higher rate. Germany therefore reaches 67 and not 75. Elterngeld is a SHARED household pot, so it is not doubled. GERMANY HAS NO STATUTORY PAID PATERNITY LEAVE: the Familienstartzeit bill transposing EU Directive 2019/1158 art. 4 was still not law on 2026-08-01, so zero paternity weeks. Excluded: ElterngeldPlus (an election that halves the monthly amount to double the months) and the Partnerschaftsbonus (conditional on both parents working reduced hours)." }, "tax.income_tax_tariff_32a": { "source": "§ 32a Abs. 1 EStG, Veranlagungszeitraum 2026 https://www.gesetze-im-internet.de/estg/__32a.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Statutory continuous tariff, identical nationwide (see berlin for the verbatim zone text and the coefficient cross-check). Replaces the 5-bracket approximation that flattened the 17,799-69,878 progression zone to 33.5%." }, "tax.werbungskosten_pauschbetrag": { "source": "§ 9a Satz 1 Nr. 1 Buchst. a EStG https://www.gesetze-im-internet.de/estg/__9a.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Arbeitnehmer-Pauschbetrag EUR 1,230 per employed person. National, identical to berlin." }, "tax.sonderausgaben_pauschbetrag": { "source": "§ 10c EStG https://www.gesetze-im-internet.de/estg/__10c.html", "last_updated": "2026-07-31", "confidence": "high", "note": "EUR 36, doubled on Zusammenveranlagung. National, identical to berlin." }, "tax.vorsorge_deductible_rate_rv": { "source": "§ 10 Abs. 1 Nr. 2 und Abs. 3 EStG https://www.gesetze-im-internet.de/estg/__10.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Employee RV share 9.3%, 100% deductible since 2023, against the RV Beitragsbemessungsgrenze. National, identical to berlin." }, "tax.vorsorge_deductible_rate_kvpv": { "source": "§ 10 Abs. 1 Nr. 3 und Abs. 4 Satz 4 EStG https://www.gesetze-im-internet.de/estg/__10.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Basis KV 8.75% x 0.96 (Krankengeld haircut) + PV 1.8% = 10.2%, against the KV/PV Beitragsbemessungsgrenze; ALV not deductible. National, identical to berlin." }, "tax.property_tax_rate_pct": { "source": "Düsseldorf Grundsteuer B Hebesatz 374% (NRW Bundesmodell) × 0.31‰ ≈ 0.12%", "last_updated": "2026-05-26", "confidence": "high", "note": "Notably low Hebesatz among major German cities." }, "tax.sales_tax_rate": { "source": "German VAT 19%", "last_updated": "2026-05-26", "confidence": "high" }, "housing.utilities_monthly": { "source": "DMB NRW Betriebskostenspiegel 2024 (avg 2.71 EUR/sqm/mo, heating+WW 1.32, https://www.mieterbund-nrw.de/aktuelles2/pm-bksnrw2024); Numbeo Duesseldorf 85-sqm mean 318.99 (range 230-450, fetched 2026-08-08)", "last_updated": "2026-08-08", "confidence": "high", "note": "2BR (~83 sqm) reference incl. household electricity. VERIFIED 2026-08-08, value CONFIRMED at 303: implies a 2.89 EUR/sqm wedge (+6.6% over the NRW average, a normal big-city premium), between the bottom-up 288-299 (2.71-2.85 x 83 + ~63 electricity) and the Numbeo mean 319. The 2026-08-02 'not credible vs Munich' flag mis-read the basis the same way as Leipzig's (household electricity sits outside the Warmmiete wedge). Derivation: docs/superpowers/research/2026-08-08-utilities-values.md" }, "housing._summary": { "source": "Investropa + Numbeo Düsseldorf 2026", "last_updated": "2026-05-26", "confidence": "high" }, "healthcare._summary": { "source": "GKV statutory insurance (premium in payroll); Eigenanteil ~€300/yr", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Düsseldorf Elternbeitragssatzung 2025 + kita.nrw.de (Ü3 free)", "last_updated": "2026-05-26", "confidence": "medium", "note": "Preschool (age 3+) free in NRW; only U3 income-scaled." }, "food._summary": { "source": "Numbeo Düsseldorf 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Deutschlandticket €58/mo (nationwide)", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Düsseldorf", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "DUS->US transatlantic economy 2025-26 (direct service)", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Kindergeld €259/mo = €3,108/yr (national, inherited)", "last_updated": "2026-05-26", "confidence": "high", "note": "German Kindergeld (EUR 259/mo per child = EUR 3,108/yr, which is the stored value): universal, NOT means-tested, and it does not taper at professional incomes. Corrected 2026-08-02 from a stale \"EUR 255/mo\", which is the 2025 rate; the Steuerfortentwicklungsgesetz stepped Kindergeld to EUR 255 on 2025-01-01 and EUR 259 on 2026-01-01. The stored value was already correct and was NOT changed." }, "non_cash._summary": { "source": "German labour law (national, inherited); NRW public holidays 11 (Land-specific)", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Düsseldorf: EUR 74,741, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 74,741 / 0.867 = USD 86,206; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3448, R5-rounded down to 0.34. REPLACES 0.61 (2026-08-11), and the interim 0.42 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. REFUTES the stored 0.61. Measured 0.829 of Berlin; the stored pair implied 1.017. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.61 was 1.0167 of Berlin's 0.6; that ratio x Berlin's re-derived 0.42 = 0.4270, R5-rounded down to 0.42. Berlin's own figure: EUR 91,048 / 0.867 = USD 105,015, / USD 250,000 = 0.4201. This is a BASIS move, not a re-measurement: Düsseldorf's rank inside Germany is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.61, finance/trade hub (Henkel, Metro, Vodafone DE)." }, "tax.solidaritaetszuschlag_threshold_married": { "source": "Soli Freigrenze 2026 joint assessment EUR 40,700 (2 x 20,350 single; tk.de Solidaritaetszuschlag); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.social_insurance_rate_rvalv": { "source": "RV 9.3% + ALV 1.3% employee share = 10.6%, runs to the higher Beitragsbemessungsgrenze (lohn-info.de Sozialversicherungsbeitraege 2026); verified tax audit 2026-07-14", "confidence": "high", "note": "Carved out of the combined 21.15% employee share; the KV/PV remainder (10.55%) stops at the lower ceiling.", "last_updated": "2026-07-17" }, "tax.social_insurance_cap_rvalv": { "source": "Beitragsbemessungsgrenze RV/ALV 2026 = EUR 101,400 (bundesregierung.de/breg-de/aktuelles/beitragsgemessungsgrenzen-2386514); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.pv_childless_surcharge_rate": { "source": "PV Kinderlosenzuschlag § 55 Abs. 3 SGB XI: +0.6pp employee-borne for childless members 23+ (dak.de Beitrag zur Pflegeversicherung Zuschlag Kinderlose; vdek.com Aenderungen Gesundheitswesen 2026); found by the 2026-08-06 corrections wave", "confidence": "high", "note": "Applied when the household has no children; the under-23 exemption is not modeled (ICP is older). Deductible with the rest of the Basis-Pflegeversicherung.", "last_updated": "2026-08-06" }, "transit.monthly_pass": { "source": "https://www.bundesregierung.de/breg-de/aktuelles/deutschlandticket-2134074", "last_updated": "2026-08-05", "confidence": "high", "note": "Deutschlandticket EUR 63/mo since 2026-01-01 (VMK decision 2025-09-18). Was 58 through Dec 2025." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Destatis Gesundheitsausgabenrechnung 2024 (private households + NPISH EUR 65bn) + Destatis EVS 2023 Konsumausgaben nach Haushaltstyp + GKV Zuzahlungen aggregate EUR 5.18bn (BMG via Aerzteblatt) + SGB V paras 61-62", "last_updated": "2026-08-08", "confidence": "medium", "note": "Level essentially CONFIRMED at EUR 311 per adult (778/capita gross, ~40% in scope after netting long-term care 46%, dental 9%, vision ~3pp). What Germany needed was a curve, not a level. The curve is steep in adults and nearly flat in children because under-18s are exempt from every Zuzahlung except transport AND each child subtracts EUR 9,756 from the income base before the 2%-of-gross Belastungsgrenze, so adding a child LOWERS the household's statutory ceiling in absolute euros. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Destatis Gesundheitsausgabenrechnung 2024 (private households + NPISH EUR 65bn) + Destatis EVS 2023 Konsumausgaben nach Haushaltstyp + GKV Zuzahlungen aggregate EUR 5.18bn (BMG via Aerzteblatt) + SGB V paras 61-62", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, urb_ctran indicator TT1019V, city DE011C. https://ec.europa.eu/eurostat/databrowser/view/urb_ctran/default/table?lang=en", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average time of journey to work (minutes)", "year": "2016", "note": "26.2 min, BASIS YEAR 2016. CITY geography, narrower than our metro unit, so biased LOW and flattering. Survey self-report, standard half-up rounding. METRO-LEVEL GERMAN SERIES: ESTABLISHED HARD NEGATIVE, do not re-run. MiD 2023 Tabelle 6 carries RegioStaR7 breakdowns of population, trips, distance and total time but NO duration by purpose; the national banner table W9 Wegedauer is all-purpose and sits on an \"Ohne regelmaessige berufliche Wege\" base; the small-area products regionalise headline indicators only; and Eurostat publishes no functional-urban-area rows for TT1019V. SrV 2023 was rejected because it publishes an ALL-PURPOSE trip duration, and swapping an all-purpose duration for a work-purpose one is a UNIT error, which is worse than the geography error we accepted. SrV is therefore used only as rank-order corroboration, and it agrees with Eurostat against our old stored spread: both put Frankfurt near the BOTTOM of the big five, where we had it second highest." }, "non_cash.sick_leave": { "source": "Entgeltfortzahlungsgesetz (EFZG) §3 (employer phase); SGB V §47 (Krankengeld; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "EStG s.10 Abs 3 https://www.gesetze-im-internet.de/estg/__10.html ; SVBezGrV 2026 s.4 Abs 1 https://www.gesetze-im-internet.de/svbezgrv_2026/BJNR1160A0025.html", "last_updated": "2026-08-11", "confidence": "high", "note": "Ruerup/Basisrente Hoechstbetrag, 2026: EUR 30,825.60 single (knappschaftliche RV BBG 124,800 x 24.7%). A DEDUCTION ceiling and a COMBINED one: employee AND employer statutory pension contributions consume the same pot. less_employer_pct 0.186 = 2 x 9.3% and employer_base_cap 101,400 is the allgemeine RV BBG, because computeGermanyTaxes already deducts the employee 9.3% via vorsorge; a naked max would hand a EUR 95k earner about EUR 8.8k of phantom deduction. Riester (EUR 2,100) and bAV Entgeltumwandlung (8% of the BBG, EUR 8,112) are deliberately NOT stacked. Married couples: the statute doubles the ceiling to 61,651.20 even on one income; our per-earner cap under-deducts there, which errs toward higher tax." } }, "stuttgart": { "housing.median_rent_by_bedroom": { "source": "Landeshauptstadt Stuttgart, Statistisches Amt and Amt fuer Stadtplanung und Wohnen, report Wohnungsmarkt Stuttgart 2025 (Angebotsmiete H1 2024, netto kalt, city area) https://www.stuttgart.de/service/aktuelle-meldungen/2025/juli/stadt-legt-neuen-bericht-zum-wohnungsmarkt-vor ; official Mietspiegel Stuttgart 2025/2026, Grundwert size gradient, \"Im Mietspiegel werden Nettokaltmieten ausgewiesen\" https://www.stuttgart.de/medien/ibs/mietspiegel_2025_2026.pdf ; per-room gradient VALUE Marktdatenbank 2026 https://mietspiegeltabelle.de/mietspiegel-stuttgart/", "last_updated": "2026-08-07", "confidence": "high", "note": "CORRECTED 2026-08-02 from 850/1150/1450/1850, replacing a \"Nationsbase + JLL + Numbeo\" citation with no URL. The city own 2025 report gives an H1 2024 Angebotsmiete of 15.51 EUR/sqm netto kalt against a Mietspiegel ortsuebliche Vergleichsmiete of 11.15, a gap of 4.36 or 39%; uprating 15.51 two years at the observed pace lands near the VALUE 2026 figure of 16.43, so the two agree. Per-room rates 19.31/16.41/15.53/15.50 x 1.0587 = 20.44/17.37/16.44/16.41. Studio 850 to 860, 1BR 1150 to 1040, 2BR 1450 to 1400, 3BR 1850 to 1720. Stuttgart was the ladder that had been claimed to over-scale and does not: its 1-to-2BR and 2-to-3BR steps sat near the 25th percentile of the whole 178-city dataset, flatter than median, and the correction is a modest re-basing. Recorded conflict, unresolved on purpose: the city own report ranks Stuttgart FOURTH on asking rents \"hinter Muenchen, Frankfurt am Main und Berlin\", so below Berlin, while the 2026 VALUE cut puts Stuttgart at 16.43 against Berlin 15.68. The two differ by about 5%, which is inside vintage and source noise. The official Mietspiegel Grundwert table gives a clean size gradient indexed to the 40-to-70 sqm band: 21-30 sqm 1.272, 30-40 1.133, 40-70 1.000, 70-90 0.943, 90-115 0.920, 115+ 0.922, identical in shape across all four building-age columns. Per-room rates come from the VALUE Marktdatenbank 2026 cut published at mietspiegeltabelle.de, explicitly nettokalt and explicitly market/asking rather than the qualified municipal Mietspiegel. That is the same database family the Berlin correction leant on. Leipzig proves what it serves: its 10.45 equals VALUE own published MEAN of 10.49 against a median of 10.00, so this basis runs about 5% above median, which is the conservative direction. Method is the shipped Berlin recipe re-run per city, so all eight German cities sit on one footing: rent = size x city per-room EUR/sqm x 1.0587. The 1.0587 is Berlin own uplift over this same source (Berlin shipped a 16.6 base against 15.68 here), and re-running it reproduces Berlin 760/970/1360/1750 to the euro. Sizes are held constant across cities at 42/60/85/105 sqm on the German Zimmer convention (our 1BR = 2-Zimmer-Wohnung), because the product prices an equivalent standard of living rather than how much space local households ration themselves into at local prices. Duesseldorf publishes real offered-flat sizes of 37/58/83/111, so the fixed 42 makes the studio the most conservative cell and the fixed 105 makes the 3BR mildly generous on size; the two roughly cancel. Basis is Angebotsmiete (new lease), netto kalt, never the Mietspiegel stock rent. Rents exclude utilities, insurance and internet, which are separate fields. Values are a CEILING per the conservative value-bias rule, not central estimates. Working: docs/superpowers/research/2026-08-02-german-city-rents.md SIZE LADDER CORRECTED 2026-08-07 (small-fixes wave 1, coordinated eight-city pass incl. Berlin): the common dwelling-size ladder moves from the inherited 42/60/85/105 sqm to 37/58/83/111 sqm, Duesseldorf's PUBLISHED offered-flat sizes (Wohnungsmarktbericht 2025 Abb. 54, MFH column, n=10,370 listings; the only published per-Zimmer size table in the German set, and the conservative top of the seven-city evidence range at 3-Zi 83 and 4-Zi 111). Root cause of the old ladder: German official statistics count a closed kitchen over 6 sqm as a room, so census room distributions ran one room high against listing Zimmer counts; the 42 sqm studio was priced at the 1-Zimmer premium rate measured on 33-37 sqm stock (double-counting the small-flat premium) and the 4-Zimmer was undersized at 105 vs the real ~111. Values 860/1040/1400/1720 -> 760/1010/1360/1820. Per-room rates and the 1.0587 uplift are UNCHANGED; sizes remain COMMON across all eight German cities because the product prices an equivalent standard of living. Working: docs/superpowers/research/2026-08-07-small-fixes.md" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "non_cash.parental_leave_paid_pct": { "source": "SGB V para 24i (Mutterschaftsgeld) https://www.gesetze-im-internet.de/sgb_5/__24i.html ; MuSchG para 20 (Zuschuss zum Mutterschaftsgeld, employer tops up to full net) https://www.gesetze-im-internet.de/muschg_2018/__20.html", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the percentage-paid remainder of the 67 weeks: the 14.1 Mutterschutz weeks (6 pre-birth + 8 post-birth), where para 24i SGB V pays EUR 13/day and para 20 MuSchG obliges the employer to top up to full net pay with no ceiling. Was 72, a blend that averaged the capped Elterngeld slice into this field; since 2026-08-06 the 52.87 Elterngeld weeks live in non_cash.parental_leave_capped (65% up to EUR 1,800/month, nothing above EUR 175,000 household income) and parentalLeaveEffectiveRate() blends the slices at the user's salary, so every reader-facing surface already prices the cap and the cliff." }, "non_cash.parental_leave_capped": { "source": "BEEG para 2 (Elterngeld: 65% of prior income, minimum EUR 300 and maximum EUR 1,800 per month) https://www.gesetze-im-internet.de/beeg/__2.html ; BEEG para 1 Abs. 8 (no Elterngeld above EUR 175,000 combined zu versteuerndes Einkommen) https://www.gesetze-im-internet.de/beeg/__1.html ; MuSchG para 20 (the 14 Mutterschutz weeks stay OUTSIDE this block: the employer net top-up has no ceiling) https://www.gesetze-im-internet.de/muschg_2018/__20.html", "last_updated": "2026-08-06", "confidence": "high", "note": "52.87 of the 67 weeks: the Elterngeld pot (60.87 wk) minus the 8 overlap weeks counted at the Mutterschutz rate. The EUR 1,800/month cap (unindexed since Elterngeld's 2007 introduction, re-confirmed 2026-08-06) binds above ~EUR 33,000 gross, i.e. at every ICP salary: at EUR 90,000 this slice pays ~24% of gross and the blended Berlin rate is ~40%; at EUR 150,000 the slice pays ~14% and the blend is ~32%, versus the 72 this field carried before the 2026-08-06 caps rebuild. BEEG para 1 Abs. 8 zeroes the slice for households above EUR 175,000: the engine compares HOUSEHOLD GROSS against the ceiling as a conservative proxy for zu versteuerndes Einkommen (gross exceeds taxable income, so the cliff bites earlier, the direction less favorable to Germany). Same conservative direction on the rate basis: the statute pays 65% of Elterngeld-Netto while the engine applies 65% to gross, and above ~EUR 33,000 gross the cap binds under both readings, so the distinction is immaterial at any ICP salary. Floor EUR 300/month (BEEG para 2). Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Germany section) and 2026-08-06-caps-engine-build.md." }, "non_cash.parental_leave_weeks": { "source": "MuSchG para 3 (Schutzfristen, 6 weeks before and 8 after) https://www.gesetze-im-internet.de/muschg_2018/__3.html ; BEEG para 4 Abs. 3 (12 Monatsbetraege Basiselterngeld plus 2 Partnermonate, gemeinsam) and para 4 Abs. 4 Satz 3 https://www.gesetze-im-internet.de/beeg/__4.html", "last_updated": "2026-08-01", "confidence": "high", "note": "67 wk = 6 additive pre-birth Mutterschutz weeks + the 14-month Elterngeld pot (60.87 wk). Was 61, the Elterngeld pot alone with Mutterschutz omitted. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. OVERLAP RULE APPLIED: the 8 post-birth Mutterschutz weeks are NOT stacked on top, because BEEG para 4 Abs. 4 Satz 3 deems Lebensmonate carrying Mutterschaftsgeld to be consumed Basiselterngeld months (gelten als Monate, fuer die dieser Elternteil Basiselterngeld bezieht). They are counted once, at the higher rate. Germany therefore reaches 67 and not 75. Elterngeld is a SHARED household pot, so it is not doubled. GERMANY HAS NO STATUTORY PAID PATERNITY LEAVE: the Familienstartzeit bill transposing EU Directive 2019/1158 art. 4 was still not law on 2026-08-01, so zero paternity weeks. Excluded: ElterngeldPlus (an election that halves the monthly amount to double the months) and the Partnerschaftsbonus (conditional on both parents working reduced hours)." }, "tax.income_tax_tariff_32a": { "source": "§ 32a Abs. 1 EStG, Veranlagungszeitraum 2026 https://www.gesetze-im-internet.de/estg/__32a.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Statutory continuous tariff, identical nationwide (see berlin for the verbatim zone text and the coefficient cross-check). Replaces the 5-bracket approximation that flattened the 17,799-69,878 progression zone to 33.5%." }, "tax.werbungskosten_pauschbetrag": { "source": "§ 9a Satz 1 Nr. 1 Buchst. a EStG https://www.gesetze-im-internet.de/estg/__9a.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Arbeitnehmer-Pauschbetrag EUR 1,230 per employed person. National, identical to berlin." }, "tax.sonderausgaben_pauschbetrag": { "source": "§ 10c EStG https://www.gesetze-im-internet.de/estg/__10c.html", "last_updated": "2026-07-31", "confidence": "high", "note": "EUR 36, doubled on Zusammenveranlagung. National, identical to berlin." }, "tax.vorsorge_deductible_rate_rv": { "source": "§ 10 Abs. 1 Nr. 2 und Abs. 3 EStG https://www.gesetze-im-internet.de/estg/__10.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Employee RV share 9.3%, 100% deductible since 2023, against the RV Beitragsbemessungsgrenze. National, identical to berlin." }, "tax.vorsorge_deductible_rate_kvpv": { "source": "§ 10 Abs. 1 Nr. 3 und Abs. 4 Satz 4 EStG https://www.gesetze-im-internet.de/estg/__10.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Basis KV 8.75% x 0.96 (Krankengeld haircut) + PV 1.8% = 10.2%, against the KV/PV Beitragsbemessungsgrenze; ALV not deductible. National, identical to berlin." }, "tax.property_tax_rate_pct": { "source": "Stuttgart Grundsteuer B Hebesatz 160% (Baden-Württemberg Bodenwertmodell)", "last_updated": "2026-05-26", "confidence": "medium", "note": "Bodenwertmodell (land-value), not Bundesmodell — set to German-norm ~0.15% effective." }, "tax.sales_tax_rate": { "source": "German VAT 19%", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Nationsbase + JLL + Numbeo Stuttgart 2025-26", "last_updated": "2026-05-26", "confidence": "high" }, "healthcare._summary": { "source": "GKV statutory insurance (premium in payroll); Eigenanteil ~€300/yr", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "stuttgart.de Elternbeiträge 2025 (income-scaled, no blanket subsidy; ~30% rise planned Sep 2026)", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Stuttgart 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Deutschlandticket €58/mo (nationwide)", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Stuttgart", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "STR->US transatlantic economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Kindergeld €259/mo = €3,108/yr (national, inherited)", "last_updated": "2026-05-26", "confidence": "high", "note": "German Kindergeld (EUR 259/mo per child = EUR 3,108/yr, which is the stored value): universal, NOT means-tested, and it does not taper at professional incomes. Corrected 2026-08-02 from a stale \"EUR 255/mo\", which is the 2025 rate; the Steuerfortentwicklungsgesetz stepped Kindergeld to EUR 255 on 2025-01-01 and EUR 259 on 2026-01-01. The stored value was already correct and was NOT changed." }, "non_cash._summary": { "source": "German labour law (national, inherited); Baden-Württemberg public holidays 12 (Land-specific)", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Stuttgart: EUR 84,496, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 84,496 / 0.867 = USD 97,458; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3898, R5-rounded down to 0.38. REPLACES 0.65 (2026-08-11), and the interim 0.45 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. REFUTES the stored 0.65. Measured 0.937 of Berlin; the stored pair implied 1.083, i.e. Stuttgart 8% ABOVE Berlin where it measures 6% below. The likely cause is that the note reasoned about automotive engineering pay, not software. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.65 was 1.0833 of Berlin's 0.6; that ratio x Berlin's re-derived 0.42 = 0.4550, R5-rounded down to 0.45. Berlin's own figure: EUR 91,048 / 0.867 = USD 105,015, / USD 250,000 = 0.4201. This is a BASIS move, not a re-measurement: Stuttgart's rank inside Germany is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.65, automotive/engineering premium (Porsche, Bosch, Mercedes)." }, "tax.solidaritaetszuschlag_threshold_married": { "source": "Soli Freigrenze 2026 joint assessment EUR 40,700 (2 x 20,350 single; tk.de Solidaritaetszuschlag); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.social_insurance_rate_rvalv": { "source": "RV 9.3% + ALV 1.3% employee share = 10.6%, runs to the higher Beitragsbemessungsgrenze (lohn-info.de Sozialversicherungsbeitraege 2026); verified tax audit 2026-07-14", "confidence": "high", "note": "Carved out of the combined 21.15% employee share; the KV/PV remainder (10.55%) stops at the lower ceiling.", "last_updated": "2026-07-17" }, "tax.social_insurance_cap_rvalv": { "source": "Beitragsbemessungsgrenze RV/ALV 2026 = EUR 101,400 (bundesregierung.de/breg-de/aktuelles/beitragsgemessungsgrenzen-2386514); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.pv_childless_surcharge_rate": { "source": "PV Kinderlosenzuschlag § 55 Abs. 3 SGB XI: +0.6pp employee-borne for childless members 23+ (dak.de Beitrag zur Pflegeversicherung Zuschlag Kinderlose; vdek.com Aenderungen Gesundheitswesen 2026); found by the 2026-08-06 corrections wave", "confidence": "high", "note": "Applied when the household has no children; the under-23 exemption is not modeled (ICP is older). Deductible with the rest of the Basis-Pflegeversicherung.", "last_updated": "2026-08-06" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Destatis Gesundheitsausgabenrechnung 2024 (private households + NPISH EUR 65bn) + Destatis EVS 2023 Konsumausgaben nach Haushaltstyp + GKV Zuzahlungen aggregate EUR 5.18bn (BMG via Aerzteblatt) + SGB V paras 61-62", "last_updated": "2026-08-08", "confidence": "medium", "note": "Level essentially CONFIRMED at EUR 311 per adult (778/capita gross, ~40% in scope after netting long-term care 46%, dental 9%, vision ~3pp). What Germany needed was a curve, not a level. The curve is steep in adults and nearly flat in children because under-18s are exempt from every Zuzahlung except transport AND each child subtracts EUR 9,756 from the income base before the 2%-of-gross Belastungsgrenze, so adding a child LOWERS the household's statutory ceiling in absolute euros. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Destatis Gesundheitsausgabenrechnung 2024 (private households + NPISH EUR 65bn) + Destatis EVS 2023 Konsumausgaben nach Haushaltstyp + GKV Zuzahlungen aggregate EUR 5.18bn (BMG via Aerzteblatt) + SGB V paras 61-62", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, urb_ctran indicator TT1019V, city DE007C. https://ec.europa.eu/eurostat/databrowser/view/urb_ctran/default/table?lang=en", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average time of journey to work (minutes)", "year": "2016", "note": "26.5 min rounded half-up to 27, which is the same rounding used on all eight German rows and not a conservative thumb. BASIS YEAR 2016. CITY geography, narrower than our metro unit, so biased LOW and flattering. Survey self-report. METRO-LEVEL GERMAN SERIES: ESTABLISHED HARD NEGATIVE, do not re-run. MiD 2023 Tabelle 6 carries RegioStaR7 breakdowns of population, trips, distance and total time but NO duration by purpose; the national banner table W9 Wegedauer is all-purpose and sits on an \"Ohne regelmaessige berufliche Wege\" base; the small-area products regionalise headline indicators only; and Eurostat publishes no functional-urban-area rows for TT1019V. SrV 2023 was rejected because it publishes an ALL-PURPOSE trip duration, and swapping an all-purpose duration for a work-purpose one is a UNIT error, which is worse than the geography error we accepted. SrV is therefore used only as rank-order corroboration, and it agrees with Eurostat against our old stored spread: both put Frankfurt near the BOTTOM of the big five, where we had it second highest." }, "non_cash.sick_leave": { "source": "Entgeltfortzahlungsgesetz (EFZG) §3 (employer phase); SGB V §47 (Krankengeld; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "housing.utilities_monthly": { "source": "Numbeo 85 sqm 275.42 (tight 196 to 320 range) + the municipal-fee model 288, cross-checked against the DMB national floor, https://mieterbund.de/aktuelles/meldungen/aktueller-betriebskostenspiegel-fuer-deutschland/", "confidence": "medium", "note": "2BR reference. The genuinely cheap city of the eight: BW's 160% Bodenwertmodell Hebesatz and the lowest waste fee in the German set put it 6% below the national model. 285 sits between the two sources, resolved up from Numbeo.", "last_updated": "2026-08-08" }, "tax.retirement_deduction_cap": { "source": "EStG s.10 Abs 3 https://www.gesetze-im-internet.de/estg/__10.html ; SVBezGrV 2026 s.4 Abs 1 https://www.gesetze-im-internet.de/svbezgrv_2026/BJNR1160A0025.html", "last_updated": "2026-08-11", "confidence": "high", "note": "Ruerup/Basisrente Hoechstbetrag, 2026: EUR 30,825.60 single (knappschaftliche RV BBG 124,800 x 24.7%). A DEDUCTION ceiling and a COMBINED one: employee AND employer statutory pension contributions consume the same pot. less_employer_pct 0.186 = 2 x 9.3% and employer_base_cap 101,400 is the allgemeine RV BBG, because computeGermanyTaxes already deducts the employee 9.3% via vorsorge; a naked max would hand a EUR 95k earner about EUR 8.8k of phantom deduction. Riester (EUR 2,100) and bAV Entgeltumwandlung (8% of the BBG, EUR 8,112) are deliberately NOT stacked. Married couples: the statute doubles the ceiling to 61,651.20 even on one income; our per-earner cap under-deducts there, which errs toward higher tax." } }, "leipzig": { "housing.median_rent_by_bedroom": { "source": "Stadt Leipzig, Amt fuer Statistik und Wahlen, Analysen zur Stadtgesellschaft 04/2026, \"Eine Analyse der Mietpreisentwicklung in Leipzig\" (published 2026-07-01; median Angebotsmiete; Grundmiete = netto kalt; VALUE AG Marktdatenbank Stand 15.03.2026; city area) https://static.leipzig.de/fileadmin/mediendatenbank/leipzig-de/Stadt/02.1_Dez1_Allgemeine_Verwaltung/12_Statistik_und_Wahlen/Analysen_zur_Stadtgesellschaft/AzSG_13_Eine_Analyse_der_Mietpreisentwicklung_in_Leipzig.pdf ; index https://statistik.leipzig.de/statpubl/index.aspx?cat=1&rub=4 ; per-room gradient VALUE Marktdatenbank 2026 https://mietspiegeltabelle.de/mietspiegel-leipzig/ ; dwelling stock Leipzig-Informationssystem https://statistik.leipzig.de/statcity/table.aspx?cat=6&rub=1", "last_updated": "2026-08-07", "confidence": "high", "note": "CORRECTED 2026-08-02 from 480/600/900/1200, replacing a \"Homeday + lohntastik + Numbeo\" citation. The city own analysis, published five weeks before this edit, gives a 2025 median Angebotsmiete of 10.00 EUR/sqm with a mean of 10.49, up about 25% from 8.01 in 2022, and confirms the basis is Grundmiete, that is netto kalt, with Nebenkosten explicitly excluded. Per-room rates 10.88/10.29/10.16/11.44 x 1.0587 = 11.52/10.89/10.76/12.11. Studio 480 unchanged, 1BR 600 to 650, 2BR 900 to 910, 3BR 1200 to 1270. Leipzig STANDS in substance: three of four values move by 8% or less and the studio does not move at all, so the prior claim that Leipzig was clean is confirmed on level even though its citation was not. Leipzig is also the city that proves what the cross-city gradient source actually serves: mietspiegeltabelle 10.45 equals the city published VALUE MEAN of 10.49 to within 0.04 against a median of 10.00, so that basis is a mean and runs about 5% above median across all seven cities, which is the conservative direction. Stock contrast is the widest in the German set and the reason the Mietspiegel had to be refused: asking 10.00 against a Mietspiegel 2025-2027 value of 6.56 is +52.4%, and against the Buergerumfrage stock rent of 7.32 is +36.6%; the city notes the two series were still equivalent in 2014. Leipzig has no current official Mietspiegel on its own publication index, which ends at 2016, so the stock contrast uses the Buergerumfrage. Average dwelling across the whole stock is 70.4 sqm, but that includes owner-occupied and vacant units and is not a per-Zimmer size. Per-room rates come from the VALUE Marktdatenbank 2026 cut published at mietspiegeltabelle.de, explicitly nettokalt and explicitly market/asking rather than the qualified municipal Mietspiegel. That is the same database family the Berlin correction leant on. Leipzig proves what it serves: its 10.45 equals VALUE own published MEAN of 10.49 against a median of 10.00, so this basis runs about 5% above median, which is the conservative direction. Method is the shipped Berlin recipe re-run per city, so all eight German cities sit on one footing: rent = size x city per-room EUR/sqm x 1.0587. The 1.0587 is Berlin own uplift over this same source (Berlin shipped a 16.6 base against 15.68 here), and re-running it reproduces Berlin 760/970/1360/1750 to the euro. Sizes are held constant across cities at 42/60/85/105 sqm on the German Zimmer convention (our 1BR = 2-Zimmer-Wohnung), because the product prices an equivalent standard of living rather than how much space local households ration themselves into at local prices. Duesseldorf publishes real offered-flat sizes of 37/58/83/111, so the fixed 42 makes the studio the most conservative cell and the fixed 105 makes the 3BR mildly generous on size; the two roughly cancel. Basis is Angebotsmiete (new lease), netto kalt, never the Mietspiegel stock rent. Rents exclude utilities, insurance and internet, which are separate fields. Values are a CEILING per the conservative value-bias rule, not central estimates. Working: docs/superpowers/research/2026-08-02-german-city-rents.md SIZE LADDER CORRECTED 2026-08-07 (small-fixes wave 1, coordinated eight-city pass incl. Berlin): the common dwelling-size ladder moves from the inherited 42/60/85/105 sqm to 37/58/83/111 sqm, Duesseldorf's PUBLISHED offered-flat sizes (Wohnungsmarktbericht 2025 Abb. 54, MFH column, n=10,370 listings; the only published per-Zimmer size table in the German set, and the conservative top of the seven-city evidence range at 3-Zi 83 and 4-Zi 111). Root cause of the old ladder: German official statistics count a closed kitchen over 6 sqm as a room, so census room distributions ran one room high against listing Zimmer counts; the 42 sqm studio was priced at the 1-Zimmer premium rate measured on 33-37 sqm stock (double-counting the small-flat premium) and the 4-Zimmer was undersized at 105 vs the real ~111. Values 480/650/910/1270 -> 430/630/890/1340. Per-room rates and the 1.0587 uplift are UNCHANGED; sizes remain COMMON across all eight German cities because the product prices an equivalent standard of living. Working: docs/superpowers/research/2026-08-07-small-fixes.md" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "non_cash.parental_leave_paid_pct": { "source": "SGB V para 24i (Mutterschaftsgeld) https://www.gesetze-im-internet.de/sgb_5/__24i.html ; MuSchG para 20 (Zuschuss zum Mutterschaftsgeld, employer tops up to full net) https://www.gesetze-im-internet.de/muschg_2018/__20.html", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the percentage-paid remainder of the 67 weeks: the 14.1 Mutterschutz weeks (6 pre-birth + 8 post-birth), where para 24i SGB V pays EUR 13/day and para 20 MuSchG obliges the employer to top up to full net pay with no ceiling. Was 72, a blend that averaged the capped Elterngeld slice into this field; since 2026-08-06 the 52.87 Elterngeld weeks live in non_cash.parental_leave_capped (65% up to EUR 1,800/month, nothing above EUR 175,000 household income) and parentalLeaveEffectiveRate() blends the slices at the user's salary, so every reader-facing surface already prices the cap and the cliff." }, "non_cash.parental_leave_capped": { "source": "BEEG para 2 (Elterngeld: 65% of prior income, minimum EUR 300 and maximum EUR 1,800 per month) https://www.gesetze-im-internet.de/beeg/__2.html ; BEEG para 1 Abs. 8 (no Elterngeld above EUR 175,000 combined zu versteuerndes Einkommen) https://www.gesetze-im-internet.de/beeg/__1.html ; MuSchG para 20 (the 14 Mutterschutz weeks stay OUTSIDE this block: the employer net top-up has no ceiling) https://www.gesetze-im-internet.de/muschg_2018/__20.html", "last_updated": "2026-08-06", "confidence": "high", "note": "52.87 of the 67 weeks: the Elterngeld pot (60.87 wk) minus the 8 overlap weeks counted at the Mutterschutz rate. The EUR 1,800/month cap (unindexed since Elterngeld's 2007 introduction, re-confirmed 2026-08-06) binds above ~EUR 33,000 gross, i.e. at every ICP salary: at EUR 90,000 this slice pays ~24% of gross and the blended Berlin rate is ~40%; at EUR 150,000 the slice pays ~14% and the blend is ~32%, versus the 72 this field carried before the 2026-08-06 caps rebuild. BEEG para 1 Abs. 8 zeroes the slice for households above EUR 175,000: the engine compares HOUSEHOLD GROSS against the ceiling as a conservative proxy for zu versteuerndes Einkommen (gross exceeds taxable income, so the cliff bites earlier, the direction less favorable to Germany). Same conservative direction on the rate basis: the statute pays 65% of Elterngeld-Netto while the engine applies 65% to gross, and above ~EUR 33,000 gross the cap binds under both readings, so the distinction is immaterial at any ICP salary. Floor EUR 300/month (BEEG para 2). Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Germany section) and 2026-08-06-caps-engine-build.md." }, "non_cash.parental_leave_weeks": { "source": "MuSchG para 3 (Schutzfristen, 6 weeks before and 8 after) https://www.gesetze-im-internet.de/muschg_2018/__3.html ; BEEG para 4 Abs. 3 (12 Monatsbetraege Basiselterngeld plus 2 Partnermonate, gemeinsam) and para 4 Abs. 4 Satz 3 https://www.gesetze-im-internet.de/beeg/__4.html", "last_updated": "2026-08-01", "confidence": "high", "note": "67 wk = 6 additive pre-birth Mutterschutz weeks + the 14-month Elterngeld pot (60.87 wk). Was 61, the Elterngeld pot alone with Mutterschutz omitted. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. OVERLAP RULE APPLIED: the 8 post-birth Mutterschutz weeks are NOT stacked on top, because BEEG para 4 Abs. 4 Satz 3 deems Lebensmonate carrying Mutterschaftsgeld to be consumed Basiselterngeld months (gelten als Monate, fuer die dieser Elternteil Basiselterngeld bezieht). They are counted once, at the higher rate. Germany therefore reaches 67 and not 75. Elterngeld is a SHARED household pot, so it is not doubled. GERMANY HAS NO STATUTORY PAID PATERNITY LEAVE: the Familienstartzeit bill transposing EU Directive 2019/1158 art. 4 was still not law on 2026-08-01, so zero paternity weeks. Excluded: ElterngeldPlus (an election that halves the monthly amount to double the months) and the Partnerschaftsbonus (conditional on both parents working reduced hours)." }, "tax.income_tax_tariff_32a": { "source": "§ 32a Abs. 1 EStG, Veranlagungszeitraum 2026 https://www.gesetze-im-internet.de/estg/__32a.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Statutory continuous tariff, identical nationwide (see berlin for the verbatim zone text and the coefficient cross-check). Replaces the 5-bracket approximation that flattened the 17,799-69,878 progression zone to 33.5%." }, "tax.werbungskosten_pauschbetrag": { "source": "§ 9a Satz 1 Nr. 1 Buchst. a EStG https://www.gesetze-im-internet.de/estg/__9a.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Arbeitnehmer-Pauschbetrag EUR 1,230 per employed person. National, identical to berlin." }, "tax.sonderausgaben_pauschbetrag": { "source": "§ 10c EStG https://www.gesetze-im-internet.de/estg/__10c.html", "last_updated": "2026-07-31", "confidence": "high", "note": "EUR 36, doubled on Zusammenveranlagung. National, identical to berlin." }, "tax.vorsorge_deductible_rate_rv": { "source": "§ 10 Abs. 1 Nr. 2 und Abs. 3 EStG https://www.gesetze-im-internet.de/estg/__10.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Employee RV share 9.3%, 100% deductible since 2023, against the RV Beitragsbemessungsgrenze. National, identical to berlin." }, "tax.vorsorge_deductible_rate_kvpv": { "source": "§ 10 Abs. 1 Nr. 3 und Abs. 4 Satz 4 EStG https://www.gesetze-im-internet.de/estg/__10.html", "last_updated": "2026-07-31", "confidence": "high", "note": "Basis KV 8.75% x 0.96 (Krankengeld haircut) + PV 1.8% = 10.2%, against the KV/PV Beitragsbemessungsgrenze; ALV not deductible. National, identical to berlin." }, "tax.property_tax_rate_pct": { "source": "Leipzig Grundsteuer B Hebesatz 450% (Saxony Bundesmodell) × 0.31‰ ≈ 0.14%", "last_updated": "2026-05-26", "confidence": "high" }, "tax.sales_tax_rate": { "source": "German VAT 19%", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Homeday + lohntastik + Numbeo Leipzig 2025-26", "last_updated": "2026-05-26", "confidence": "high" }, "healthcare._summary": { "source": "GKV statutory insurance (premium in payroll); Eigenanteil ~€300/yr", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Stadt Leipzig Kita-Elternbeiträge (June 2025 rates, income-scaled)", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Leipzig 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Deutschlandticket €58/mo (nationwide; replaces LVB local €110 pass)", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Leipzig", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "LEJ->US transatlantic (via hub) economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Kindergeld €259/mo = €3,108/yr (national, inherited)", "last_updated": "2026-05-26", "confidence": "high", "note": "German Kindergeld (EUR 259/mo per child = EUR 3,108/yr, which is the stored value): universal, NOT means-tested, and it does not taper at professional incomes. Corrected 2026-08-02 from a stale \"EUR 255/mo\", which is the 2025 rate; the Steuerfortentwicklungsgesetz stepped Kindergeld to EUR 255 on 2025-01-01 and EUR 259 on 2026-01-01. The stored value was already correct and was NOT changed." }, "non_cash._summary": { "source": "German labour law (national, inherited); Saxony public holidays 11 (incl. Buß- und Bettag, Land-specific)", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Leipzig: EUR 65,035, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "EUR 65,035 / 0.867 = USD 75,012; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3000, R5-rounded down to 0.3. REPLACES 0.5 (2026-08-11), and the interim 0.35 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Thin: p90 only 14.8% above the median against Berlin's 51%. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.5 was 0.8333 of Berlin's 0.6; that ratio x Berlin's re-derived 0.42 = 0.3500, R5-rounded down to 0.35. Berlin's own figure: EUR 91,048 / 0.867 = USD 105,015, / USD 250,000 = 0.4201. This is a BASIS move, not a re-measurement: Leipzig's rank inside Germany is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.50, eastern Germany lower-wage market, growing startup scene." }, "tax.solidaritaetszuschlag_threshold_married": { "source": "Soli Freigrenze 2026 joint assessment EUR 40,700 (2 x 20,350 single; tk.de Solidaritaetszuschlag); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.social_insurance_rate_rvalv": { "source": "RV 9.3% + ALV 1.3% employee share = 10.6%, runs to the higher Beitragsbemessungsgrenze (lohn-info.de Sozialversicherungsbeitraege 2026); verified tax audit 2026-07-14", "confidence": "high", "note": "Carved out of the combined 21.15% employee share; the KV/PV remainder (10.55%) stops at the lower ceiling.", "last_updated": "2026-07-17" }, "tax.social_insurance_cap_rvalv": { "source": "Beitragsbemessungsgrenze RV/ALV 2026 = EUR 101,400 (bundesregierung.de/breg-de/aktuelles/beitragsgemessungsgrenzen-2386514); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "tax.pv_childless_surcharge_rate": { "source": "PV Kinderlosenzuschlag § 55 Abs. 3 SGB XI: +0.6pp employee-borne for childless members 23+ (dak.de Beitrag zur Pflegeversicherung Zuschlag Kinderlose; vdek.com Aenderungen Gesundheitswesen 2026); found by the 2026-08-06 corrections wave", "confidence": "high", "note": "Applied when the household has no children; the under-23 exemption is not modeled (ICP is older). Deductible with the rest of the Basis-Pflegeversicherung.", "last_updated": "2026-08-06" }, "transit.monthly_pass": { "source": "https://www.bundesregierung.de/breg-de/aktuelles/deutschlandticket-2134074", "last_updated": "2026-08-05", "confidence": "high", "note": "Deutschlandticket EUR 63/mo since 2026-01-01 (VMK decision 2025-09-18). Was 58 through Dec 2025." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Destatis Gesundheitsausgabenrechnung 2024 (private households + NPISH EUR 65bn) + Destatis EVS 2023 Konsumausgaben nach Haushaltstyp + GKV Zuzahlungen aggregate EUR 5.18bn (BMG via Aerzteblatt) + SGB V paras 61-62", "last_updated": "2026-08-08", "confidence": "medium", "note": "Level essentially CONFIRMED at EUR 311 per adult (778/capita gross, ~40% in scope after netting long-term care 46%, dental 9%, vision ~3pp). What Germany needed was a curve, not a level. The curve is steep in adults and nearly flat in children because under-18s are exempt from every Zuzahlung except transport AND each child subtracts EUR 9,756 from the income base before the 2%-of-gross Belastungsgrenze, so adding a child LOWERS the household's statutory ceiling in absolute euros. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Destatis Gesundheitsausgabenrechnung 2024 (private households + NPISH EUR 65bn) + Destatis EVS 2023 Konsumausgaben nach Haushaltstyp + GKV Zuzahlungen aggregate EUR 5.18bn (BMG via Aerzteblatt) + SGB V paras 61-62", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, urb_ctran indicator TT1019V, city DE008C. https://ec.europa.eu/eurostat/databrowser/view/urb_ctran/default/table?lang=en", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average time of journey to work (minutes)", "year": "2016", "note": "25.4 min, BASIS YEAR 2016. CITY geography, narrower than our metro unit, so biased LOW and flattering. Survey self-report, standard half-up rounding. METRO-LEVEL GERMAN SERIES: ESTABLISHED HARD NEGATIVE, do not re-run. MiD 2023 Tabelle 6 carries RegioStaR7 breakdowns of population, trips, distance and total time but NO duration by purpose; the national banner table W9 Wegedauer is all-purpose and sits on an \"Ohne regelmaessige berufliche Wege\" base; the small-area products regionalise headline indicators only; and Eurostat publishes no functional-urban-area rows for TT1019V. SrV 2023 was rejected because it publishes an ALL-PURPOSE trip duration, and swapping an all-purpose duration for a work-purpose one is a UNIT error, which is worse than the geography error we accepted. SrV is therefore used only as rank-order corroboration, and it agrees with Eurostat against our old stored spread: both put Frankfurt near the BOTTOM of the big five, where we had it second highest." }, "non_cash.sick_leave": { "source": "Entgeltfortzahlungsgesetz (EFZG) §3 (employer phase); SGB V §47 (Krankengeld; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "EStG s.10 Abs 3 https://www.gesetze-im-internet.de/estg/__10.html ; SVBezGrV 2026 s.4 Abs 1 https://www.gesetze-im-internet.de/svbezgrv_2026/BJNR1160A0025.html", "last_updated": "2026-08-11", "confidence": "high", "note": "Ruerup/Basisrente Hoechstbetrag, 2026: EUR 30,825.60 single (knappschaftliche RV BBG 124,800 x 24.7%). A DEDUCTION ceiling and a COMBINED one: employee AND employer statutory pension contributions consume the same pot. less_employer_pct 0.186 = 2 x 9.3% and employer_base_cap 101,400 is the allgemeine RV BBG, because computeGermanyTaxes already deducts the employee 9.3% via vorsorge; a naked max would hand a EUR 95k earner about EUR 8.8k of phantom deduction. Riester (EUR 2,100) and bAV Entgeltumwandlung (8% of the BBG, EUR 8,112) are deliberately NOT stacked. Married couples: the statute doubles the ceiling to 61,651.20 even on one income; our per-earner cap under-deducts there, which errs toward higher tax." } }, "marseille": { "non_cash.parental_leave_paid_pct": { "source": "service-public.gouv.fr F2265 (indemnites journalieres de maternite) https://www.service-public.gouv.fr/particuliers/vosdroits/F2265 ; F39685 (conge supplementaire de naissance) https://www.service-public.gouv.fr/particuliers/vosdroits/F39685", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 37 weeks into non_cash.parental_leave_capped (the 83 blend up to the IJ daily ceiling), so no percentage-paid weeks remain on this field. Was 83 with three caps ignored per the old field convention." }, "non_cash.parental_leave_weeks": { "source": "Conge supplementaire de naissance, service-public.gouv.fr https://www.service-public.gouv.fr/particuliers/vosdroits/F39685 (verifie le 03 juin 2026) ; conge de maternite https://www.service-public.gouv.fr/particuliers/vosdroits/F2265 ; conge de paternite et d'accueil de l'enfant, 25 jours calendaires (4 employer-paid + 21 CPAM) https://www.service-public.gouv.fr/particuliers/vosdroits/F3156 ; conge parental d'education https://www.service-public.gouv.fr/particuliers/vosdroits/F2280", "last_updated": "2026-08-01", "confidence": "high", "note": "37 wk = conge de maternite 16 (mother) + conge de paternite et d'accueil 3.6 (25 calendar days, father) + conge supplementaire de naissance 17.4 (2 months per parent x 2). Was 16, maternity alone. MAJOR CHANGE IN FORCE: the conge supplementaire de naissance (CSN) applies from 1 July 2026 to children born on or after 1 January 2026, giving each parent 1 or 2 months as a personal non-transferable right, so it SUMS across both parents. It must be taken AFTER maternity and paternity leave, so those weeks stack rather than offset. PreParE is EXCLUDED from the headline under the overlap rule: service-public states the CSN cannot be cumulated with PreParE for the same period, and the CSN pays roughly six times more, so the weeks are counted once at the higher rate. Excluded: the 3 jours ouvrables of the OLD employer-paid conge de naissance (0.5 wk of noise, and omitting it is the less favourable reading); the conge parental d'education, which service-public confirms is unpaid absent a collective agreement; multiple-birth uplifts." }, "childcare.free_tier_start_age": { "source": "service-public.gouv.fr F1898 (Direction de l'information legale et administrative): \"L'instruction est obligatoire pour tous les enfants, francais et etrangers, a partir de 3 ans et jusqu'a l'age de 16 ans revolus\" (loi du 26 juillet 2019 pour une ecole de la confiance) https://www.service-public.fr/particuliers/vosdroits/F1898 ; education.gouv.fr https://www.education.gouv.fr/l-inscription-l-ecole-maternelle-8651", "last_updated": "2026-08-01", "confidence": "high", "note": "Ecole maternelle runs 3-6, is free, and has been COMPULSORY since 2019, so participation at 3-5 is at or above 96% (Eurostat). Creche (0-3) is paid. Before this, ages 1-5 all billed one rate, so a 1-year-old in creche was billed the maternelle wraparound rate." }, "tax.federal_brackets": { "source": "France 2026 IRPEF: national, peer-identical with Paris", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "UNPI / IndiceVille Marseille taxe foncière ~0.55% effective of market value", "last_updated": "2026-05-26", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "French VAT (TVA) 20%", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Investropa + Numbeo Marseille 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Assurance Maladie + employer mutuelle (inherited from Paris baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "CAF barème EAJE 2025; école maternelle free age 3+; CMG subsidy at ICP (high-income) tier", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Marseille 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "RTM monthly pass ~EUR 40", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo + Expatistan Marseille", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "MRS->US transatlantic economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Allocations familiales (national, inherited; 2-kid high-income)", "last_updated": "2026-05-26", "confidence": "high", "note": "French allocations familiales: means-tested/income-modulated (from the 2nd child), retaining a reduced high-income tier; modeled value is the high-income modulated rate." }, "non_cash._summary": { "source": "French labour law (national, inherited); 11 public holidays", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Marseille. Its basis is converted with the rest of France and its rank inside France is preserved. Prior source: Derived, not observed: no all-levels total-comp measurement exists for Marseille. Basis re-derived with its country anchor Paris; Marseille's stored ratio to Paris is preserved. Prior source: Glassdoor Marseille SWE 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.43 x 0.4255 = 0.1830, R5-rounded down to 0.18. 0.4255 is the MEDIAN of (re-derived / previous) across the 4 France cities that DO carry an all-levels measurement (Lyon 0.48 to 0.2; Nantes 0.47 to 0.2; Paris 0.55 to 0.29; Toulouse 0.5 to 0.2), so Marseille moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Marseille's rank inside France is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.43 was 0.7818 of Paris's 0.55; that ratio x Paris's re-derived 0.29 = 0.2267, R5-rounded down to 0.22. Paris's own figure: EUR 64,372 / 0.867 = USD 74,247, / USD 250,000 = 0.2970. This is a BASIS move, not a re-measurement: Marseille's rank inside France is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.43, below Paris/Lyon." }, "childcare.subsidy_per_kid_annual": { "source": "paris.fr quotient familial / cantine 2025; CAF PSU crèche", "last_updated": "2026-07-17", "note": "French childcare (crèche PSU / périscolaire + cantine) is means-tested by income/quotient familial; top-tranche households pay the full/top rate, so the modeled prices are net top-tranche costs and the means-tested subsidy is 0. Was 2,200-2,400 (a mid-income CMG/PSU level that a top earner does not receive)." }, "childcare.preschool_monthly": { "source": "Ville de Marseille, restauration scolaire: EUR 4.70/meal for CAF quotient familial > 1,850 (the top tranche), plus a EUR 0.50/month admin fee https://www.marseille.fr/index.php/actualites/inscription-cantine", "last_updated": "2026-08-01", "confidence": "low", "note": "VALUE UNCHANGED AT 120, and the least trustworthy of the five French cities. Only the cantine leg is sourced: Marseille is a 4-DAY week = 144 days, so 4.70 x 144 + 6 = EUR 683/yr = EUR 57/month. Marseille sets periscolaire prices by conseil-municipal deliberation (refs 21/049/VDV and 22/030/VDV, both 2021/22 and likely stale) rather than publishing a grid, and superminot.marseille.fr refused connection, so the morning, evening and Wednesday legs could NOT be priced. 120 is retained because it sits above the sourced floor and lowering it on missing evidence would flatter France. Marseille needs its own sourcing pass. Free tuition, so the modelled cost is the working-parent wraparound at the TOP quotient-familial tranche: cantine plus periscolaire plus, on a 4-day-week commune, Wednesday care. Stored value is the ANNUAL total divided by 12 (computeChildcareCost multiplies by 12); French communes bill over the ~10-month school year. School days: the annee scolaire is 36 weeks (Code de l'education art. L521-1) = 144 days on a 4-day week, 162 on a 4.5-day week. Sourcing and the full per-slot grids: docs/superpowers/research/2026-08-01-france-cantine-fees.md." }, "transit.statutory_employer_share_pct": { "source": "https://entreprendre.service-public.gouv.fr/vosdroits/F37900 (Code du travail L3261-2)", "last_updated": "2026-08-05", "confidence": "high", "note": "Every French employer must reimburse 50% of a public-transit subscription (prise en charge obligatoire). The engine nets this share on the pass branch and the row label discloses it. Statutory-share convention, owner decision 2026-08-04; Belgium checked and does NOT qualify (sectoral agreements, not one statute)." }, "non_cash.parental_leave_capped": { "source": "service-public.gouv.fr F2265: IJ maternite/paternite plafonnee a 104,02 EUR par jour (2026), salaire journalier de base capped at the PASS of 4 005 EUR/month https://www.service-public.gouv.fr/particuliers/vosdroits/F2265 ; F39685: CSN paid 70% month 1 / 60% month 2 of a base capped at the same PASS https://www.service-public.gouv.fr/particuliers/vosdroits/F39685", "last_updated": "2026-08-06", "confidence": "high", "note": "Conservative single blend of TWO statutory cap mechanisms: the 19.6 IJ weeks (maternity 16 + paternity 3.6) cap the PAYOUT at EUR 104.02/day, and the 17.4 CSN weeks cap the BASE at EUR 4,005/month before the 70%/60% rate (a weekly ceiling around EUR 555 to 647). We store one block, 83% up to EUR 104.02/day (EUR 728.14/wk), the best-known ceiling, which reads the CSN weeks slightly LOW, the direction less favorable to France. At EUR 90,000 the effective rate is ~42%, matching the 40-45% the 2026-08-01 note already estimated. If the engine ever supports two capped blocks, France is a candidate to split. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (France)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "DREES Fiche 21, reste a charge des menages measured after BOTH Assurance Maladie and the complementaire (EUR 292 per inhabitant, 2024), netted of dental and optical + ameli participation forfaitaire and franchises medicales, EUR 100/yr cap each", "last_updated": "2026-08-08", "confidence": "medium", "note": "DREES measures exactly our basis, which matters in France more than anywhere: the mutuelle has been employer-mandated since ANI 2016 and is a PREMIUM, so the honest French out-of-pocket figure is the residual after it. Under-18s are statutorily exempt from both participations and are covered by 100% Sante, so a child adds ~40% of an adult. All French cities are normalised onto one curve: the previous Lyon-300 against Paris/Toulouse/Nantes/Marseille-350 split had no stated basis. A Paris depassements uplift is queued, not shipped. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "DREES Fiche 21, reste a charge des menages measured after BOTH Assurance Maladie and the complementaire (EUR 292 per inhabitant, 2024), netted of dental and optical + ameli participation forfaitaire and franchises medicales, EUR 100/yr cap each", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, indicator TT1019V, greater city. https://ec.europa.eu/eurostat/api/dissemination/statistics/1.0/data/urb_ctran?indic_ur=TT1019V&format=JSON&lang=EN", "last_updated": "2026-08-08", "confidence": "low", "statistic": "Average time of journey to work", "year": "2012", "note": "Urban Audit 25.2. The previous 47 matches no published figure located anywhere. Same French-wave caveat as Lyon." }, "non_cash.sick_leave": { "source": "Indemnités journalières de sécurité sociale (IJSS): Code de la sécurité sociale; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": "https://www.echangesassurances.org/actualites/indemnites-journalieres-maladie-plafond-juillet-2026", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "service-public.gouv.fr, Plafond epargne retraite (PER) https://www.service-public.gouv.fr/particuliers/vosdroits/F14709?lang=fr ; impots.gouv.fr https://www.impots.gouv.fr/particulier/questions/que-signifie-la-somme-imprimee-dans-la-rubrique-plafond-epargne-retraite-de", "last_updated": "2026-08-11", "confidence": "high", "note": "PER individuel deduction from revenu net global, CGI art. 163 quatervicies, for versements paid in 2026 (plafond computed on 2025 income, PASS 2025 = EUR 47,100): the greater of 10% of prior-year revenus d'activite capped at 8 x PASS (EUR 37,680) or 10% of PASS (EUR 4,710). BASIS TRAP HANDLED: the statutory 10% runs on activity income NET of cotisations sociales and after the 10% abattement, which is 0.10 x (1 - 0.113 - 0.068) x 0.90 = 7.37% of GROSS, so 0.0737 is used. It shipped as 0.074 on 2026-08-11 and was corrected the same day: rounding UP was the one value in the 69-country set that went the generous way, and the house rule is to err less favorable to the country; applying 0.10 to gross would overstate the ceiling by about a quarter. earnings_cap omitted as redundant (max binds first, around EUR 511k of gross). The EUR 4,710 floor has no engine key and omitting it is the conservative direction. 5-year carry-forward is unmodelable." } }, "toulouse": { "non_cash.parental_leave_paid_pct": { "source": "service-public.gouv.fr F2265 (indemnites journalieres de maternite) https://www.service-public.gouv.fr/particuliers/vosdroits/F2265 ; F39685 (conge supplementaire de naissance) https://www.service-public.gouv.fr/particuliers/vosdroits/F39685", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 37 weeks into non_cash.parental_leave_capped (the 83 blend up to the IJ daily ceiling), so no percentage-paid weeks remain on this field. Was 83 with three caps ignored per the old field convention." }, "non_cash.parental_leave_weeks": { "source": "Conge supplementaire de naissance, service-public.gouv.fr https://www.service-public.gouv.fr/particuliers/vosdroits/F39685 (verifie le 03 juin 2026) ; conge de maternite https://www.service-public.gouv.fr/particuliers/vosdroits/F2265 ; conge de paternite et d'accueil de l'enfant, 25 jours calendaires (4 employer-paid + 21 CPAM) https://www.service-public.gouv.fr/particuliers/vosdroits/F3156 ; conge parental d'education https://www.service-public.gouv.fr/particuliers/vosdroits/F2280", "last_updated": "2026-08-01", "confidence": "high", "note": "37 wk = conge de maternite 16 (mother) + conge de paternite et d'accueil 3.6 (25 calendar days, father) + conge supplementaire de naissance 17.4 (2 months per parent x 2). Was 16, maternity alone. MAJOR CHANGE IN FORCE: the conge supplementaire de naissance (CSN) applies from 1 July 2026 to children born on or after 1 January 2026, giving each parent 1 or 2 months as a personal non-transferable right, so it SUMS across both parents. It must be taken AFTER maternity and paternity leave, so those weeks stack rather than offset. PreParE is EXCLUDED from the headline under the overlap rule: service-public states the CSN cannot be cumulated with PreParE for the same period, and the CSN pays roughly six times more, so the weeks are counted once at the higher rate. Excluded: the 3 jours ouvrables of the OLD employer-paid conge de naissance (0.5 wk of noise, and omitting it is the less favourable reading); the conge parental d'education, which service-public confirms is unpaid absent a collective agreement; multiple-birth uplifts." }, "childcare.after_school_monthly": { "source": "Toulouse Metropole, Recueil des tarifs janvier 2026 (tranche 11, revenus > EUR 7,200.01/month, the top tranche): restauration scolaire EUR 6.60/day for one child; accueil du matin EUR 0.75/day; accueil du soir EUR 1.89/day; mercredi apres-midi EUR 2.50 https://metropole.toulouse.fr/sites/toulouse-fr/files/2026-02/recueil_tarifs_janvier2026.pdf", "last_updated": "2026-08-01", "confidence": "high", "note": "French communes price periscolaire and the cantine off ONE tariff schedule that covers maternelle and elementaire alike, so the school-age rate is set equal to the maternelle rate rather than left at a separate legacy value. Aligned 2026-08-01 with the same sourced annual total. Was 80." }, "childcare.free_tier_start_age": { "source": "service-public.gouv.fr F1898 (Direction de l'information legale et administrative): \"L'instruction est obligatoire pour tous les enfants, francais et etrangers, a partir de 3 ans et jusqu'a l'age de 16 ans revolus\" (loi du 26 juillet 2019 pour une ecole de la confiance) https://www.service-public.fr/particuliers/vosdroits/F1898 ; education.gouv.fr https://www.education.gouv.fr/l-inscription-l-ecole-maternelle-8651", "last_updated": "2026-08-01", "confidence": "high", "note": "Ecole maternelle runs 3-6, is free, and has been COMPULSORY since 2019, so participation at 3-5 is at or above 96% (Eurostat). Creche (0-3) is paid. Before this, ages 1-5 all billed one rate, so a 1-year-old in creche was billed the maternelle wraparound rate." }, "tax.federal_brackets": { "source": "France 2026 IRPEF: national, peer-identical with Paris", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Détaxe-foncière Toulouse: combined rate 57.38% => ~1.15% effective of market value (among France's highest)", "last_updated": "2026-05-26", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "French VAT (TVA) 20%", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Trackstone + Numbeo Toulouse 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Assurance Maladie + employer mutuelle (inherited from Paris baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Toulouse Métropole crèche tarifs 2025; école maternelle free age 3+; CMG at ICP tier", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Toulouse 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Tisséo monthly pass EUR 55", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Toulouse", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "TLS->US transatlantic (via hub) economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Allocations familiales (national, inherited; 2-kid high-income)", "last_updated": "2026-05-26", "confidence": "high", "note": "French allocations familiales: means-tested/income-modulated (from the 2nd child), retaining a reduced high-income tier; modeled value is the high-income modulated rate." }, "non_cash._summary": { "source": "French labour law (national, inherited); 11 public holidays", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Toulouse: EUR 43,847, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 43,847 / 0.867 = USD 50,573; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2023, R5-rounded down to 0.2. REPLACES 0.5 (2026-08-11), and the interim 0.26 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.5 was 0.9091 of Paris's 0.55; that ratio x Paris's re-derived 0.29 = 0.2636, R5-rounded down to 0.26. Paris's own figure: EUR 64,372 / 0.867 = USD 74,247, / USD 250,000 = 0.2970. This is a BASIS move, not a re-measurement: Toulouse's rank inside France is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.50, aerospace hub above the French average." }, "childcare.subsidy_per_kid_annual": { "source": "paris.fr quotient familial / cantine 2025; CAF PSU crèche", "last_updated": "2026-07-17", "note": "French childcare (crèche PSU / périscolaire + cantine) is means-tested by income/quotient familial; top-tranche households pay the full/top rate, so the modeled prices are net top-tranche costs and the means-tested subsidy is 0. Was 2,200-2,400 (a mid-income CMG/PSU level that a top earner does not receive)." }, "childcare.preschool_monthly": { "source": "Toulouse Metropole, Recueil des tarifs janvier 2026 (tranche 11, revenus > EUR 7,200.01/month, the top tranche): restauration scolaire EUR 6.60/day for one child; accueil du matin EUR 0.75/day; accueil du soir EUR 1.89/day; mercredi apres-midi EUR 2.50 https://metropole.toulouse.fr/sites/toulouse-fr/files/2026-02/recueil_tarifs_janvier2026.pdf", "note": "Toulouse bands on plain household MONTHLY INCOME with no division by parts, unlike the Paris/Lyon quotient familial, so a EUR 140k household lands in tranche 11 (on a 3-part QF reading it would have fallen to tranche 8 / EUR 4.00). 4.5-day week = 162 days. 6.60 x 162 = 1,069 + 0.75 x 162 = 122 + 1.89 x 162 = 306 + 2.50 x 36 Wednesdays = 90, total EUR 1,587/yr; 1587/12 = 132.25. Was 80/mo, unsourced. Free tuition, so the modelled cost is the working-parent wraparound at the TOP quotient-familial tranche: cantine plus periscolaire plus, on a 4-day-week commune, Wednesday care. Stored value is the ANNUAL total divided by 12 (computeChildcareCost multiplies by 12); French communes bill over the ~10-month school year. School days: the annee scolaire is 36 weeks (Code de l'education art. L521-1) = 144 days on a 4-day week, 162 on a 4.5-day week. Sourcing and the full per-slot grids: docs/superpowers/research/2026-08-01-france-cantine-fees.md.", "last_updated": "2026-08-01", "confidence": "high" }, "transit.statutory_employer_share_pct": { "source": "https://entreprendre.service-public.gouv.fr/vosdroits/F37900 (Code du travail L3261-2)", "last_updated": "2026-08-05", "confidence": "high", "note": "Every French employer must reimburse 50% of a public-transit subscription (prise en charge obligatoire). The engine nets this share on the pass branch and the row label discloses it. Statutory-share convention, owner decision 2026-08-04; Belgium checked and does NOT qualify (sectoral agreements, not one statute)." }, "non_cash.parental_leave_capped": { "source": "service-public.gouv.fr F2265: IJ maternite/paternite plafonnee a 104,02 EUR par jour (2026), salaire journalier de base capped at the PASS of 4 005 EUR/month https://www.service-public.gouv.fr/particuliers/vosdroits/F2265 ; F39685: CSN paid 70% month 1 / 60% month 2 of a base capped at the same PASS https://www.service-public.gouv.fr/particuliers/vosdroits/F39685", "last_updated": "2026-08-06", "confidence": "high", "note": "Conservative single blend of TWO statutory cap mechanisms: the 19.6 IJ weeks (maternity 16 + paternity 3.6) cap the PAYOUT at EUR 104.02/day, and the 17.4 CSN weeks cap the BASE at EUR 4,005/month before the 70%/60% rate (a weekly ceiling around EUR 555 to 647). We store one block, 83% up to EUR 104.02/day (EUR 728.14/wk), the best-known ceiling, which reads the CSN weeks slightly LOW, the direction less favorable to France. At EUR 90,000 the effective rate is ~42%, matching the 40-45% the 2026-08-01 note already estimated. If the engine ever supports two capped blocks, France is a candidate to split. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (France)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "DREES Fiche 21, reste a charge des menages measured after BOTH Assurance Maladie and the complementaire (EUR 292 per inhabitant, 2024), netted of dental and optical + ameli participation forfaitaire and franchises medicales, EUR 100/yr cap each", "last_updated": "2026-08-08", "confidence": "medium", "note": "DREES measures exactly our basis, which matters in France more than anywhere: the mutuelle has been employer-mandated since ANI 2016 and is a PREMIUM, so the honest French out-of-pocket figure is the residual after it. Under-18s are statutorily exempt from both participations and are covered by 100% Sante, so a child adds ~40% of an adult. All French cities are normalised onto one curve: the previous Lyon-300 against Paris/Toulouse/Nantes/Marseille-350 split had no stated basis. A Paris depassements uplift is queued, not shipped. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "DREES Fiche 21, reste a charge des menages measured after BOTH Assurance Maladie and the complementaire (EUR 292 per inhabitant, 2024), netted of dental and optical + ameli participation forfaitaire and franchises medicales, EUR 100/yr cap each", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, indicator TT1019V, greater city. https://ec.europa.eu/eurostat/api/dissemination/statistics/1.0/data/urb_ctran?indic_ur=TT1019V&format=JSON&lang=EN", "last_updated": "2026-08-08", "confidence": "low", "statistic": "Average time of journey to work", "year": "2012-2013", "note": "Urban Audit 24.1 and 24.31. Same French-wave caveat as Lyon." }, "non_cash.sick_leave": { "source": "Indemnités journalières de sécurité sociale (IJSS): Code de la sécurité sociale; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": "https://www.echangesassurances.org/actualites/indemnites-journalieres-maladie-plafond-juillet-2026", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "service-public.gouv.fr, Plafond epargne retraite (PER) https://www.service-public.gouv.fr/particuliers/vosdroits/F14709?lang=fr ; impots.gouv.fr https://www.impots.gouv.fr/particulier/questions/que-signifie-la-somme-imprimee-dans-la-rubrique-plafond-epargne-retraite-de", "last_updated": "2026-08-11", "confidence": "high", "note": "PER individuel deduction from revenu net global, CGI art. 163 quatervicies, for versements paid in 2026 (plafond computed on 2025 income, PASS 2025 = EUR 47,100): the greater of 10% of prior-year revenus d'activite capped at 8 x PASS (EUR 37,680) or 10% of PASS (EUR 4,710). BASIS TRAP HANDLED: the statutory 10% runs on activity income NET of cotisations sociales and after the 10% abattement, which is 0.10 x (1 - 0.113 - 0.068) x 0.90 = 7.37% of GROSS, so 0.0737 is used. It shipped as 0.074 on 2026-08-11 and was corrected the same day: rounding UP was the one value in the 69-country set that went the generous way, and the house rule is to err less favorable to the country; applying 0.10 to gross would overstate the ceiling by about a quarter. earnings_cap omitted as redundant (max binds first, around EUR 511k of gross). The EUR 4,710 floor has no engine key and omitting it is the conservative direction. 5-year carry-forward is unmodelable." } }, "nantes": { "non_cash.parental_leave_paid_pct": { "source": "service-public.gouv.fr F2265 (indemnites journalieres de maternite) https://www.service-public.gouv.fr/particuliers/vosdroits/F2265 ; F39685 (conge supplementaire de naissance) https://www.service-public.gouv.fr/particuliers/vosdroits/F39685", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 37 weeks into non_cash.parental_leave_capped (the 83 blend up to the IJ daily ceiling), so no percentage-paid weeks remain on this field. Was 83 with three caps ignored per the old field convention." }, "non_cash.parental_leave_weeks": { "source": "Conge supplementaire de naissance, service-public.gouv.fr https://www.service-public.gouv.fr/particuliers/vosdroits/F39685 (verifie le 03 juin 2026) ; conge de maternite https://www.service-public.gouv.fr/particuliers/vosdroits/F2265 ; conge de paternite et d'accueil de l'enfant, 25 jours calendaires (4 employer-paid + 21 CPAM) https://www.service-public.gouv.fr/particuliers/vosdroits/F3156 ; conge parental d'education https://www.service-public.gouv.fr/particuliers/vosdroits/F2280", "last_updated": "2026-08-01", "confidence": "high", "note": "37 wk = conge de maternite 16 (mother) + conge de paternite et d'accueil 3.6 (25 calendar days, father) + conge supplementaire de naissance 17.4 (2 months per parent x 2). Was 16, maternity alone. MAJOR CHANGE IN FORCE: the conge supplementaire de naissance (CSN) applies from 1 July 2026 to children born on or after 1 January 2026, giving each parent 1 or 2 months as a personal non-transferable right, so it SUMS across both parents. It must be taken AFTER maternity and paternity leave, so those weeks stack rather than offset. PreParE is EXCLUDED from the headline under the overlap rule: service-public states the CSN cannot be cumulated with PreParE for the same period, and the CSN pays roughly six times more, so the weeks are counted once at the higher rate. Excluded: the 3 jours ouvrables of the OLD employer-paid conge de naissance (0.5 wk of noise, and omitting it is the less favourable reading); the conge parental d'education, which service-public confirms is unpaid absent a collective agreement; multiple-birth uplifts." }, "childcare.after_school_monthly": { "source": "Nantes Metropole, accueil periscolaire tarifs (top of the published range): matin EUR 3.77/session, midi incl. meal EUR 7.35/session, soir EUR 4.04/session https://metropole.nantes.fr/mes-services-mon-quotidien/s-informer-sur-l-accueil-periscolaire-et-ses-tarifs", "last_updated": "2026-08-01", "confidence": "medium", "note": "French communes price periscolaire and the cantine off ONE tariff schedule that covers maternelle and elementaire alike, so the school-age rate is set equal to the maternelle rate rather than left at a separate legacy value. Aligned 2026-08-01 with the same sourced annual total. Was 80." }, "childcare.free_tier_start_age": { "source": "service-public.gouv.fr F1898 (Direction de l'information legale et administrative): \"L'instruction est obligatoire pour tous les enfants, francais et etrangers, a partir de 3 ans et jusqu'a l'age de 16 ans revolus\" (loi du 26 juillet 2019 pour une ecole de la confiance) https://www.service-public.fr/particuliers/vosdroits/F1898 ; education.gouv.fr https://www.education.gouv.fr/l-inscription-l-ecole-maternelle-8651", "last_updated": "2026-08-01", "confidence": "high", "note": "Ecole maternelle runs 3-6, is free, and has been COMPULSORY since 2019, so participation at 3-5 is at or above 96% (Eurostat). Creche (0-3) is paid. Before this, ages 1-5 all billed one rate, so a 1-year-old in creche was billed the maternelle wraparound rate." }, "tax.federal_brackets": { "source": "France 2026 IRPEF: national, peer-identical with Paris", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Nantes Métropole taxe foncière combined 52.75% => ~0.48% effective of market value", "last_updated": "2026-05-26", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "French VAT (TVA) 20%", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Numbeo + DVF notarial Nantes 2025-26", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Assurance Maladie + employer mutuelle (inherited from Paris baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Nantes Métropole crèche/périscolaire tarifs; école maternelle free age 3+; CMG at ICP tier", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Nantes 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "TAN monthly pass EUR 55", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Nantes", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "NTE->US transatlantic (via hub) economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Allocations familiales (national, inherited; 2-kid high-income)", "last_updated": "2026-05-26", "confidence": "high", "note": "French allocations familiales: means-tested/income-modulated (from the 2nd child), retaining a reduced high-income tier; modeled value is the high-income modulated rate." }, "non_cash._summary": { "source": "French labour law (national, inherited); 11 public holidays", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Nantes: EUR 44,286, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "EUR 44,286 / 0.867 = USD 51,080; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2043, R5-rounded down to 0.2. REPLACES 0.47 (2026-08-11), and the interim 0.24 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Percentile collision: its all-levels p75 equals its senior median, which two independent percentiles do not do by chance. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.47 was 0.8545 of Paris's 0.55; that ratio x Paris's re-derived 0.29 = 0.2478, R5-rounded down to 0.24. Paris's own figure: EUR 64,372 / 0.867 = USD 74,247, / USD 250,000 = 0.2970. This is a BASIS move, not a re-measurement: Nantes's rank inside France is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.47, growing western-France tech hub." }, "childcare.subsidy_per_kid_annual": { "source": "paris.fr quotient familial / cantine 2025; CAF PSU crèche", "last_updated": "2026-07-17", "note": "French childcare (crèche PSU / périscolaire + cantine) is means-tested by income/quotient familial; top-tranche households pay the full/top rate, so the modeled prices are net top-tranche costs and the means-tested subsidy is 0. Was 2,200-2,400 (a mid-income CMG/PSU level that a top earner does not receive)." }, "childcare.preschool_monthly": { "source": "Nantes Metropole, accueil periscolaire tarifs (top of the published range): matin EUR 3.77/session, midi incl. meal EUR 7.35/session, soir EUR 4.04/session https://metropole.nantes.fr/mes-services-mon-quotidien/s-informer-sur-l-accueil-periscolaire-et-ses-tarifs", "confidence": "medium", "note": "Nantes is the most expensive of the five because it charges a real per-session price for all three daily slots, where Lyon gives the morning free and bills the evening as a flat annual forfait. 4.5-day week = 162 days. (3.77 + 7.35 + 4.04) x 162 = EUR 2,456/yr; 2456/12 = 204.67. Was 80/mo, unsourced. CONFIDENCE MEDIUM: Nantes publishes a min/max range but not the tranche grid, so the QF threshold for the top of the range is not pinned; non-residents pay the maximum regardless. Free tuition, so the modelled cost is the working-parent wraparound at the TOP quotient-familial tranche: cantine plus periscolaire plus, on a 4-day-week commune, Wednesday care. Stored value is the ANNUAL total divided by 12 (computeChildcareCost multiplies by 12); French communes bill over the ~10-month school year. School days: the annee scolaire is 36 weeks (Code de l'education art. L521-1) = 144 days on a 4-day week, 162 on a 4.5-day week. Sourcing and the full per-slot grids: docs/superpowers/research/2026-08-01-france-cantine-fees.md.", "last_updated": "2026-08-01" }, "transit.statutory_employer_share_pct": { "source": "https://entreprendre.service-public.gouv.fr/vosdroits/F37900 (Code du travail L3261-2)", "last_updated": "2026-08-05", "confidence": "high", "note": "Every French employer must reimburse 50% of a public-transit subscription (prise en charge obligatoire). The engine nets this share on the pass branch and the row label discloses it. Statutory-share convention, owner decision 2026-08-04; Belgium checked and does NOT qualify (sectoral agreements, not one statute)." }, "non_cash.parental_leave_capped": { "source": "service-public.gouv.fr F2265: IJ maternite/paternite plafonnee a 104,02 EUR par jour (2026), salaire journalier de base capped at the PASS of 4 005 EUR/month https://www.service-public.gouv.fr/particuliers/vosdroits/F2265 ; F39685: CSN paid 70% month 1 / 60% month 2 of a base capped at the same PASS https://www.service-public.gouv.fr/particuliers/vosdroits/F39685", "last_updated": "2026-08-06", "confidence": "high", "note": "Conservative single blend of TWO statutory cap mechanisms: the 19.6 IJ weeks (maternity 16 + paternity 3.6) cap the PAYOUT at EUR 104.02/day, and the 17.4 CSN weeks cap the BASE at EUR 4,005/month before the 70%/60% rate (a weekly ceiling around EUR 555 to 647). We store one block, 83% up to EUR 104.02/day (EUR 728.14/wk), the best-known ceiling, which reads the CSN weeks slightly LOW, the direction less favorable to France. At EUR 90,000 the effective rate is ~42%, matching the 40-45% the 2026-08-01 note already estimated. If the engine ever supports two capped blocks, France is a candidate to split. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (France)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "DREES Fiche 21, reste a charge des menages measured after BOTH Assurance Maladie and the complementaire (EUR 292 per inhabitant, 2024), netted of dental and optical + ameli participation forfaitaire and franchises medicales, EUR 100/yr cap each", "last_updated": "2026-08-08", "confidence": "medium", "note": "DREES measures exactly our basis, which matters in France more than anywhere: the mutuelle has been employer-mandated since ANI 2016 and is a PREMIUM, so the honest French out-of-pocket figure is the residual after it. Under-18s are statutorily exempt from both participations and are covered by 100% Sante, so a child adds ~40% of an adult. All French cities are normalised onto one curve: the previous Lyon-300 against Paris/Toulouse/Nantes/Marseille-350 split had no stated basis. A Paris depassements uplift is queued, not shipped. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "DREES Fiche 21, reste a charge des menages measured after BOTH Assurance Maladie and the complementaire (EUR 292 per inhabitant, 2024), netted of dental and optical + ameli participation forfaitaire and franchises medicales, EUR 100/yr cap each", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, indicator TT1019V, greater city. https://ec.europa.eu/eurostat/api/dissemination/statistics/1.0/data/urb_ctran?indic_ur=TT1019V&format=JSON&lang=EN", "last_updated": "2026-08-08", "confidence": "low", "statistic": "Average time of journey to work", "year": "2012-2013", "note": "Urban Audit 23.88 and 23.93. Same French-wave caveat as Lyon." }, "non_cash.sick_leave": { "source": "Indemnités journalières de sécurité sociale (IJSS): Code de la sécurité sociale; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": "https://www.echangesassurances.org/actualites/indemnites-journalieres-maladie-plafond-juillet-2026", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "service-public.gouv.fr, Plafond epargne retraite (PER) https://www.service-public.gouv.fr/particuliers/vosdroits/F14709?lang=fr ; impots.gouv.fr https://www.impots.gouv.fr/particulier/questions/que-signifie-la-somme-imprimee-dans-la-rubrique-plafond-epargne-retraite-de", "last_updated": "2026-08-11", "confidence": "high", "note": "PER individuel deduction from revenu net global, CGI art. 163 quatervicies, for versements paid in 2026 (plafond computed on 2025 income, PASS 2025 = EUR 47,100): the greater of 10% of prior-year revenus d'activite capped at 8 x PASS (EUR 37,680) or 10% of PASS (EUR 4,710). BASIS TRAP HANDLED: the statutory 10% runs on activity income NET of cotisations sociales and after the 10% abattement, which is 0.10 x (1 - 0.113 - 0.068) x 0.90 = 7.37% of GROSS, so 0.0737 is used. It shipped as 0.074 on 2026-08-11 and was corrected the same day: rounding UP was the one value in the 69-country set that went the generous way, and the house rule is to err less favorable to the country; applying 0.10 to gross would overstate the ceiling by about a quarter. earnings_cap omitted as redundant (max binds first, around EUR 511k of gross). The EUR 4,710 floor has no engine key and omitting it is the conservative direction. 5-year carry-forward is unmodelable." } }, "turin": { "childcare.free_tier_start_age": { "source": "ISTAT / Dipartimento per le politiche della famiglia / Universita Ca Foscari, I servizi educativi per l'infanzia in Italia, a.e. 2023/2024, p.39: \"La scuola d'infanzia ... se a titolarita pubblica e sostanzialmente gratuita (fatta eccezione per la quota della mensa)\" https://www.istat.it/wp-content/uploads/2025/05/report_infanzia_2023_2024.pdf", "last_updated": "2026-08-01", "confidence": "high", "note": "Scuola dell'infanzia runs 3-6 and is free in state schools; asilo nido (0-3) is paid. Participation from age 3 to compulsory school is 94.7% (2023, Eurostat via the same report). Before this, ages 1-5 all billed one rate, so the bucket straddled paid nido and free infanzia." }, "non_cash.parental_leave_paid_pct": { "source": "D.Lgs 151/2001 art. 22 c.1 (indennita pari all 80 per cento for congedo di maternita, testo in vigore dal 28-5-2003), art. 29 c.1 (100 per cento for congedo di paternita), art. 34 c.1 (3 months at 80 per cento in alternativa tra i genitori, the remainder at 30 per cento) https://www.normattiva.it/uri-res/N2Ls?urn:nir:stato:decreto.legislativo:2001-03-26;151 ; Legge 207/2024 art. 1 commi 217-218 (the third month at 80%) ; D.Lgs 151/2001 art. 23 (no massimale for dependent employees) https://www.normattiva.it/uri-res/N2Ls?urn:nir:stato:decreto.legislativo:2001-03-26;151", "last_updated": "2026-08-06", "confidence": "high", "note": "60% blended (exact 59.9): 34.7 wk at 80% (maternity plus the 3 uplifted parental months) + 2 wk at 100% (paternity) + 26 wk at 30% (the parental tail). This LOWERS the stored 80, because that figure was the maternity rate applied to maternity weeks only. Congedo di maternita was NOT raised to 100%: art. 22 is untouched since 2003, and the 100% figure in circulation comes from collective agreements, which are excluded. Cap research 2026-08-06: genuinely UNCAPPED for standard dependent employees at every tier. D.Lgs 151/2001 art. 23 defines the indennita off the actual retribuzione media globale giornaliera with no massimale language, and the Legge 335/1995 massimale contributivo (EUR 122,295 for 2026, INPS circolare 47/2026) applies ONLY to Gestione Separata workers, not lavoratori dipendenti, so Italy keeps a plain percentage representation, a positive citable finding rather than a research gap. Separate income-eligibility cliff (EUR 19,885.13/yr, 2.5x minimum pension) on part of the 30% tail is flagged for its own pass; months 10-11 are already excluded from the stored 62. See docs/superpowers/research/2026-08-06-parental-caps-europe.md (Italy)." }, "non_cash.parental_leave_weeks": { "source": "D.Lgs 151/2001 Testo Unico consolidated on Normattiva, testo in vigore dal 1-1-2026 https://www.normattiva.it/uri-res/N2Ls?urn:nir:stato:decreto.legislativo:2001-03-26;151; art. 16 (congedo di maternita, 5 mesi), art. 27-bis (congedo di paternita obbligatorio, 10 giorni lavorativi), art. 32 and 34 (congedo parentale)", "last_updated": "2026-08-01", "confidence": "high", "note": "62 wk = congedo di maternita 21.7 + congedo di paternita obbligatorio 2 + congedo parentale 9 paid months (3 non-transferable per parent plus 3 in alternativa tra loro). Was 22, which was congedo di maternita alone and dropped the mandatory 10 paternity days and all 9 paid parental months. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Excluded: months 10-11 of congedo parentale, which art. 34 c.3 means-tests at 2.5x the minimum pension and our ICP fails; CCNL top-ups." }, "tax.inbound_regime": { "source": "Regime impatriati, art. 5 D.Lgs. 209/2023 (transfers of residence from 2024). Agenzia delle Entrate: 'i redditi ... prodotti in Italia entro il limite annuo di 600.000 euro concorrono alla formazione del reddito complessivo limitatamente al 50%', so the EUR 600,000 is a cap on the exempted base, not a cliff. Five tax periods; 3 prior tax periods of non-Italian residency (6 or 7 if the same employer or group); elevata qualificazione o specializzazione; 4-year residency commitment with clawback plus interest. https://www.agenziaentrate.gov.it/portale/lavoratori-impatriati-209-2023/infogen-lavoratori-impatriati-209-2023 ; normattiva urn:nir:stato:decreto.legislativo:2023-12-27;209~art5", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: values verified UNCHANGED for 2026 (0.50 / 600,000 / 5 years). The only 2026 amendment is art. 22 L. 132/2025 (in force 2025-10-10), which merely widened the qualification test to include applied AI research. Notes harmonised across all five Italian cities: Bologna, Naples and Turin still carried the PRE-2024 condition '2+ yrs prior non-IT residency', which is wrong (the current statute is 3 tax periods, 6 or 7 within the same employer group). Caught by the new tax.inbound_regime entry in data/_national_fields.json. The 60% exclusion for a worker with a minor child is DELIBERATELY left out: it requires the child to reside in Italy for the whole benefit period, which we cannot test, and the conservative tiebreaker takes the lower exclusion. WATCH 2027: normattiva shows art. 5 abrogated by D.Lgs. 117/2026 (Testo unico imposte sui redditi, GU n. 152 of 2026-07-03) with effect 2027-01-01; the regime moves to art. 225. The art. 225 text could not be retrieved, so treat the consolidation as presumptively substance-neutral. TODO (before 2027-01): re-verify the art. 225 text." }, "tax.filing_basis": { "source": "PwC WWTS - Italy (Tax administration): Modello Redditi PF cannot be filed jointly; each spouse taxed on own income (the joint Modello 730 does not aggregate income).", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Italy 2026 IRPEF: national, peer-identical with Milan", "last_updated": "2026-05-26", "confidence": "high" }, "tax.addizionale_regionale_rate": { "source": "Regione Piemonte addizionale IRPEF 2025: effective ~2.75% at EUR 45k", "last_updated": "2026-05-26", "confidence": "high" }, "tax.addizionale_comunale_rate": { "source": "Città di Torino addizionale comunale 1.1% (EUR 28-50k band)", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "IMU exempt for primary residence; nominal 0.1 placeholder (matches Milan)", "last_updated": "2026-05-26", "confidence": "medium" }, "housing._summary": { "source": "Investropa + Numbeo Turin 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "SSN universal + supplemental (inherited from Milan baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Comune di Torino nidi tariffe 2025/26; bonus asilo nido at ICP (high-ISEE) tier EUR 1,500", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Turin 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "GTT Formula U monthly pass EUR 38", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Turin", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "TRN->US transatlantic (via MXP hub) economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Assegno Unico (national, inherited; high-ISEE)", "last_updated": "2026-05-26", "confidence": "high", "note": "Italian Assegno Unico Universale: means-tested by ISEE with a guaranteed minimum (~EUR 57/mo/kid) paid at any income; modeled value is the high-ISEE floor (survives)." }, "non_cash._summary": { "source": "Italian labour law (national, inherited); 12 public holidays (incl. patron saint)", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Turin: EUR 38,448, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 38,448 / 0.867 = USD 44,346; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1774, R5-rounded down to 0.17. REPLACES 0.25 (2026-08-11), and the interim 0.14 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.25 was 0.8333 of Milan's 0.3; that ratio x Milan's re-derived 0.17 = 0.1417, R5-rounded down to 0.14. Milan's own figure: EUR 37,601 / 0.867 = USD 43,369, / USD 250,000 = 0.1735. This is a BASIS move, not a re-measurement: Turin's rank inside Italy is unchanged and still rests on its prior evidence, so confidence is medium. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.25, industrial north (Stellantis, Leonardo); below Milan. Was 0.38. Re-measured on the peer basis (levels.fyi senior SWE median total comp / USD 220,000 Seattle senior reference): Turin senior median EUR 48,698 = USD 56,169 / 220,000 = 0.2553, rounded DOWN. RAL, gross annual, mensilita included. Ties with Rome (0.2517 derived); the 1.4% measured gap is inside the noise of a self-reported sample, so the old 0.38-over-0.37 ordering is not defended. Research: docs/superpowers/research/2026-08-10-milan-comp-index.md" }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-21", "note": "Zeroed for 3-6: the Italian \"bonus asilo nido\" applies to 0-3 nido and is means-tested (only a small floor survives for a high earner); it does not apply to free 3-6 infanzia." }, "childcare.preschool_monthly": { "source": "Citta di Torino, tariffe refezione scolastica scuola dell'infanzia a.s. 2026/2027, top fascia ISEE monthly tariff EUR 157,00 (full rate; EUR 118,00 is the sibling rate) https://www.comune.torino.it/media/2489", "note": "Flat MONTHLY tariff, not per meal, applied from 1 September each school year (Sept-June, ~10 months) = EUR 1,570/yr; 1570/12 = 130.83. Statale assumed; a comunale place adds an enrolment fee of EUR 141/yr (EUR 11.75/mo) at the top band. Was 350 (paritaria/bilingual tier). Applies the owner decision of 2026-07-18 that Spain and Italy follow the statutory free convention, now that the age boundary makes it applicable to the free ages only. REPLACES the 2026-07-21 modelling, which priced Italy at the paritaria/bilingual private tier on the argument that January-only enrolment and Italian-only instruction push international families private. That argument holds equally for Catalan-medium Barcelona and for France, Belgium and Germany, where the repo already models the free public path, so Italy was the only country priced private. Stored value is the ANNUAL out-of-pocket divided by 12, because computeChildcareCost multiplies by 12; the invoice figure is a school-year figure over ~10 months. Sourcing: docs/superpowers/research/2026-08-01-italy-mensa-fees.md.", "last_updated": "2026-08-01", "confidence": "high" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "National ticket framework on specialist visits and diagnostics plus the pharmaceutical co-payment, both regionally set; esenzione per reddito thresholds sit far below our ICP household", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 raised from 400 to 480 on the batch-D research. The ticket on specialist visits and diagnostics plus the pharmaceutical co-payment are both REGIONALLY set, and the esenzione per reddito income thresholds sit far below our ICP household, so neither the adults nor the children are exempt and the curve follows the no-child-discount shape. Italy has no annual out-of-pocket ceiling. Naples is the fifth Italian city, not Florence. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "National ticket framework on specialist visits and diagnostics plus the pharmaceutical co-payment, both regionally set; esenzione per reddito thresholds sit far below our ICP household", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, indicator TT1019V, greater city. https://ec.europa.eu/eurostat/api/dissemination/statistics/1.0/data/urb_ctran?indic_ur=TT1019V&format=JSON&lang=EN", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average time of journey to work", "year": "2011", "note": "Urban Audit 33.29. Same ISTAT work-or-study caveat." }, "non_cash.sick_leave": { "source": "INPS indennità di malattia (general private-sector \"industria\" regime); Codice; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Agenzia delle Entrate, Risposta n. 76/2024 https://www.agenziaentrate.gov.it/portale/documents/20143/5946162/Risposta+n.+76_2024.pdf/f0933165-e52e-a519-8ca9-85d3f5162f23 ; Risoluzione n. 25/2025 https://www.agenziaentrate.gov.it/portale/documents/20143/8874453/Risoluzione+n.+25_2025/58473993-686e-9f56-723a-034c8a1dd3f6 (D.Lgs. 252/2005 art. 8 c.4)", "last_updated": "2026-08-11", "confidence": "high", "note": "Previdenza complementare deduction from reddito complessivo, EUR 5,164.57/yr, fixed in statute since 2007 (it is 10 million lire) and not indexed. A COMBINED ceiling: employer contributions to the fondo pensione count against the same figure. TFR conferred to the fund does NOT count and is not deductible, having never been taxed as income. less_employer_pct is deliberately left unset, which is the GENEROUS reading, rather than inventing an employer rate; flagged for a later pass. The \"lavoratori di prima occupazione\" extra platform in years 6-25 is deliberately unmodeled." } }, "naples": { "childcare.free_tier_start_age": { "source": "ISTAT / Dipartimento per le politiche della famiglia / Universita Ca Foscari, I servizi educativi per l'infanzia in Italia, a.e. 2023/2024, p.39: \"La scuola d'infanzia ... se a titolarita pubblica e sostanzialmente gratuita (fatta eccezione per la quota della mensa)\" https://www.istat.it/wp-content/uploads/2025/05/report_infanzia_2023_2024.pdf", "last_updated": "2026-08-01", "confidence": "high", "note": "Scuola dell'infanzia runs 3-6 and is free in state schools; asilo nido (0-3) is paid. Participation from age 3 to compulsory school is 94.7% (2023, Eurostat via the same report). Before this, ages 1-5 all billed one rate, so the bucket straddled paid nido and free infanzia." }, "non_cash.parental_leave_paid_pct": { "source": "D.Lgs 151/2001 art. 22 c.1 (indennita pari all 80 per cento for congedo di maternita, testo in vigore dal 28-5-2003), art. 29 c.1 (100 per cento for congedo di paternita), art. 34 c.1 (3 months at 80 per cento in alternativa tra i genitori, the remainder at 30 per cento) https://www.normattiva.it/uri-res/N2Ls?urn:nir:stato:decreto.legislativo:2001-03-26;151 ; Legge 207/2024 art. 1 commi 217-218 (the third month at 80%) ; D.Lgs 151/2001 art. 23 (no massimale for dependent employees) https://www.normattiva.it/uri-res/N2Ls?urn:nir:stato:decreto.legislativo:2001-03-26;151", "last_updated": "2026-08-06", "confidence": "high", "note": "60% blended (exact 59.9): 34.7 wk at 80% (maternity plus the 3 uplifted parental months) + 2 wk at 100% (paternity) + 26 wk at 30% (the parental tail). This LOWERS the stored 80, because that figure was the maternity rate applied to maternity weeks only. Congedo di maternita was NOT raised to 100%: art. 22 is untouched since 2003, and the 100% figure in circulation comes from collective agreements, which are excluded. Cap research 2026-08-06: genuinely UNCAPPED for standard dependent employees at every tier. D.Lgs 151/2001 art. 23 defines the indennita off the actual retribuzione media globale giornaliera with no massimale language, and the Legge 335/1995 massimale contributivo (EUR 122,295 for 2026, INPS circolare 47/2026) applies ONLY to Gestione Separata workers, not lavoratori dipendenti, so Italy keeps a plain percentage representation, a positive citable finding rather than a research gap. Separate income-eligibility cliff (EUR 19,885.13/yr, 2.5x minimum pension) on part of the 30% tail is flagged for its own pass; months 10-11 are already excluded from the stored 62. See docs/superpowers/research/2026-08-06-parental-caps-europe.md (Italy)." }, "non_cash.parental_leave_weeks": { "source": "D.Lgs 151/2001 Testo Unico consolidated on Normattiva, testo in vigore dal 1-1-2026 https://www.normattiva.it/uri-res/N2Ls?urn:nir:stato:decreto.legislativo:2001-03-26;151; art. 16 (congedo di maternita, 5 mesi), art. 27-bis (congedo di paternita obbligatorio, 10 giorni lavorativi), art. 32 and 34 (congedo parentale)", "last_updated": "2026-08-01", "confidence": "high", "note": "62 wk = congedo di maternita 21.7 + congedo di paternita obbligatorio 2 + congedo parentale 9 paid months (3 non-transferable per parent plus 3 in alternativa tra loro). Was 22, which was congedo di maternita alone and dropped the mandatory 10 paternity days and all 9 paid parental months. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Excluded: months 10-11 of congedo parentale, which art. 34 c.3 means-tests at 2.5x the minimum pension and our ICP fails; CCNL top-ups." }, "tax.inbound_regime": { "source": "Regime impatriati, art. 5 D.Lgs. 209/2023 (transfers of residence from 2024). Agenzia delle Entrate: 'i redditi ... prodotti in Italia entro il limite annuo di 600.000 euro concorrono alla formazione del reddito complessivo limitatamente al 50%', so the EUR 600,000 is a cap on the exempted base, not a cliff. Five tax periods; 3 prior tax periods of non-Italian residency (6 or 7 if the same employer or group); elevata qualificazione o specializzazione; 4-year residency commitment with clawback plus interest. https://www.agenziaentrate.gov.it/portale/lavoratori-impatriati-209-2023/infogen-lavoratori-impatriati-209-2023 ; normattiva urn:nir:stato:decreto.legislativo:2023-12-27;209~art5", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: values verified UNCHANGED for 2026 (0.50 / 600,000 / 5 years). The only 2026 amendment is art. 22 L. 132/2025 (in force 2025-10-10), which merely widened the qualification test to include applied AI research. Notes harmonised across all five Italian cities: Bologna, Naples and Turin still carried the PRE-2024 condition '2+ yrs prior non-IT residency', which is wrong (the current statute is 3 tax periods, 6 or 7 within the same employer group). Caught by the new tax.inbound_regime entry in data/_national_fields.json. The 60% exclusion for a worker with a minor child is DELIBERATELY left out: it requires the child to reside in Italy for the whole benefit period, which we cannot test, and the conservative tiebreaker takes the lower exclusion. WATCH 2027: normattiva shows art. 5 abrogated by D.Lgs. 117/2026 (Testo unico imposte sui redditi, GU n. 152 of 2026-07-03) with effect 2027-01-01; the regime moves to art. 225. The art. 225 text could not be retrieved, so treat the consolidation as presumptively substance-neutral. TODO (before 2027-01): re-verify the art. 225 text." }, "tax.filing_basis": { "source": "PwC WWTS - Italy (Tax administration): Modello Redditi PF cannot be filed jointly; each spouse taxed on own income (the joint Modello 730 does not aggregate income).", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Italy 2026 IRPEF: national, peer-identical with Milan", "last_updated": "2026-05-26", "confidence": "high" }, "tax.addizionale_regionale_rate": { "source": "Regione Campania addizionale IRPEF 2025: effective ~2.64% at EUR 45k (among Italy's highest)", "last_updated": "2026-05-26", "confidence": "high" }, "tax.addizionale_comunale_rate": { "source": "Comune di Napoli addizionale comunale 1.0% (raised from 0.8% Jan 2025)", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "IMU exempt for primary residence; nominal 0.1 placeholder (matches Milan)", "last_updated": "2026-05-26", "confidence": "medium" }, "housing._summary": { "source": "Numbeo + Immobiliare.it Naples 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "SSN universal + supplemental (inherited from Milan baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Comune di Napoli nido tariffe; bonus asilo nido at ICP (high-ISEE) tier EUR 1,500", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Naples 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "ANM integrated monthly pass EUR 42", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Naples", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "NAP->US transatlantic (via hub) economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Assegno Unico (national, inherited; high-ISEE)", "last_updated": "2026-05-26", "confidence": "high", "note": "Italian Assegno Unico Universale: means-tested by ISEE with a guaranteed minimum (~EUR 57/mo/kid) paid at any income; modeled value is the high-ISEE floor (survives)." }, "non_cash._summary": { "source": "Italian labour law (national, inherited); 12 public holidays (incl. patron saint)", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Naples: EUR 34,755, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "EUR 34,755 / 0.867 = USD 40,087; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1603, R5-rounded down to 0.16. REPLACES 0.21 (2026-08-11), and the interim 0.11 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. The levels.fyi slug is napoli-ita. The 2026-08-10 Italian pass recorded a 404 for the English form and derived Naples from a ratio instead; this is the first measurement Naples has ever had. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.21 was 0.7000 of Milan's 0.3; that ratio x Milan's re-derived 0.17 = 0.1190, R5-rounded down to 0.11. Milan's own figure: EUR 37,601 / 0.867 = USD 43,369, / USD 250,000 = 0.1735. This is a BASIS move, not a re-measurement: Naples's rank inside Italy is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.21, southern Italy, lower salaries. Was 0.30. NOT measured: levels.fyi returns 404 for Naples at both all-levels and senior, so this is the one Italian row that could not be re-derived from a source. Its stored 0.30 was 0.714 of stored Milan, and that within-Italy ratio is the only information the row ever carried; applied to the corrected Milan 0.30 it gives 0.214. Sanity check: 0.21 implies EUR 40,056 senior total comp, which sits below measured Bologna (EUR 44,546) as it must. Confidence dropped to low because the value is inherited, not observed. Research: docs/superpowers/research/2026-08-10-milan-comp-index.md" }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-21", "note": "Zeroed for 3-6: the Italian \"bonus asilo nido\" applies to 0-3 nido and is means-tested (only a small floor survives for a high earner); it does not apply to free 3-6 infanzia." }, "childcare.preschool_monthly": { "source": "Comune di Napoli, tariffe refezione scolastica a.s. 2025/2026: EUR 4,70 per meal at the top band plus a one-off EUR 15/yr https://www.comune.napoli.it", "note": "Per meal ordered. EUR 4.70 x ~200 meal days + EUR 15 = EUR 955/yr; 955/12 = 79.58. 2025/26 edition; the 2026/27 page was not yet published. Was 250 (paritaria/bilingual tier). Applies the owner decision of 2026-07-18 that Spain and Italy follow the statutory free convention, now that the age boundary makes it applicable to the free ages only. REPLACES the 2026-07-21 modelling, which priced Italy at the paritaria/bilingual private tier on the argument that January-only enrolment and Italian-only instruction push international families private. That argument holds equally for Catalan-medium Barcelona and for France, Belgium and Germany, where the repo already models the free public path, so Italy was the only country priced private. Stored value is the ANNUAL out-of-pocket divided by 12, because computeChildcareCost multiplies by 12; the invoice figure is a school-year figure over ~10 months. Sourcing: docs/superpowers/research/2026-08-01-italy-mensa-fees.md.", "last_updated": "2026-08-01", "confidence": "high" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "National ticket framework on specialist visits and diagnostics plus the pharmaceutical co-payment, both regionally set; esenzione per reddito thresholds sit far below our ICP household", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 raised from 400 to 480 on the batch-D research. The ticket on specialist visits and diagnostics plus the pharmaceutical co-payment are both REGIONALLY set, and the esenzione per reddito income thresholds sit far below our ICP household, so neither the adults nor the children are exempt and the curve follows the no-child-discount shape. Italy has no annual out-of-pocket ceiling. Naples is the fifth Italian city, not Florence. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "National ticket framework on specialist visits and diagnostics plus the pharmaceutical co-payment, both regionally set; esenzione per reddito thresholds sit far below our ICP household", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, indicator TT1019V, greater city. https://ec.europa.eu/eurostat/api/dissemination/statistics/1.0/data/urb_ctran?indic_ur=TT1019V&format=JSON&lang=EN", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average time of journey to work", "year": "2011", "note": "Urban Audit 28.11. The previous 45 matches no published year located. Same ISTAT work-or-study caveat." }, "non_cash.sick_leave": { "source": "INPS indennità di malattia (general private-sector \"industria\" regime); Codice; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Agenzia delle Entrate, Risposta n. 76/2024 https://www.agenziaentrate.gov.it/portale/documents/20143/5946162/Risposta+n.+76_2024.pdf/f0933165-e52e-a519-8ca9-85d3f5162f23 ; Risoluzione n. 25/2025 https://www.agenziaentrate.gov.it/portale/documents/20143/8874453/Risoluzione+n.+25_2025/58473993-686e-9f56-723a-034c8a1dd3f6 (D.Lgs. 252/2005 art. 8 c.4)", "last_updated": "2026-08-11", "confidence": "high", "note": "Previdenza complementare deduction from reddito complessivo, EUR 5,164.57/yr, fixed in statute since 2007 (it is 10 million lire) and not indexed. A COMBINED ceiling: employer contributions to the fondo pensione count against the same figure. TFR conferred to the fund does NOT count and is not deductible, having never been taxed as income. less_employer_pct is deliberately left unset, which is the GENEROUS reading, rather than inventing an employer rate; flagged for a later pass. The \"lavoratori di prima occupazione\" extra platform in years 6-25 is deliberately unmodeled." } }, "bologna": { "childcare.free_tier_start_age": { "source": "ISTAT / Dipartimento per le politiche della famiglia / Universita Ca Foscari, I servizi educativi per l'infanzia in Italia, a.e. 2023/2024, p.39: \"La scuola d'infanzia ... se a titolarita pubblica e sostanzialmente gratuita (fatta eccezione per la quota della mensa)\" https://www.istat.it/wp-content/uploads/2025/05/report_infanzia_2023_2024.pdf", "last_updated": "2026-08-01", "confidence": "high", "note": "Scuola dell'infanzia runs 3-6 and is free in state schools; asilo nido (0-3) is paid. Participation from age 3 to compulsory school is 94.7% (2023, Eurostat via the same report). Before this, ages 1-5 all billed one rate, so the bucket straddled paid nido and free infanzia." }, "non_cash.parental_leave_paid_pct": { "source": "D.Lgs 151/2001 art. 22 c.1 (indennita pari all 80 per cento for congedo di maternita, testo in vigore dal 28-5-2003), art. 29 c.1 (100 per cento for congedo di paternita), art. 34 c.1 (3 months at 80 per cento in alternativa tra i genitori, the remainder at 30 per cento) https://www.normattiva.it/uri-res/N2Ls?urn:nir:stato:decreto.legislativo:2001-03-26;151 ; Legge 207/2024 art. 1 commi 217-218 (the third month at 80%) ; D.Lgs 151/2001 art. 23 (no massimale for dependent employees) https://www.normattiva.it/uri-res/N2Ls?urn:nir:stato:decreto.legislativo:2001-03-26;151", "last_updated": "2026-08-06", "confidence": "high", "note": "60% blended (exact 59.9): 34.7 wk at 80% (maternity plus the 3 uplifted parental months) + 2 wk at 100% (paternity) + 26 wk at 30% (the parental tail). This LOWERS the stored 80, because that figure was the maternity rate applied to maternity weeks only. Congedo di maternita was NOT raised to 100%: art. 22 is untouched since 2003, and the 100% figure in circulation comes from collective agreements, which are excluded. Cap research 2026-08-06: genuinely UNCAPPED for standard dependent employees at every tier. D.Lgs 151/2001 art. 23 defines the indennita off the actual retribuzione media globale giornaliera with no massimale language, and the Legge 335/1995 massimale contributivo (EUR 122,295 for 2026, INPS circolare 47/2026) applies ONLY to Gestione Separata workers, not lavoratori dipendenti, so Italy keeps a plain percentage representation, a positive citable finding rather than a research gap. Separate income-eligibility cliff (EUR 19,885.13/yr, 2.5x minimum pension) on part of the 30% tail is flagged for its own pass; months 10-11 are already excluded from the stored 62. See docs/superpowers/research/2026-08-06-parental-caps-europe.md (Italy)." }, "non_cash.parental_leave_weeks": { "source": "D.Lgs 151/2001 Testo Unico consolidated on Normattiva, testo in vigore dal 1-1-2026 https://www.normattiva.it/uri-res/N2Ls?urn:nir:stato:decreto.legislativo:2001-03-26;151; art. 16 (congedo di maternita, 5 mesi), art. 27-bis (congedo di paternita obbligatorio, 10 giorni lavorativi), art. 32 and 34 (congedo parentale)", "last_updated": "2026-08-01", "confidence": "high", "note": "62 wk = congedo di maternita 21.7 + congedo di paternita obbligatorio 2 + congedo parentale 9 paid months (3 non-transferable per parent plus 3 in alternativa tra loro). Was 22, which was congedo di maternita alone and dropped the mandatory 10 paternity days and all 9 paid parental months. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Excluded: months 10-11 of congedo parentale, which art. 34 c.3 means-tests at 2.5x the minimum pension and our ICP fails; CCNL top-ups." }, "tax.inbound_regime": { "source": "Regime impatriati, art. 5 D.Lgs. 209/2023 (transfers of residence from 2024). Agenzia delle Entrate: 'i redditi ... prodotti in Italia entro il limite annuo di 600.000 euro concorrono alla formazione del reddito complessivo limitatamente al 50%', so the EUR 600,000 is a cap on the exempted base, not a cliff. Five tax periods; 3 prior tax periods of non-Italian residency (6 or 7 if the same employer or group); elevata qualificazione o specializzazione; 4-year residency commitment with clawback plus interest. https://www.agenziaentrate.gov.it/portale/lavoratori-impatriati-209-2023/infogen-lavoratori-impatriati-209-2023 ; normattiva urn:nir:stato:decreto.legislativo:2023-12-27;209~art5", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: values verified UNCHANGED for 2026 (0.50 / 600,000 / 5 years). The only 2026 amendment is art. 22 L. 132/2025 (in force 2025-10-10), which merely widened the qualification test to include applied AI research. Notes harmonised across all five Italian cities: Bologna, Naples and Turin still carried the PRE-2024 condition '2+ yrs prior non-IT residency', which is wrong (the current statute is 3 tax periods, 6 or 7 within the same employer group). Caught by the new tax.inbound_regime entry in data/_national_fields.json. The 60% exclusion for a worker with a minor child is DELIBERATELY left out: it requires the child to reside in Italy for the whole benefit period, which we cannot test, and the conservative tiebreaker takes the lower exclusion. WATCH 2027: normattiva shows art. 5 abrogated by D.Lgs. 117/2026 (Testo unico imposte sui redditi, GU n. 152 of 2026-07-03) with effect 2027-01-01; the regime moves to art. 225. The art. 225 text could not be retrieved, so treat the consolidation as presumptively substance-neutral. TODO (before 2027-01): re-verify the art. 225 text." }, "tax.filing_basis": { "source": "PwC WWTS - Italy (Tax administration): Modello Redditi PF cannot be filed jointly; each spouse taxed on own income (the joint Modello 730 does not aggregate income).", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Italy 2026 IRPEF: national, peer-identical with Milan", "last_updated": "2026-05-26", "confidence": "high" }, "tax.addizionale_regionale_rate": { "source": "Regione Emilia-Romagna addizionale IRPEF 2025 (LR 1/2025): effective ~2.11% at EUR 45k", "last_updated": "2026-05-26", "confidence": "high" }, "tax.addizionale_comunale_rate": { "source": "Comune di Bologna addizionale comunale 0.8%", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "IMU exempt for primary residence; nominal 0.1 placeholder (matches Milan)", "last_updated": "2026-05-26", "confidence": "medium" }, "housing._summary": { "source": "Immobiliare.it + Numbeo Bologna 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "SSN universal + supplemental (inherited from Milan baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Comune di Bologna nidi tariffe 2025/26 (Italy's priciest municipal nido); bonus asilo nido at ICP tier EUR 1,500", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Bologna 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "TPER monthly pass EUR 39 (from Mar 2025)", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Bologna", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "BLQ->US transatlantic (via hub) economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Assegno Unico (national, inherited; high-ISEE)", "last_updated": "2026-05-26", "confidence": "high", "note": "Italian Assegno Unico Universale: means-tested by ISEE with a guaranteed minimum (~EUR 57/mo/kid) paid at any income; modeled value is the high-ISEE floor (survives)." }, "non_cash._summary": { "source": "Italian labour law (national, inherited); 12 public holidays (incl. patron saint)", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Bologna: EUR 35,287, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "EUR 35,287 / 0.867 = USD 40,700; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1628, R5-rounded down to 0.16. REPLACES 0.23 (2026-08-11), and the interim 0.13 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Percentile collision: its all-levels median equals its senior p25. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.23 was 0.7667 of Milan's 0.3; that ratio x Milan's re-derived 0.17 = 0.1303, R5-rounded down to 0.13. Milan's own figure: EUR 37,601 / 0.867 = USD 43,369, / USD 250,000 = 0.1735. This is a BASIS move, not a re-measurement: Bologna's rank inside Italy is unchanged and still rests on its prior evidence, so confidence is medium. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.23, wealthy north (university/biomedical/packaging-valley). Was 0.40. Re-measured on the peer basis (levels.fyi senior SWE median total comp / USD 220,000 Seattle senior reference): Bologna senior median EUR 44,546 = USD 51,378 / 220,000 = 0.2335. RAL, gross annual, mensilita included. A wealthy local economy is not the same as a high software-pay market: Bologna measures BELOW Turin and Rome and drops from 2nd to 4th inside Italy. Research: docs/superpowers/research/2026-08-10-milan-comp-index.md" }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit 2026-07-17", "last_updated": "2026-07-21", "note": "Zeroed for 3-6: the Italian \"bonus asilo nido\" applies to 0-3 nido and is means-tested (only a small floor survives for a high earner); it does not apply to free 3-6 infanzia." }, "childcare.preschool_monthly": { "source": "Comune di Bologna, Opuscolo informativo servizi integrativi scolastici a.s. 2025/2026, TARIFFE REFEZIONE p.8: EUR 6,80 per meal with no ISEE presented (EUR 6,40 at ISEE >= 35.000); underlying act Delibera di Giunta P.G.n. 833964/2023 del 18/12/2023 https://assets.culturabologna.it/d74f0e0c-b513-4f1b-a774-5c197a65038c-opuscolo-informativo-servizi-integrativi-scol-a-s-2025-2026-1.pdf", "note": "Strictly per meal ordered. EUR 6.80 (the no-ISEE row, the higher one) x ~200 meal days = EUR 1,360/yr; 1360/12 = 113.33. A comunale place instead charges an all-in tariffa di frequenza that the delibera caps at the meal rate per day, so 6.80/day bounds either route. Sibling discounts ignored. 2025/26 edition; the 2026/27 table was not retrievable. Was 350 (paritaria/bilingual tier). Applies the owner decision of 2026-07-18 that Spain and Italy follow the statutory free convention, now that the age boundary makes it applicable to the free ages only. REPLACES the 2026-07-21 modelling, which priced Italy at the paritaria/bilingual private tier on the argument that January-only enrolment and Italian-only instruction push international families private. That argument holds equally for Catalan-medium Barcelona and for France, Belgium and Germany, where the repo already models the free public path, so Italy was the only country priced private. Stored value is the ANNUAL out-of-pocket divided by 12, because computeChildcareCost multiplies by 12; the invoice figure is a school-year figure over ~10 months. Sourcing: docs/superpowers/research/2026-08-01-italy-mensa-fees.md.", "last_updated": "2026-08-01", "confidence": "high" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "National ticket framework on specialist visits and diagnostics plus the pharmaceutical co-payment, both regionally set; esenzione per reddito thresholds sit far below our ICP household", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 raised from 400 to 480 on the batch-D research. The ticket on specialist visits and diagnostics plus the pharmaceutical co-payment are both REGIONALLY set, and the esenzione per reddito income thresholds sit far below our ICP household, so neither the adults nor the children are exempt and the curve follows the no-child-discount shape. Italy has no annual out-of-pocket ceiling. Naples is the fifth Italian city, not Florence. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "National ticket framework on specialist visits and diagnostics plus the pharmaceutical co-payment, both regionally set; esenzione per reddito thresholds sit far below our ICP household", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Eurostat Urban Audit, indicator TT1019V, greater city. https://ec.europa.eu/eurostat/api/dissemination/statistics/1.0/data/urb_ctran?indic_ur=TT1019V&format=JSON&lang=EN", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average time of journey to work", "year": "2011", "note": "Urban Audit 31.93. Largest understatement in the Southern Europe shard, and it runs opposite to Naples and Marseille, so the stored set was not uniformly biased. Same ISTAT work-or-study caveat." }, "non_cash.sick_leave": { "source": "INPS indennità di malattia (general private-sector \"industria\" regime); Codice; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Agenzia delle Entrate, Risposta n. 76/2024 https://www.agenziaentrate.gov.it/portale/documents/20143/5946162/Risposta+n.+76_2024.pdf/f0933165-e52e-a519-8ca9-85d3f5162f23 ; Risoluzione n. 25/2025 https://www.agenziaentrate.gov.it/portale/documents/20143/8874453/Risoluzione+n.+25_2025/58473993-686e-9f56-723a-034c8a1dd3f6 (D.Lgs. 252/2005 art. 8 c.4)", "last_updated": "2026-08-11", "confidence": "high", "note": "Previdenza complementare deduction from reddito complessivo, EUR 5,164.57/yr, fixed in statute since 2007 (it is 10 million lire) and not indexed. A COMBINED ceiling: employer contributions to the fondo pensione count against the same figure. TFR conferred to the fund does NOT count and is not deductible, having never been taxed as income. less_employer_pct is deliberately left unset, which is the GENEROUS reading, rather than inventing an employer rate; flagged for a later pass. The \"lavoratori di prima occupazione\" extra platform in years 6-25 is deliberately unmodeled." } }, "yokohama": { "childcare.free_tier_start_age": { "source": "Kodomo Katei Cho (Children and Families Agency), free early childhood education and care policy: \"3 sai kara 5 sai made no subete no kodomotachi no riyoryo ga musho ni narimasu\" (usage fees become free for all children aged 3 to 5), in force 2019-10-01 https://www.cfa.go.jp/policies/kokoseido/mushouka/", "confidence": "high", "note": "Free tuition covers ages 3-5. Ages 0-2 stay fee-liable except for municipal-resident-tax-exempt households, with fees set by household tax band x municipality x hours, so the paid rate applies across the whole 0-2 class rather than to infants only. Before this, ages 1-5 billed the free-tier rate. Watch item: Tokyo has been funding wards to drop 0-2 fees (second and later children since FY2023, first children from 2025-09-01 in adopting wards incl. Meguro, Nerima, Kita), so the Tokyo infant rate may now be high for a growing share of the city; ward coverage is unconfirmed and the value is left alone.", "last_updated": "2026-08-01" }, "tax.filing_basis": { "source": "PwC WWTS - Japan (Tax administration): All income tax returns are filed on an individual basis in Japan; joint tax returns are not permitted.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.national_brackets": { "source": "Japan 2026 national income tax: peer-identical with Tokyo", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Fixed asset tax 1.4% (national standard)", "last_updated": "2026-05-26", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Consumption tax 10%", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Numbeo + realestate.co.jp Yokohama 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Japan National Health Insurance (inherited from Tokyo baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Yokohama hoikuen income-scaled infant fee; ages 3-5 free since 2019; gakudo after-school", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Yokohama 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Yokohama subway/JR commuter pass", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Yokohama", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "Tokyo/Yokohama->US west coast economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Jidou teate (national, inherited)", "last_updated": "2026-05-26", "confidence": "high", "note": "Japanese jido teate: income cap abolished Oct 2024, now universal, NOT means-tested." }, "non_cash._summary": { "source": "Japan labour law (national, inherited); 16 public holidays", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Yokohama. Its basis is converted with the rest of Japan and its rank inside Japan is preserved. Prior source: Derived, not observed: no all-levels total-comp measurement exists for Yokohama. Basis re-derived with its country anchor Tokyo; Yokohama's stored ratio to Tokyo is preserved. Prior source: Levels.fyi/japan-dev Yokohama SWE 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.34 x 0.5946 = 0.2022, R5-rounded down to 0.2. 0.5946 is the MEDIAN of (re-derived / previous) across the 1 Japan city that DO carry an all-levels measurement (Tokyo 0.37 to 0.22), so Yokohama moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Yokohama's rank inside Japan is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.34 was 0.9189 of Tokyo's 0.37; that ratio x Tokyo's re-derived 0.22 = 0.2022, R5-rounded down to 0.2. Tokyo's own figure: JPY 8,920,499 / 158.02 = USD 56,452, / USD 250,000 = 0.2258. This is a BASIS move, not a re-measurement: Yokohama's rank inside Japan is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.45, Tokyo metro commuter belt. Rebased 2026-08-11 with its country anchor (comp-index sweep A moved Oslo, Stockholm, Tokyo and Singapore); the documented ratio to the anchor is preserved, so this note's relative claim still holds." }, "healthcare.statutory_employee": { "source": "kyoukaikenpo.or.jp FY2026 per-prefecture insurance rate tables (Tokyo 9.85% total, employee half 4.925%; Kanagawa/Aichi/Osaka/Fukuoka within 9.85-10.13%); nenkin.go.jp standard remuneration grade table (health grade 50 = JPY 1,390,000/month) and the JPY 5,730,000 fiscal-year bonus ceiling (1 April to 31 March) shared by health, kaigo and the support levy; mhlw.go.jp FY2026 kodomo-kosodate shienkin 0.23% total / 0.115% employee, first charged on the April 2026 premium.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll. Tokyo's 9.85% is used as the national proxy because the block is enforced identical across a country's cities and the spread across our five metros is 0.07pp of employee rate. DELIBERATE EXCLUSIONS: kaigo 0.81% employee is age-gated to 40-64 and computeStatutoryHealthContribution cannot read age_bracket, so charging it would be wrong for the sub-40 half of our ICP (at the ceiling it would add about JPY 181,500/yr); the 0.115% kodomo-kosodate shienkin rides the same grades and bonus ceiling but funds child support, not health (about JPY 25,800/yr at the ceiling). The 2025 pension reform raising the pension ceiling to JPY 750,000 is not yet in force. TODO (phase-ins 2027-09, 2028-09, 2029-09): apply each step only when it takes effect." }, "non_cash.parental_leave_weeks": { "source": "Ikuji-Kaigo Kyugyo Ho art. 5 (childcare leave until the child turns 1, PER PARENT, non-transferable); Labor Standards Act art. 65 (maternity 6 wk pre + 8 wk post); Employment Insurance Act (ikuji kyugyo kyufu-kin); 2026 parameter references gyomu-keisan.jp (Reiwa 8), tfp-group.co.jp", "last_updated": "2026-08-06", "confidence": "high", "note": "110 wk = mother 14 (sanzen-sango kyugyo) + mother ikuji kyugyo from the end of the postnatal 8 wk to the first birthday (44 wk) + father ikuji kyugyo from birth to the first birthday (52 wk). The leave is PER PARENT and non-transferable, so both spans are summed, which is why parental_leave_both_parents is true (the Korea 120-wk precedent, an identical construction). The 4-wk shussei-ji ikuji kyugyo (post-birth papa leave) is taken WITHIN the father's span and its benefit days count against his 180-day 67% clock, so it adds nothing. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. Excluded: the 13% top-up to 80% for 28 days and the papa-mama 14-month extension (conditional on both parents' behaviour), the 18/24-month extension when no daycare place is found (conditional tail), and the JPY 500,000 shussan ikuji ichijikin (a lump sum, not leave pay). Was 52/67, the mother's span alone, as the old note itself admitted." }, "non_cash.parental_leave_paid_pct": { "source": "Health Insurance Act art. 102 (shussan teate-kin, 2/3 of standard remuneration); Employment Insurance Act art. 61-7 (ikuji kyugyo kyufu-kin, 67% first 180 days of each parent's leave, 50% after); ceilings per MHLW leaflet LL080731 ho 02 (2026-08-01) via https://roumu.com/pdf/2026080314.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "We store 0: since the 2026-08-07 Japan caps apply, all 110 weeks live in non_cash.parental_leave_capped (60% up to JPY 297,720/month), so no percentage-paid remainder weeks exist and a leftover headline rate would describe zero weeks. The old 60 blend (65.4 wk at 67 + 44.6 wk at 50, maternity at 66.7) is preserved as the capped block's pct. Research: docs/superpowers/research/2026-08-07-japan-parental-caps.md" }, "non_cash.parental_leave_capped": { "source": "MHLW leaflet LL080731 ho 02 (Reiwa 8 revision, in force 2026-08-01): ikuji kyugyo kyufu-kin shikyu jogen-gaku JPY 332,454/30-day unit at 67% and JPY 248,100 at 50%, wage-month base cap (kyugyo kaishi-ji chingin getsugaku jogen) JPY 496,200; fetched 2026-08-07 via https://roumu.com/pdf/2026080314.pdf (announcement https://roumu.com/archives/132696.html; MHLW hub https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000160564_00050.html); prior-year values confirmed against MHLW LL070722 https://www.mhlw.go.jp/content/001520023.pdf. Statutes: Employment Insurance Act art. 61-7 (ikuji kyugyo kyufu-kin), Health Insurance Act art. 102 (shussan teate-kin)", "last_updated": "2026-08-07", "confidence": "high", "note": "All 110 weeks flattened to ONE capped block, elbow-exact at the wage-month base cap: pct 60 (duration-weighted blend, rounded down) x JPY 496,200 = cap 297,720/month. Maternity (14 wk at 2/3, own std-remuneration ceiling JPY 1.39M/month) is FOLDED UNDER the shared ikuji ceiling per the Austria fold + Chile/Kazakhstan flatten precedents; true duration-weighted ceiling is ~298,064, so the stored 297,720 understates Japan (conservative). The two ikuji phases (67% first 180 days per parent, 50% after) share the same 496,200 elbow, so their flatten is lossless. No floor stored: the statutory kagen (96,090 wage-month) implies ~57,654/month, sub-ICP-relevant; omitting it is the less favorable direction. Ceilings revise every Aug 1 (indexed to average wages) - annual refresh row. Research: docs/superpowers/research/2026-08-07-japan-parental-caps.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "MHLW coinsurance schedule (30% for ages 6-69, 20% under 6) + the income-banded kougaku ryouyouhi monthly high-cost ceiling + Tokyo ward-level child medical subsidy", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 held at 50,000, structurally plausible against the high-cost ceiling, which is mid-revision (rising August 2026) and so usable only as an approximate check. Tokyo's ward child medical subsidy makes children close to free in practice, but ward income limits leave real uncertainty for a high earner, so the child increment is small rather than zero. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "MHLW coinsurance schedule (30% for ages 6-69, 20% under 6) + the income-banded kougaku ryouyouhi monthly high-cost ceiling + Tokyo ward-level child medical subsidy", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Statistics Bureau of Japan, Shakai Seikatsu Kihon Chosa 2021, rank12, Kanagawa-ken. https://www.stat.go.jp/data/shakai/2021/rank/zuhyou/rank/rank12.xlsx", "last_updated": "2026-08-08", "confidence": "high", "statistic": "1-nichi atari no tsukin tsugaku jikan", "year": "2021", "note": "CONVERSION: 1h40m = 100 min ROUND TRIP / 2 = 50.0 one way. Same school-bundling and age-10 caveats as Tokyo. Survey self-report." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "NTA tax answer 1135 https://www.nta.go.jp/taxes/shiraberu/taxanswer/shotoku/1135.htm ; MHLW iDeCo contribution limits https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/nenkin/nenkin/kyoshutsu/taishousha.html", "last_updated": "2026-08-11", "confidence": "medium", "note": "iDeCo, deductible in full under the small-enterprise-mutual-aid category; the binding ceiling lives in the pension law and is stated MONTHLY, converted x12 here. Category-2 insured WITH a corporate DC/DB plan: JPY 20,000/mo = 240,000/yr. Without one it is 23,000/mo = 276,000/yr. The lower figure is the house-rule choice, because a senior engineer at a large employer usually has a corporate pension; that is a BASIS choice, not a magnitude, and it is recorded in docs/corrections-ledger.md. Separate from the already-modeled kosei nenkin. Life- and annuity-insurance premium relief is deliberately not stacked. REFRESH ITEM: from 1 Dec 2026 the limit becomes 62,000/mo less employer DC/DB and the 20,000 sub-cap is abolished." } }, "nagoya": { "childcare.free_tier_start_age": { "source": "Kodomo Katei Cho (Children and Families Agency), free early childhood education and care policy: \"3 sai kara 5 sai made no subete no kodomotachi no riyoryo ga musho ni narimasu\" (usage fees become free for all children aged 3 to 5), in force 2019-10-01 https://www.cfa.go.jp/policies/kokoseido/mushouka/", "confidence": "high", "note": "Free tuition covers ages 3-5. Ages 0-2 stay fee-liable except for municipal-resident-tax-exempt households, with fees set by household tax band x municipality x hours, so the paid rate applies across the whole 0-2 class rather than to infants only. Before this, ages 1-5 billed the free-tier rate. Watch item: Tokyo has been funding wards to drop 0-2 fees (second and later children since FY2023, first children from 2025-09-01 in adopting wards incl. Meguro, Nerima, Kita), so the Tokyo infant rate may now be high for a growing share of the city; ward coverage is unconfirmed and the value is left alone.", "last_updated": "2026-08-01" }, "tax.filing_basis": { "source": "PwC WWTS - Japan (Tax administration): All income tax returns are filed on an individual basis in Japan; joint tax returns are not permitted.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.national_brackets": { "source": "Japan 2026 national income tax: peer-identical with Tokyo", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Fixed asset tax 1.4% (national standard)", "last_updated": "2026-05-26", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Consumption tax 10%", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Numbeo + Hitoshige Nagoya 2025-26", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Japan National Health Insurance (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Hoikuen income-scaled; ages 3-5 free; gakudo after-school", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Nagoya 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Nagoya subway commuter pass", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Nagoya", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "NGO->US west coast economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Jidou teate (national, inherited)", "last_updated": "2026-05-26", "confidence": "high", "note": "Japanese jido teate: income cap abolished Oct 2024, now universal, NOT means-tested." }, "non_cash._summary": { "source": "Japan labour law (national, inherited); 16 public holidays", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Nagoya. Its basis is converted with the rest of Japan and its rank inside Japan is preserved. Prior source: Derived, not observed: no all-levels total-comp measurement exists for Nagoya. Basis re-derived with its country anchor Tokyo; Nagoya's stored ratio to Tokyo is preserved. Prior source: Levels.fyi Nagoya SWE 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.32 x 0.5946 = 0.1903, R5-rounded down to 0.19. 0.5946 is the MEDIAN of (re-derived / previous) across the 1 Japan city that DO carry an all-levels measurement (Tokyo 0.37 to 0.22), so Nagoya moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Nagoya's rank inside Japan is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.32 was 0.8649 of Tokyo's 0.37; that ratio x Tokyo's re-derived 0.22 = 0.1903, R5-rounded down to 0.19. Tokyo's own figure: JPY 8,920,499 / 158.02 = USD 56,452, / USD 250,000 = 0.2258. This is a BASIS move, not a re-measurement: Nagoya's rank inside Japan is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.42, manufacturing/Toyota hub, below Tokyo. Rebased 2026-08-11 with its country anchor (comp-index sweep A moved Oslo, Stockholm, Tokyo and Singapore); the documented ratio to the anchor is preserved, so this note's relative claim still holds." }, "healthcare.statutory_employee": { "source": "kyoukaikenpo.or.jp FY2026 per-prefecture insurance rate tables (Tokyo 9.85% total, employee half 4.925%; Kanagawa/Aichi/Osaka/Fukuoka within 9.85-10.13%); nenkin.go.jp standard remuneration grade table (health grade 50 = JPY 1,390,000/month) and the JPY 5,730,000 fiscal-year bonus ceiling (1 April to 31 March) shared by health, kaigo and the support levy; mhlw.go.jp FY2026 kodomo-kosodate shienkin 0.23% total / 0.115% employee, first charged on the April 2026 premium.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll. Tokyo's 9.85% is used as the national proxy because the block is enforced identical across a country's cities and the spread across our five metros is 0.07pp of employee rate. DELIBERATE EXCLUSIONS: kaigo 0.81% employee is age-gated to 40-64 and computeStatutoryHealthContribution cannot read age_bracket, so charging it would be wrong for the sub-40 half of our ICP (at the ceiling it would add about JPY 181,500/yr); the 0.115% kodomo-kosodate shienkin rides the same grades and bonus ceiling but funds child support, not health (about JPY 25,800/yr at the ceiling). The 2025 pension reform raising the pension ceiling to JPY 750,000 is not yet in force. TODO (phase-ins 2027-09, 2028-09, 2029-09): apply each step only when it takes effect." }, "non_cash.parental_leave_weeks": { "source": "Ikuji-Kaigo Kyugyo Ho art. 5 (childcare leave until the child turns 1, PER PARENT, non-transferable); Labor Standards Act art. 65 (maternity 6 wk pre + 8 wk post); Employment Insurance Act (ikuji kyugyo kyufu-kin); 2026 parameter references gyomu-keisan.jp (Reiwa 8), tfp-group.co.jp", "last_updated": "2026-08-06", "confidence": "high", "note": "110 wk = mother 14 (sanzen-sango kyugyo) + mother ikuji kyugyo from the end of the postnatal 8 wk to the first birthday (44 wk) + father ikuji kyugyo from birth to the first birthday (52 wk). The leave is PER PARENT and non-transferable, so both spans are summed, which is why parental_leave_both_parents is true (the Korea 120-wk precedent, an identical construction). The 4-wk shussei-ji ikuji kyugyo (post-birth papa leave) is taken WITHIN the father's span and its benefit days count against his 180-day 67% clock, so it adds nothing. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. Excluded: the 13% top-up to 80% for 28 days and the papa-mama 14-month extension (conditional on both parents' behaviour), the 18/24-month extension when no daycare place is found (conditional tail), and the JPY 500,000 shussan ikuji ichijikin (a lump sum, not leave pay). Was 52/67, the mother's span alone, as the old note itself admitted." }, "non_cash.parental_leave_paid_pct": { "source": "Health Insurance Act art. 102 (shussan teate-kin, 2/3 of standard remuneration); Employment Insurance Act art. 61-7 (ikuji kyugyo kyufu-kin, 67% first 180 days of each parent's leave, 50% after); ceilings per MHLW leaflet LL080731 ho 02 (2026-08-01) via https://roumu.com/pdf/2026080314.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "We store 0: since the 2026-08-07 Japan caps apply, all 110 weeks live in non_cash.parental_leave_capped (60% up to JPY 297,720/month), so no percentage-paid remainder weeks exist and a leftover headline rate would describe zero weeks. The old 60 blend (65.4 wk at 67 + 44.6 wk at 50, maternity at 66.7) is preserved as the capped block's pct. Research: docs/superpowers/research/2026-08-07-japan-parental-caps.md" }, "non_cash.parental_leave_capped": { "source": "MHLW leaflet LL080731 ho 02 (Reiwa 8 revision, in force 2026-08-01): ikuji kyugyo kyufu-kin shikyu jogen-gaku JPY 332,454/30-day unit at 67% and JPY 248,100 at 50%, wage-month base cap (kyugyo kaishi-ji chingin getsugaku jogen) JPY 496,200; fetched 2026-08-07 via https://roumu.com/pdf/2026080314.pdf (announcement https://roumu.com/archives/132696.html; MHLW hub https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000160564_00050.html); prior-year values confirmed against MHLW LL070722 https://www.mhlw.go.jp/content/001520023.pdf. Statutes: Employment Insurance Act art. 61-7 (ikuji kyugyo kyufu-kin), Health Insurance Act art. 102 (shussan teate-kin)", "last_updated": "2026-08-07", "confidence": "high", "note": "All 110 weeks flattened to ONE capped block, elbow-exact at the wage-month base cap: pct 60 (duration-weighted blend, rounded down) x JPY 496,200 = cap 297,720/month. Maternity (14 wk at 2/3, own std-remuneration ceiling JPY 1.39M/month) is FOLDED UNDER the shared ikuji ceiling per the Austria fold + Chile/Kazakhstan flatten precedents; true duration-weighted ceiling is ~298,064, so the stored 297,720 understates Japan (conservative). The two ikuji phases (67% first 180 days per parent, 50% after) share the same 496,200 elbow, so their flatten is lossless. No floor stored: the statutory kagen (96,090 wage-month) implies ~57,654/month, sub-ICP-relevant; omitting it is the less favorable direction. Ceilings revise every Aug 1 (indexed to average wages) - annual refresh row. Research: docs/superpowers/research/2026-08-07-japan-parental-caps.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "MHLW coinsurance schedule (30% for ages 6-69, 20% under 6) + the income-banded kougaku ryouyouhi monthly high-cost ceiling + Tokyo ward-level child medical subsidy", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 held at 50,000, structurally plausible against the high-cost ceiling, which is mid-revision (rising August 2026) and so usable only as an approximate check. Tokyo's ward child medical subsidy makes children close to free in practice, but ward income limits leave real uncertainty for a high earner, so the child increment is small rather than zero. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "MHLW coinsurance schedule (30% for ages 6-69, 20% under 6) + the income-banded kougaku ryouyouhi monthly high-cost ceiling + Tokyo ward-level child medical subsidy", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Statistics Bureau of Japan, Shakai Seikatsu Kihon Chosa 2021, rank12, Aichi-ken. https://www.stat.go.jp/data/shakai/2021/rank/zuhyou/rank/rank12.xlsx", "last_updated": "2026-08-08", "confidence": "high", "statistic": "1-nichi atari no tsukin tsugaku jikan", "year": "2021", "note": "CONVERSION: 1h18m = 78 min ROUND TRIP / 2 = 39.0 one way. Same caveats. Survey self-report." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "NTA tax answer 1135 https://www.nta.go.jp/taxes/shiraberu/taxanswer/shotoku/1135.htm ; MHLW iDeCo contribution limits https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/nenkin/nenkin/kyoshutsu/taishousha.html", "last_updated": "2026-08-11", "confidence": "medium", "note": "iDeCo, deductible in full under the small-enterprise-mutual-aid category; the binding ceiling lives in the pension law and is stated MONTHLY, converted x12 here. Category-2 insured WITH a corporate DC/DB plan: JPY 20,000/mo = 240,000/yr. Without one it is 23,000/mo = 276,000/yr. The lower figure is the house-rule choice, because a senior engineer at a large employer usually has a corporate pension; that is a BASIS choice, not a magnitude, and it is recorded in docs/corrections-ledger.md. Separate from the already-modeled kosei nenkin. Life- and annuity-insurance premium relief is deliberately not stacked. REFRESH ITEM: from 1 Dec 2026 the limit becomes 62,000/mo less employer DC/DB and the 20,000 sub-cap is abolished." } }, "fukuoka": { "childcare.free_tier_start_age": { "source": "Kodomo Katei Cho (Children and Families Agency), free early childhood education and care policy: \"3 sai kara 5 sai made no subete no kodomotachi no riyoryo ga musho ni narimasu\" (usage fees become free for all children aged 3 to 5), in force 2019-10-01 https://www.cfa.go.jp/policies/kokoseido/mushouka/", "confidence": "high", "note": "Free tuition covers ages 3-5. Ages 0-2 stay fee-liable except for municipal-resident-tax-exempt households, with fees set by household tax band x municipality x hours, so the paid rate applies across the whole 0-2 class rather than to infants only. Before this, ages 1-5 billed the free-tier rate. Watch item: Tokyo has been funding wards to drop 0-2 fees (second and later children since FY2023, first children from 2025-09-01 in adopting wards incl. Meguro, Nerima, Kita), so the Tokyo infant rate may now be high for a growing share of the city; ward coverage is unconfirmed and the value is left alone.", "last_updated": "2026-08-01" }, "tax.filing_basis": { "source": "PwC WWTS - Japan (Tax administration): All income tax returns are filed on an individual basis in Japan; joint tax returns are not permitted.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.national_brackets": { "source": "Japan 2026 national income tax: peer-identical with Tokyo", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Fixed asset tax 1.4% (national standard)", "last_updated": "2026-05-26", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Consumption tax 10%", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Numbeo + Bamboo Routes Fukuoka 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Japan National Health Insurance (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Fukuoka hoikuen income-scaled; ages 3-5 free; gakudo after-school ~Y4,000", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Fukuoka 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Fukuoka City Subway Chika Pass Y12,340", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Fukuoka", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "FUK->US west coast economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Jidou teate (national, inherited)", "last_updated": "2026-05-26", "confidence": "high", "note": "Japanese jido teate: income cap abolished Oct 2024, now universal, NOT means-tested." }, "non_cash._summary": { "source": "Japan labour law (national, inherited); 16 public holidays", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Fukuoka. Its basis is converted with the rest of Japan and its rank inside Japan is preserved. Prior source: Derived, not observed: no all-levels total-comp measurement exists for Fukuoka. Basis re-derived with its country anchor Tokyo; Fukuoka's stored ratio to Tokyo is preserved. Prior source: Levels.fyi Fukuoka SWE 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.29 x 0.5946 = 0.1724, R5-rounded down to 0.17. 0.5946 is the MEDIAN of (re-derived / previous) across the 1 Japan city that DO carry an all-levels measurement (Tokyo 0.37 to 0.22), so Fukuoka moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Fukuoka's rank inside Japan is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.29 was 0.7838 of Tokyo's 0.37; that ratio x Tokyo's re-derived 0.22 = 0.1724, R5-rounded down to 0.17. Tokyo's own figure: JPY 8,920,499 / 158.02 = USD 56,452, / USD 250,000 = 0.2258. This is a BASIS move, not a re-measurement: Fukuoka's rank inside Japan is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.38, regional hub / startup-visa city, lower cost. Rebased 2026-08-11 with its country anchor (comp-index sweep A moved Oslo, Stockholm, Tokyo and Singapore); the documented ratio to the anchor is preserved, so this note's relative claim still holds." }, "healthcare.statutory_employee": { "source": "kyoukaikenpo.or.jp FY2026 per-prefecture insurance rate tables (Tokyo 9.85% total, employee half 4.925%; Kanagawa/Aichi/Osaka/Fukuoka within 9.85-10.13%); nenkin.go.jp standard remuneration grade table (health grade 50 = JPY 1,390,000/month) and the JPY 5,730,000 fiscal-year bonus ceiling (1 April to 31 March) shared by health, kaigo and the support levy; mhlw.go.jp FY2026 kodomo-kosodate shienkin 0.23% total / 0.115% employee, first charged on the April 2026 premium.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll. Tokyo's 9.85% is used as the national proxy because the block is enforced identical across a country's cities and the spread across our five metros is 0.07pp of employee rate. DELIBERATE EXCLUSIONS: kaigo 0.81% employee is age-gated to 40-64 and computeStatutoryHealthContribution cannot read age_bracket, so charging it would be wrong for the sub-40 half of our ICP (at the ceiling it would add about JPY 181,500/yr); the 0.115% kodomo-kosodate shienkin rides the same grades and bonus ceiling but funds child support, not health (about JPY 25,800/yr at the ceiling). The 2025 pension reform raising the pension ceiling to JPY 750,000 is not yet in force. TODO (phase-ins 2027-09, 2028-09, 2029-09): apply each step only when it takes effect." }, "non_cash.parental_leave_weeks": { "source": "Ikuji-Kaigo Kyugyo Ho art. 5 (childcare leave until the child turns 1, PER PARENT, non-transferable); Labor Standards Act art. 65 (maternity 6 wk pre + 8 wk post); Employment Insurance Act (ikuji kyugyo kyufu-kin); 2026 parameter references gyomu-keisan.jp (Reiwa 8), tfp-group.co.jp", "last_updated": "2026-08-06", "confidence": "high", "note": "110 wk = mother 14 (sanzen-sango kyugyo) + mother ikuji kyugyo from the end of the postnatal 8 wk to the first birthday (44 wk) + father ikuji kyugyo from birth to the first birthday (52 wk). The leave is PER PARENT and non-transferable, so both spans are summed, which is why parental_leave_both_parents is true (the Korea 120-wk precedent, an identical construction). The 4-wk shussei-ji ikuji kyugyo (post-birth papa leave) is taken WITHIN the father's span and its benefit days count against his 180-day 67% clock, so it adds nothing. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. Excluded: the 13% top-up to 80% for 28 days and the papa-mama 14-month extension (conditional on both parents' behaviour), the 18/24-month extension when no daycare place is found (conditional tail), and the JPY 500,000 shussan ikuji ichijikin (a lump sum, not leave pay). Was 52/67, the mother's span alone, as the old note itself admitted." }, "non_cash.parental_leave_paid_pct": { "source": "Health Insurance Act art. 102 (shussan teate-kin, 2/3 of standard remuneration); Employment Insurance Act art. 61-7 (ikuji kyugyo kyufu-kin, 67% first 180 days of each parent's leave, 50% after); ceilings per MHLW leaflet LL080731 ho 02 (2026-08-01) via https://roumu.com/pdf/2026080314.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "We store 0: since the 2026-08-07 Japan caps apply, all 110 weeks live in non_cash.parental_leave_capped (60% up to JPY 297,720/month), so no percentage-paid remainder weeks exist and a leftover headline rate would describe zero weeks. The old 60 blend (65.4 wk at 67 + 44.6 wk at 50, maternity at 66.7) is preserved as the capped block's pct. Research: docs/superpowers/research/2026-08-07-japan-parental-caps.md" }, "non_cash.parental_leave_capped": { "source": "MHLW leaflet LL080731 ho 02 (Reiwa 8 revision, in force 2026-08-01): ikuji kyugyo kyufu-kin shikyu jogen-gaku JPY 332,454/30-day unit at 67% and JPY 248,100 at 50%, wage-month base cap (kyugyo kaishi-ji chingin getsugaku jogen) JPY 496,200; fetched 2026-08-07 via https://roumu.com/pdf/2026080314.pdf (announcement https://roumu.com/archives/132696.html; MHLW hub https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000160564_00050.html); prior-year values confirmed against MHLW LL070722 https://www.mhlw.go.jp/content/001520023.pdf. Statutes: Employment Insurance Act art. 61-7 (ikuji kyugyo kyufu-kin), Health Insurance Act art. 102 (shussan teate-kin)", "last_updated": "2026-08-07", "confidence": "high", "note": "All 110 weeks flattened to ONE capped block, elbow-exact at the wage-month base cap: pct 60 (duration-weighted blend, rounded down) x JPY 496,200 = cap 297,720/month. Maternity (14 wk at 2/3, own std-remuneration ceiling JPY 1.39M/month) is FOLDED UNDER the shared ikuji ceiling per the Austria fold + Chile/Kazakhstan flatten precedents; true duration-weighted ceiling is ~298,064, so the stored 297,720 understates Japan (conservative). The two ikuji phases (67% first 180 days per parent, 50% after) share the same 496,200 elbow, so their flatten is lossless. No floor stored: the statutory kagen (96,090 wage-month) implies ~57,654/month, sub-ICP-relevant; omitting it is the less favorable direction. Ceilings revise every Aug 1 (indexed to average wages) - annual refresh row. Research: docs/superpowers/research/2026-08-07-japan-parental-caps.md" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "MHLW coinsurance schedule (30% for ages 6-69, 20% under 6) + the income-banded kougaku ryouyouhi monthly high-cost ceiling + Tokyo ward-level child medical subsidy", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 held at 50,000, structurally plausible against the high-cost ceiling, which is mid-revision (rising August 2026) and so usable only as an approximate check. Tokyo's ward child medical subsidy makes children close to free in practice, but ward income limits leave real uncertainty for a high earner, so the child increment is small rather than zero. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "MHLW coinsurance schedule (30% for ages 6-69, 20% under 6) + the income-banded kougaku ryouyouhi monthly high-cost ceiling + Tokyo ward-level child medical subsidy", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Statistics Bureau of Japan, Shakai Seikatsu Kihon Chosa 2021, rank12, Fukuoka-ken. https://www.stat.go.jp/data/shakai/2021/rank/zuhyou/rank/rank12.xlsx", "last_updated": "2026-08-08", "confidence": "high", "statistic": "1-nichi atari no tsukin tsugaku jikan", "year": "2021", "note": "CONVERSION: 1h09m = 69 min ROUND TRIP / 2 = 34.5 one way. Same caveats. Survey self-report." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "NTA tax answer 1135 https://www.nta.go.jp/taxes/shiraberu/taxanswer/shotoku/1135.htm ; MHLW iDeCo contribution limits https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/nenkin/nenkin/kyoshutsu/taishousha.html", "last_updated": "2026-08-11", "confidence": "medium", "note": "iDeCo, deductible in full under the small-enterprise-mutual-aid category; the binding ceiling lives in the pension law and is stated MONTHLY, converted x12 here. Category-2 insured WITH a corporate DC/DB plan: JPY 20,000/mo = 240,000/yr. Without one it is 23,000/mo = 276,000/yr. The lower figure is the house-rule choice, because a senior engineer at a large employer usually has a corporate pension; that is a BASIS choice, not a magnitude, and it is recorded in docs/corrections-ledger.md. Separate from the already-modeled kosei nenkin. Life- and annuity-insurance premium relief is deliberately not stacked. REFRESH ITEM: from 1 Dec 2026 the limit becomes 62,000/mo less employer DC/DB and the 20,000 sub-cap is abolished." } }, "guadalajara": { "tax.federal_brackets": { "source": "ISR tarifa anual 2026 inflation-adjusted limits (taxsummaries.pwc.com/mexico); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Predial Guadalajara ~0.25% (cadastral base low)", "last_updated": "2026-05-26", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "IVA 16%", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "TheLatinvestor + Numbeo Guadalajara 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Mercado Libre Inmuebles listing-count CDF, 2026-08-18 (departamentos en renta, fixed recamaras path segment + _PriceRange_0MXN-cap filter, bind-proven by zero-results collapse; base-page facet counts found STALE, filtered-URL counts authoritative; USD listings converted into MXN bands, not dropped; asking MXN/month per unit): metro pool GDL+Zapopan 2BR P25/med/P75 = 20,063/26,800/33,833 (n=298) -> 0.749/1.262; city-proper 2BR 0.816/1.141 (n=136); Zapopan alone 0.666/1.485 (n=162).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24. Verdict SPLIT leaning TOO NARROW: on the metro pool (the pool the stored metro-scale medians plausibly cover; Zapopan holds ~53% of the pooled formal 2BR stock) the lower edge breaches tolerance (-0.07) in the flattering direction and the upper edge holds; the breach is ROBUST to the portal's known bias (cheap-tail truncation and the un-pooled cheaper municipios Tlaquepaque/Tonala can only push the true lower edge further down). Applied 0.75/1.26 (nearest rounding of the metro pair). The metro is bimodal ACROSS municipios (city 0.82/1.14 vs Zapopan 0.67/1.49). 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-mexico-colombia.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (B1). NOTE the stored 2BR median (21,000) sits almost exactly at the measured metro P25 (20,063), not the metro median (26,800) - a LEVEL question, flagged, out of scope here." }, "healthcare._summary": { "source": "IMSS + private insurance (inherited from Mexico City baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Private guardería; no cash subsidy for private-sector tech families", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Guadalajara 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Guadalajara transit pass ~MXN 570", "last_updated": "2026-05-26", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo Guadalajara", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "GDL->US economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No federal child allowance for tech income (national, inherited; 0)", "last_updated": "2026-05-26", "confidence": "high" }, "non_cash._summary": { "source": "Mexican labour law (national, inherited); vacation 12, holidays 7", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Guadalajara. Its basis is converted with the rest of Mexico and its rank inside Mexico is preserved. Prior source: Glassdoor/Alcor Guadalajara SWE 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.27 x 0.7667 = 0.2070, R5-rounded down to 0.2. 0.7667 is the MEDIAN of (re-derived / previous) across the 1 Mexico city that DO carry an all-levels measurement (Mexico City 0.3 to 0.23), so Guadalajara moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Guadalajara's rank inside Mexico is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.27, 'Silicon Valley of Mexico', below CDMX." }, "tax.filing_basis": { "source": "SAT - ISR assessed per individual; no joint returns", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "non_cash.parental_leave_weeks": { "source": "LSS art. 101 (IMSS maternity subsidy, 100% of last daily contribution salary, 42+42 days); LFT art. 170; LFT art. 132 XXVII bis (5 working days paternity, employer-paid); UMA 2026 = MXN 117.31/day (INEGI comunicado 1/26; DOF 2026-01-09)", "last_updated": "2026-08-06", "confidence": "high", "note": "13 = 12 weeks IMSS maternity (84 days at 100%) + 5 working days of employer-paid paternity (1 week; the 20-day reform was never published in the DOF, so 5 days remains the law in 2026), counted per the Brazil precedent for short paternity. Was 12: maternity only. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "LSS art. 101 (IMSS maternity subsidy, 100% of last daily contribution salary, 42+42 days); LFT art. 170; LFT art. 132 XXVII bis (5 working days paternity, employer-paid); UMA 2026 = MXN 117.31/day (INEGI comunicado 1/26; DOF 2026-01-09)", "last_updated": "2026-08-06", "confidence": "high", "note": "100 is the statutory headline for both blocks. Cap arithmetic for the standing caps-convention item: the IMSS maternity subsidy base is capped at 25 UMA/day (UMA 2026 = MXN 117.31) = MXN 2,932.75/day, about MXN 1.07M/yr, so it binds only at the top of the CDMX tech band (at MXN 1.2M the effective maternity rate is about 89%); paternity is employer-paid at full salary, uncapped." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Mexican private consultation pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "CHECKED for the premium-as-out-of-pocket error and CLEAR: the value sits below any real private premium and above the IMSS contribution, which is consistent with cash-pay private consultations and not with a mis-booked premium. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Mexican private consultation pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "INEGI, Censo 2020, movilidad cotidiana, TIE_TRASLADO_TRAB, Zona Metropolitana de Guadalajara (7 municipios). https://www.inegi.org.mx/contenidos/programas/ccpv/2020/tabulados/ampliado/cpv2020_a_jal_10_movilidad_cotidiana.xlsx", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Tiempo de traslado al trabajo", "year": "2020", "note": "One way; band-midpoint mean over the metropolitan zone." }, "non_cash.sick_leave": { "source": "Ley del Seguro Social (LSS) Art. 96 (payment start / 52-week + 26-week extension),; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": "https://leyes-mx.com/ley_federal_del_trabajo/42.htm", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "SAT, Aportaciones complementarias de retiro https://www.sat.gob.mx/minisitio/DeduccionesPersonales/aportaciones_complementarias.html (LISR art. 151 fr. V) ; INEGI UMA 2026", "last_updated": "2026-08-11", "confidence": "high", "note": "Aportaciones complementarias de retiro, deductible at 10% of cumulative income up to 5 UMA anual. DENOMINATION TRAP HANDLED: the UMA \"anual\" is NOT diario x 365. INEGI computes mensual = diario x 30.4 and anual = mensual x 12 (= x 364.8), so UMA 2026 diario 117.31 gives anual 42,794.64 and 5 UMA = MXN 213,973.20 (2025 was 206,418.60). LISR art. 151 fr. V sits OUTSIDE the general 15% / 5-UMA personal-deduction cap and stands alone. Binds exactly at a 10% contribution." } }, "monterrey": { "tax.federal_brackets": { "source": "ISR tarifa anual 2026 inflation-adjusted limits (taxsummaries.pwc.com/mexico); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Predial Monterrey ~0.15% (cadastral base low)", "last_updated": "2026-05-26", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "IVA 16%", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "TheLatinvestor + Numbeo Monterrey 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Mercado Libre Inmuebles listing-count CDF, 2026-08-18 (departamentos en renta, fixed recamaras path segment + _PriceRange_0MXN-cap filter, bind-proven by zero-results collapse; base-page facet counts found STALE - facet said 303 where the filtered URL returns 550, card-audited as all genuine 2BR; USD listings converted into MXN bands, not dropped; asking MXN/month per unit): metro pool MTY+SPGG 2BR P25/med/P75 = 25,815/33,569/45,799 (n=769) -> 0.769/1.364; city-proper 2BR 0.800/1.189 (n=550); SPGG alone 0.841/1.122 (n=219) at ~1.9x the city-proper median.", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24. Verdict SPLIT adjudicated to the METRO pool (the repo's city = metro convention; much of the formal stock and the entire top tail live in San Pedro Garza Garcia, which is internally compressed but sits at ~1.9x the city median - the metro is bimodal ACROSS municipios): pooled upper edge +0.12 beyond tolerance in the flattering direction; city-proper alone ROUGHLY HOLDS. Applied 0.77/1.36 (nearest). Cheaper metro municipios (Guadalupe, Apodaca, San Nicolas, Santa Catarina, Escobedo) were NOT pooled; adding them widens BOTH edges further, so the applied pair is conservative. 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-mexico-colombia.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (B2). NOTE the stored 2BR median (25,300) sits between the measured city-proper P25 (23,900) and median (29,900) and does not clearly identify its pool - a LEVEL question, flagged, out of scope here." }, "healthcare._summary": { "source": "IMSS + private insurance (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Private guardería; no cash subsidy for private-sector tech families", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Monterrey 2026 (priciest MX metro)", "last_updated": "2026-05-26", "confidence": "medium", "note": "Groceries 8500/adult is high end vs CDMX 7000 — flag if recalibrating." }, "transit._summary": { "source": "Monterrey metro pass MXN 576", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Monterrey", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "MTY->US economy 2025-26 (near US border)", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No federal child allowance for tech income (national, inherited; 0)", "last_updated": "2026-05-26", "confidence": "high" }, "non_cash._summary": { "source": "Mexican labour law (national, inherited); vacation 12, holidays 7", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Monterrey. Its basis is converted with the rest of Mexico and its rank inside Mexico is preserved. Prior source: Glassdoor/Howdy Monterrey SWE 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.3 x 0.7667 = 0.2300, R5-rounded down to 0.23. 0.7667 is the MEDIAN of (re-derived / previous) across the 1 Mexico city that DO carry an all-levels measurement (Mexico City 0.3 to 0.23), so Monterrey moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Monterrey's rank inside Mexico is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.30, wealthiest Mexican city, corporate/industrial HQ." }, "tax.filing_basis": { "source": "SAT - ISR assessed per individual; no joint returns", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "non_cash.parental_leave_weeks": { "source": "LSS art. 101 (IMSS maternity subsidy, 100% of last daily contribution salary, 42+42 days); LFT art. 170; LFT art. 132 XXVII bis (5 working days paternity, employer-paid); UMA 2026 = MXN 117.31/day (INEGI comunicado 1/26; DOF 2026-01-09)", "last_updated": "2026-08-06", "confidence": "high", "note": "13 = 12 weeks IMSS maternity (84 days at 100%) + 5 working days of employer-paid paternity (1 week; the 20-day reform was never published in the DOF, so 5 days remains the law in 2026), counted per the Brazil precedent for short paternity. Was 12: maternity only. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "LSS art. 101 (IMSS maternity subsidy, 100% of last daily contribution salary, 42+42 days); LFT art. 170; LFT art. 132 XXVII bis (5 working days paternity, employer-paid); UMA 2026 = MXN 117.31/day (INEGI comunicado 1/26; DOF 2026-01-09)", "last_updated": "2026-08-06", "confidence": "high", "note": "100 is the statutory headline for both blocks. Cap arithmetic for the standing caps-convention item: the IMSS maternity subsidy base is capped at 25 UMA/day (UMA 2026 = MXN 117.31) = MXN 2,932.75/day, about MXN 1.07M/yr, so it binds only at the top of the CDMX tech band (at MXN 1.2M the effective maternity rate is about 89%); paternity is employer-paid at full salary, uncapped." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Mexican private consultation pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "CHECKED for the premium-as-out-of-pocket error and CLEAR: the value sits below any real private premium and above the IMSS contribution, which is consistent with cash-pay private consultations and not with a mis-booked premium. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Mexican private consultation pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "INEGI, Censo 2020, movilidad cotidiana, TIE_TRASLADO_TRAB, Zona Metropolitana de Monterrey (16 municipios). https://www.inegi.org.mx/contenidos/programas/ccpv/2020/tabulados/ampliado/cpv2020_a_nl_10_movilidad_cotidiana.xlsx", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Tiempo de traslado al trabajo", "year": "2020", "note": "One way; band-midpoint mean over the metropolitan zone." }, "non_cash.sick_leave": { "source": "Ley del Seguro Social (LSS) Art. 96 (payment start / 52-week + 26-week extension),; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": "https://leyes-mx.com/ley_federal_del_trabajo/42.htm", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "SAT, Aportaciones complementarias de retiro https://www.sat.gob.mx/minisitio/DeduccionesPersonales/aportaciones_complementarias.html (LISR art. 151 fr. V) ; INEGI UMA 2026", "last_updated": "2026-08-11", "confidence": "high", "note": "Aportaciones complementarias de retiro, deductible at 10% of cumulative income up to 5 UMA anual. DENOMINATION TRAP HANDLED: the UMA \"anual\" is NOT diario x 365. INEGI computes mensual = diario x 30.4 and anual = mensual x 12 (= x 364.8), so UMA 2026 diario 117.31 gives anual 42,794.64 and 5 UMA = MXN 213,973.20 (2025 was 206,418.60). LISR art. 151 fr. V sits OUTSIDE the general 15% / 5-UMA personal-deduction cap and stands alone. Binds exactly at a 10% contribution." } }, "rio-de-janeiro": { "tax.filing_basis": { "source": "PwC WWTS - Brazil (Tax administration): spouses may elect to file jointly or separately; individual is the two-earner default.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Brazil 2026 IRPF: peer-identical with São Paulo", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "IPTU Rio ~0.85% effective", "last_updated": "2026-05-26", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "ICMS/ISS (inherited from São Paulo baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "TheLatinvestor + Numbeo Rio 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Zap Imoveis glue-API listing-count CDF, 2026-08-18 (apartments for rent, city proper, addressCity/State Rio de Janeiro; priceMax RE-PROVEN this run to filter BARE ALUGUEL on Rio itself - 2Q count 2,012 at priceMax=3000 vs 1,467 at rentalTotalPriceMax=3000, price separate from monthlyCondoFee/yearlyIptu; bedroom bins exact; asking BRL/month): 2 quartos P25/med/P75 = 1,496/2,987/5,909 (n=3,977) -> 0.501/1.978, robust to interpolation (cluster reading 0.499/1.970); 1 quarto 0.481/1.675 (n=1,712, studios fold in).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW by the widest margin in the whole program (2Q -0.319/+0.738; wider than Sao Paulo's 0.563/1.616) - the classic Zona Sul/Barra premium vs Zona Norte/Oeste cheap dual market in one citywide pool. Applied 0.50/1.98 (nearest; BOTH per-city guard exceptions - p25 floor and p75 bound). Caveats: informal cheap tail truncated (true lower edge lower still); multi-agency duplication fattens the premium tail (no duplicate-free corroboration pool was run for Rio, unlike Sao Paulo/Brasilia - recorded). 2 quartos anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-brazil.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (C). NOTE stored 2BR median 4,050 vs measured pool median ~2,987 - a LEVEL question, flagged, out of scope here." }, "healthcare._summary": { "source": "SUS public + private plano de saúde (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Private creche; public waitlisted, no cash subsidy for tech families", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Rio 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Rio RioCard monthly ~R$255", "last_updated": "2026-05-26", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo Rio", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "GIG->US economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Bolsa Família means-tested, 0 for tech income (national, inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "non_cash._summary": { "source": "Brazilian labour law (national, inherited); 30-day statutory vacation = 22 working days", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Rio de Janeiro. Its basis is converted with the rest of Brazil and its rank inside Brazil is preserved. Prior source: Glassdoor Rio SWE 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.3 x 0.4688 = 0.1406, R5-rounded down to 0.14. 0.4688 is the MEDIAN of (re-derived / previous) across the 1 Brazil city that DO carry an all-levels measurement (São Paulo 0.32 to 0.15), so Rio de Janeiro moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Rio de Janeiro's rank inside Brazil is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.30, just below São Paulo." }, "non_cash.parental_leave_capped": { "source": "Lei 8.213/91 art. 72 (salario-maternidade = remuneracao integral, paid by the empresa and compensated against INSS contributions, art. 72 par. 1); CF art. 248 (benefits outside the RGPS teto observe the art. 37 XI ceiling, the STF-minister subsidio); STF ADI 1946 (salario-maternidade exempt from the RGPS teto); subsidio = R$ 46,366.19/month from 1 Feb 2025, Lei 14.520/2023 https://www.planalto.gov.br/ccivil_03/_ato2023-2026/2023/lei/l14520.htm (final step; the Dec-2025 judiciary raise covers servidores, not ministers)", "last_updated": "2026-08-17", "confidence": "high", "note": "The 17.143 maternity weeks pay 100% of remuneration up to the constitutional teto of R$ 46,366.19/month (R$ 556,394.28/yr); the 0.857-wk paternity remainder stays on parental_leave_paid_pct 100 (ADCT art. 10 par. 1, employer-paid at full salary, no ceiling). The ceiling that binds a CLT empregada is CF art. 248's STF-minister subsidio, NOT the RGPS teto (R$ 8,475.55 in 2026), per ADI 1946; the RGPS teto does bind the other claimant categories our ICP is not in. Refutation branch checked and rejected: no statute or binding precedent obliges the employer to top up above the teto, and modelling the cap is the conservative call. At the R$ 782,000/yr base scenario the maternity slice pays 71.2% and the blended 18-wk rate is 72.6%; below R$ 556k/yr it stays an honest 100%. UNIT: the subsidio is a MONTHLY figure, revised only by lei (last step 1 Feb 2025). Research: docs/superpowers/research/2026-08-17-parental-leave-caps.md." }, "non_cash.parental_leave_weeks": { "source": "CF art. 7 XVIII (120 dias); CLT art. 392; Lei 8.213/91 arts. 71-73 (salario-maternidade = remuneracao integral); ADCT art. 10 par. 1 (5 days paternity, in force through 2026); Lei 15.371 phase-in: Senado https://www12.senado.leg.br/noticias/materias/2026/04/01/sancionado-o-aumento-gradual-da-licenca-paternidade", "last_updated": "2026-08-06", "confidence": "high", "note": "18 wk = 120 days maternity at 100% (17.143 wk; capped at the constitutional STF-minister teto, BRL 46,366.19/month, modelled in non_cash.parental_leave_capped as of 2026-08-17; the 2026-08-06 judgment that the teto is irrelevant to any ICP was corrected, it binds at the USD-anchored scenarios this site prices, e.g. the R$ 782,000/yr base scenario) + 5 days paternity at full salary, employer-paid (0.71 wk); 17.86 rounds to 18 (the Germany 66.87-to-67 convention). Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. WATCH ITEM: Lei 15.371 (sanctioned 2026-03-31) phases paternity to 10 days from 2027-01-01 (18.6 wk), 15 from 2028 (19.3), 20 from 2029 (20), with an INSS-reimbursed salario-paternidade at full remuneration; re-touch each January. Excluded: Programa Empresa Cidada 180/20-day extensions (employer opt-in via tax incentive) and collective-agreement top-ups." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Brazilian private consultation and co-payment pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "CHECKED for the premium-as-out-of-pocket error and CLEAR: SUS is free at the point of use and these values are verified far below any real plano de saude premium, so they are co-pays and cash consultations rather than the plan itself. Two city tiers preserved. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Brazilian private consultation and co-payment pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "IBGE, Censo Demografico 2022, SIDRA table 10331, Regiao Metropolitana do Rio de Janeiro (22 municipios). https://sidra.ibge.gov.br/tabela/10331", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Tempo habitual de deslocamento de casa para o trabalho", "year": "2022", "note": "One way by construction; band-midpoint mean. Longest in Brazil." }, "non_cash.sick_leave": { "source": "Lei 8.213/1991 (Plano de Benefícios da Previdência Social) , Art. 60 §3º (employer; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": "https://www.gov.br/inss/pt-br/direitos-e-deveres/inscricao-e-contribuicao/tabela-de-contribuicao-mensal", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Receita Federal, PGBL e VGBL https://www.gov.br/receitafederal/pt-br/acesso-a-informacao/perguntas-frequentes/imposto-de-renda/dirpf/declaracao/pgvl-vgbl (Lei 9.532/97 art. 11)", "last_updated": "2026-08-11", "confidence": "high", "note": "PGBL contributions are deductible up to 12% of rendimentos tributaveis, i.e. 12% of gross for an employee. No monetary ceiling, so no max key. VGBL is NOT deductible. The deduction requires the COMPLETE declaration and this value assumes it, which is consistent with the branch: Sao Paulo carries no standard_deduction, so the engine already models a complete/itemised basis, and at BRL 300k the 12% (R$36,000) beats the simplificado cap (R$17,640) so choosing complete is genuinely rational. Lei 15.270/2025, in force 1 Jan 2026, leaves art. 11 untouched. Never binds at 6% or 10%: pure correctness insurance." } }, "belo-horizonte": { "tax.filing_basis": { "source": "PwC WWTS - Brazil (Tax administration): spouses may elect to file jointly or separately; individual is the two-earner default.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Brazil 2026 IRPF: peer-identical with São Paulo", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "IPTU Belo Horizonte ~0.75% effective", "last_updated": "2026-05-26", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "ICMS/ISS (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "myside.com.br + Numbeo Belo Horizonte 2025-26", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Zap Imoveis glue-API listing-count CDF, 2026-08-18 (apartments for rent, addressCity Belo Horizonte / addressState Minas Gerais, bare aluguel semantics proven on Rio this run and carried; bedroom bins exact; asking BRL/month): 2 quartos -> 0.708/1.423 (n=2,233); 1 quarto 0.754/1.241 (n includes studios; borderline-inside).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW at the primary slice (2Q -0.112/+0.183; the 1Q slice is borderline-inside and the primary decides). Applied 0.71/1.42 (nearest). Single portal; informal cheap tail truncated - the true spread is wider, the fix conservative. 2 quartos anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-brazil.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (C). Level observations logged out of scope." }, "healthcare._summary": { "source": "SUS public + private (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Private creche (O Tempo BH survey 2025); no cash subsidy for tech families", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Belo Horizonte 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "BH bus monthly ~R$345", "last_updated": "2026-05-26", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo Belo Horizonte", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "CNF->US economy 2025-26 (via hub)", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Bolsa Família means-tested, 0 for tech income (national, inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "non_cash._summary": { "source": "Brazilian labour law (national, inherited); 22 working-day vacation", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Belo Horizonte. Its basis is converted with the rest of Brazil and its rank inside Brazil is preserved. Prior source: Levels.fyi Belo Horizonte SWE 2026 (San Pedro Valley)", "last_updated": "2026-08-11", "confidence": "low", "note": "0.28 x 0.4688 = 0.1313, R5-rounded down to 0.13. 0.4688 is the MEDIAN of (re-derived / previous) across the 1 Brazil city that DO carry an all-levels measurement (São Paulo 0.32 to 0.15), so Belo Horizonte moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Belo Horizonte's rank inside Brazil is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.28, real tech cluster (Google/Coinbase offices)." }, "non_cash.parental_leave_capped": { "source": "Lei 8.213/91 art. 72 (salario-maternidade = remuneracao integral, paid by the empresa and compensated against INSS contributions, art. 72 par. 1); CF art. 248 (benefits outside the RGPS teto observe the art. 37 XI ceiling, the STF-minister subsidio); STF ADI 1946 (salario-maternidade exempt from the RGPS teto); subsidio = R$ 46,366.19/month from 1 Feb 2025, Lei 14.520/2023 https://www.planalto.gov.br/ccivil_03/_ato2023-2026/2023/lei/l14520.htm (final step; the Dec-2025 judiciary raise covers servidores, not ministers)", "last_updated": "2026-08-17", "confidence": "high", "note": "The 17.143 maternity weeks pay 100% of remuneration up to the constitutional teto of R$ 46,366.19/month (R$ 556,394.28/yr); the 0.857-wk paternity remainder stays on parental_leave_paid_pct 100 (ADCT art. 10 par. 1, employer-paid at full salary, no ceiling). The ceiling that binds a CLT empregada is CF art. 248's STF-minister subsidio, NOT the RGPS teto (R$ 8,475.55 in 2026), per ADI 1946; the RGPS teto does bind the other claimant categories our ICP is not in. Refutation branch checked and rejected: no statute or binding precedent obliges the employer to top up above the teto, and modelling the cap is the conservative call. At the R$ 782,000/yr base scenario the maternity slice pays 71.2% and the blended 18-wk rate is 72.6%; below R$ 556k/yr it stays an honest 100%. UNIT: the subsidio is a MONTHLY figure, revised only by lei (last step 1 Feb 2025). Research: docs/superpowers/research/2026-08-17-parental-leave-caps.md." }, "non_cash.parental_leave_weeks": { "source": "CF art. 7 XVIII (120 dias); CLT art. 392; Lei 8.213/91 arts. 71-73 (salario-maternidade = remuneracao integral); ADCT art. 10 par. 1 (5 days paternity, in force through 2026); Lei 15.371 phase-in: Senado https://www12.senado.leg.br/noticias/materias/2026/04/01/sancionado-o-aumento-gradual-da-licenca-paternidade", "last_updated": "2026-08-06", "confidence": "high", "note": "18 wk = 120 days maternity at 100% (17.143 wk; capped at the constitutional STF-minister teto, BRL 46,366.19/month, modelled in non_cash.parental_leave_capped as of 2026-08-17; the 2026-08-06 judgment that the teto is irrelevant to any ICP was corrected, it binds at the USD-anchored scenarios this site prices, e.g. the R$ 782,000/yr base scenario) + 5 days paternity at full salary, employer-paid (0.71 wk); 17.86 rounds to 18 (the Germany 66.87-to-67 convention). Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. WATCH ITEM: Lei 15.371 (sanctioned 2026-03-31) phases paternity to 10 days from 2027-01-01 (18.6 wk), 15 from 2028 (19.3), 20 from 2029 (20), with an INSS-reimbursed salario-paternidade at full remuneration; re-touch each January. Excluded: Programa Empresa Cidada 180/20-day extensions (employer opt-in via tax incentive) and collective-agreement top-ups." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Brazilian private consultation and co-payment pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "CHECKED for the premium-as-out-of-pocket error and CLEAR: SUS is free at the point of use and these values are verified far below any real plano de saude premium, so they are co-pays and cash consultations rather than the plan itself. Two city tiers preserved. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Brazilian private consultation and co-payment pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "IBGE, Censo Demografico 2022, SIDRA table 10331, Regiao Metropolitana de Belo Horizonte (34 municipios). https://sidra.ibge.gov.br/tabela/10331", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Tempo habitual de deslocamento de casa para o trabalho", "year": "2022", "note": "One way by construction; band-midpoint mean. The only Brazilian city that moved UP, which is why the Brazilian error was not a uniform scale factor." }, "non_cash.sick_leave": { "source": "Lei 8.213/1991 (Plano de Benefícios da Previdência Social) , Art. 60 §3º (employer; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": "https://www.gov.br/inss/pt-br/direitos-e-deveres/inscricao-e-contribuicao/tabela-de-contribuicao-mensal", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Receita Federal, PGBL e VGBL https://www.gov.br/receitafederal/pt-br/acesso-a-informacao/perguntas-frequentes/imposto-de-renda/dirpf/declaracao/pgvl-vgbl (Lei 9.532/97 art. 11)", "last_updated": "2026-08-11", "confidence": "high", "note": "PGBL contributions are deductible up to 12% of rendimentos tributaveis, i.e. 12% of gross for an employee. No monetary ceiling, so no max key. VGBL is NOT deductible. The deduction requires the COMPLETE declaration and this value assumes it, which is consistent with the branch: Sao Paulo carries no standard_deduction, so the engine already models a complete/itemised basis, and at BRL 300k the 12% (R$36,000) beats the simplificado cap (R$17,640) so choosing complete is genuinely rational. Lei 15.270/2025, in force 1 Jan 2026, leaves art. 11 untouched. Never binds at 6% or 10%: pure correctness insurance." } }, "mumbai": { "tax.federal_brackets": { "source": "India FY 2026-27 new regime + 4% cess, peer-identical with Bangalore. Boundaries corrected 2026-08-01 (see the Bangalore tax.federal_brackets note): the INR 75,000 standard deduction is folded in by shifting EVERY slab and surcharge boundary, not just the first. https://www.incometax.gov.in/iec/foportal/help/individual/return-applicable-1", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "MCGM capital-value property tax ~0.2% effective", "last_updated": "2026-05-26", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "GST (inherited from Bangalore baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Numbeo + NoBroker Mumbai 2026 (India's priciest housing)", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "MagicBricks listing-count CDF, 2026-08-17 (flats for rent, Greater Mumbai city id 4320 - Thane/Navi Mumbai excluded; BudgetMax sweep, inclusive, operates on the rent field with deposit separate, verified against embedded JSON; asking INR/month per unit): 2 BHK P25/med/P75 = 58,727/76,447/113,758 (N=9,249) -> 0.768/1.488, applied 0.77/1.49; 1 BHK 0.761/1.318 (N=3,568). Independent corroboration: NoBroker owner-direct MMR-wide 2 BHK 0.793/1.313 (n=1,322).", "last_updated": "2026-08-17", "confidence": "medium", "note": "REBASED 2026-08-17 (ledger cluster housing-0817) from the calibrated default 0.82/1.24, which understates Mumbai's upper tail badly (+0.25 at 2 BHK on the primary pool; both independent portals put the upper edge out of tolerance, and every lower edge sits on the low side of the default). Same basis as the band's claim (citywide listing pool, fixed BHK, per-unit monthly asking rent). MagicBricks is primary: right geography (Greater Mumbai, our city) and fine-grained sweeps; NoBroker (MMR-wide, owner-direct mid-market, no luxury belt) corroborates direction with a narrower top - the 1.31-vs-1.49 gap is a real between-pool difference (broker duplication + the Worli/Bandra/Juhu luxury belt fatten MB's tail; read 1.49 with that caveat). Dual/heterogeneous-market class, same treatment as Mexico City and Bangkok. 2 BHK anchor per the flagship convention (2 BHK = 2 bedrooms + hall + kitchen). Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-17-housing-band-fixes-mumbai.md; decision: docs/superpowers/research/2026-08-17-housing-band-fixes.md (S3). NOTE the stored 2BR median (115,000 INR) sits at roughly the measured 2 BHK p75 - only defensible as an unwritten expat-corridor basis; a LEVEL question, flagged, out of scope here." }, "healthcare._summary": { "source": "Private insurance (inherited from Bangalore baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "DaycareFee Mumbai; private creche, no public subsidy for tech families", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Mumbai 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Mumbai local train + metro pass", "last_updated": "2026-05-26", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo + ZoloStays Mumbai", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "BOM->US economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No statutory child allowance (national, inherited; 0)", "last_updated": "2026-05-26", "confidence": "high" }, "non_cash._summary": { "source": "Indian labour law (national, inherited); ~18 leave days, 15 holidays", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Mumbai: INR 1,996,084, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "INR 1,996,084 / 95.48 = USD 20,906; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.0836, R5-rounded down to 0.08. REPLACES 0.21 (2026-05-26), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.18, financial capital, high COL relative to tech pay." }, "tax.filing_basis": { "source": "Income-tax Act - individual assessee; no joint/spousal returns", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "tax.surcharge_brackets": { "source": "IN-1: new-regime high-earner surcharge on the income tax; 0 (<50L) / 10% (50L-1cr) / 15% (1cr-2cr) / 25% (>2cr, the new-regime cap), applied as a step on total income with marginal relief; 4% cess then runs on tax+surcharge. PwC IN taxes-on-personal-income; Taxmann surcharge/cess.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.social_security_cap": { "source": "EPFO International Worker framework (EPF Scheme 1952 para 83, 2008 amendment, carried forward into the EPF Scheme, 2026 notified 1 July 2026 under the Code on Social Security), described identically across AscentHR, cgaa.org, gconnect.in's EPFO brochure summary and taxaaram.com; SSA list per Dezan Shira / ExpatOrbit, UK signing 10 Feb 2026 per KPMG Flash Alert; US absence per india-briefing.com (\"pursuing\"). epfindia.gov.in itself was unreachable on two separate attempts (DNS, then connection timeout on re-verification 2026-08-06); the gap is network access from our environment, not thin sourcing. docs/superpowers/research/2026-08-06-forced-savings-foreigner-rules.md", "last_updated": "2026-08-06", "confidence": "medium-high", "note": "The INR 15,000/mo wage ceiling (cap INR 180,000/yr, max employee contribution INR 21,600) is the DOMESTIC basis. A foreign-passport International Worker at an EPF-covered establishment contributes 12% on FULL salary with no ceiling, unless a home-country SSA plus Certificate of Coverage exempts them. Engine seam: computeIndiaTaxes reads state.in_epf_intl_worker, absent or false = this capped basis, true = uncapped 12%. The target side of a comparison defaults to the uncapped IW basis when the SOURCE country holds no SSA with India (INDIA_SSA_COUNTRIES in app.js; the US has none, the UK signed Feb 2026); SSA-source movers keep this capped basis rather than the full COC exemption, which is paperwork-gated (Chile Law 18.156 precedent, conservative rule). At INR 5,000,000 gross the two bases are 21,600 vs 600,000 of employee EPF. Confidence medium-high: the mechanism is uniform across independent sources but the EPFO primary text was not reached; a Karnataka HC challenge to the uncapped rule exists and EPFO has stated the framework stands. TODO (2026-08-06): the EPF Scheme, 2026 applies the INR 15,000 ceiling to international workers in some provisions while others still read \"total wages\"; EPFO called this a typographical and drafting inconsistency, said no proposal exists to change the international-worker framework, and is weighing a corrigendum (Business Standard, 16 July 2026). Re-check this cap once the corrigendum issues. Re-checked 2026-08-07: the corrigendum has still NOT issued (latest reporting remains Business Standard 2026-07-16; EPFO maintains the IW framework is carried forward unchanged, so the uncapped IW default stands), and the DOMESTIC INR 15,000/mo ceiling is also confirmed still in force (the proposed rise to 21k is delayed indefinitely; the SC four-month directive of Jan 2026 has produced no enacted decision). The watch now rides docs/data-refresh-calendar.md rather than this dated note." }, "non_cash.parental_leave_weeks": { "source": "Maternity Benefit Act 1961 s.5(3) as amended 2017: 26 weeks at 100% of average daily wage, employer-paid; Act text via state-government mirror (wb.gov.in PDF; central gov.in domains unreachable at check, flagged)", "last_updated": "2026-08-06", "confidence": "high", "note": "26 = maternity only, and on the household-inclusive basis this IS the household total: India has no statutory paid paternity or parental leave for private-sector employees as of 2026 (CCS Rule 43-A covers government employees only; the 2017 Paternity Benefit Bill lapsed). Caveats: establishments with 10+ employees; 26 weeks for the first two children, 12 for the third and beyond (we store the first-child case). Confirmed on the basis 2026-08-06. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "Maternity Benefit Act 1961 s.5(3) as amended 2017: 26 weeks at 100% of average daily wage, employer-paid; Act text via state-government mirror (wb.gov.in PDF; central gov.in domains unreachable at check, flagged)", "last_updated": "2026-08-06", "confidence": "high", "note": "100% of the average daily wage (3-month average), employer-paid, no monetary cap. Confirmed 2026-08-06." }, "non_cash.vacation_days_statutory": { "source": "Tamil Nadu Shops and Establishments Act 1947 s.25; Delhi Shops and Establishments Act 1954; Telangana Shops and Establishments Act 1988 s.30; Karnataka Shops and Commercial Establishments Act 1961 s.15; Maharashtra Shops and Establishments Act 2017 s.18(3); Factories Act 1948 s.79", "last_updated": "2026-08-08", "confidence": "medium", "note": "CORRECTED 2026-08-08, 18 to 12. India has NO national annual-leave statute for office workers, so this nationally-enforced field has to pick one number for six cities in five states. The old 18 was sourced to Karnataka, and Karnataka's own Shops and Commercial Establishments Act 1961 s.15 does not say 18: it grants one day per twenty days worked, the same accrual formula as Factories Act 1948 s.79, whose administered output is about 15. So 18 was unsupported by the statute it cited. Verified range across the states we actually model: Tamil Nadu (Chennai) s.25 gives a flat 12, Delhi 15, Telangana (Hyderabad) s.30 a flat 15, and Karnataka (Bangalore) plus Maharashtra (Mumbai, Pune) s.18(3) both run the 1-in-20 accrual that lands near 15. We store 12, the lower end, per rule 5 and because Tamil Nadu's is a clean section-specific floor covering one of our own cities rather than an extrapolation. LIVE ALTERNATIVE: 15 is equally defensible and is the modal figure across five of the six cities; flip this if the owner prefers representativeness to conservatism. Also checked: the OSH Code 2020 s.32 is in force from 2025-11-21 with rules from May 2026, but it is scoped to factory-type establishments and does not yet displace the state Shops Acts for commercial or IT employers. No six-day or calendar conversion applies: the flat states state days directly, and the accrual denominator counts actual attendance days." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series", "last_updated": "2026-08-08", "confidence": "low", "note": "Almost all Indian health spending is out of pocket, and employer group cover is a premium that is not booked here. Two city tiers preserved. FLAGGED for a future look: claims and premium data both hint row 1 may be low for a Bangalore or Mumbai private-hospital household. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. MoveInSync gives 55 (2019) and about 63 (2023) on a biased subpopulation; not cited." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-F.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Income Tax Department, New tax regime vs old tax regime FAQs https://www.incometax.gov.in/iec/foportal/help/new-tax-vs-old-tax-regime-faqs", "last_updated": "2026-08-11", "confidence": "high", "note": "No deduction under the regime the engine models. The NEW regime (s.115BAC) is the statutory default from AY 2024-25 and disallows Chapter-VIA deductions except 80CCD(2), 80CCH and 80JJAA, so 80C (where EPF would sit) and the extra 80CCD(1B) NPS deduction are both unavailable; 80CCD(2) is EMPLOYER NPS money, not the employee's own. computeIndiaTaxes models the new regime (slabs shifted by the INR 75,000 standard deduction) and charges EPF in payroll without removing it from the base, which is correct there. If an old-regime path is ever added, the value there is INR 200,000 and it must net off the EPF already flowing through payroll." } }, "hyderabad": { "tax.federal_brackets": { "source": "India FY 2026-27 new regime + 4% cess, peer-identical with Bangalore. Boundaries corrected 2026-08-01 (see the Bangalore tax.federal_brackets note): the INR 75,000 standard deduction is folded in by shifting EVERY slab and surcharge boundary, not just the first. https://www.incometax.gov.in/iec/foportal/help/individual/return-applicable-1", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "GHMC property tax ~0.25% effective", "last_updated": "2026-05-26", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "GST (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Numbeo + NoBroker Hyderabad 2026 (HITEC City corridor)", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "MagicBricks listing-count CDF, 2026-08-18 (SERVER_PRELOADED_STATE_ resultCount; card price = monthly rent, deposit separate as bookingAmtExact, cap inclusive on the rent field re-confirmed on this city at cap 25,000; apartments only; asking INR/month): 2 BHK P25/med/P75 = 17,095/25,893/42,250 (n=1,846) -> 0.660/1.632; 1 BHK 0.581/1.327 (n=542).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW on both edges at both slices, with the widest upper-edge miss measured in any band run to date (2 BHK +0.392). Applied 0.66/1.63 (nearest; the guard test carries a documented per-city p75 exception - never a loosened global bound). The CDF shows a distinct luxury shelf (~16% of the pool between 45-60k INR - the Gachibowli/HITEC City premium belt): a bimodal market a symmetric band cannot fully represent, same shape class as Mumbai. Informal-tail truncation means the true spread is wider still - conservative. 2 BHK anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-india.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (E). NOTE the stored medians track the portal medians closely here (28,000 vs 25,893; 18,000 vs 17,786) - recorded, no level flag needed." }, "healthcare._summary": { "source": "Private insurance (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Private daycare; no public subsidy for tech families", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Hyderabad 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Hyderabad Metro Deluxe pass", "last_updated": "2026-05-26", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo Hyderabad", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "HYD->US economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No statutory child allowance (national, inherited; 0)", "last_updated": "2026-05-26", "confidence": "high" }, "non_cash._summary": { "source": "Indian labour law (national, inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Hyderabad: INR 3,084,259, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "INR 3,084,259 / 95.48 = USD 32,303; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1292, R5-rounded down to 0.12. REPLACES 0.19 (2026-05-26), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.19, major tech hub (Microsoft/Google/Amazon campuses), cheaper than Mumbai." }, "tax.filing_basis": { "source": "Income-tax Act - individual assessee; no joint/spousal returns", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "tax.surcharge_brackets": { "source": "IN-1: new-regime high-earner surcharge on the income tax; 0 (<50L) / 10% (50L-1cr) / 15% (1cr-2cr) / 25% (>2cr, the new-regime cap), applied as a step on total income with marginal relief; 4% cess then runs on tax+surcharge. PwC IN taxes-on-personal-income; Taxmann surcharge/cess.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.social_security_cap": { "source": "EPFO International Worker framework (EPF Scheme 1952 para 83, 2008 amendment, carried forward into the EPF Scheme, 2026 notified 1 July 2026 under the Code on Social Security), described identically across AscentHR, cgaa.org, gconnect.in's EPFO brochure summary and taxaaram.com; SSA list per Dezan Shira / ExpatOrbit, UK signing 10 Feb 2026 per KPMG Flash Alert; US absence per india-briefing.com (\"pursuing\"). epfindia.gov.in itself was unreachable on two separate attempts (DNS, then connection timeout on re-verification 2026-08-06); the gap is network access from our environment, not thin sourcing. docs/superpowers/research/2026-08-06-forced-savings-foreigner-rules.md", "last_updated": "2026-08-06", "confidence": "medium-high", "note": "The INR 15,000/mo wage ceiling (cap INR 180,000/yr, max employee contribution INR 21,600) is the DOMESTIC basis. A foreign-passport International Worker at an EPF-covered establishment contributes 12% on FULL salary with no ceiling, unless a home-country SSA plus Certificate of Coverage exempts them. Engine seam: computeIndiaTaxes reads state.in_epf_intl_worker, absent or false = this capped basis, true = uncapped 12%. The target side of a comparison defaults to the uncapped IW basis when the SOURCE country holds no SSA with India (INDIA_SSA_COUNTRIES in app.js; the US has none, the UK signed Feb 2026); SSA-source movers keep this capped basis rather than the full COC exemption, which is paperwork-gated (Chile Law 18.156 precedent, conservative rule). At INR 5,000,000 gross the two bases are 21,600 vs 600,000 of employee EPF. Confidence medium-high: the mechanism is uniform across independent sources but the EPFO primary text was not reached; a Karnataka HC challenge to the uncapped rule exists and EPFO has stated the framework stands. TODO (2026-08-06): the EPF Scheme, 2026 applies the INR 15,000 ceiling to international workers in some provisions while others still read \"total wages\"; EPFO called this a typographical and drafting inconsistency, said no proposal exists to change the international-worker framework, and is weighing a corrigendum (Business Standard, 16 July 2026). Re-check this cap once the corrigendum issues. Re-checked 2026-08-07: the corrigendum has still NOT issued (latest reporting remains Business Standard 2026-07-16; EPFO maintains the IW framework is carried forward unchanged, so the uncapped IW default stands), and the DOMESTIC INR 15,000/mo ceiling is also confirmed still in force (the proposed rise to 21k is delayed indefinitely; the SC four-month directive of Jan 2026 has produced no enacted decision). The watch now rides docs/data-refresh-calendar.md rather than this dated note." }, "non_cash.parental_leave_weeks": { "source": "Maternity Benefit Act 1961 s.5(3) as amended 2017: 26 weeks at 100% of average daily wage, employer-paid; Act text via state-government mirror (wb.gov.in PDF; central gov.in domains unreachable at check, flagged)", "last_updated": "2026-08-06", "confidence": "high", "note": "26 = maternity only, and on the household-inclusive basis this IS the household total: India has no statutory paid paternity or parental leave for private-sector employees as of 2026 (CCS Rule 43-A covers government employees only; the 2017 Paternity Benefit Bill lapsed). Caveats: establishments with 10+ employees; 26 weeks for the first two children, 12 for the third and beyond (we store the first-child case). Confirmed on the basis 2026-08-06. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "Maternity Benefit Act 1961 s.5(3) as amended 2017: 26 weeks at 100% of average daily wage, employer-paid; Act text via state-government mirror (wb.gov.in PDF; central gov.in domains unreachable at check, flagged)", "last_updated": "2026-08-06", "confidence": "high", "note": "100% of the average daily wage (3-month average), employer-paid, no monetary cap. Confirmed 2026-08-06." }, "non_cash.vacation_days_statutory": { "source": "Tamil Nadu Shops and Establishments Act 1947 s.25; Delhi Shops and Establishments Act 1954; Telangana Shops and Establishments Act 1988 s.30; Karnataka Shops and Commercial Establishments Act 1961 s.15; Maharashtra Shops and Establishments Act 2017 s.18(3); Factories Act 1948 s.79", "last_updated": "2026-08-08", "confidence": "medium", "note": "CORRECTED 2026-08-08, 18 to 12. India has NO national annual-leave statute for office workers, so this nationally-enforced field has to pick one number for six cities in five states. The old 18 was sourced to Karnataka, and Karnataka's own Shops and Commercial Establishments Act 1961 s.15 does not say 18: it grants one day per twenty days worked, the same accrual formula as Factories Act 1948 s.79, whose administered output is about 15. So 18 was unsupported by the statute it cited. Verified range across the states we actually model: Tamil Nadu (Chennai) s.25 gives a flat 12, Delhi 15, Telangana (Hyderabad) s.30 a flat 15, and Karnataka (Bangalore) plus Maharashtra (Mumbai, Pune) s.18(3) both run the 1-in-20 accrual that lands near 15. We store 12, the lower end, per rule 5 and because Tamil Nadu's is a clean section-specific floor covering one of our own cities rather than an extrapolation. LIVE ALTERNATIVE: 15 is equally defensible and is the modal figure across five of the six cities; flip this if the owner prefers representativeness to conservatism. Also checked: the OSH Code 2020 s.32 is in force from 2025-11-21 with rules from May 2026, but it is scoped to factory-type establishments and does not yet displace the state Shops Acts for commercial or IT employers. No six-day or calendar conversion applies: the flat states state days directly, and the accrual denominator counts actual attendance days." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series", "last_updated": "2026-08-08", "confidence": "low", "note": "Almost all Indian health spending is out of pocket, and employer group cover is a premium that is not booked here. Two city tiers preserved. FLAGGED for a future look: claims and premium data both hint row 1 may be low for a Bangalore or Mumbai private-hospital household. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. MoveInSync spans 50 (2019), 56 (2023) and 45 (its own GCC subset, Q1 2025), an 11-minute spread inside a single vendor, which is why it is not citable." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-F.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Income Tax Department, New tax regime vs old tax regime FAQs https://www.incometax.gov.in/iec/foportal/help/new-tax-vs-old-tax-regime-faqs", "last_updated": "2026-08-11", "confidence": "high", "note": "No deduction under the regime the engine models. The NEW regime (s.115BAC) is the statutory default from AY 2024-25 and disallows Chapter-VIA deductions except 80CCD(2), 80CCH and 80JJAA, so 80C (where EPF would sit) and the extra 80CCD(1B) NPS deduction are both unavailable; 80CCD(2) is EMPLOYER NPS money, not the employee's own. computeIndiaTaxes models the new regime (slabs shifted by the INR 75,000 standard deduction) and charges EPF in payroll without removing it from the base, which is correct there. If an old-regime path is ever added, the value there is INR 200,000 and it must net off the EPF already flowing through payroll." } }, "pune": { "tax.federal_brackets": { "source": "India FY 2026-27 new regime + 4% cess, peer-identical with Bangalore. Boundaries corrected 2026-08-01 (see the Bangalore tax.federal_brackets note): the INR 75,000 standard deduction is folded in by shifting EVERY slab and surcharge boundary, not just the first. https://www.incometax.gov.in/iec/foportal/help/individual/return-applicable-1", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "PMC capital-value property tax ~0.25% effective", "last_updated": "2026-05-26", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "GST (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Numbeo + NoBroker Pune 2026 (Hinjewadi IT corridor)", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Private insurance (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Private daycare; no public subsidy for tech families", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Pune 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "PMPML pass", "last_updated": "2026-05-26", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo Pune", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "Pune->US (via BOM) economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No statutory child allowance (national, inherited; 0)", "last_updated": "2026-05-26", "confidence": "high" }, "non_cash._summary": { "source": "Indian labour law (national, inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Pune: INR 2,024,606, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "INR 2,024,606 / 95.48 = USD 21,205; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.0848, R5-rounded down to 0.08. REPLACES 0.18 (2026-05-26), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.18, major IT/auto hub." }, "tax.filing_basis": { "source": "Income-tax Act - individual assessee; no joint/spousal returns", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "tax.surcharge_brackets": { "source": "IN-1: new-regime high-earner surcharge on the income tax; 0 (<50L) / 10% (50L-1cr) / 15% (1cr-2cr) / 25% (>2cr, the new-regime cap), applied as a step on total income with marginal relief; 4% cess then runs on tax+surcharge. PwC IN taxes-on-personal-income; Taxmann surcharge/cess.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.social_security_cap": { "source": "EPFO International Worker framework (EPF Scheme 1952 para 83, 2008 amendment, carried forward into the EPF Scheme, 2026 notified 1 July 2026 under the Code on Social Security), described identically across AscentHR, cgaa.org, gconnect.in's EPFO brochure summary and taxaaram.com; SSA list per Dezan Shira / ExpatOrbit, UK signing 10 Feb 2026 per KPMG Flash Alert; US absence per india-briefing.com (\"pursuing\"). epfindia.gov.in itself was unreachable on two separate attempts (DNS, then connection timeout on re-verification 2026-08-06); the gap is network access from our environment, not thin sourcing. docs/superpowers/research/2026-08-06-forced-savings-foreigner-rules.md", "last_updated": "2026-08-06", "confidence": "medium-high", "note": "The INR 15,000/mo wage ceiling (cap INR 180,000/yr, max employee contribution INR 21,600) is the DOMESTIC basis. A foreign-passport International Worker at an EPF-covered establishment contributes 12% on FULL salary with no ceiling, unless a home-country SSA plus Certificate of Coverage exempts them. Engine seam: computeIndiaTaxes reads state.in_epf_intl_worker, absent or false = this capped basis, true = uncapped 12%. The target side of a comparison defaults to the uncapped IW basis when the SOURCE country holds no SSA with India (INDIA_SSA_COUNTRIES in app.js; the US has none, the UK signed Feb 2026); SSA-source movers keep this capped basis rather than the full COC exemption, which is paperwork-gated (Chile Law 18.156 precedent, conservative rule). At INR 5,000,000 gross the two bases are 21,600 vs 600,000 of employee EPF. Confidence medium-high: the mechanism is uniform across independent sources but the EPFO primary text was not reached; a Karnataka HC challenge to the uncapped rule exists and EPFO has stated the framework stands. TODO (2026-08-06): the EPF Scheme, 2026 applies the INR 15,000 ceiling to international workers in some provisions while others still read \"total wages\"; EPFO called this a typographical and drafting inconsistency, said no proposal exists to change the international-worker framework, and is weighing a corrigendum (Business Standard, 16 July 2026). Re-check this cap once the corrigendum issues. Re-checked 2026-08-07: the corrigendum has still NOT issued (latest reporting remains Business Standard 2026-07-16; EPFO maintains the IW framework is carried forward unchanged, so the uncapped IW default stands), and the DOMESTIC INR 15,000/mo ceiling is also confirmed still in force (the proposed rise to 21k is delayed indefinitely; the SC four-month directive of Jan 2026 has produced no enacted decision). The watch now rides docs/data-refresh-calendar.md rather than this dated note." }, "non_cash.parental_leave_weeks": { "source": "Maternity Benefit Act 1961 s.5(3) as amended 2017: 26 weeks at 100% of average daily wage, employer-paid; Act text via state-government mirror (wb.gov.in PDF; central gov.in domains unreachable at check, flagged)", "last_updated": "2026-08-06", "confidence": "high", "note": "26 = maternity only, and on the household-inclusive basis this IS the household total: India has no statutory paid paternity or parental leave for private-sector employees as of 2026 (CCS Rule 43-A covers government employees only; the 2017 Paternity Benefit Bill lapsed). Caveats: establishments with 10+ employees; 26 weeks for the first two children, 12 for the third and beyond (we store the first-child case). Confirmed on the basis 2026-08-06. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "Maternity Benefit Act 1961 s.5(3) as amended 2017: 26 weeks at 100% of average daily wage, employer-paid; Act text via state-government mirror (wb.gov.in PDF; central gov.in domains unreachable at check, flagged)", "last_updated": "2026-08-06", "confidence": "high", "note": "100% of the average daily wage (3-month average), employer-paid, no monetary cap. Confirmed 2026-08-06." }, "non_cash.vacation_days_statutory": { "source": "Tamil Nadu Shops and Establishments Act 1947 s.25; Delhi Shops and Establishments Act 1954; Telangana Shops and Establishments Act 1988 s.30; Karnataka Shops and Commercial Establishments Act 1961 s.15; Maharashtra Shops and Establishments Act 2017 s.18(3); Factories Act 1948 s.79", "last_updated": "2026-08-08", "confidence": "medium", "note": "CORRECTED 2026-08-08, 18 to 12. India has NO national annual-leave statute for office workers, so this nationally-enforced field has to pick one number for six cities in five states. The old 18 was sourced to Karnataka, and Karnataka's own Shops and Commercial Establishments Act 1961 s.15 does not say 18: it grants one day per twenty days worked, the same accrual formula as Factories Act 1948 s.79, whose administered output is about 15. So 18 was unsupported by the statute it cited. Verified range across the states we actually model: Tamil Nadu (Chennai) s.25 gives a flat 12, Delhi 15, Telangana (Hyderabad) s.30 a flat 15, and Karnataka (Bangalore) plus Maharashtra (Mumbai, Pune) s.18(3) both run the 1-in-20 accrual that lands near 15. We store 12, the lower end, per rule 5 and because Tamil Nadu's is a clean section-specific floor covering one of our own cities rather than an extrapolation. LIVE ALTERNATIVE: 15 is equally defensible and is the modal figure across five of the six cities; flip this if the owner prefers representativeness to conservatism. Also checked: the OSH Code 2020 s.32 is in force from 2025-11-21 with rules from May 2026, but it is scoped to factory-type establishments and does not yet displace the state Shops Acts for commercial or IT employers. No six-day or calendar conversion applies: the flat states state days directly, and the accrual denominator counts actual attendance days." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series", "last_updated": "2026-08-08", "confidence": "low", "note": "Almost all Indian health spending is out of pocket, and employer group cover is a premium that is not booked here. Two city tiers preserved. FLAGGED for a future look: claims and premium data both hint row 1 may be low for a Bangalore or Mumbai private-hospital household. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. MoveInSync gave 50 min in 2019 and dropped Pune from its 2023 set. The stored 40 sits 10 below even that, so it may be genuinely low." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-F.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Income Tax Department, New tax regime vs old tax regime FAQs https://www.incometax.gov.in/iec/foportal/help/new-tax-vs-old-tax-regime-faqs", "last_updated": "2026-08-11", "confidence": "high", "note": "No deduction under the regime the engine models. The NEW regime (s.115BAC) is the statutory default from AY 2024-25 and disallows Chapter-VIA deductions except 80CCD(2), 80CCH and 80JJAA, so 80C (where EPF would sit) and the extra 80CCD(1B) NPS deduction are both unavailable; 80CCD(2) is EMPLOYER NPS money, not the employee's own. computeIndiaTaxes models the new regime (slabs shifted by the INR 75,000 standard deduction) and charges EPF in payroll without removing it from the base, which is correct there. If an old-regime path is ever added, the value there is INR 200,000 and it must net off the EPF already flowing through payroll." } }, "perth": { "non_cash.parental_leave_flat": { "source": "DSS Paid Parental Leave Guide 3.1: the daily NMW amount for PLP is 7.6 times the hourly NMW; from 1 July 2026 the hourly NMW is AUD 26.44, the daily amount is AUD 200.94 and the weekly rate is AUD 1,004.70 https://guides.dss.gov.au/paid-parental-leave-guide/3/1 ; Paid Parental Leave Act 2010 ss 6, 65, 11D; Fair Work Act 2009 Part 2-6", "last_updated": "2026-08-01", "confidence": "high", "note": "AUD 1,004.70 per week (AUD 200.94 per payable day x 5), in force from 1 July 2026 and reindexed each 1 July with the Fair Work Commission national minimum wage order. PLP is taxable. Annualised that is about AUD 52,244, which against a typical Australian tech salary is roughly a 29% replacement rate and falls as pay rises. TWO ITEMS DELIBERATELY NOT IN THIS FIGURE. First, the PPL superannuation contribution for births from 1 July 2025 adds 12% of the PLP amount plus nominal interest, paid by the Commissioner of Taxation into the parent super fund, which raises the cash value but adds no weeks. Second, an income-test CLIFF: a claimant qualifies if EITHER their own adjusted taxable income is at or under AUD 186,487 OR the couple combined is at or under AUD 386,525 (2025-26 reference year, indexed each 1 July). A household on AUD 250,000 qualifies on the family test however it splits, but two earners at AUD 200,000 each fail both tests and receive NOTHING, because PLP is all-or-nothing rather than tapered. ATI includes RSUs and reportable super, which pushes tech households toward that cliff. https://guides.dss.gov.au/paid-parental-leave-guide/2/3" }, "non_cash.parental_leave_paid_pct": { "source": "DSS Paid Parental Leave Guide 3.1, Current rate of PLP https://guides.dss.gov.au/paid-parental-leave-guide/3/1", "last_updated": "2026-08-01", "confidence": "high", "note": "0 because Australia has NO percentage-paid weeks. PLP is paid at the national minimum wage as a flat daily amount, so the stored 100 was a fabricated headline rate for a benefit that never tracked salary. The flat amount lives in non_cash.parental_leave_flat and the rate is derived at the household own salary. Never display this 0 on its own." }, "non_cash.parental_leave_weeks": { "source": "DSS Paid Parental Leave Guide 2.9, Maximum PLP entitlements (Guide v1.99, released 1 July 2026) https://guides.dss.gov.au/paid-parental-leave-guide/2/9 ; Paid Parental Leave Act 2010 ss 31AB, 31ABA", "last_updated": "2026-08-01", "confidence": "high", "note": "26 wk = 130 payable days at 5 payable weekdays per PLP week. Was 22, which is the 110-day rule for births from 1 July 2024, so the repo was two annual steps stale. The statutory ladder is 100 days from 1 Jul 2023, 110 from 1 Jul 2024, 120 from 1 Jul 2025 and 130 from 1 Jul 2026, and a child born today attracts 130. COUNTED ONCE, NOT SUMMED: PLP is a single family entitlement for the child (the maximum number of days that can be received for a child for all claimants for the child), and for a partnered claimant the 20 reserved days are carved OUT of the 130 rather than added on top, so two parents claiming together still receive 130 between them. Excluded: the 12 to 24 months of Fair Work Act unpaid parental leave; reserved-period exemptions for ADF deployment or a deceased partner. Dad and Partner Pay no longer exists separately, having been absorbed into PLP in July 2023." }, "tax.filing_basis": { "source": "PwC Worldwide Tax Summaries - Australia (Individual, Tax administration): no joint assessment or joint filing in Australia; spouses assessed individually.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "FY2026-27 stage: bracket-2 15% from 2026-07-01, drops to 14% 2027-07-01 (ato.gov.au personal income tax cuts); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "City of Perth council rates (GRV) ~0.24% effective", "last_updated": "2026-05-26", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "GST 10% (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "REIWA + Numbeo Perth 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Medicare + private (inherited from Sydney baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Gross LDC fees; Child Care Subsidy normalized to AU peer scale (~9000/kid at ICP)", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Perth 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Transperth Go Anywhere Fare", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Perth", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "PER->US west coast economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "FTB phases out at tech income (national, inherited; 0)", "last_updated": "2026-05-26", "confidence": "high" }, "non_cash._summary": { "source": "Australian NES (national, inherited); public holidays vary by state", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Perth. Its basis is converted with the rest of Australia and its rank inside Australia is preserved. Prior source: Derived, not observed: no all-levels total-comp measurement exists for Perth. Basis re-derived with its country anchor Sydney; Perth's stored ratio to Sydney is preserved. Prior source: PayScale Perth SWE 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.65 x 0.5556 = 0.3611, R5-rounded down to 0.36. 0.5556 is the MEDIAN of (re-derived / previous) across the 3 Australia cities that DO carry an all-levels measurement (Brisbane 0.65 to 0.34; Melbourne 0.72 to 0.4; Sydney 0.75 to 0.45), so Perth moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Perth's rank inside Australia is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.65 was 0.8667 of Sydney's 0.75; that ratio x Sydney's re-derived 0.45 = 0.3900, R5-rounded down to 0.39. Sydney's own figure: AUD 161,818 / 1.42 = USD 113,956, / USD 250,000 = 0.4558. This is a BASIS move, not a re-measurement: Perth's rank inside Australia is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.65, mining/resources economy + tech." }, "childcare.subsidy_per_kid_annual": { "source": "servicesaustralia.gov.au Child Care Subsidy 2025-26 (90% up to 85,279; 0% at 535,279)", "last_updated": "2026-07-17", "note": "Child Care Subsidy at the solved household: CCS% = 90 - (income - 85,279)/5,000. At the median published perth household (~470k -> 13%) applied to per-kid gross childcare. CCS is means-tested; it survives partially at these incomes (unlike FTB, which is 0). The audit's 36-48% guess assumed a lower household; the engine solves perth higher. Was a flat over-covering 8,500-10,000." }, "healthcare.statutory_employee": { "source": "ATO, Medicare levy: 2% of taxable income with no upper income limit (Medicare Levy Act 1986). ATO Medicare levy surcharge income thresholds and rates, 2026-27: singles AUD 105,000 / 123,000 / 164,000 and families AUD 210,000 / 246,000 / 328,000 at 1% / 1.25% / 1.5%, with the family threshold rising AUD 1,500 per dependent child after the first (2025-26 was 101,000 / 118,000 / 158,000 and 202,000 / 236,000 / 316,000). The base levy is assessed individually; the surcharge tests combined income for couples.", "last_updated": "2026-07-31", "confidence": "medium", "note": "DISPLAY ONLY. NOT capped, and that is the load-bearing fact for Australia: it is one of very few systems in this dataset where the health contribution never flattens. BASE MISMATCH: the levy is charged on taxable income, while computeStatutoryHealthContribution applies the rate to gross, so the displayed figure runs a few percent above the charged one. That is the conservative direction and is the same base issue flagged for every country in this rollout. CAVEAT: ato.gov.au and legislation.gov.au returned 403 to every fetch attempt this session, so the 2% rate and the 2026-27 surcharge tiers were read from indexed snippets of the live ATO pages and cross-checked against two independent professional tax references, hence medium confidence. The 2% rate itself is long-standing and uncontested. 2026-27 Medicare levy LOW-INCOME thresholds were not confirmed (historically legislated retroactively); irrelevant at ICP income." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "AIHW health expenditure 2023-24 individual out-of-pocket, AUD 1,634 per person and ~1,338 excluding dental + PBS co-payment and the family-pooled PBS Safety Net", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 RAISED from 800 on primary-source evidence, the largest upward level move outside Austria. Medicare and the PBS are national with no state benefit variation, and the previous Brisbane-versus-rest spread had no documented basis in _meta or the city JSONs, so it is removed. Private hospital cover is a PREMIUM and stays in the premium field. Children add ~30% of an adult: frequently bulk-billed, lower PBS costs, and the family-pooled Safety Net pulls the whole household toward the concessional co-pay. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "AIHW health expenditure 2023-24 individual out-of-pocket, AUD 1,634 per person and ~1,338 excluding dental + PBS co-payment and the family-pooled PBS Safety Net", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "HILDA Survey (Melbourne Institute), average daily commuting time by capital city, 2017 wave. https://theconversation.com/australian-city-workers-average-commute-has-blown-out-to-66-minutes-a-day-how-does-yours-compare-120598", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average daily commuting time (round trip)", "year": "2017", "note": "CONVERSION: 59 min/day round trip / 2 = 29.5 one way. The previous 38 made Perth Australia's longest commute, ahead of Sydney, which no source supports." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "ATO, Contributions caps https://www.ato.gov.au/tax-rates-and-codes/key-superannuation-rates-and-thresholds/contributions-caps ; Maximum contributions base https://www.ato.gov.au/businesses-and-organisations/super-for-employers/paying-super-on-payday/what-payments-are-qualifying-earnings/maximum-contributions-base ; Super guarantee https://www.ato.gov.au/tax-rates-and-codes/key-superannuation-rates-and-thresholds/super-guarantee", "last_updated": "2026-08-11", "confidence": "high", "note": "Concessional contributions cap AUD 32,500 for 2026-27, a CONTRIBUTION ceiling shared with compulsory employer Super Guarantee, so the employee has only what SG leaves: 12% of gross up to the AUD 270,830 maximum contribution base (annual from 1 July 2026 under Payday Super). Room is zero at and above 270,830. The 5-year carry-forward is deliberately ignored. SEPARATE, UNFIXED: concessional contributions are taxed 15% inside the fund (30% under Division 293 above AUD 250,000), so treating the allowed portion as fully pre-tax still flatters Australia." } }, "adelaide": { "non_cash.parental_leave_flat": { "source": "DSS Paid Parental Leave Guide 3.1: the daily NMW amount for PLP is 7.6 times the hourly NMW; from 1 July 2026 the hourly NMW is AUD 26.44, the daily amount is AUD 200.94 and the weekly rate is AUD 1,004.70 https://guides.dss.gov.au/paid-parental-leave-guide/3/1 ; Paid Parental Leave Act 2010 ss 6, 65, 11D; Fair Work Act 2009 Part 2-6", "last_updated": "2026-08-01", "confidence": "high", "note": "AUD 1,004.70 per week (AUD 200.94 per payable day x 5), in force from 1 July 2026 and reindexed each 1 July with the Fair Work Commission national minimum wage order. PLP is taxable. Annualised that is about AUD 52,244, which against a typical Australian tech salary is roughly a 29% replacement rate and falls as pay rises. TWO ITEMS DELIBERATELY NOT IN THIS FIGURE. First, the PPL superannuation contribution for births from 1 July 2025 adds 12% of the PLP amount plus nominal interest, paid by the Commissioner of Taxation into the parent super fund, which raises the cash value but adds no weeks. Second, an income-test CLIFF: a claimant qualifies if EITHER their own adjusted taxable income is at or under AUD 186,487 OR the couple combined is at or under AUD 386,525 (2025-26 reference year, indexed each 1 July). A household on AUD 250,000 qualifies on the family test however it splits, but two earners at AUD 200,000 each fail both tests and receive NOTHING, because PLP is all-or-nothing rather than tapered. ATI includes RSUs and reportable super, which pushes tech households toward that cliff. https://guides.dss.gov.au/paid-parental-leave-guide/2/3" }, "non_cash.parental_leave_paid_pct": { "source": "DSS Paid Parental Leave Guide 3.1, Current rate of PLP https://guides.dss.gov.au/paid-parental-leave-guide/3/1", "last_updated": "2026-08-01", "confidence": "high", "note": "0 because Australia has NO percentage-paid weeks. PLP is paid at the national minimum wage as a flat daily amount, so the stored 100 was a fabricated headline rate for a benefit that never tracked salary. The flat amount lives in non_cash.parental_leave_flat and the rate is derived at the household own salary. Never display this 0 on its own." }, "non_cash.parental_leave_weeks": { "source": "DSS Paid Parental Leave Guide 2.9, Maximum PLP entitlements (Guide v1.99, released 1 July 2026) https://guides.dss.gov.au/paid-parental-leave-guide/2/9 ; Paid Parental Leave Act 2010 ss 31AB, 31ABA", "last_updated": "2026-08-01", "confidence": "high", "note": "26 wk = 130 payable days at 5 payable weekdays per PLP week. Was 22, which is the 110-day rule for births from 1 July 2024, so the repo was two annual steps stale. The statutory ladder is 100 days from 1 Jul 2023, 110 from 1 Jul 2024, 120 from 1 Jul 2025 and 130 from 1 Jul 2026, and a child born today attracts 130. COUNTED ONCE, NOT SUMMED: PLP is a single family entitlement for the child (the maximum number of days that can be received for a child for all claimants for the child), and for a partnered claimant the 20 reserved days are carved OUT of the 130 rather than added on top, so two parents claiming together still receive 130 between them. Excluded: the 12 to 24 months of Fair Work Act unpaid parental leave; reserved-period exemptions for ADF deployment or a deceased partner. Dad and Partner Pay no longer exists separately, having been absorbed into PLP in July 2023." }, "tax.filing_basis": { "source": "PwC Worldwide Tax Summaries - Australia (Individual, Tax administration): no joint assessment or joint filing in Australia; spouses assessed individually.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "FY2026-27 stage: bracket-2 15% from 2026-07-01, drops to 14% 2027-07-01 (ato.gov.au personal income tax cuts); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "City of Adelaide council rates ~0.30% effective", "last_updated": "2026-05-26", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "GST 10% (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "REIWA-equiv + Numbeo Adelaide 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Medicare + private (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Gross LDC fees; CCS normalized to AU peer scale (~8500/kid at ICP)", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Adelaide 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Adelaide Metro pass", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Adelaide", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "ADL->US west coast economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "FTB phases out at tech income (national, inherited; 0)", "last_updated": "2026-05-26", "confidence": "high" }, "non_cash._summary": { "source": "Australian NES (national, inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Adelaide. Its basis is converted with the rest of Australia and its rank inside Australia is preserved. Prior source: Derived, not observed: no all-levels total-comp measurement exists for Adelaide. Basis re-derived with its country anchor Sydney; Adelaide's stored ratio to Sydney is preserved. Prior source: PayScale Adelaide SWE 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.6 x 0.5556 = 0.3333, R5-rounded down to 0.33. 0.5556 is the MEDIAN of (re-derived / previous) across the 3 Australia cities that DO carry an all-levels measurement (Brisbane 0.65 to 0.34; Melbourne 0.72 to 0.4; Sydney 0.75 to 0.45), so Adelaide moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Adelaide's rank inside Australia is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.6 was 0.8000 of Sydney's 0.75; that ratio x Sydney's re-derived 0.45 = 0.3600, R5-rounded down to 0.36. Sydney's own figure: AUD 161,818 / 1.42 = USD 113,956, / USD 250,000 = 0.4558. This is a BASIS move, not a re-measurement: Adelaide's rank inside Australia is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.60, smaller market + lower cost." }, "childcare.subsidy_per_kid_annual": { "source": "servicesaustralia.gov.au Child Care Subsidy 2025-26 (90% up to 85,279; 0% at 535,279)", "last_updated": "2026-07-17", "note": "Child Care Subsidy at the solved household: CCS% = 90 - (income - 85,279)/5,000. At the median published adelaide household (~427k -> 22%) applied to per-kid gross childcare. CCS is means-tested; it survives partially at these incomes (unlike FTB, which is 0). The audit's 36-48% guess assumed a lower household; the engine solves adelaide higher. Was a flat over-covering 8,500-10,000." }, "healthcare.statutory_employee": { "source": "ATO, Medicare levy: 2% of taxable income with no upper income limit (Medicare Levy Act 1986). ATO Medicare levy surcharge income thresholds and rates, 2026-27: singles AUD 105,000 / 123,000 / 164,000 and families AUD 210,000 / 246,000 / 328,000 at 1% / 1.25% / 1.5%, with the family threshold rising AUD 1,500 per dependent child after the first (2025-26 was 101,000 / 118,000 / 158,000 and 202,000 / 236,000 / 316,000). The base levy is assessed individually; the surcharge tests combined income for couples.", "last_updated": "2026-07-31", "confidence": "medium", "note": "DISPLAY ONLY. NOT capped, and that is the load-bearing fact for Australia: it is one of very few systems in this dataset where the health contribution never flattens. BASE MISMATCH: the levy is charged on taxable income, while computeStatutoryHealthContribution applies the rate to gross, so the displayed figure runs a few percent above the charged one. That is the conservative direction and is the same base issue flagged for every country in this rollout. CAVEAT: ato.gov.au and legislation.gov.au returned 403 to every fetch attempt this session, so the 2% rate and the 2026-27 surcharge tiers were read from indexed snippets of the live ATO pages and cross-checked against two independent professional tax references, hence medium confidence. The 2% rate itself is long-standing and uncontested. 2026-27 Medicare levy LOW-INCOME thresholds were not confirmed (historically legislated retroactively); irrelevant at ICP income." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "AIHW health expenditure 2023-24 individual out-of-pocket, AUD 1,634 per person and ~1,338 excluding dental + PBS co-payment and the family-pooled PBS Safety Net", "last_updated": "2026-08-08", "confidence": "medium", "note": "Row 1 RAISED from 800 on primary-source evidence, the largest upward level move outside Austria. Medicare and the PBS are national with no state benefit variation, and the previous Brisbane-versus-rest spread had no documented basis in _meta or the city JSONs, so it is removed. Private hospital cover is a PREMIUM and stays in the premium field. Children add ~30% of an adult: frequently bulk-billed, lower PBS costs, and the family-pooled Safety Net pulls the whole household toward the concessional co-pay. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "AIHW health expenditure 2023-24 individual out-of-pocket, AUD 1,634 per person and ~1,338 excluding dental + PBS co-payment and the family-pooled PBS Safety Net", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "HILDA Survey (Melbourne Institute), average daily commuting time by capital city, 2017 wave, reported in The Conversation 2019-07-29. https://theconversation.com/australian-city-workers-average-commute-has-blown-out-to-66-minutes-a-day-how-does-yours-compare-120598", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average daily commuting time (round trip)", "year": "2017", "note": "CONVERSION: published 56 min/day is a ROUND TRIP by HILDA's own construction ('time spent travelling to and from work in a typical week divided by the usual number of days worked'), so 56/2 = 28.0 one way. The ABS publishes no commute time at all, only method of travel, so every Australian figure in circulation traces to HILDA. The newer 2023 wave was deliberately NOT used: it includes home workers at zero commute, which is the wrong basis for this field." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "ATO, Contributions caps https://www.ato.gov.au/tax-rates-and-codes/key-superannuation-rates-and-thresholds/contributions-caps ; Maximum contributions base https://www.ato.gov.au/businesses-and-organisations/super-for-employers/paying-super-on-payday/what-payments-are-qualifying-earnings/maximum-contributions-base ; Super guarantee https://www.ato.gov.au/tax-rates-and-codes/key-superannuation-rates-and-thresholds/super-guarantee", "last_updated": "2026-08-11", "confidence": "high", "note": "Concessional contributions cap AUD 32,500 for 2026-27, a CONTRIBUTION ceiling shared with compulsory employer Super Guarantee, so the employee has only what SG leaves: 12% of gross up to the AUD 270,830 maximum contribution base (annual from 1 July 2026 under Payday Super). Room is zero at and above 270,830. The 5-year carry-forward is deliberately ignored. SEPARATE, UNFIXED: concessional contributions are taxed 15% inside the fund (30% under Division 293 above AUD 250,000), so treating the allowed portion as fully pre-tax still flatters Australia." } }, "busan": { "non_cash.parental_leave_paid_pct": { "source": "Geunrogijunbeop art. 74(4) (employer pays 100% of tongsang imgeum for maternity days 1-60, the EI benefit only discharges the employer up to its amount) via easylaw.go.kr ; Namnyeogoyongpyeongdeungbeop art. 18-2 (paternity 20 days, same employer-pay construction)", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the 11.31-week remainder: maternity days 1-60 (8.57 wk) plus the 20-day paternity leave (2.86 wk), both employer-paid wage continuation that the EI reimbursement cannot reduce, genuinely uncapped at any salary (11.43 wk exact, clipped to the 120-week stored total). The other 108.69 weeks moved to non_cash.parental_leave_capped in the 2026-08-06 caps rebuild. Was 91, the whole-blend headline with every monetary ceiling ignored." }, "non_cash.parental_leave_weeks": { "source": "고용보험법 시행령 제95조 (육아휴직 급여) https://law.go.kr/ ; 근로기준법 제74조 (출산전후휴가 90일) ; 남녀고용평등법 제18조의2 (배우자 출산휴가 20일) and 제19조 (육아휴직 1년 per parent) ; 고용24 육아휴직급여 https://www.work24.go.kr/cm/c/f/1100/selecSystInfo.do?systClId=SC00000251&systId=SI00000402", "last_updated": "2026-08-01", "confidence": "high", "note": "120 wk = 출산전후휴가 12.9 (mother) + 배우자 출산휴가 2.9 (father) + 육아휴직 12 months x 2 parents (52.2 x 2). Was 13, which held maternity alone and dropped 육아휴직 entirely. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Conservative: 12 months per parent, not the 18 available only when BOTH parents each take 3 months or more; 20 working days counted as 2.86 wk rather than the 4 calendar weeks of actual absence; single birth, not multiples." }, "tax.federal_brackets": { "source": "South Korea 2026 income tax + 10% local surcharge: peer-identical with Seoul", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Korea jaesan-se ~0.20% effective", "last_updated": "2026-05-26", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "VAT 10% (inherited from Seoul baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Numbeo + BambooRoutes Busan 2026 (wolse monthly-rent basis)", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "National Health Insurance (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Heavily subsidized (Nuri voucher, near-free 0-5); net parent fee + after-school", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Busan 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Busan unlimited monthly pass ₩45,000", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Busan", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "PUS->US west coast economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Korea child allowance (national, inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "non_cash._summary": { "source": "Korean labour law (national, inherited); 15 leave, 15 holidays", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Busan. Its basis is converted with the rest of South Korea and its rank inside South Korea is preserved. Prior source: Levels.fyi Busan SWE 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.41 x 0.4727 = 0.1938, R5-rounded down to 0.19. 0.4727 is the MEDIAN of (re-derived / previous) across the 1 South Korea city that DO carry an all-levels measurement (Seoul 0.55 to 0.26), so Busan moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Busan's rank inside South Korea is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.41, 2nd city, below Seoul." }, "tax.filing_basis": { "source": "KR-5: no joint filing; spouses file separate returns. PwC KR tax administration: https://taxsummaries.pwc.com/republic-of-korea/individual/tax-administration", "last_updated": "2026-07-26", "confidence": "high" }, "tax.earned_income_deduction_tiers": { "source": "KR-1: earned-income deduction (근로소득공제), tiered 70/40/15/5/2% by KRW band. PwC KR deductions: https://taxsummaries.pwc.com/republic-of-korea/individual/deductions", "last_updated": "2026-07-26", "confidence": "high" }, "tax.earned_income_deduction_cap": { "source": "KR-1: earned-income deduction total capped at KRW 20M. PwC KR deductions (same URL).", "last_updated": "2026-07-26", "confidence": "high" }, "tax.personal_deduction_per_head": { "source": "KR-2: basic personal deduction KRW 1.5M per head. PwC KR deductions.", "last_updated": "2026-07-26", "confidence": "high", "note": "Only the taxpayer head is modelled; dependent (spouse/child) heads are omitted — conservative (overstates tax) and avoids double-counting kids in the individual-filing clone." }, "tax.np_rate": { "source": "KR-4: National Pension employee rate 4.75% for 2026 (pension reform: 9.0%->9.5% total from 2026-01-01). PwC KR other-taxes; koreantaxexpert 2026; Korea Herald/Lockton pension reform.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.np_cap": { "source": "KR-4: National Pension standard-income cap KRW 6.59M/mo = 79.08M/yr wage base (from Jul 2026). PwC KR other-taxes; koreantaxexpert.", "last_updated": "2026-07-26", "confidence": "high", "note": "Value rule: the higher (from-Jul-2026) cap is the less-favorable choice for the country; annual model uses the 12x monthly cap." }, "tax.nhi_ltc_rate": { "source": "KR-4: National Health Insurance incl. Long-term Care, employee ~4.0674% for 2026 (uncapped at ICP income). PwC KR other-taxes (4.0674%); koreantaxexpert (NHI 3.595% + LTC 0.4656%).", "last_updated": "2026-07-26", "confidence": "medium", "note": "The exact 2026 NHI rate is \"tentative\"; the conservative (higher, PwC) combined figure is used per the value rule." }, "tax.ei_rate": { "source": "KR-4: Employment Insurance employee 0.9% (uncapped). PwC KR other-taxes; koreantaxexpert.", "last_updated": "2026-07-26", "confidence": "high" }, "healthcare.statutory_employee": { "source": "MOHW press release '2026 NHI premium rate set at 7.19%' (mohw.go.kr board.es list_no=1487279), which states the employee and employer shares are 3.595% each; MOHW press release '2026 long-term care insurance rate 0.9448%' (list_no=1487817), which sets LTC at 13.14% of the health premium and 0.9448% of income across both parties, so 0.4724% employee (3.595% x 13.14% = 0.4724%, matching 0.9448% / 2). Ceiling: MOHW notice 2025-222 amending the monthly upper and lower health premium limits, issued 2025-12-24 and effective 2026-01-01; the monthly premium ceiling of KRW 9,183,480 implies a wage base of 9,183,480 / 0.0719 = KRW 127,725,730/month = KRW 1,532,708,760/year.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll (tax.nhi_ltc_rate 0.040674, charged uncapped by computeSouthKoreaTaxes; the display cap is far above any modeled salary so the two agree in practice). LTC is included because the regime definition covers long-term care and it is compulsory with no opt-out. CAVEAT ON THE CAP: the rates are primary from MOHW, but the KRW 9,183,480 premium ceiling reached us through five independent Korean outlets citing MOHW notice 2025-222, whose own figures sit in an unopenable attachment, and nhis.or.kr was unreachable. The derived annual wage base is therefore secondary-sourced and approximate. It does not bind at any income this calculator models, so the imprecision costs nothing." }, "childcare.subsidy_per_kid_annual": { "source": "Korean boyukryo / aihaengbok childcare voucher (Ministry of Health and Welfare); double-subtraction correction 2026-08-02", "last_updated": "2026-08-02", "confidence": "medium", "note": "Set to 0 on 2026-08-02, was KRW 6,000,000 (new meta entry; the value carried no provenance before). The old value was the childcare voucher, which the government pays direct to the centre rather than to the household. The monthly rates already carry only the residual parents actually pay, the teukbyeol-hwaldongbi extra-activity and meal charges. Busan bills KRW 50,000/month of residual and the old subsidy clamped that to zero. This field holds CASH THAT REACHES THE HOUSEHOLD. Where the monthly rate it nets against is already the price the household actually pays, the value is 0 and the relief is described here instead. See docs/superpowers/research/2026-08-02-childcare-double-subtraction.md." }, "non_cash.parental_leave_capped": { "source": "Goyongbohumbeop art. 70(4) + Enforcement Decree art. 95(1) via easylaw.go.kr (fetched 2026-08-06): yugahyujik ceilings KRW 2.5M/mo months 1-3, 2.0M months 4-6, 1.6M months 7-12 (in force since 2025-01-01, current 2026); maternity days 61-90 ceiling KRW 2.2M/mo (2026-01-01). Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (South Korea) and docs/superpowers/research/2026-08-06-pl-caps-east-asia.md", "last_updated": "2026-08-06", "confidence": "high", "note": "Multi-phase caps flattened to ONE duration-weighted record because the engine holds a single capped block: (4.29 wk x 2.2M + 26.1 x 2.5M + 26.1 x 2.0M + 52.2 x 1.6M) / 108.69 = KRW 1,935,845/month, stored rounded down to 1,935,000; pct 90.4 is the same weighting of the 100/100/100/80 phase rates. At ICP salaries every phase cap binds, so the weighted form reproduces the exact statutory total payout; between the lowest and highest cap it can overstate by up to ~4% at sub-ICP salaries, the flatten cost recorded here. Excluded per the basis convention: the 6+6 bumo-yugahyujik enhanced ceilings (conditional on both parents' behaviour). At KRW 100M the blended effective rate is ~30%, versus the 91 headline this replaces." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "NHIS coinsurance by care setting (20% inpatient, 30-60% outpatient by facility level) + the income-decile-banded annual out-of-pocket ceiling", "last_updated": "2026-08-08", "confidence": "low", "note": "LEVEL NOT RE-VERIFIED in the 2026-08-08 sweep; row 1 is held and only the household curve was added. Korea's child discount is narrow, mainly under-6 inpatient at 5%, so its child increment is one of the steepest in the Asia-Pacific batch. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "NHIS coinsurance by care setting (20% inpatient, 30-60% outpatient by facility level) + the income-decile-banded annual out-of-pocket ceiling", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Statistics Korea, Population and Housing Census 2020, commuting time by province, via KIRI. https://kiri.or.kr/PDF/weeklytrend/20211206/trend20211206_8.pdf", "last_updated": "2026-08-08", "confidence": "high", "statistic": "시도별 통근·통학 소요시간", "year": "2020", "note": "30.3 min, published one way, no conversion. Metro city proper. Same work-and-school blending caveat as Seoul. Survey self-report (census)." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "National Tax Service, pension account tax credit https://www.nts.go.kr/nts/cm/cntnts/cntntsView.do?cntntsId=7875", "last_updated": "2026-08-11", "confidence": "high", "note": "No base deduction. Pension-account relief is a 세액공제, deducted from the computed global income tax: contribution base KRW 6,000,000 (pension savings) or 9,000,000 combined with retirement-pension DC/IRP, at 15% below the income thresholds and 12% at our ICP's income. Setting 0 costs the user a real benefit worth at most about KRW 1,080,000 of tax; modeling it needs a tax-credit key, not this one. computeSouthKoreaTaxes already subtracts the full mandatory payroll, so nothing double-counts." } }, "incheon": { "non_cash.parental_leave_paid_pct": { "source": "Geunrogijunbeop art. 74(4) (employer pays 100% of tongsang imgeum for maternity days 1-60, the EI benefit only discharges the employer up to its amount) via easylaw.go.kr ; Namnyeogoyongpyeongdeungbeop art. 18-2 (paternity 20 days, same employer-pay construction)", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the 11.31-week remainder: maternity days 1-60 (8.57 wk) plus the 20-day paternity leave (2.86 wk), both employer-paid wage continuation that the EI reimbursement cannot reduce, genuinely uncapped at any salary (11.43 wk exact, clipped to the 120-week stored total). The other 108.69 weeks moved to non_cash.parental_leave_capped in the 2026-08-06 caps rebuild. Was 91, the whole-blend headline with every monetary ceiling ignored." }, "non_cash.parental_leave_weeks": { "source": "고용보험법 시행령 제95조 (육아휴직 급여) https://law.go.kr/ ; 근로기준법 제74조 (출산전후휴가 90일) ; 남녀고용평등법 제18조의2 (배우자 출산휴가 20일) and 제19조 (육아휴직 1년 per parent) ; 고용24 육아휴직급여 https://www.work24.go.kr/cm/c/f/1100/selecSystInfo.do?systClId=SC00000251&systId=SI00000402", "last_updated": "2026-08-01", "confidence": "high", "note": "120 wk = 출산전후휴가 12.9 (mother) + 배우자 출산휴가 2.9 (father) + 육아휴직 12 months x 2 parents (52.2 x 2). Was 13, which held maternity alone and dropped 육아휴직 entirely. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Conservative: 12 months per parent, not the 18 available only when BOTH parents each take 3 months or more; 20 working days counted as 2.86 wk rather than the 4 calendar weeks of actual absence; single birth, not multiples." }, "tax.federal_brackets": { "source": "South Korea 2026 income tax + 10% local surcharge: peer-identical with Seoul", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Korea jaesan-se ~0.25% effective", "last_updated": "2026-05-26", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "VAT 10% (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Numbeo + BambooRoutes Incheon 2026 (Songdo IT district)", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "National Health Insurance (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Heavily subsidized (Nuri voucher); subsidy aligned to Seoul ₩6M baseline", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Incheon 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Incheon Climate Card ₩62,000", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Incheon", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "ICN->US west coast economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Korea child allowance (national, inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "non_cash._summary": { "source": "Korean labour law (national, inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Incheon. Its basis is converted with the rest of South Korea and its rank inside South Korea is preserved. Prior source: Levels.fyi Incheon SWE 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.45 x 0.4727 = 0.2127, R5-rounded down to 0.21. 0.4727 is the MEDIAN of (re-derived / previous) across the 1 South Korea city that DO carry an all-levels measurement (Seoul 0.55 to 0.26), so Incheon moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Incheon's rank inside South Korea is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.45, Seoul Capital Area / Songdo commuter belt." }, "tax.filing_basis": { "source": "KR-5: no joint filing; spouses file separate returns. PwC KR tax administration: https://taxsummaries.pwc.com/republic-of-korea/individual/tax-administration", "last_updated": "2026-07-26", "confidence": "high" }, "tax.earned_income_deduction_tiers": { "source": "KR-1: earned-income deduction (근로소득공제), tiered 70/40/15/5/2% by KRW band. PwC KR deductions: https://taxsummaries.pwc.com/republic-of-korea/individual/deductions", "last_updated": "2026-07-26", "confidence": "high" }, "tax.earned_income_deduction_cap": { "source": "KR-1: earned-income deduction total capped at KRW 20M. PwC KR deductions (same URL).", "last_updated": "2026-07-26", "confidence": "high" }, "tax.personal_deduction_per_head": { "source": "KR-2: basic personal deduction KRW 1.5M per head. PwC KR deductions.", "last_updated": "2026-07-26", "confidence": "high", "note": "Only the taxpayer head is modelled; dependent (spouse/child) heads are omitted — conservative (overstates tax) and avoids double-counting kids in the individual-filing clone." }, "tax.np_rate": { "source": "KR-4: National Pension employee rate 4.75% for 2026 (pension reform: 9.0%->9.5% total from 2026-01-01). PwC KR other-taxes; koreantaxexpert 2026; Korea Herald/Lockton pension reform.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.np_cap": { "source": "KR-4: National Pension standard-income cap KRW 6.59M/mo = 79.08M/yr wage base (from Jul 2026). PwC KR other-taxes; koreantaxexpert.", "last_updated": "2026-07-26", "confidence": "high", "note": "Value rule: the higher (from-Jul-2026) cap is the less-favorable choice for the country; annual model uses the 12x monthly cap." }, "tax.nhi_ltc_rate": { "source": "KR-4: National Health Insurance incl. Long-term Care, employee ~4.0674% for 2026 (uncapped at ICP income). PwC KR other-taxes (4.0674%); koreantaxexpert (NHI 3.595% + LTC 0.4656%).", "last_updated": "2026-07-26", "confidence": "medium", "note": "The exact 2026 NHI rate is \"tentative\"; the conservative (higher, PwC) combined figure is used per the value rule." }, "tax.ei_rate": { "source": "KR-4: Employment Insurance employee 0.9% (uncapped). PwC KR other-taxes; koreantaxexpert.", "last_updated": "2026-07-26", "confidence": "high" }, "healthcare.statutory_employee": { "source": "MOHW press release '2026 NHI premium rate set at 7.19%' (mohw.go.kr board.es list_no=1487279), which states the employee and employer shares are 3.595% each; MOHW press release '2026 long-term care insurance rate 0.9448%' (list_no=1487817), which sets LTC at 13.14% of the health premium and 0.9448% of income across both parties, so 0.4724% employee (3.595% x 13.14% = 0.4724%, matching 0.9448% / 2). Ceiling: MOHW notice 2025-222 amending the monthly upper and lower health premium limits, issued 2025-12-24 and effective 2026-01-01; the monthly premium ceiling of KRW 9,183,480 implies a wage base of 9,183,480 / 0.0719 = KRW 127,725,730/month = KRW 1,532,708,760/year.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll (tax.nhi_ltc_rate 0.040674, charged uncapped by computeSouthKoreaTaxes; the display cap is far above any modeled salary so the two agree in practice). LTC is included because the regime definition covers long-term care and it is compulsory with no opt-out. CAVEAT ON THE CAP: the rates are primary from MOHW, but the KRW 9,183,480 premium ceiling reached us through five independent Korean outlets citing MOHW notice 2025-222, whose own figures sit in an unopenable attachment, and nhis.or.kr was unreachable. The derived annual wage base is therefore secondary-sourced and approximate. It does not bind at any income this calculator models, so the imprecision costs nothing." }, "childcare.subsidy_per_kid_annual": { "source": "Korean boyukryo / aihaengbok childcare voucher (Ministry of Health and Welfare); double-subtraction correction 2026-08-02", "last_updated": "2026-08-02", "confidence": "medium", "note": "Set to 0 on 2026-08-02, was KRW 6,000,000 (new meta entry; the value carried no provenance before). The old value was the childcare voucher, which the government pays direct to the centre rather than to the household. The monthly rates already carry only the residual parents actually pay, the teukbyeol-hwaldongbi extra-activity and meal charges. Incheon bills KRW 30,000/month of residual and the old subsidy clamped that to zero. This field holds CASH THAT REACHES THE HOUSEHOLD. Where the monthly rate it nets against is already the price the household actually pays, the value is 0 and the relief is described here instead. See docs/superpowers/research/2026-08-02-childcare-double-subtraction.md." }, "non_cash.parental_leave_capped": { "source": "Goyongbohumbeop art. 70(4) + Enforcement Decree art. 95(1) via easylaw.go.kr (fetched 2026-08-06): yugahyujik ceilings KRW 2.5M/mo months 1-3, 2.0M months 4-6, 1.6M months 7-12 (in force since 2025-01-01, current 2026); maternity days 61-90 ceiling KRW 2.2M/mo (2026-01-01). Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (South Korea) and docs/superpowers/research/2026-08-06-pl-caps-east-asia.md", "last_updated": "2026-08-06", "confidence": "high", "note": "Multi-phase caps flattened to ONE duration-weighted record because the engine holds a single capped block: (4.29 wk x 2.2M + 26.1 x 2.5M + 26.1 x 2.0M + 52.2 x 1.6M) / 108.69 = KRW 1,935,845/month, stored rounded down to 1,935,000; pct 90.4 is the same weighting of the 100/100/100/80 phase rates. At ICP salaries every phase cap binds, so the weighted form reproduces the exact statutory total payout; between the lowest and highest cap it can overstate by up to ~4% at sub-ICP salaries, the flatten cost recorded here. Excluded per the basis convention: the 6+6 bumo-yugahyujik enhanced ceilings (conditional on both parents' behaviour). At KRW 100M the blended effective rate is ~30%, versus the 91 headline this replaces." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "NHIS coinsurance by care setting (20% inpatient, 30-60% outpatient by facility level) + the income-decile-banded annual out-of-pocket ceiling", "last_updated": "2026-08-08", "confidence": "low", "note": "LEVEL NOT RE-VERIFIED in the 2026-08-08 sweep; row 1 is held and only the household curve was added. Korea's child discount is narrow, mainly under-6 inpatient at 5%, so its child increment is one of the steepest in the Asia-Pacific batch. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "NHIS coinsurance by care setting (20% inpatient, 30-60% outpatient by facility level) + the income-decile-banded annual out-of-pocket ceiling", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Statistics Korea, Population and Housing Census 2020, commuting time by province, via KIRI. https://kiri.or.kr/PDF/weeklytrend/20211206/trend20211206_8.pdf", "last_updated": "2026-08-08", "confidence": "high", "statistic": "시도별 통근·통학 소요시간", "year": "2020", "note": "35.0 min, published one way, no conversion. Metro city proper. Same work-and-school blending caveat as Seoul. Survey self-report (census)." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "National Tax Service, pension account tax credit https://www.nts.go.kr/nts/cm/cntnts/cntntsView.do?cntntsId=7875", "last_updated": "2026-08-11", "confidence": "high", "note": "No base deduction. Pension-account relief is a 세액공제, deducted from the computed global income tax: contribution base KRW 6,000,000 (pension savings) or 9,000,000 combined with retirement-pension DC/IRP, at 15% below the income thresholds and 12% at our ICP's income. Setting 0 costs the user a real benefit worth at most about KRW 1,080,000 of tax; modeling it needs a tax-credit key, not this one. computeSouthKoreaTaxes already subtracts the full mandatory payroll, so nothing double-counts." } }, "seville": { "childcare.free_tier_start_age": { "last_updated": "2026-08-01", "confidence": "high", "source": "Ley Organica 2/2006 de Educacion (LOE), consolidated text, art. 15.2: \"El segundo ciclo de la educacion infantil sera gratuito\"; art. 14.1: \"El primero comprende hasta los tres anos, y el segundo, desde los tres a los seis anos de edad\" https://www.boe.es/buscar/doc.php?id=BOE-A-2006-7899 . Enrolment: Ministerio de Educacion FP y Deportes, 2025-09-09, \"La escolarizacion en el segundo ciclo de educacion infantil (3 a 5 anos) es en Espana practicamente universal, con un 98% de los menores matriculados\" https://www.educacionfpydeportes.gob.es/prensa/actualidad/2025/09/20250909-panoramaeducacion2025.html", "note": "Segundo ciclo (3-6) is free by statute in public and concertado schools and 98% enrolled; primer ciclo (0-3) is neither free nor universal. Before this, ages 1-5 all billed one rate, so the bucket straddled both: a 1-2 year old in paid guarderia was billed the free-tier meal fee. Entry to P3 is by BIRTH-YEAR COHORT (children born in year N start the September of year N+3), so a December-born child starts free schooling at 2 years 9 months; whole-number ages make 3 the correct modelling boundary." }, "non_cash.parental_leave_paid_pct": { "source": "Real Decreto-ley 9/2025 preamble (cien por cien replacement) https://www.boe.es/eli/es/rdl/2025/07/29/9 ; LGSS art. 179 (base reguladora) and art. 148 (tope maximo de la base de cotizacion)", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 38 weeks into non_cash.parental_leave_capped (100% of a base capped at the base maxima de cotizacion), so no percentage-paid weeks remain on this field. Was 100 with the cap ignored per the old field convention. Watch item unchanged: the RDL preamble records a government commitment to extend to 20 weeks per parent, which is not law and is not stored." }, "non_cash.parental_leave_weeks": { "source": "Estatuto de los Trabajadores art. 48.4 as amended by Real Decreto-ley 9/2025 de 29 de julio, in force 31/07/2025 https://www.boe.es/buscar/act.php?id=BOE-A-2025-15741 ; consolidated ET https://www.boe.es/buscar/act.php?id=BOE-A-2015-11430", "last_updated": "2026-08-01", "confidence": "high", "note": "38 wk = 19 per parent x 2 (6 obligatory post-partum + 11 usable to age 12 months + 2 usable to age 8). ET art. 48.4 states este derecho es individual de la persona trabajadora sin que pueda transferirse su ejercicio al otro progenitor. Spain has no separate maternity leave: both parents hold the identical permiso por nacimiento y cuidado. Was 16, one parent on the pre-2025 duration. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Excluded: the 8-week permiso parental of art. 48 bis, which RDL 9/2025 confirms remains unpaid; the 32-week monoparental variant; multiple-birth uplifts." }, "tax.inbound_regime": { "source": "Beckham Law, art. 93.2.e) LIRPF (Ley 35/2006, as amended by Ley 28/2022 Ley de Startups and Ley 7/2024): marginal two-band scale on Spanish employment income, 24% up to EUR 600,000 and 47% on the excess. Duration = the tax period of the move plus the five following (6 years); 5 prior tax periods of non-residency; election on modelo 149. AEAT Manual practico IRPF: https://sede.agenciatributaria.gob.es/Sede/ayuda/manuales-videos-folletos/manuales-practicos/irpf-2024/c15-calculo-impuesto-determinacion-cuotas-integras/regimen-fiscal-especial-aplicable-trabajadores.html ; AEAT Regimen especial impatriados art. 93: https://sede.agenciatributaria.gob.es/Sede/ayuda/manuales-videos-folletos/manuales-practicos/manual-tributacion-no-residentes/regimenes-opcionales/regimen-especial-impatriados.html", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31 OVER-GRANT FIX. The block declared flat_rate 0.24 with no max_income and no flat_rate_above, so 24% ran to infinity even though this very entry described the 47% band. Wired max_income 600000 + flat_rate_above 0.47; the engine blends them into one effective rate. Barcelona single filer at EUR 1,000,000: tax 243,887 -> 335,887. Nothing at or below EUR 600,000 moves." }, "tax.federal_brackets": { "source": "Art. 63.1 LIRPF, escala estatal 2026 (via guiafiscal.es/irpf); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.state_brackets": { "source": "Junta de Andalucía autonómica IRPF scale 2025 (9.5-22.5%)", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "IBI Sevilla 0.6758% of cadastral ~0.40% effective", "last_updated": "2026-05-26", "confidence": "medium" }, "housing._summary": { "source": "Numbeo + Investropa Seville 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "SNS universal + private (inherited from Madrid baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Private guardería 0-3; 3-6 free public; means-tested cash benefit 0 at tech income. Junta de Andalucía escolarización 0-3 (juntadeandalucia.es): age-2 tuition made universal-free for 2025-26 (age <1 free only below 0.5-0.75 IPREM per cápita, not ICP-relevant), aula matinal/aula de tarde extras still renta-scaled and non-zero for high earners", "last_updated": "2026-07-17", "confidence": "low", "note": "preschool_monthly raised from EUR 0 to a rough EUR 180/mo blend — the old 0 assumed the whole age 1-5 bucket was free, but only age 2 base tuition is (extras aren't, and ages 1/3/4/5 aren't). subsidy_per_kid_annual correctly stays 0 (no cash beca in this region, it's a fee waiver not representable as a per-kid check). Precision here is a follow-up item, not independently ground-truthed" }, "food._summary": { "source": "Numbeo Seville 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "TUSSAM monthly pass ~EUR 35", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Seville", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "SVQ/MAD->US economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Means-tested, 0 at tech income (national, inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "non_cash._summary": { "source": "Spanish labour law (national, inherited); 22 vacation, ~14 holidays", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Seville: EUR 33,714, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "EUR 33,714 / 0.867 = USD 38,886; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1555, R5-rounded down to 0.15. REPLACES 0.33 (2026-08-11), and the interim 0.19 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Senior/all-levels 1.84, the widest in the Spanish set. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.33 was 0.8250 of Madrid's 0.4; that ratio x Madrid's re-derived 0.24 = 0.1980, R5-rounded down to 0.19. Madrid's own figure: EUR 53,912 / 0.867 = USD 62,182, / USD 250,000 = 0.2487. This is a BASIS move, not a re-measurement: Seville's rank inside Spain is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.33, Andalusia, below Madrid." }, "tax.social_security_cap": { "confidence": "high", "source": "Seguridad Social base maxima 2026 EUR 5,101.20/mo (taxsummaries.pwc.com/spain); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "tax.filing_basis": { "source": "Agencia Tributaria - individual IRPF (the engine files spouses individually; the elective declaracion conjunta is excluded)", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "childcare.preschool_monthly": { "last_updated": "2026-08-01", "confidence": "high", "source": "Junta de Andalucia, precios de los servicios complementarios, curso 2025-2026 (one regional price for all Andalusian public schools): comedor escolar EUR 5,54/dia; aula matinal \"Precio mensual: 23,91 euros. Precio por dia: 1,82 euros\"; actividades extraescolares EUR 17,51/mes por actividad. Published via BOJA and reproduced on juntadeandalucia.es school pages for curso 2025-2026.", "note": "EUR 5.54/day x 15 = EUR 83.10 comedor, plus aula matinal EUR 23.91/month over 10 months = EUR 19.93, total EUR 103.03. Was 160. Andalucia is the only region in the set that sets aula matinal centrally, which is why this leg is high confidence here and medium elsewhere. Extraescolares (EUR 17.51/month per activity) are excluded as elective. Reduced comedor rates exist only for beca recipients, which a EUR 140k household is not. Free tuition, so the modelled cost is the real residual a dual-earner family pays: comedor escolar plus aula matinal. Stored value is the ANNUAL total divided by 12 (computeChildcareCost multiplies by 12), NOT the school-year invoice, which runs ~10 months. Comedor annualised as daily rate x 180 dias lectivos / 12 = daily x 15; 180 is the conservative end of the 175-181 range (independently implied day counts: ~177 Catalonia, ~181 Andalucia). Applies the owner decision of 2026-07-18 that Spain follows the statutory free convention, now that the age boundary makes it applicable to the free ages only. Sourcing: docs/superpowers/research/2026-08-01-childcare-free-tier-boundary.md and 2026-08-01-spain-wraparound-fees.md." }, "non_cash.parental_leave_capped": { "source": "Orden PJC/297/2026, de 30 de marzo, art. 2.1: tope maximo de la base de cotizacion 5.101,20 euros mensuales desde el 1 de enero de 2026 https://www.boe.es/diario_boe/txt.php?id=BOE-A-2026-7296 ; LGSS art. 179 (the base reguladora is the contingencias comunes contribution base) and art. 148 (that base is capped at the tope maximo)", "last_updated": "2026-08-06", "confidence": "high", "note": "All 38 weeks pay 100% of a base capped at EUR 5,101.20/month (EUR 61,214.40/yr), one clean segment with no phase split: above that gross the benefit is the cap, not the salary. At EUR 90,000 the effective rate is ~68%, at EUR 150,000 ~41%, versus the naked 100% this replaces. cap_amount doubles as the payout ceiling because the rate is 100. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Spain)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "RD 16/2012 pharmaceutical co-payment tramos, with our ICP household in the uncapped 50-60% band + Ministerio de Sanidad SCS household out-of-pocket with dental at ~32%", "last_updated": "2026-08-08", "confidence": "medium", "note": "Spain is the European exception on the child rule: a dependent minor registered on a working parent's card pays the SAME co-payment percentage as the parent, so the sub-linearity here is a volume argument (free well-child checks and the vaccination schedule never touch the pharmacy co-pay), not a statutory discount. Spain has NO statutory annual ceiling for the household we model; the monthly caps introduced in 2026 apply only below EUR 35,000 of income. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "RD 16/2012 pharmaceutical co-payment tramos, with our ICP household in the uncapped 50-60% band + Ministerio de Sanidad SCS household out-of-pocket with dental at ~32%", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "INE Spain, Urban Audit indicator 6.4, Area Urbana Funcional de Sevilla. https://www.ine.es/jaxiT3/Tabla.htm?t=69327", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Duracion media del desplazamiento al trabajo (solo ida)", "year": "2011", "note": "AUF value 23.2. Largest Spanish correction; the previous 33 exceeded even Madrid." }, "non_cash.sick_leave": { "source": "Ley General de la Seguridad Social, current consolidated text Real Decreto; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Agencia Tributaria, Manual Renta 2025 s.8.2.2.6 https://sede.agenciatributaria.gob.es/Sede/ayuda/manuales-videos-folletos/manuales-ayuda-presentacion/irpf-2025/8-cumplimentacion-irpf/8_2-base-liquidable-general-base-ahorro/8_2_2-reducciones-aportaciones-prevision-social/8_2_2_6-aportaciones-anuales-maximas-limite-reduccion.html", "last_updated": "2026-08-11", "confidence": "high", "note": "Reduccion de la base imponible general for aportaciones a sistemas de prevision social (Ley 35/2006 arts. 51-52), EUR 1,500/yr. A base deduction, not a credit, and individual. NOT 8,500: the +8,500 increment requires EMPLOYER contributions to a plan de empleo, which we do not model, and worker contributions only ride along under a matching coefficient table. pct_of_earnings deliberately omitted: the outer 30%-of-net-earned-income bound never binds at EUR 75-95k and 0.30 of GROSS would be both inert and wrong-based. SEPARATE pre-existing understatement, out of scope: computeSpainTaxes does not treat cotizaciones as a gasto deducible." } }, "malaga": { "childcare.free_tier_start_age": { "last_updated": "2026-08-01", "confidence": "high", "source": "Ley Organica 2/2006 de Educacion (LOE), consolidated text, art. 15.2: \"El segundo ciclo de la educacion infantil sera gratuito\"; art. 14.1: \"El primero comprende hasta los tres anos, y el segundo, desde los tres a los seis anos de edad\" https://www.boe.es/buscar/doc.php?id=BOE-A-2006-7899 . Enrolment: Ministerio de Educacion FP y Deportes, 2025-09-09, \"La escolarizacion en el segundo ciclo de educacion infantil (3 a 5 anos) es en Espana practicamente universal, con un 98% de los menores matriculados\" https://www.educacionfpydeportes.gob.es/prensa/actualidad/2025/09/20250909-panoramaeducacion2025.html", "note": "Segundo ciclo (3-6) is free by statute in public and concertado schools and 98% enrolled; primer ciclo (0-3) is neither free nor universal. Before this, ages 1-5 all billed one rate, so the bucket straddled both: a 1-2 year old in paid guarderia was billed the free-tier meal fee. Entry to P3 is by BIRTH-YEAR COHORT (children born in year N start the September of year N+3), so a December-born child starts free schooling at 2 years 9 months; whole-number ages make 3 the correct modelling boundary." }, "non_cash.parental_leave_paid_pct": { "source": "Real Decreto-ley 9/2025 preamble (cien por cien replacement) https://www.boe.es/eli/es/rdl/2025/07/29/9 ; LGSS art. 179 (base reguladora) and art. 148 (tope maximo de la base de cotizacion)", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 38 weeks into non_cash.parental_leave_capped (100% of a base capped at the base maxima de cotizacion), so no percentage-paid weeks remain on this field. Was 100 with the cap ignored per the old field convention. Watch item unchanged: the RDL preamble records a government commitment to extend to 20 weeks per parent, which is not law and is not stored." }, "non_cash.parental_leave_weeks": { "source": "Estatuto de los Trabajadores art. 48.4 as amended by Real Decreto-ley 9/2025 de 29 de julio, in force 31/07/2025 https://www.boe.es/buscar/act.php?id=BOE-A-2025-15741 ; consolidated ET https://www.boe.es/buscar/act.php?id=BOE-A-2015-11430", "last_updated": "2026-08-01", "confidence": "high", "note": "38 wk = 19 per parent x 2 (6 obligatory post-partum + 11 usable to age 12 months + 2 usable to age 8). ET art. 48.4 states este derecho es individual de la persona trabajadora sin que pueda transferirse su ejercicio al otro progenitor. Spain has no separate maternity leave: both parents hold the identical permiso por nacimiento y cuidado. Was 16, one parent on the pre-2025 duration. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Excluded: the 8-week permiso parental of art. 48 bis, which RDL 9/2025 confirms remains unpaid; the 32-week monoparental variant; multiple-birth uplifts." }, "tax.inbound_regime": { "source": "Beckham Law, art. 93.2.e) LIRPF (Ley 35/2006, as amended by Ley 28/2022 Ley de Startups and Ley 7/2024): marginal two-band scale on Spanish employment income, 24% up to EUR 600,000 and 47% on the excess. Duration = the tax period of the move plus the five following (6 years); 5 prior tax periods of non-residency; election on modelo 149. AEAT Manual practico IRPF: https://sede.agenciatributaria.gob.es/Sede/ayuda/manuales-videos-folletos/manuales-practicos/irpf-2024/c15-calculo-impuesto-determinacion-cuotas-integras/regimen-fiscal-especial-aplicable-trabajadores.html ; AEAT Regimen especial impatriados art. 93: https://sede.agenciatributaria.gob.es/Sede/ayuda/manuales-videos-folletos/manuales-practicos/manual-tributacion-no-residentes/regimenes-opcionales/regimen-especial-impatriados.html", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31 OVER-GRANT FIX. The block declared flat_rate 0.24 with no max_income and no flat_rate_above, so 24% ran to infinity even though this very entry described the 47% band. Wired max_income 600000 + flat_rate_above 0.47; the engine blends them into one effective rate. Barcelona single filer at EUR 1,000,000: tax 243,887 -> 335,887. Nothing at or below EUR 600,000 moves." }, "tax.federal_brackets": { "source": "Art. 63.1 LIRPF, escala estatal 2026 (via guiafiscal.es/irpf); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.state_brackets": { "source": "Junta de Andalucía autonómica IRPF scale 2025 (9.5-22.5%)", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "IBI Málaga 0.4510% of cadastral ~0.45% effective", "last_updated": "2026-05-26", "confidence": "medium" }, "housing._summary": { "source": "Numbeo + Idealista Málaga 2026 (expat/nomad rent surge)", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "SNS universal + private (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Private guardería 0-3; 3-6 free public; means-tested cash benefit 0. Junta de Andalucía escolarización 0-3 (juntadeandalucia.es): age-2 tuition made universal-free for 2025-26 (age <1 free only below 0.5-0.75 IPREM per cápita, not ICP-relevant), aula matinal/aula de tarde extras still renta-scaled and non-zero for high earners", "last_updated": "2026-07-17", "confidence": "low", "note": "preschool_monthly raised from EUR 0 to a rough EUR 200/mo blend — the old 0 assumed the whole age 1-5 bucket was free, but only age 2 base tuition is (extras aren't, and ages 1/3/4/5 aren't). subsidy_per_kid_annual correctly stays 0 (no cash beca in this region, it's a fee waiver not representable as a per-kid check). Precision here is a follow-up item, not independently ground-truthed" }, "food._summary": { "source": "Numbeo Málaga 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "EMT Málaga monthly pass ~EUR 24", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Málaga", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "AGP/MAD->US economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Means-tested, 0 at tech income (national, inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "non_cash._summary": { "source": "Spanish labour law (national, inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Málaga: EUR 46,123, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 46,123 / 0.867 = USD 53,198; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2128, R5-rounded down to 0.21. REPLACES 0.35 (2026-08-11), and the interim 0.21 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.35 was 0.8750 of Madrid's 0.4; that ratio x Madrid's re-derived 0.24 = 0.2100, R5-rounded down to 0.21. Madrid's own figure: EUR 53,912 / 0.867 = USD 62,182, / USD 250,000 = 0.2487. This is a BASIS move, not a re-measurement: Málaga's rank inside Spain is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.35, growing tech hub but Andalusia wage base." }, "tax.social_security_cap": { "confidence": "high", "source": "Seguridad Social base maxima 2026 EUR 5,101.20/mo (taxsummaries.pwc.com/spain); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "tax.filing_basis": { "source": "Agencia Tributaria - individual IRPF (the engine files spouses individually; the elective declaracion conjunta is excluded)", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "childcare.preschool_monthly": { "last_updated": "2026-08-01", "confidence": "high", "source": "Junta de Andalucia, precios de los servicios complementarios, curso 2025-2026: comedor escolar EUR 5,54/dia; aula matinal EUR 23,91/mes (EUR 1,82/dia). ONE regional price applies to every Andalusian public school, so Malaga and Seville are identical by construction.", "note": "EUR 5.54/day x 15 = EUR 83.10 comedor, plus aula matinal EUR 23.91/month over 10 months = EUR 19.93, total EUR 103.03. Was 175. Malaga and Seville now match, as the single regional price schedule requires; they previously differed (175 vs 160) with no basis. NOTE a REMAINING inconsistency this pass did not fix: the paid-tier rates still differ (Malaga 550 vs Seville 400) under the same regional regime, and the boundary change makes them bill three years instead of one. Neither has a sourced replacement, so neither was guessed. Free tuition, so the modelled cost is the real residual a dual-earner family pays: comedor escolar plus aula matinal. Stored value is the ANNUAL total divided by 12 (computeChildcareCost multiplies by 12), NOT the school-year invoice, which runs ~10 months. Comedor annualised as daily rate x 180 dias lectivos / 12 = daily x 15; 180 is the conservative end of the 175-181 range (independently implied day counts: ~177 Catalonia, ~181 Andalucia). Applies the owner decision of 2026-07-18 that Spain follows the statutory free convention, now that the age boundary makes it applicable to the free ages only. Sourcing: docs/superpowers/research/2026-08-01-childcare-free-tier-boundary.md and 2026-08-01-spain-wraparound-fees.md." }, "non_cash.parental_leave_capped": { "source": "Orden PJC/297/2026, de 30 de marzo, art. 2.1: tope maximo de la base de cotizacion 5.101,20 euros mensuales desde el 1 de enero de 2026 https://www.boe.es/diario_boe/txt.php?id=BOE-A-2026-7296 ; LGSS art. 179 (the base reguladora is the contingencias comunes contribution base) and art. 148 (that base is capped at the tope maximo)", "last_updated": "2026-08-06", "confidence": "high", "note": "All 38 weeks pay 100% of a base capped at EUR 5,101.20/month (EUR 61,214.40/yr), one clean segment with no phase split: above that gross the benefit is the cap, not the salary. At EUR 90,000 the effective rate is ~68%, at EUR 150,000 ~41%, versus the naked 100% this replaces. cap_amount doubles as the payout ceiling because the rate is 100. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Spain)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "RD 16/2012 pharmaceutical co-payment tramos, with our ICP household in the uncapped 50-60% band + Ministerio de Sanidad SCS household out-of-pocket with dental at ~32%", "last_updated": "2026-08-08", "confidence": "medium", "note": "Spain is the European exception on the child rule: a dependent minor registered on a working parent's card pays the SAME co-payment percentage as the parent, so the sub-linearity here is a volume argument (free well-child checks and the vaccination schedule never touch the pharmacy co-pay), not a statutory discount. Spain has NO statutory annual ceiling for the household we model; the monthly caps introduced in 2026 apply only below EUR 35,000 of income. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "RD 16/2012 pharmaceutical co-payment tramos, with our ICP household in the uncapped 50-60% band + Ministerio de Sanidad SCS household out-of-pocket with dental at ~32%", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "INE Spain, Urban Audit indicator 6.4, Area Urbana Funcional de Malaga. https://www.ine.es/jaxiT3/Tabla.htm?t=69327", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Duracion media del desplazamiento al trabajo (solo ida)", "year": "2011", "note": "AUF value 21.5." }, "non_cash.sick_leave": { "source": "Ley General de la Seguridad Social, current consolidated text Real Decreto; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Agencia Tributaria, Manual Renta 2025 s.8.2.2.6 https://sede.agenciatributaria.gob.es/Sede/ayuda/manuales-videos-folletos/manuales-ayuda-presentacion/irpf-2025/8-cumplimentacion-irpf/8_2-base-liquidable-general-base-ahorro/8_2_2-reducciones-aportaciones-prevision-social/8_2_2_6-aportaciones-anuales-maximas-limite-reduccion.html", "last_updated": "2026-08-11", "confidence": "high", "note": "Reduccion de la base imponible general for aportaciones a sistemas de prevision social (Ley 35/2006 arts. 51-52), EUR 1,500/yr. A base deduction, not a credit, and individual. NOT 8,500: the +8,500 increment requires EMPLOYER contributions to a plan de empleo, which we do not model, and worker contributions only ride along under a matching coefficient table. pct_of_earnings deliberately omitted: the outer 30%-of-net-earned-income bound never binds at EUR 75-95k and 0.30 of GROSS would be both inert and wrong-based. SEPARATE pre-existing understatement, out of scope: computeSpainTaxes does not treat cotizaciones as a gasto deducible." } }, "aarhus": { "tax.inbound_regime": { "source": "Forskerskatteordningen, KSL ss 48E-F: flat 27% after 8% AM-bidrag = 32.84% all-in, up to 84 months, no upper cap. KSL s 48E stk. 3 nr. 3 sets a grundbelob of DKK 59,500/mo at 2010 level, indexed under PSL s 20, 'hvortil kommer bidrag efter lov om Arbejdsmarkedets Tillaegspension'. Indexed 2026 figure DKK 65,400/mo measured BEFORE AM-bidrag per Den juridiske vejledning C.F.6.1.4 (https://info.skat.dk/data.aspx?oid=1977289), plus the employee ATP share of DKK 99/mo (borger.dk, ATP-satser for privat virksomhed, effective 2026-01-01). Ten-year prior tax-liability bar, KSL s 48E stk. 3 nr. 1 (https://info.skat.dk/data.aspx?oid=2350238). skat.dk tax-scheme-for-researchers.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: min_income 785,000 -> 786,000 = (65,400 + 99) x 12, rounded up. The 2026 monthly figure was LOWERED from DKK 78,000 (Ivaerksaetterpakke), so the direction in the old prose was right. Two prose corrections: (a) this is NOT note-only, the block carries flat_rate 0.27 and the engine applies it; (b) the regime note used to claim the scheme cuts the rate 'from ~52-56% down to about 33%', which compares a marginal rate to an effective one. Our own Denmark integration test pins the ordinary effective rate near 40% and the gap at ~7.1pp, so the note now says 40%. NOT PRIMARILY SOURCED: the transitional provision limiting DKK 65,400 to employments commencing on or after 2026-01-01 (retsinformation.dk and ft.dk both 403); four Danish secondary sources agree, and it does not affect a current-year arrival." }, "tax.filing_basis": { "source": "Skattestyrelsen (skat.dk) - individual taxation; unused personal allowance transferable between spouses", "last_updated": "2026-07-11", "note": "Activates the per-filer split in compute() for two-earner households." }, "tax.federal_brackets": { "source": "Denmark 2026 national income tax: peer-identical with Copenhagen", "last_updated": "2026-05-26", "confidence": "high" }, "tax.kommunal_rate": { "source": "Aarhus kommuneskat 24.52% (Skatteministeriet, unchanged 2025-26)", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Ejendomsværdiskat + grundskyld ~0.65% effective", "last_updated": "2026-05-26", "confidence": "medium" }, "housing._summary": { "source": "Numbeo + DinBoligAgent Aarhus 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Tax-funded universal (inherited from Copenhagen baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "aarhus.dk 2026 vuggestue/børnehave net parent fee (~25% share)", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Aarhus 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Midttrafik 30-day pass DKK 435", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Aarhus", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "AAR/CPH->US economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Børnecheck (national, inherited)", "last_updated": "2026-05-26", "confidence": "high", "note": "Danish borne- og ungeydelse: universal below a high-income threshold, means-tested down 2% above the ~DKK 892k topskat base; largely survives, mildly reduced for the very highest earners." }, "non_cash._summary": { "source": "Danish labour law (national, inherited); 25 vacation, 11 holidays", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Aarhus. Its basis is converted with the rest of Denmark and its rank inside Denmark is preserved. Prior source: Derived, not observed: no all-levels total-comp measurement exists for Aarhus. Basis re-derived with its country anchor Copenhagen; Aarhus's stored ratio to Copenhagen is preserved. Prior source: Levels.fyi Aarhus SWE 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.55 x 0.6774 = 0.3726, R5-rounded down to 0.37. 0.6774 is the MEDIAN of (re-derived / previous) across the 1 Denmark city that DO carry an all-levels measurement (Copenhagen 0.62 to 0.42), so Aarhus moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Aarhus's rank inside Denmark is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.55 was 0.8871 of Copenhagen's 0.62; that ratio x Copenhagen's re-derived 0.42 = 0.3726, R5-rounded down to 0.37. Copenhagen's own figure: DKK 695,268 / 6.47 = USD 107,460, / USD 250,000 = 0.4298. This is a BASIS move, not a re-measurement: Aarhus's rank inside Denmark is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.55, 2nd city below Copenhagen." }, "tax.personal_allowance": { "confidence": "high", "source": "Personfradrag 2026 DKK 54,100 (skat.dk); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "tax.personfradrag": { "source": "Personfradrag 2026 DKK 54,100: https://www.exploringdenmark.com/danish-tax-rates-2026/ (cross-checked against SKAT)", "source_name": "SKAT / exploringdenmark", "last_updated": "2026-07-23", "confidence": "high", "note": "Applied INSIDE computeDenmarkTaxes against personal income (0.92 x gross), not against gross. standard_deduction_single/married are therefore held at 0 so compute() passes gross through; a non-zero value there would double-subtract the allowance. Correct-first item #3, 2026-07-17 resolutions doc." }, "tax.mellemskat_rate": { "source": "skat.dk 2026 schedule: bundskat/mellemskat/topskat/toptopskat https://skat.dk/en-us/help/botton-bracket-middle-bracket-top-bracket-and-additional-top-bracket-tax", "source_name": "SKAT (skat.dk)", "last_updated": "2026-07-23", "confidence": "high", "note": "2026 three-tier reform: mellemskat 7.5% on personal income above DKK 641,200 (i.e. after the 8% AM-bidrag). Replaced the pre-2026 single 15% topskat together with the two keys below." }, "tax.mellemskat_threshold": { "source": "skat.dk 2026 schedule: bundskat/mellemskat/topskat/toptopskat https://skat.dk/en-us/help/botton-bracket-middle-bracket-top-bracket-and-additional-top-bracket-tax", "source_name": "SKAT (skat.dk)", "last_updated": "2026-07-23", "confidence": "high", "note": "DKK 641,200 measured on personal income AFTER AM-bidrag. Gross-basis equivalent 696,957 (= 641,200 / 0.92), which confirms the basis." }, "tax.toptopskat_rate": { "source": "skat.dk 2026 schedule: bundskat/mellemskat/topskat/toptopskat https://skat.dk/en-us/help/botton-bracket-middle-bracket-top-bracket-and-additional-top-bracket-tax", "source_name": "SKAT (skat.dk)", "last_updated": "2026-07-23", "confidence": "high", "note": "2026 reform: toptopskat 5% on personal income above DKK 2,592,700. Does not bind at ICP salaries but is modeled for completeness." }, "tax.toptopskat_threshold": { "source": "skat.dk 2026 schedule: bundskat/mellemskat/topskat/toptopskat https://skat.dk/en-us/help/botton-bracket-middle-bracket-top-bracket-and-additional-top-bracket-tax", "source_name": "SKAT (skat.dk)", "last_updated": "2026-07-23", "confidence": "high", "note": "DKK 2,592,700 on personal income after AM-bidrag. Gross-basis equivalent 2,818,152." }, "tax.topskat_rate": { "source": "skat.dk 2026 schedule: bundskat/mellemskat/topskat/toptopskat https://skat.dk/en-us/help/botton-bracket-middle-bracket-top-bracket-and-additional-top-bracket-tax", "source_name": "SKAT (skat.dk)", "last_updated": "2026-07-23", "confidence": "high", "note": "CUT to 7.5% by the 2026 reform (was 15%). Now the middle of three upper tiers, not the top one." }, "tax.topskat_threshold": { "source": "skat.dk 2026 schedule: bundskat/mellemskat/topskat/toptopskat https://skat.dk/en-us/help/botton-bracket-middle-bracket-top-bracket-and-additional-top-bracket-tax", "source_name": "SKAT (skat.dk)", "last_updated": "2026-07-23", "confidence": "high", "note": "RAISED to DKK 777,900 on personal income after AM-bidrag by the 2026 reform (was 640,108 applied to the wrong base). Gross-basis equivalent 845,543." }, "tax.standard_deduction_single": { "source": "skat.dk 2026 schedule: bundskat/mellemskat/topskat/toptopskat https://skat.dk/en-us/help/botton-bracket-middle-bracket-top-bracket-and-additional-top-bracket-tax", "source_name": "SKAT (skat.dk)", "last_updated": "2026-07-23", "confidence": "high", "note": "Held at 0 BY DESIGN. Denmark applies its personfradrag inside computeDenmarkTaxes against personal income (0.92 x gross); leaving a lump here would double-subtract it. See tax.personfradrag." }, "tax.standard_deduction_married": { "source": "skat.dk 2026 schedule: bundskat/mellemskat/topskat/toptopskat https://skat.dk/en-us/help/botton-bracket-middle-bracket-top-bracket-and-additional-top-bracket-tax", "source_name": "SKAT (skat.dk)", "last_updated": "2026-07-23", "confidence": "high", "note": "Held at 0 BY DESIGN. Denmark applies its personfradrag inside computeDenmarkTaxes against personal income (0.92 x gross); leaving a lump here would double-subtract it. See tax.personfradrag." }, "tax.beskaeftigelsesfradrag_rate": { "source": "DK-1: beskæftigelsesfradrag 12.75% of employment income (2026), a ligningsmæssigt fradrag reducing the kommuneskat base ONLY (not personlig indkomst / bundskat / mellemskat / topskat). SKAT employment-and-job-allowances; PwC DK deductions.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.beskaeftigelsesfradrag_cap": { "source": "DK-1: beskæftigelsesfradrag maximum DKK 63,300 (2026). SKAT; SKM faktaark.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.jobfradrag_rate": { "source": "DK-1: jobfradrag 4.5% of employment income over the threshold (2026), municipal base only. SKAT employment-and-job-allowances.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.jobfradrag_threshold": { "source": "DK-1: jobfradrag applies to employment income over DKK 235,200 (2026). SKAT.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.jobfradrag_cap": { "source": "DK-1: jobfradrag maximum DKK 3,100 (2026). SKAT.", "last_updated": "2026-07-26", "confidence": "high" }, "non_cash.parental_leave_weeks": { "source": "borger.dk, Loenmodtager paa barsel (ny orlovsmodel) https://www.borger.dk/familie-og-boern/barsel-oversigt/barsel-loenmodtagere-ny-orlovsmodel ; barselsloven (2022 model, births from 2 Aug 2022)", "last_updated": "2026-08-06", "confidence": "high", "note": "52 wk = mother 4 pre-birth + 24 after + father/co-parent 24 (2 at birth, 9 earmarked non-transferable and 13 transferable inside each 24). Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. The both-parents flag stays off per the Sweden precedent: reserved weeks inside a standard mother-plus-father split, not a summed child-rearing entitlement driving the count." }, "non_cash.parental_leave_paid_pct": { "source": "borger.dk: barselsdagpenge are computed from your hourly wage and capped at the hojeste sats, DKK 5,085/wk before tax in 2026", "last_updated": "2026-08-06", "confidence": "high", "note": "0 percentage-paid weeks: the statute pays your wage up to a flat cap, and the cap binds for any ICP (it is reached at ~DKK 22,041/month gross), so the whole entitlement lives in non_cash.parental_leave_flat. The previous 50 was an admitted guess that blended employer top-ups into a statutory field." }, "non_cash.parental_leave_flat": { "source": "borger.dk https://www.borger.dk/familie-og-boern/barsel-oversigt/barsel-loenmodtagere-ny-orlovsmodel : hojeste sats DKK 5,085/wk before tax, 2026 (DKK 137.43/hr x 37)", "last_updated": "2026-08-06", "confidence": "high", "note": "52 x 5,085 = DKK 264,420 total, about a 29% effective rate at a DKK 900,000 gross. Owner call 2026-08-05 under the conservative rule: state dagpenge only. Funktionaerloven para 7 half pay from the employer (salaried employees, 4 wk pre-birth to 14 wk post) is EXCLUDED from this value: counting it is more generous and carries the higher evidence bar, and the para 7 details were not verified this pass. Flag it as a possible refinement, not a defect." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Danmarks Statistik health expenditure DKK 262.9bn (2023) at 13.9% household out-of-pocket + sundhed.dk medicintilskud tiers and the DKK 4,850 egenbetalingsloft", "last_updated": "2026-08-08", "confidence": "medium", "note": "The old DKK 4,250 was 88% of the medicine egenbetalingsloft, which is only reached at DKK 22,596 of gross annual medicine spend, so it was maximum-flavoured rather than an average. Denmark is the only Nordic country where children add material cost: under-18s get a 60% medicine subsidy from the first krone, which is a discount rather than an exemption, while GP and hospital care are free for everyone in gruppe 1. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Danmarks Statistik health expenditure DKK 262.9bn (2023) at 13.9% household out-of-pocket + sundhed.dk medicintilskud tiers and the DKK 4,850 egenbetalingsloft", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "DTU Transportvaneundersogelsen, faktaark om pendling 2016-2019, Region Midtjylland. https://www.man.dtu.dk/-/media/institutter/management/scientific-advice/transportvaneundersoegelsen/t-u/tu-udgivelser/faktaark/2016-19_faktaark_pendling.pdf", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Gennemsnitligt tidsforbrug paa pendlerrejser", "year": "2016-2019", "note": "CONVERSION: 53 min ROUND TRIP / 2 = 26.5 one way. Region rather than metro is the weak link here, since rural Jutland inflates it." }, "non_cash.sick_leave": { "source": "Sygedagpengeloven (Consolidated Act on Sickness Benefits) , §31 (arbejdsgiverperiode; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.detfagligehus.dk/...", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Skat.dk, Fradrag for indbetalinger til pension https://skat.dk/borger/pension-og-efterloen/fradrag-for-indbetalinger-til-pension ; Livsvarig livrente https://skat.dk/borger/pension-og-efterloen/livsvarig-livrente ; Juridisk vejledning C.A.10.2.1.3.3 (PBL s.18) https://info.skat.dk/data.aspx?oid=2048243", "last_updated": "2026-08-11", "confidence": "medium", "note": "JUDGMENT CALL. Ratepension plus ophoerende livrente carry a combined DKK 68,700 ceiling for 2026, shared with employer-administered schemes; overshoot draws a 20% afgift. Livsvarig livrente is genuinely UNCAPPED where the saver signs a 10-year-or-longer agreement, so the strictly-correct value for that saver is no key at all. 68,700 is chosen because our ICP is a mobile tech worker who would not sign a 10-year binding Danish lifetime-annuity commitment, and it is the conservative side. Aldersopsparing gives no deduction. The opfyldningsfradrag DKK 63,200 is for lump-sum engangsindskud, not a general annual ceiling. NEIGHBOURING BUG, out of scope: the Danish branch charges AM-bidrag on the post-retirement-deduction figure; AM-bidrag is levied on full gross." } }, "espoo": { "tax.filing_basis": { "source": "Vero (vero.fi) - spouses taxed as individuals on earned income", "last_updated": "2026-07-11", "note": "Activates the per-filer split in compute() for two-earner households." }, "tax.federal_brackets": { "source": "Finland 2026 national income tax: peer-identical with Helsinki", "last_updated": "2026-05-26", "confidence": "high" }, "tax.state_income_tax_rate": { "source": "Espoo kunnallisvero 5.30% post-reform = 0.179 pre-reform-basis (matches Helsinki's 0.185 model basis)", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Kiinteistövero ~0.55%", "last_updated": "2026-05-26", "confidence": "medium" }, "housing._summary": { "source": "Numbeo + Investropa Espoo 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Public universal + private (inherited from Helsinki baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Espoo varhaiskasvatus income-capped (max ~EUR 311/mo)", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Espoo 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "HSL B/C monthly pass ~EUR 72", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Espoo", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "HEL->US economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Lapsilisä (national, inherited)", "last_updated": "2026-05-26", "confidence": "high", "note": "Finnish lapsilisa (child benefit): universal, NOT means-tested." }, "non_cash._summary": { "source": "Finnish labour law (national, inherited); 25 vacation, 11 holidays", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Espoo: EUR 62,437, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 62,437 / 0.867 = USD 72,015; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2881, R5-rounded down to 0.28. REPLACES 0.46 (2026-08-11), and the interim 0.33 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Measures 0.851 of Helsinki, so the stored ordering (Espoo above Helsinki) is refuted. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.46 was 1.0222 of Helsinki's 0.45; that ratio x Helsinki's re-derived 0.33 = 0.3373, R5-rounded down to 0.33. Helsinki's own figure: EUR 73,328 / 0.867 = USD 84,577, / USD 250,000 = 0.3383. This is a BASIS move, not a re-measurement: Espoo's rank inside Finland is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.46, Helsinki metro tech HQs." }, "tax.state_tax_brackets": { "source": "Pre-reform-basis national scale coherent with the 18.5%-basis municipal rate (netsalaire.com/en/finland/income-tax; cross-checked countrytaxcalc.com, PwC); verified tax audit 2026-07-14", "confidence": "high", "note": "The old post-reform scale (top 41.75/44%) double-counted the pre-reform municipal 18.5%. National tax at EUR 138k now ~27,628 (benchmark).", "last_updated": "2026-07-17" }, "tax.social_insurance_rate": { "source": "Employee 2026: TyEL 7.30% + health insurance 1.98% + unemployment 0.89% = 10.17% (varma.fi social security contributions 2026 PDF); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "healthcare.statutory_employee": { "source": "vero.fi, Vakuutetun sairausvakuutusmaksu (guidance 338472): 2026 employee sairaanhoitomaksu 1.10% of earned income and paivarahamaksu 0.88% above an annual earned-income floor of about EUR 17,255, together the 1.98% employee sairausvakuutusmaksu. Corroborated by stm.fi, 'Sairausvakuutusmaksut nousevat hieman vuonna 2026'. Neither component carries a ceiling; the absence is consistent across both sources, whose purpose is to state the full calculation. Public treatment is delivered by the 21 hyvinvointialueet out of general taxation, not out of this contribution.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll. UNCAPPED rests on the consistent absence of a ceiling clause in two primary sources rather than on an explicit 'there is no cap' sentence. The 0.88% paivarahamaksu is deliberately excluded: it funds cash daily allowances, closer to disability insurance than to health care. THE FRAMING MATTERS MORE THAN THE NUMBER HERE: 1.10% is a genuinely small earmarked contribution, and a reader who takes it for the cost of Finnish health care will badly flatter Finland, so the note says outright where the money actually comes from. Also for 2026: the age-graded employee TyEL contribution is abolished and flat at all ages, and employee unemployment insurance rose from 0.59% to 0.89%." }, "non_cash.parental_leave_weeks": { "source": "Kela https://www.kela.fi/web/selkokieli/vanhempainpaivarahat (raskausraha 40 arkipaivaa; vanhempainraha 320 arkipaivaa per child, 160 per parent, up to 63 transferable); 2022 family-leave reform, births since 4 Sep 2022", "last_updated": "2026-08-06", "confidence": "high", "note": "60 wk = raskausraha 40 arkipaivaa (~6.5 wk, starting pre-birth, a benefit SEPARATE from and additional to vanhempainraha) + vanhempainraha 320 arkipaivaa (a family pot counted once: a single parent receives all 320) = 360 arkipaivaa at Kela's 6 payable days a week = 60. The stored 46 was the pre-2022 system (aitiysraha 105 + vanhempainraha 158 days). REFUTATION recorded: Duunitori's claim that raskausraha sits inside the birthing parent's 160-day quota contradicts Kela's own page, which lists them as distinct benefits; Kela wins. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md." }, "non_cash.parental_leave_paid_pct": { "source": "Kela vanhempainpaivarahat; sairausvakuutuslaki 11 luku. 2026 degressive formula: ~70% of annual income EUR 14,282-43,567, 40% of the slice to 67,113, 25% above, minimum EUR 31.99/day", "last_updated": "2026-08-06", "confidence": "medium-high", "note": "70 is the statutory headline. The formula is degressive rather than flat-capped, so no parental_leave_flat block: at a EUR 90,000 Helsinki salary the daily allowance is ~EUR 152, about 51% effective; the full 70 applies only below ~EUR 43,500. Raskausraha pays at the same daily-allowance level in the post-2022 system (medium confidence on that sub-point)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Kela maksukatto (EUR 815) and laeaekekatto (EUR 636.12) + OECD Finland out-of-pocket per capita, netted of dental (14%) and income-tested long-term-care client fees", "last_updated": "2026-08-08", "confidence": "medium", "note": "The old EUR 700 sat between Finland's two ceilings and modelled one. Held down deliberately by tyoeterveyshuolto, which gives most employed Finns employer-funded primary care free at the point of use, and our ICP is exactly that cohort. Finland's child rule is distinct in Europe: a minor's fees accrue to a PARENT's maksukatto rather than opening a new one, so a child adds real cost that then saturates against the parent's cap. Weakest country in its batch. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Kela maksukatto (EUR 815) and laeaekekatto (EUR 636.12) + OECD Finland out-of-pocket per capita, netted of dental (14%) and income-tested long-term-care client fees", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Traficom, Henkiloliikennetutkimus 2021, seuturaportti Helsingin seutu. https://www.traficom.fi/files/media/file/HLT%202021%20seuturaportti%20Helsingin%20seutu.pdf", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Keskimaarainen matka-aika, tyo", "year": "2021", "note": "GEOGRAPHY CALL: Espoo takes the Helsinki metro figure, because no municipality-level duration is published (the by-area table gives distance, not time). Espoo's work trips are the LONGEST of the core municipalities (17.3 km against 15.3 for the region), so the previous arrangement of Espoo 28 below Helsinki 30 was backwards." }, "non_cash.sick_leave": { "source": "Työsopimuslaki (Employment Contracts Act, 55/2001) Chapter 2 Section 11 (employer; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Vero.fi, Pitkaaikaissaastamissopimuksen ja vapaaehtoisen yksiloellisen elaekevakuutuksen verotus https://www.vero.fi/syventavat-vero-ohjeet/ohje-hakusivu/48212/pitkaaikaissaastamissopimuksen_ja_vapaa/ ; VM, Vapaaehtoisen elaekesaastamisen verotuki poistuu vuonna 2027 https://vm.fi/-/vapaaehtoisen-elakesaastamisen-verotuki-poistuu-vuonna-2027", "last_updated": "2026-08-11", "confidence": "high", "note": "No earned-income deduction. The EUR 5,000/yr relief for vapaaehtoinen yksiloellinen elaekevakuutus and PS-sopimus is deducted from PAAOMATULOISTA, capital income; with no capital income it converts to the erityinen alijaeaemaehyvitys, a CREDIT worth about EUR 1,400. It never reduces the earned-income base, which is what this field represents. Abolished entirely from 2027. SEPARATE KNOWN GAP, out of scope: computeFinlandTaxes deducts nothing at all, including the mandatory TyEL and unemployment-insurance employee contributions, which ARE deductible in Finland and run the other way." } }, "tampere": { "tax.filing_basis": { "source": "Vero (vero.fi) - spouses taxed as individuals on earned income", "last_updated": "2026-07-11", "note": "Activates the per-filer split in compute() for two-earner households." }, "tax.federal_brackets": { "source": "Finland 2026 national income tax: peer-identical with Helsinki", "last_updated": "2026-05-26", "confidence": "high" }, "tax.state_income_tax_rate": { "source": "Tampere kunnallisvero 7.60% post-reform = 0.202 pre-reform-basis (matches Helsinki model)", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Kiinteistövero ~0.55%", "last_updated": "2026-05-26", "confidence": "medium" }, "housing._summary": { "source": "Numbeo Tampere 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Public universal + private (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Tampere varhaiskasvatus income-capped (~EUR 300/mo)", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Tampere 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Nysse 30-day pass EUR 32.50", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Tampere", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "HEL->US economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Lapsilisä (national, inherited)", "last_updated": "2026-05-26", "confidence": "high", "note": "Finnish lapsilisa (child benefit): universal, NOT means-tested." }, "non_cash._summary": { "source": "Finnish labour law (national, inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Tampere: EUR 50,412, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "EUR 50,412 / 0.867 = USD 58,145; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2326, R5-rounded down to 0.23. REPLACES 0.4 (2026-08-11), and the interim 0.29 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Compressed ladder: median only 6.1% above its own p25. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.4 was 0.8889 of Helsinki's 0.45; that ratio x Helsinki's re-derived 0.33 = 0.2933, R5-rounded down to 0.29. Helsinki's own figure: EUR 73,328 / 0.867 = USD 84,577, / USD 250,000 = 0.3383. This is a BASIS move, not a re-measurement: Tampere's rank inside Finland is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.40, engineering hub below Helsinki." }, "tax.state_tax_brackets": { "source": "Pre-reform-basis national scale coherent with the 18.5%-basis municipal rate (netsalaire.com/en/finland/income-tax; cross-checked countrytaxcalc.com, PwC); verified tax audit 2026-07-14", "confidence": "high", "note": "The old post-reform scale (top 41.75/44%) double-counted the pre-reform municipal 18.5%. National tax at EUR 138k now ~27,628 (benchmark).", "last_updated": "2026-07-17" }, "tax.social_insurance_rate": { "source": "Employee 2026: TyEL 7.30% + health insurance 1.98% + unemployment 0.89% = 10.17% (varma.fi social security contributions 2026 PDF); verified tax audit 2026-07-14", "confidence": "high", "last_updated": "2026-07-17" }, "healthcare.statutory_employee": { "source": "vero.fi, Vakuutetun sairausvakuutusmaksu (guidance 338472): 2026 employee sairaanhoitomaksu 1.10% of earned income and paivarahamaksu 0.88% above an annual earned-income floor of about EUR 17,255, together the 1.98% employee sairausvakuutusmaksu. Corroborated by stm.fi, 'Sairausvakuutusmaksut nousevat hieman vuonna 2026'. Neither component carries a ceiling; the absence is consistent across both sources, whose purpose is to state the full calculation. Public treatment is delivered by the 21 hyvinvointialueet out of general taxation, not out of this contribution.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll. UNCAPPED rests on the consistent absence of a ceiling clause in two primary sources rather than on an explicit 'there is no cap' sentence. The 0.88% paivarahamaksu is deliberately excluded: it funds cash daily allowances, closer to disability insurance than to health care. THE FRAMING MATTERS MORE THAN THE NUMBER HERE: 1.10% is a genuinely small earmarked contribution, and a reader who takes it for the cost of Finnish health care will badly flatter Finland, so the note says outright where the money actually comes from. Also for 2026: the age-graded employee TyEL contribution is abolished and flat at all ages, and employee unemployment insurance rose from 0.59% to 0.89%." }, "non_cash.parental_leave_weeks": { "source": "Kela https://www.kela.fi/web/selkokieli/vanhempainpaivarahat (raskausraha 40 arkipaivaa; vanhempainraha 320 arkipaivaa per child, 160 per parent, up to 63 transferable); 2022 family-leave reform, births since 4 Sep 2022", "last_updated": "2026-08-06", "confidence": "high", "note": "60 wk = raskausraha 40 arkipaivaa (~6.5 wk, starting pre-birth, a benefit SEPARATE from and additional to vanhempainraha) + vanhempainraha 320 arkipaivaa (a family pot counted once: a single parent receives all 320) = 360 arkipaivaa at Kela's 6 payable days a week = 60. The stored 46 was the pre-2022 system (aitiysraha 105 + vanhempainraha 158 days). REFUTATION recorded: Duunitori's claim that raskausraha sits inside the birthing parent's 160-day quota contradicts Kela's own page, which lists them as distinct benefits; Kela wins. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md." }, "non_cash.parental_leave_paid_pct": { "source": "Kela vanhempainpaivarahat; sairausvakuutuslaki 11 luku. 2026 degressive formula: ~70% of annual income EUR 14,282-43,567, 40% of the slice to 67,113, 25% above, minimum EUR 31.99/day", "last_updated": "2026-08-06", "confidence": "medium-high", "note": "70 is the statutory headline. The formula is degressive rather than flat-capped, so no parental_leave_flat block: at a EUR 90,000 Helsinki salary the daily allowance is ~EUR 152, about 51% effective; the full 70 applies only below ~EUR 43,500. Raskausraha pays at the same daily-allowance level in the post-2022 system (medium confidence on that sub-point)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Kela maksukatto (EUR 815) and laeaekekatto (EUR 636.12) + OECD Finland out-of-pocket per capita, netted of dental (14%) and income-tested long-term-care client fees", "last_updated": "2026-08-08", "confidence": "medium", "note": "The old EUR 700 sat between Finland's two ceilings and modelled one. Held down deliberately by tyoeterveyshuolto, which gives most employed Finns employer-funded primary care free at the point of use, and our ICP is exactly that cohort. Finland's child rule is distinct in Europe: a minor's fees accrue to a PARENT's maksukatto rather than opening a new one, so a child adds real cost that then saturates against the parent's cap. Weakest country in its batch. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Kela maksukatto (EUR 815) and laeaekekatto (EUR 636.12) + OECD Finland out-of-pocket per capita, netted of dental (14%) and income-tested long-term-care client fees", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Traficom, Henkiloliikennetutkimus 2021, seuturaportti Tampereen kaupunkiseutu. https://www.traficom.fi/files/media/file/HLT%202021%20seuturaportti%20Tampereen%20kaupunkiseutu.pdf", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Keskimaarainen matka-aika, tyo", "year": "2021", "note": "Directly published mean, one way, all modes, 8-municipality city region, work-only." }, "non_cash.sick_leave": { "source": "Työsopimuslaki (Employment Contracts Act, 55/2001) Chapter 2 Section 11 (employer; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Vero.fi, Pitkaaikaissaastamissopimuksen ja vapaaehtoisen yksiloellisen elaekevakuutuksen verotus https://www.vero.fi/syventavat-vero-ohjeet/ohje-hakusivu/48212/pitkaaikaissaastamissopimuksen_ja_vapaa/ ; VM, Vapaaehtoisen elaekesaastamisen verotuki poistuu vuonna 2027 https://vm.fi/-/vapaaehtoisen-elakesaastamisen-verotuki-poistuu-vuonna-2027", "last_updated": "2026-08-11", "confidence": "high", "note": "No earned-income deduction. The EUR 5,000/yr relief for vapaaehtoinen yksiloellinen elaekevakuutus and PS-sopimus is deducted from PAAOMATULOISTA, capital income; with no capital income it converts to the erityinen alijaeaemaehyvitys, a CREDIT worth about EUR 1,400. It never reduces the earned-income base, which is what this field represents. Abolished entirely from 2027. SEPARATE KNOWN GAP, out of scope: computeFinlandTaxes deducts nothing at all, including the mandatory TyEL and unemployment-insurance employee contributions, which ARE deductible in Finland and run the other way." } }, "antwerp": { "childcare.free_tier_start_age": { "source": "Eurydice (European Commission), Belgium - Flemish Community, Early childhood education and care: \"From age 2-and-a-half, children are legally entitled to pre-primary education (ISCED 020)\" https://eurydice.eacea.ec.europa.eu/national-education-systems/belgium-flemish-community/early-childhood-education-and-care", "confidence": "high", "note": "Statutory entitlement starts at 2 years 6 months. Kid ages in the model are whole numbers, so the boundary is set to 3, the conservative rounding: the whole of age 2 keeps billing the paid kinderopvang rate rather than the free kleuteronderwijs rate. Before this, ages 1-5 all billed the free-tier rate, so a 1-year-old in income-tariff kinderopvang was charged nothing. Rates themselves are unchanged and still owe a review: the 2025 inkomenstarief ceiling is EUR 35.89/day (about EUR 718-790/month) for a top-income family, above all three modelled infant rates, and the free-tier residual is not zero either (statutory maximumfactuur EUR 55/year plus school meals at roughly EUR 76-95/month). See docs/superpowers/research/2026-08-01-childcare-free-tier-boundary.md.", "last_updated": "2026-08-01" }, "tax.payroll_deductible": { "source": "Batch B base-deduction convention (tax audit 2026-07-14, verified): the IPP brackets apply to income net of the deductible employee RSZ (13.07%), so the ladder runs on gross minus RSZ (computeBelgiumTaxes, gated by this flag); RSZ itself stays on full gross. Source: PwC BE individual/deductions.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.federal_brackets": { "source": "Belgium 2026 federal income tax: peer-identical with Brussels", "last_updated": "2026-05-26", "confidence": "high" }, "tax.communal_surcharge_rate": { "source": "Antwerp aanvullende gemeentebelasting 7.0% (FOD Financiën 2025)", "last_updated": "2026-05-26", "confidence": "high" }, "tax.inbound_regime": { "source": "Bijzonder belastingstelsel voor ingekomen belastingplichtigen (BBIB / RSII). Wet-loi van 18 december 2025 houdende diverse bepalingen, arts. 15-18, Belgisch Staatsblad 30-12-2025, applicable to remuneration paid or allocated from 1 January 2025: expense exclusion raised 30% -> 35% of gross remuneration, the ~EUR 90k tax cap abolished (it survives for social security), and the minimum gross remuneration cut from EUR 75,000 to more than EUR 70,000. Circulaire 2026/C/51 of 1 April 2026 (Administration generale de la Fiscalite) on the retroactive application, contract-adjustment deadline 30 June 2026. Cross-checked Claeys & Engels, EY Belgium and BDO.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31 OVER-GRANT FIX. This entry already said the floor had been lowered to ~EUR 70k, but no min_income was declared, so a 35% exclusion applied at ANY salary. Wired min_income 70000, the exact statutory figure. Brussels single filer at EUR 69,999: tax 16,808 -> 29,015 (ordinary). OPEN QUESTION, not resolved here: the allowance is legally paid bovenop de bezoldiging (on top of gross remuneration), so reduction_pct 0.35 sits between the contract-gross reading (allowance is extra tax-free cash, we understate the break) and the total-package reading (taxable share is 1/1.35 = 74.07%, true carve-out 25.93%, we overstate it). Left at 0.35 pending a modeling decision." }, "tax.property_tax_rate_pct": { "source": "Onroerende voorheffing ~0.9% effective (Flanders urban)", "last_updated": "2026-05-26", "confidence": "medium" }, "housing._summary": { "source": "Numbeo + Immoweb Antwerp 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "INAMI/RIZIV universal + private (inherited from Brussels baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Flemish kinderopvang income-scaled (net fee); kleuterschool free 2.5+", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Antwerp 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "De Lijn Omnipas ~EUR 35/mo", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Antwerp", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "BRU->US economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Groeipakket Flanders (regional; inherited from baseline)", "last_updated": "2026-05-26", "confidence": "medium", "note": "Belgian Groeipakket / allocations familiales: universal (regional), NOT means-tested (social supplements are extra)." }, "non_cash._summary": { "source": "Belgian labour law (national, inherited); 20 vacation, 10 holidays", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Antwerp: EUR 66,830, fetched live 2026-08-11 and recorded at sweep B F8c of docs/superpowers/research/2026-08-11-comp-index-sweep-b.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "EUR 66,830 / 0.867 = USD 77,082; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3083, R5-rounded down to 0.3. REPLACES 0.48 (2026-05-26), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Antwerp by 60%. The measured p50 was 0.642 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.48, port/logistics, below Brussels. CONFIDENCE LOW, and the reason is recorded rather than smoothed over: levels.fyi ranks Brussels (EUR 53,895) BELOW Antwerp (66,830), Ghent (62,057) and Leuven (58,586), which sweep B F8c calls implausible for the capital and EU-institution hub and attributes to sample composition (Brussels senior p25-p75 spread 2.5x against Helsinki 1.6x). All three values come from that one source on one basis, so the country is internally consistent, but its ORDERING is the weakest in the 2026-08-11 re-derivation." }, "tax.filing_basis": { "source": "PwC Worldwide Tax Summaries: Belgium, Individual: Sample personal income tax calculation (spouses taxed separately on all earned income / décumul). https://taxsummaries.pwc.com/belgium/individual/sample-personal-income-tax-calculation", "last_updated": "2026-07-26", "note": "BE-3 (closeout 2026-07-26): spouses taxed separately (décumul) though filing one joint return." }, "tax.professional_expense_rate": { "source": "PwC Worldwide Tax Summaries: Belgium, Individual: Deductions (lump-sum professional expenses = 30% of gross taxable income for employees). https://taxsummaries.pwc.com/belgium/individual/deductions", "last_updated": "2026-07-26", "note": "BE-1 (closeout 2026-07-26): 30% professional-expense forfait rate." }, "tax.professional_expense_cap": { "source": "PwC Worldwide Tax Summaries: Belgium, Individual: Deductions (30% forfait capped at EUR 5,930, AY2026/IY2025). https://taxsummaries.pwc.com/belgium/individual/deductions", "last_updated": "2026-07-26", "note": "BE-1 (closeout 2026-07-26): forfait cap EUR 5,930 (confirmed current cap; IY2026 indexation not yet published. TODO (when it publishes): apply the small uplift; the cap stays EUR 5,930 until then)." }, "tax.tax_free_allowance": { "source": "PwC BE: Belgian tax reform 2026 (quotité exemptée / belastingvrije som = EUR 11,180 for income year 2026, up from EUR 10,910). https://news.pwc.be/belgian-tax-reform-2026-tax-and-social-measures-of-the-programme-law/", "last_updated": "2026-07-26", "note": "BE-2 (closeout 2026-07-26): tax-free allowance for income year 2026." }, "tax.tax_free_allowance_credit_rate": { "source": "PwC BE: quotité exemptée credited at the bottom-bracket 25% rate (the allowance falls entirely within the first IPP band). https://news.pwc.be/belgian-tax-reform-2026-tax-and-social-measures-of-the-programme-law/", "last_updated": "2026-07-26", "note": "BE-2 (closeout 2026-07-26): credited at 25% → reduction ~EUR 2,795." }, "non_cash.parental_leave_weeks": { "source": "belgium.be (conge de maternite, 15 semaines); INAMI maternity amounts https://www.inami.fgov.be/fr/themes/grossesse-et-naissance/montants-et-plafonds-des-indemnites/salaries-chomeurs/montant-maximal-de-votre-indemnite-pendant-un-repos-de-maternite-salariee ; ONEM conges thematiques montants https://www.onem.be/documentation/montants/interruption-de-carriere-credit-temps/conges-thematiques", "last_updated": "2026-08-06", "confidence": "high", "note": "54 wk = repos de maternite 15 (6 prenatal of which 1 mandatory, 9 postnatal mandatory) + conge de naissance 20 days (4 wk, the other parent's own right) + conge parental 4 months PER PARENT via ONEM until the child is 12 (34.78 wk summed across both parents, which is why parental_leave_both_parents is true; Netherlands/Ireland precedent). Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. Was 15/82, maternity alone with the first-30-days rate applied to everything." }, "non_cash.parental_leave_paid_pct": { "source": "INAMI: 82% of uncapped gross for the first 30 days, then 75% of salary capped at EUR 186.79/day from 1 Mar 2026 (max EUR 140.09/day, 6 payable days/wk); conge de naissance 3 days at 100% employer-paid then 17 days at 82% capped", "last_updated": "2026-08-06", "confidence": "high", "note": "79 = blend across the 19.2 percentage-paid weeks: maternity 4.29 wk at 82 (uncapped) + 10.71 wk at 75 (capped, max EUR 840.54/wk) + naissance 0.6 wk at 100 + 3.4 wk at 82 (capped): (352 + 803 + 60 + 279) / 19.0 = 79. The EUR 186.79/day ceiling sits near EUR 58,000/yr, so the capped phases pay their daily maxima for an ICP." }, "non_cash.parental_leave_flat": { "source": "ONEM conges thematiques montants, index 01.03.2026: EUR 1,058.86 gross a month (EUR 951.60 net after the 11.11% withholding) for full-time interruption https://www.onem.be/documentation/montants/interruption-de-carriere-credit-temps/conges-thematiques", "last_updated": "2026-08-06", "confidence": "high", "note": "EUR 243.53/wk = 1,058.86 / 4.348. Four months per parent, so 8 months = 34.78 wk for the household. A flat sum regardless of salary, roughly 15-19% of an ICP wage, which is exactly what the flat block exists to price honestly." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "INAMI/RIZIV maximum a facturer 2026 income bands (top band EUR 2,194.20 per household, child sub-cap EUR 792.35) + Statbel household health expenditure", "last_updated": "2026-08-08", "confidence": "medium", "note": "The old EUR 450 was neither a maximum nor a sourced average, just a round number that landed near the right single-adult figure and was then applied flat to households. Belgium has no child exemption, so the curve scales with headcount, shaded down slightly because the maximum a facturer is a HOUSEHOLD ceiling that does not grow with family size. Note it caps tickets moderateurs only, not fee or room supplements, which are the larger half of Belgian out-of-pocket spend. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "INAMI/RIZIV maximum a facturer 2026 income bands (top band EUR 2,194.20 per household, child sub-cap EUR 792.35) + Statbel household health expenditure", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Belgium has no usable modern commute-duration series: Statbel publishes nothing on commute TIME, the federal Diagnostic home-to-work survey (7th edition, 2024-25) reports mode share and distance only, and the BeMob report PDFs sit behind JavaScript rendering with no JSON API. Eurostat Urban Audit TT1019V does carry a Belgian value (31.0 for BE002C) but its BASIS YEAR IS 1991, and a thirty-five-year-old figure attached to a live URL would read as verified while saying nothing about commuting today, so it was DELIBERATELY REJECTED rather than shipped. OVG Vlaanderen 7 (2023-24) was CHECKED and REJECTED: its duration figures are DAY-LEVEL and ALL-PURPOSE (22.3 min, Flanders), which is the same unit error as SrV, so adopting it would have swapped one defect for another. FORWARD POINTER: OVG 8 is the source to watch, if it publishes duration by motive." }, "non_cash.sick_leave": { "source": "Arbeidsovereenkomstenwet (Employment Contracts Act, 3 July 1978) art. 52; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.ejustice.just.fgov.be", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "SPF Finances, Epargne-pension https://fin.belgium.be/fr/particuliers/avantages-fiscaux/epargne-pension", "last_updated": "2026-08-11", "confidence": "high", "note": "No base deduction. Epargne-pension is a REDUCTION D'IMPOT applied after the tax is computed and never touches the taxable base: 30% on up to EUR 1,050 (max EUR 315) or 25% on up to EUR 1,350 (max EUR 337.50) for income year 2026. Epargne a long terme is also a credit (30%, ceiling 6% of net professional income plus EUR 183.6, absolute max EUR 2,450). Occupational 2nd-pillar (groepsverzekering) employee contributions DO reduce taxable pay but are scheme-specific and we model no employer scheme. computeBelgiumTaxes already nets the mandatory RSZ 13.07% separately, so nothing here double-counts." } }, "ghent": { "childcare.free_tier_start_age": { "source": "Eurydice (European Commission), Belgium - Flemish Community, Early childhood education and care: \"From age 2-and-a-half, children are legally entitled to pre-primary education (ISCED 020)\" https://eurydice.eacea.ec.europa.eu/national-education-systems/belgium-flemish-community/early-childhood-education-and-care", "confidence": "high", "note": "Statutory entitlement starts at 2 years 6 months. Kid ages in the model are whole numbers, so the boundary is set to 3, the conservative rounding: the whole of age 2 keeps billing the paid kinderopvang rate rather than the free kleuteronderwijs rate. Before this, ages 1-5 all billed the free-tier rate, so a 1-year-old in income-tariff kinderopvang was charged nothing. Rates themselves are unchanged and still owe a review: the 2025 inkomenstarief ceiling is EUR 35.89/day (about EUR 718-790/month) for a top-income family, above all three modelled infant rates, and the free-tier residual is not zero either (statutory maximumfactuur EUR 55/year plus school meals at roughly EUR 76-95/month). See docs/superpowers/research/2026-08-01-childcare-free-tier-boundary.md.", "last_updated": "2026-08-01" }, "tax.payroll_deductible": { "source": "Batch B base-deduction convention (tax audit 2026-07-14, verified): the IPP brackets apply to income net of the deductible employee RSZ (13.07%), so the ladder runs on gross minus RSZ (computeBelgiumTaxes, gated by this flag); RSZ itself stays on full gross. Source: PwC BE individual/deductions.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.federal_brackets": { "source": "Belgium 2026 federal income tax: peer-identical with Brussels", "last_updated": "2026-05-26", "confidence": "high" }, "tax.communal_surcharge_rate": { "source": "Ghent aanvullende gemeentebelasting 6.9% (Stad Gent 2025; drops to 6.5% from 2027)", "last_updated": "2026-05-26", "confidence": "high" }, "tax.inbound_regime": { "source": "Bijzonder belastingstelsel voor ingekomen belastingplichtigen (BBIB / RSII). Wet-loi van 18 december 2025 houdende diverse bepalingen, arts. 15-18, Belgisch Staatsblad 30-12-2025, applicable to remuneration paid or allocated from 1 January 2025: expense exclusion raised 30% -> 35% of gross remuneration, the ~EUR 90k tax cap abolished (it survives for social security), and the minimum gross remuneration cut from EUR 75,000 to more than EUR 70,000. Circulaire 2026/C/51 of 1 April 2026 (Administration generale de la Fiscalite) on the retroactive application, contract-adjustment deadline 30 June 2026. Cross-checked Claeys & Engels, EY Belgium and BDO.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31 OVER-GRANT FIX. This entry already said the floor had been lowered to ~EUR 70k, but no min_income was declared, so a 35% exclusion applied at ANY salary. Wired min_income 70000, the exact statutory figure. Brussels single filer at EUR 69,999: tax 16,808 -> 29,015 (ordinary). OPEN QUESTION, not resolved here: the allowance is legally paid bovenop de bezoldiging (on top of gross remuneration), so reduction_pct 0.35 sits between the contract-gross reading (allowance is extra tax-free cash, we understate the break) and the total-package reading (taxable share is 1/1.35 = 74.07%, true carve-out 25.93%, we overstate it). Left at 0.35 pending a modeling decision." }, "tax.property_tax_rate_pct": { "source": "Onroerende voorheffing ~0.9% effective", "last_updated": "2026-05-26", "confidence": "medium" }, "housing._summary": { "source": "Numbeo + Immoweb Ghent 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "INAMI/RIZIV universal + private (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Flemish kinderopvang income-scaled; kleuterschool free 2.5+", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Ghent 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "De Lijn Omnipas EUR 62/mo", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Ghent", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "BRU->US economy 2025-26", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Groeipakket Flanders (regional; inherited)", "last_updated": "2026-05-26", "confidence": "medium", "note": "Belgian Groeipakket / allocations familiales: universal (regional), NOT means-tested (social supplements are extra)." }, "non_cash._summary": { "source": "Belgian labour law (national, inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Ghent: EUR 62,057, fetched live 2026-08-11 and recorded at sweep B F8c of docs/superpowers/research/2026-08-11-comp-index-sweep-b.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "EUR 62,057 / 0.867 = USD 71,577; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2863, R5-rounded down to 0.28. REPLACES 0.47 (2026-05-26), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference; the anchor is ALL-LEVELS over USD 250,000, so the old value overstated Ghent by 68%. The measured p50 was 0.609 of old index x anchor. Re-derived 2026-08-11 on the declared anchor. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.47, Flemish tech/university hub. CONFIDENCE LOW, and the reason is recorded rather than smoothed over: levels.fyi ranks Brussels (EUR 53,895) BELOW Antwerp (66,830), Ghent (62,057) and Leuven (58,586), which sweep B F8c calls implausible for the capital and EU-institution hub and attributes to sample composition (Brussels senior p25-p75 spread 2.5x against Helsinki 1.6x). All three values come from that one source on one basis, so the country is internally consistent, but its ORDERING is the weakest in the 2026-08-11 re-derivation." }, "tax.filing_basis": { "source": "PwC Worldwide Tax Summaries: Belgium, Individual: Sample personal income tax calculation (spouses taxed separately on all earned income / décumul). https://taxsummaries.pwc.com/belgium/individual/sample-personal-income-tax-calculation", "last_updated": "2026-07-26", "note": "BE-3 (closeout 2026-07-26): spouses taxed separately (décumul) though filing one joint return." }, "tax.professional_expense_rate": { "source": "PwC Worldwide Tax Summaries: Belgium, Individual: Deductions (lump-sum professional expenses = 30% of gross taxable income for employees). https://taxsummaries.pwc.com/belgium/individual/deductions", "last_updated": "2026-07-26", "note": "BE-1 (closeout 2026-07-26): 30% professional-expense forfait rate." }, "tax.professional_expense_cap": { "source": "PwC Worldwide Tax Summaries: Belgium, Individual: Deductions (30% forfait capped at EUR 5,930, AY2026/IY2025). https://taxsummaries.pwc.com/belgium/individual/deductions", "last_updated": "2026-07-26", "note": "BE-1 (closeout 2026-07-26): forfait cap EUR 5,930 (confirmed current cap; IY2026 indexation not yet published. TODO (when it publishes): apply the small uplift; the cap stays EUR 5,930 until then)." }, "tax.tax_free_allowance": { "source": "PwC BE: Belgian tax reform 2026 (quotité exemptée / belastingvrije som = EUR 11,180 for income year 2026, up from EUR 10,910). https://news.pwc.be/belgian-tax-reform-2026-tax-and-social-measures-of-the-programme-law/", "last_updated": "2026-07-26", "note": "BE-2 (closeout 2026-07-26): tax-free allowance for income year 2026." }, "tax.tax_free_allowance_credit_rate": { "source": "PwC BE: quotité exemptée credited at the bottom-bracket 25% rate (the allowance falls entirely within the first IPP band). https://news.pwc.be/belgian-tax-reform-2026-tax-and-social-measures-of-the-programme-law/", "last_updated": "2026-07-26", "note": "BE-2 (closeout 2026-07-26): credited at 25% → reduction ~EUR 2,795." }, "non_cash.parental_leave_weeks": { "source": "belgium.be (conge de maternite, 15 semaines); INAMI maternity amounts https://www.inami.fgov.be/fr/themes/grossesse-et-naissance/montants-et-plafonds-des-indemnites/salaries-chomeurs/montant-maximal-de-votre-indemnite-pendant-un-repos-de-maternite-salariee ; ONEM conges thematiques montants https://www.onem.be/documentation/montants/interruption-de-carriere-credit-temps/conges-thematiques", "last_updated": "2026-08-06", "confidence": "high", "note": "54 wk = repos de maternite 15 (6 prenatal of which 1 mandatory, 9 postnatal mandatory) + conge de naissance 20 days (4 wk, the other parent's own right) + conge parental 4 months PER PARENT via ONEM until the child is 12 (34.78 wk summed across both parents, which is why parental_leave_both_parents is true; Netherlands/Ireland precedent). Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. Was 15/82, maternity alone with the first-30-days rate applied to everything." }, "non_cash.parental_leave_paid_pct": { "source": "INAMI: 82% of uncapped gross for the first 30 days, then 75% of salary capped at EUR 186.79/day from 1 Mar 2026 (max EUR 140.09/day, 6 payable days/wk); conge de naissance 3 days at 100% employer-paid then 17 days at 82% capped", "last_updated": "2026-08-06", "confidence": "high", "note": "79 = blend across the 19.2 percentage-paid weeks: maternity 4.29 wk at 82 (uncapped) + 10.71 wk at 75 (capped, max EUR 840.54/wk) + naissance 0.6 wk at 100 + 3.4 wk at 82 (capped): (352 + 803 + 60 + 279) / 19.0 = 79. The EUR 186.79/day ceiling sits near EUR 58,000/yr, so the capped phases pay their daily maxima for an ICP." }, "non_cash.parental_leave_flat": { "source": "ONEM conges thematiques montants, index 01.03.2026: EUR 1,058.86 gross a month (EUR 951.60 net after the 11.11% withholding) for full-time interruption https://www.onem.be/documentation/montants/interruption-de-carriere-credit-temps/conges-thematiques", "last_updated": "2026-08-06", "confidence": "high", "note": "EUR 243.53/wk = 1,058.86 / 4.348. Four months per parent, so 8 months = 34.78 wk for the household. A flat sum regardless of salary, roughly 15-19% of an ICP wage, which is exactly what the flat block exists to price honestly." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "INAMI/RIZIV maximum a facturer 2026 income bands (top band EUR 2,194.20 per household, child sub-cap EUR 792.35) + Statbel household health expenditure", "last_updated": "2026-08-08", "confidence": "medium", "note": "The old EUR 450 was neither a maximum nor a sourced average, just a round number that landed near the right single-adult figure and was then applied flat to households. Belgium has no child exemption, so the curve scales with headcount, shaded down slightly because the maximum a facturer is a HOUSEHOLD ceiling that does not grow with family size. Note it caps tickets moderateurs only, not fee or room supplements, which are the larger half of Belgian out-of-pocket spend. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "INAMI/RIZIV maximum a facturer 2026 income bands (top band EUR 2,194.20 per household, child sub-cap EUR 792.35) + Statbel household health expenditure", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Same Belgian gap as Antwerp. The Eurostat Urban Audit value for BE003C is 34.0 on a 1991 basis and was deliberately rejected; adopting it would have moved this city by +6 on thirty-five-year-old evidence. OVG Vlaanderen 7 (2023-24) was CHECKED and REJECTED: its duration figures are DAY-LEVEL and ALL-PURPOSE (22.3 min, Flanders), which is the same unit error as SrV, so adopting it would have swapped one defect for another. FORWARD POINTER: OVG 8 is the source to watch, if it publishes duration by motive." }, "non_cash.sick_leave": { "source": "Arbeidsovereenkomstenwet (Employment Contracts Act, 3 July 1978) art. 52; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": "https://www.ejustice.just.fgov.be", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "SPF Finances, Epargne-pension https://fin.belgium.be/fr/particuliers/avantages-fiscaux/epargne-pension", "last_updated": "2026-08-11", "confidence": "high", "note": "No base deduction. Epargne-pension is a REDUCTION D'IMPOT applied after the tax is computed and never touches the taxable base: 30% on up to EUR 1,050 (max EUR 315) or 25% on up to EUR 1,350 (max EUR 337.50) for income year 2026. Epargne a long terme is also a credit (30%, ceiling 6% of net professional income plus EUR 183.6, absolute max EUR 2,450). Occupational 2nd-pillar (groepsverzekering) employee contributions DO reduce taxable pay but are scheme-specific and we model no employer scheme. computeBelgiumTaxes already nets the mandatory RSZ 13.07% separately, so nothing here double-counts." } }, "manchester": { "non_cash.parental_leave_flat": { "source": "Social Security Benefits Up-rating Order 2026 (SI 2026/148), arts. 9-10: for GBP 187.18 substitute GBP 194.32, applied to SMP, SPP, adoption pay, ShPP and neonatal pay https://www.legislation.gov.uk/uksi/2026/148/contents/made ; gov.uk https://www.gov.uk/maternity-pay-leave/pay", "last_updated": "2026-08-01", "confidence": "high", "note": "GBP 194.32 per week for tax year 2026/27, effective 5 April 2026 for SMP and 6 April 2026 for SPP and ShPP; the prior rate was GBP 187.18. Covers SMP weeks 7 to 39 (33 weeks) plus the 2 weeks of Statutory Paternity Pay. Statute pays the LOWER of the flat rate and 90% of average weekly earnings, so an ICP always receives the flat rate. Note: gov.uk/employers-maternity-pay-leave still showed the stale 187.18 on 2026-08-01; cite the employee-facing pages. The 2026/27 Lower Earnings Limit was not sourced and is not load-bearing, since our user clears it." }, "non_cash.parental_leave_paid_pct": { "source": "gov.uk Statutory Maternity Pay: the first 6 weeks are 90% of average weekly earnings before tax, with no cap https://www.gov.uk/maternity-pay-leave/pay", "last_updated": "2026-08-01", "confidence": "high", "note": "90% applies to the SIX percentage-paid weeks only. The other 35 weeks are flat-rate and live in non_cash.parental_leave_flat. The stored 40 was a hand-blend that folded the flat weeks in as though they tracked salary, and it also averaged across 13 unpaid weeks." }, "non_cash.parental_leave_weeks": { "source": "gov.uk Statutory Maternity Pay https://www.gov.uk/maternity-pay-leave/pay ; Statutory Paternity Pay https://www.gov.uk/paternity-pay-leave/pay ; Shared Parental Leave and Pay https://www.gov.uk/shared-parental-leave-and-pay ; unpaid parental leave https://www.gov.uk/parental-leave/entitlement", "last_updated": "2026-08-01", "confidence": "high", "note": "41 PAID wk = SMP 39 (6 at 90% of average weekly earnings plus 33 flat) + Statutory Paternity Pay 2 (flat). Was 52, which INCLUDED the 13 unpaid weeks of the maternity year and violated the field definition. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. Shared Parental Leave adds ZERO weeks: gov.uk requires the mother to take less than the 39 weeks of maternity pay and use the rest as ShPP, so it redistributes her untaken entitlement and leaves the household ceiling at 39. Excluded: the 13-week unpaid SMP tail; 18 unpaid weeks per parent of parental leave; Neonatal Care Pay, which is conditional on the baby needing care; occupational maternity pay, which is contractual rather than statutory." }, "tax.inbound_regime": { "source": "4-year FIG regime (from 2025-04-06) relieves FOREIGN income and gains only, not a UK salary: HMRC RFIG41000 https://www.gov.uk/hmrc-internal-manuals/residence-and-fig-regime-manual/rfig41000 . Overseas Workday Relief covers only earnings for duties NOT performed in the UK (EIM43565) and is capped from 2025-04-06 at the lower of 30% of relevant qualifying employment income or GBP 300,000 (EIM43600); conditions incl. 10 immediately preceding tax years of non-UK residence (EIM43560). https://www.gov.uk/hmrc-internal-manuals/employment-income-manual/eim43600", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: re-verified against HMRC manuals. No salary relief remains correct for 2026: a resident performing all duties in the UK gets nothing from FIG (foreign income only) and nothing from OWR (non-UK duties only). Note sharpened with the new GBP 300,000 / 30% OWR cap and the 10-year prior non-residence test. No percentage discount added, deliberately: OWR turns on a UK/non-UK workday split we never ask for, and the conservative default for a relocating employee is 100% UK workdays." }, "tax.federal_brackets": { "source": "UK 2026 income tax (England rates): peer-identical with London", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Manchester council tax Band D £1,968 ~0.52% effective of market value", "last_updated": "2026-05-26", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "VAT 20% (inherited from London baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Varbes + ONS + Numbeo Manchester 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "NHS universal (inherited from London baseline)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Private nursery; 15-30 free hours for 3-4yr; expensive UK childcare", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Manchester 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "Bee Network bus travelcard £80/mo", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Manchester", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "MAN->US economy 2025-26 (direct)", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "gov.uk Child Benefit, what you'll get https://www.gov.uk/child-benefit/what-youll-get ; HMRC rates by tax year https://www.gov.uk/government/publications/rates-and-allowances-tax-credits-child-benefit-and-guardians-allowance/tax-credits-child-benefit-and-guardians-allowance ; High Income Child Benefit Charge https://www.gov.uk/child-benefit-tax-charge (all fetched 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "STATUTORY additional-child rate, 2026-27 tax year: GBP 17.90 a week x 52 = 930.80 a year. The eldest or only child draws govt_benefits.child_allowance_eldest_annual instead. Means-tested by the ENGINE, not in this field: govt_benefits.child_allowance_taper carries the High Income Child Benefit Charge and computeGovtBenefits() withdraws the benefit per household on the higher earner's gross. Until 2026-09-12 this field stored 0 on the assumption that every published UK household's higher earner was over GBP 80,000. That holds for the USD 280,000 family scenario (London primary about GBP 142k, still 0 through the taper) and fails for the USD 150,000 two-earner persona (London higher earner GBP 74,063 keeps 30%, GBP 701.22 a year) and for any calculator user under the ceiling. Ledger decision 2026-09-12; research docs/superpowers/research/2026-09-12-child-benefit-audit-fixes.md." }, "govt_benefits.child_allowance_eldest_annual": { "source": "gov.uk Child Benefit, what you'll get https://www.gov.uk/child-benefit/what-youll-get ; HMRC rates by tax year https://www.gov.uk/government/publications/rates-and-allowances-tax-credits-child-benefit-and-guardians-allowance/tax-credits-child-benefit-and-guardians-allowance ; High Income Child Benefit Charge https://www.gov.uk/child-benefit-tax-charge (all fetched 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "Eldest or only child, 2026-27 tax year: GBP 27.05 a week x 52 = 1,406.60 a year. Unit: ANNUAL, from the statutory WEEKLY rate (HMRC pays every four weeks, 13 payments a year). Drawn by the first eligible child; every further child draws child_allowance_per_kid_annual. Same means-testing as that field: the taper block withdraws it." }, "govt_benefits.child_allowance_taper": { "source": "gov.uk Child Benefit, what you'll get https://www.gov.uk/child-benefit/what-youll-get ; HMRC rates by tax year https://www.gov.uk/government/publications/rates-and-allowances-tax-credits-child-benefit-and-guardians-allowance/tax-credits-child-benefit-and-guardians-allowance ; High Income Child Benefit Charge https://www.gov.uk/child-benefit-tax-charge (all fetched 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "High Income Child Benefit Charge: threshold GBP 60,000 of adjusted net income for tax years from 2024-25, 1% of the benefit for every GBP 200 over it, 100% at GBP 80,000 or more. Tested on the HIGHER earner (\"whoever has the higher income is responsible for paying the tax charge\"), never on the household total, so basis is higher_earner. The engine tests gross salary plus RSU as the conservative proxy for adjusted net income (ANI is net of pension contributions and Gift Aid, so it is lower and keeps more benefit); at the USD 150,000 persona that is a 70% charge on gross against about 48% on ANI. Whole-percent steps rounded down, as HMRC computes it. A deterministic statute, so no conservative-value tiebreak was applied to the BASIS; the gross proxy is the one conservative call inside it." }, "non_cash._summary": { "source": "UK labour law (national, inherited); 28 statutory leave incl bank holidays, 8 public holidays", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Manchester: GBP 53,650, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "GBP 53,650 / 0.742 = USD 72,305; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2892, R5-rounded down to 0.28. REPLACES 0.55 (2026-08-11), and the interim 0.38 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.55 was 0.7333 of London's 0.75; that ratio x London's re-derived 0.53 = 0.3887, R5-rounded down to 0.38. London's own figure: GBP 99,953 / 0.742 = USD 134,708, / USD 250,000 = 0.5388. This is a BASIS move, not a re-measurement: Manchester's rank inside United Kingdom is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.55, biggest UK tech hub outside London." }, "tax.paye_brackets": { "confidence": "high", "source": "HMRC income tax rates; additional-rate threshold 125,140 taxable + PA taper over 100k (gov.uk/income-tax-rates); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "tax.filing_basis": { "source": "HMRC - independent taxation of spouses since 1990", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "childcare.subsidy_per_kid_annual": { "source": "https://www.gov.uk/free-childcare-if-working", "last_updated": "2026-08-05", "note": "ZEROED 2026-08-05 (CC5): the free-hours entitlement is a provider-funded fee reduction, never household cash, so under the framework it belongs inside preschool_monthly, same defect class as the 08-02 double-subtraction batch. Working: docs/superpowers/research/2026-08-05-cc5-brazil-vienna-sourcing.md", "confidence": "high" }, "childcare.preschool_monthly": { "source": "Coram Family and Childcare Survey 2026 (regional tables, derivation in the research file)", "last_updated": "2026-08-05", "confidence": "high", "note": "Corrected 2026-08-05 from a too-low figure to the Coram-derived regional out-of-pocket for a 3-4yo with the surviving entitlement; the GBP 100k per-parent cliff kills 30h/TFC for every published kid-bearing scenario (England), so 15h universal only. Value 995." }, "childcare.infant_monthly": { "source": "Coram Family and Childcare Survey 2026 (regional 50h tables x51/12)", "last_updated": "2026-08-05", "confidence": "high", "note": "Corrected 2026-08-05: full price for an under-3 (no entitlement at ICP income). Value 1510." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "ONS Healthcare expenditure, UK Health Accounts 2023 and 2024 https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthcaresystem/bulletins/ukhealthaccounts/2023and2024 (household out-of-pocket GBP 46bn in 2024, of which ONS states ~38% is long-term care) + NHS England prescription charge and prepayment certificate", "last_updated": "2026-08-08", "confidence": "medium", "note": "Almost NONE of this is NHS cost-sharing, and the _meta note has to say so: realized England prescription charges are only about GBP 10-12 per capita because ~89% of items are dispensed free. The figure is overwhelmingly OTC medicines, self-pay private treatment against NHS waiting lists, and private physiotherapy. The old GBP 400 sat ABOVE the entire statutory ceiling at every household size (England prepayment certificate GBP 114.50 per adult, family of four GBP 229 since children are free, and GBP 0 in Scotland, Wales and Northern Ireland). One UK-wide value is retained this pass; Edinburgh's free prescriptions are a real but small (GBP 10-15/yr) difference and the city-level split is queued. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "ONS Healthcare expenditure, UK Health Accounts 2023 and 2024 https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthcaresystem/bulletins/ukhealthaccounts/2023and2024 (household out-of-pocket GBP 46bn in 2024, of which ONS states ~38% is long-term care) + NHS England prescription charge and prepayment certificate", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UK Department for Transport, National Travel Survey tables NTS9914a and NTS9914b, North West ITL1 region and Urban Conurbation class. https://assets.publishing.service.gov.uk/media/68a4371132d2c63f869343d0/nts9914.ods", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average trip duration by trip purpose and region of residence, commuting", "year": "2024", "note": "DERIVED. North West two-year mean is 27.7, but that region includes Cumbria, Lancashire and Merseyside and so understates Greater Manchester; the conurbation adjustment from NTS9914b brings it to about 30." }, "non_cash.sick_leave": { "source": "Social Security Contributions and Benefits Act 1992, s.151 (employer's duty to pay; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": "https://www.gov.uk/statutory-sick-pay/eligibility", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "HMRC, Pension schemes rates and allowances https://www.gov.uk/government/publications/rates-and-allowances-pension-schemes/pension-schemes-rates ; Pensions Tax Manual PTM057100 https://www.gov.uk/hmrc-internal-manuals/pensions-tax-manual/ptm057100 ; https://www.gov.uk/tax-on-your-private-pension/annual-allowance", "last_updated": "2026-08-11", "confidence": "high", "note": "Annual allowance, 2025-26 and 2026-27 (identical on every figure). A COMBINED input ceiling covering everything paid in by the member or anyone else including the employer, enforced by the annual allowance charge, plus the separate rule that relief cannot exceed 100% of relevant UK earnings. Taper is GBP 1 per GBP 2 of ADJUSTED income above 260,000 down to a 10,000 floor; our gross understates adjusted income (which adds employer contributions) so the taper bites later than reality. No less_employer_pct: auto-enrolment minimum is 3% of qualifying earnings, about GBP 1,321, and anything larger would be an assumption. Never binds at our incomes." } }, "birmingham": { "non_cash.parental_leave_flat": { "source": "Social Security Benefits Up-rating Order 2026 (SI 2026/148), arts. 9-10: for GBP 187.18 substitute GBP 194.32, applied to SMP, SPP, adoption pay, ShPP and neonatal pay https://www.legislation.gov.uk/uksi/2026/148/contents/made ; gov.uk https://www.gov.uk/maternity-pay-leave/pay", "last_updated": "2026-08-01", "confidence": "high", "note": "GBP 194.32 per week for tax year 2026/27, effective 5 April 2026 for SMP and 6 April 2026 for SPP and ShPP; the prior rate was GBP 187.18. Covers SMP weeks 7 to 39 (33 weeks) plus the 2 weeks of Statutory Paternity Pay. Statute pays the LOWER of the flat rate and 90% of average weekly earnings, so an ICP always receives the flat rate. Note: gov.uk/employers-maternity-pay-leave still showed the stale 187.18 on 2026-08-01; cite the employee-facing pages. The 2026/27 Lower Earnings Limit was not sourced and is not load-bearing, since our user clears it." }, "non_cash.parental_leave_paid_pct": { "source": "gov.uk Statutory Maternity Pay: the first 6 weeks are 90% of average weekly earnings before tax, with no cap https://www.gov.uk/maternity-pay-leave/pay", "last_updated": "2026-08-01", "confidence": "high", "note": "90% applies to the SIX percentage-paid weeks only. The other 35 weeks are flat-rate and live in non_cash.parental_leave_flat. The stored 40 was a hand-blend that folded the flat weeks in as though they tracked salary, and it also averaged across 13 unpaid weeks." }, "non_cash.parental_leave_weeks": { "source": "gov.uk Statutory Maternity Pay https://www.gov.uk/maternity-pay-leave/pay ; Statutory Paternity Pay https://www.gov.uk/paternity-pay-leave/pay ; Shared Parental Leave and Pay https://www.gov.uk/shared-parental-leave-and-pay ; unpaid parental leave https://www.gov.uk/parental-leave/entitlement", "last_updated": "2026-08-01", "confidence": "high", "note": "41 PAID wk = SMP 39 (6 at 90% of average weekly earnings plus 33 flat) + Statutory Paternity Pay 2 (flat). Was 52, which INCLUDED the 13 unpaid weeks of the maternity year and violated the field definition. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. Shared Parental Leave adds ZERO weeks: gov.uk requires the mother to take less than the 39 weeks of maternity pay and use the rest as ShPP, so it redistributes her untaken entitlement and leaves the household ceiling at 39. Excluded: the 13-week unpaid SMP tail; 18 unpaid weeks per parent of parental leave; Neonatal Care Pay, which is conditional on the baby needing care; occupational maternity pay, which is contractual rather than statutory." }, "tax.inbound_regime": { "source": "4-year FIG regime (from 2025-04-06) relieves FOREIGN income and gains only, not a UK salary: HMRC RFIG41000 https://www.gov.uk/hmrc-internal-manuals/residence-and-fig-regime-manual/rfig41000 . Overseas Workday Relief covers only earnings for duties NOT performed in the UK (EIM43565) and is capped from 2025-04-06 at the lower of 30% of relevant qualifying employment income or GBP 300,000 (EIM43600); conditions incl. 10 immediately preceding tax years of non-UK residence (EIM43560). https://www.gov.uk/hmrc-internal-manuals/employment-income-manual/eim43600", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: re-verified against HMRC manuals. No salary relief remains correct for 2026: a resident performing all duties in the UK gets nothing from FIG (foreign income only) and nothing from OWR (non-UK duties only). Note sharpened with the new GBP 300,000 / 30% OWR cap and the 10-year prior non-residence test. No percentage discount added, deliberately: OWR turns on a UK/non-UK workday split we never ask for, and the conservative default for a relocating employee is 100% UK workdays." }, "tax.federal_brackets": { "source": "UK 2026 income tax (England rates): peer-identical with London", "last_updated": "2026-05-26", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Birmingham council tax Band D £1,960 ~0.65% effective of market value", "last_updated": "2026-05-26", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "VAT 20% (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "housing._summary": { "source": "Investropa + ONS + Numbeo Birmingham 2026", "last_updated": "2026-05-26", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "NHS universal (inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "childcare._summary": { "source": "Private nursery; free-hours schemes for 3-4yr", "last_updated": "2026-05-26", "confidence": "medium" }, "food._summary": { "source": "Numbeo Birmingham 2026", "last_updated": "2026-05-26", "confidence": "medium" }, "transit._summary": { "source": "NX Bus / West Midlands travelcard £64/mo", "last_updated": "2026-05-26", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Birmingham", "last_updated": "2026-05-26", "confidence": "medium" }, "travel_home._summary": { "source": "BHX->US economy 2025-26 (via hub)", "last_updated": "2026-05-26", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "gov.uk Child Benefit, what you'll get https://www.gov.uk/child-benefit/what-youll-get ; HMRC rates by tax year https://www.gov.uk/government/publications/rates-and-allowances-tax-credits-child-benefit-and-guardians-allowance/tax-credits-child-benefit-and-guardians-allowance ; High Income Child Benefit Charge https://www.gov.uk/child-benefit-tax-charge (all fetched 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "STATUTORY additional-child rate, 2026-27 tax year: GBP 17.90 a week x 52 = 930.80 a year. The eldest or only child draws govt_benefits.child_allowance_eldest_annual instead. Means-tested by the ENGINE, not in this field: govt_benefits.child_allowance_taper carries the High Income Child Benefit Charge and computeGovtBenefits() withdraws the benefit per household on the higher earner's gross. Until 2026-09-12 this field stored 0 on the assumption that every published UK household's higher earner was over GBP 80,000. That holds for the USD 280,000 family scenario (London primary about GBP 142k, still 0 through the taper) and fails for the USD 150,000 two-earner persona (London higher earner GBP 74,063 keeps 30%, GBP 701.22 a year) and for any calculator user under the ceiling. Ledger decision 2026-09-12; research docs/superpowers/research/2026-09-12-child-benefit-audit-fixes.md." }, "govt_benefits.child_allowance_eldest_annual": { "source": "gov.uk Child Benefit, what you'll get https://www.gov.uk/child-benefit/what-youll-get ; HMRC rates by tax year https://www.gov.uk/government/publications/rates-and-allowances-tax-credits-child-benefit-and-guardians-allowance/tax-credits-child-benefit-and-guardians-allowance ; High Income Child Benefit Charge https://www.gov.uk/child-benefit-tax-charge (all fetched 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "Eldest or only child, 2026-27 tax year: GBP 27.05 a week x 52 = 1,406.60 a year. Unit: ANNUAL, from the statutory WEEKLY rate (HMRC pays every four weeks, 13 payments a year). Drawn by the first eligible child; every further child draws child_allowance_per_kid_annual. Same means-testing as that field: the taper block withdraws it." }, "govt_benefits.child_allowance_taper": { "source": "gov.uk Child Benefit, what you'll get https://www.gov.uk/child-benefit/what-youll-get ; HMRC rates by tax year https://www.gov.uk/government/publications/rates-and-allowances-tax-credits-child-benefit-and-guardians-allowance/tax-credits-child-benefit-and-guardians-allowance ; High Income Child Benefit Charge https://www.gov.uk/child-benefit-tax-charge (all fetched 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "High Income Child Benefit Charge: threshold GBP 60,000 of adjusted net income for tax years from 2024-25, 1% of the benefit for every GBP 200 over it, 100% at GBP 80,000 or more. Tested on the HIGHER earner (\"whoever has the higher income is responsible for paying the tax charge\"), never on the household total, so basis is higher_earner. The engine tests gross salary plus RSU as the conservative proxy for adjusted net income (ANI is net of pension contributions and Gift Aid, so it is lower and keeps more benefit); at the USD 150,000 persona that is a 70% charge on gross against about 48% on ANI. Whole-percent steps rounded down, as HMRC computes it. A deterministic statute, so no conservative-value tiebreak was applied to the BASIS; the gross proxy is the one conservative call inside it." }, "non_cash._summary": { "source": "UK labour law (national, inherited)", "last_updated": "2026-05-26", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Birmingham: GBP 47,726, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "GBP 47,726 / 0.742 = USD 64,321; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2573, R5-rounded down to 0.25. REPLACES 0.52 (2026-08-11), and the interim 0.36 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.52 was 0.6933 of London's 0.75; that ratio x London's re-derived 0.53 = 0.3675, R5-rounded down to 0.36. London's own figure: GBP 99,953 / 0.742 = USD 134,708, / USD 250,000 = 0.5388. This is a BASIS move, not a re-measurement: Birmingham's rank inside United Kingdom is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.52, UK 2nd city, growing finance/tech." }, "tax.paye_brackets": { "confidence": "high", "source": "HMRC income tax rates; additional-rate threshold 125,140 taxable + PA taper over 100k (gov.uk/income-tax-rates); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "tax.filing_basis": { "source": "HMRC - independent taxation of spouses since 1990", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "childcare.subsidy_per_kid_annual": { "source": "https://www.gov.uk/free-childcare-if-working", "last_updated": "2026-08-05", "note": "ZEROED 2026-08-05 (CC5): the free-hours entitlement is a provider-funded fee reduction, never household cash, so under the framework it belongs inside preschool_monthly, same defect class as the 08-02 double-subtraction batch. Working: docs/superpowers/research/2026-08-05-cc5-brazil-vienna-sourcing.md", "confidence": "high" }, "childcare.preschool_monthly": { "source": "Coram Family and Childcare Survey 2026 (regional tables, derivation in the research file)", "last_updated": "2026-08-05", "confidence": "high", "note": "Corrected 2026-08-05 from a too-low figure to the Coram-derived regional out-of-pocket for a 3-4yo with the surviving entitlement; the GBP 100k per-parent cliff kills 30h/TFC for every published kid-bearing scenario (England), so 15h universal only. Value 1095." }, "childcare.infant_monthly": { "source": "Coram Family and Childcare Survey 2026 (regional 50h tables x51/12)", "last_updated": "2026-08-05", "confidence": "high", "note": "Corrected 2026-08-05: full price for an under-3 (no entitlement at ICP income). Value 1485." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "ONS Healthcare expenditure, UK Health Accounts 2023 and 2024 https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthcaresystem/bulletins/ukhealthaccounts/2023and2024 (household out-of-pocket GBP 46bn in 2024, of which ONS states ~38% is long-term care) + NHS England prescription charge and prepayment certificate", "last_updated": "2026-08-08", "confidence": "medium", "note": "Almost NONE of this is NHS cost-sharing, and the _meta note has to say so: realized England prescription charges are only about GBP 10-12 per capita because ~89% of items are dispensed free. The figure is overwhelmingly OTC medicines, self-pay private treatment against NHS waiting lists, and private physiotherapy. The old GBP 400 sat ABOVE the entire statutory ceiling at every household size (England prepayment certificate GBP 114.50 per adult, family of four GBP 229 since children are free, and GBP 0 in Scotland, Wales and Northern Ireland). One UK-wide value is retained this pass; Edinburgh's free prescriptions are a real but small (GBP 10-15/yr) difference and the city-level split is queued. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "ONS Healthcare expenditure, UK Health Accounts 2023 and 2024 https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthcaresystem/bulletins/ukhealthaccounts/2023and2024 (household out-of-pocket GBP 46bn in 2024, of which ONS states ~38% is long-term care) + NHS England prescription charge and prepayment certificate", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UK Department for Transport, National Travel Survey table NTS9914a, West Midlands ITL1 region. https://assets.publishing.service.gov.uk/media/68a4371132d2c63f869343d0/nts9914.ods", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average trip duration by trip purpose and region of residence, commuting", "year": "2024", "note": "An NTS trip is one way, built from stages, so walking and waiting are inside it and no halving applies. Two-year mean of the West Midlands region (30.5 in 2023, 27.5 in 2024) is 29.0. The region is wider than the metro and includes rural Shropshire and Herefordshire, so it is biased short as a metro proxy; DfT's own Urban Conurbation class ran 34.2 against 29.7 for all areas in 2023." }, "non_cash.sick_leave": { "source": "Social Security Contributions and Benefits Act 1992, s.151 (employer's duty to pay; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": "https://www.gov.uk/statutory-sick-pay/eligibility", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "HMRC, Pension schemes rates and allowances https://www.gov.uk/government/publications/rates-and-allowances-pension-schemes/pension-schemes-rates ; Pensions Tax Manual PTM057100 https://www.gov.uk/hmrc-internal-manuals/pensions-tax-manual/ptm057100 ; https://www.gov.uk/tax-on-your-private-pension/annual-allowance", "last_updated": "2026-08-11", "confidence": "high", "note": "Annual allowance, 2025-26 and 2026-27 (identical on every figure). A COMBINED input ceiling covering everything paid in by the member or anyone else including the employer, enforced by the annual allowance charge, plus the separate rule that relief cannot exceed 100% of relevant UK earnings. Taper is GBP 1 per GBP 2 of ADJUSTED income above 260,000 down to a 10,000 floor; our gross understates adjusted income (which adds employer contributions) so the taper bites later than reality. No less_employer_pct: auto-enrolment minimum is 3% of qualifying earnings, about GBP 1,321, and anything larger would be an assumption. Never binds at our incomes." } }, "calgary": { "tax.filing_basis": { "source": "PwC WWTS - Canada (Individual, Tax administration): married taxpayers file separately; joint returns are not allowed.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Canada 2025 federal income tax + CPP/EI: peer-identical with Toronto", "last_updated": "2026-05-27", "confidence": "high" }, "tax.provincial_brackets": { "source": "Alberta 2026 brackets/BPA (taxtips.ca/taxrates/ab.htm); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "City of Calgary 2025 residential mill rate 6.1803 = 0.618%", "last_updated": "2026-05-27", "confidence": "high" }, "housing._summary": { "source": "WOWA + Zumper + Numbeo Calgary 2026", "last_updated": "2026-05-27", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Canada Medicare universal (inherited from Toronto baseline)", "last_updated": "2026-05-27", "confidence": "high" }, "childcare._summary": { "source": "Alberta CWELCC flat ~$326/mo 2025 (moving to $10/day)", "last_updated": "2026-05-27", "confidence": "medium" }, "food._summary": { "source": "Numbeo Calgary 2026", "last_updated": "2026-05-27", "confidence": "medium" }, "transit._summary": { "source": "Calgary Transit monthly pass $118 (2025)", "last_updated": "2026-05-27", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Calgary", "last_updated": "2026-05-27", "confidence": "medium" }, "travel_home._summary": { "source": "YYC->US economy 2025-26", "last_updated": "2026-05-27", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "canada.ca Canada Child Benefit calculation sheet (July 2025-June 2026)", "last_updated": "2026-07-17", "note": "CCB is means-tested on adjusted family net income: a 2-child mixed-age family reaches $0 by ~CAD 236k AFNI. The published CAD 231k-339k households receive ~$0 (residual <$150/kid only at the very low end), and Ontario/Alberta have no high-income floor. Set to 0 (was a flat 3,000)." }, "non_cash._summary": { "source": "Canada labour law (federal/provincial, inherited)", "last_updated": "2026-05-27", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Calgary: CAD 114,345, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "CAD 114,345 / 1.4 = USD 81,675; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3267, R5-rounded down to 0.32. REPLACES 0.72 (2026-08-11), and the interim 0.38 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.72 was 0.9600 of Toronto's 0.75; that ratio x Toronto's re-derived 0.4 = 0.3840, R5-rounded down to 0.38. Toronto's own figure: CAD 142,630 / 1.4 = USD 101,879, / USD 250,000 = 0.4075. This is a BASIS move, not a re-measurement: Calgary's rank inside Canada is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.72, energy + growing tech; Alberta low-tax." }, "tax.sales_tax_rate": { "confidence": "high", "source": "Alberta levies GST 5% only, no PST (alberta.ca/taxes-levies-overview); fixed copy-paste from Toronto, tax audit 2026-07-14", "last_updated": "2026-07-17" }, "tax.bpa_phaseout": { "source": "Federal BPA supplement phases 16,452 -> 14,829 over net income 181,440-258,482 (CRA 2026); https://www.canada.ca/en/revenue-agency/programs/about-canada-revenue-agency-cra/federal-government-budgets/basic-personal-amount.html; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "BPA modeled as a non-refundable credit at the lowest bracket rate with this federal supplement phase-out." }, "childcare.subsidy_per_kid_annual": { "source": "CWELCC (Canada-Wide Early Learning and Child Care) provincial fee schedules; double-subtraction correction 2026-08-02", "last_updated": "2026-08-02", "confidence": "medium", "note": "Set to 0 on 2026-08-02, was CAD 3,900. The old value was the CWELCC fee reduction. CWELCC is delivered as a lower posted fee at a participating centre, not as cash to the household, and childcare.preschool_monthly already carries the reduced rate, so the reduction was netted twice. CAD 326/month IS the post-CWELCC Alberta rate, and the CAD 3,900 subsidy left an annual bill of CAD 12. This was the clearest proof of the double-subtraction anywhere in the data. This field holds CASH THAT REACHES THE HOUSEHOLD. Where the monthly rate it nets against is already the price the household actually pays, the value is 0 and the relief is described here instead. See docs/superpowers/research/2026-08-02-childcare-double-subtraction.md. Open follow-up: the Calgary monthly rate itself is a blend of the capped and uncapped rates and still owes its own sourcing pass." }, "non_cash.parental_leave_weeks": { "source": "Employment Insurance Act ss. 22-23: maternity 15 wk + standard parental up to 35 wk per parent (40 when shared) https://www.canada.ca/en/services/benefits/ei/ei-maternity-parental.html", "last_updated": "2026-08-06", "confidence": "high", "note": "50 = EI maternity 15 (mother) + standard parental 35, a shared pot counted once at the one-parent maximum. The 5 extra weeks that exist only when the second parent also claims are excluded as conditional on both parents' behaviour (same ruling as Portugal's sharing bonus). The extended option (up to 61 wk one parent, at 33%) was rejected: near-identical capped dollars stretched over 26 more weeks would inflate the weeks column. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. Quebec runs QPIP instead, so montreal legitimately diverges (city-varying by design in data/_national_fields.json)." }, "non_cash.parental_leave_paid_pct": { "source": "canada.ca, EI maximum insurable earnings for 2026: CAD 68,900 https://www.canada.ca/en/employment-social-development/programs/ei/ei-list/ei-employers/premium-reduction-program/2026-maximum-insurable-earnings.html", "last_updated": "2026-08-06", "confidence": "high", "note": "55% of insurable earnings for both maternity and standard parental weeks, capped at the 2026 MIE of CAD 68,900, max CAD 729/wk. Against a CAD 130,000 Toronto tech salary the cap makes the effective rate about 29%; the 55 is real only below the MIE. Headline kept per field convention; the capped-percentage shape has no honest flat-block representation, so the arithmetic lives here." }, "housing.homeowners_insurance_pct_of_value": { "source": "Provincial average premium (rates.ca Q2 2026, MyChoice city data, IBC) over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "CIHI national health expenditure household out-of-pocket + provincial drug plans: Quebec RAMQ covers a registered parent's children free, Ontario Trillium and BC Fair PharmaCare are income-deductible-based", "last_updated": "2026-08-08", "confidence": "medium", "note": "Medically necessary physician and hospital care is free everywhere, so the whole of this field is prescription drugs and allied health, which is PROVINCIAL, and the existing per-city spread is kept for that reason. Employer drug plans are a benefit and a premium, not out-of-pocket. Montreal carries a much smaller child increment than the other cities because of the RAMQ free-child rule. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "CIHI national health expenditure household out-of-pocket + provincial drug plans: Quebec RAMQ covers a registered parent's children free, Ontario Trillium and BC Fair PharmaCare are income-deductible-based", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Statistics Canada, average commute times for the 15 largest census metropolitan areas, May 2025 (The Daily, 2025-08-26, Table 4). https://www150.statcan.gc.ca/n1/daily-quotidien/250826/t004a-eng.htm", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Commuting duration, one way", "year": "2025", "note": "Calgary CMA 27.7. StatCan writes '(one way)' into its own definition and excludes people who work most hours at home. Previous value tracked the COVID-depressed May 2021 Census." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "CRA, MP, DB, RRSP, DPSP, ALDA, TFSA limits and the YMPE https://www.canada.ca/en/revenue-agency/services/tax/registered-plans-administrators/pspa/mp-rrsp-dpsp-tfsa-limits-ympe.html ; https://www.canada.ca/en/revenue-agency/services/tax/individuals/topics/rrsps-related-plans/contributing-a-rrsp-prpp/contributions-affect-your-rrsp-prpp-deduction-limit.html", "last_updated": "2026-08-11", "confidence": "high", "note": "RRSP deduction limit (line 20800), 2026: 18% of prior-year earned income to an annual dollar limit. Pinned at 33,810 via the statutory identity RRSP limit year X = money-purchase limit year X-1 (CRA confirms 2025 MP = 33,810). The PENSION ADJUSTMENT is deliberately NOT modeled, so this is the no-workplace-DB case; a DB-plan member has materially less room. Carry-forward of unused room is also unmodeled. Never binds at our incomes; the 18% governs below CAD 187,833." } }, "ottawa": { "tax.filing_basis": { "source": "PwC WWTS - Canada (Individual, Tax administration): married taxpayers file separately; joint returns are not allowed.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Canada 2025 federal income tax + CPP/EI: peer-identical with Toronto", "last_updated": "2026-05-27", "confidence": "high" }, "tax.provincial_brackets": { "source": "Ontario 2026 brackets/BPA (taxtips.ca/taxrates/on.htm); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Ottawa 2026 residential mill rate 1.2271%", "last_updated": "2026-05-27", "confidence": "high" }, "housing._summary": { "source": "WOWA + DoorInsight + Numbeo Ottawa 2026", "last_updated": "2026-05-27", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Canada Medicare universal (inherited)", "last_updated": "2026-05-27", "confidence": "high" }, "childcare._summary": { "source": "Ontario CWELCC ~$22/day cap (City of Ottawa)", "last_updated": "2026-05-27", "confidence": "medium" }, "food._summary": { "source": "Numbeo Ottawa 2026", "last_updated": "2026-05-27", "confidence": "medium" }, "transit._summary": { "source": "OC Transpo monthly PRESTO pass $125.50", "last_updated": "2026-05-27", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Ottawa", "last_updated": "2026-05-27", "confidence": "medium" }, "travel_home._summary": { "source": "YOW->US economy 2025-26", "last_updated": "2026-05-27", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "canada.ca Canada Child Benefit calculation sheet (July 2025-June 2026)", "last_updated": "2026-07-17", "note": "CCB is means-tested on adjusted family net income: a 2-child mixed-age family reaches $0 by ~CAD 236k AFNI. The published CAD 231k-339k households receive ~$0 (residual <$150/kid only at the very low end), and Ontario/Alberta have no high-income floor. Set to 0 (was a flat 3,000)." }, "non_cash._summary": { "source": "Canada labour law (federal/Ontario, inherited)", "last_updated": "2026-05-27", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Ottawa: CAD 116,334, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "CAD 116,334 / 1.4 = USD 83,096; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3324, R5-rounded down to 0.33. REPLACES 0.71 (2026-08-11), and the interim 0.37 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.71 was 0.9467 of Toronto's 0.75; that ratio x Toronto's re-derived 0.4 = 0.3787, R5-rounded down to 0.37. Toronto's own figure: CAD 142,630 / 1.4 = USD 101,879, / USD 250,000 = 0.4075. This is a BASIS move, not a re-measurement: Ottawa's rank inside Canada is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.71, federal govt + Kanata tech corridor." }, "tax.surtax_threshold_1": { "source": "Ontario surtax 20% of basic ON tax over 5,818 (2026); https://www.taxtips.ca/taxrates/on.htm; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "First Ontario surtax threshold (20%)." }, "tax.surtax_threshold_2": { "source": "Ontario surtax +36% of basic ON tax over 7,446 (2026); https://www.taxtips.ca/taxrates/on.htm; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "Second Ontario surtax threshold (36%)." }, "tax.health_premium": { "source": "Ontario Health Premium, stepped income-tested levy capped at 900; https://www.ontario.ca/page/health-premium; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "Gates the stepped OHP (600/750/900 plateaus) in computeCanadaTaxes." }, "tax.bpa_phaseout": { "source": "Federal BPA supplement phases 16,452 -> 14,829 over net income 181,440-258,482 (CRA 2026); https://www.canada.ca/en/revenue-agency/programs/about-canada-revenue-agency-cra/federal-government-budgets/basic-personal-amount.html; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "BPA modeled as a non-refundable credit at the lowest bracket rate with this federal supplement phase-out." }, "childcare.subsidy_per_kid_annual": { "source": "CWELCC (Canada-Wide Early Learning and Child Care) provincial fee schedules; double-subtraction correction 2026-08-02", "last_updated": "2026-08-02", "confidence": "medium", "note": "Set to 0 on 2026-08-02, was CAD 5,500. The old value was the CWELCC fee reduction. CWELCC is delivered as a lower posted fee at a participating centre, not as cash to the household, and childcare.preschool_monthly already carries the reduced rate, so the reduction was netted twice. The modelled CAD 420/month is already below Ontario’s CAD 22/day cap; netting CAD 5,500 on top drove the annual bill negative before the floor caught it. This field holds CASH THAT REACHES THE HOUSEHOLD. Where the monthly rate it nets against is already the price the household actually pays, the value is 0 and the relief is described here instead. See docs/superpowers/research/2026-08-02-childcare-double-subtraction.md. Open follow-up: the Ottawa monthly rate itself is a blend of the capped and uncapped rates and still owes its own sourcing pass." }, "non_cash.parental_leave_weeks": { "source": "Employment Insurance Act ss. 22-23: maternity 15 wk + standard parental up to 35 wk per parent (40 when shared) https://www.canada.ca/en/services/benefits/ei/ei-maternity-parental.html", "last_updated": "2026-08-06", "confidence": "high", "note": "50 = EI maternity 15 (mother) + standard parental 35, a shared pot counted once at the one-parent maximum. The 5 extra weeks that exist only when the second parent also claims are excluded as conditional on both parents' behaviour (same ruling as Portugal's sharing bonus). The extended option (up to 61 wk one parent, at 33%) was rejected: near-identical capped dollars stretched over 26 more weeks would inflate the weeks column. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. Quebec runs QPIP instead, so montreal legitimately diverges (city-varying by design in data/_national_fields.json)." }, "non_cash.parental_leave_paid_pct": { "source": "canada.ca, EI maximum insurable earnings for 2026: CAD 68,900 https://www.canada.ca/en/employment-social-development/programs/ei/ei-list/ei-employers/premium-reduction-program/2026-maximum-insurable-earnings.html", "last_updated": "2026-08-06", "confidence": "high", "note": "55% of insurable earnings for both maternity and standard parental weeks, capped at the 2026 MIE of CAD 68,900, max CAD 729/wk. Against a CAD 130,000 Toronto tech salary the cap makes the effective rate about 29%; the 55 is real only below the MIE. Headline kept per field convention; the capped-percentage shape has no honest flat-block representation, so the arithmetic lives here." }, "housing.homeowners_insurance_pct_of_value": { "source": "Provincial average premium (rates.ca Q2 2026, MyChoice city data, IBC) over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "CIHI national health expenditure household out-of-pocket + provincial drug plans: Quebec RAMQ covers a registered parent's children free, Ontario Trillium and BC Fair PharmaCare are income-deductible-based", "last_updated": "2026-08-08", "confidence": "medium", "note": "Medically necessary physician and hospital care is free everywhere, so the whole of this field is prescription drugs and allied health, which is PROVINCIAL, and the existing per-city spread is kept for that reason. Employer drug plans are a benefit and a premium, not out-of-pocket. Montreal carries a much smaller child increment than the other cities because of the RAMQ free-child rule. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "CIHI national health expenditure household out-of-pocket + provincial drug plans: Quebec RAMQ covers a registered parent's children free, Ontario Trillium and BC Fair PharmaCare are income-deductible-based", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Statistics Canada, average commute times for the 15 largest CMAs, May 2025 (The Daily, 2025-08-26, Table 4). https://www150.statcan.gc.ca/n1/daily-quotidien/250826/t004a-eng.htm", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Commuting duration, one way", "year": "2025", "note": "Ottawa-Gatineau CMA 26.4, spanning Ontario and Quebec, which is the correct metro geography for this slug." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "CRA, MP, DB, RRSP, DPSP, ALDA, TFSA limits and the YMPE https://www.canada.ca/en/revenue-agency/services/tax/registered-plans-administrators/pspa/mp-rrsp-dpsp-tfsa-limits-ympe.html ; https://www.canada.ca/en/revenue-agency/services/tax/individuals/topics/rrsps-related-plans/contributing-a-rrsp-prpp/contributions-affect-your-rrsp-prpp-deduction-limit.html", "last_updated": "2026-08-11", "confidence": "high", "note": "RRSP deduction limit (line 20800), 2026: 18% of prior-year earned income to an annual dollar limit. Pinned at 33,810 via the statutory identity RRSP limit year X = money-purchase limit year X-1 (CRA confirms 2025 MP = 33,810). The PENSION ADJUSTMENT is deliberately NOT modeled, so this is the no-workplace-DB case; a DB-plan member has materially less room. Carry-forward of unused room is also unmodeled. Never binds at our incomes; the 18% governs below CAD 187,833." } }, "edinburgh": { "non_cash.parental_leave_flat": { "source": "Social Security Benefits Up-rating Order 2026 (SI 2026/148), arts. 9-10: for GBP 187.18 substitute GBP 194.32, applied to SMP, SPP, adoption pay, ShPP and neonatal pay https://www.legislation.gov.uk/uksi/2026/148/contents/made ; gov.uk https://www.gov.uk/maternity-pay-leave/pay", "last_updated": "2026-08-01", "confidence": "high", "note": "GBP 194.32 per week for tax year 2026/27, effective 5 April 2026 for SMP and 6 April 2026 for SPP and ShPP; the prior rate was GBP 187.18. Covers SMP weeks 7 to 39 (33 weeks) plus the 2 weeks of Statutory Paternity Pay. Statute pays the LOWER of the flat rate and 90% of average weekly earnings, so an ICP always receives the flat rate. Note: gov.uk/employers-maternity-pay-leave still showed the stale 187.18 on 2026-08-01; cite the employee-facing pages. The 2026/27 Lower Earnings Limit was not sourced and is not load-bearing, since our user clears it." }, "non_cash.parental_leave_paid_pct": { "source": "gov.uk Statutory Maternity Pay: the first 6 weeks are 90% of average weekly earnings before tax, with no cap https://www.gov.uk/maternity-pay-leave/pay", "last_updated": "2026-08-01", "confidence": "high", "note": "90% applies to the SIX percentage-paid weeks only. The other 35 weeks are flat-rate and live in non_cash.parental_leave_flat. The stored 40 was a hand-blend that folded the flat weeks in as though they tracked salary, and it also averaged across 13 unpaid weeks." }, "non_cash.parental_leave_weeks": { "source": "gov.uk Statutory Maternity Pay https://www.gov.uk/maternity-pay-leave/pay ; Statutory Paternity Pay https://www.gov.uk/paternity-pay-leave/pay ; Shared Parental Leave and Pay https://www.gov.uk/shared-parental-leave-and-pay ; unpaid parental leave https://www.gov.uk/parental-leave/entitlement", "last_updated": "2026-08-01", "confidence": "high", "note": "41 PAID wk = SMP 39 (6 at 90% of average weekly earnings plus 33 flat) + Statutory Paternity Pay 2 (flat). Was 52, which INCLUDED the 13 unpaid weeks of the maternity year and violated the field definition. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md and 2026-08-01-pl-germany-sweden.md / 2026-08-01-pl-uk-ireland.md. Shared Parental Leave adds ZERO weeks: gov.uk requires the mother to take less than the 39 weeks of maternity pay and use the rest as ShPP, so it redistributes her untaken entitlement and leaves the household ceiling at 39. Excluded: the 13-week unpaid SMP tail; 18 unpaid weeks per parent of parental leave; Neonatal Care Pay, which is conditional on the baby needing care; occupational maternity pay, which is contractual rather than statutory." }, "tax.inbound_regime": { "source": "4-year FIG regime (from 2025-04-06) relieves FOREIGN income and gains only, not a UK salary: HMRC RFIG41000 https://www.gov.uk/hmrc-internal-manuals/residence-and-fig-regime-manual/rfig41000 . Overseas Workday Relief covers only earnings for duties NOT performed in the UK (EIM43565) and is capped from 2025-04-06 at the lower of 30% of relevant qualifying employment income or GBP 300,000 (EIM43600); conditions incl. 10 immediately preceding tax years of non-UK residence (EIM43560). https://www.gov.uk/hmrc-internal-manuals/employment-income-manual/eim43600", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: re-verified against HMRC manuals. No salary relief remains correct for 2026: a resident performing all duties in the UK gets nothing from FIG (foreign income only) and nothing from OWR (non-UK duties only). Note sharpened with the new GBP 300,000 / 30% OWR cap and the 10-year prior non-residence test. No percentage discount added, deliberately: OWR turns on a UK/non-UK workday split we never ask for, and the conservative default for a relocating employee is 100% UK workdays." }, "tax.paye_brackets": { "source": "Scottish top-rate threshold 125,140 taxable + PA taper over 100k (gov.scot income tax); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Edinburgh council tax Band D £1,563.51 ~0.47% effective", "last_updated": "2026-05-27", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "VAT 20% (UK-wide, inherited from London baseline)", "last_updated": "2026-05-27", "confidence": "high" }, "housing._summary": { "source": "ESPC + Numbeo Edinburgh 2025-26", "last_updated": "2026-05-27", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "NHS Scotland universal (inherited)", "last_updated": "2026-05-27", "confidence": "high" }, "childcare._summary": { "source": "Scotland 1140 free hours for 3-4yr; private nursery net fee", "last_updated": "2026-05-27", "confidence": "medium" }, "food._summary": { "source": "Numbeo Edinburgh 2026", "last_updated": "2026-05-27", "confidence": "medium" }, "transit._summary": { "source": "Lothian Buses Ridacard £74/mo", "last_updated": "2026-05-27", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Edinburgh", "last_updated": "2026-05-27", "confidence": "medium" }, "travel_home._summary": { "source": "EDI->US economy 2025-26 (direct routes)", "last_updated": "2026-05-27", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "gov.uk Child Benefit, what you'll get https://www.gov.uk/child-benefit/what-youll-get ; HMRC rates by tax year https://www.gov.uk/government/publications/rates-and-allowances-tax-credits-child-benefit-and-guardians-allowance/tax-credits-child-benefit-and-guardians-allowance ; High Income Child Benefit Charge https://www.gov.uk/child-benefit-tax-charge (all fetched 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "STATUTORY additional-child rate, 2026-27 tax year: GBP 17.90 a week x 52 = 930.80 a year. The eldest or only child draws govt_benefits.child_allowance_eldest_annual instead. Means-tested by the ENGINE, not in this field: govt_benefits.child_allowance_taper carries the High Income Child Benefit Charge and computeGovtBenefits() withdraws the benefit per household on the higher earner's gross. Until 2026-09-12 this field stored 0 on the assumption that every published UK household's higher earner was over GBP 80,000. That holds for the USD 280,000 family scenario (London primary about GBP 142k, still 0 through the taper) and fails for the USD 150,000 two-earner persona (London higher earner GBP 74,063 keeps 30%, GBP 701.22 a year) and for any calculator user under the ceiling. Ledger decision 2026-09-12; research docs/superpowers/research/2026-09-12-child-benefit-audit-fixes.md." }, "govt_benefits.child_allowance_eldest_annual": { "source": "gov.uk Child Benefit, what you'll get https://www.gov.uk/child-benefit/what-youll-get ; HMRC rates by tax year https://www.gov.uk/government/publications/rates-and-allowances-tax-credits-child-benefit-and-guardians-allowance/tax-credits-child-benefit-and-guardians-allowance ; High Income Child Benefit Charge https://www.gov.uk/child-benefit-tax-charge (all fetched 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "Eldest or only child, 2026-27 tax year: GBP 27.05 a week x 52 = 1,406.60 a year. Unit: ANNUAL, from the statutory WEEKLY rate (HMRC pays every four weeks, 13 payments a year). Drawn by the first eligible child; every further child draws child_allowance_per_kid_annual. Same means-testing as that field: the taper block withdraws it." }, "govt_benefits.child_allowance_taper": { "source": "gov.uk Child Benefit, what you'll get https://www.gov.uk/child-benefit/what-youll-get ; HMRC rates by tax year https://www.gov.uk/government/publications/rates-and-allowances-tax-credits-child-benefit-and-guardians-allowance/tax-credits-child-benefit-and-guardians-allowance ; High Income Child Benefit Charge https://www.gov.uk/child-benefit-tax-charge (all fetched 2026-09-12)", "last_updated": "2026-09-12", "confidence": "high", "note": "High Income Child Benefit Charge: threshold GBP 60,000 of adjusted net income for tax years from 2024-25, 1% of the benefit for every GBP 200 over it, 100% at GBP 80,000 or more. Tested on the HIGHER earner (\"whoever has the higher income is responsible for paying the tax charge\"), never on the household total, so basis is higher_earner. The engine tests gross salary plus RSU as the conservative proxy for adjusted net income (ANI is net of pension contributions and Gift Aid, so it is lower and keeps more benefit); at the USD 150,000 persona that is a 70% charge on gross against about 48% on ANI. Whole-percent steps rounded down, as HMRC computes it. A deterministic statute, so no conservative-value tiebreak was applied to the BASIS; the gross proxy is the one conservative call inside it." }, "non_cash._summary": { "source": "UK labour law (national, inherited); NHS, statutory leave UK-wide", "last_updated": "2026-05-27", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Edinburgh: GBP 64,629, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "GBP 64,629 / 0.742 = USD 87,101; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3484, R5-rounded down to 0.34. REPLACES 0.58 (2026-08-11), and the interim 0.4 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.58 was 0.7733 of London's 0.75; that ratio x London's re-derived 0.53 = 0.4099, R5-rounded down to 0.4. London's own figure: GBP 99,953 / 0.742 = USD 134,708, / USD 250,000 = 0.5388. This is a BASIS move, not a re-measurement: Edinburgh's rank inside United Kingdom is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.58, UK 2nd financial centre." }, "tax.filing_basis": { "source": "HMRC - independent taxation of spouses since 1990", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "childcare.subsidy_per_kid_annual": { "source": "https://www.gov.uk/free-childcare-if-working", "last_updated": "2026-08-05", "note": "ZEROED 2026-08-05 (CC5): the free-hours entitlement is a provider-funded fee reduction, never household cash, so under the framework it belongs inside preschool_monthly, same defect class as the 08-02 double-subtraction batch. Working: docs/superpowers/research/2026-08-05-cc5-brazil-vienna-sourcing.md", "confidence": "high" }, "childcare.preschool_monthly": { "source": "Coram Family and Childcare Survey 2026 (regional tables, derivation in the research file)", "last_updated": "2026-08-05", "confidence": "high", "note": "Corrected 2026-08-05 from a too-low figure to the Coram-derived regional out-of-pocket for a 3-4yo with the surviving entitlement; the GBP 100k per-parent cliff kills 30h/TFC for every published kid-bearing scenario (England), so 15h universal only. Value 790." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "ONS Healthcare expenditure, UK Health Accounts 2023 and 2024 https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthcaresystem/bulletins/ukhealthaccounts/2023and2024 (household out-of-pocket GBP 46bn in 2024, of which ONS states ~38% is long-term care) + NHS England prescription charge and prepayment certificate", "last_updated": "2026-08-08", "confidence": "medium", "note": "Almost NONE of this is NHS cost-sharing, and the _meta note has to say so: realized England prescription charges are only about GBP 10-12 per capita because ~89% of items are dispensed free. The figure is overwhelmingly OTC medicines, self-pay private treatment against NHS waiting lists, and private physiotherapy. The old GBP 400 sat ABOVE the entire statutory ceiling at every household size (England prepayment certificate GBP 114.50 per adult, family of four GBP 229 since children are free, and GBP 0 in Scotland, Wales and Northern Ireland). One UK-wide value is retained this pass; Edinburgh's free prescriptions are a real but small (GBP 10-15/yr) difference and the city-level split is queued. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "ONS Healthcare expenditure, UK Health Accounts 2023 and 2024 https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthcaresystem/bulletins/ukhealthaccounts/2023and2024 (household out-of-pocket GBP 46bn in 2024, of which ONS states ~38% is long-term care) + NHS England prescription charge and prepayment certificate", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Transport Scotland, Scottish Household Survey travel diary, Transport and Travel in Scotland 2024, table TD6a (journeys to work by duration). https://www.transport.gov.scot/media/kokbappi/travel-diary-tables-transport-and-travel-in-scotland-2024.xlsx", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Journeys to work by duration, banded", "year": "2024", "note": "DERIVED, and the weakest number in the sweep. No Edinburgh commute-time figure exists in any official source: Scotland's census collects method only, and its one travel table is 'work or study', a different quantity. Scotland's banded one-way mean is 25.6; England's conurbation premium of about +15% (NTS9914b, 34.2 conurbation vs 29.7 all areas) was applied to reach 29. The previous 35 would have put a compact 530k city above Manchester and within 5 minutes of London." }, "non_cash.sick_leave": { "source": "Social Security Contributions and Benefits Act 1992, s.151 (employer's duty to pay; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": "https://www.gov.uk/statutory-sick-pay/eligibility", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "HMRC, Pension schemes rates and allowances https://www.gov.uk/government/publications/rates-and-allowances-pension-schemes/pension-schemes-rates ; Pensions Tax Manual PTM057100 https://www.gov.uk/hmrc-internal-manuals/pensions-tax-manual/ptm057100 ; https://www.gov.uk/tax-on-your-private-pension/annual-allowance", "last_updated": "2026-08-11", "confidence": "high", "note": "Annual allowance, 2025-26 and 2026-27 (identical on every figure). A COMBINED input ceiling covering everything paid in by the member or anyone else including the employer, enforced by the annual allowance charge, plus the separate rule that relief cannot exceed 100% of relevant UK earnings. Taper is GBP 1 per GBP 2 of ADJUSTED income above 260,000 down to a 10,000 floor; our gross understates adjusted income (which adds employer contributions) so the taper bites later than reality. No less_employer_pct: auto-enrolment minimum is 3% of qualifying earnings, about GBP 1,321, and anything larger would be an assumption. Never binds at our incomes." } }, "bilbao": { "childcare.free_tier_start_age": { "last_updated": "2026-08-01", "confidence": "high", "source": "Ley Organica 2/2006 de Educacion (LOE), consolidated text, art. 15.2: \"El segundo ciclo de la educacion infantil sera gratuito\"; art. 14.1: \"El primero comprende hasta los tres anos, y el segundo, desde los tres a los seis anos de edad\" https://www.boe.es/buscar/doc.php?id=BOE-A-2006-7899 . Enrolment: Ministerio de Educacion FP y Deportes, 2025-09-09, \"La escolarizacion en el segundo ciclo de educacion infantil (3 a 5 anos) es en Espana practicamente universal, con un 98% de los menores matriculados\" https://www.educacionfpydeportes.gob.es/prensa/actualidad/2025/09/20250909-panoramaeducacion2025.html", "note": "Segundo ciclo (3-6) is free by statute in public and concertado schools and 98% enrolled; primer ciclo (0-3) is neither free nor universal. Before this, ages 1-5 all billed one rate, so the bucket straddled both: a 1-2 year old in paid guarderia was billed the free-tier meal fee. Entry to P3 is by BIRTH-YEAR COHORT (children born in year N start the September of year N+3), so a December-born child starts free schooling at 2 years 9 months; whole-number ages make 3 the correct modelling boundary." }, "non_cash.parental_leave_paid_pct": { "source": "Real Decreto-ley 9/2025 preamble (cien por cien replacement) https://www.boe.es/eli/es/rdl/2025/07/29/9 ; LGSS art. 179 (base reguladora) and art. 148 (tope maximo de la base de cotizacion)", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 38 weeks into non_cash.parental_leave_capped (100% of a base capped at the base maxima de cotizacion), so no percentage-paid weeks remain on this field. Was 100 with the cap ignored per the old field convention. Watch item unchanged: the RDL preamble records a government commitment to extend to 20 weeks per parent, which is not law and is not stored." }, "non_cash.parental_leave_weeks": { "source": "Estatuto de los Trabajadores art. 48.4 as amended by Real Decreto-ley 9/2025 de 29 de julio, in force 31/07/2025 https://www.boe.es/buscar/act.php?id=BOE-A-2025-15741 ; consolidated ET https://www.boe.es/buscar/act.php?id=BOE-A-2015-11430", "last_updated": "2026-08-01", "confidence": "high", "note": "38 wk = 19 per parent x 2 (6 obligatory post-partum + 11 usable to age 12 months + 2 usable to age 8). ET art. 48.4 states este derecho es individual de la persona trabajadora sin que pueda transferirse su ejercicio al otro progenitor. Spain has no separate maternity leave: both parents hold the identical permiso por nacimiento y cuidado. Was 16, one parent on the pre-2025 duration. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. Excluded: the 8-week permiso parental of art. 48 bis, which RDL 9/2025 confirms remains unpaid; the 32-week monoparental variant; multiple-birth uplifts." }, "tax.federal_brackets": { "source": "Bizkaia FORAL IRPF 2025 (Hacienda Foral, NF 13/2013 mod NF 10/2023): unified 23-49% scale REPLACING estatal+autonómica", "last_updated": "2026-05-27", "confidence": "high", "note": "Foral territory: federal_brackets holds the full foral scale; state_brackets is a 0-rate passthrough so compute sums to foral-only." }, "tax.state_brackets": { "source": "Foral passthrough (0%: no separate autonómica layer in Basque Country)", "last_updated": "2026-05-27", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "IBI Bilbao 0.183% of cadastral (Ordenanzas Fiscales 2025)", "last_updated": "2026-05-27", "confidence": "medium" }, "housing._summary": { "source": "Numbeo + Idealista Bilbao 2026", "last_updated": "2026-05-27", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Osakidetza/SNS universal + private (inherited from Madrid baseline)", "last_updated": "2026-05-27", "confidence": "high" }, "childcare._summary": { "source": "Public Haurreskolak free 0-2 (Basque Govt); private guardería ~€425/mo", "last_updated": "2026-05-27", "confidence": "medium" }, "food._summary": { "source": "Numbeo Bilbao 2026", "last_updated": "2026-05-27", "confidence": "medium" }, "transit._summary": { "source": "Bilbao Metro + Barik monthly ~€45", "last_updated": "2026-05-27", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Bilbao", "last_updated": "2026-05-27", "confidence": "medium" }, "travel_home._summary": { "source": "BIO->US (via hub) economy 2025-26", "last_updated": "2026-05-27", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Means-tested, 0 at tech income (national, inherited)", "last_updated": "2026-05-27", "confidence": "high" }, "non_cash._summary": { "source": "Spanish labour law (national, inherited)", "last_updated": "2026-05-27", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Bilbao: EUR 40,005, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-south.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 40,005 / 0.867 = USD 46,142; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1846, R5-rounded down to 0.18. REPLACES 0.4 (2026-08-11), and the interim 0.24 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.4 was 1.0000 of Madrid's 0.4; that ratio x Madrid's re-derived 0.24 = 0.2400, R5-rounded down to 0.24. Madrid's own figure: EUR 53,912 / 0.867 = USD 62,182, / USD 250,000 = 0.2487. This is a BASIS move, not a re-measurement: Bilbao's rank inside Spain is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.40, on par with Madrid; Basque engineering base." }, "tax.social_security_cap": { "confidence": "high", "source": "Seguridad Social base maxima 2026 EUR 5,101.20/mo (taxsummaries.pwc.com/spain); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "tax.filing_basis": { "source": "Agencia Tributaria - individual IRPF (the engine files spouses individually; the elective declaracion conjunta is excluded)", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "childcare.preschool_monthly": { "source": "Comedor: Orden de 9 de septiembre de 2024, de la Consejera de Educacion (Gobierno Vasco / Eusko Jaurlaritza), comensal habitual EUR 4,60/dia (no habitual EUR 5,20) https://www.euskadi.eus/comedores-escolares/web01-a2elikal/es/ ; the same EUR 4,60 daily maximum recurs in an explicitly 2025-26 euskadi.eus document. Goiztiarrak (aula matinal): priced per school by the guraso elkartea, not centrally; the only Bizkaia public-school tariff found for 2025-26 is Landako Eskola, Durango, 08:00-09:00 https://landakoeskola.org .", "last_updated": "2026-08-01", "confidence": "low", "note": "EUR 4.60/day x 15 = EUR 69.00 comedor, plus goiztiarrak EUR 60/month over 10 months = EUR 50.00, total EUR 119.00. Was 200, an unsourced blend. A dual-earner family using the comedor daily is a comensal habitual and pays 4,60, not the 5,20 occasional rate. CONFIDENCE LOW, and this is the weakest of the six Spanish cities: NO Bilbao-city public school publishes a goiztiarrak tariff, so the EUR 60 figure comes from Durango, elsewhere in Bizkaia, at the top of a EUR 40-60 range. It is used at the top per the conservative value bias and because it excludes breakfast, which the food block already charges. Basque dies lectius were not separately verified, so the Madrid-confirmed 180 is used. WATCH: no 2025-26-specific comedor Orden was found, only the September 2024 Orden plus a 2025-26 becas document reusing the same figure. Stored value is the ANNUAL total divided by 12 (computeChildcareCost multiplies by 12), NOT the school-year invoice, which runs ~10 months. Applies the owner decision of 2026-07-18 that Spain follows the statutory free convention, now that the age boundary makes it applicable to the free ages only. Sourcing: docs/superpowers/research/2026-08-01-childcare-free-tier-boundary.md and 2026-08-01-spain-wraparound-fees.md." }, "non_cash.parental_leave_capped": { "source": "Orden PJC/297/2026, de 30 de marzo, art. 2.1: tope maximo de la base de cotizacion 5.101,20 euros mensuales desde el 1 de enero de 2026 https://www.boe.es/diario_boe/txt.php?id=BOE-A-2026-7296 ; LGSS art. 179 (the base reguladora is the contingencias comunes contribution base) and art. 148 (that base is capped at the tope maximo)", "last_updated": "2026-08-06", "confidence": "high", "note": "All 38 weeks pay 100% of a base capped at EUR 5,101.20/month (EUR 61,214.40/yr), one clean segment with no phase split: above that gross the benefit is the cap, not the salary. At EUR 90,000 the effective rate is ~68%, at EUR 150,000 ~41%, versus the naked 100% this replaces. cap_amount doubles as the payout ceiling because the rate is 100. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Spain)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "RD 16/2012 pharmaceutical co-payment tramos, with our ICP household in the uncapped 50-60% band + Ministerio de Sanidad SCS household out-of-pocket with dental at ~32%", "last_updated": "2026-08-08", "confidence": "medium", "note": "Spain is the European exception on the child rule: a dependent minor registered on a working parent's card pays the SAME co-payment percentage as the parent, so the sub-linearity here is a volume argument (free well-child checks and the vaccination schedule never touch the pharmacy co-pay), not a statutory discount. Spain has NO statutory annual ceiling for the household we model; the monthly caps introduced in 2026 apply only below EUR 35,000 of income. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "RD 16/2012 pharmaceutical co-payment tramos, with our ICP household in the uncapped 50-60% band + Ministerio de Sanidad SCS household out-of-pocket with dental at ~32%", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "INE Spain, Urban Audit indicator 6.4, Area Urbana Funcional de Bilbao. https://www.ine.es/jaxiT3/Tabla.htm?t=69327", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Duracion media del desplazamiento al trabajo (solo ida)", "year": "2011", "note": "AUF value 22.9." }, "non_cash.sick_leave": { "source": "Ley General de la Seguridad Social, current consolidated text Real Decreto; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-B.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Agencia Tributaria, Manual Renta 2025 s.8.2.2.6 https://sede.agenciatributaria.gob.es/Sede/ayuda/manuales-videos-folletos/manuales-ayuda-presentacion/irpf-2025/8-cumplimentacion-irpf/8_2-base-liquidable-general-base-ahorro/8_2_2-reducciones-aportaciones-prevision-social/8_2_2_6-aportaciones-anuales-maximas-limite-reduccion.html", "last_updated": "2026-08-11", "confidence": "high", "note": "Reduccion de la base imponible general for aportaciones a sistemas de prevision social (Ley 35/2006 arts. 51-52), EUR 1,500/yr. A base deduction, not a credit, and individual. NOT 8,500: the +8,500 increment requires EMPLOYER contributions to a plan de empleo, which we do not model, and worker contributions only ride along under a matching coefficient table. pct_of_earnings deliberately omitted: the outer 30%-of-net-earned-income bound never binds at EUR 75-95k and 0.30 of GROSS would be both inert and wrong-based. SEPARATE pre-existing understatement, out of scope: computeSpainTaxes does not treat cotizaciones as a gasto deducible." } }, "lausanne": { "tax.federal_brackets": { "source": "ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Exact 2026 Grundtarif (kalte-Progression vintage confirmed against the 2025 circular). 2026-08-07: restored the missing 5.94% band 76,200-82,100 (the stored table jumped 2.97% to 6.6%, overstating by up to CHF 38.94/yr). Applied to taxable income = gross minus the statutory deduction stack (see tax.payroll_deductible)." }, "tax.cantonal_brackets": { "source": "ESTV Feuille cantonale VD etat 1.1.2026 (art. 47 LI bareme indexe 2026; LA 2026 coefficient 155%; LRIPP art. 4 -5% 2026 canton-only) https://www.estv2.admin.ch/stp/kb/vd-fr.pdf ; Ville de Lausanne arrete d imposition 2025-2029 (78.5) https://www.lausanne.ch/officiel/administration/finances-et-mobilite/finances/impots/coefficient-taux-arrete-imposition.html", "last_updated": "2026-08-07", "confidence": "high", "note": "RECALIBRATED 2026-08-07: exact art. 47 LI 2026 bareme (1% from the first franc, top base 15.5%, NO zero band) x 2.2575 combined = 1.55 cantonal coefficient x 0.95 LRIPP rebate (canton-only) + 0.785 Lausanne communal. REFUTES the prior ladder, a from-gross calculator fit (zero band to 14,500, implied top base 11.5%) that understated ~29% at taxable 105,000 (14,929.34 vs official 20,942.83). Married stays quotient 1.8 (art. 43 al. 2 LI, confirmed 2026); child quotient parts (+0.5/child) NOT modeled - conservative. Probe checkpoints reproduced to the cent before storing." }, "tax.property_tax_rate_pct": { "source": "Vaud impôt foncier 1.5‰ = 0.15%", "last_updated": "2026-05-27", "confidence": "high" }, "housing._summary": { "source": "Numbeo + RealAdvisor Lausanne 2026", "last_updated": "2026-05-27", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Mandatory private LAMal (inherited from Zurich baseline)", "last_updated": "2026-05-27", "confidence": "high" }, "childcare._summary": { "source": "Vaud crèche income-scaled, expensive (~CHF 2,200/mo net)", "last_updated": "2026-05-27", "confidence": "medium" }, "food._summary": { "source": "Numbeo Lausanne 2026", "last_updated": "2026-05-27", "confidence": "medium" }, "transit._summary": { "source": "tl monthly pass ~CHF 78", "last_updated": "2026-05-27", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Lausanne", "last_updated": "2026-05-27", "confidence": "medium" }, "travel_home._summary": { "source": "GVA->US economy 2025-26", "last_updated": "2026-05-27", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Cantonal family allowance (inherited; cantonal varies)", "last_updated": "2026-05-27", "confidence": "medium", "note": "Swiss Kinderzulagen / allocations familiales: cantonal, universal minimum (>= CHF 200-250/mo), NOT means-tested." }, "non_cash._summary": { "source": "Swiss labour law (national, inherited); 20 vacation, mandatory-private healthcare", "last_updated": "2026-05-27", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Lausanne: CHF 103,070, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "CHF 103,070 / CHF 143,634 (Zurich, fetched the SAME DAY) = 0.7176 of Zurich; x Zurich's 0.6832 = 0.4903, R5-rounded down to 0.49. A SAME-DAY RATIO is used here rather than the direct CHF/FX/250,000 division used elsewhere, and only here, because Zurich is the one country anchor whose stored table has drifted materially: stored CHF 137,843 against CHF 143,634 live on 2026-08-11, 4.2% apart. Dividing a fresh satellite figure by 250,000 while the anchor sits on a 4.2%-stale table leaks that gap into every Swiss satellite in the FLATTERING direction. Every other same-day anchor drifted under 3% and in the conservative direction (Oslo 2.8%, Sweden 1.6%, Denmark 1.3%, Germany 1.0%), so none of them needed this. REPLACES 0.83 (2026-08-11), and the interim 0.53 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Tight ladder, modest sample. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.83 was 0.7905 of Zurich's 1.05; that ratio x Zurich's re-derived 0.68 = 0.5375, R5-rounded down to 0.53. Zurich's own figure: CHF 137,843 / 0.807 = USD 170,809, / USD 250,000 = 0.6832. This is a BASIS move, not a re-measurement: Lausanne's rank inside Switzerland is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.83, below Zurich, ~Geneva; EPFL/tech." }, "childcare.subsidy_per_kid_annual": { "source": "bern.ch Betreuungsgutscheine; bs.ch Betreuungsbeitraege; lausanne.ch Reseau-L 2025", "last_updated": "2026-07-17", "note": "Municipal childcare subsidy is means-tested and 0 for the CHF 200k+ household: Bern Betreuungsgutscheine cut off at CHF 160k massgebendes Einkommen; Lausanne Reseau-L means-tested reduction hits the max tariff; Basel-Stadt subsidized places are income-prioritized/capacity-limited, so a high earner is modeled at the private-Kita market rate (infant_monthly). Was 8,000/4,800/2,400." }, "tax.bvg": { "source": "BVG statutory minimum, 2026 parameters (unchanged from 2025): entry threshold CHF 22,680, coordination deduction CHF 26,460, insured-salary upper limit CHF 90,720, coordinated salary clamped to 3,780-64,260. ASGA Pensionskasse, Merkblatt Renten und Grenzbetraege per 1.1.2026 https://www.asga.ch/wp-content/uploads/downloads/merkblaetter/merkblaetter-d/renten-und-grenzbetraege_2026_d.pdf ; BSV https://www.bsv.admin.ch/de/altersvorsorge-beruflichen-vorsorge (confirms the 22,680 threshold, savings from age 25, and the BVG Art. 66 employer-parity rule); age credits BVG Art. 16: 7/10/15/18% of coordinated salary by age band (Sicherheitsfonds BVG, sfbvg.ch Sparprozess), employee pays half: 3.5/5/7.5/9%.", "last_updated": "2026-08-06", "confidence": "high", "note": "Statutory MINIMUM only, deliberately: real Swiss professional plans insure above the mandatory band (ueberobligatorisch) and add plan-specific risk/admin premiums, so a typical payslip deducts MORE than this charge; everything above the legal floor is excluded under the statutory-minimum convention (the same rule that keeps Swedish ITP and Danish employer pensions out entirely). The charge docks take-home inside taxes.payroll and is carved to the forced-savings own-money row (funded, FZG-vested, portable). It runs on the same wage base as the AHV/ALV slice and never reduces the income-tax base: the CH-1 deduction floor stays CHF 13,500/27,000, which errs against Switzerland. Applied per earner (entry threshold and coordination deduction per person); the 18-24 bracket pays 0 because retirement credits start at 25 (BVG Art. 7). Decision log: docs/superpowers/research/2026-08-06-pension-omission-pair.md" }, "tax.cantonal_splitting_divisor": { "source": "Vaud LI (BLV 642.11) art. 43 al. 2, Base legislative vaudoise, consolidated version in force 01.05.2026 https://prestations.vd.ch/pub/blv-publication/actes/consolide/642.11", "last_updated": "2026-08-01", "confidence": "high", "married_variant": "divisor", "note": "Upgraded 2026-08-01 from a simulator citation to the statute itself. LI art. 43 al. 1 divides taxable income by the family parts, and al. 2 sets those parts at 1 for a single filer and \"1,8 pour les epoux vivant en menage commun\". Art. 49 al. 4 names the value: \"le quotient familial des epoux sans enfant (art. 43)\". Art. 47 is one rate schedule for all filers and no married ordinary schedule exists, so lausanne must not be given a tax.cantonal_brackets_married array. Exact for a childless couple, a simplification once children arrive, since al. 2 adds 0.5 per child and al. 3 caps that reduction. The parts are pure ratios and are not indexed, unlike the art. 47 bareme. Vaud LRIPP (BLV 642.12) cuts cantonal income tax from 01.01.2027." }, "non_cash.parental_leave_weeks": { "source": "BSV, Maternity, paternity and adoption leave and allowances https://www.bsv.admin.ch/en/income-compensation-parents ; LECA/EOG; ahv-iv.ch leaflets 6.02 (maternity) and 6.04 (paternity)", "last_updated": "2026-08-06", "confidence": "high", "note": "16 wk = 14 maternity (EO, continuous from birth) + 2 paternity (10 working days within 6 months, federal statute since 1 Jan 2021), each parent's own entitlement; no federal shared scheme exists. The previously stored value omitted the 2 paternity weeks its own note cited. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md." }, "non_cash.parental_leave_paid_pct": { "source": "BSV https://www.bsv.admin.ch/en/income-compensation-parents : up to 80% of average income, max. CHF 220 per day, for both the maternity and paternity allowances", "last_updated": "2026-08-06", "confidence": "high", "note": "80% capped at CHF 220/day = CHF 1,540/wk, a cap reached at about CHF 99,000/yr, so every ICP salary receives the flat maximum: at CHF 130,000 the effective rate is about 62% and falls as salary rises. Kept as pct 80 because below the threshold it is a true 80%; the capped-percentage shape has no honest flat-block representation, so the cap arithmetic lives here." }, "tax.federal_brackets_married": { "source": "ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Exact 2026 Verheiratetentarif (Art. 36 Abs. 2 DBG), marginal form with the 11.5% flat crossover at 941,300. Replaces the divisor-2 full-splitting approximation, which UNDERSTATED federal married tax (exact 7,579.60 vs splitting 3,185.42 at joint taxable 166,920) - the federal tariff is harsher than splitting. All checkpoints re-derived by hand; see docs/superpowers/research/2026-08-08-ch-federal-worksheet.md." }, "tax.payroll_deductible": { "source": "Art. 33 Abs. 1 lit. d + f DBG SR 642.11, consolidation in force 2026-01-01 (fedlex)", "last_updated": "2026-08-07", "confidence": "high", "note": "Employee AHV/IV/EO, ALV and BVG contributions are fully deductible from the income-tax base (salary-certificate net is the taxable base). Gates the per-earner SI+BVG deduction in computeSwitzerlandTaxes; the charged amounts are the deducted amounts, so the FORCED_SAVINGS carve cannot double-count. Replaces the frozen CHF 13,500/27,000 CH-1 floor (2026-07-26), which bundled SI at one income level." }, "tax.professional_forfait": { "source": "Berufskostenverordnung SR 642.118.1 Art. 7 + Anhang Ziff. 1 (fedlex); 2026 min/max confirmed in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Uebrige Berufskosten pauschale: 3% of NET salary (post-SI/BVG), min CHF 2,000 / max CHF 4,000, per earner with income. Commuting and meal deductions excluded (variable/behavioral, conservative)." }, "tax.insurance_deduction_single": { "source": "Art. 33 Abs. 1 lit. g DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amount in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Versicherungsabzug CHF 1,800 (taxpayer WITH pillar-2 contributions, the employee case). Real KVG premiums exceed this cap and some cantons deduct more (BS 4,200, GE full premiums) - federal amount kept for the one-base model, conservative." }, "tax.insurance_deduction_married": { "source": "Art. 33 Abs. 1 lit. g DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amount in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Versicherungsabzug CHF 3,700 married (WITH pillar-2 amounts). Per-child +700 supplement excluded (conservative)." }, "tax.married_deduction": { "source": "Art. 35 Abs. 1 lit. c DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amount in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Verheiratetenabzug CHF 2,800 (2026), automatic for married couples." }, "tax.two_earner_deduction": { "source": "Art. 33 Abs. 2 DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amounts in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Zweitverdienerabzug: 50% of the lower earned income, min CHF 8,600 / max CHF 14,100, capped at the lower income; applies only when both spouses have earned income. Federal amount used for the one-base model (cantonal equivalents are smaller: BS 1,100, ZH ~6,100); partially offset by the larger cantonal insurance deductions we exclude - see the reasoning in docs/superpowers/research/2026-08-08-tax-structural-ch-ie.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "BAG Kostenbeteiligung https://www.bag.admin.ch/de/krankenversicherung-kostenbeteiligung-fuer-in-der-schweiz-wohnhafte-versicherte + BAG OKP statutory cost-sharing aggregate CHF 5.7bn (2024) over 8.96m insured, adult/child split from BAG age-group net benefits", "last_updated": "2026-08-08", "confidence": "high", "note": "Blended CHF 636 per insured person; children ~CHF 140 each (zero franchise, halved CHF 350 Selbstbehalt cap, exempt from the hospital per-diem), leaving CHF 749 per adult, plus a ~7% conservative uplift for non-KVG self-pay that is neither dental nor long-term care. REPLACES a value that was a single adult's statutory MAXIMUM: the old CHF 1,200 was above the CHF 1,000 minimum-franchise ceiling and was billed unchanged to a family of four whose statutory maximum is CHF 2,700. Geneva keeps its existing 1.083 premium, which is a utilisation difference, not a rules difference: cost-sharing law is federal and identical in every canton. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "BAG Kostenbeteiligung https://www.bag.admin.ch/de/krankenversicherung-kostenbeteiligung-fuer-in-der-schweiz-wohnhafte-versicherte + BAG OKP statutory cost-sharing aggregate CHF 5.7bn (2024) over 8.96m insured, adult/child split from BAG age-group net benefits", "last_updated": "2026-08-08", "confidence": "high", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Swiss Federal Statistical Office (BFS), PxWeb cube px-x-2105000000_206, Agglomeration Lausanne. https://www.pxweb.bfs.admin.ch/pxweb/de/px-x-2105000000_206/px-x-2105000000_206/px-x-2105000000_206.px", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Durchschnittliche Arbeitswegzeit, in Minuten (ein Hinweg)", "year": "2020", "note": "33.0 min. Agglomeration geography, matching our metro basis." }, "non_cash.sick_leave": { "source": "Code of Obligations / Obligationenrecht (OR/CO) Art. 324a, paragraphs 1-4.; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Steuerverwaltung Kanton Bern, Taxinfo Saeule 3a https://taxinfo.sv.fin.be.ch/taxinfo/a2830da2-1a00-4928-b58b-74da5d48a598 ; EFD/ESTV confirmation of the 2026 amounts, 17 Nov 2025 https://www.estv.admin.ch/de/newnsb/xgRMirCsezICX4rtof9Lm", "last_updated": "2026-08-11", "confidence": "high", "note": "Saeule 3a kleiner Abzug, CHF 7,258 for 2026 (unchanged from 2025), for anyone who also contributes to a Pensionskasse, which our ICP does (both CH cities carry a bvg block). The grosser Abzug (20% of income to CHF 36,288, no 2nd pillar) is NOT our case. Employee-only; ESTV states the maximum deduction is simultaneously the contribution limit. 2nd-pillar Einkauf is also deductible but is bounded by a personal Einkaufspotenzial we cannot know, so it is excluded." } }, "bern": { "tax.federal_brackets": { "source": "ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Exact 2026 Grundtarif (kalte-Progression vintage confirmed against the 2025 circular). 2026-08-07: restored the missing 5.94% band 76,200-82,100 (the stored table jumped 2.97% to 6.6%, overstating by up to CHF 38.94/yr). Applied to taxable income = gross minus the statutory deduction stack (see tax.payroll_deductible)." }, "tax.cantonal_brackets": { "source": "TaxInfo Kanton Bern, Art. 42 tariff tables ab Steuerjahr 2026: single https://www.taxinfo.sv.fin.be.ch/taxinfo/f244ba97-5848-473f-9037-0621b32f8836 married https://www.taxinfo.sv.fin.be.ch/taxinfo/36f0ac7e-6293-4612-90dd-e4a1cb2f5915 ; Steueranlagen 2026 (kumuliert Stadt Bern 4.515) https://www.fin.be.ch/content/dam/fin/dokumente/de/finanzverwaltung/finanzausgleich/publikationen/fv-stat-steueranlage_2026_1.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "RECALIBRATED 2026-08-07: exact Art. 42 Abs. 2 single tariff (14 marginal tranches, 1.95% from the first franc, NO zero band, top 6.5%) x 4.515 (Kanton 2.975 + Stadt Bern 1.54, 2026, church excluded). REFUTES the prior 7-band ladder, a from-gross calculator fit that understated ~9.5% at taxable 105,000 (19,487.02 vs official 21,537.68). Anchors reproduced EXACT pre-store: einfache 50k=1,959.75 / 105k=4,770.25 / 167k=8,363.30." }, "tax.property_tax_rate_pct": { "source": "Stadt Bern Liegenschaftssteuer 1.5‰ = 0.15%", "last_updated": "2026-05-27", "confidence": "high" }, "housing._summary": { "source": "Numbeo + RealAdvisor Bern 2026", "last_updated": "2026-05-27", "confidence": "high" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Mandatory private LAMal (inherited)", "last_updated": "2026-05-27", "confidence": "high" }, "childcare._summary": { "source": "Bern kiBon voucher system, income-scaled (~CHF 2,200/mo net)", "last_updated": "2026-05-27", "confidence": "medium" }, "food._summary": { "source": "Numbeo Bern 2026", "last_updated": "2026-05-27", "confidence": "medium" }, "transit._summary": { "source": "Bernmobil Libero monthly pass CHF 82", "last_updated": "2026-05-27", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Bern", "last_updated": "2026-05-27", "confidence": "medium" }, "travel_home._summary": { "source": "ZRH/GVA->US economy 2025-26 (via Zurich)", "last_updated": "2026-05-27", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Cantonal family allowance (inherited)", "last_updated": "2026-05-27", "confidence": "medium", "note": "Swiss Kinderzulagen / allocations familiales: cantonal, universal minimum (>= CHF 200-250/mo), NOT means-tested." }, "non_cash._summary": { "source": "Swiss labour law (national, inherited)", "last_updated": "2026-05-27", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Bern: CHF 110,311, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "CHF 110,311 / CHF 143,634 (Zurich, fetched the SAME DAY) = 0.7680 of Zurich; x Zurich's 0.6832 = 0.5247, R5-rounded down to 0.52. A SAME-DAY RATIO is used here rather than the direct CHF/FX/250,000 division used elsewhere, and only here, because Zurich is the one country anchor whose stored table has drifted materially: stored CHF 137,843 against CHF 143,634 live on 2026-08-11, 4.2% apart. Dividing a fresh satellite figure by 250,000 while the anchor sits on a 4.2%-stale table leaks that gap into every Swiss satellite in the FLATTERING direction. Every other same-day anchor drifted under 3% and in the conservative direction (Oslo 2.8%, Sweden 1.6%, Denmark 1.3%, Germany 1.0%), so none of them needed this. REPLACES 0.82 (2026-08-11), and the interim 0.53 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Tight ladder, modest sample. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.82 was 0.7810 of Zurich's 1.05; that ratio x Zurich's re-derived 0.68 = 0.5310, R5-rounded down to 0.53. Zurich's own figure: CHF 137,843 / 0.807 = USD 170,809, / USD 250,000 = 0.6832. This is a BASIS move, not a re-measurement: Bern's rank inside Switzerland is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.82, govt-heavy, below Zurich." }, "childcare.subsidy_per_kid_annual": { "source": "bern.ch Betreuungsgutscheine; bs.ch Betreuungsbeitraege; lausanne.ch Reseau-L 2025", "last_updated": "2026-07-17", "note": "Municipal childcare subsidy is means-tested and 0 for the CHF 200k+ household: Bern Betreuungsgutscheine cut off at CHF 160k massgebendes Einkommen; Lausanne Reseau-L means-tested reduction hits the max tariff; Basel-Stadt subsidized places are income-prioritized/capacity-limited, so a high earner is modeled at the private-Kita market rate (infant_monthly). Was 8,000/4,800/2,400." }, "tax.bvg": { "source": "BVG statutory minimum, 2026 parameters (unchanged from 2025): entry threshold CHF 22,680, coordination deduction CHF 26,460, insured-salary upper limit CHF 90,720, coordinated salary clamped to 3,780-64,260. ASGA Pensionskasse, Merkblatt Renten und Grenzbetraege per 1.1.2026 https://www.asga.ch/wp-content/uploads/downloads/merkblaetter/merkblaetter-d/renten-und-grenzbetraege_2026_d.pdf ; BSV https://www.bsv.admin.ch/de/altersvorsorge-beruflichen-vorsorge (confirms the 22,680 threshold, savings from age 25, and the BVG Art. 66 employer-parity rule); age credits BVG Art. 16: 7/10/15/18% of coordinated salary by age band (Sicherheitsfonds BVG, sfbvg.ch Sparprozess), employee pays half: 3.5/5/7.5/9%.", "last_updated": "2026-08-06", "confidence": "high", "note": "Statutory MINIMUM only, deliberately: real Swiss professional plans insure above the mandatory band (ueberobligatorisch) and add plan-specific risk/admin premiums, so a typical payslip deducts MORE than this charge; everything above the legal floor is excluded under the statutory-minimum convention (the same rule that keeps Swedish ITP and Danish employer pensions out entirely). The charge docks take-home inside taxes.payroll and is carved to the forced-savings own-money row (funded, FZG-vested, portable). It runs on the same wage base as the AHV/ALV slice and never reduces the income-tax base: the CH-1 deduction floor stays CHF 13,500/27,000, which errs against Switzerland. Applied per earner (entry threshold and coordination deduction per person); the 18-24 bracket pays 0 because retirement credits start at 25 (BVG Art. 7). Decision log: docs/superpowers/research/2026-08-06-pension-omission-pair.md" }, "tax.cantonal_brackets_married": { "source": "TaxInfo Kanton Bern, Art. 42 tariff tables ab Steuerjahr 2026: single https://www.taxinfo.sv.fin.be.ch/taxinfo/f244ba97-5848-473f-9037-0621b32f8836 married https://www.taxinfo.sv.fin.be.ch/taxinfo/36f0ac7e-6293-4612-90dd-e4a1cb2f5915 ; Steueranlagen 2026 (kumuliert Stadt Bern 4.515) https://www.fin.be.ch/content/dam/fin/dokumente/de/finanzverwaltung/finanzausgleich/publikationen/fv-stat-steueranlage_2026_1.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "RECALIBRATED 2026-08-07: exact Art. 42 Abs. 1 married tariff x 4.515. The official Verheiratetentarif is its OWN curve - same boundaries as single only up to 140,100 (lower rates), divergent above; it is HIGHER than doubled-single splitting, so the prior doubled-threshold approximation additionally understated married Bern (-17.7% at taxable 167,000: 27,450.04 vs official 33,357.72). Anchors EXACT pre-store: einfache 100k=3,815.90 / 167k=7,388.20." }, "non_cash.parental_leave_weeks": { "source": "BSV, Maternity, paternity and adoption leave and allowances https://www.bsv.admin.ch/en/income-compensation-parents ; LECA/EOG; ahv-iv.ch leaflets 6.02 (maternity) and 6.04 (paternity)", "last_updated": "2026-08-06", "confidence": "high", "note": "16 wk = 14 maternity (EO, continuous from birth) + 2 paternity (10 working days within 6 months, federal statute since 1 Jan 2021), each parent's own entitlement; no federal shared scheme exists. The previously stored value omitted the 2 paternity weeks its own note cited. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md." }, "non_cash.parental_leave_paid_pct": { "source": "BSV https://www.bsv.admin.ch/en/income-compensation-parents : up to 80% of average income, max. CHF 220 per day, for both the maternity and paternity allowances", "last_updated": "2026-08-06", "confidence": "high", "note": "80% capped at CHF 220/day = CHF 1,540/wk, a cap reached at about CHF 99,000/yr, so every ICP salary receives the flat maximum: at CHF 130,000 the effective rate is about 62% and falls as salary rises. Kept as pct 80 because below the threshold it is a true 80%; the capped-percentage shape has no honest flat-block representation, so the cap arithmetic lives here." }, "tax.federal_brackets_married": { "source": "ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Exact 2026 Verheiratetentarif (Art. 36 Abs. 2 DBG), marginal form with the 11.5% flat crossover at 941,300. Replaces the divisor-2 full-splitting approximation, which UNDERSTATED federal married tax (exact 7,579.60 vs splitting 3,185.42 at joint taxable 166,920) - the federal tariff is harsher than splitting. All checkpoints re-derived by hand; see docs/superpowers/research/2026-08-08-ch-federal-worksheet.md." }, "tax.payroll_deductible": { "source": "Art. 33 Abs. 1 lit. d + f DBG SR 642.11, consolidation in force 2026-01-01 (fedlex)", "last_updated": "2026-08-07", "confidence": "high", "note": "Employee AHV/IV/EO, ALV and BVG contributions are fully deductible from the income-tax base (salary-certificate net is the taxable base). Gates the per-earner SI+BVG deduction in computeSwitzerlandTaxes; the charged amounts are the deducted amounts, so the FORCED_SAVINGS carve cannot double-count. Replaces the frozen CHF 13,500/27,000 CH-1 floor (2026-07-26), which bundled SI at one income level." }, "tax.professional_forfait": { "source": "Berufskostenverordnung SR 642.118.1 Art. 7 + Anhang Ziff. 1 (fedlex); 2026 min/max confirmed in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Uebrige Berufskosten pauschale: 3% of NET salary (post-SI/BVG), min CHF 2,000 / max CHF 4,000, per earner with income. Commuting and meal deductions excluded (variable/behavioral, conservative)." }, "tax.insurance_deduction_single": { "source": "Art. 33 Abs. 1 lit. g DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amount in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Versicherungsabzug CHF 1,800 (taxpayer WITH pillar-2 contributions, the employee case). Real KVG premiums exceed this cap and some cantons deduct more (BS 4,200, GE full premiums) - federal amount kept for the one-base model, conservative." }, "tax.insurance_deduction_married": { "source": "Art. 33 Abs. 1 lit. g DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amount in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Versicherungsabzug CHF 3,700 married (WITH pillar-2 amounts). Per-child +700 supplement excluded (conservative)." }, "tax.married_deduction": { "source": "Art. 35 Abs. 1 lit. c DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amount in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Verheiratetenabzug CHF 2,800 (2026), automatic for married couples." }, "tax.two_earner_deduction": { "source": "Art. 33 Abs. 2 DBG SR 642.11, consolidation in force 2026-01-01 (fedlex); 2026 amounts in ESTV Rundschreiben Nr. 215 (2-215-D-2025-d, 2025-09-11) https://www.estv.admin.ch/dam/de/sd-web/vFK3ntWLQ4s4/2-215-D-2025-d.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "Zweitverdienerabzug: 50% of the lower earned income, min CHF 8,600 / max CHF 14,100, capped at the lower income; applies only when both spouses have earned income. Federal amount used for the one-base model (cantonal equivalents are smaller: BS 1,100, ZH ~6,100); partially offset by the larger cantonal insurance deductions we exclude - see the reasoning in docs/superpowers/research/2026-08-08-tax-structural-ch-ie.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "BAG Kostenbeteiligung https://www.bag.admin.ch/de/krankenversicherung-kostenbeteiligung-fuer-in-der-schweiz-wohnhafte-versicherte + BAG OKP statutory cost-sharing aggregate CHF 5.7bn (2024) over 8.96m insured, adult/child split from BAG age-group net benefits", "last_updated": "2026-08-08", "confidence": "high", "note": "Blended CHF 636 per insured person; children ~CHF 140 each (zero franchise, halved CHF 350 Selbstbehalt cap, exempt from the hospital per-diem), leaving CHF 749 per adult, plus a ~7% conservative uplift for non-KVG self-pay that is neither dental nor long-term care. REPLACES a value that was a single adult's statutory MAXIMUM: the old CHF 1,200 was above the CHF 1,000 minimum-franchise ceiling and was billed unchanged to a family of four whose statutory maximum is CHF 2,700. Geneva keeps its existing 1.083 premium, which is a utilisation difference, not a rules difference: cost-sharing law is federal and identical in every canton. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "BAG Kostenbeteiligung https://www.bag.admin.ch/de/krankenversicherung-kostenbeteiligung-fuer-in-der-schweiz-wohnhafte-versicherte + BAG OKP statutory cost-sharing aggregate CHF 5.7bn (2024) over 8.96m insured, adult/child split from BAG age-group net benefits", "last_updated": "2026-08-08", "confidence": "high", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Swiss Federal Statistical Office (BFS), PxWeb cube px-x-2105000000_206, Agglomeration Bern. https://www.pxweb.bfs.admin.ch/pxweb/de/px-x-2105000000_206/px-x-2105000000_206/px-x-2105000000_206.px", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Durchschnittliche Arbeitswegzeit, in Minuten (ein Hinweg)", "year": "2020", "note": "35.2 min. Agglomeration geography, matching our metro basis. Replaces a value of 22 that was 13 minutes short." }, "non_cash.sick_leave": { "source": "Code of Obligations / Obligationenrecht (OR/CO) Art. 324a, paragraphs 1-4.; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-A.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Steuerverwaltung Kanton Bern, Taxinfo Saeule 3a https://taxinfo.sv.fin.be.ch/taxinfo/a2830da2-1a00-4928-b58b-74da5d48a598 ; EFD/ESTV confirmation of the 2026 amounts, 17 Nov 2025 https://www.estv.admin.ch/de/newnsb/xgRMirCsezICX4rtof9Lm", "last_updated": "2026-08-11", "confidence": "high", "note": "Saeule 3a kleiner Abzug, CHF 7,258 for 2026 (unchanged from 2025), for anyone who also contributes to a Pensionskasse, which our ICP does (both CH cities carry a bvg block). The grosser Abzug (20% of income to CHF 36,288, no 2nd pillar) is NOT our case. Employee-only; ESTV states the maximum deduction is simultaneously the contribution limit. 2nd-pillar Einkauf is also deductible but is bounded by a personal Einkaufspotenzial we cannot know, so it is excluded." } }, "beijing": { "tax.filing_basis": { "source": "PwC WWTS - China (Tax administration): There is no joint tax return in China. Husbands and wives are assessed and taxed separately.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.payroll_deductible": { "source": "Batch B base-deduction convention (tax audit 2026-07-14, verified): China assesses income tax on income net of the employee social contribution, so the progressive scale runs on taxable_income - payroll (computeChinaTaxes (IIT), gated by this flag).", "last_updated": "2026-07-26", "confidence": "high" }, "tax.federal_brackets": { "source": "PRC IIT comprehensive-income schedule (3-45%, 7 brackets) on income net of ¥60k standard deduction: national, peer-identical", "last_updated": "2026-05-29", "confidence": "high" }, "tax.social_security_cap": { "source": "Beijing SI contribution-base ceiling ¥35,811/mo x 12, cycle 2025-07 to 2026-06 (关于2025年度各项社会保险缴费工资基数上下限的通告, Beijing HRSS 2025-09-18: 月缴费基数上限35811元). Confirmed current 2026-08-07. https://rsj.beijing.gov.cn/xxgk/2024zcwj/202509/t20250918_4204880.html", "last_updated": "2026-08-07", "confidence": "high" }, "housing._summary": { "source": "Numbeo + Lianjia Beijing 2026 (city-centre rent)", "last_updated": "2026-05-29", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Public UEBMI (employer + employee payroll); modest out-of-pocket for tech worker", "last_updated": "2026-05-29", "confidence": "medium" }, "childcare._summary": { "source": "Private nursery + public kindergarten Beijing 2026", "last_updated": "2026-05-29", "confidence": "low" }, "food._summary": { "source": "Numbeo Beijing 2026", "last_updated": "2026-05-29", "confidence": "medium" }, "transit._summary": { "source": "Beijing metro monthly ~¥240; private car ownership", "last_updated": "2026-05-29", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo Beijing sports & leisure", "last_updated": "2026-05-29", "confidence": "medium" }, "travel_home._summary": { "source": "PEK/PKX->US economy round-trip 2025-26", "last_updated": "2026-05-29", "confidence": "low" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No statutory national child cash allowance in China = 0 (national, inherited)", "last_updated": "2026-05-29", "confidence": "high" }, "non_cash._summary": { "source": "PRC Labour Law (national): 5 statutory annual leave, 13 public holidays. Parental leave is PER-CITY since batch 4 (2026-08-07): provincial extensions + paternity + yuerjia on the national 98-day base, capped by the local 3x-average-wage ceiling; see the parental_leave_* entries.", "last_updated": "2026-08-07", "confidence": "high" }, "non_cash.parental_leave_weeks": { "source": "Beijing Population and Family Planning Regulation (北京市人口与计划生育条例, amended 2021-11-26): 98 national days + 60-day extension, 15-day paternity, 5 working days yuerjia per parent per year to age 3 https://www.bjrd.gov.cn/xwzx/ymjj/rkyjhsytlxg/202202/t20220228_2618289.html ; State Council Order 619 art. 7 (98-day national base); fund/employer split of the extension: Beijing Fangshan district gov 2025-03-05 https://www.bjfsh.gov.cn/ztzx/2025/fsqganx/jkfw/fmh/202503/t20250305_40086674.shtml", "last_updated": "2026-08-07", "confidence": "high", "note": "26.71 = mother's 158 calendar days (98 + 60 extension) = 22.57 wk, + father's 15 calendar-day paternity = 2.14 wk, + one year's yuerjia of 5 WORKING days per parent (2 x 1.0 wk). Household-inclusive basis (2026-08-01): statutory paid weeks around one birth, non-transferable entitlements summed. Yuerjia counted for the birth year only, not the 3-year cumulative total (ruling D3). Excluded: difficult-birth/multiple +15d uplifts and the discretionary employer-consent 1-3 extra months (D10). Was 14/100, the naked national floor. Research: docs/superpowers/research/2026-08-07-china-parental-batch4-beijing.md + rulings in -batch4.md." }, "non_cash.parental_leave_paid_pct": { "source": "Beijing official Q&A: wages may not be reduced during the extension / paternity / yuerjia days (条例规定休假期间工资不得降低) https://www.beijing.gov.cn/zhengce/zcjd/zcwd/rkyjhsytl/ ; Beijing Second Intermediate Court guidance 2023-03 https://m.bjnews.com.cn/detail/168018095314190.html", "last_updated": "2026-08-07", "confidence": "high", "note": "100 covers only the 8.42 employer-paid remainder weeks (the last 30 extension days + 15 paternity days + 10 household yuerjia working days), all unconditional statutory wage obligations at the employee's own wage. The 18.29 fund-paid weeks live in parental_leave_capped." }, "non_cash.parental_leave_capped": { "source": "京医保发〔2019〕32号 (allowance = employer-unit average monthly contribution wage / 30 x days; 就高补差 top-up rule) via https://zc.51shebao.com/detail/803328 ; ceiling: Beijing HRSS 2025 contribution-base notice, upper limit CNY 35,811/mo from 2025-07 https://rsj.beijing.gov.cn/xxgk/2024zcwj/202509/t20250918_4204880.html", "last_updated": "2026-08-07", "confidence": "high", "note": "128 fund-paid days (98 national + 30 of the 60-day extension; Beijing did NOT extend fund coverage to 158, refutation recorded). pct 100 capped at the 35,811/mo payout ceiling (3x Beijing average wage, 2025-07 to 2026-06 cycle, July refresh). The formula rides the EMPLOYER-average wage, so storing the ceiling is already the most favorable reading; the conditional employer top-up to own wage is deliberately NOT banked on top (ruling D1). Tenure: under 9 contribution months the employer fronts the allowance and the fund reimburses at 12 cumulative months, so a zero-tenure mover is made whole; no haircut encoded (D8)." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Beijing: CNY 651,694, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "CNY 651,694 / 6.76 = USD 96,404; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3856, R5-rounded down to 0.38. REPLACES 0.62 (2026-05-29), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Annual. Do not multiply by 13 to 16 months. Excludes the wu-xian-yi-jin employer contributions. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.62; top-tier mainland tech hub (ByteDance/Baidu)." }, "healthcare.statutory_employee": { "source": "Beijing Basic Medical Insurance Regulation (beijing.gov.cn/gongkai/zfxxgk/zc/gz/202112/t20211215_2561945.html) art. 10, 'zhigong an benren shangyinian yue pingjun gongzi de 2% jiaona jiben yiliao baoxian fei', and art. 38, employer 1% plus CNY 3/month per employee for the large-expense mutual aid fund. Ceiling: Jing Ren She Fa (2025) No. 11, joint HRSS / Medical Insurance / Tax notice (rsj.beijing.gov.cn/xxgk/2024zcwj/202509/t20250918_4204880.html and ybj.beijing.gov.cn/tzgg2022/202509/t20250919_4205869.html): from July 2025 the monthly contribution base upper limit is CNY 35,811 and the lower limit CNY 7,162, one unified figure across all five insurances.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll (medical 2% sits inside the modeled 10.5% employee social insurance rate). China is on _subnational_contribution in data/_healthcare_regime.json: the 2% rate is the national standard and uniform across all five metros, but the contribution BASE CEILING is a local figure at 300% of the local average wage, so the blocks are deliberately not identical. The flat CNY 3/month mutual-aid levy is excluded from `rate` because it is a fixed amount, not a share of pay; it is about CNY 36/yr against a CNY 8,595 ceiling. CYCLE WARNING: the 2025-07 to 2026-06 figure is the one in force; Beijing had not published the 2026-27 ceiling as of 2026-07-31 and its own 2026-06-15 wage-declaration notice says the limits will be announced separately. Recheck in September 2026. A secondary claim of CNY 6,821/month for 2026 was found and REJECTED: it sits below Beijing's own floor of 7,162." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Urban Employee Basic Medical Insurance deductible, coinsurance and annual ceiling, plus the personal medical account", "last_updated": "2026-08-08", "confidence": "low", "note": "THE ONE COUNTRY IN THE DATASET WHERE CHILDREN STEEPEN THE CURVE INSTEAD OF FLATTENING IT, and a future editor should not normalise it toward the pattern everyone else follows. A working parent's UEBMI does not cover their children, who need URRBMI separately and get weaker reimbursement, so a Chinese child costs ~70% of an adult against 20-35% almost everywhere else. LEVEL NOT RE-VERIFIED in the 2026-08-08 sweep. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Urban Employee Basic Medical Insurance deductible, coinsurance and annual ceiling, plus the personal medical account", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Baidu Huiyan and China Academy of Urban Planning and Design, 2024 中国主要城市通勤监测报告 (China Major Cities Commuting Monitoring Report). https://huiyan.baidu.com/boswebsite/cms/report/2024tongqin/", "last_updated": "2026-08-08", "confidence": "high", "statistic": "单程平均通勤时耗 (average one-way commute time)", "year": "2024", "note": "Published ONE WAY, door to door (门到门), all modes (全方式), morning peak, no conversion. MEASUREMENT METHOD: DEVICE-DERIVED from mobile location data, unlike every other sourced city in this field, which is survey self-report; self-report generally runs longer than device-measured, so the two are not interchangeable. GEOGRAPHY: 中心城区 (central urban district), NARROWER than our metro, excluding the longest suburban commutes and therefore biasing the value LOW. That residual bias runs AGAINST our usual conservatism and is accepted only because no metro-wide Chinese series exists." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "MOF / STA Announcement No. 21 of 2024 http://szs.mof.gov.cn/zhengcefabu/202412/t20241213_3949608.htm", "last_updated": "2026-08-11", "confidence": "high", "note": "Private pension (个人养老金), CNY 12,000/yr, deducted as actually paid from comprehensive income. A deduction, not a credit; individual; the limit is stated ANNUALLY in the announcement (the CNY 1,000/month framing is a product convention, not the statute). EET treatment: deductible in, 3% flat on withdrawal. Shebao is already off the base via payroll_deductible, so 12,000 is the private-pension figure ALONE. Enterprise annuity (Circular 103 of 2013, an extra 4% of the contribution-wage base) is rejected for the default because it requires an employer annuity plan most private employers do not run. Housing Provident Fund is separately out of scope." } }, "shanghai": { "tax.filing_basis": { "source": "PwC WWTS - China (Tax administration): There is no joint tax return in China. Husbands and wives are assessed and taxed separately.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.payroll_deductible": { "source": "Batch B base-deduction convention (tax audit 2026-07-14, verified): China assesses income tax on income net of the employee social contribution, so the progressive scale runs on taxable_income - payroll (computeChinaTaxes (IIT), gated by this flag).", "last_updated": "2026-07-26", "confidence": "high" }, "tax.federal_brackets": { "source": "PRC IIT comprehensive-income schedule (3-45%, 7 brackets) on income net of ¥60k standard deduction: national, peer-identical", "last_updated": "2026-05-29", "confidence": "high" }, "tax.social_security_cap": { "source": "Shanghai SI contribution-base ceiling ¥37,302/mo x 12 = 447,624, cycle 2025-07-01 to 2026-06-30 (本市调整2025年度社保缴费基数上下限, Shanghai HRSS 2025-09-18: 上限调整为37302元/月，下限7460元/月; 2024 all-caliber avg wage 12,434/mo). Replaces the two-cycles-stale 36,549 (2023-07 cycle); executes ruling D7 of docs/superpowers/research/2026-08-07-china-parental-batch4.md. https://rsj.sh.gov.cn/tgsgg_17341/20250918/t0035_1435637.html", "last_updated": "2026-08-07", "confidence": "high" }, "housing._summary": { "source": "Numbeo + Lianjia Shanghai 2026 (city-centre rent)", "last_updated": "2026-05-29", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Public UEBMI (employer + employee payroll); modest out-of-pocket", "last_updated": "2026-05-29", "confidence": "medium" }, "childcare._summary": { "source": "Private nursery + public kindergarten Shanghai 2026", "last_updated": "2026-05-29", "confidence": "low" }, "food._summary": { "source": "Numbeo Shanghai 2026", "last_updated": "2026-05-29", "confidence": "medium" }, "transit._summary": { "source": "Shanghai metro monthly ~¥200; private car ownership", "last_updated": "2026-05-29", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo Shanghai sports & leisure", "last_updated": "2026-05-29", "confidence": "medium" }, "travel_home._summary": { "source": "PVG/SHA->US economy round-trip 2025-26", "last_updated": "2026-05-29", "confidence": "low" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No statutory national child cash allowance in China = 0 (national, inherited)", "last_updated": "2026-05-29", "confidence": "high" }, "non_cash._summary": { "source": "PRC Labour Law (national): 5 statutory annual leave, 13 public holidays. Parental leave is PER-CITY since batch 4 (2026-08-07): provincial extensions + paternity + yuerjia on the national 98-day base, capped by the local 3x-average-wage ceiling; see the parental_leave_* entries.", "last_updated": "2026-08-07", "confidence": "high" }, "non_cash.parental_leave_weeks": { "source": "Shanghai Population and Family Planning Regulation (上海市人口与计划生育条例, amended 2021-11-25) art. 31: 98 national days + 60-day shengyujia, 10-day paternity, 5 days yuerjia per parent per year to age 3 https://www.shanghai.gov.cn/jcsfbrkcqjhfzzh/20230621/c440fb200a9b48da87aed931e369792f.html ; implementation koujing 市卫健委 2021-11-25 (Qingpu district mirror) https://www.shqp.gov.cn/wsjkw/wsjkw/upload/202112/1202_170513_376.pdf", "last_updated": "2026-08-07", "confidence": "high", "note": "26.0 = mother's 158 calendar days (98 + 60, all fund-paid) = 22.57 wk, + father's 10 calendar-day paternity (taken continuously during the maternity leave) = 1.43 wk, + one year's yuerjia of 5 working days per parent (2 x 1.0 wk; statute silent on day type, working-day treatment is the administered reading, medium-high). Household-inclusive basis; yuerjia birth-year only (ruling D3). Was 14/100. Research: docs/superpowers/research/2026-08-07-china-parental-batch4-shanghai.md." }, "non_cash.parental_leave_paid_pct": { "source": "条例 art. 31: paternity and yuerjia wages 按照本人正常出勤应得的工资发给 (employer-paid at the normal-attendance wage); base rules 上海市企业工资支付办法 (HRSS interpretation re-issued 2026-07-31) https://rsj.sh.gov.cn/tzcjd_17352_17352/20260731/t0035_1442849.html", "last_updated": "2026-08-07", "confidence": "high", "note": "100 covers only the 3.43 employer-paid remainder weeks (10 paternity days + 10 household yuerjia working days) at the contract normal-attendance wage base, which excludes annual bonus and overtime. The 22.57 fund-paid weeks live in parental_leave_capped." }, "non_cash.parental_leave_capped": { "source": "上海市城镇生育保险办法 + successor medical-insurance notices: allowance = employer prior-year average monthly wage / 30 x days, fund pays at most 300% of the citywide average wage; 2025-07-01 to 2026-06-30 ceiling CNY 37,302/mo (citywide average 12,434) per the official adjustment quoted at https://sh.bendibao.com/shsi/2023629/274144.shtm", "last_updated": "2026-08-07", "confidence": "high", "note": "All 158 maternity days are fund-paid in Shanghai (生育假享受产假同等待遇, implementation koujing 2021-11-25). pct 100 capped at the 37,302/mo payout ceiling. The statutory employer 补差 top-up to the employee's own maternity-wage standard is deliberately NOT banked (ruling D1: formula rides the employer-average wage; the ceiling is already the favorable reading). D7 follow-up EXECUTED 2026-08-07 (small-fixes wave 1): tax.social_security_cap restated 438,588 -> 447,624 against the primary rsj.sh.gov.cn notice; parental cap and tax cap now agree at 37,302/mo. Tenure: <12 cumulative / 9 continuous months = pro-rata fund payment + employer advances the remainder (reimbursed later); mover made whole, no haircut (D8)." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Shanghai: CNY 597,112, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "CNY 597,112 / 6.76 = USD 88,330; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.3533, R5-rounded down to 0.35. REPLACES 0.63 (2026-05-29), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Measures BELOW Beijing, reversing the stored order (Shanghai 0.63 against Beijing 0.62). Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.63; highest mainland tech comp (finance + tech)." }, "healthcare.statutory_employee": { "source": "Shanghai Medical Insurance Bureau notice on continuing the phased reduction of employee basic medical insurance rates, effective 2026-03-01 to 2027-02-28 (ybj.sh.gov.cn/gfxwj/20260313/921e047144694b61b6df8ca0c5ef2cfc.html): 'geren jiaofei bili wei 2%'; corroborated by the Bureau FAQ dated 2026-06-29. Ceiling: Shanghai HRSS Bureau, 2025 social insurance contribution base adjustment published 2025-09-18 (rsj.sh.gov.cn/tgsgg_17341/20250918/t0035_1435637.html): from 2025-07-01 the upper limit is CNY 37,302/month and the lower CNY 7,460, against a 2024 all-caliber city average wage of CNY 12,434/month. Implementing rules for employee basic medical insurance (2025-08-29) confirm the medical base follows the same social insurance base.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll (medical 2% sits inside the modeled 10.5% employee social insurance rate). China is on _subnational_contribution in data/_healthcare_regime.json: the 2% rate is the national standard and uniform across all five metros, but the contribution BASE CEILING is a local figure at 300% of the local average wage, so the blocks are deliberately not identical. NOTE FOR THE TAX BLOCK, not fixed here: tax.social_security_cap for Shanghai is CNY 438,588 (36,549/month), which is the 2024-25 cycle. The current figure is 37,302/month. CYCLE WARNING: the 2026-27 ceiling was not published as of 2026-07-31; the previous one appeared about eleven weeks after its effective date. No employee mutual-aid levy found: this is a negative finding from a direct read of the governing implementing rules rather than an explicit statement that none exists. The trade-union mutual aid product sold in Shanghai is opt-in, not statutory." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Urban Employee Basic Medical Insurance deductible, coinsurance and annual ceiling, plus the personal medical account", "last_updated": "2026-08-08", "confidence": "low", "note": "THE ONE COUNTRY IN THE DATASET WHERE CHILDREN STEEPEN THE CURVE INSTEAD OF FLATTENING IT, and a future editor should not normalise it toward the pattern everyone else follows. A working parent's UEBMI does not cover their children, who need URRBMI separately and get weaker reimbursement, so a Chinese child costs ~70% of an adult against 20-35% almost everywhere else. LEVEL NOT RE-VERIFIED in the 2026-08-08 sweep. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Urban Employee Basic Medical Insurance deductible, coinsurance and annual ceiling, plus the personal medical account", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Baidu Huiyan and China Academy of Urban Planning and Design, 2024 中国主要城市通勤监测报告. https://huiyan.baidu.com/boswebsite/cms/report/2024tongqin/", "last_updated": "2026-08-08", "confidence": "high", "statistic": "单程平均通勤时耗", "year": "2024", "note": "Published one way, door to door, all modes, no conversion. DEVICE-DERIVED (mobile location data), not survey self-report. GEOGRAPHY: 中心城区 central urban district, narrower than our metro, so the value is biased LOW, which runs against our usual conservatism. The previous 50 was unsourced and aggregator-shaped." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "MOF / STA Announcement No. 21 of 2024 http://szs.mof.gov.cn/zhengcefabu/202412/t20241213_3949608.htm", "last_updated": "2026-08-11", "confidence": "high", "note": "Private pension (个人养老金), CNY 12,000/yr, deducted as actually paid from comprehensive income. A deduction, not a credit; individual; the limit is stated ANNUALLY in the announcement (the CNY 1,000/month framing is a product convention, not the statute). EET treatment: deductible in, 3% flat on withdrawal. Shebao is already off the base via payroll_deductible, so 12,000 is the private-pension figure ALONE. Enterprise annuity (Circular 103 of 2013, an extra 4% of the contribution-wage base) is rejected for the default because it requires an employer annuity plan most private employers do not run. Housing Provident Fund is separately out of scope." } }, "shenzhen": { "tax.filing_basis": { "source": "PwC WWTS - China (Tax administration): There is no joint tax return in China. Husbands and wives are assessed and taxed separately.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.payroll_deductible": { "source": "Batch B base-deduction convention (tax audit 2026-07-14, verified): China assesses income tax on income net of the employee social contribution, so the progressive scale runs on taxable_income - payroll (computeChinaTaxes (IIT), gated by this flag).", "last_updated": "2026-07-26", "confidence": "high" }, "tax.federal_brackets": { "source": "PRC IIT comprehensive-income schedule (3-45%, 7 brackets) on income net of ¥60k standard deduction: national, peer-identical", "last_updated": "2026-05-29", "confidence": "high" }, "tax.social_security_cap": { "source": "Guangdong provincial pension contribution-base ceiling ¥27,549/mo x 12, effective 2025-07-01 (广东省人社厅/财政厅/税务局 2025年职工基本养老保险缴费基数上下限通知; Shenzhen-area floor 4,775). Confirmed current 2026-08-07, resolving ruling D6's stale-figure concern. NOTE: this is the provincial pension-branch cap (3x provincial full-caliber avg wage), NOT 3x Shenzhen avg wage; medical/UI branches carry higher city-specific caps a one-cap model cannot express, so employee SI is slightly understated above the pension cap (known limitation, flagged for the annual SI refresh). https://hrss.gd.gov.cn/zwgk/xxgkml/bmwj/qtwj/shbz/content/post_4789618.html", "last_updated": "2026-08-07", "confidence": "high" }, "housing._summary": { "source": "Numbeo + Lianjia Shenzhen 2026 (city-centre rent)", "last_updated": "2026-05-29", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Public UEBMI (employer + employee payroll); modest out-of-pocket", "last_updated": "2026-05-29", "confidence": "medium" }, "childcare._summary": { "source": "Private nursery + public kindergarten Shenzhen 2026", "last_updated": "2026-05-29", "confidence": "low" }, "food._summary": { "source": "Numbeo Shenzhen 2026", "last_updated": "2026-05-29", "confidence": "medium" }, "transit._summary": { "source": "Shenzhen metro monthly ~¥150; private car ownership", "last_updated": "2026-05-29", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo Shenzhen sports & leisure", "last_updated": "2026-05-29", "confidence": "medium" }, "travel_home._summary": { "source": "SZX/HKG->US economy round-trip 2025-26", "last_updated": "2026-05-29", "confidence": "low" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No statutory national child cash allowance in China = 0 (national, inherited)", "last_updated": "2026-05-29", "confidence": "high" }, "non_cash._summary": { "source": "PRC Labour Law (national): 5 statutory annual leave, 13 public holidays. Parental leave is PER-CITY since batch 4 (2026-08-07): provincial extensions + paternity + yuerjia on the national 98-day base, capped by the local 3x-average-wage ceiling; see the parental_leave_* entries.", "last_updated": "2026-08-07", "confidence": "high" }, "non_cash.parental_leave_weeks": { "source": "Guangdong Population and Family Planning Regulation (广东省人口与计划生育条例, amended 2021-12) art. 30: 80-day reward leave + 15-day paternity + 10 days yuerjia per parent per year to age 3; 广东省职工生育保险规定 (provincial decree, eff. 2021-10-01) art. 16, full-text mirror at Shenzhen Medical Security Bureau https://hsa.sz.gov.cn/zwgk/zcfgjzcjd/zcfg/sybx/content/post_1931446.html ; GD HRSS + Health Commission implementation notice 2023-01-11 https://hrss.gd.gov.cn/zwgk/xxgkml/bmwj/gfxwj/content/post_4082488.html", "last_updated": "2026-08-07", "confidence": "high", "note": "27.57 = mother's 178 calendar days (98 national fund-paid + 80 Guangdong reward leave employer-paid) = 25.43 wk, + father's 15-day paternity = 2.14 wk. Identical to guangzhou by provincial law (Shenzhen implements the provincial decree with no deviation, confirmed via the SZ Medical Security Bureau notice). Yuerjia EXCLUDED (ruling D4: statutory pay floor is local minimum wage only). Paternity as CALENDAR days (D5). Was 14/100. Research: docs/superpowers/research/2026-08-07-china-parental-batch4-guangdong.md." }, "non_cash.parental_leave_paid_pct": { "source": "条例 art. 30 para 1 verbatim: 女方享受八十日的奖励假，男方享受十五日的陪产假。在规定假期内照发工资，不影响福利待遇和全勤评奖 (via GD HRSS notice + gz.gov.cn FAQ)", "last_updated": "2026-08-07", "confidence": "high", "note": "100 covers only the 13.57 employer-paid remainder weeks (80 reward days + 15 paternity days) at full normal wage, an unconditional statutory obligation. The 14 fund-paid weeks live in parental_leave_capped." }, "non_cash.parental_leave_capped": { "source": "广东省职工生育保险规定 arts. 15-17: allowance = employer prior-year average monthly wage / 30 x the art.-16 days (98 normal birth; the 80 reward days are NOT fund days); employer advances monthly, surplus to employee, shortfall topped up. Dec-2025 direct-payment notice (Shenzhen, top-up preserved verbatim): https://hsa.sz.gov.cn/xwzx/xwdt/content/post_12523157.html . Ceiling: stored city 3x-local-average-wage SI ceiling CNY 27,549/mo (2025 vintage)", "last_updated": "2026-08-07", "confidence": "medium", "note": "98 fund-paid days only; the fund-coverage-of-reward-leave hypothesis was REFUTED (decree art. 16 enumerates 98 / +30 difficult / +15 per extra multiple). pct 100 capped at 27,549/mo: the batch-4 research verified the formula but not a fresher city ceiling, so the stored 2025-vintage SI-ceiling figure is applied (ruling D6, medium confidence, refresh with the annual SI-ceiling cycle). Employer top-up not banked (D1). Tenure: under 1 year cumulative contributions the employer advances the allowance and claims reimbursement once 12 months accrue; mover made whole, no haircut (D8)." }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Shenzhen. Its basis is converted with the rest of China and its rank inside China is preserved. Prior source: Levels.fyi/Glassdoor Shenzhen SWE 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.62 x 0.6129 = 0.3800, R5-rounded down to 0.38. 0.6129 is the MEDIAN of (re-derived / previous) across the 2 China cities that DO carry an all-levels measurement (Beijing 0.62 to 0.38; Shanghai 0.63 to 0.35), so Shenzhen moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Shenzhen's rank inside China is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.62; Tencent/Huawei/DJI hardware+software hub." }, "healthcare.statutory_employee": { "source": "Shenzhen Medical Security Measures, Government Order No. 358, effective 2023-10-01, art. 8 (hsa.sz.gov.cn/zwgk/zcfgjzcjd/zcfg/ylbx/content/post_10826882.html): Tier 1 is 8% of the base, employer 6% and employee 2%; Tier 2 is 2% total, employer 1.5% and employee 0.5%. Art. 8 also sets the ceiling at 300% of the city all-caliber average monthly wage from two years prior. Art. 14: enrollees and employers pay nothing extra for serious-illness cover. 2026 base notice: medical and maternity contribution base lower limit CNY 6,727 and upper limit CNY 33,633, running 2026-01-01 to 2026-12-31. A 2026 Bureau explainer confirms only the EMPLOYER Tier-1 rate moved, from 5% back to 6% as pandemic relief expired; the employee 2% was never touched.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll (medical 2% sits inside the modeled 10.5% employee social insurance rate). China is on _subnational_contribution in data/_healthcare_regime.json: the 2% rate is the national standard and uniform across all five metros, but the contribution BASE CEILING is a local figure at 300% of the local average wage, so the blocks are deliberately not identical. Shenzhen is the clearest case for the subnational list: its MEDICAL ceiling (CNY 33,633/month, calendar 2026) is structurally decoupled from its pension ceiling, and tax.social_security_cap for Shenzhen is CNY 330,588 (27,549/month), which is the pension figure. That is not an error in the tax block, it is a different ceiling for a different insurance. The pension comparison figure itself is secondary-sourced; only the structural fact that the two cycles differ is primary. Tier 2 (0.5% employee) is deliberately excluded: Tier 1 is mandatory for hukou holders and the standard comparison, and Tier 2 carries materially lower benefits." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Urban Employee Basic Medical Insurance deductible, coinsurance and annual ceiling, plus the personal medical account", "last_updated": "2026-08-08", "confidence": "low", "note": "THE ONE COUNTRY IN THE DATASET WHERE CHILDREN STEEPEN THE CURVE INSTEAD OF FLATTENING IT, and a future editor should not normalise it toward the pattern everyone else follows. A working parent's UEBMI does not cover their children, who need URRBMI separately and get weaker reimbursement, so a Chinese child costs ~70% of an adult against 20-35% almost everywhere else. LEVEL NOT RE-VERIFIED in the 2026-08-08 sweep. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Urban Employee Basic Medical Insurance deductible, coinsurance and annual ceiling, plus the personal medical account", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Baidu Huiyan and China Academy of Urban Planning and Design, 2024 中国主要城市通勤监测报告. https://huiyan.baidu.com/boswebsite/cms/report/2024tongqin/", "last_updated": "2026-08-08", "confidence": "high", "statistic": "单程平均通勤时耗", "year": "2024", "note": "Published one way, door to door, all modes, no conversion. DEVICE-DERIVED, not survey self-report. GEOGRAPHY: 中心城区, narrower than our metro, biasing the value LOW." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "MOF / STA Announcement No. 21 of 2024 http://szs.mof.gov.cn/zhengcefabu/202412/t20241213_3949608.htm", "last_updated": "2026-08-11", "confidence": "high", "note": "Private pension (个人养老金), CNY 12,000/yr, deducted as actually paid from comprehensive income. A deduction, not a credit; individual; the limit is stated ANNUALLY in the announcement (the CNY 1,000/month framing is a product convention, not the statute). EET treatment: deductible in, 3% flat on withdrawal. Shebao is already off the base via payroll_deductible, so 12,000 is the private-pension figure ALONE. Enterprise annuity (Circular 103 of 2013, an extra 4% of the contribution-wage base) is rejected for the default because it requires an employer annuity plan most private employers do not run. Housing Provident Fund is separately out of scope." } }, "guangzhou": { "tax.filing_basis": { "source": "PwC WWTS - China (Tax administration): There is no joint tax return in China. Husbands and wives are assessed and taxed separately.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.payroll_deductible": { "source": "Batch B base-deduction convention (tax audit 2026-07-14, verified): China assesses income tax on income net of the employee social contribution, so the progressive scale runs on taxable_income - payroll (computeChinaTaxes (IIT), gated by this flag).", "last_updated": "2026-07-26", "confidence": "high" }, "tax.federal_brackets": { "source": "PRC IIT comprehensive-income schedule (3-45%, 7 brackets) on income net of ¥60k standard deduction: national, peer-identical", "last_updated": "2026-05-29", "confidence": "high" }, "tax.social_security_cap": { "source": "Guangdong provincial pension contribution-base ceiling ¥27,549/mo x 12, effective 2025-07-01 (广东省人社厅/财政厅/税务局 2025年职工基本养老保险缴费基数上下限通知; Guangzhou/省直 floor 5,510). Confirmed current 2026-08-07, resolving ruling D6's stale-figure concern. NOTE: provincial pension-branch cap (3x provincial full-caliber avg wage), NOT 3x Guangzhou avg wage; medical/UI branches carry higher city-specific caps a one-cap model cannot express, so employee SI is slightly understated above the pension cap (known limitation, flagged for the annual SI refresh). https://hrss.gd.gov.cn/zwgk/xxgkml/bmwj/qtwj/shbz/content/post_4789618.html", "last_updated": "2026-08-07", "confidence": "high" }, "housing._summary": { "source": "Numbeo + Lianjia Guangzhou 2026 (city-centre rent)", "last_updated": "2026-05-29", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Public UEBMI (employer + employee payroll); modest out-of-pocket", "last_updated": "2026-05-29", "confidence": "medium" }, "childcare._summary": { "source": "Private nursery + public kindergarten Guangzhou 2026", "last_updated": "2026-05-29", "confidence": "low" }, "food._summary": { "source": "Numbeo Guangzhou 2026", "last_updated": "2026-05-29", "confidence": "medium" }, "transit._summary": { "source": "Guangzhou metro monthly ~¥150; private car ownership", "last_updated": "2026-05-29", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo Guangzhou sports & leisure", "last_updated": "2026-05-29", "confidence": "medium" }, "travel_home._summary": { "source": "CAN->US economy round-trip 2025-26", "last_updated": "2026-05-29", "confidence": "low" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No statutory national child cash allowance in China = 0 (national, inherited)", "last_updated": "2026-05-29", "confidence": "high" }, "non_cash._summary": { "source": "PRC Labour Law (national): 5 statutory annual leave, 13 public holidays. Parental leave is PER-CITY since batch 4 (2026-08-07): provincial extensions + paternity + yuerjia on the national 98-day base, capped by the local 3x-average-wage ceiling; see the parental_leave_* entries.", "last_updated": "2026-08-07", "confidence": "high" }, "non_cash.parental_leave_weeks": { "source": "Guangdong Population and Family Planning Regulation (广东省人口与计划生育条例, amended 2021-12) art. 30: 80-day reward leave + 15-day paternity + 10 days yuerjia per parent per year to age 3; 广东省职工生育保险规定 (provincial decree, eff. 2021-10-01) art. 16 https://www.gd.gov.cn/zwgk/wjk/qbwj/yfl/content/post_3475284.html ; GD HRSS + Health Commission implementation notice 2023-01-11 https://hrss.gd.gov.cn/zwgk/xxgkml/bmwj/gfxwj/content/post_4082488.html", "last_updated": "2026-08-07", "confidence": "high", "note": "27.57 = mother's 178 calendar days (98 national fund-paid + 80 Guangdong reward leave employer-paid) = 25.43 wk, + father's 15-day paternity = 2.14 wk. Yuerjia EXCLUDED (ruling D4): Guangdong's statutory pay floor for it is the LOCAL MINIMUM WAGE only, full pay merely encouraged (triple-confirmed), a token against an ICP wage per the flat/token convention. Paternity counted as CALENDAR days (statute says 日, conservative reading, ruling D5). Household-inclusive basis. Was 14/100. Research: docs/superpowers/research/2026-08-07-china-parental-batch4-guangdong.md." }, "non_cash.parental_leave_paid_pct": { "source": "条例 art. 30 para 1 verbatim: 女方享受八十日的奖励假，男方享受十五日的陪产假。在规定假期内照发工资，不影响福利待遇和全勤评奖 (via GD HRSS notice + gz.gov.cn FAQ)", "last_updated": "2026-08-07", "confidence": "high", "note": "100 covers only the 13.57 employer-paid remainder weeks (80 reward days + 15 paternity days) at full normal wage, an unconditional statutory obligation. The 14 fund-paid weeks live in parental_leave_capped." }, "non_cash.parental_leave_capped": { "source": "广东省职工生育保险规定 arts. 15-17: allowance = employer prior-year average monthly wage / 30 x the art.-16 days (98 normal birth; the 80 reward days are NOT fund days); employer advances monthly, surplus to employee, shortfall topped up. Dec-2025 direct-payment notice (Guangzhou, top-up preserved): https://www.gz.gov.cn/xw/tzgg/content/post_10569765.html . Ceiling: stored city 3x-local-average-wage SI ceiling CNY 27,549/mo (2025 vintage)", "last_updated": "2026-08-07", "confidence": "medium", "note": "98 fund-paid days only; the fund-coverage-of-reward-leave hypothesis was REFUTED (decree art. 16 enumerates 98 / +30 difficult / +15 per extra multiple). pct 100 capped at 27,549/mo: the batch-4 research verified the formula but not a fresher city ceiling, so the stored 2025-vintage SI-ceiling figure is applied (ruling D6, medium confidence, refresh with the annual SI-ceiling cycle). Employer top-up not banked (D1). Tenure: under 1 year cumulative contributions the employer advances the allowance and claims reimbursement once 12 months accrue; mover made whole, no haircut (D8)." }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Guangzhou. Its basis is converted with the rest of China and its rank inside China is preserved. Prior source: Levels.fyi/Glassdoor Guangzhou SWE 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.52 x 0.6129 = 0.3187, R5-rounded down to 0.31. 0.6129 is the MEDIAN of (re-derived / previous) across the 2 China cities that DO carry an all-levels measurement (Beijing 0.62 to 0.38; Shanghai 0.63 to 0.35), so Guangzhou moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Guangzhou's rank inside China is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.52; trade/manufacturing base, lighter pure-tech than SZ/SH." }, "healthcare.statutory_employee": { "source": "Sui Yi Bao Gui Zi (2022) No. 1, notice on employee medical and maternity insurance funding standards for Guangzhou, in force 2022-12-01 to 2026-12-31 and therefore governing all of 2026 (gz.gov.cn/gfxwj/sbmgfxwj/gzsylbzj/content/post_8689834.html): the individual pays 2% of the contribution base monthly, the employer 6%; the ceiling is 300% of the city all-caliber average monthly wage from two years prior; the large medical expense subsidy is paid from the pooled fund with no additional contribution from employer or enrollee. The 2024-25 rate reduction (Sui Yi Bao Fa (2024) No. 5) cut only employer and flexibly-employed rates through 2025-12-31 and reverted on 2026-01-01; the 2% individual rate was never touched.", "last_updated": "2026-07-31", "confidence": "medium", "note": "DISPLAY ONLY, already inside taxes.payroll (medical 2% sits inside the modeled 10.5% employee social insurance rate). China is on _subnational_contribution in data/_healthcare_regime.json: the 2% rate is the national standard and uniform across all five metros, but the contribution BASE CEILING is a local figure at 300% of the local average wage, so the blocks are deliberately not identical. WEAKEST NUMBER IN THE CHINA SET: the 2026 ceiling of CNY 31,170/month (floor 6,234) could NOT be pinned to a primary gov.cn document despite repeated attempts across the Bureau notice board, the gov.cn disclosure search platform and the municipal tax notice board. It rests on six mutually consistent secondary sources plus an arithmetic cross-check (6,234 x 2% = CNY 124.68/month, matching the independently reported statutory minimum individual payment exactly). The FORMULA behind it is primary. Note ybj.gz.gov.cn does not resolve; Guangzhou medical insurance content lives on the municipal portal at gz.gov.cn/gzybj/. tax.social_security_cap for Guangzhou (CNY 330,588 = 27,549/month) is the Guangdong PROVINCIAL pension figure and correctly differs from this medical ceiling. A union-run CNY 95/period hospitalization mutual-aid product exists and is opt-in, not statutory." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Urban Employee Basic Medical Insurance deductible, coinsurance and annual ceiling, plus the personal medical account", "last_updated": "2026-08-08", "confidence": "low", "note": "THE ONE COUNTRY IN THE DATASET WHERE CHILDREN STEEPEN THE CURVE INSTEAD OF FLATTENING IT, and a future editor should not normalise it toward the pattern everyone else follows. A working parent's UEBMI does not cover their children, who need URRBMI separately and get weaker reimbursement, so a Chinese child costs ~70% of an adult against 20-35% almost everywhere else. LEVEL NOT RE-VERIFIED in the 2026-08-08 sweep. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Urban Employee Basic Medical Insurance deductible, coinsurance and annual ceiling, plus the personal medical account", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Baidu Huiyan and China Academy of Urban Planning and Design, 2024 中国主要城市通勤监测报告. https://huiyan.baidu.com/boswebsite/cms/report/2024tongqin/", "last_updated": "2026-08-08", "confidence": "high", "statistic": "单程平均通勤时耗", "year": "2024", "note": "Published one way, door to door, all modes, no conversion. DEVICE-DERIVED, not survey self-report. GEOGRAPHY: 中心城区, narrower than our metro, biasing the value LOW." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "MOF / STA Announcement No. 21 of 2024 http://szs.mof.gov.cn/zhengcefabu/202412/t20241213_3949608.htm", "last_updated": "2026-08-11", "confidence": "high", "note": "Private pension (个人养老金), CNY 12,000/yr, deducted as actually paid from comprehensive income. A deduction, not a credit; individual; the limit is stated ANNUALLY in the announcement (the CNY 1,000/month framing is a product convention, not the statute). EET treatment: deductible in, 3% flat on withdrawal. Shebao is already off the base via payroll_deductible, so 12,000 is the private-pension figure ALONE. Enterprise annuity (Circular 103 of 2013, an extra 4% of the contribution-wage base) is rejected for the default because it requires an employer annuity plan most private employers do not run. Housing Provident Fund is separately out of scope." } }, "hangzhou": { "tax.filing_basis": { "source": "PwC WWTS - China (Tax administration): There is no joint tax return in China. Husbands and wives are assessed and taxed separately.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.payroll_deductible": { "source": "Batch B base-deduction convention (tax audit 2026-07-14, verified): China assesses income tax on income net of the employee social contribution, so the progressive scale runs on taxable_income - payroll (computeChinaTaxes (IIT), gated by this flag).", "last_updated": "2026-07-26", "confidence": "high" }, "tax.federal_brackets": { "source": "PRC IIT comprehensive-income schedule (3-45%, 7 brackets) on income net of ¥60k standard deduction: national, peer-identical", "last_updated": "2026-05-29", "confidence": "high" }, "tax.social_security_cap": { "source": "Zhejiang provincial SI contribution-base ceiling ¥25,299/mo x 12, 2025 cycle from 2025-01-01 (浙人社发〔2025〕52号: band 4,986-25,299; 300% of 2024 weighted all-unit avg wage; continues into 2026 until the next annual notice). Confirmed current 2026-08-07 via Zhejiang tax bureau Q&A. https://zhejiang.chinatax.gov.cn/art/2025/12/11/art_13314_645797.html", "last_updated": "2026-08-07", "confidence": "high" }, "housing._summary": { "source": "Numbeo + Lianjia Hangzhou 2026 (city-centre rent)", "last_updated": "2026-05-29", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Public UEBMI (employer + employee payroll); modest out-of-pocket", "last_updated": "2026-05-29", "confidence": "medium" }, "childcare._summary": { "source": "Private nursery + public kindergarten Hangzhou 2026", "last_updated": "2026-05-29", "confidence": "low" }, "food._summary": { "source": "Numbeo Hangzhou 2026", "last_updated": "2026-05-29", "confidence": "medium" }, "transit._summary": { "source": "Hangzhou metro monthly ~¥120; private car ownership", "last_updated": "2026-05-29", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo Hangzhou sports & leisure", "last_updated": "2026-05-29", "confidence": "medium" }, "travel_home._summary": { "source": "HGH/PVG->US economy round-trip 2025-26", "last_updated": "2026-05-29", "confidence": "low" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No statutory national child cash allowance in China = 0 (national, inherited)", "last_updated": "2026-05-29", "confidence": "high" }, "non_cash._summary": { "source": "PRC Labour Law (national): 5 statutory annual leave, 13 public holidays. Parental leave is PER-CITY since batch 4 (2026-08-07): provincial extensions + paternity + yuerjia on the national 98-day base, capped by the local 3x-average-wage ceiling; see the parental_leave_* entries.", "last_updated": "2026-08-07", "confidence": "high" }, "non_cash.parental_leave_weeks": { "source": "Zhejiang Population and Family Planning Regulation (浙江省人口与计划生育条例, amended 2021-11-25), official Q&A via Hangzhou gov portal: 158 days first child (calendar, 自然日), 15-day working-day paternity, 10 working days yuerjia per parent per year to age 3 https://www.hzsc.gov.cn/art/2021/12/1/art_1229290183_3974633.html ; 浙江省生育保险办法 浙政发〔2024〕14号 (eff. 2024-07-01) https://www.zjxc.gov.cn/art/2024/7/5/art_1229785649_1903902.html", "last_updated": "2026-08-07", "confidence": "high", "note": "29.57 = mother's 158 calendar days (98 + 60 Zhejiang extension, first child) = 22.57 wk, + father's 15 WORKING-day paternity (excludes holidays/rest days per the official Q&A) = 3.0 wk, + one year's yuerjia of 10 working days per parent (2 x 2.0 wk). Household-inclusive basis; yuerjia birth-year only (ruling D3); the 188-day second/third-child block excluded (one-child basis, D10). Was 14/100. Research: docs/superpowers/research/2026-08-07-china-parental-batch4-zhejiang.md." }, "non_cash.parental_leave_paid_pct": { "source": "条例 official Q&A: 护理假/育儿假期间的 工资、奖金和其他福利待遇由用人单位照发 (employer pays wage, bonus and benefits in full) https://www.hzsc.gov.cn/art/2021/12/1/art_1229290183_3974633.html", "last_updated": "2026-08-07", "confidence": "high", "note": "100 covers only the 7.0 employer-paid remainder weeks (15 working-day paternity + 20 household yuerjia working days) at full own wage including bonus. The 22.57 fund-paid weeks live in parental_leave_capped." }, "non_cash.parental_leave_capped": { "source": "浙政发〔2024〕14号: 生育津贴 = employer prior-year average monthly wage / 30 x the leave days INCLUDING the provincial extension; base bounded 60-300% of provincial average wage. Ceiling: 浙人社发〔2025〕52号, 2025 upper limit CNY 25,299/mo, via Zhejiang tax bureau Q&A 2025-12-11 https://zhejiang.chinatax.gov.cn/art/2025/12/11/art_13314_645797.html", "last_updated": "2026-08-07", "confidence": "high", "note": "All 158 maternity days fund-paid (Zhejiang funds the extension, unlike Guangdong). pct 100 capped at 25,299/mo (2025 cycle, provincial figure used by Hangzhou, still operative Aug 2026 pending the next annual notice). Zhejiang has NO statutory employer top-up: 省政府令第402号 (2023-12-29) DELETED the former soft may-top-up clause and the 2024 办法 is silent, so unlike Beijing/Shanghai/Guangdong this cap is the true ceiling on statutory pay (refutation recorded). Tenure: 6 continuous LOCAL contribution months (before birth, or accrued birth-to-claim) with retroactive back-pay; an enrolled zero-tenure mover is made whole, later; no haircut encoded (D8). A birth before ever enrolling gets nothing from the fund; the employer then owes the pre-leave wage (uninsured branch, 女职工劳动保护办法 art. 16)." }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Hangzhou. Its basis is converted with the rest of China and its rank inside China is preserved. Prior source: Levels.fyi/Glassdoor Hangzhou SWE 2026", "last_updated": "2026-08-11", "confidence": "low", "note": "0.55 x 0.6129 = 0.3371, R5-rounded down to 0.33. 0.6129 is the MEDIAN of (re-derived / previous) across the 2 China cities that DO carry an all-levels measurement (Beijing 0.62 to 0.38; Shanghai 0.63 to 0.35), so Hangzhou moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Hangzhou's rank inside China is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.55; Alibaba/Ant HQ, strong but below BJ/SH/SZ." }, "healthcare.statutory_employee": { "source": "Ceiling: State Taxation Administration Zhejiang Province hot Q&A dated 2025-12-11 (zhejiang.chinatax.gov.cn/art/2025/12/11/art_13314_645797.html): from 2025-01-01 the contribution base upper limit is 300% of the 2024 weighted average wage, CNY 25,299/month. Underlying wage from Zhe Ren She Fa (2025) No. 52, five-department joint notice: 2024 Zhejiang weighted average wage across non-private and private units CNY 101,194 (101,194 / 12 x 3 = 25,298.5). Scope and unified base from the 2026-01-22 joint notice (zhejiang.chinatax.gov.cn/art/2026/1/22/art_8414_84395.html): applies to Zhejiang Province excluding Ningbo, and one contribution-wage declaration covers all insurance types. Rate: Hang Zheng Ban (2013) No. 8 art. 6 states the employee 2% verbatim; it was superseded by Hang Zheng Han (2025) No. 88 effective 2026-01-01, whose live text was unreachable, so the continuing 2% rests on convergent 2025 and 2026 reporting citing the joint Hangzhou Medical Insurance Bureau and Tax Bureau notice.", "last_updated": "2026-07-31", "confidence": "medium", "note": "DISPLAY ONLY, already inside taxes.payroll (medical 2% sits inside the modeled 10.5% employee social insurance rate). China is on _subnational_contribution in data/_healthcare_regime.json: the 2% rate is the national standard and uniform across all five metros, but the contribution BASE CEILING is a local figure at 300% of the local average wage, so the blocks are deliberately not identical. MESSIEST OF THE FIVE. (1) The 2% rate is convergent-secondary for 2026, not a clean current-law quote: zfgb.hangzhou.gov.cn returns 403 and the two mirrors are JS-obfuscated or login-gated, so Hang Zheng Han (2025) No. 88 could not be read. Five independent 2025-2026 sources agree on 2% with no conflicting figure anywhere. (2) TWO COMPETING CEILINGS: CNY 25,299 is confirmed primary; a rumored CNY 26,730 successor appears in secondary sources but matches no gov.cn page, and a third-party tracker claiming to follow official announcements still showed 25,299 as of 2026-07-11. We take the CONFIRMED figure. That is the lower of the two and therefore not the conservative one, but an unverified number is not conservative, it is just unverified. Recheck. (3) GENUINE DATA GAP: the superseded 2013 rule carried a flat CNY 3/month plus a 0.2% pooled-fund transfer for serious illness. Whether either survived into the 2026 rewrite is BLOCKED, not assumed zero." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Urban Employee Basic Medical Insurance deductible, coinsurance and annual ceiling, plus the personal medical account", "last_updated": "2026-08-08", "confidence": "low", "note": "THE ONE COUNTRY IN THE DATASET WHERE CHILDREN STEEPEN THE CURVE INSTEAD OF FLATTENING IT, and a future editor should not normalise it toward the pattern everyone else follows. A working parent's UEBMI does not cover their children, who need URRBMI separately and get weaker reimbursement, so a Chinese child costs ~70% of an adult against 20-35% almost everywhere else. LEVEL NOT RE-VERIFIED in the 2026-08-08 sweep. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Urban Employee Basic Medical Insurance deductible, coinsurance and annual ceiling, plus the personal medical account", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Baidu Huiyan and China Academy of Urban Planning and Design, 2024 中国主要城市通勤监测报告. https://huiyan.baidu.com/boswebsite/cms/report/2024tongqin/", "last_updated": "2026-08-08", "confidence": "high", "statistic": "单程平均通勤时耗", "year": "2024", "note": "Published one way, door to door, all modes, no conversion. DEVICE-DERIVED, not survey self-report. GEOGRAPHY: 中心城区, narrower than our metro, biasing the value LOW." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "MOF / STA Announcement No. 21 of 2024 http://szs.mof.gov.cn/zhengcefabu/202412/t20241213_3949608.htm", "last_updated": "2026-08-11", "confidence": "high", "note": "Private pension (个人养老金), CNY 12,000/yr, deducted as actually paid from comprehensive income. A deduction, not a credit; individual; the limit is stated ANNUALLY in the announcement (the CNY 1,000/month framing is a product convention, not the statute). EET treatment: deductible in, 3% flat on withdrawal. Shebao is already off the base via payroll_deductible, so 12,000 is the private-pension figure ALONE. Enterprise annuity (Circular 103 of 2013, an extra 4% of the contribution-wage base) is rejected for the default because it requires an employer annuity plan most private employers do not run. Housing Provident Fund is separately out of scope." } }, "santiago": { "non_cash.parental_leave_paid_pct": { "source": "Codigo del Trabajo art. 195 / 197 bis; SUSESO subsidio maternal topes (73.2 UF base subsidio, 66 UF permiso postnatal parental full-time) via chileatiende.gob.cl and crececontigo.gob.cl", "last_updated": "2026-08-06", "confidence": "medium", "note": "We store 0: the 2026-08-06 caps rebuild converted Chile to the flat shape because the tope (66 to 73.2 UF/month, roughly USD 35,000 to 39,000/yr) sits BELOW the site's USD 60k ICP salary floor, so the old 100% headline never existed for our reader (the 2026-08-01 owner rule for always-binding caps). The 2026-08-01 note already flagged the tope as ignored per the old field convention; this closes that gap. An 89.9-UF alternate reading was checked and refuted as the general AFP/health contribution tope, a different cap." }, "non_cash.parental_leave_weeks": { "source": "Codigo del Trabajo art. 195 (descanso prenatal 6 and postnatal 12, permiso de paternidad 5 dias) and art. 197 bis (permiso postnatal parental, 12 semanas) https://www.dt.gob.cl/portal/1626/w3-article-99872.html ; SUSESO https://www.suseso.gob.cl/606/w3-propertyvalue-10294.html", "last_updated": "2026-08-01", "confidence": "high", "note": "31 wk = prenatal 6 + postnatal 12 + permiso postnatal parental 12 + permiso de paternidad 1 (5 days). Was 24. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. Caps ignored; unpaid and conditional weeks excluded. See docs/superpowers/research/2026-08-01-parental-leave-basis-and-flat-rate.md. The permiso postnatal parental is a SHARED pot held by the mother, of which at most 6 weeks transfer to the father from week 7, so it is counted once and NOT doubled. Conservative: the 12-week full-time option, not the 18-week half-time variant at 50%. Watch item: a bill extending postnatal leave to a year passed a first reading in July 2025 and is not law." }, "tax.filing_basis": { "source": "PwC WWTS - Chile (Individual); SII Impuesto Unico de Segunda Categoria - salary taxed per employee (unico). Community-property joint return applies only to non-salary Global Complementario income (out of scope).", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.payroll_deductible": { "source": "Batch B base-deduction convention (tax audit 2026-07-14, verified): Chile assesses income tax on income net of the employee social contribution, so the progressive scale runs on taxable_income - payroll (computeChileTaxes (IUSC), gated by this flag).", "last_updated": "2026-07-26", "confidence": "high" }, "tax.federal_brackets": { "source": "SII Impuesto Unico de Segunda Categoria 2025 (8 brackets 0-40%, UTA-denominated, annualized to CLP at UTA ~847,056)", "last_updated": "2026-05-29", "confidence": "high" }, "tax.social_security_cap": { "source": "Chile 2025 taxable-earnings ceiling 90 UF/mo (UF ~40,577) -> ~CLP 43.8M/yr for AFP + health", "last_updated": "2026-05-29", "confidence": "medium" }, "housing._summary": { "source": "Numbeo + Portalinmobiliario Santiago 2026 (Las Condes/Providencia rent)", "last_updated": "2026-05-29", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Isapre private plan premiums (7%+ of wage) + out-of-pocket; Fonasa public alternative", "last_updated": "2026-05-29", "confidence": "medium" }, "childcare._summary": { "source": "Private jardin infantil + sala cuna Santiago 2026", "last_updated": "2026-05-29", "confidence": "low" }, "food._summary": { "source": "Numbeo Santiago 2026", "last_updated": "2026-05-29", "confidence": "medium" }, "transit._summary": { "source": "Red Metropolitana de Movilidad (Metro/bus) monthly ~CLP 40k; private car ownership", "last_updated": "2026-05-29", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo Santiago sports & leisure", "last_updated": "2026-05-29", "confidence": "medium" }, "travel_home._summary": { "source": "SCL->US economy round-trip 2025-26", "last_updated": "2026-05-29", "confidence": "low" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Asignacion familiar is means-tested and ~0 for high earners (national, inherited)", "last_updated": "2026-05-29", "confidence": "high" }, "non_cash._summary": { "source": "Codigo del Trabajo (national, inherited): 15 working-day annual leave, 16 public holidays, 24-week (postnatal incl. parental) leave at 100% via subsidio", "last_updated": "2026-05-29", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Santiago: CLP 32,751,912, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "CLP 32,751,912 / 928.34 = USD 35,280; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1411, R5-rounded down to 0.14. REPLACES 0.37 (2026-05-29), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Chile has no statutory 13th, so a 12-month annualisation is correct here. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.37; top LatAm tech market alongside major MX/BR hubs." }, "non_cash.parental_leave_flat": { "source": "SUSESO / chileatiende.gob.cl: subsidio base capped at 73.2 UF/month (18 prenatal + postnatal weeks), permiso postnatal parental full-time capped at 66 UF/month (12 weeks); UF = CLP 40,844.79 (Banco Central, 2026-08-06). Research: docs/superpowers/research/2026-08-06-parental-caps-americas-africa.md (Chile) and docs/superpowers/research/2026-08-06-parental-caps-southern-cone.md", "last_updated": "2026-08-06", "confidence": "medium", "note": "CLP 620,000/week: the LOWER 66 UF/month tier (66 x 40,844.79 = CLP 2,695,756/month, x12/52 = CLP 622,098/week, stored rounded down) flattened across all 31 weeks. Conservative three ways: the first 18 weeks actually cap at the higher 73.2 UF, the 1 paternity week is uncapped employer pay, and the rounding goes down. UF is inflation-indexed and floats daily, so this is a snapshot; the annual refresh re-derives it. Confidence medium: the qualitative always-binds conclusion is robust across 5+ sources including two government portals, but the exact UF topes live in SUSESO circulars we could not fetch as primary text." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "CASEN commute questions analysed in Razmilic, Centro de Estudios Publicos, Puntos de Referencia 475. https://static.cepchile.cl/uploads/cepchile/2022/09/pder475_srazmilic.pdf", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Tiempo de viaje al trabajo", "year": "2015", "note": "Gran Santiago, one way, all modes. Stored value confirmed." }, "non_cash.sick_leave": { "source": "DFL N°44 of 1978 (Ministerio del Trabajo y Previsión Social, \"fija normas comunes; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "low", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Superintendencia de Pensiones, APV regimen B, 600 UF annual ceiling ; UF value at 1 Aug 2026 = CLP 40,844.79", "last_updated": "2026-08-11", "confidence": "high", "note": "APV regimen B, the base-deduction route: a combined ceiling of 600 UF per calendar year, which is CLP 24,506,874 at the 1 Aug 2026 UF. Regimen A is a 15% state bonificacion, a credit, and is not representable in this field. THE UNDERLYING UNIT MOVES DAILY, because the UF is inflation-indexed, so this figure is a snapshot and belongs on the refresh calendar. Never binds at 6% or 10%: correctness insurance. Sourced to the pension regulator rather than SII, whose PDF would not parse." } }, "bogota": { "tax.filing_basis": { "source": "PwC WWTS - Colombia (Tax administration): Joint returns are not permitted. Income splitting is not allowed.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "DIAN cedula-general schedule (7 brackets 0-39%, UVT-denominated) at 2026 UVT = COP 52,374 (Res. 000238/2025). standard_deduction approximates the 25% labour exemption + mandatory-contribution + 40%-of-net/1340-UVT benefits cap (~COP 85M, within ~2pp of statutory across ICP range)", "last_updated": "2026-05-29", "confidence": "medium" }, "tax.social_security_cap": { "source": "Employee salud 4% + pension 4% + FSP 1% on the IBC (gross), capped at 25 SMMLV/mo; 2026 SMMLV COP 1,750,905 -> ~COP 525.3M/yr", "last_updated": "2026-05-29", "confidence": "high" }, "housing._summary": { "source": "Numbeo Bogota 2026 (Chapinero/Chico/Usaquen city-centre rent)", "last_updated": "2026-05-29", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Fincaraiz listing-count CDF, 2026-08-18 (arriendo apartamentos, Bogota D.C., path-segment filters proven binding after the query-param form was caught as a silent no-op; price filter boundary-probed to bind on canon de arriendo EXCLUDING administracion; asking COP/month per unit): 2BR P25/med/P75 = 1,497,900/2,428,700/4,802,900 (n=2,417) -> 0.617/1.977; 1BR 0.72/1.47 (n=1,813).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, which is TOO NARROW for this dual-market city in the flattering direction on BOTH edges (2BR -0.20/+0.74 - the largest miss measured in the program; visibly bimodal CDF with a 1.7M-2.4M plateau against a long 3M-12M tail, the estrato dual-market signature). Applied 0.62/1.98 (nearest rounding; the guard test carries a documented per-city p75 exception - never a loosened global bound). Pool spans estratos 2-6 (sampled stratum fields); estrato 1/informal absent, so the TRUE citywide spread is wider still - the fix is conservative. CAVEATS: canon-sin-administracion ratios applied to a stored median whose admin basis is UNRECORDED (admin adds ~10-18% where broken out; ratio transfer approximately holds if admin is roughly proportional) - a LEVEL basis question, flagged, out of scope. 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-mexico-colombia.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (B3)." }, "healthcare._summary": { "source": "Medicina prepagada (Sura/Colmedica) on top of payroll-funded EPS", "last_updated": "2026-05-29", "confidence": "medium" }, "childcare._summary": { "source": "Private jardin infantil + preschool Bogota 2026 (Numbeo)", "last_updated": "2026-05-29", "confidence": "low" }, "food._summary": { "source": "Numbeo Bogota 2026", "last_updated": "2026-05-29", "confidence": "medium" }, "transit._summary": { "source": "TransMilenio/SITP monthly ~COP 160k; private car ownership", "last_updated": "2026-05-29", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo Bogota sports & leisure", "last_updated": "2026-05-29", "confidence": "medium" }, "travel_home._summary": { "source": "BOG->Miami economy round-trip 2025-26", "last_updated": "2026-05-29", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "ICBF subsidies means-tested for low income -> ~0 for high earners (national, inherited)", "last_updated": "2026-05-29", "confidence": "high" }, "non_cash._summary": { "source": "Codigo Sustantivo del Trabajo (national, inherited): 15 working-day annual leave, 18 public holidays, 18-week maternity leave at 100% via EPS", "last_updated": "2026-05-29", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Bogotá: COP 111,699,111, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "COP 111,699,111 / 3142.82 = USD 35,541; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1422, R5-rounded down to 0.14. REPLACES 0.18 (2026-05-29), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Prima de servicios treated as already inside, which is the lower read. Closes sweep B's finding that Bogota anchored itself to a Bangalore value Bangalore does not hold. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.18; below CDMX/Sao Paulo (~0.30), ~at/above Bangalore (0.20)." }, "non_cash.parental_leave_weeks": { "source": "Ley 1822/2017 (licencia de maternidad 18 semanas, CST art. 236); Ley 2114/2021 (licencia de paternidad 2 semanas). In-force 2026 state confirmed: the June-2025 labor reform kept 14 days (La República) and Infobae (Feb 2026) confirms 'dos semanas actualmente vigentes'; the DNP unemployment-ladder extension toward 5 weeks never triggered (Dec-2025 rate 10.54% still in draft)", "last_updated": "2026-08-07", "confidence": "high", "note": "20 wk = 18 maternity (mother) + 2.0 paternity (Ley 2114 'dos semanas', the father's own paid entitlement at 100%). Was 18, maternity only. REFUTATION recorded: elempleo.com's '5 semanas for 2026' is the never-reached statutory maximum, not the in-force figure. The licencia parental compartida (mother transfers her last 6 weeks) and flexible (part-time) options are elective redistributions of existing weeks -> add nothing, excluded. Mover gates pass (EPS pro-rata only when gestation is not fully covered; paternity needs ~2 weeks' cotización). Household-inclusive basis, batch 5; research: docs/superpowers/research/2026-08-07-pl-b5-latam.md." }, "non_cash.parental_leave_paid_pct": { "source": "CST art. 236 (licencia remunerada con el salario que devengue); Ley 100/1993 art. 18 as amended (IBC cap 25 SMMLV). Research: docs/superpowers/research/2026-08-06-parental-caps-americas-africa.md (Colombia) and docs/superpowers/research/2026-08-06-parental-caps-andean-central-america.md", "last_updated": "2026-08-06", "confidence": "high", "note": "100% confirmed honest for the WORKER on the 2026-08-06 cap sweep: CST art. 236 entitles her to her actual salary. The 25x-SMMLV IBC ceiling (equal to the stored tax.social_security_cap) caps only the EPS-to-employer reimbursement; the employer absorbs the gap. One uncorroborated secondary claimed a reduced payout for salario-integral contracts; the literal statute text wins and the claim is recorded as an open question, not applied." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "DANE, Gran Encuesta Integrada de Hogares, tiempo de ida al lugar de trabajo, reported via ANIF. https://www.larepublica.co/economia/en-las-principales-capitales-del-pais-ir-al-trabajo-dura-mas-de-30-minutos-en-promedio-4227750", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Tiempo de ida al lugar de trabajo", "year": "2024", "note": "DANE states 49 for Bogota. The statistic name says IDA (outbound), so it is one way. Newspaper URL is the accessible conduit to the DANE/ANIF figure; the underlying statistic is DANE's. Bogota's Encuesta bienal de cultura ciudadana 2025 was rejected as a source because its 'una hora' framing is a round-trip restatement." }, "non_cash.sick_leave": { "source": "Código Sustantivo del Trabajo (CST) Art. 227 (auxilio monetario por incapacidad,; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": "https://www.gerencie.com", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Estatuto Tributario art. 336 (40% / 1,340 UVT cap on exempt income plus deductions) ; UVT 2026 = COP 52,374", "last_updated": "2026-08-11", "confidence": "high", "note": "A real allowance exists but the engine already grants MORE than the statutory maximum, so a nonzero cap here would double-count. ET art. 336 caps all exempt income plus deductions at 40% of net income, hard-capped at 1,340 UVT = COP 70,181,160 for 2026. Worked at a Bogota senior gross of COP 180m: mandatory INCRNGO 16.2m plus the capped stack 65.5m gives a statutory maximum relief of COP 81.7m, LESS than the flat COP 85,000,000 standard_deduction the engine already applies, with voluntary contributions fully used. At COP 250m there is about 7.7m of headroom but it is income-dependent and not expressible as a fixed max. THE HONEST FIX is replacing the 85m lump with an explicit stack, which is a Colombia-branch change, not this field." } }, "buenos-aires": { "tax.federal_brackets": { "source": "AFIP/ARCA Impuesto a las Ganancias 2025 (9 brackets 5-35%, Art. 94 escala, indexed semiannually by RIPTE). standard_deduction approximates ganancia no imponible (MNI) + deduccion especial (3.5x MNI for employees, Art. 30(c) ap. 2) + conyuge/hijo deductions; aportes (17%, capped) are separately deducted from the income-tax base in the engine", "last_updated": "2026-05-29", "confidence": "medium" }, "tax.social_security_cap": { "source": "Employee aportes 17% (11% jubilacion SIPA + 3% PAMI/INSSJP + 3% obra social) on remuneracion, capped at the base imponible maxima ~ARS 3.82M/mo (2025) -> ~ARS 45.88M/yr", "last_updated": "2026-05-29", "confidence": "medium" }, "housing._summary": { "source": "Numbeo + Zonaprop Buenos Aires 2026 (Palermo/Recoleta/Belgrano rent)", "last_updated": "2026-05-29", "confidence": "low", "note": "Peso inflation makes ARS COL stale fast; re-verify quarterly." }, "housing.rent_percentiles": { "source": "Mercado Libre AR listing-count CDF, 2026-08-18 (departamentos en alquiler, CABA pool, fixed dormitorios path facet; the values_with_conversion block proves ONE unified price dimension with USD-priced listings converted at ~1,492 ARS/USD - identical bucket counts under ARS and USD labels, sum = pool; expensas separate; asking ARS/month per unit; slice-page totals authoritative over stale base-page facets): 2 dormitorios P25/med/P75 = 950,202/1,294,937/2,037,179 (n=1,918) -> 0.734/1.573; 1 dormitorio 0.820/1.370 (n=4,746).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW on both 2BR edges (-0.086/+0.333, the upper edge 5.5x tolerance) - a full dual-market miss: the post-2024-deregulation CABA market splits an ARS mass segment from a USD-priced premium segment, which the portal's own conversion folds into one distribution. Applied 0.73/1.57 (nearest). Zonaprop 403s and Argenprop soft-blocks after a few requests (URLs+dates recorded; its one pre-block datapoint shows a similar pool size). Informal/GBA cheap stock outside the CABA pool - the true spread is wider, the fix conservative. IMPORTANT dormitorios not ambientes: the portal facet is true bedrooms (3 ambientes = 2 dormitorios cross-check recorded). 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-southern-cone.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (F2). NOTE the stored ARS median levels read stale-low against the measured pool (~24-34% below market per the agent) - inflation drift, a LEVEL question, flagged, out of scope here." }, "healthcare._summary": { "source": "Prepaga (OSDE/Swiss Medical/Galeno) premiums on top of payroll-funded obra social", "last_updated": "2026-05-29", "confidence": "low" }, "childcare._summary": { "source": "Private jardin maternal + preescolar Buenos Aires 2026", "last_updated": "2026-05-29", "confidence": "low" }, "food._summary": { "source": "Numbeo Buenos Aires 2026", "last_updated": "2026-05-29", "confidence": "low" }, "transit._summary": { "source": "SUBE monthly (subte/colectivo) ~ARS 35k; private car ownership", "last_updated": "2026-05-29", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Buenos Aires sports & leisure", "last_updated": "2026-05-29", "confidence": "low" }, "travel_home._summary": { "source": "EZE->Miami economy round-trip 2025-26", "last_updated": "2026-05-29", "confidence": "low" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Asignaciones Familiares (AUH/SUAF) means-tested by income ceiling -> ~0 for high earners (national, inherited)", "last_updated": "2026-05-29", "confidence": "high" }, "non_cash._summary": { "source": "Ley de Contrato de Trabajo (national, inherited): 14 calendar-day (~10 working-day) annual leave minimum, 16 feriados, ~13-week (90-day) maternity leave at 100% via ANSES", "last_updated": "2026-05-29", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Buenos Aires: ARS 59,353,348, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "ARS 59,353,348 / 1488.45 = USD 39,876; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1595, R5-rounded down to 0.15. REPLACES 0.3 (2026-05-29), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Observation month and FX basis (official against parallel) are both unpublished, and pooled ARS under high inflation is not a stable quantity. At the early-August 2026 official rate of about 1,494 this is roughly USD 39,700. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.30; strong LatAm tech talent pool but USD comp depressed by peso; ~at Sao Paulo/CDMX." }, "tax.filing_basis": { "source": "PwC Worldwide Tax Summaries: Argentina, Individual: Tax administration (the earnings of each spouse are assessable separately and individually). https://taxsummaries.pwc.com/argentina/individual/tax-administration", "last_updated": "2026-07-26", "note": "AR-1 (closeout 2026-07-26): Ganancias is individual; each spouse assessed separately with their own deductions." }, "non_cash.parental_leave_paid_pct": { "source": "Ley 24.714 art. 11: asignacion por maternidad = 100% of the salary the worker would have earned, with maternity expressly carved out of the income ceiling that gates other ANSES family allowances. Research: docs/superpowers/research/2026-08-06-parental-caps-americas-africa.md (Argentina) and docs/superpowers/research/2026-08-06-parental-caps-southern-cone.md", "last_updated": "2026-08-06", "confidence": "high", "note": "100% confirmed genuinely UNCAPPED on the 2026-08-06 cap sweep: Ley 24.714 art. 11 structurally exempts maternity from the ANSES upper income cutoff, so the rate is real at every ICP salary. Recorded so nobody re-runs the lookup; five independent 2026 sources agree it is paid sin tope." }, "non_cash.parental_leave_weeks": { "source": "LCT (Ley 20.744 t.o., InfoLEG consolidation current through Ley 27.802 B.O. 6/3/2026) art. 177: 90 calendar days maternity (pre-birth electable down to 10 days since Ley Bases 27.742 art. 93, B.O. 8/7/2024; 90-day total unchanged) + art. 158(a) 'dos (2) dias corridos' paternity, employer-paid per art. 159. Research: docs/superpowers/research/2026-08-08-pl-b7-latam.md", "last_updated": "2026-08-08", "confidence": "high", "note": "13.1 wk = 90/7 (12.86, mother) + 2/7 (0.29, father), rounded down from 13.14. Was 13, maternity-only; the paternity add is the batch-7 household-inclusive correction. Ley 27.802 (B.O. 6/3/2026) amended ~65 LCT spots but left arts. 158/177 untouched (verified in the consolidation itself); no paternity-extension bill enacted through 2026-08-08 (refutation recorded). Gate: Ley 24.714 art. 11 asignacion por maternidad requires 3 months' continuous tenure - the zero-tenure mover (~7-10.5 months at leave start) passes, caveat only. CBA paternity extensions are non-statutory, excluded; CABA's 10-day paternity is public-sector only." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "INDEC, Encuesta Nacional de Uso del Tiempo 2021, resultados definitivos, Gran Buenos Aires. https://www.indec.gob.ar/ftp/cuadros/sociedad/enut_2021_resultados_definitivos.pdf", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Tiempo de traslado al trabajo, ida y vuelta", "year": "2021", "note": "CONVERSION: published 95 min per day is explicitly IDA Y VUELTA (round trip); 95/2 = 47.5 one way." }, "non_cash.sick_leave": { "source": "Ley de Contrato de Trabajo (LCT), Ley 20.744 (t.o. 1976) , Art. 208 (accidentes y; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": "https://www.globalization-partners.com/globalpedia/argentina-employer-of-record/", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "ARCA (ex-AFIP), deducciones generales 2026, tope for seguros de retiro privado", "last_updated": "2026-08-11", "confidence": "high", "note": "Deduction for seguros de retiro privado, ARS 753,472.14 for 2026, about USD 506, roughly a fifth of a 6% contribution at an ARS 60m gross. The phase-1 lead that the deduction had been eliminated is REFUTED; it survives, it is just tiny. The same figure is published for life insurance and it could not be confirmed whether the two topes are shared or separate; treating them separately is the generous reading and is flagged. FASTEST-AGEING VALUE IN THIS FIELD: IPC-indexed, so it is stale within a year." } }, "lima": { "tax.filing_basis": { "source": "PwC WWTS - Peru; SUNAT: fifth-category (employment) income of each spouse is declared independently; only common-property income may optionally be attributed to one spouse.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "SUNAT renta de quinta categoría 2025 (5 brackets 8-30%, UIT-denominated at 2025 UIT = PEN 5,350, applied to income net of the 7-UIT deduction). standard_deduction = 7 UIT (37,450); the optional +3-UIT receipted-expense deduction is omitted", "last_updated": "2026-05-29", "confidence": "high" }, "tax.social_security_cap": { "source": "Employee AFP ~12.7% (10% fund + ~1.4% comisión + 1.37% prima de seguro) on gross; only the prima sliver is capped (SBS tope) so modeled UNCAPPED (cap null). EsSalud 9% health is employer-paid (employee medicare_rate 0)", "last_updated": "2026-05-29", "confidence": "medium" }, "housing._summary": { "source": "Urbania + Adondevivir Lima 2026 (Miraflores/San Isidro/Barranco rent)", "last_updated": "2026-05-29", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Private EPS/clinic plan (Pacífico/Rímac) on top of payroll-funded EsSalud", "last_updated": "2026-05-29", "confidence": "low" }, "childcare._summary": { "source": "Private cuna/guardería + nido Lima 2026 (Cuna Más is poverty-targeted → 0 subsidy for high earners)", "last_updated": "2026-05-29", "confidence": "low" }, "food._summary": { "source": "Numbeo Lima 2026 (Wong/Vivanda upper-middle basket)", "last_updated": "2026-05-29", "confidence": "medium" }, "transit._summary": { "source": "Metropolitano + Metro Línea 1 + combi blend ~PEN 120/mo; private car ownership", "last_updated": "2026-05-29", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Lima sports & leisure", "last_updated": "2026-05-29", "confidence": "low" }, "travel_home._summary": { "source": "LIM->Miami economy round-trip 2025-26", "last_updated": "2026-05-29", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No universal statutory child allowance for formal high earners → 0 (national, inherited)", "last_updated": "2026-05-29", "confidence": "high" }, "non_cash._summary": { "source": "Peru labor law (national, inherited): 30 calendar-day (~22 working-day) annual leave (D.Leg. 713), 16 feriados, 98-day (~14-week) maternity leave at 100% via EsSalud", "last_updated": "2026-05-29", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Lima: PEN 83,231, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "PEN 83,231 / 3.39 = USD 24,552; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.0982, R5-rounded down to 0.09. REPLACES 0.18 (2026-05-29), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Peru pays two statutory gratificaciones, so a formal-contract monthly divisor is about 14, not 12. Thin sample. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.18; smaller/lower-paying than CDMX/Sao Paulo (~0.30), ~at Bogotá (0.18)/Bangalore (0.20)." }, "non_cash.parental_leave_paid_pct": { "source": "Ley 26790 / EsSalud subsidio por maternidad, Regimen 728: subsidio on ACTUAL average pay of the last 12 months, no tope maximo expreso. Research: docs/superpowers/research/2026-08-06-parental-caps-americas-africa.md (Peru) and docs/superpowers/research/2026-08-06-parental-caps-andean-central-america.md", "last_updated": "2026-08-08", "confidence": "high", "note": "100% confirmed uncapped for the general private-sector regime (728) on the 2026-08-06 cap sweep. WARNING for future editors: the S/1,650/month cap that appears in searches belongs to Regimen CAS (public-sector service contracts, 30% of UIT) and does NOT apply to a private-sector professional. Confidence medium-high: EsSalud's own page is silent on any cap and four secondaries converge, but a readable primary Ley 26790 text was not fetched. UPGRADE 2026-08-08: Ley 26790 primary text now read in full - art. 12 contains NO monetary ceiling for the maternity subsidy (Regimen 728); the no-tope verdict moves medium-high -> high. Paternity days are employer-paid at full salary (Ley 30807)." }, "non_cash.parental_leave_weeks": { "source": "EsSalud subsidio por maternidad: 98 days (49 pre + 49 post), official page essalud.gob.pe/maternidad/; Ley 30807 (2018, El Peruano 2018-07-05): licencia por paternidad 10 consecutive CALENDAR days, employer-paid full salary. Ley 26790 primary PDF read in full (essalud.gob.pe/transparencia/pdf/publicacion/ley26790.pdf). Research: docs/superpowers/research/2026-08-08-pl-b6-latam-romania.md", "last_updated": "2026-08-08", "confidence": "high", "note": "15.43 wk = 14.0 maternity (98/7) + 1.43 paternity (10 dias calendario consecutivos /7). Was 14, maternity only. Special paternity cases (20d premature/multiple, 30d severe) are conditional, excluded. REFUTATION recorded: the 15-day paternity extension is NOT law - the March-2024 autografa was observed (returned) by the Executive in July 2024, and the 2026 bills (12680/13310/13151) died in commission when the legislature ended 2026-06-24; still 10 days. EsSalud gate (Ley 26790 art. 10, primary: 3 consecutive contribution months, or 4 non-consecutive within 6, before the causal month + employed at conception and at subsidy start): the mover passes cleanly, no caveat. both_parents not set." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "INEI, Encuesta Nacional de Uso del Tiempo 2024, reported in BCRP working paper DT-2025-009. https://www.bcrp.gob.pe/docs/publicaciones/documentos-de-trabajo/dt-2025/dt-2025-009.pdf", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Tiempo de traslado al trabajo, ida y vuelta", "year": "2024", "note": "CONVERSION: published 1.7 hours per day is IDA Y VUELTA (round trip); 102/2 = 51 one way. Stored value effectively confirmed. The widely repeated '107 minutos' figure was rejected as a round-trip restatement." }, "non_cash.sick_leave": { "source": "Ley 26790 (Ley de Modernización de la Seguridad Social en Salud) and its reglamento; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": "https://www.perucontable.com/laboral/subsidio-por-incapacidad-temporal/", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "SUNAT Informe N. 000054-2025-SUNAT/7T0000 ; SUNAT 3-UIT additional deduction list", "last_updated": "2026-08-11", "confidence": "high", "note": "No employment-base deduction. The AFP is not deductible (the 7-UIT deduction is in lieu, and the engine comment at the Peru branch already says so). The Ley 32123 headline is a false friend: secondary sources describe the new voluntary-contribution benefit as \"deducible\", but SUNAT states it exempts the RETURNS (dividendos, intereses, rendimientos) while funds stay in the system, which is investment-income relief, not employment-base relief. SUNAT's 3-UIT qualifying-expense list is closed and contains no AFP line." } }, "quito": { "tax.filing_basis": { "source": "PwC WWTS - Ecuador (Individual): joint filing is not permitted; each individual files a separate return.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "SRI impuesto a la renta 2025 tabla para personas naturales (9 brackets 0-37%, USD-denominated; first band to ~USD 12,081 is the basic exemption). standard_deduction = 0: the IESS deduction and gastos-personales rebaja are modeled explicitly (see tax.payroll_deductible and tax.tax_credit_annual), not folded into the ladder.", "last_updated": "2026-05-29", "confidence": "high" }, "tax.social_security_cap": { "source": "Employee IESS aporte personal 9.45% on materia gravada with no ceiling → modeled UNCAPPED (cap null). Health funded within IESS (employee medicare_rate 0). Ecuador is USD-dollarized (no FX conversion)", "last_updated": "2026-05-29", "confidence": "high" }, "tax.payroll_deductible": { "source": "The IESS aporte personal (9.45%) is deductible from the income-tax base: base imponible = ingreso gravado menos aportes personales al IESS (LRTI, applied by SRI). This flag turns on the base-net-of-IESS convention inside computeEcuadorTaxes; above the exempt band it always cuts tax.", "source_name": "IESS (Instituto Ecuatoriano de Seguridad Social) / SRI", "last_updated": "2026-07-23", "confidence": "high", "note": "Correct-first item #6 of the 2026-07-17 resolutions doc. Ecuador-only key. The four countries that reuse computeEcuadorTaxes (Georgia, Armenia, Azerbaijan, Uzbekistan) omit it and keep the full-gross ladder, so their output is unchanged." }, "tax.tax_credit_annual": { "source": "Rebaja por gastos personales (SRI, a tax CREDIT since the 2021 reform): 18% of the lesser of declared personal expenses or a canastas-basicas cap. Single filer with no dependents = 7 canastas; canasta basica Jan 2026 = USD 821.80, so max rebaja = 0.18 x 7 x 821.80 = USD 1,035.47. Non-refundable, so federal income tax floors at 0.", "source_name": "SRI (Servicio de Rentas Internas); cap cross-checked at hlbecuador.com gastos-personales-2026", "last_updated": "2026-07-23", "confidence": "medium", "note": "MED-HIGH per the resolutions doc: the 1,035.47 single-filer figure is firm, but the cap scales with dependents (9 canastas / USD 1,331.32 at 1 dependent, 11 / USD 1,627.16 at 2) which the single stored figure cannot express (the credit is understated for parents, the conservative direction), and Quito's stacked two-earner household draws a single household credit. Ecuador-only key; absent for the four reusing countries. TODO (each January): re-verify the canasta basica." }, "housing._summary": { "source": "Plusvalia + Properati Quito 2026 (Cumbayá/González Suárez/La Carolina rent)", "last_updated": "2026-05-29", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Properati listing-count CDF, 2026-08-18 (apartments for rent, Quito DMQ incl. valley belt, fixed habitaciones filter verified binding, pagination via path segment after two silent no-op params were caught and documented; asking USD/month per unit, alicuota separate): 2BR P25/med/P75 -> 0.684/1.316 (n=230); 1BR 0.625/1.312 (n=58, thin; 0.77 lower if one 10x-multiposted micro-suite building is deduped).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, which is TOO NARROW for this dual-market city in the flattering direction (2BR lower edge -0.136, more than 2x tolerance; upper +0.076). Applied 0.68/1.32, measured pair rounded outward. Sensitivity: city-proper-only still fails the lower edge (0.699); valleys-only reads 0.850/1.400. Pool caveats: single portal (Plusvalia 403s to automated fetch, 2026-08-17/18); Sur de Quito is ~2% of the pool vs ~1/3 of the population, so the TRUE citywide spread is wider than measured - the applied pair is conservative. 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw price vector: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-andes.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (A2). NOTE the stored median's own source names the premium corridor (Cumbaya/Gonzalez Suarez/La Carolina) while the CDF covers the whole DMQ pool - a LEVEL basis question, flagged, out of scope here." }, "healthcare._summary": { "source": "Private seguro médico (Salud S.A./BMI) on top of payroll-funded IESS", "last_updated": "2026-05-29", "confidence": "low" }, "childcare._summary": { "source": "Private guardería + jardín Quito 2026 (CNH/CIBV are poverty-targeted → 0 subsidy for high earners)", "last_updated": "2026-05-29", "confidence": "low" }, "food._summary": { "source": "Numbeo Quito 2026 (Supermaxi upper-middle basket)", "last_updated": "2026-05-29", "confidence": "medium" }, "transit._summary": { "source": "Quito Metro + Trolebús/Ecovía ~USD 25/mo; private car ownership", "last_updated": "2026-05-29", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Quito sports & leisure", "last_updated": "2026-05-29", "confidence": "low" }, "travel_home._summary": { "source": "UIO->Miami economy round-trip 2025-26", "last_updated": "2026-05-29", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No universal statutory child allowance for formal high earners → 0 (national, inherited)", "last_updated": "2026-05-29", "confidence": "high" }, "non_cash._summary": { "source": "Ecuador Código del Trabajo (national, inherited): 15 calendar-day (~11 working-day) annual leave, 11 feriados nacionales, 12-week maternity leave at 100% (IESS subsidio)", "last_updated": "2026-05-29", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Quito: USD 32,400, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "USD 32,400 / 1 = USD 32,400; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1296, R5-rounded down to 0.12. REPLACES 0.14 (2026-05-29), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Ecuador is dollarized, so USD IS the local currency here and this figure must NOT be converted. Very thin: the p25 sits 2.5x below the median. The 13th and 14th salaries are already inside. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.14; smaller tech market than Lima (0.18)/Bogotá (0.18); dollarized so figures are directly USD." }, "non_cash.parental_leave_paid_pct": { "source": "IESS maternity subsidy: 75% of last salary for 12 weeks, employer tops up 25% to full pay; IESS Salarios de Aportacion page states only a minimum base, no maximum. Research: docs/superpowers/research/2026-08-06-parental-caps-americas-africa.md (Ecuador) and docs/superpowers/research/2026-08-06-parental-caps-andean-central-america.md", "last_updated": "2026-08-06", "confidence": "high", "note": "100% confirmed uncapped end-to-end on the 2026-08-06 cap sweep (IESS 75% leg has no contribution ceiling, employer 25% top-up is ordinary wage continuation; stored tax.social_security_cap is independently null). REFUTATION recorded: a \"5x SBU\" contribution-ceiling claim surfaced in search synthesis twice, but IESS's own page has no maximum language and the cited article does not contain the claim; do not re-import it." }, "non_cash.parental_leave_weeks": { "source": "Codigo del Trabajo art. 152 as amended by the Ley Organica del Derecho al Cuidado Humano (R.O. Supl. 309, 12-V-2023; reform text read verbatim: 'plazo de diez dias' -> 'plazo de quince (15) dias'): 12 weeks maternity + 15 calendar days paternity 'contados desde la fecha del parto'. Consolidated art. 152 cross-checked via the Camara de Comercio de Quito consulta + four 2026 guides. Research: docs/superpowers/research/2026-08-08-pl-b7-latam.md", "last_updated": "2026-08-08", "confidence": "high", "note": "14.1 wk = 12 (mother) + 15/7 (2.14, father), rounded down from 14.14. Was 12, maternity-only, and the pre-2023 10-day paternity figure is STALE (ecuadorlegalonline still shows it - do not re-import). The Economia Violeta mother-father sharing option was DEROGATED by the same 2023 law: no shared pot. Conditional add-ons excluded (paternity +5 c-section/multiples, +8 premature, 25 d severe illness; maternity +10 multiples). Lactancia 6-hour workday for 15 months = working-time reduction, excluded; the 15-month optional leave (art. 152.1) is unpaid, excluded. IESS gate: 12 continuous monthly contributions pre-birth - a ~9-12-month mover can fail, but the consequence is a PAYER SHIFT (CT art. 153: employer owes remuneracion completa when IESS does not cover), not a rate cut. WATCH: an Asamblea bill to double maternity leave is under discussion (2026-08-06 press), not enacted." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series", "last_updated": "2026-08-08", "confidence": "low", "note": "FLAGGED for a future look: 9.5x the national mean, with no country-specific source found to explain the gap. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. INEC publishes no commute-duration series and the Quito mobility plan carries no mean." }, "non_cash.sick_leave": { "source": "Código del Trabajo (employer's first 3 days); Ley de Seguridad Social / Reglamento; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": "https://www.ecuadorlegalonline.com/laboral/subsidio-por-enfermedad/", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "SRI, rebaja por gastos personales (the SRI page would not load; cross-checked against the engine)", "last_updated": "2026-08-11", "confidence": "medium", "note": "No deduction. Ecuador relieves personal spending through a REBAJA against the tax, and retirement saving is not one of the covered categories. Cross-check that validates the engine rather than the value: 18% x (7 x canasta 821.80) = USD 1,035.47, matching quito.json's existing tax_credit_annual of 1035.47 to the cent. MEDIUM because SRI's own page returned an error and the finding rests on second-best sources." } }, "montevideo": { "tax.filing_basis": { "source": "DGI: individual is the default IRPF basis; nucleo familiar (joint) is an optional year-by-year election for employment income, and individual is the two-earner-beneficial default.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "DGI IRPF Categoría II 2025 (8 brackets 0-36%, BPC-denominated at 2025 BPC = UYU 6,576/mo → 78,912/yr, annualized). standard_deduction = 0: the real two-step impuesto-sobre-deducciones credit is approximated by running the scale on full income (validated within +1.8pp for an ICP earner)", "last_updated": "2026-05-29", "confidence": "high" }, "tax.social_security_cap": { "source": "Employee BPS blended ~19%: jubilación 15% (tope jubilatorio ~UYU 3.27M/yr) + FONASA health 4.5-8% by family (separately capped ~UYU 100k/yr) + FRL 0.1%, modeled as one rate on min(gross, jubilática cap). FONASA folded into payroll (medicare_rate 0); on-top-of-mutualista copays in healthcare", "last_updated": "2026-05-29", "confidence": "medium" }, "housing._summary": { "source": "InfoCasas + Suárez Figari Montevideo 2025 (Pocitos/Punta Carretas/Carrasco rent)", "last_updated": "2026-05-29", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Out-of-pocket on top of FONASA mutualista: emergencia móvil + tasas moderadoras + partial private seguro", "last_updated": "2026-05-29", "confidence": "low" }, "childcare._summary": { "source": "Private guardería + jardín Montevideo 2025 (AFAM is poverty-targeted → 0 subsidy for high earners)", "last_updated": "2026-05-29", "confidence": "low" }, "food._summary": { "source": "Numbeo Montevideo 2026 (Tienda Inglesa/Disco upper-middle basket)", "last_updated": "2026-05-29", "confidence": "medium" }, "transit._summary": { "source": "Montevideo STM boleto (no unlimited pass; ~46 rides/mo modeled) + private car ownership", "last_updated": "2026-05-29", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Montevideo sports & leisure + Punta del Este/interior vacation", "last_updated": "2026-05-29", "confidence": "low" }, "travel_home._summary": { "source": "MVD->Miami economy round-trip 2025-26", "last_updated": "2026-05-29", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Asignación Familiar is means-tested (AFAM-PE) → 0 for a high earner (national, inherited)", "last_updated": "2026-05-29", "confidence": "high" }, "non_cash._summary": { "source": "Uruguay labor law (national, inherited): 20 working-day annual leave (Ley 12.590), ~12 feriados, 14-week maternity leave at 100% via BPS (Ley 19.161); paternity 13 days", "last_updated": "2026-05-29", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Montevideo: UYU 1,818,802, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "UYU 1,818,802 / 40.16 = USD 45,289; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1812, R5-rounded down to 0.18. REPLACES 0.22 (2026-05-29), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. The page itself says 'not enough data for top locations'. Large USD-paid offshore cohort. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.22; strong nearshore/outsourcing ecosystem (Globant/MercadoLibre/dLocal) lifts it above Lima/Bogotá (0.18) but absolute salaries sit below Buenos Aires/Santiago/São Paulo (~0.30)." }, "childcare.preschool_monthly": { "source": "docs/superpowers/research/2026-08-05-santiago-montevideo-spotcheck.md (2ab38ecc adjudication + refinement)", "last_updated": "2026-08-05", "confidence": "medium", "note": "Raised 15,000 to 16,500 UYU/mo (owner decision 2026-08-05): the elite-tier report was rejected under rule 2, but the sourcing pass supported a modest upward refinement; conservative direction for a destination city." }, "non_cash.parental_leave_paid_pct": { "source": "Ley 19.161 arts. 6 and 9 (impo.com.uy primary text): subsidio at 100% with only a FLOOR (nunca inferior a 2 BPC/mes), no ceiling clause; BPS benefit page concurs. Research: docs/superpowers/research/2026-08-06-parental-caps-americas-africa.md (Uruguay) and docs/superpowers/research/2026-08-06-parental-caps-southern-cone.md", "last_updated": "2026-08-08", "confidence": "high", "note": "74 = the duration-weighted household blend (14 wk maternity @ 100 + 2.86 wk paternity @ 100 + 18.0 wk medio-horario @ 50) / 34.86 = 74.18, rounded down (2026-08-08). Every component is genuinely UNCAPPED: subsidio = 100% of the last-6-months average + aguinaldo/salario-vacacional quota parts with only a FLOOR (the 15-BPC claim is the AFAP pension split - refuted on the 2026-08-06 cap sweep from primary text, not re-raised). Was 100 over the maternity-only 14 weeks. A8 note: the 6-month reference window is continuous prior income - no mover discount." }, "non_cash.parental_leave_weeks": { "source": "Ley 19.161 consolidated primary text (impo.com.uy/bases/leyes/19161-2013): art. 2 maternity 6 wk pre + 8 wk post; art. 8 as amended by Ley 20.312 (promulgated 2024-08-02; from 2026-01-01 dependent workers: 17 BPS subsidy days + 3 employer days of Ley 18.345 = 20 continuous days paternity); arts. 12-16 subsidio parental para cuidados (half-time, either parent alternating, until child age 6 months). BPS official pages 8958/8959. Research: docs/superpowers/research/2026-08-08-pl-b6-latam-romania.md", "last_updated": "2026-08-08", "confidence": "high", "note": "34.9 wk = 14.0 maternity + 2.86 paternity (20/7; the Ley 20.312 phase-in completed 2026-01-01, superseding the 13-day figure) + 18.0 wk of the half-time parental span (from the day after the maternity subsidy ends to child age 6 months: 26.07 - 8 postnatal wk = 18.07, stored 18.0 conservative; the span pays HALF the subsidy for a half schedule - see paid_pct). The half-time subsidy is usable by EITHER parent alternating = ONE shared pot, counted once (Sweden precedent) - both_parents NOT set. Was 14, maternity only. No tenure gate for dependent employees (status-based; only self-employed/monotributistas need contribution currency) - the mover passes everything. Conditional variants (multiple/premature 18-wk maternity, 30-day paternity, 9-month art. 2-BIS extension) excluded." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED for the exact unit, value retained. Partial evidence from the metropolitan mobility material suggests it may be high, but nothing states a one-way work-trip mean." }, "non_cash.sick_leave": { "source": "Ley 14.407 (1975), subsidio por enfermedad, administered by BPS (Banco de Previsión; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "low", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "DGI, IRPF Categoria II deduction list", "last_updated": "2026-08-11", "confidence": "medium", "note": "No voluntary-retirement deduction in the IRPF Categoria II deduction list, which is closed. The engine runs Uruguay on full income with standard_deduction 0, approximating the real credit-based system to about +1.8pp. HIGH on the mechanism (IRPF relief is credit-shaped and the deduction list is closed), MEDIUM on the completeness of the exclusion list." } }, "san-jose": { "tax.filing_basis": { "source": "PwC WWTS - Costa Rica: legislation does not contemplate joint filing; if both spouses earn income they submit individual returns (joint returns prohibited). A spouse/child credit is claimable by only one spouse.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Decreto Ejecutivo 45333-H (Gaceta 229, 2025-12-05), impuesto sobre la renta del trabajo 2026: 5 monthly tramos 0-25% (₡918k/₡1.347M/₡2.364M/₡4.727M per month) annualized ×12. First band is the exempt minimum; small fixed family credits (~₡20.5k/kid + ₡31k cónyuge per yr) omitted as immaterial vs an ICP tax bill", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_cap": { "source": "CCSS cuota obrera 2026 = 10.83% (SEM 5.5% + IVM 4.33% + Banco Popular 1.0%). Salaried CCSS is UNCAPPED (only a minimum contribution base exists, no maximum), so social_security_cap is null (BDO CR 2026 CCSS contributions; OECD, no IVM ceiling). The cuota is an independent withholding that does NOT reduce the income-tax base: the salary-tax tramos apply to salario bruto (reverse Batch B correction, verified tax audit 2026-07-14).", "last_updated": "2026-07-26", "confidence": "high" }, "housing._summary": { "source": "Numbeo San José 2026 (city-centre rent) nudged to the Escazú/Santa Ana professional tier; buy prices estimated from ₡/m²", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Private seguro on top of CCSS (Caja) for expat/professional tier; OOP for private consults/labs to skip Caja queues (Pacific Prime / ExpatDen 2025)", "last_updated": "2026-05-30", "confidence": "low" }, "childcare._summary": { "source": "Private guardería/preescolar San José 2025-26 (Numbeo + International Schools DB); IMAS/CEN-CINAI subsidies poverty-targeted → 0 for high earners", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo San José 2026 (Automercado/Walmart upper-middle basket) + Wise", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "San José autobús monthly + private car (high import-tax vehicle + fuel) 2025-26", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo San José leisure + domestic beach/eco vacation (Guanacaste/Manuel Antonio)", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "SJO->US/Europe economy long-haul round-trip 2025-26", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Costa Rica cash transfers (Avancemos, IMAS/Bono) are means-tested for poverty → 0 for a high earner (national, inherited)", "last_updated": "2026-05-30", "confidence": "high" }, "non_cash._summary": { "source": "Código de Trabajo (national, inherited): 2 weeks (10 working-day) annual leave after 50 weeks (Art. 153), 9 paid mandatory feriados, 4-month (17-week) maternity at 100% (employer 50% + CCSS 50%); mandatory aguinaldo (13th-month) noted but unmodeled", "last_updated": "2026-05-30", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, San José: CRC 27,175,759, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "CRC 27,175,759 / 454.48 = USD 59,795; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2392, R5-rounded down to 0.23. REPLACES 0.2 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Sample dominated by US shared-service centres; the top payer, Meta, pays 4x the median. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.20; established nearshore hub (Intel/Amazon/HP/Microsoft) puts it on par with Montevideo (0.22) and above Lima/Bogotá (0.18)." }, "non_cash.parental_leave_weeks": { "source": "Código de Trabajo art. 95 (licencia de maternidad: 1 month pre + 3 months post); Ley 10.211 (2022) art. 95 paternity: 2 days per week for the first 4 weeks = 8 days, 50% CCSS + 50% employer (three convergent 2026 guides; CCSS paid 28,455 such licenses in 2024)", "last_updated": "2026-08-07", "confidence": "medium-high", "note": "18.6 wk = 17 maternity (4 months) + 1.6 paternity (8 days ruled WORKING days /5: the entitlement is days off work taken 2 per week while employed; ruling recorded in the batch-5 cluster file, supersedes the earlier ~1.14 calendar sketch). Was 17, maternity only. Both slices pay 100% (50% CCSS + 50% employer). Mover gate passes (3 cuotas in 6 months). No-cap verdict re-settled 2026-08-07: MTSS official brochure (full-salary framing), the Nov-2024 Reglamento reform (base = last 12 reported salaries, no ceiling) and 2026 guides are all cap-free; the stray 'CRC 1,659,000 tope' claim is untraceable and REFUTED. Caveat: Reglamento arts. 40-54 full text unfetchable today (pgrweb/sinalevi TLS error, CCSS PDFs failed, MTSS criterio 403), hence medium-high. Research: docs/superpowers/research/2026-08-07-pl-b5-latam.md." }, "non_cash.parental_leave_paid_pct": { "source": "CCSS Reglamento para el Otorgamiento de Licencias e Incapacidades: the trabajadora tiene derecho a devengar el mismo salario que recibiria si estuviera laborando, CCSS 50% + employer 50% (obtained via a cuc.ac.cr PDF mirror; ccss.sa.cr itself ECONNREFUSED on every attempt). Research: docs/superpowers/research/2026-08-06-lt-cr-verification.md and 2026-08-06-parental-caps-americas-africa.md (Costa Rica)", "last_updated": "2026-08-06", "confidence": "high", "note": "100% confirmed uncapped for asalariados on the 2026-08-06 re-verification, upgrading the earlier MEDIUM call: the CCSS regulation text guarantees the same salary as if working, and every maximo in the document is a duration limit, not an amount ceiling; the tiered escala with ceiling-shaped figures belongs to the independientes (self-employed) regime only. Caveat: the regulation PDF is the 2014 edition, current-ness corroborated by 2026 secondaries. The Ley 10.211 paternity days flagged here were stored in batch 5 (2026-08-07): 8 working days = 1.6 wk at the same 100% split, see the weeks entry." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. This slug is San Jose, COSTA RICA. INEC Costa Rica and the Gran Area Metropolitana transport studies carry no one-way work-trip mean." }, "non_cash.sick_leave": { "source": "Ley Constitutiva de la CCSS; Reglamento del Seguro de Salud de la CCSS (governs the; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "low", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Ley 7983 art. 71 (Ley de Proteccion al Trabajador); Hacienda's reglamento would not load", "last_updated": "2026-08-11", "confidence": "medium", "note": "Ley 7983 art. 71 exempts \"los aportes que realicen los patronos Y los trabajadores\" to complementary pension schemes under ONE 10% ceiling, so it is a percentage with no monetary cap. UNRESOLVED and flagged rather than invented: a stricter reading would carve the employer legs (FCL 3%, ROP obligatorio) out via less_employer_pct, but which employer legs sit inside the 10% could not be sourced. The engine does not deduct CCSS from the income-tax base (deliberate, see the branch comment), so this is additive. Binds exactly at a 10% contribution and not at 6%." } }, "manila": { "tax.filing_basis": { "source": "PwC WWTS - Philippines (Tax administration): a husband and wife file one consolidated return, but the tax is computed separately per spouse.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "TRAIN Law (RA 10963) graduated income-tax schedule effective 2023-01-01: 0% ≤₱250k, 15%/20%/25%/30% bands, 35% >₱8M (BIR). Single == married (Philippines has no joint filing); the ₱250k first band absorbs the old personal/additional exemptions", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_cap": { "source": "Combined mandatory employee contributions modeled as one ~9.5% rate capped ~₱562k/yr: SSS 5% on MSC (cap ₱35k/mo, 2025 schedule) + PhilHealth 2.5% employee share (5% premium split, ceiling ₱100k/mo) + Pag-IBIG/HDMF 2% (cap ₱10k/mo). These are EXCLUDED from taxable income → modeled Argentina-shape (federal base = gross − contributions)", "last_updated": "2026-05-30", "confidence": "medium" }, "housing._summary": { "source": "Numbeo Manila/Metro Manila 2026 (BGC/Makati professional tier) for rent; condo buy prices from ₱/sqm BGC-Makati", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "DotProperty listing-count CDF, 2026-08-18 (condos for rent, Metro Manila NCR, bedroom filter is a path segment - the ?bedrooms= GET param is a silent no-op, caught; max_price binds, inclusive, consistency-proven by the min/max complement overlap at the 100k round point; studios excluded from 1BR by construction; price-on-request listings excluded from N; asking PHP/month): 2BR P25/med/P75 = 28,389/71,796/119,689 (n=9,725) -> 0.395/1.667; 1BR 0.623/1.603 (n=10,047).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW by ~7x tolerance on BOTH 2BR edges - the pool is explicitly bimodal: a mass-market economic-condo segment (Urban Deca-class, 2BR 15-25k, a 1,218-listing jump in the single 20k->25k band) against a CBD/expat segment (BGC/Makati/Rockwell, 22% of the pool in the 100k-200k band alone). Applied 0.40/1.67 (nearest; BOTH per-city guard exceptions). Duplication fattens the top but cannot explain the lower-edge miss (-0.425), which breaks the verdict alone. Condo-only pool: the apartment/house verticals and informal market (which skew cheap) are excluded - the true citywide spread is wider, the fix conservative. Lamudi PH bot-blocked (recorded), single portal. 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-sea-1.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (J). NOTE stored 1BR (45,000) sits ~34% above the measured condo-pool median (33,551) and stored 2BR ~11% above (71,796) - a LEVEL observation, flagged, out of scope here." }, "healthcare._summary": { "source": "Private HMO (Maxicare/Intellicare) on top of PhilHealth for the professional tier; OOP for private consults/labs (Pacific Prime / ExpatDen 2025)", "last_updated": "2026-05-30", "confidence": "low" }, "childcare._summary": { "source": "Private daycare/preschool Metro Manila 2025-26 (Numbeo + international preschool DB); public/ECCD subsidies poverty-targeted → 0 for high earners", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Manila 2026 (S&R/Landmark upper-middle basket) + Wise", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "Manila MRT/LRT/bus + private car (fuel + parking, heavy congestion) 2025-26", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Manila leisure + domestic island vacation (Palawan/Boracay/Cebu)", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "MNL->US/Europe economy long-haul round-trip 2025-26", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Philippines has no universal child cash benefit; 4Ps conditional cash transfer is poverty-targeted → 0 for a high earner (national, inherited)", "last_updated": "2026-05-30", "confidence": "high" }, "non_cash._summary": { "source": "Labor Code (national, inherited): 5-day Service Incentive Leave (Art. 95) as the statutory minimum, 12 regular holidays (Proclamation), 105-day (15-week) maternity at 100% under RA 11210 (Expanded Maternity Leave). No separate statutory paid sick leave; SSS sickness benefit covers prolonged certified illness", "last_updated": "2026-05-30", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Manila: PHP 905,396, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "PHP 905,396 / 61.24 = USD 14,784; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.0591, R5-rounded down to 0.05. REPLACES 0.18 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Corroborated by PSA OWS to within 1.4%, and already includes the 13th month. One of only two markets in the APAC batch where levels.fyi did NOT run above the domestic median. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.18; large IT-BPM/offshore-dev hub but comp sits below Latin nearshore peers, on par with Lima/Bogotá (0.18) and below Montevideo (0.22)." }, "non_cash.parental_leave_weeks": { "source": "RA 11210 (105-day maternity) + RA 8187 (7-day paternity, first 4 deliveries). Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (Philippines) and docs/superpowers/research/2026-08-06-pl-caps-southeast-asia.md", "last_updated": "2026-08-08", "confidence": "high", "note": "16 wk = 105 days maternity + 7 days paternity (112/7). Was 15, maternity only. The mother may reallocate up to 7 of her 105 days to the father; a reallocation inside the same pot, not additional days, so not double-counted. Re-verified on the household basis 2026-08-08 (RA 11210 + RA 8187 primary via lawphil): the 7-day maternal allocation is black-letter 'of said benefits' - a transfer inside the 105, counted once. RA 8187's 'seven (7) days' ruled CALENDAR: the statute and the official DOLE 2023 Handbook say plain 'days'; only unverifiable IRR reproductions say 'working days' (if ever primary-verified that reading would give 16.4). The June-2026 'PAPA Leave Act' (105-day paternity) is a FILED BILL, not law. SSS gate (3 contributions in the 12 months before the semester of birth): work-to-term mover passes with ~6 contributions; conceive-immediately narrow-pass caveat only. Research: docs/superpowers/research/2026-08-08-parental-leave-batch6.md" }, "non_cash.parental_leave_paid_pct": { "source": "RA 11210 s.5: full pay = SSS benefit + MANDATORY employer salary differential (\"any difference... shall be borne by the employer\"), PwC PH + Respicio & Co. concur; SSS benefit-computation MSC ceiling PHP 20,000/month (sss.gov.ph, fetched 2026-08-06) caps only the SSS-reimbursed portion; RA 8187 paternity is direct employer pay", "last_updated": "2026-08-08", "confidence": "high", "note": "100% confirmed honest for the HOUSEHOLD on the 2026-08-06 cap sweep, REFUTING the working hypothesis that the SSS wage ceiling caps what the family receives: the salary differential is a statutory employer obligation, not optional top-up, so a standard-size Manila employer owes the gap above the SSS PHP 666.67/day maximum. The narrow DOLE exemptions (10-or-fewer workers, micro-business, distressed) cannot apply to our ICP's employer. The PHP 20,000 MSC is distinct from the PHP 35,000 total contribution ceiling; do not conflate. Re-verified 2026-08-08: RA 11210 s.5(c)/s.6 mandatory salary differential confirmed primary; the DOLE exemption categories cannot reach an ICP-sized employer." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. The JICA MUCEP Person Trip Survey (2014) is on our unit (one way, origin to destination, walk time at both ends included) but states 70 min, all-purpose, over the larger Mega Manila area. Work trips run longer than all-purpose (pushing up) while Metro Manila proper is smaller than the MUCEP area (pushing down), and there is no principled way to net those, so 60 is not attributed to MUCEP." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-F.md", "source_url": "https://www.respicio.ph/bar/2025/labor-law-and-social-legislation/labor-standards/leaves/labor-code/service-incentive-leave", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Bangko Sentral ng Pilipinas, PERA FAQs on the tax credit https://www.bsp.gov.ph/Pages/InclusiveFinance/PERA_FAQs_TaxCredit.aspx (RA 9505)", "last_updated": "2026-08-11", "confidence": "high", "note": "No base deduction. PERA caps contributions at PHP 100,000/yr (200,000 for OFWs) and pays a 5% INCOME TAX CREDIT against tax due, expressly not a deduction from the base. The engine already subtracts mandatory payroll from the base, which is separate." } }, "jakarta": { "tax.federal_brackets": { "source": "UU No. 7 Tahun 2021 (UU HPP) art. III no. 7, replacing UU PPh Pasal 17(1)(a); full text on DJP, status Aktif https://www.pajak.go.id/id/peraturan/harmonisasi-peraturan-perpajakan", "last_updated": "2026-08-01", "confidence": "high", "married_variant": "none", "note": "VERIFIED 2026-08-01 against the statute: 5% to Rp60m, 15% to Rp250m, 25% to Rp500m, 30% to Rp5bn, 35% above. Indonesia has ONE Article 17 rate table with no marital qualifier, because the family is the tax unit. The Penjelasan to Pasal 8 puts the whole family \"sebagai satu kesatuan ekonomis\", income combined, one NPWP (PER-7/PJ/2025 Pasal 4). Even the PH and MT separate-assessment statuses combine net income, compute once on this same table, then apportion pro rata (Pasal 8(3)). The whole married/single difference lives in PTKP, which we carry in standard_deduction_*." }, "tax.social_security_cap": { "source": "BPJS modeled bespoke as three employee contributions: JHT 2% (uncapped, deductible), JP 1% (cap Rp133,035,600/yr 2025 wage ceiling, deductible), BPJS Kesehatan 1% (cap Rp144M/yr salary ceiling, NOT deductible). biaya jabatan occupational deduction 5% capped Rp6M/yr also reduces the income-tax base", "last_updated": "2026-05-30", "confidence": "medium" }, "housing._summary": { "source": "Numbeo Jakarta 2026 (South Jakarta/SCBD-Kebayoran professional tier) for rent; apartment buy prices from Rp/sqm central Jakarta", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Pinhome weighted stratified harvest, 2026-08-18 (apartments for rent, DKI Jakarta; the RENT-PERIOD TRAP is real and was solved by partition: the market splits into a monthly-term pool (8,627 listings, 100% monthly-priced) and a yearly-term pool (8,007, ~77% yearly-priced) via binding min/maxRentTermMonth params; every price normalized to monthly IDR via per-listing JSON-LD period flags (P1Y/P1M), never the raw mixed field; window-count weighted quantiles): MONTHLY-TERM pool 2BR -> 0.529/2.059 (n_harv=138 repr. ~4,623); yearly-term 2BR 0.608/1.910 (n_harv=150 repr. ~4,077); combined 0.643/2.571. Applied pair = the monthly-term pool, the stored medians' own universe.", "last_updated": "2026-08-18", "confidence": "medium-low", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW in EVERY defensible pool definition (2BR edges out by 3-22x tolerance; the only narrow slice is mass-market studios at 0.86/1.14). Applied 0.53/2.06 (nearest; per-city p75 guard exception 2.1, the widest honest pair in the repo). POOL ADJUDICATION (bangkok like-for-like precedent): the stored medians track the MONTHLY-term universe (stored 1BR 7jt = monthly-pool 7.0jt exact; 0BR 4.5 vs 3.5; 2BR 13 vs 8.5 above; all far above yearly equivalents 4.2/4.8), so the monthly-term pair is the like-for-like choice; the combined-pool 2.571 conflates the term-market gap into the ratio and was NOT applied. Jakarta is dual on TWO axes (term markets of equal size + mass-vs-expat within each), recorded. Caveats: within-window sampling (n_harv 138; window weights exact, within-window shape sampled - no plausible sampling error changes the verdict at 3-22x); single portal (Rumah123/99.co/Lamudi blocked, recorded); kos/kontrakan informal stock absent - true spread wider, fix conservative. 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw windows: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-sea-1.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (J). NOTE the stored medians' monthly/serviced basis is NOT recorded in _meta and the yearly-contract local market sits at roughly HALF those levels - the loudest LEVEL flag of this run, out of scope here." }, "healthcare._summary": { "source": "Private insurance on top of BPJS Kesehatan for the professional tier; OOP for private hospital/specialist (Pacific Prime / ExpatDen 2025)", "last_updated": "2026-05-30", "confidence": "low" }, "childcare._summary": { "source": "Private daycare/preschool (TK/PAUD) Jakarta 2025-26 (Numbeo + international preschool DB); public subsidies poverty-targeted → 0 for high earners", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Jakarta 2026 (Ranch Market/Grand Lucky upper-middle basket) + Wise", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "Jakarta MRT/TransJakarta/KRL + private car (fuel + parking, heavy congestion) 2025-26", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Jakarta leisure + domestic vacation (Bali/Yogyakarta/Lombok)", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "CGK->US/Europe economy long-haul round-trip 2025-26", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Indonesia has no universal child cash benefit; PKH conditional cash transfer is poverty-targeted → 0 for a high earner (national, inherited)", "last_updated": "2026-05-30", "confidence": "high" }, "non_cash._summary": { "source": "UU Ketenagakerjaan / UU Cipta Kerja (national, inherited): 12-day statutory annual leave minimum, ~17 national holidays (incl. joint-leave cuti bersama), 13-week (3-month) maternity at 100% wage. Prolonged certified sick leave paid on a 100/75/50/25% sliding scale", "last_updated": "2026-05-30", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Jakarta: IDR 247,281,434, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "IDR 247,281,434 / 18075.57 = USD 13,680; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.0547, R5-rounded down to 0.05. REPLACES 0.12 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Annual and ALREADY includes THR. Do not add a 13th month. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.12; large dev market (Gojek/Tokopedia/Traveloka) but local-currency comp sits below Manila (0.18); nudged above the raw Levels ratio (~0.09) for the multinational ICP segment." }, "tax.federal_brackets_married": { "source": "UU No. 7 Tahun 2021 (UU HPP) Pasal 17(1)(a) https://www.pajak.go.id/id/peraturan/harmonisasi-peraturan-perpajakan", "last_updated": "2026-08-01", "confidence": "high", "married_variant": "none", "note": "Byte-identical to tax.federal_brackets ON PURPOSE, verified 2026-08-01: no married rate schedule exists in Indonesian law. Do not \"fix\" this by inventing widened bands. The married relief is the PTKP allowance in standard_deduction_married, subtracted before the table. Worth raising separately: standard_deduction_married is Rp67.5m, which is PTKP K/2 (married plus two dependants) applied to every couple, where K/0 is Rp58.5m." }, "tax.jp_rate": { "source": "BPJS Ketenagakerjaan program rules via acclime.com Indonesia guide (indonesia.acclime.com/guides/mandatory-healthcare-social-security): foreign workers (TKA, 6+ months) enrol in BPJS Ketenagakerjaan and pay JHT, but \"Pension benefit (JP)... applies to Indonesian citizens only\". docs/superpowers/research/2026-08-06-forced-savings-foreigner-rules.md", "last_updated": "2026-08-06", "confidence": "medium-high", "note": "JP (Jaminan Pensiun) 1% employee, capped at the JP wage ceiling, covers Indonesian citizens ONLY; a foreign worker pays JHT (2%) and BPJS Kesehatan (1%) but never JP. Engine seam: computeIndonesiaTaxes reads state.id_bpjs_resident, absent or true = citizen (JHT+JP+health), false = foreign worker (JHT+health, and JP leaves the PPh deductible base too, since a contribution never paid cannot be deducted). The target side of a comparison defaults to foreign worker, the source side to citizen; rankings and single-city callers carry no flag and stay on the citizen basis. Charging JP to the modeled mover overstated payroll by up to ~1pp of gross below the JP cap (~IDR 133M), ~0.2pp at a 600M tech salary where the cap binds. Confidence medium-high: one strong compliance-advisory source with an exact quote; bpjsketenagakerjaan.go.id itself not reached." }, "non_cash.parental_leave_weeks": { "source": "UU Ketenagakerjaan art. 84 (3 months maternity at full wage) + art. 93(2)(c) (2 days paternity); UU KIA 4/2024 confirms the same baselines. Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (Indonesia) and docs/superpowers/research/2026-08-06-pl-caps-southeast-asia.md", "last_updated": "2026-08-08", "confidence": "high", "note": "13.29 wk = 13 maternity + 0.29 paternity (2 days / 7). Was 13, maternity only. EXCLUDED per the basis convention: the UU KIA extension to 6 months (medical-certificate-gated, not a healthy-birth entitlement) and the extra 3 discretionary paternity days (employer agreement, not statute). Re-verified on the household basis 2026-08-08 from the Kemnaker JDIH primary text: art. 93(4)(e) pays '2 (dua) hari' - plain 'hari', and the statute writes 'hari kerja' where it means working days (art. 79 annual leave), so 2 CALENDAR days = 0.29 wk is the black-letter reading (not 2/5). UU KIA 4/2024 re-confirmed: base unchanged; months 4-6 are medical-certificate-gated (excluded); the +3 paternity days are 'dapat diberikan' discretionary (excluded); the union '5 hari' gloss refuted for our basis. No tenure gate (day-1 employer obligation, criminally enforced). Research: docs/superpowers/research/2026-08-08-parental-leave-batch6.md" }, "non_cash.parental_leave_paid_pct": { "source": "UU Ketenagakerjaan arts. 84 and 93: upah penuh (full wages), direct employer obligation, no BPJS cash transfer and no insured-wage ceiling (hukumonline.com legal clinic explicit: the law sets a minimum, not a ceiling)", "last_updated": "2026-08-08", "confidence": "high", "note": "100% confirmed genuinely UNCAPPED on the 2026-08-06 cap sweep: employer wage continuation with no ceiling found across ~15 sources. The task's staleness worry was unfounded; the UU KIA 2024 change is conditional, not the base entitlement. Re-verified 2026-08-08: upah penuh direct employer liability confirmed primary; no BPJS cash maternity benefit through 2026." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. BPS does publish Statistik Komuter Jabodetabek 2023, which is on our unit (one way, home to main activity), and its bands imply a mean of 49 to 53. It was NOT adopted because BPS counts only cross-kabupaten/kota commuters (4.4M of 29.6M aged 5+), excluding the intra-district working majority, so the series is biased long; corrected to all employed commuters it lands near 38 to 42. The stored 45 sits between the two and matches neither, so it is recorded as an honest gap rather than given a citation that does not state it." }, "non_cash.sick_leave": { "source": "Undang-Undang No. 13 Tahun 2003 tentang Ketenagakerjaan (Manpower Law), Pasal 93; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-F.md", "source_url": "https://www.hukumonline.com/klinik/...", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." } }, "kuala-lumpur": { "tax.federal_brackets": { "source": "LHDN (Inland Revenue Board) Malaysia YA2024/2025 resident PIT schedule: 0% ≤RM5k, then 1/3/6/11/19/25/26/28/30% bands to >RM2M (LHDN tax-rate table). Single == married rate ladder (Malaysia has no joint-filing ladder change); personal reliefs modeled via standard_deduction (RM9k individual + RM4k EPF/life relief = RM13k single; +RM4k spouse relief = RM17k married)", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_cap": { "source": "EPF (KWSP) modeled in payroll as 11% employee statutory rate, uncapped (no wage ceiling on the employee share). SOCSO+EIS modeled together at ~0.7% employee on wages capped RM72k/yr (PERKESO 2025 ceiling, raised from RM5k/mo). EPF relief already folded into standard_deduction so it is not re-deducted from the income-tax base", "last_updated": "2026-05-30", "confidence": "high" }, "housing._summary": { "source": "Numbeo Kuala Lumpur 2026 (KLCC/Mont Kiara/Bangsar professional-tier condo) for rent; condo buy prices from RM/sqft central KL", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "mudah.my listing-count CDF, 2026-08-18 (apartments/condos for rent, WP Kuala Lumpur - Selangor excluded; the price key is literally monthly_rent, MYR/month; filters verified binding; pool includes cheap flat stock, not only premium condos): 2BR P25/med/P75 -> 0.768/1.319 (n=1,685); 1BR 0.799/1.298; 3BR 0.793/1.293 (n=4,784).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24. SPLIT verdict adjudicated to a fix: the primary 2BR upper edge is out beyond tolerance (+0.079) while 1BR and 3BR hold, and ALL SIX measured edges lean wider than the default; the breach is ROBUST to the portal's known bias (informal/room stock absent means the true spread is wider still), the same adjudication class as Guadalajara this cluster. Applied 0.77/1.32 (nearest). Single portal (iProperty/PropertyGuru bot-blocked, recorded). 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-sea-1.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (J). NOTE the stored medians sit 53-164% above the mudah pool (which includes cheap flats) - a pool-basis LEVEL question, flagged, out of scope here." }, "healthcare._summary": { "source": "Private insurance on top of the public (MOH) system for the professional tier; OOP for private hospital/specialist (Pacific Prime / iMoney Malaysia 2025)", "last_updated": "2026-05-30", "confidence": "low" }, "childcare._summary": { "source": "Private taska/tadika (daycare/kindergarten) KL 2025-26 (Numbeo + international preschool DB); public subsidies income-targeted → 0 for high earners", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Kuala Lumpur 2026 (Village Grocer/Jaya Grocer upper-middle basket) + Wise", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "KL MRT/LRT/Monorail + private car (fuel + Touch'n Go tolls + parking) 2025-26", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Kuala Lumpur leisure + domestic vacation (Penang/Langkawi/Borneo)", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "KUL->US/Europe economy long-haul round-trip 2025-26", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Malaysia has no universal child cash benefit; STR/Sumbangan Tunai Rahmah is income-targeted → 0 for a high earner (national, inherited)", "last_updated": "2026-05-30", "confidence": "high" }, "non_cash._summary": { "source": "Employment Act 1955 (national, inherited): paid annual leave 8/12/16 days by tenure (16 at >5yr), ~15 federal+state public holidays, 14-week maternity at 100% (Employment Act 2022 amendment). Paid sick leave 14-22 days/yr by tenure + 60 days hospitalisation", "last_updated": "2026-05-30", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Kuala Lumpur: MYR 85,870, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "MYR 85,870 / 4.09 = USD 20,995; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.0840, R5-rounded down to 0.08. REPLACES 0.18 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Corroborated: the RM 6,500/month domestic median times 13 is about RM 84,500. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.18; KL dev market (Grab/Shopee/regional MNC hubs) puts local-currency comp level with Manila and above Jakarta (0.12); raw Levels ratio ~0.17 nudged to 0.18 for the multinational ICP segment." }, "tax.filing_basis": { "source": "LHDN (hasil.gov.my) - separate assessment; individual filing", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "tax.epf_foreigner_rate": { "source": "KWSP, Contribution For Non-Malaysian Citizen Employees (kwsp.gov.my/en/employer/responsibilities/non-malaysian-citizen-employees; the page 403s automated fetches, content cross-confirmed verbatim via the KWSP news page, blog.kakitangan.com, payrollpanda.my and briohr.com, all agreeing exactly). docs/superpowers/research/2026-08-06-forced-savings-foreigner-rules.md", "last_updated": "2026-08-06", "confidence": "high", "note": "Flat 2% employee EPF for non-Malaysian citizen employees, MANDATORY from 1 October 2025 wages (contribution month November 2025); before that, foreigner EPF was opt-in and typically 0%. Citizens/PRs and pre-1 Aug 1998 EPF members stay at the 11% in tax.social_security_rate; a foreigner already contributing voluntarily at 11% may keep that rate on request, so 2% is the mandatory default, not a universal truth. Engine seam: computeMalaysiaTaxes reads state.my_epf_resident, absent or true = citizen 11%, false = this rate. The target side of a comparison defaults to foreigner, the source side to citizen; rankings and single-city callers carry no flag and stay on the citizen basis. A RATE swap, not an exemption: either contribution lands in the member's own EPF account, so the forced-savings display follows whichever rate applies." }, "non_cash.parental_leave_weeks": { "source": "Employment Act 1955 s.37 as amended by the Employment (Amendment) Act 2022, effective 1 Jan 2023: maternity 98 days; paternity 7 consecutive days (jtksm.mohr.gov.my PDF corroborates). Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (Malaysia) and docs/superpowers/research/2026-08-06-pl-caps-southeast-asia.md", "last_updated": "2026-08-08", "confidence": "high", "note": "15 wk = 98 days maternity + 7 days paternity (105/7). Was 14, maternity only. REFUTATION recorded: the \"128 days\" figure circulating in 2026 is the proposed EPCB post-maternity allowance, still under consideration, not law; 98 days stands. Re-verified on the household basis 2026-08-08 from the official MOHR consolidated Act 265: s.37 98 consecutive days at ordinary rate of pay; s.60FA 7 consecutive (calendar) days paternity, married to the mother, first 5 confinements. First Schedule para 1A verified: above RM4,000/month only overtime/premium provisions drop out - Part IX maternity and s.60FA apply at EVERY wage level (genuinely uncapped across the ICP band). LOUD MOVER CAVEAT: s.60FA requires 12 months with the SAME employer at leave commencement - a conceive-immediately mover loses the 1.0 paternity week; work-to-term passes (caveat-only per the Israel/Lithuania precedent, no haircut). Maternity allowance gate (90 days' employment in the 9 months + employed within 4 months of confinement) passes trivially. Research: docs/superpowers/research/2026-08-08-parental-leave-batch6.md" }, "non_cash.parental_leave_paid_pct": { "source": "Employment Act 1955: ordinary rate of pay (100%), employer-paid; SOCSO/PERKESO does not pay employed women's maternity (its maternity-adjacent benefit sits in the self-employment scheme SKSPS only)", "last_updated": "2026-08-08", "confidence": "high", "note": "100% confirmed genuinely UNCAPPED on the 2026-08-06 cap sweep: direct employer wage continuation, no wage ceiling in any source checked, and no SOCSO involvement for a standard salaried KL employee. Re-verified 2026-08-08 (MOHR consolidated Act): ordinary rate of pay, employer-paid, no wage ceiling at any level (First Schedule para 1A scope finding)." }, "non_cash.vacation_days_statutory": { "source": "Employment Act 1955 (Act 265) s.60E(1)(a), Ministry of Human Resources consolidated text as at 2023-01-01 https://jtksm.mohr.gov.my/sites/default/files/2023-11/Akta%20Kerja%201955%20(Akta%20265).pdf", "last_updated": "2026-08-08", "confidence": "high", "note": "CORRECTED 2026-08-08, 16 to 8. s.60E(1) is a three-rung tenure ladder: 8 days under two years of service, 12 days from two to five years, 16 days at five years or more. We had stored the TOP rung. The floor is 8. No conversion applies, the statute states flat day counts. s.60E(1A) makes the leave additive to rest days and paid holidays. Coverage checked: the First Schedule exempts only domestic employees from Part XII, and the RM4,000 salary carve-out lists other subsections but not s.60E, so a white-collar KL employee is inside the floor." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. DOSM has no locatable series. One academic paper reported second-hand at about 41.6 min is a lead rather than a source; if it verifies, the stored value is about 7 low." }, "non_cash.sick_leave": { "source": "Employment Act 1955 (Act 265), s.60F \"Sick leave\"; First Schedule (s.2(1)) for; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-F.md", "source_url": "https://lom.agc.gov.my/ilims/upload/portal/akta/outputaktap/1741197_BI/Act%20265_FINAL_as%20at%201%20Jan%202023%20(30.3.23", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Private Pension Administrator Malaysia (PRS central administrator appointed by the Securities Commission), PRS tax relief https://www.ppa.my/joining-prs/prs-tax-relief/", "last_updated": "2026-08-11", "confidence": "medium", "note": "PRS / deferred annuity relief, RM3,000 (YA2012-YA2030), the only new money available. The headline RM7,000 splits into an RM4,000 EPF bucket and this RM3,000 bucket; voluntary EPF self-contribution sits INSIDE the RM4,000, which mandatory 11% EPF already exhausts above about RM36.4k gross, and that RM4,000 is already baked into the engine's standard_deduction, so counting it again would double-count. SOURCING GAP, stated plainly: no LHDN-hosted page was readable (the relief table 404s on both paths and Public Ruling 9/2021 refused the connection)." } }, "ho-chi-minh-city": { "tax.filing_basis": { "source": "PwC WWTS - Vietnam; Vietnam.vn: no joint declaration mechanism; each spouse is assessed separately (joint filing is only a legislative proposal).", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Vietnam PIT Law 109/2025/QH15 (new 5-bracket resident schedule effective tax-year 2026): 5/10/20/30/35% on monthly taxable bands ≤10M/30M/60M/100M/>100M, annualized ×12. Single == married rate ladder (Vietnam taxes individuals; no joint filing). Personal/dependent deductions (Resolution 110/2025: VND 15.5M/mo self = 186M/yr; VND 6.2M/mo per dependent) modeled via standard_deduction (186M single; +1 spouse-dependent 74.4M = 260.4M married)", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_cap": { "source": "Mandatory employee contributions modeled bespoke (two caps, both deductible from PIT base): Social+Health Insurance 9.5% capped at 20× the reference salary (lương cơ sở VND 2.34M/mo from Jul-2025 → cap VND 561.6M/yr); Unemployment Insurance 1% capped at 20× the Region I minimum wage (VND 5.31M/mo from Jan-2026 per Decree 293/2025 → cap VND 1,274.4M/yr). HCMC = Region I (EY/PwC Vietnam 2025-26)", "last_updated": "2026-05-30", "confidence": "high" }, "housing._summary": { "source": "Numbeo Ho Chi Minh City 2026 (District 2 Thảo Điền / District 7 / Bình Thạnh localized-professional tier) for rent; condo buy prices from VND/sqm central HCMC", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Cho Tot gateway API full price harvest, 2026-08-18 (gateway.chotot.com/v1/public/ad-listing, region 13000 = pre-merger HCMC proper, cg=1010 can ho chung cu; price proven VND/month against 'N trieu/thang' strings; rooms = phong ngu verified; EXACT quantiles of the harvested price vector, no interpolation): 2BR pooled P25/med/P75 = 8.0M/10.5M/14.0M VND (n=4,287) -> 0.762/1.333, bootstrap CI on the upper edge 1.333-1.400 entirely outside tolerance; chung-cu-only slice 0.739/1.304 (n=3,298), 3BR upper 1.467.", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW (2BR upper edge +0.093 with the bootstrap CI entirely out; the like-for-like chung-cu slice is out on BOTH edges). Applied 0.76/1.33 (nearest, pooled slice - the citywide apartment pool is the band's claim). Pool caveats: Cho Tot is a VND-only local portal with NO premium/USD expat stock (Batdongsan 403s both fetchers, recorded), and the phong tro segment (>=10k ads) sits outside the apartment pool - BOTH truncations mean the true citywide spread is wider than measured, so the fix is conservative. Composition finding recorded: the mini/serviced segment is tight (~0.82/1.18) while proper chung cu is wide - composition, not noise, drives the spread. 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs + CIs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-vietnam.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (D1). NOTE the stored 2BR median (24.0M) sits far above the budget-skewed Cho Tot pool median (10.5M pooled / 11.5M chung cu) - different universes; a LEVEL question, flagged loudly, out of scope here." }, "healthcare._summary": { "source": "Private domestic insurance (Bảo Việt/PVI tier) on top of public BHYT for the professional tier; OOP for private clinic/specialist (Pacific Prime Vietnam 2025-26): NOT the international expat plan", "last_updated": "2026-05-30", "confidence": "low" }, "childcare._summary": { "source": "Private taska/mầm non (daycare/kindergarten) HCMC 2025-26 (Numbeo + bilingual preschool DB, mid-tier not international); public preschool tuition exemption does not reach private-school families", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Ho Chi Minh City 2026 (upper-middle local basket) + Wise", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "HCMC Metro Line 1 + bus pass; private car (high import/registration tax) 2025-26: note most locals ride motorbikes, so the car line overstates typical professional transport", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Ho Chi Minh City leisure + domestic vacation (Đà Nẵng/Phú Quốc/Hà Nội)", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "SGN->US/Europe economy long-haul round-trip 2025-26", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Vietnam has no universal child cash benefit; a one-off SI maternity lump sum exists but no ongoing child allowance → 0 for a high earner (national, inherited)", "last_updated": "2026-05-30", "confidence": "high" }, "non_cash._summary": { "source": "Labour Code 2019 + Social Insurance Law (national, inherited): 12-day statutory minimum paid annual leave (+1/5yr tenure), 12 paid public holidays (2026), 26-week (6-month) maternity at 100% funded by social insurance. Statutory sick pay via SI up to 30 days/yr at 75%", "last_updated": "2026-05-30", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Ho Chi Minh City: VND 332,647,673, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "VND 332,647,673 / 26179.98 = USD 12,706; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.0508, R5-rounded down to 0.05. REPLACES 0.14 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. All top employers sit BELOW this median, and senior/all-levels is 2.62 with the senior median landing exactly on the all-levels p90, which cannot happen legitimately. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.14; HCMC MNC/regional-hub tech comp (~VND 600-800M total for the ICP) sits just above Jakarta (0.12), below Manila/KL (0.18); raw Levels ratio ~0.13." }, "non_cash.parental_leave_weeks": { "source": "Law on Social Insurance 41/2024 (in force 2025-07-01) art. 53: paternity \"05 ngày làm việc\" (working days), full statute text via the government portal, fetched 2026-08-07; 6 months maternity. Research: docs/superpowers/research/2026-08-07-pl-b5-africa-gulf-sea.md, 2026-08-06-parental-caps-asia-mideast.md (Vietnam), 2026-08-06-pl-caps-southeast-asia.md", "last_updated": "2026-08-07", "confidence": "high", "note": "27.0 wk = 26.0 maternity (mother, 6 months) + 1.0 paternity (5 WORKING days / 5 per the house convention; the prior 0.71 used the calendar /7 conversion, corrected in batch 5 on the statute's own 'ngày làm việc' wording; C-section/twin uplifts excluded per the basis convention). The 6-in-12-months contribution gate passes for an employed-on-arrival mover by the birth date. The one-off 2x-reference-level birth grant (art. 58(4), VND 5.06M) is a lump allowance, not wage replacement, and stays excluded (Japan precedent)." }, "non_cash.parental_leave_paid_pct": { "source": "Law on Social Insurance: benefit = 100% of the average SI-contribution salary of the preceding 6 months", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 26.71 weeks into non_cash.parental_leave_capped (100% of an SI wage capped at 20x base salary), so no percentage-paid weeks remain on this field." }, "non_cash.parental_leave_capped": { "source": "Decree 161/2026/ND-CP: reference level VND 2,530,000/month from 1 July 2026; SI contribution-and-benefit wage ceiling = 20x = VND 50,600,000/month (multi-source corroborated: vinatpt.com, vietnam.vn, laodong.vn; re-verified 2026-08-07). Research: docs/superpowers/research/2026-08-07-pl-b5-africa-gulf-sea.md, 2026-08-06-parental-caps-asia-mideast.md (Vietnam)", "last_updated": "2026-08-07", "confidence": "high", "note": "All 27.0 weeks pay 100% of an SI-contribution wage structurally capped at VND 50,600,000/month (the ceiling flows through because premiums cannot be levied above it; 20x the Decree 161/2026 reference level VND 2,530,000, re-verified 2026-08-07). Binds only above ~VND 607M/yr, the upper half of the Hanoi/HCMC tech band: at VND 900M/yr the effective rate is ~67.5%, below it the 100% is genuine. cap_amount doubles as the payout ceiling because the rate is 100. The ceiling moves with the reference level (last step 1 Jul 2026). Weeks 26.71 -> 27.0 in batch 5 (paternity working-day conversion)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "BHYT cost-sharing structure, including a 0% co-payment for children under age 6", "last_updated": "2026-08-08", "confidence": "medium", "note": "Child-discount shape applied on a directly-sourced government child exemption. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "BHYT cost-sharing structure, including a 0% co-payment for children under age 6", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Media figures are consistent with 32 but none names a survey, so none is cited." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Decree 253/2026/ND-CP and Circular 87/2026/TT-BTC (issued 30 June 2026, effective 1 July 2026), via Baker McKenzie https://www.bakermckenzie.com/en/insight/publications/2026/07/vietnam-decree-circular-implementing-personal-income-tax-law and corroborated by KPMG, Vialto, Forvis Mazars and Parker Russell", "last_updated": "2026-08-11", "confidence": "high", "note": "Voluntary pension fund contributions, deductible up to VND 3,000,000 per MONTH, annualised x12 here. LEAD REFUTED: the VND 1,000,000/month in Circular 111/2013 is superseded and the ceiling TRIPLED for the 2026 tax year. The 3m/month is shared with life insurance premiums AND with any employer contribution to the same fund; 36m assumes neither. Binds hard at every senior HCMC/Hanoi salary. Decree 253 itself could not be fetched (thuvienphapluat and luatvietnam both 403), hence the law-firm citation chain." } }, "riyadh": { "tax.federal_brackets": { "source": "Saudi Arabia levies NO personal income tax on employment income (nationals or expats); no sub-national/municipal income tax", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "GOSI: Saudi nationals contribute, but non-Saudi expats (ICP) have 0% employee social insurance; only the employer pays 2% occupational-hazard → employee rate 0", "last_updated": "2026-05-30", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Saudi VAT 15% (ZATCA; raised from 5% in Jul-2020)", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "No annual property tax on owner-occupied residential; 5% Real Estate Transaction Tax applies at purchase only → annual rate 0", "last_updated": "2026-05-30", "confidence": "high" }, "housing._summary": { "source": "Numbeo Riyadh 2026 + Aqar/Bayut SA (Al Olaya / Al Malqa / KAFD professional districts)", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "CCHI-mandated employer private health insurance (expats not in the public system); OOP/dental Pacific Prime + Numbeo Riyadh 2025-26", "last_updated": "2026-05-30", "confidence": "medium" }, "childcare._summary": { "source": "Riyadh private nursery / international preschool 2025-26 (Numbeo + international-school DB, mid-tier)", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Riyadh 2026 (upper-middle basket) + Wise", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "Riyadh Metro (opened 2024-25) monthly pass + private car (low fuel cost) 2025-26", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Riyadh leisure + domestic vacation (Jeddah / AlUla / Eastern Province)", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "RUH->US/Europe economy long-haul round-trip 2025-26", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No cash child allowance for expats (Citizen's Account / family support is Saudi-nationals only) → 0", "last_updated": "2026-05-30", "confidence": "high" }, "non_cash._summary": { "source": "Saudi Labor Law: 21-day statutory minimum paid annual leave (30 after 5 yrs tenure), ~13 paid public holidays, 10-week maternity at 100%, sick leave Art. 117 (30d full / 60d at 75% / 30d unpaid). No public universal healthcare for expats", "last_updated": "2026-05-30", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Riyadh: SAR 227,919, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "SAR 227,919 / 3.75 = USD 60,778; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2431, R5-rounded down to 0.24. REPLACES 0.7 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. No income tax; expat gross is about net. Allowance inclusion unknown (housing about 25%, transport about 10% typical). Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.70; 0% income tax + SAR-USD peg lift the USD-equivalent index; package-band anchored just below Dubai (0.85)." }, "non_cash.parental_leave_weeks": { "source": "Royal Decree M/44 (1446H) amending the Saudi Labor Law, in force 2025-02-19: maternity 12 weeks (art. 151, six weeks post-birth mandatory); paternity 3 plain days within 7 days of birth (calendar, /7). 87/7 = 12.43 -> 12.4 (house 1dp). Implementing Regulations (Ministerial Decision 115921/1446) alter neither; no later decree (batch-8 audit 2026-08-08). King & Spalding + Clyde & Co Apr-2025 leave series + Shehata; hrsd.gov.sa still 403s. Research: docs/superpowers/research/2026-08-08-pl-b8-gulf.md", "last_updated": "2026-08-08", "confidence": "high", "note": "12.4 wk = 84 days maternity + 3 days paternity. Secondaries disagree on the paternity article number (113 vs 98) - flagged, not outcome-relevant. Unpaid delayed-birth overrun and the optional 1-month extension are excluded (unpaid)." }, "non_cash.parental_leave_paid_pct": { "source": "Saudi Labor Law art. 151 as amended by M/44 (gazetted Umm al-Qura 2024-08-23, in force 2025-02-19): 'fully paid maternity leave of twelve weeks' with NO tenure condition - art. 152 (the tenure pay tier) was REPEALED IN ITS ENTIRETY. MHRSD publication ('Maternity leave with full pay for working women') + King & Spalding + Clyde & Co + Shehata, convergent. Paternity 3 days full pay. Employer-paid, no monetary cap. Research: docs/superpowers/research/2026-08-08-pl-b8-gulf.md", "last_updated": "2026-08-08", "confidence": "high", "note": "OVERTURN of the 2026-08-06 mover-tier pricing (51.7) on new primary-grade evidence: the under-1-year half-pay tier no longer exists post-M/44, so the zero-tenure mover takes the full 100% (the sweep's source also mis-stated the HISTORIC art. 152, which was unpaid <1y / half 1-3y / full 3+y - 'full at 1 year' was never the rule; commoner-law.com refuted). GOSI maternity compensation (live 2025-07-01) is a FUNDER SUBSTITUTION: for 12-month contributors (Saudi or not) GOSI pays ~3 months on behalf of the employer, not stackable with salary; below the gate the employer's unconditional art. 151 obligation stands - no metric impact." }, "non_cash.vacation_days_statutory": { "source": "Labor Law, Royal Decree M/51, arts. 104, 109 and 112", "last_updated": "2026-08-08", "confidence": "medium", "note": "CORRECTED 2026-08-08, 21 to 15, and this one is a deliberate tie-break rather than a clean fact. Art. 109(1) grants not less than 21 days, rising to 30 after five consecutive years with one employer, so 21 is the floor tier. Art. 109 gives NO calendar-versus-working qualifier, unlike Egypt's art. 124 which excludes rest days explicitly, and the secondary sources genuinely conflict: some describe Saudi annual leave as counted in calendar days, others say weekends never consume the balance. The two live readings are 15 (calendar span, 21 x 5/7) and 21 (already working days, no change). We take 15 because rule 5 of the corrections ledger says to pick the value less favourable to the country where latitude remains, and because 15 is how our Azerbaijan and South Africa 21-day entitlements are already converted. REVISIT if the Implementing Regulations or an HRSD source ever answers the unit question; 21 is the live alternative, not a discarded one. Art. 112 keeps public holidays separate, so they stay additive either way." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Gulf mandatory employer-provided health cover (Council of Cooperative Health Insurance), with its published co-payment structure", "last_updated": "2026-08-08", "confidence": "medium", "note": "CHECKED for the premium-as-out-of-pocket error and CLEAR: the value is co-pay-sized. Dependant coverage is a premium and is not booked here. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Gulf mandatory employer-provided health cover (Council of Cooperative Health Insurance), with its published co-payment structure", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Not reached within the sweep. Lead: the GASTAT Household Transport Survey." }, "non_cash.sick_leave": { "source": "Saudi Labor Law, Article 117 (Royal Decree M/51, as amended).; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-H.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "PwC Worldwide Tax Summaries, Saudi Arabia (reviewed 29 Jul 2026): \"Income tax is not imposed on an individual's earnings if they are derived only from employment\"", "last_updated": "2026-08-11", "confidence": "high", "note": "Recorded for explicitness only; it changes no output by construction. Saudi Arabia dispatches to the same zero-return function as the UAE, which never reads taxable_income. GOSI is social insurance, not income tax." } }, "cape-town": { "tax.filing_basis": { "source": "PwC WWTS - South Africa (Tax administration): spouses are taxed separately and must submit separate returns.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "SARS 2024/25 PIT: 7 brackets 18%-45%; primary rebate R17,235 (modeled as a tax credit, not a 0% band)", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "UIF: 1% employee contribution (employer matches 1%), capped at monthly earnings of R17,712 → annual ceiling R212,544; not deductible from PIT base", "last_updated": "2026-05-30", "confidence": "high" }, "tax.sales_tax_rate": { "source": "South Africa VAT 15% (the proposed 2025 increase was withdrawn and the rate held at 15%)", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "City of Cape Town municipal property rates ~0.72% effective (cents-in-the-rand on valuation, residential rebate applied)", "last_updated": "2026-05-30", "confidence": "medium" }, "housing._summary": { "source": "Numbeo Cape Town 2026 + Property24 (City Bowl / Atlantic Seaboard / Southern Suburbs professional areas)", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Private medical scheme (Discovery/Momentum): no NHI in force yet, professionals use private cover; OOP/dental Numbeo + Pacific Prime 2025-26", "last_updated": "2026-05-30", "confidence": "medium" }, "childcare._summary": { "source": "Cape Town private crèche / preschool / aftercare 2025-26 (Numbeo + provider listings, mid-tier)", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Cape Town 2026 (upper-middle basket) + Wise", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "MyCiTi bus monthly + private car (Cape Town has no comprehensive metro) 2025-26", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Cape Town leisure + domestic vacation (Garden Route / Kruger / Winelands)", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "CPT->US/Europe economy long-haul round-trip 2025-26", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "SA Child Support Grant (~R530/month) is means-tested and falls away well below an ICP tech income → 0", "last_updated": "2026-05-30", "confidence": "high" }, "non_cash._summary": { "source": "BCEA: 15 statutory paid annual-leave days (3 weeks on a 5-day week), 12 public holidays, ~4 months parental leave at UIF ~38% replacement, sick leave 30 days per 36-month cycle. No universal public healthcare (NHI not yet operative)", "last_updated": "2026-05-30", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Cape Town. Its basis is converted with the rest of South Africa and its rank inside South Africa is preserved. Prior source: Levels.fyi/Glassdoor + OfferZen Cape Town tech 2025-26", "last_updated": "2026-08-11", "confidence": "low", "note": "0.18 x 0.7222 = 0.1300, R5-rounded down to 0.13. 0.7222 is the MEDIAN of (re-derived / previous) across the 1 South Africa city that DO carry an all-levels measurement (Johannesburg 0.18 to 0.13), so Cape Town moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Cape Town's rank inside South Africa is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.18; ZAR-denominated tech pay is low in USD terms — ranks with São Paulo (0.32) above and Bangalore/Manila/Bogotá (~0.18-0.20) alongside." }, "non_cash.parental_leave_weeks": { "source": "UI Act 63/2001 s.13(3)(a) (1 credit day per 5 employed days, 365-day cap) + s.13(5); BCEA s25A (10 consecutive days parental leave) + UI Amendment Act 10/2016 s.12(3)(c) (parental benefit at the flat 66%); Van Wyk v Minister of Employment and Labour [2025] ZACC 20", "last_updated": "2026-08-07", "confidence": "medium-high", "note": "11.86 wk = 10.43 maternity + 1.43 s25A parental (10 consecutive calendar days, the father's own UIF-paid entitlement), MOVER-PRICED per the standing zero-tenure ruling: a ~12-month mover accrues ~73 UIF credit days (1 per 5 employed days), below the 121-day maternity maximum, so the credits are the binding number. s.13(5) prevents maternity draws from consuming credits for later unemployment; it does not exempt maternity from the accrual limit (statute reading corroborated by Leave Network + 2026 practitioner consensus, hence medium-high). Tenured alternative 18.71 wk recorded in the flat-block note. Was 17, the un-gated maternity maximum with the father's s25A days dropped. Household-inclusive basis; research: docs/superpowers/research/2026-08-07-pl-b5-africa-gulf-sea.md." }, "non_cash.parental_leave_paid_pct": { "source": "Unemployment Insurance Amendment Act 10 of 2016 (maternity benefits at a flat 66% of earnings, off the 38-60% unemployment IRR scale; parental leave benefit at the same flat 66% per s.12(3)(c)); UIF earnings ceiling ZAR 17,712/month, unchanged since 1 June 2021", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild converted South Africa to the flat shape because the UIF ceiling (ZAR 212,544/yr) sits 5-13x below any ICP salary here, so the benefit is always the capped amount. The 38 this replaces was ALSO the wrong statutory rate: it was the unemployment-benefit IRR floor, not the flat 66% maternity rate the 2016 amendment moved maternity onto. Correctness bug recorded in docs/corrections-ledger.md (internal:za-parental-38pct)." }, "non_cash.parental_leave_flat": { "source": "UI Act 63/2001 as amended by Act 10/2016: maternity at 66% of earnings up to the UIF earnings ceiling of ZAR 17,712/month = ZAR 212,544/yr (unchanged since 1 June 2021; equals the stored tax.social_security_cap, an independent internal corroboration). Research: docs/superpowers/research/2026-08-06-parental-caps-americas-africa.md (South Africa) and docs/superpowers/research/2026-08-06-parental-caps-southern-east-africa.md", "last_updated": "2026-08-06", "confidence": "high", "note": "ZAR 2,697.67/week = 66% x (212,544 / 52), covering all 11.86 weeks: 10.43 wk maternity + 1.43 wk BCEA s25A parental leave (10 consecutive days, the other parent's own entitlement, UIF-paid at the same flat 66% per UI Amendment Act 10/2016 s.12(3)(c) — the batch-5 audit stored the previously flagged follow-up). The cap always binds at ICP salaries (~5.7% effective at the $150k basis). Weeks are MOVER-PRICED (2026-08-07): UIF benefits are credit-day limited, s.13(3)(a) accrues 1 benefit day per 5 employed days, and s.13(5) only stops maternity CONSUMING credits, it does not exempt maternity from the accrual limit; a ~12-month mover holds ~73 credit days = 10.43 wk, not the 121-day (17.29 wk) statutory maximum. Tenured alternative (605+ employed days, e.g. a second child in-country): 17.29 + 1.43 = 18.71 wk, recorded not stored. Van Wyk watch item stands: ConCourt [2025] ZACC 20 (3 Oct 2025) confirmed the gendered-structure invalidity, suspended ~36 months to around Oct 2028; current structure remains law in force, re-check before 2028. Research: docs/superpowers/research/2026-08-07-pl-b5-africa-gulf-sea.md." }, "non_cash.vacation_days_statutory": { "source": "Basic Conditions of Employment Act 75 of 1997 s.20(2)", "last_updated": "2026-08-08", "confidence": "medium-high", "note": "RE-VERIFIED 2026-08-08, 15 stands. s.20(2)(a) grants 21 CONSECUTIVE days of annual leave per leave cycle, with s.20(2)(b) offering the alternative of one day per 17 days worked. 21 consecutive days is three weeks, which is 15 five-day weekdays, and that is what we store. The Act itself frames the entitlement in consecutive days, so this is one of the calendar-to-weekday conversions that was already applied correctly." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita plus medical-scheme co-payment and self-payment-gap structure", "last_updated": "2026-08-08", "confidence": "medium", "note": "Medical-scheme CONTRIBUTIONS are a premium and are not booked here; this is the scheme's co-payments and the self-payment gap. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita plus medical-scheme co-payment and self-payment-gap structure", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Statistics South Africa, National Household Travel Survey 2020, tables 4.22 and 4.3, Western Cape. https://www.statssa.gov.za/publications/P0320/P03202020.pdf", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Total time travelled to place of work", "year": "2020", "note": "Mode-blind mean of 49.2 computed the same way as Johannesburg. Western Cape is diluted by rural areas, so this is a FLOOR for the Cape Town metro rather than a centred estimate." }, "non_cash.sick_leave": { "source": "Basic Conditions of Employment Act 75 of 1997 (BCEA), s.22 (sick leave entitlement); Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-H.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "SARS, What are the s11F annual allowable deductions https://www.sars.gov.za/faq/faq-what-are-s11f-annual-allowable-deductions/ ; https://www.sars.gov.za/latest-news/retirement-fund-contribution-deductions-section-11f2a/ ; Budget 2026 FAQ https://www.sars.gov.za/about/sars-tax-and-customs-system/budget/budget-2026-frequently-asked-questions/", "last_updated": "2026-08-11", "confidence": "medium", "note": "S11F: the lesser of the monetary cap, 27.5% of the greater of remuneration or taxable income, and taxable income before the deduction. Pension plus provident plus RA combined; excess carries forward (unmodeled). THE CAP MOVED AND THE HOUSE RULE DECIDED IT: SARS's Budget 2026 FAQ says R430,000 from 1 Mar 2026, the first change since 2016, but SARS's own still-live s11F pages say R350,000 and the increase was OMITTED from the Rates and Monetary Amounts Act gazetted 1 Apr 2026, with Treasury implementing under the ministerial-announcement provision. R350,000 is used (lower cap, higher tax). REFRESH ITEM. Employer contributions DO count against the same ceiling, but less_employer_pct is deliberately not encoded because it would require inventing an employer rate and employee 6-10% plus a typical employer 5-10% still sits under 27.5%. Does not bind at realistic South African salaries." } }, "istanbul": { "tax.filing_basis": { "source": "PwC WWTS - Turkey: each individual files an individual return; joint assessment is not permitted.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "GİB wage-income gelir vergisi 2026 schedule (Communiqué No. 332, Official Gazette 31-Dec-2025): 15/20/27/35/40% at 190k/400k/1.5M/5.3M cumulative-annual thresholds (wage-specific bands, more generous than other-income)", "last_updated": "2026-05-30", "confidence": "high" }, "tax.min_wage_exemption_annual": { "source": "Asgari ücret istisnası (Law 7349, 2022): income up to the gross minimum wage is income-tax-exempt for ALL employees, applied as a credit; 2026 gross min wage TRY 33,030/mo net of 15% SGK → annual income-tax base TRY 336,906", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "SGK employee 14% (pension+health) + 1% unemployment (işsizlik) = 15%; 2026 ceiling (tavan) raised to 9× gross min wage = TRY 297,270/mo → TRY 3,567,240/yr; SGK is deductible from the income-tax base", "last_updated": "2026-05-30", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Turkey KDV (VAT) standard rate 20% (raised from 18% mid-2023)", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Emlak vergisi residential 0.1% doubled to 0.2% inside metropolitan municipalities (Istanbul); applied to tax-assessed value, which runs below market", "last_updated": "2026-05-30", "confidence": "medium" }, "housing._summary": { "source": "Numbeo Istanbul 2026-Q2 + Kadıköy/Beşiktaş/Şişli/Ataşehir district rent surveys (professional-tier, slightly below Numbeo city-centre)", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Full private özel sağlık sigortası premiums (on top of universal SGK/GSS) + OOP/dental, Turkish insurer 2025-26 price surveys", "last_updated": "2026-05-30", "confidence": "medium" }, "childcare._summary": { "source": "Istanbul private kreş / preschool / aftercare 2025-26 (Numbeo + local provider listings, professional tier)", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Istanbul 2026 grocery + restaurant indices (professional basket)", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "Istanbulkart monthly (İETT/Metro İstanbul) + private car all-in 2025-26", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Istanbul leisure + domestic vacation (Antalya/Bodrum/Cappadocia)", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "IST->US/W.Europe economy long-haul round-trip 2026 (Skyscanner/Kayak), blended", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No private-sector cash child allowance for a high earner (means-tested support falls away; the AGİ child component was abolished with the 2022 min-wage-exemption reform) → 0", "last_updated": "2026-05-30", "confidence": "high" }, "non_cash._summary": { "source": "Labor Law 4857: 14 statutory paid annual-leave days (1-5 yrs service), 14 public holidays, 16 weeks maternity at SGK ~67% replacement (a May-2026 reform extends post-birth leave toward 24 weeks total; modeled at the established 16), sick leave via SGK temporary-incapacity benefit. Universal SGK/GSS healthcare", "last_updated": "2026-05-30", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Istanbul: TRY 1,579,052, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-europe.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "TRY 1,579,052 / 47.51 = USD 33,236; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1329, R5-rounded down to 0.13. REPLACES 0.2 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. TRY pooled across an inflationary period with no published vintage, so the lira figure is not a stable quantity. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.20; the lira's depreciation has gutted USD-equivalent local tech pay — sits just above Bangalore (0.20)/Manila (0.18), below São Paulo (0.32). Most FX-volatile field." }, "non_cash.parental_leave_weeks": { "source": "Law 7578 (Resmi Gazete 1 May 2026, Sayi 33240, in force 1 May 2026): maternity 16 -> 24 weeks (8 pre + 16 post birth), SGK pays the full span (alomaliye.com fetched 2026-08-06; SGK Genelge 2026/13); paternity 5 -> 10 days, employer-paid. Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (Turkey) and docs/superpowers/research/2026-08-06-pl-caps-israel-turkey.md", "last_updated": "2026-08-08", "confidence": "high", "note": "25.43 wk = 168 days maternity + 10 days paternity (178/7). Was 16, the pre-reform maternity-only figure the old _summary note already flagged as at risk. Single-birth case stored; the multiple-birth 26-week variant excluded. Batch-7 audit (2026-08-08): paternity 10 days ruled CALENDAR (Is Kanunu mazeret-izni convention; PwC 7578 bulletin + practitioner convergents) - /7 stands. SGK gate: 90 days of short-term premiums in the year before birth - the mover passes comfortably; the full 24 weeks including the 8 pre-birth are SGK-paid (paid pre-birth weeks count per the house basis). MOVER-EXCLUDED: the ISKUR yarim calisma odenegi (60/120/180 half-time days at minimum wage after maternity, Law 4447 Ek Madde 5, verified still in force post-7578) requires 600 days of unemployment-insurance premiums in the last 3 years - the ~270-360-day mover fails this discrete gate, so it is excluded from the store per the South Africa precedent; the tenured alternative is recorded in the cluster research file, not stored. 24-month earnings averaging dilutes a mover's daily base at the margin - continuous reference-window dilution, not priced (A8)." }, "non_cash.parental_leave_paid_pct": { "source": "Is Kanunu 4857 Ek Madde 2 as amended by Law 7578: 10-day babalik izni, employer-paid at full wage, no SGK intermediation (\"iscinin ucretinde kesinti yapilamaz\")", "last_updated": "2026-08-06", "confidence": "high", "note": "100 covers only the 1.43-week paternity remainder, direct employer wage continuation, genuinely uncapped. The 24 maternity weeks live in non_cash.parental_leave_capped. Was 67, the maternity-only rate. The engine blend reproduces ~68.9% at low salaries, matching the research's 68% conservative blended proposal." }, "non_cash.parental_leave_capped": { "source": "5510 sayili Kanun art. 17 (benefit = 2/3 of average daily prime esas kazanc of the last 3 months) cross-referencing art. 78 (2026 daily earnings ceiling TRY 9,909.00 = 9x the TRY 1,101/day gross minimum wage, raised from 7.5x on 1 Jan 2026; EY Türkiye Sirküler 7, verginet, cottgroup, alomaliye, muhasebetr concur). Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (Turkey)", "last_updated": "2026-08-06", "confidence": "high", "note": "Payout ceiling 2/3 x TRY 9,909 = TRY 6,606.00/calendar day on the 24 SGK-paid maternity weeks (monthly equivalent TRY 198,180 payout off the TRY 297,270 input ceiling, the same figure as the stored tax.social_security_cap, structurally identical by statute). Does not bind inside the TRY 800k-2.5M ICP band (binds above ~TRY 3.57M/yr input), but this is the fastest-moving threshold in the dataset: the ceiling multiplier and minimum wage both moved Jan 2026, next revision expected Jan 2027, and no mid-year 2026 revision is planned (verified). Re-check at every annual refresh." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "SGK cost-sharing structure, including free prescriptions and enrolment for children", "last_updated": "2026-08-08", "confidence": "medium", "note": "Child-discount shape applied on a directly-sourced government child exemption, one of only three countries in the long tail where that was verified rather than assumed. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "SGK cost-sharing structure, including free prescriptions and enrolment for children", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Istanbul Buyuksehir Belediyesi, Istanbul Metropoliten Alani Entegre Kentsel Ulasim Ana Plani (IUAP), Final Rapor, section 2.3. https://www.ibb.gov.tr/tr-TR/kurumsal/Birimler/ulasimPlanlama/Documents/%C4%B0UAP_Ana_Raporu.pdf", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Ortalama Yolculuk Suresi, ev-is (average journey time, home to work)", "year": "2006 survey base", "note": "Door-to-door by the plan's own definition. Mean of 52.8 fitted from the published quantiles, triangulated against 48 and 50.8 from other cuts, giving a band of 48 to 53. Moovit's Istanbul figure of 77 min was rejected as public-transport riders only. TUIK's Zaman Kullanim Arastirmasi was identified as a round-trip time-use statistic and rejected." }, "non_cash.sick_leave": { "source": "Law No. 5510 (Sosyal Sigortalar ve Genel Sağlık Sigortası Kanunu), Art. 18; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-H.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "GVK art. 63/3 as amended by Law No. 6327 of 29 June 2012 (in force 1 Jan 2013) ; GIB 2026 hayat-sahis sigorta primleri infographic https://cdn.gib.gov.tr/api/gibportal-file/file/getFile?objectKey=DUYURU/UNIVERSAL/2026/2026_hayatsahis_sigortaprim_info.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "BES contributions by employer or employee \"can in no way be deducted\" from the taxable wage base. The relief was replaced by devlet katkisi, a 25% STATE PAYMENT into the account capped at the annual gross minimum wage, which is a subsidy and not representable in this field. The surviving wage-base deduction is sahis sigortasi only (15% of monthly gross, capped at the annual minimum wage), which is protection insurance rather than retirement; GIB's own 2026 infographic is titled for it with BES absent by design." } }, "bucharest": { "tax.federal_brackets": { "source": "Romania flat personal income tax (impozit pe venit) 10% on salary, levied on gross net of mandatory social contributions; the IT-sector income-tax exemption was eliminated for 2025+ (Law 296/2023 + 2025 fiscal measures)", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "Employee CAS (pension) 25% + CASS (health) 10% = 35% of gross, uncapped on employment income; both deductible from the 10% PIT base (so effective combined burden ≈ 41.5% for a high earner with no personal deduction)", "last_updated": "2026-05-30", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Romania TVA (VAT) standard rate raised from 19% to 21% effective 1 Aug 2025 (Official Gazette, GD 602/2025); reduced rates merged into 11%", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Impozit pe clădiri (residential building tax), Bucharest local council ~0.08–0.2% of taxable value; modeled 0.1%", "last_updated": "2026-05-30", "confidence": "medium" }, "housing._summary": { "source": "Numbeo Bucharest 2026-Q2 + Sector 1–3 (Floreasca/Dorobanți/Cotroceni) professional-tier rent surveys, RON/mo", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Private medical subscriptions (Regina Maria / MedLife / Sanador) on top of universal CASS/CNAS coverage; insurer 2025-26 price lists + OOP/dental", "last_updated": "2026-05-30", "confidence": "medium" }, "childcare._summary": { "source": "Bucharest private creșă / grădiniță / after-school 2025-26 (provider listings, professional tier); public crèche heavily subsidized", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Bucharest 2026 grocery + restaurant indices (professional basket)", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "STB Bucharest integrated monthly pass (metro + surface) + private car all-in 2025-26", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Bucharest leisure + domestic vacation (Black Sea / Brașov / Transylvania)", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "OTP->US/W.Europe round-trip 2026 (Skyscanner/Kayak); cheap intra-EU via Wizz/Ryanair, long-haul blended", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Alocația de stat pentru copii: universal (non-means-tested) state child allowance ~RON 300/mo for ages 2–18 → ~RON 3,600/yr per kid", "last_updated": "2026-05-30", "confidence": "high" }, "non_cash._summary": { "source": "Romanian Labour Code: 20 statutory paid annual-leave days, ~15 legal public holidays, parental leave (concediu creștere copil) payable up to the child's 2nd birthday at 85% of prior net income (capped); universal CASS/CNAS healthcare", "last_updated": "2026-05-30", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Bucharest: RON 199,629, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-europe.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "RON 199,629 / 4.56 = USD 43,778; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1751, R5-rounded down to 0.17. REPLACES 0.3 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.30; one of CEE's largest dev hubs — sits just below Warsaw/Prague, above São Paulo (0.32) on PPP but below on USD TC; pairs with Poland/Czechia/Hungary." }, "healthcare.statutory_employee": { "source": "Codul fiscal (Legea 227/2015) Titlul V: art. 156 sets CASS at 10% for employees; art. 157 defines the base as castigul brut with no ceiling clause; art. 170 by contrast sets explicit 6, 12 and 24 minimum-wage ceilings for non-salary income, which is the textual proof that the omission in art. 157 is deliberate. Art. 138 sets employee CAS at 25% and art. 139 gives it the same uncapped salary base. Art. 78(2)(a) makes both deductible before the 10% income tax. Consolidated code current to 2025-12-17 per ANAF (static.anaf.ro/static/10/Anaf/Legislatie_R/Codfiscal.htm). Sector exemptions for IT, construction, agriculture and food were removed by OUG 156/2024 (income tax, from 2025-01-01) and Legea 141/2025 (CASS, from 2025-08-01).", "last_updated": "2026-07-31", "confidence": "medium", "note": "DISPLAY ONLY, already inside taxes.payroll. UNCAPPED is proven by contrast rather than by a 'there is no cap' sentence: the legislature wrote explicit ceilings into art. 170 a few articles away and did not repeat them in art. 157. CAVEAT ON THE SOURCING ROUTE: legislatie.just.ro refused connections all session and ANAF's consolidated code is too large for a fetcher to return article bodies, so ANAF confirmed the article numbering and currency date while the verbatim text came from an article-indexed mirror, corroborated by PwC Worldwide Tax Summaries. Nothing contradicted it. The 2025 sector-exemption repeal dates are secondary-sourced. A refuted figure worth recording so nobody re-imports it: one ANAF fetch returned a 10.5/15.8 CAS split, which is the pre-2018 regime from before OUG 79/2017 moved contributions onto employees." }, "non_cash.parental_leave_paid_pct": { "source": "OUG 111/2010 as amended (indemnizatia pentru cresterea copilului, 85% of average net income over the last 12 of 24 months)", "last_updated": "2026-08-08", "confidence": "medium-high", "note": "We store 100: it rides only the 2.0 paternity remainder weeks (Legea 210/1999 as am. OUG 117/2022: 10 working days at full salary, employer-paid, uncapped). The 104 maternity+CIC weeks live in non_cash.parental_leave_capped. Was 0 when the whole 104-week window sat in the capped block with no remainder (2026-08-06 caps rebuild); the 2026-08-08 basis audit added the paternity weeks and the maternity tier." }, "non_cash.parental_leave_capped": { "source": "Maternity tier: OUG 158/2005 - calculation base capped at 12x the gross national minimum wage (RON 4,325/month from 2026-07-01 -> payout cap 85% x 51,900 = RON 44,115/month; H1-2026 min wage was 4,050). CIC tier: OUG 111/2010 - ceiling RON 8,500/month gross, unchanged since Sep 2022 (gov.ro release; 2026 convergents financiarul.ro / NN / estimez). Research: docs/superpowers/research/2026-08-08-pl-b6-latam-romania.md", "last_updated": "2026-08-08", "confidence": "medium-high", "note": "cap_amount RON 14,664/month is the DURATION-WEIGHTED payout ceiling over the 104 capped weeks: (18 wk maternity x 44,115 + 86 wk CIC x 8,500) / 104 (the Korea/Kazakhstan flatten precedent). EXACT wherever both caps bind - annual gross >= RON 622,800 (~$136.6k); the RON 684,000 $150k basis qualifies (effective 27.1%, verified identical computed tier-by-tier). Between ~RON 141k and 622,800 the flatten OVERSTATES the maternity slice (18/104 of the window) - recorded and accepted per the house flatten precedent. The in-force-at-research-date minimum wage governs per the batch-3 mid-year-step ruling (4,050 would give 14,179, recorded). The cap TRACKS 12x minimum wage - re-check each January and July (refresh-calendar row added 2026-08-08). Caveats: CASS 10% is withheld from the CIC indemnity since 2025-08-01 (Legea 141/2025; net at the 8,500 cap = 7,650); the maternity indemnity is CASS-exempt but bears CAS 25% per the better-supported reading (estimez vs contabss conflict recorded); CIC floor ~RON 1,650 (2.5x ISR, exact 2026 ISR not pinned) never binds at ICP; CIC rides NET income while maternity rides GROSS - both caps bind at ICP salaries so the base distinction does not affect the stored shape." }, "non_cash.parental_leave_weeks": { "source": "OUG 158/2005 (concediu de maternitate: 126 calendar days = 63 pre + 63 post, min 42 postnatal mandatory, 85% of last-6-months gross) + Legea 210/1999 as am. OUG 117/2022 (paternity 10 working days, employer-paid full salary) + OUG 111/2010 (CIC to child age 2; 2-month non-transferable other-parent quota per OUG 164/2022, M.Of. 1173/2022-12-07). legislatie.just.ro and lege5.ro unreachable (dead fetches logged); 3+ convergent secondaries per figure. Research: docs/superpowers/research/2026-08-08-pl-b6-latam-romania.md", "last_updated": "2026-08-08", "confidence": "medium-high", "note": "106 wk = 18.0 maternity + 86.0 CIC + 2.0 paternity (10 working days /5; the +5 puericultura-course days are conditional, excluded). CONSERVATIVE ELECTION (recorded ruling): the mother may shift prenatal maternity days postnatal, so the household total spans 106 (all 126 days taken postnatal; CIC = 104 - 18 = 86) to 115 (standard 63/63 split; CIC = 95 plus 9 paid pre-birth weeks additive per the Austria precedent) - the stored 106 is the MINIMUM of the statutory election range. CIC is one shared window with the 2-month other-parent quota INSIDE it ('1 month' claims are pre-OUG-164/2022, refuted) - counted once, both_parents NOT set (Sweden precedent). Was 104 = the CIC window alone with maternity and paternity dropped. Stimulent de insertie excluded (paid while working, not leave pay). LOUD MOVER CAVEAT: CIC eligibility = 12 months of taxable income within the 2 years before birth - a conceive-immediately mover (~9-10 months at birth) narrowly misses; work-to-term passes; EU-origin movers aggregate periods under Reg. 883/2004 (caveat-only per the Israel/Lithuania precedent). The maternity gate (secondaries conflict: 1 vs 6 months stage in the last 12) passes either way for the mover." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "CNAS cost-sharing structure, including free prescriptions for under-18s", "last_updated": "2026-08-08", "confidence": "medium", "note": "Child-discount shape applied on a directly-sourced government child exemption. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "CNAS cost-sharing structure, including free prescriptions for under-18s", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Planul de Mobilitate Urbana Durabila Bucuresti-Ilfov 2016-2030, Tabel 7-8. https://www.adrbi.ro/media/2354/pmud_2016-2030.pdf", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Durata calatorie spre EGC (travel time to employment growth zones)", "year": "2015", "note": "42.77 min one way for the Bucuresti-Ilfov metro region. SUMP fallback, because INS publishes no travel-time series at all. Stored value confirmed." }, "non_cash.sick_leave": { "source": "OUG 158/2005 privind concediile și indemnizațiile de asigurări sociale de sănătate, as amended by **Legea 141/2025** (rate-tiering reform) and **OUG 91/2025** (introduced the 1-day carență/waiting day, in force 1 Feb 2026 - 31 Dec 2027, i.e. CURRENTLY ACTIVE today). Administered by FNUASS (Fondul Na; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-C.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Codul fiscal art. 78 alin. (2) lit. a) https://www.noulcodfiscal.ro/titlu-4/capitol-3/articol-78.html ; 2026 status https://agerpres.ro/economic/2026/02/05/ministerul-finantelor-propune-deduceri-fiscale-pentru-pensiile-ocupationale-incepand-cu-2026-proiect--1525339", "last_updated": "2026-08-11", "confidence": "medium", "note": "Voluntary pension contributions under Legea 204/2006 borne by the EMPLOYEE are subtracted from monthly gross in computing the taxable base, capped at the leu equivalent of EUR 400/yr. Art. 78 is the employee-borne rule; art. 76(4)(s) is the employer-borne twin at the same EUR 400 and most search results land there by mistake. Voluntary health insurance has its own separate EUR 400. Conversion is house-ruled twice: a Senate proposal to triple the ceiling to EUR 1,200 was NOT enacted, and on FX the repo's own build/fx-rates.json implies RON 2,104 while the BNR rate implies about 2,032, so RON 2,000 is the lowest defensible figure and matches what Romanian practitioners quote. Statute defines conversion at the BNR rate on the last day of the salary month." }, "healthcare.premium_monthly_by_family_size": { "source": "Romanian statutory health insurance (CASS) is an employee contribution of 10% of gross, which this file already carries in tax.medicare_rate 0.10 alongside CAS pension 25% in tax.social_security_rate.", "last_updated": "2026-09-01", "confidence": "high", "note": "ZEROED 2026-09-01 from 180/340/480/600/720 RON per month. Same double-count as belgrade, found in the same pass by sweeping every European city whose statutory health is payroll-funded rather than only the city that was reported. CASS was withheld inside tax.medicare_rate and charged again here as a premium. No user reported this one; it is recorded in docs/corrections-ledger.md as found-on-the-way. Basis now matches the seven peer cities that already carried 0." } }, "sofia": { "tax.filing_basis": { "source": "PwC WWTS - Bulgaria (Individual): individual assessment; flat 10% PIT. (Immaterial: the flag only corrects the per-earner capped social-insurance base, ~0.3pp on two earners.)", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Bulgaria flat personal income tax (данък върху доходите) 10% on annual taxable base (gross net of the employee social/health contributions); flat since 2008, unchanged for 2026", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "Employee share of social + health insurance ≈ 13.78% of gross (pension/DOO ~10.58% incl. universal pension, health/NЗ OK 3.2%), levied up to the monthly max insurable income ceiling (BGN 4,130/mo ≈ €2,112 → ~€28,224/yr in 2026); both deductible before the 10% flat PIT", "last_updated": "2026-05-30", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Bulgaria ДДС (VAT) standard rate 20%; reduced 9% for hotels/restaurants/books (NRA, unchanged for 2026)", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Sofia Municipality annual real-estate tax (данък сгради) ~1.625‰ + waste fee; combined modeled ~0.19% of tax value", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.utilities_monthly": { "source": "Numbeo Sofia 85-sqm mean EUR 143.20 (range 102-220, fetched 2026-08-07); innovires.com Sofia cost-of-living guide 2026-04-09 (85-sqm basic utilities avg EUR 120-150/mo across the year, winter 150-200, https://innovires.com/tax-residency/blog/cost-of-living-sofia-expats.html); EWRC 2026-07-01 tariffs: Toplofikacia Sofia 73.30 EUR/MWh (capital.bg 2026-07-02), household electricity +3.05% (sofiaglobe.com 2026-07-01), Sofiyska Voda ~4 BGN/m3 from 2026-01-01", "last_updated": "2026-08-08", "confidence": "medium-high", "note": "2BR (~83 sqm) reference incl. household electricity. Cut 180 -> 150 closing the adjudicated user flag (8c7dcc79): stored value sat 26% above Numbeo's own mean on the same 85-sqm basis. Higher sourcing bar for a cut met twice over Numbeo: innovires annual band tops at 150, and the EWRC tariff bottom-up (district heat ~40-45 + electricity ~23-27 + water ~16 = ~80-90/mo core) sits far below it. 150 = top of every central band, conservative margin retained. Derivation: docs/superpowers/research/2026-08-08-utilities-values.md" }, "housing._summary": { "source": "Numbeo Sofia 2026-Q2 + imot.bg professional-tier (Lozenets/Center/Iztok) rent & purchase surveys, converted to EUR (Bulgaria adopted the euro 1 Jan 2026 at the fixed BGN 1.95583/€ peg)", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Private health packages (Bulstrad/DZI) on top of universal NЗOK coverage; insurer 2025-26 price lists + OOP/dental, EUR", "last_updated": "2026-05-30", "confidence": "medium" }, "childcare._summary": { "source": "Sofia private детска градина / ясла / after-school 2025-26 (provider listings, professional tier); municipal nurseries near-free but waitlisted", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Sofia 2026 grocery + restaurant indices (professional basket), EUR", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "Sofia Urban Mobility Centre integrated monthly pass (metro + surface) + private car all-in 2025-26, EUR", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Sofia leisure + domestic vacation (Black Sea / Bansko / Rila), EUR", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "SOF->US/W.Europe round-trip 2026 (Skyscanner/Kayak); cheap intra-EU via Wizz/Ryanair, long-haul blended, EUR", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Месечни помощи за дете (child allowance) is means-tested with an income ceiling far below a senior-tech salary → modeled 0 for the ICP earner", "last_updated": "2026-05-30", "confidence": "high" }, "non_cash._summary": { "source": "Bulgarian Labour Code: minimum 20 statutory paid annual-leave days, 12 official public holidays, paid maternity/parental leave for 410 days (~59 weeks) at 90% of average daily gross; universal NЗOK healthcare", "last_updated": "2026-05-30", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Sofia: EUR 27,095, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-europe.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 27,095 / 0.867 = USD 31,251; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1250, R5-rounded down to 0.12. REPLACES 0.28 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. CURRENCY CORRECTION. levels.fyi still labels Sofia in BGN and shows 52,993. Bulgaria adopted the euro on 2026-01-01 at the fixed irrevocable 1.95583 and the lev ceased to be legal tender 2026-02-01, so the real figure is EUR 27,095. Reading the raw 52,993 under our EUR label would have overstated Sofia by 96%. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.28; slightly below Bucharest (0.30) — comparable CEE dev hub, low cost base, capped social makes net take-home unusually high for the region; pairs with Romania/Poland/Czechia." }, "healthcare.statutory_employee": { "source": "Zakon za byudzheta na darzhavnoto obshtestveno osiguryavane za 2026 (2026 State Social Insurance Budget Act) art. 9, via the Darzhaven vestnik document portal dv.parliament.bg (idMat=244982), which sets the maximum monthly insurable income at EUR 2,111.64 to 2026-07-31 and EUR 2,300.00 from 2026-08-01, holding at EUR 2,300 for 2027 and 2028. Health contribution 8% total with a 60/40 employer/employee split under Zakon za zdravnoto osiguryavane art. 29, giving the employee 3.2%. Euro adoption 2026-01-01 at the fixed rate EUR 1 = BGN 1.95583.", "last_updated": "2026-07-31", "confidence": "medium", "note": "DISPLAY ONLY, already inside taxes.payroll (health 3.2% sits inside the modeled 13.78% employee rate). The EUR 27,600 ceiling is the forward-looking one and is also the higher, therefore conservative, of the two defensible 2026 readings; the exact calendar-2026 prorated base is EUR 26,281.48 (7 months at 2,111.64 plus 5 at 2,300). CAVEAT: the ceiling is primary from the budget act, but the 8% rate and its 60/40 split rest on convergent secondary sources, because nra.bg and nssi.bg refused connections. It is a longstanding unamended provision with no dissent among sources, hence medium rather than high. SEPARATE ISSUE, not fixed here: tax.social_security_cap for Sofia is EUR 28,224, which does not reconcile with its own citation (BGN 4,130/month = EUR 2,111.64 = EUR 25,340/yr) and is now superseded by the August step to EUR 2,300/month." }, "non_cash.parental_leave_paid_pct": { "source": "KSO (Social Insurance Code) art. 49: obezshtetenie za bremennost i razhdane at 90% of average gross insurance income over the prior 24 months, paid by NOI", "last_updated": "2026-08-08", "confidence": "high", "note": "We store 0: all 121.5 weeks live in the blocks - 60.7 wk in non_cash.parental_leave_capped (90% under the NOI payout ceiling) and 60.8 wk in non_cash.parental_leave_flat (the ZBDOO flat amount) - so no percentage-paid weeks remain on this field. Before the batch-7 household restate this was the same 0 over 59 capped-only weeks." }, "non_cash.parental_leave_capped": { "source": "Darzhaven vestnik issue 68 / 2026-07-28 (2026 State Social Insurance Budget Act): maximalen osiguritelen dohod EUR 2,300/month from 1 August 2026 (Jan-Jul 2026 interim: EUR 2,111.64 = BGN 4,130 at the retired peg, DV 113/2025-12-23); corroborated by kik-info.com, accountingnews.bg, 3e-news.net", "last_updated": "2026-08-08", "confidence": "high", "note": "cap_amount stores the PAYOUT ceiling per the caps convention: 90% x the EUR 2,300/month maximum insurable income = EUR 2,070/month, on the whole 59-week (410-day) benefit. Bulgaria adopted the euro on 2026-01-01 (BGN retired at 1.95583), so 2026 figures are native EUR. The mid-year step (2,111.64 Jan-Jul, 2,300 from 1 Aug) is current law, not a judgment call, so the in-force figure is stored. Refutation recorded: innovires.com's \"BGN 4,600 / ~EUR 2,352\" could not be corroborated and contradicts the Gazette-cited figure; rejected. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Bulgaria). Batch-7 restate (2026-08-08): block weeks moved 59 -> 60.7 - the father's 15-day birth benefit (KSO art. 50(6), 12-month gate like the mother's) rides the same 90% rate and NOI ceiling, so it folds into this block; the cap figure itself is unchanged and current. An additional statutory cap (daily benefit <= the earner's average daily NET pay) exists but the insurable-income ceiling binds first at ICP pay." }, "non_cash.parental_leave_weeks": { "source": "KSO arts. 48a-50 + KT art. 163 (maternity 410 days, 45 pre-birth; NOI 90% of the 24-month average) ; KSO art. 50(6) father's 15 calendar days at the same 90%/ceiling ; KT art. 164 / KSO art. 53 child-raising leave to age 2 at the ZBDOO flat benefit ; KSO art. 53g / KT art. 164v father's own 2 months (nssi.bg primary pages fetched 2026-08-08) ; ZBDOO 2026 adopted 22 Jul 2026 (DV consistent with DV 68/2026). Research: docs/superpowers/research/2026-08-08-pl-b7-europe-caucasus.md", "last_updated": "2026-08-08", "confidence": "high (structure and rates) / medium-high (flat-amount effective date)", "note": "121.5 wk household total = capped stream 60.7 wk (maternity 410 d + father 15 d, both at 90% under the NOI ceiling) + flat stream 60.8 wk (child-raising to age 2: 365 postnatal maternity days put the child at ~12 months when maternity ends, so ~365 days remain to the 2nd birthday = 52.1 wk, ONE alternating pot counted once, transfers excluded + the father's OWN non-transferable art. 53g 2 months = 8.7 wk, usable child ages 2-8 - deferred but statutory paid, Sweden multi-year-pot precedent; anti-double-dip art. 164v(2) bites only transfer households). Was 59, the maternity benefit alone. Art. 54 return-to-work election (75% of the flat benefit while working) excluded as an alternative election. LOUD mover caveat: all four components carry a 12-month CUMULATIVE insurance gate ('prekasnat ili neprekasnat' - any employer, any career; EU movers aggregate under Reg 883/2004); a zero-career non-EU mover conceiving on arrival could miss the opening months. Not priced, per the A8 + Lithuania/Romania precedent." }, "non_cash.parental_leave_flat": { "source": "ZBDOO 2026 (adopted 22 Jul 2026; bgonair/tribune coverage: child-raising benefit EUR 460.17/month = BGN 900 at the retired peg) ; nssi.bg art. 53 (dete do 2-godishna vazrast) + art. 53g (dete do 8-godishna vazrast ot bashtata) pages, fetched 2026-08-08. Research: docs/superpowers/research/2026-08-08-pl-b7-europe-caucasus.md", "last_updated": "2026-08-08", "confidence": "high (amount) / medium-high (effective-date nuance)", "note": "weekly_amount 106.19 = EUR 460.17/month x 12/52. Covers the age-2 child-raising span (52.1 wk, either-parent pot counted once) + the father's own art. 53g 2-month block (8.7 wk), both at the same ZBDOO flat amount. The Jan-Jul 2026 interim figure was EUR 398.81 (BGN 780 carry-over); the store carries the in-force-at-research-date figure per the mid-year-step ruling. The amount re-sets in each annual ZBDOO - tracked in docs/data-refresh-calendar.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Sofia Sustainable Urban Mobility Plan 2019-2035 (Plan za ustoychiva gradska mobilnost). https://sofiaplan.bg/wp-content/uploads/2022/05/SUMP-21-05-2019.pdf", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Average daily travel time, converted to a work trip", "year": "2019", "note": "CONVERSION, recorded in full: published 64 min per day across 2.67 trips per day gives 23.97 min per all-purpose trip, multiplied by a 1.231 work-purpose uplift derived in the research file, gives 29.5. SUMP fallback, because NSI publishes no travel-time series." }, "non_cash.sick_leave": { "source": "Кодекс на труда (Labour Code) Art. 40 (employer-paid first 3 working days) + Кодекс за социално осигуряване (Social Insurance Code, КСО) Art. 40-42 (NSSI/НОИ cash benefit, обезщетение за временна неработоспособност поради общо заболяване).; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-C.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "ZDDFL art. 19, NRA https://nra.bg/wps/wcm/connect/agency/site/taxes/danak-vurhu-dohodite-na-fizicheski-lica/danachni-oblekcheniya/danachni-oblekcheniq (article text quoted verbatim via a secondary; the NRA host refused the fetch)", "last_updated": "2026-08-11", "confidence": "medium", "note": "Personal contributions to supplementary voluntary pension insurance REDUCE the sum of the annual tax bases, up to 10% of that sum. No absolute cap, so no max key. BASIS TRAP HANDLED: the 10% is of the TAX BASE, not of gross, and the Bulgarian tax base is gross minus mandatory employee social and health contributions (about 13.78% up to the insurable-income ceiling), so 0.10 x (1 - 0.1378) = 0.0862 is the strict floor, rounded to 0.086 against gross. A bare 0.10 on gross would over-deduct. Voluntary health plus life insurance get a separate 10%. Low stakes: Bulgarian PIT is a flat 10%. Currency is EUR after the 1 Jan 2026 adoption but the rule is percentage-only." } }, "lagos": { "tax.filing_basis": { "source": "PwC WWTS - Nigeria: PITA / Nigeria Tax Act assess each individual via the State IRS; there is no joint filing.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Nigeria Tax Act 2025 (effective 2026-01-01): new 6-band progressive PIT 0/15/18/21/23/25% (first ₦800k at 0%, top rate above ₦50M); the Consolidated Relief Allowance was abolished and replaced by a 'rent relief' = lower of ₦500k or 20% of rent (modeled as standard_deduction, renter assumption). PwC/KPMG 2026 tax summaries", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "Contributory Pension Scheme employee share 8% of pensionable pay (basic+housing+transport), uncapped; deductible from the PIT base. NHF (2.5%) is often not withheld in private-sector tech so modeled pension-only. PenCom / Nigeria Tax Act 2025", "last_updated": "2026-05-30", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "Nigeria VAT 7.5%: UNCHANGED for 2026 (the proposed rise to 10%+ was dropped in the legislature; the Nigeria Tax Act keeps 7.5% and zero-rates more essentials). FIRS / NRS", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Lagos Land Use Charge owner-occupied residential 0.076% of assessed value less a 25% owner-occupier relief → effective ~0.057%/yr (2026 LUC schedule, Lagos State)", "last_updated": "2026-05-30", "confidence": "medium" }, "housing._summary": { "source": "Numbeo Lagos 2026 + local listings, professional Lekki/Victoria-Island/Ikoyi tier (down-adjusted from Numbeo's serviced-VI luxury centre figures), NGN/mo monthly-equivalent (Lagos typically demands 1–2 years' rent upfront)", "last_updated": "2026-05-30", "confidence": "low" }, "housing.rent_percentiles": { "source": "NigeriaPropertyCentre listing-count CDF, 2026-08-18 (flats for rent, Lagos; QUOTE PERIOD PER ANNUM uniform - 22/22 sampled price strings 'per annum', zero 'per month', so ratios are computed on yearly quotes (period-invariant) and levels normalized /12; bedrooms and maxprice GET params verified binding by zero-collapse; bare annual rent, service charge separate; NGN): 2BR P25/med/P75 = 4.792M/9.208M/16.652M NGN/yr (= 399,400/767,400/1,387,600 monthly, n=6,114) -> 0.520/1.809; 1BR 0.574/1.687 (n=3,869).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW by 5-9x tolerance on both edges at both slices - the sharpest island/mainland dual market in the program. Applied 0.52/1.81 (nearest; BOTH per-city guard exceptions - p25 floor and p75 bound - never loosened globally). POOL COMPOSITION recorded: the island/peninsula corridor (Lekki 56%, +Ajah/VI/Ikoyi = ~85% of the 2BR pool) dominates while the mainland mass market barely lists (Ikeja 131, Yaba 149, Surulere 80 of 6,117) - underrepresenting the cheap mainland RAISES the measured p25, so the true citywide lower edge is even lower than 0.520; the fix is conservative. 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-africa.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (H). NOTE the stored 2BR median (530,000/mo) sits below the measured pool median (767,400/mo) while stored 1BR (350,000) sits ABOVE measured (275,400) - inverted divergence, a LEVEL question, flagged, out of scope here." }, "healthcare._summary": { "source": "Private HMO family plans (Hygeia / Reliance / Avon): the ICP reality since NHIS/NHIA statutory cover is thin (universal_healthcare: false); insurer 2026 price lists + OOP/dental, NGN", "last_updated": "2026-05-30", "confidence": "medium" }, "childcare._summary": { "source": "Lagos professional-tier private creche/preschool/after-school 2026 (NOT the ₦11M/yr expat international-school tier: that is a separate sensitivity); no public subsidy", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Lagos 2026 grocery + restaurant indices, professional basket (food inflation has been very high ~+580% over 5 yrs), NGN", "last_updated": "2026-05-30", "confidence": "low" }, "transit._summary": { "source": "Lagos monthly transit + private car all-in 2026; fuel subsidy removed 2023 (petrol ~₦950–1,150/L) so car running costs rose sharply; heavy congestion raises fuel burn, NGN", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Lagos leisure + domestic vacation, NGN", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "LOS→US/W.Europe round-trip 2026 (Lagos–JFK / Lagos–London economy, Skyscanner/airline), NGN", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Nigeria has no universal (non-means-tested) government child allowance → modeled 0", "last_updated": "2026-05-30", "confidence": "high" }, "non_cash._summary": { "source": "Nigeria Labour Act: statutory paid annual leave only 6 working days (§18; one of the world's lowest floors; professional tech firms voluntarily offer 20–25, but vacation_days_statutory tracks the statutory minimum per our cross-country convention), ~12 public holidays, maternity 12 weeks at 50% of pay (§54 statutory floor; no federal statutory paternity leave), sick leave up to 12 working days; NHIS coverage thin so universal_healthcare: false", "last_updated": "2026-05-30", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Lagos: NGN 7,976,291, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "NGN 7,976,291 / 1361.41 = USD 5,859; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.0234, R5-rounded down to 0.02. REPLACES 0.12 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. p25 to p90 spans 7.4x, which is two pay systems (naira-local and USD-remote) sharing one bucket; the median describes neither. Page renders 'not enough data for top companies'. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.12; below Bangalore (0.20) and Manila (0.18) — naira depreciation has gutted USD-equivalent local pay. USD-paid remote-for-foreign-employer Lagos engineers are a separate, higher-earning cohort not captured here." }, "non_cash.parental_leave_weeks": { "source": "Labour Act Cap L1 LFN 2004 s.54(1): 6 weeks pre + 6 weeks post confinement, stated in WEEKS (jurist.ng + commonlii black-letter). Batch-8 in-force check 2026-08-08: still the governing statute (ICLG Nigeria 2026-03-20); the July-2026 'labour reform' is a ministerial AGENDA, not an act, and the NSSTF Bill touches social-security acts, not maternity. Research: docs/superpowers/research/2026-08-08-pl-b8-africa.md", "last_updated": "2026-08-08", "confidence": "high", "note": "No private-sector paternity anywhere in the Labour Act: federal civil servants' 14 days, Lagos State's 10 working days and Enugu's are PUBLIC-sector only, excluded. The s.54(1) pay is conditional on six months' continuous service - the work-to-term mover passes at birth, caveat only. WATCH: ILO gave Nigeria a Sept-2026 deadline, so a 2026-27 overhaul is plausible." }, "non_cash.parental_leave_paid_pct": { "source": "Labour Act s.54(1): maternity pay \"not less than fifty per cent of the wages she would have earned\", a statutory FLOOR on actual wages with no ceiling. Batch-8 audit 2026-08-08: no statute raised the floor; 'full pay' claims in HR secondaries conflate Public Service Rules and private bank CBAs (non-statutory). Research: docs/superpowers/research/2026-08-08-pl-b8-africa.md and docs/superpowers/research/2026-08-06-parental-caps-americas-africa.md (Nigeria)", "last_updated": "2026-08-08", "confidence": "high", "note": "50% confirmed on the 2026-08-06 cap sweep as the correct conservative statutory-minimum figure, uncapped at every income level (floor language, not a ceiling). Lagos State's 16wk/100% applies only to Lagos State GOVERNMENT employees, not our private-sector ICP. Refuted: the \"2022 Employees Compensation Act\" hypothesis; the actual ECA is 2010, NSITF workplace-injury only, no maternity provision." }, "non_cash.vacation_days_statutory": { "source": "Labour Act (Nigeria) Cap. L1 s.18 https://lawsofnigeria.placng.org/laws/L1.pdf", "last_updated": "2026-08-08", "confidence": "high", "note": "RE-VERIFIED 2026-08-08, 6 stands. s.18: after twelve months of continuous service a worker is entitled to 'a holiday with full pay of at least six working days' (twelve for workers under sixteen). One of the lowest statutory floors in the dataset and it is correct." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained and deliberately NOT halved, because one-way versus round-trip could not be settled. Three candidates examined and rejected on the record: LAMATA STMP's 105 min is a travel-demand MODEL forecast and public-transport only; Osoba 2015 (2,500 households) publishes bands with no mean; the 'two hours daily' press figure is round-trip journalism. Probably too high, but nothing supports a specific replacement." }, "non_cash.sick_leave": { "source": "Labour Act (Cap. L1, LFN 2004), Section 16.; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-H.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Nigeria Tax Act 2025 s.30(2)(a)(iii) (in force 1 Jan 2026) ; PenCom 2018 Voluntary Contribution Guidelines (one-third of monthly salary, per the Labour Act) ; PwC Worldwide Tax Summaries, Nigeria", "last_updated": "2026-08-11", "confidence": "low", "note": "LOW, and the ceiling is REGULATORY rather than tax. Contributions under the Pension Reform Act are among the eligible deductions, and PwC states mandatory and/or voluntary employee contributions to approved schemes are deductible; the binding limit is PenCom's restriction of voluntary contributions to one-third of monthly salary. No monetary cap, so no max key. UNRESOLVED FORK: several 2026 secondary guides say only the employee's mandatory 8% is deductible, and the verbatim s.30(2) text could not be obtained. The stakes are low (about 2pp of gross at a 10% contribution) and an absent key would be equally defensible; the conservative alternative is { pct_of_earnings: 0.08 }. Note the mandatory 8% runs on basic plus housing plus transport, roughly 5% of gross, not on gross." } }, "nairobi": { "tax.filing_basis": { "source": "PwC WWTS - Kenya (Tax administration): spouses can file separate self-assessment returns (individual assessment).", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "KRA PAYE annual bands 10/25/30/32.5/35%: the 32.5% (KES 6M–9.6M) and 35% (>KES 9.6M) top bands were added by the Finance Act 2023 (effective 2024-07-01); lower bands 10% to KES 288k, 25% to 388k, 30% to 6M. KRA PAYE schedule / PwC 2026 Kenya tax summary", "last_updated": "2026-05-30", "confidence": "high" }, "tax.personal_relief": { "source": "KRA personal relief KES 28,800/yr (KES 2,400/mo): a flat tax CREDIT subtracted from computed PAYE (not a 0% band, not a base deduction), floored at 0. KRA", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "NSSF Act 2013 Tier I+II employee 6% of pensionable pay to the Upper Earnings Limit KES 108,000/mo (annual cap 1,296,000 → max KES 77,760/yr); DEDUCTIBLE from the PAYE base per the Tax Laws (Amendment) Act 2024 (Dec 2024). NSSF / KRA", "last_updated": "2026-05-30", "confidence": "high" }, "tax.medicare_rate": { "source": "SHIF 2.75% (Social Health Insurance Act 2023, replaced NHIF 2024-10) + Affordable Housing Levy 1.5% = 4.25%, both uncapped and DEDUCTIBLE from the PAYE base per the Tax Laws (Amendment) Act 2024: modeled together as medicare_rate. SHA / KRA", "last_updated": "2026-05-30", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Kenya VAT standard rate 16% (VAT Act 2013). KRA", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Nairobi City County land rates ~0.15% of assessed land value (County valuation roll / rating); residential property tax burden is low. Nairobi County", "last_updated": "2026-05-30", "confidence": "low" }, "housing._summary": { "source": "Numbeo Nairobi 2026 + local listings, professional Kilimani/Westlands/Lavington tier, KES/mo", "last_updated": "2026-05-30", "confidence": "low" }, "housing.rent_percentiles": { "source": "BuyRentKenya rank-read percentiles, 2026-08-18 (apartments for rent, Nairobi; the max_price GET param is a TOTAL silent no-op - caught and replaced with a superior instrument: sort=lowest-price + pagination = direct rank reads off the 22-organic-per-page JSON-LD, promoted cards excluded; KES/month, mixed service-charge inclusion recorded): 2BR P25/med/P75 -> 0.692/1.538 (n=1,581); 1BR 0.800/1.600 (n=605).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW (2BR lower edge -0.128 = 2x tolerance, upper +0.298 = 5x; 1BR lower edge holds, upper +0.360 out). Applied 0.69/1.54 (nearest). Pool is the formal Kilimani-Lavington-Westlands corridor; the informal mass market is absent, so the measured spread is a lower bound - conservative. Mixed service-charge inclusion in listed prices recorded as a basis caveat (direction ambiguous, small). 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + rank reads: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-africa.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (H). NOTE stored medians (105k/65k) sit below measured pool medians (130k/75k) - a LEVEL observation, flagged, out of scope here." }, "healthcare._summary": { "source": "Private cover family plans (AAR / Jubilee / Britam): the ICP reality since SHIF statutory cover is thin (universal_healthcare: false); insurer 2026 price lists + OOP/dental, KES", "last_updated": "2026-05-30", "confidence": "medium" }, "childcare._summary": { "source": "Nairobi professional-tier private daycare/kindergarten/after-school 2026 (NOT the expat international-school tier); no public subsidy", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Nairobi 2026 grocery + restaurant indices, professional basket, KES", "last_updated": "2026-05-30", "confidence": "low" }, "transit._summary": { "source": "Nairobi matatu/bus monthly pass + private car all-in 2026; heavy congestion raises fuel burn, KES", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Nairobi leisure + domestic vacation (coast/safari), KES", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "NBO→US/W.Europe round-trip 2026 (Nairobi–JFK / Nairobi–London economy, Skyscanner/airline), KES", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Kenya has no universal (non-means-tested) government child allowance → modeled 0", "last_updated": "2026-05-30", "confidence": "high" }, "non_cash._summary": { "source": "Employment Act 2007: annual leave 21 working days (§28 statutory minimum), ~13 gazetted public holidays, maternity 14 weeks at full pay (§29; paternity 2 weeks), sick leave 7 days full + 7 days half pay (§30); SHIF coverage thin so universal_healthcare: false", "last_updated": "2026-05-30", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Nairobi: KES 3,588,096, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "KES 3,588,096 / 129.52 = USD 27,703; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1108, R5-rounded down to 0.11. REPLACES 0.16 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. p25 to p90 spans 5.8x. The p75 (KES 6.67M) lands exactly on the USD-remote cohort, whose local mid-level counterpart is about KES 1.8M/yr. Two populations, one median. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.16; above Lagos (0.12), below Bangalore (0.20) — USD-paid remote engineers are a separate higher-earning cohort not captured here." }, "non_cash.parental_leave_paid_pct": { "source": "Employment Act 2007 s.29 (\"three months maternity leave with full pay\") and s.29(8) (\"two weeks paternity leave with full pay\"), consolidated text at new.kenyalaw.org, fetched 2026-08-06. Research: docs/superpowers/research/2026-08-06-parental-caps-americas-africa.md (Kenya) and docs/superpowers/research/2026-08-06-parental-caps-southern-east-africa.md", "last_updated": "2026-08-08", "confidence": "high", "note": "100% confirmed uncapped on the 2026-08-06 cap sweep: full pay by primary statute text, no ceiling in either provision; NSSF/SHIF are unrelated deduction schemes (re-confirmed 2026-08-08: SHIF/SHA pays no maternity income replacement; employer bears the full salary). The week-count follow-up flagged here was APPLIED 2026-08-08: weeks 14 -> 15 (see the weeks entry)." }, "non_cash.parental_leave_weeks": { "source": "Employment Act 2007 (CAP 226) s.29(1): 'three months maternity leave with full pay'; s.29(8): 'two weeks paternity leave with full pay'. Consolidated text via Sheriaplex (new.kenyalaw.org timed out - dead fetch logged); 2026-dated Kenyan secondaries corroborate. Research: docs/superpowers/research/2026-08-08-parental-leave-batch6.md", "last_updated": "2026-08-08", "confidence": "high", "note": "15 wk = 13.0 maternity (3 months, house convention; the 90-calendar-day gloss = 12.86, immaterial - statutory wording governs) + 2.0 paternity (insensitive to the calendar/working reading: 14/7 = 10wd/5 = 2). Was 14; this applies the caps-sweep follow-up flag recorded 2026-08-06. No tenure gate anywhere in the Act (7 days' written notice only) - the mover takes full value. The 2021 s.29A pre-adoptive leave is adoption-triggered, excluded. both_parents not set." }, "non_cash.vacation_days_statutory": { "source": "Employment Act 2007 (Kenya) ss. 27 and 28 https://new.kenyalaw.org/akn/ke/act/2007/11/eng@2022-12-31", "last_updated": "2026-08-08", "confidence": "medium-high", "note": "RE-VERIFIED 2026-08-08, 21 stands. s.28(a) grants 'not less than twenty-one WORKING days of leave with full pay' after twelve consecutive months. s.27 requires only one rest day in seven, so Kenya's statutory floor is nominally a six-day week, which raised the question of whether the 21 needed a 5/6 deflation. It does not: s.28 says working days without tying them to a six-day scale, so it is a portable count of days the employee would otherwise have worked, the same construction as Egypt's art. 124 and unlike Morocco's jour ouvrable, which is fixed to a single named rest day and does need converting." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series", "last_updated": "2026-08-08", "confidence": "low", "note": "FLAGGED for a future look: 56x Kenya's national per-capita out-of-pocket mean, the widest multiple in its batch. Directionally right for a private-care professional household, but the multiple itself is not independently corroborated. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "JICA/NIUPLAN Nairobi household interview survey 2013, analysed in Sustainability 11(14):3823. https://www.mdpi.com/2071-1050/11/14/3823", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Mean commute time, household interview survey", "year": "2013", "note": "JICA gives 49 against a World Bank figure of 31. JICA was taken because it is a door-to-door person-trip instrument at metro geography. The gap was tested for a mode-mix artefact by re-weighting JICA's modal times onto World Bank mode shares, which gives 47.9, so the difference is instrument-level rather than sampling. The previous 60 is unsupported by either and is likely Numbeo-derived." }, "non_cash.sick_leave": { "source": "Employment Act 2007, Section 30.; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-H.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "KRA public notice, Amendments to PAYE computation pursuant to the Tax Laws (Amendment) Act 2024 https://www.kra.go.ke/news-center/public-notices/2157-amendments-to-paye-computation-pursuant-to-the-tax-laws-amendment-act,-2024", "last_updated": "2026-08-11", "confidence": "high", "note": "The lowest of actual contributions, 30% of pensionable income, or KES 360,000/yr (\"Kshs. 360,000 per year (Kshs. 30,000 per month)\"), effective 27 Dec 2024, up from KES 240,000. OPEN QUESTION, flagged rather than assumed: whether mandatory NSSF consumes the same ceiling. NSSF is a registered scheme so probably yes, which would cut voluntary headroom further, but neither KRA nor PwC states it. Binds at 10% of KES 3.6m and 6% of KES 6.0m." } }, "cairo": { "tax.filing_basis": { "source": "PwC WWTS - Egypt (Tax administration): individual filing; there is no joint-return regime.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Egypt PIT (Income Tax Law 91/2005 as amended; 2025 schedule) is a progressive 0–27.5% scale, but a 'bracket-withdrawal' rule removes the lower bands as total income rises: above EGP 1.2M only the 25% and top 27.5% bands apply. Modeled for the ICP as a clean 2-band ladder [25% to 1.2M, 27.5% above]. PwC 2026 Egypt tax summary / Egyptian Tax Authority", "last_updated": "2026-05-30", "confidence": "medium" }, "tax.social_security_rate": { "source": "Social insurance employee share 11% of the contribution wage, capped at the maximum insurance salary (~EGP 14,500/mo in 2025 → annual cap ≈ EGP 200,400 → max ≈ EGP 22,044/yr); DEDUCTIBLE from taxable income. Social Insurance Law 148/2019 / NOSI", "last_updated": "2026-05-30", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "Egypt VAT standard rate 14% (VAT Law 67/2016). Egyptian Tax Authority", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Real estate tax is 10% of annual rental value with a generous primary-residence exemption (Real Estate Tax Law 196/2008) → effective burden on an owner-occupied home is very low; modeled ~0.1% of value", "last_updated": "2026-05-30", "confidence": "low" }, "housing._summary": { "source": "Numbeo Cairo 2026 + local listings, professional New Cairo / Maadi / Zamalek tier, EGP/mo", "last_updated": "2026-05-30", "confidence": "low" }, "housing.rent_percentiles": { "source": "Aqarmap listing-count CDF, 2026-08-18 (apartments for rent, /cairo/ = Greater Cairo pool - Cairo gov + Giza-side Haram/Faisal/Mohandessin + New Cairo/6th October/Sheikh Zayed, matching our stored basis; ?room=N verified = bedrooms via card icons and ?maxPrice binding by collapse test, seven other bedroom-param spellings recorded as silent no-ops; EGP/month verified on cards at bottom, middle and top of the distribution, USD compound stock pre-converted): 2BR P25/med/P75 = 3,864/5,002/8,803 (n=7,254) -> 0.772/1.760; 3BR 0.750/1.713 (n=17,048 - Cairo's dominant listed class, the secondary slice; 1BR thin ~1k and serviced-skewed).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW decisively (2BR upper edge +0.52, the starkest dual-market case measured; 3BR corroborates at +0.47; lower edges -0.048 within / -0.07 marginal). Applied 0.77/1.76 (nearest; per-city p75 guard exception). The dual market is explicit in one pool: mass popular-district stock at 2.5-6k EGP against compound/expat stock at 20-125k (top-tail spot check: all genuine New Cairo compound monthly asks, no period/currency artifacts); the verdict survives a furnished-stripping counterfactual (everything above 22k removed still gives 1.42). No furnished filter exists on the portal - mixed pool recorded. Old-rent-law sitting stock and informal rentals are outside the pool: the TRUE spread is wider, the fix conservative. 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-mena.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (I). NOTE the stored 2BR median (22,000) sits at the ~86.5th percentile of the measured Greater-Cairo pool - a formal/expat-segment level basis, flagged LOUDLY as a separate level question, out of scope here." }, "healthcare._summary": { "source": "Private cover family plans (the ICP reality while the new Universal Health Insurance System is still phasing in nationally: universal_healthcare: false); insurer 2026 price lists + OOP/dental, EGP", "last_updated": "2026-05-30", "confidence": "medium" }, "childcare._summary": { "source": "Cairo professional-tier private nursery/KG/after-school 2026 (NOT the expat international-school tier); no public subsidy", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Cairo 2026 grocery + restaurant indices, professional basket, EGP", "last_updated": "2026-05-30", "confidence": "low" }, "transit._summary": { "source": "Cairo metro/bus monthly pass + private car all-in 2026; heavy congestion raises fuel burn, EGP", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Cairo leisure + domestic vacation (Red Sea / North Coast), EGP", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "CAI→US/W.Europe round-trip 2026 (Cairo–JFK / Cairo–London economy, Skyscanner/airline), EGP", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Egypt has no universal (non-means-tested) government child allowance → modeled 0 (Takaful & Karama is means-tested social-protection cash, not for the ICP)", "last_updated": "2026-05-30", "confidence": "high" }, "non_cash._summary": { "source": "Labour Law 14/2025: annual leave 21 days after one year of service (statutory floor), 14 official public holidays, maternity ~4 months (~17 weeks) at full pay via social insurance, sick leave up to 180 days/yr at 75%; the UHIS rollout is incomplete so universal_healthcare: false", "last_updated": "2026-05-30", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Cairo: EGP 502,333, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "EGP 502,333 / 50.84 = USD 9,881; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.0395, R5-rounded down to 0.03. REPLACES 0.13 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Cohort ambiguous: a large USD-paid foreign-remote population is overrepresented, and pooled EGP spans several devaluation regimes. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.13; above Lagos (0.12), below Nairobi (0.16) — USD-paid remote engineers are a separate higher-earning cohort not captured here." }, "non_cash.parental_leave_paid_pct": { "source": "Social Insurance Law art. 77 (state reimburses the employer 75% of the capped insured wage, EGP 16,700/month in 2026) + Labour Law 14/2025 (in force 2025-09-01; employer owes the full wage for the 120-day maternity leave). Research: docs/superpowers/research/2026-08-06-parental-caps-americas-africa.md (Egypt) and docs/superpowers/research/2026-08-06-parental-caps-north-west-africa.md", "last_updated": "2026-08-08", "confidence": "medium-high", "note": "100% confirmed honest for the EMPLOYEE on the 2026-08-06 cap sweep: the insured-wage cap limits only the government-to-employer reimbursement, invisible to the paycheck; the employer statutorily owes full wage. Egypt patterns with Ghana (employer wage continuation), not Morocco (hard state-fund cap). Confidence high on the structure, medium on details (no full primary statute text located). New 1-day paid paternity under Law 14/2025 is immaterial to the week count. Re-verified 2026-08-08 under Labour Law 14/2025: the whole 120-day period is statutorily 'fully paid'; employer wage continuation unchanged, employee-facing pay uncapped (the SI art. 77 ceiling still limits only the state-to-employer reimbursement)." }, "non_cash.parental_leave_weeks": { "source": "Labour Law No. 14 of 2025 (in force 1 Sep 2025, replacing Law 12/2003) art. 54: 120 calendar days fully paid maternity, min 45 days post-delivery, up to 3 uses; plus the law's new 1 paid paternity day on the day of the child's birth. EY global tax alert + ICLG briefing + Africa-HR convergent (Arabic gazette not machine-readable). Research: docs/superpowers/research/2026-08-08-parental-leave-batch6.md", "last_updated": "2026-08-08", "confidence": "high", "note": "17.3 wk = 120/7 = 17.14 maternity + the 1-day paternity (0.2 wk at working-day /5; the day-of-birth calendar reading 0.14 changes nothing at 1 dp: 17.28 vs 17.34, stored 17.3 rounded down). Was 17, the Law 12/2003-era store. The old 10-month service gate is ABOLISHED - entitlement from day one, so the zero-tenure mover takes full value, no caveat. The 2-year childcare leave is unpaid, excluded. both_parents not set (maternity + 1-day paternity, marker-meaningless)." }, "non_cash.vacation_days_statutory": { "source": "Labour Law No. 14 of 2025 art. 124 (in force 2025-09-01, repealing Law 12/2003 in full)", "last_updated": "2026-08-08", "confidence": "high", "note": "CORRECTED 2026-08-08, 21 to 15. THE GOVERNING LAW CHANGED: Labour Law No. 14 of 2025 took effect 2025-09-01 and repealed Law 12/2003, the law our old value came from. Art. 124 sets 15 days in the FIRST year of service, 21 from the second year, 30 at ten years or age 50, and 45 for persons with disabilities. The floor is 15; under the old law 21 was the full-year figure and 15 was a sub-year proration, so the structure inverted. Art. 124 also settles the unit outright by granting the leave 'excluding official holidays, public occasions, and weekly rest days', which makes the figures pure working days needing no conversion. The cleanest statutory read in the region." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. JICA's CREATS study measures exactly our unit door to door (access plus wait plus in-vehicle plus egress); the codebook defining it was located but not the volume publishing the aggregate. Worth flagging: a World Bank Greater Cairo document contains a 2002 projection that the average journey TO AND FROM work would exceed an hour and a half, which is round trip AND a forecast, so if the stored 60 traces there it is doubly wrong." }, "non_cash.sick_leave": { "source": "Labour Law No. 14 of 2025 (establishes the leave RIGHT; replaces the 2003 law,; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-H.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "PwC Worldwide Tax Summaries, Egypt, Individual deductions (reviewed 04 Feb 2026) https://taxsummaries.pwc.com/egypt/individual/deductions", "last_updated": "2026-08-11", "confidence": "medium", "note": "Employee contributions to private insurance funds plus life, health and pension-entitlement premiums, COMBINED, cannot exceed 15% of net revenue or EGP 10,000, whichever is lower. The EGP 10,000 is annual (up from 4,000) and is SHARED with life and health premiums, so real retirement headroom is often nil and 10,000 is already the generous end. MEDIUM and sourced second-best: no readable Egyptian Tax Authority page was found. Reached at 6% of only EGP 167k gross, so it binds essentially always." } }, "casablanca": { "tax.filing_basis": { "source": "PwC WWTS - Morocco (Individual): Moroccan IR has no tax-household concept; spouses' income remains strictly separate, with no joint return or family quotient.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Morocco IR (impôt sur le revenu) 2025 annual scale per the Loi de Finances 2025 (Law 60-24): exempt to MAD 40,000, then 10% / 20% / 30% / 34% and a 37% top rate above MAD 180,000 (the LF 2025 reform raised the exempt band from 30k and cut the top rate from 38%). PwC 2026 Morocco tax summary / BDO 2025 Budget", "last_updated": "2026-05-30", "confidence": "high" }, "tax.standard_deduction_single": { "source": "Professional-expense allowance (frais professionnels): 25% of gross taxable salary for income > MAD 78,000, CAPPED at MAD 35,000/yr; the cap binds for any ICP-level earner, so modeled as a flat MAD 35,000 standard deduction. PwC Morocco (Deductions)", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "CNSS employee contribution 4.48% of gross, CAPPED at the MAD 6,000/month ceiling (annual cap MAD 72,000 → max ≈ MAD 3,226/yr); DEDUCTIBLE from the IR base. CNSS / PwC Morocco", "last_updated": "2026-05-30", "confidence": "high" }, "tax.medicare_rate": { "source": "AMO (assurance maladie obligatoire) employee health contribution 2.26% of gross, UNCAPPED and DEDUCTIBLE from the IR base: modeled in the medicare_rate slot (the uncapped half of the Kenya-style two-stream deduction). PwC Morocco / NajmiNet", "last_updated": "2026-05-30", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Morocco VAT (TVA) standard rate 20%. PwC Morocco", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Communal taxes (taxe d'habitation + taxe de services communaux) on assessed rental value with a 75% owner-occupier abatement → very low effective burden on a primary residence; modeled ~0.1% of value", "last_updated": "2026-05-30", "confidence": "low" }, "housing._summary": { "source": "Numbeo Casablanca 2026 + Mubawab/agenz listings, professional Gauthier/Maârif/Anfa tier, MAD/mo", "last_updated": "2026-05-30", "confidence": "low" }, "housing.rent_percentiles": { "source": "Avito.ma listing-count CDF, 2026-08-18 (appartements a louer, Casablanca prefecture pool; the rooms filter is PROVEN chambres (bedrooms, salon excluded) via a 'studio + 1 chambre independante' ad carrying rooms=1 - no pieces translation needed; ?price=MIN-MAX collapse-tested; meuble/vide mixed, no filter exists; syndic separate; asking MAD/month): 2BR P25/med/P75 -> 0.673/1.516 (n=7,903); 3BR 0.684/1.562 (n=2,695). Robust cuts stripping the bottom ~7% junk/gage/daily ads: 2BR 0.69/1.48-1.49, 3BR 0.69/1.52 - verdict unchanged.", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW on both edges at both slices (2BR -0.147/+0.276; 3BR -0.136/+0.322), robust to cutting the entire junk-suspect bottom. Applied 0.67/1.52 (nearest of the primary measurement; the junk-cut reading 0.69/1.48 is narrower on both edges - the wider primary read is retained per the conservative rule with the junk share recorded, since the cut is a sensitivity bound, not a proven contamination share). Prefecture-only pool (no Mohammedia/Dar Bouazza) and absent medina/informal stock mean the true metro spread is wider - conservative. Mixed meuble/vide pool recorded. 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-mena.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (I). NOTE the stored 2BR median (9,000) sits at roughly the 80-88th percentile of the measured pool (measured median ~4,600 MAD) - a formal/expat-segment level basis, flagged as a separate level question, out of scope here." }, "healthcare._summary": { "source": "Private/employer top-up cover family plans: the ICP reality since public AMO facilities are crowded (universal_healthcare: false); insurer 2026 lists + OOP/dental, MAD", "last_updated": "2026-05-30", "confidence": "medium" }, "childcare._summary": { "source": "Casablanca professional-tier private crèche/maternelle/after-school 2026 (NOT the expat international-school tier); no public subsidy", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Casablanca 2026 grocery + restaurant indices, professional basket, MAD", "last_updated": "2026-05-30", "confidence": "low" }, "transit._summary": { "source": "Casablanca tramway/bus monthly pass + private car all-in 2026 (fuel ~13.8 MAD/L, congestion), MAD", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Casablanca leisure + domestic vacation (coast/Atlas), MAD", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "CMN→US round-trip 2026 (Casablanca–JFK economy, Skyscanner/airline): the transcontinental benchmark; intra-Europe (CMN–Paris ~3,500 MAD) is far cheaper, MAD", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Morocco's CNSS allocations familiales are tied to formal low/mid wages and phase out for a high earner → modeled 0", "last_updated": "2026-05-30", "confidence": "medium" }, "non_cash._summary": { "source": "Labour Code: annual leave 18 working days (Art. 231 statutory minimum, 1.5 days/mo), 13 official public holidays, maternity 14 weeks at 100% via CNSS (paternity 3 days), sick leave via employer then CNSS ~⅔ of capped wage from day 4; private cover is the ICP norm so universal_healthcare: false", "last_updated": "2026-05-30", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Casablanca: MAD 179,426, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "MAD 179,426 / 9.35 = USD 19,190; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.0768, R5-rounded down to 0.07. REPLACES 0.14 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. p25:p75 of only 1.44x is suspiciously tight, a thin-sample signal. Describes the European offshore-delivery cohort. Moroccan job ads usually quote monthly NET, which this is not. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.14; above Cairo (0.13) and Lagos (0.12), below Nairobi (0.16) — USD-paid remote engineers are a separate higher-earning cohort not captured here." }, "non_cash.parental_leave_weeks": { "source": "Code du Travail: maternity 14 weeks (98 days, 7+7), confirmed current in 2026-dated sources (the circulating '16 weeks' is the pathological-complication extension only, refuted) + art. 269 paternity 3 plain 'jours' (no 'ouvrables') = calendar /7 = 0.43 -> +0.4 wk. Total 14.4. Research: docs/superpowers/research/2026-08-08-pl-b8-africa.md", "last_updated": "2026-08-08", "confidence": "high", "note": "The batch-8 audit closes the 2026-08-06 paternity open item: art. 270 makes the 3 days EMPLOYER-paid at the remuneration he would have earned, so it is a tiny uncapped pct stream, NOT part of the CNSS flat shape. The 15-day paternity is loi 30.22 CIVIL SERVANTS only (2023); the private-sector extension remains un-enacted social-dialogue talk as of 2026-08. CNSS gate 54 contribution days in the 10 months before leave: work-to-term mover passes, caveat only. No other paid parental/childcare streams (convergent-negative)." }, "non_cash.parental_leave_paid_pct": { "source": "Code du Travail arts. 269-270: the father's 3 birth-leave days are paid by the EMPLOYER at the remuneration he would have earned (the CNSS MAD ~692 figure caps only the employer's reimbursement, not the employee's pay); maternity sits in parental_leave_flat", "last_updated": "2026-08-08", "confidence": "high", "note": "100 describes the 0.4 paternity week (uncapped employer wage continuation). The 14 maternity weeks stay in the flat shape: the CNSS plafond (~USD 7,800/yr) sits far below any professional salary, so the household always receives the flat capped amount and the old 100% maternity headline never existed for our reader." }, "non_cash.parental_leave_flat": { "source": "CNSS plafond MAD 6,000/month (2026, unrevised in all 2026 sources) -> indemnity capped at MAD 200/day = MAD 1,400/week; employer top-up above the cap is contractual/CBA only, not statutory. Research: docs/superpowers/research/2026-08-08-pl-b8-africa.md and docs/superpowers/research/2026-08-06-parental-caps-north-west-africa.md", "last_updated": "2026-08-08", "confidence": "high", "note": "MAD 1,400/week across the 14 maternity weeks. The 3-day paternity open item is CLOSED by the batch-8 restate: it lives as the 0.4 pct week at 100 (employer-paid), outside this flat block. Plafond revision watch row added to the refresh calendar." }, "non_cash.vacation_days_statutory": { "source": "Code du Travail (Loi 65-99) arts. 231 and 232", "last_updated": "2026-08-08", "confidence": "medium-high", "note": "CORRECTED 2026-08-08, 18 to 15. Art. 231 accrues 1.5 JOURS OUVRABLES per month, so 18 jours ouvrables a year. In French and Maghreb labour law a jour ouvrable is every day except the single weekly rest day and public holidays, a SIX-day unit fixed to Sunday, as distinct from a jour ouvre which follows the worker's actual schedule. So 18 x 5/6 = 15. Confirmed by a second independent route: sources describe the 18 as roughly 21 calendar days, and 21 x 5/7 is also 15. Contrast Egypt art. 124, which excludes whichever rest days the individual actually has and therefore needs no conversion. Art. 232's seniority step-ups are a longer-service tier and stay excluded. Open nuance not acted on: Morocco added Amazigh New Year as an official holiday from 2024 (Decree 2.23.1000), so the stored public_holidays of 13 may be a pre-2024 baseline, but live 2026 sources still span 13, 15 and 17 so it was not provably stale." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. VERIFIED NEGATIVE: Morocco's RGPH 2024 census collects the home-to-work MODE only and no duration, confirmed against the HCP release. Recorded so nobody re-runs this lookup." }, "non_cash.sick_leave": { "source": "Code du Travail (Dahir n° 1-03-194) for leave/notification rules; Dahir portant loi; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-H.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "PwC Worldwide Tax Summaries, Morocco, Individual deductions (reviewed 30 Apr 2026) https://taxsummaries.pwc.com/morocco/individual/deductions (CGI art. 28-III)", "last_updated": "2026-08-11", "confidence": "medium", "note": "Contrat de retraite complementaire (e.g. CIMR) premiums are deductible at 50% of NET TAXABLE SALARY for a salary-only individual, which is our ICP; the 10%-of-global-income variant applies to people with non-salary income. That is a statutory distinction rather than a latitude call, so it was NOT downgraded to 10%. No monetary cap, so no max key. Conditions: contract of at least 8 years, benefits from age 45. Applying 50% to gross rather than net taxable salary is slightly generous but immaterial against a 6-10% contribution, where it never binds. LF 2026 changed the taxation of CIMR pensions IN PAYMENT, which does not touch the contribution ceiling. MEDIUM because it is aggregator-sourced." } }, "accra": { "tax.filing_basis": { "source": "PwC WWTS - Ghana (Tax administration): there is no joint filing for husband and wife.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Ghana PAYE graduated annual scale for resident individuals 2025 (unchanged 2026): tax-free to GHS 5,880, then 5% / 10% / 17.5% / 25% / 30% and a 35% top band above GHS 605,000 (the 35% rate added for high earners). Annualized from the official monthly bands. PwC Worldwide Tax Summaries; Ghana (Individual taxes) / GRA PAYE", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "SSNIT employee contribution 5.5% of basic salary (Tier-1 3.5% + Tier-2 2%), DEDUCTIBLE before PAYE; CAPPED at the maximum insurable earnings ceiling GHS 61,000/month → GHS 732,000/yr (eff. 1 Jan 2025). SSNIT / PwC Ghana", "last_updated": "2026-05-30", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Ghana effective consumer VAT 20% from 1 Jan 2026 (VAT Act 2025, Act 1151): the COVID-19 levy was abolished and NHIL 2.5% + GETFund 2.5% were moved onto the standard 15% VAT base (down from the old cascading ~21.9%). graphic.com.gh / vatabout.com / PwC Ghana", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Accra Metropolitan Assembly property rate on assessed value: low single-digit effective with no clean national figure; modeled ~0.1% of value as a nominal placeholder", "last_updated": "2026-05-30", "confidence": "low" }, "housing._summary": { "source": "Numbeo Accra 2026 + theafricanvestor.com / ownkey.com listings, professional Airport Residential / East Legon / Cantonments tier, GHS/mo", "last_updated": "2026-05-30", "confidence": "low" }, "housing.rent_percentiles": { "source": "meqasa listing-count CDF, 2026-08-18 (apartments for rent, Accra; meqasa normalizes USD listings to GHS at its own ~13.113 rate - derived from its card values - and displays monthly, no yearly quotes in the pool; the fmax filter compares RAW quoted amounts so ~6% day/week serviced quotes leak into low bins - a sensitivity correction stripping them maximally was computed and APPLIED; count integrity verified against dedicated slices after the card feed was caught dropping location scope under querystring filters; GHS/month): 2BR raw 0.441/1.783, max-conservative day/week-stripped 0.473/1.715 (n=1,479) - the CORRECTED pair is applied; 1BR 0.394/1.589 (n=820, serviced-heavy).", "last_updated": "2026-08-18", "confidence": "medium-low", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW by 4-9x tolerance and ROBUST to the maximum day/week correction. Applied the CORRECTED pair 0.47/1.72 rather than the raw 0.441/1.783: the raw lower edge includes a KNOWN contamination artifact (day/week serviced quotes entering low bins at raw amounts), and shipping a value known to include an artifact is not conservatism - the corrected pair still needs BOTH per-city guard exceptions. Heavy USD/serviced stock IS the top tail (composition, recorded); the informal/compound mass market never reaches the portal, so the lower edge remains a lower bound. 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs + sensitivity arithmetic: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-africa.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (H). NOTE stored 2BR (11k) tracks the measured ~12.6k, but stored 1BR (6.5k) sits far below measured ~11.7k (serviced skew: pool 1BR median nearly equals 2BR) - a LEVEL question, flagged, out of scope here." }, "healthcare._summary": { "source": "Private/employer health cover family plans: the ICP reality (public NHIS is thin for professionals, universal_healthcare: false); insurer 2026 lists + OOP/dental, GHS", "last_updated": "2026-05-30", "confidence": "medium" }, "childcare._summary": { "source": "Accra professional-tier private crèche/preschool/after-school 2026 (NOT the expat international-school tier); no public subsidy", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Accra 2026 grocery + restaurant indices, professional basket, GHS", "last_updated": "2026-05-30", "confidence": "low" }, "transit._summary": { "source": "Accra tro-tro/bus monthly spend + private car all-in 2026 (professionals drive; congestion), GHS", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Accra leisure + domestic vacation (coast/Cape Coast/Kumasi), GHS", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "ACC→Western hub round-trip 2026 (Accra–London economy, Skyscanner/airline) as the long-haul benchmark, GHS", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Ghana has no general statutory child cash allowance (LEAP is means-tested for the extreme poor) → modeled 0 for an ICP earner", "last_updated": "2026-05-30", "confidence": "medium" }, "non_cash._summary": { "source": "Labour Act 2003 (Act 651): annual leave 15 working days (§20 statutory minimum after 12 months), 13 public holidays, maternity 12 weeks at 100% employer-paid (§57; →14 weeks if caesarean/multiple); no statutory paid sick-leave minimum (per employer policy); private cover is the ICP norm so universal_healthcare: false", "last_updated": "2026-05-30", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Accra: GHS 92,041, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "GHS 92,041 / 11.69 = USD 7,873; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.0315, R5-rounded down to 0.03. REPLACES 0.13 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Page renders 'not enough data for top locations'; p25 to p90 spans 7.0x, the two-cohort signature. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.13; just below Casablanca (0.14) and Cairo (0.13)-equivalent, above Lagos (0.12) — USD-paid remote engineers are a separate higher-earning cohort not captured here." }, "non_cash.parental_leave_paid_pct": { "source": "Labour Act 2003 (Act 651) s.57: maternity leave with full pay, a direct employer-liability wage continuation; SSNIT has no maternity cash-benefit branch (SSA/ISSA cross-check). Research: docs/superpowers/research/2026-08-06-parental-caps-americas-africa.md (Ghana) and docs/superpowers/research/2026-08-06-parental-caps-north-west-africa.md", "last_updated": "2026-08-06", "confidence": "high", "note": "100% confirmed uncapped on the 2026-08-06 cap sweep: employer-liability structure, no state fund, no insured-wage ceiling in statute or any secondary checked. Watch item: the pending Labour Amendment Bill 2023 (maternity 12 -> 14 wk, 5-day paternity) was NOT enacted as of 2026-08-06; material if it passes." }, "non_cash.parental_leave_weeks": { "source": "Labour Act 2003 (Act 651) s.57, government-hosted PDF (gipc.gov.gh) read verbatim: s.57(1) 'maternity leave of at least twelve weeks', s.57(2) full remuneration. Research: docs/superpowers/research/2026-08-08-pl-b7-africa-south-asia.md", "last_updated": "2026-08-08", "confidence": "high", "note": "12 wk CONFIRMED on the household basis (batch 7): no paternity anywhere in Act 651, and no tenure gate (entitlement conditions only on a medical certificate - the mover takes full value). The Labour Bill 2024 (maternity 14 wk + paternity >=5 days per 24-month window) is NOT enacted as of 2026-08-08: absent from the 2025 passed-bills record and the May-Jul 2026 parliamentary 12-bill list; Paul Hastings (2026-02) says 'anticipated to be passed in 2026 or later' - WATCH. Contamination flag recorded: 'Labour Laws (Amendments) Act No. 4 of 2025' is TANZANIA, not Ghana. Correction to the sweep notes: the abnormal-confinement/multiples +2 wk is s.57(3) (conditional, excluded), not s.57(2)." }, "non_cash.vacation_days_statutory": { "source": "Labour Act 2003 (Act 651) s.20 https://www.gipc.gov.gh/wp-content/uploads/2023/05/LABOUR-ACT-2003-ACT-651.pdf", "last_updated": "2026-08-08", "confidence": "high", "note": "RE-VERIFIED 2026-08-08, 15 stands. s.20: 'In any undertaking every worker is entitled to not less than fifteen working days leave with full pay in any calendar year of continuous service.' A portable working-day count, no conversion needed." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Not reached within the sweep. Leads: Ghana Statistical Service 2021 census and GUMAP." }, "non_cash.sick_leave": { "source": "Labour Act, 2003 (Act 651). Section 24 is titled \"Sick Leave Not Part Of Annual; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-H.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "PwC Worldwide Tax Summaries, Ghana, Other taxes https://taxsummaries.pwc.com/ghana/individual/other-taxes ; https://www.taxlawgh.com/ghana-pensions-retirement-tax (National Pensions Act 2008, Act 766, s.112)", "last_updated": "2026-08-11", "confidence": "medium", "note": "Tier-3 voluntary contributions by contributor, employer or both are exempt up to 16.5% of the contributor's monthly income, to the extent of each party's contribution. Separate from and additional to the mandatory 18.5% tier-1/2. No monetary cap, so no max key: writing max 0 would zero out a real deduction. Stated monthly as a percentage and therefore scale-free. MEDIUM on the base as well as the source: sources split between \"monthly income\" (PwC, taxlawgh, Act 766 wording) and \"basic salary\" (trustee and HR sites), the latter being less favourable; no readable GRA page was found. Does not bind at 6% or 10%." } }, "karachi": { "tax.federal_brackets": { "source": "Pakistan salaried-individual income tax slabs TY 2026-27 (Finance Act 2026, presidential assent 25 Jun 2026, effective 1 Jul 2026): tax-free to PKR 600,000, then 1% (to 1.2M) / 11% (to 2.2M) / 20% (to 3.2M) / 25% (to 4.1M) / 29% (to 5.6M) / 32% (to 7.0M) and a 35% top band above PKR 7,000,000 (cumulative fixed amounts 6,000 / 116,000 / 316,000 / 541,000 / 976,000 / 1,424,000, all reproduced by hand). Mercans statutory alert https://mercans.com/resources/statutory-alerts/pakistan-finance-act-2026-27-1-july-2026/ ; Vialto Partners https://vialtopartners.com/regional-alerts/pakistan-employment-tax-finance-bill-2026-27-summary", "last_updated": "2026-08-07", "confidence": "high", "note": "Finance Act 2026 also ABOLISHED the s.4AB 9% high-income surcharge for salaried individuals (Vialto: top marginal impact 38.15% -> 35%), so tax.surcharge_brackets was REMOVED from this file the same day; the shared highEarnerSurcharge engine seam stays for India. Prior vintage: Finance Act 2025 (TY 2025-26) slabs 0/1/11/23/30/35 with the top band above 4.1M plus the 9% surcharge over 10M. Slabs move with every annual Finance Act (Jul 1) - see docs/data-refresh-calendar.md. Research: docs/superpowers/research/2026-08-07-tax-closeout-asia.md (F3)." }, "tax.social_security_rate": { "source": "Employee payroll modeled 0: EOBI employee contribution is a flat ~PKR 370/month (1% of the minimum wage, negligible) and provincial social security (SESSI) is employer-paid; there is no uncapped employee social tax on salaried professionals, so salaried tax ≈ pure progressive income tax. EOBI / PwC Pakistan", "last_updated": "2026-05-30", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Pakistan federal General Sales Tax on goods 18% (FBR standard rate, Finance Act 2025); Sindh provincial sales tax on services 13–15% (SRB). Modeled 18% as the headline consumption rate. FBR / SRB / PwC Pakistan", "last_updated": "2026-05-30", "confidence": "medium" }, "tax.property_tax_rate_pct": { "source": "Sindh urban immovable property tax is assessed on annual rental value (low effective rate); no clean national % of capital value: modeled ~0.1% as a nominal placeholder", "last_updated": "2026-05-30", "confidence": "low" }, "housing._summary": { "source": "Numbeo Karachi 2026 + Zameen.com listings, professional Clifton / DHA / Gulshan tier, PKR/mo", "last_updated": "2026-05-30", "confidence": "low" }, "housing.rent_percentiles": { "source": "Zameen listing-count CDF, 2026-08-18 (Rentals_Flats_Apartments, Karachi city id 2 - the URL number is the CITY ID and the name is ignored, 'Karachi-3-1' silently serves Islamabad, caught via canonical before measuring; beds_in and price_max verified binding; asking PKR/month, deposits separate): 2BR P25/med/P75 = 49,308/69,118/117,188 (n=1,021) -> 0.713/1.695, robust to round-number mass points (50k/70k/120k reading gives 0.714/1.714); 3BR 0.606/1.483 (n=1,442, secondary per the thin-1BR rule, 1BR n=73).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW at both slices (2BR -0.107/+0.455; 3BR -0.214/+0.243), with a visible bimodal shelf at 90-105k PKR separating the mass market from the DHA/Clifton belt. Applied 0.71/1.70 (nearest; per-city p75 guard exception, never a loosened bound). POOL COMPOSITION recorded: DHA+Clifton = 46.6% of the 2BR pool vs a small share of Karachi's stock - the DHA overweight fattens the measured upper tail for composition reasons, while the missing informal mass market (Korangi, Orangi, Liaquatabad barely list) truncates the lower tail; even the mass-market-internal spread (30-90k around a ~69k median) breaches the stored band on its own, so TOO NARROW stands under every reading. 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-south-asia.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (G). NOTE the stored 2BR median (110,000) sits at the measured p75 - a DHA-basis level, flagged as a separate level question, out of scope here." }, "healthcare._summary": { "source": "Private/employer health cover family plans: the ICP reality (public hospitals thin for professionals, universal_healthcare: false); insurer 2026 lists + OOP/dental, PKR", "last_updated": "2026-05-30", "confidence": "medium" }, "childcare._summary": { "source": "Karachi professional-tier private daycare/Montessori/after-school 2026 (NOT the elite international-school tier); no public subsidy", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Karachi 2026 grocery + restaurant indices, professional basket, PKR", "last_updated": "2026-05-30", "confidence": "low" }, "transit._summary": { "source": "Karachi bus/rickshaw monthly spend + private car all-in 2026 (professionals drive; limited mass transit), PKR", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Karachi leisure + domestic vacation (northern areas/Hunza/coast), PKR", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "KHI→Western hub round-trip 2026 (Karachi–Dubai/London economy, Skyscanner/airline) as the long-haul benchmark, PKR", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Pakistan has no general statutory child cash allowance (BISP/Kafaalat is means-tested for the poor) → modeled 0 for an ICP earner", "last_updated": "2026-05-30", "confidence": "medium" }, "non_cash._summary": { "source": "Sindh Terms of Employment (Standing Orders) Act 2015 + Sindh Maternity Benefit Act 2018: annual leave 14 days, ~14 public holidays, maternity 16 weeks at 100% pay, paid sick leave 16 days/yr at full pay (casual leave 10 days separate); private cover is the ICP norm so universal_healthcare: false", "last_updated": "2026-05-30", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Karachi: PKR 3,155,424, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "PKR 3,155,424 / 277.75 = USD 11,361; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.0454, R5-rounded down to 0.04. REPLACES 0.11 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. levels.fyi rendered a GLOBAL company fallback (the same Meta/CyberArk/Apple trio it showed for Manila and Colombo) and the figure is about 3.4x the domestic median. Karachi and Lahore returned byte-identical numbers. This describes USD-billing remote engineers. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.11; just below Lagos (0.12) and well below Bangalore (0.20)/Manila (0.18) — USD-paid remote/export engineers are a separate higher-earning cohort not captured here." }, "tax.filing_basis": { "source": "PwC Worldwide Tax Summaries: Pakistan, Individual: Tax administration (joint returns are not permitted; each individual files separately). https://taxsummaries.pwc.com/pakistan/individual/tax-administration", "last_updated": "2026-07-26", "note": "PK-2 (closeout 2026-07-26): no joint filing." }, "non_cash.parental_leave_paid_pct": { "source": "Sindh Maternity Benefits Act 2018 (Sindh Act XXXIX of 2018) s.7: full pay, primary gazette text read in full via clr.org.pk mirror. Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (Pakistan) and docs/superpowers/research/2026-08-06-pl-caps-south-asia.md", "last_updated": "2026-08-08", "confidence": "high", "note": "100% confirmed genuinely UNCAPPED on the 2026-08-06 cap sweep, verified against the correct (Sindh provincial, not stale federal) statute, whose 14 sections contain no ceiling and no first-two-children restriction. SESSI's separate contributory scheme caps at wages far below our ICP and is not the operative law for a Karachi professional. No paternity provision in the Act (Sindh government employees' 10 days is executive policy, out of scope). Tenure gate: 1 year of continuous employment (s.5). Re-verified 2026-08-08 against the full primary text: s.7 'the same amount of salary during the period of maternity leave', employer-paid, no ceiling anywhere in ss.1-14." }, "non_cash.parental_leave_weeks": { "source": "Sindh Maternity Benefits Act 2018 (Sindh Act XXXIX of 2018) s.3: four weeks before the expected date of delivery + twelve weeks after = 16 weeks. Primary gazette text read in full via the CLR mirror (clr.org.pk Sindh Maternity Benefits Act 2018 PDF; the official sindhlaws.gov.pk copy is a scanned-image PDF). Research: docs/superpowers/research/2026-08-08-pl-b6-south-asia.md", "last_updated": "2026-08-08", "confidence": "high", "note": "16 wk CONFIRMED on the household basis 2026-08-08 - genuinely the Sindh figure, not a federal import; s.14 gives the Act overriding effect over the 1958 Ordinance. No paternity provision anywhere in the Act (the federal Maternity and Paternity Leave Act 2023 covers federal-control establishments only, not Karachi; the 'Sindh 10-day paternity' claim circulating in search syntheses is REFUTED - its own cited source says the opposite). No children/uses limit in the Act (contrast Bangladesh). Scope: 'employer' = anyone employing 1+ people incl. private organisations - Karachi tech offices covered. LOUD MOVER CAVEAT: s.5 requires one CONTINUOUS YEAR of employment preceding the EXPECTED DELIVERY date - the tightest tenure gate in the corpus: conception within ~the first 3 months in-country, or any job switch during pregnancy, voids the paid benefit entirely (priced caveat-only per the Israel/Lithuania work-to-term precedent, no haircut); s.6 requires written notice 12 weeks pre-EDD with no saving clause. both_parents not set." }, "non_cash.vacation_days_statutory": { "source": "Sindh Shops and Commercial Establishment Act 2015 (Sindh Act XII of 2016) s.14 https://sindhlaws.gov.pk/setup/publications_SindhCode/PUB-NEW-18-000109.pdf; Factories Act 1934 s.49-B", "last_updated": "2026-08-08", "confidence": "medium-high", "note": "CORRECTED 2026-08-08, 14 to 10. Karachi office workers fall under the Sindh Shops and Commercial Establishment Act, whose s.14 grants 14 days after twelve months. The Factories Act s.49-B parallel grants 'fourteen CONSECUTIVE days, inclusive of' the weekly holiday, and Pakistani payroll practice treats both regimes as calendar days inclusive of the weekly rest day. 14 calendar days is 10 five-day weekdays, the same conversion our Argentina value uses for its 14 dias corridos." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. The JICA Karachi Transport Improvement Plan household survey (40,000 households, 2010-11) publishes five duration BANDS by home union council and no metro-wide mean; the stored value sits in the middle bands." }, "non_cash.sick_leave": { "source": "Sindh Factories Act, 2015 (Sindh Act No. XIII of 2016) s.79 \"Casual leave and sick; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-F.md", "source_url": "https://sessi.aqmsolutions.com/procedure-of-payment-of-social-security-contribution/", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "SECP Jamapunji, structure of pension funds and tax credit https://jamapunji.pk/knowledge-center/structure-pension-fund-and-tax-credit ; FBR consolidated Income Tax Ordinance 2001 to 20.02.2026 https://download1.fbr.gov.pk/Docs/2026226162211364IncomeTaxOrdinance2001-Amended-20.02.2026.pdf", "last_updated": "2026-08-11", "confidence": "medium", "note": "No base deduction. S.63 gives a TAX CREDIT at the person's average rate of tax on the lower of the actual contribution or 20% of taxable income, with a +2%/yr uplift for joiners aged 41+. Sources conflict on whether s.63 survived the Finance Act 2022 intact; a Jul 2026 Federal Tax Ombudsman order against FBR's IRIS refusing the credit implies it is live. Either way this field is 0, and we slightly OVERSTATE Pakistani tax by not modeling the credit, which is the conservative direction." } }, "dhaka": { "tax.filing_basis": { "source": "PwC WWTS - Bangladesh (Tax administration): each taxpayer files an individual return; there is no joint-filing mechanism.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Bangladesh NBR individual income tax slabs AY 2025-26 (Finance Act 2024): tax-free to BDT 350,000, then 5% (next 100k) / 10% (next 400k) / 15% (next 500k) / 20% (next 500k) / 25% (next 2,000k) and a 30% top band above BDT 3,850,000. NBR / PwC Worldwide Tax Summaries; Bangladesh (Individual taxes)", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "Employee payroll modeled 0: Bangladesh has no mandatory national social-insurance payroll deduction on salaried employees (provident fund is voluntary/employer-arranged, gratuity is employer-funded), so salaried tax ≈ pure progressive income tax. NBR / PwC Bangladesh", "last_updated": "2026-05-30", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Bangladesh standard VAT 15% (NBR, VAT and Supplementary Duty Act 2012). Modeled 15% as the headline consumption rate. NBR / PwC Bangladesh", "last_updated": "2026-05-30", "confidence": "medium" }, "tax.property_tax_rate_pct": { "source": "Dhaka City Corporation holding tax is assessed at 12% of annual rental value (not capital value); no clean national % of capital value: modeled ~0.1% as a nominal placeholder", "last_updated": "2026-05-30", "confidence": "low" }, "housing._summary": { "source": "Numbeo Dhaka 2026 + bproperty.com listings, professional Gulshan / Banani / Dhanmondi tier, BDT/mo", "last_updated": "2026-05-30", "confidence": "low" }, "housing.rent_percentiles": { "source": "Bproperty PocketBase API full price harvest, 2026-08-18 (api.bproperty.com, reproduces the SERP count within 0.5%; the front-end cities= param is a silent NO-OP and no bedroom param format works on the SERP - the API was the instrument; EXACT quartiles of a full 1,006-price harvest, no interpolation; asking BDT/month, service charge unstructured/separate): 3BR P25/med/P75 -> 0.700/1.400 (n=1,006). 2BR pool n=48 and 1BR ~nil - too thin to measure; 3BR is Dhaka's dominant listed class and the ONLY measurable slice.", "last_updated": "2026-08-18", "confidence": "medium-low", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW on both edges at the only measurable slice (3BR -0.120/+0.160). Applied 0.70/1.40. RECORDED JUDGMENT - 3BR anchor: the pair is measured at 3 beds because Dhaka's portal market lists almost exclusively 3BR (2BR n=48 not evidence-grade, its 0.43/2.71 recorded for color only); elsewhere in the program 3BR upper edges read fatter than 2BR (Istanbul, Tokyo), so the 2BR top may be slightly overstated, but the pool is ~77% Gulshan/Banani/Baridhara expat belt with Mirpur at 0.3% - the mass market is absent, the true citywide spread is far wider, and the applied pair remains conservative. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + harvest: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-south-asia.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (G). Level observations out of scope." }, "healthcare._summary": { "source": "Private/employer health cover family plans: the ICP reality (public facilities thin for professionals, universal_healthcare: false); insurer 2026 lists + OOP/dental, BDT", "last_updated": "2026-05-30", "confidence": "medium" }, "childcare._summary": { "source": "Dhaka professional-tier private daycare/preschool/after-school 2026 (NOT the elite international-school tier); no public subsidy", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Dhaka 2026 grocery + restaurant indices, professional basket, BDT", "last_updated": "2026-05-30", "confidence": "low" }, "transit._summary": { "source": "Dhaka bus/rickshaw/CNG monthly spend + private car all-in 2026 (professionals drive; metro line-6 partial), BDT", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Dhaka leisure + domestic vacation (Cox's Bazar/Sylhet), BDT", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "DAC→Western hub round-trip 2026 (Dhaka–Dubai economy, Skyscanner/airline) as the long-haul benchmark, BDT", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Bangladesh has no general statutory child cash allowance (social-safety-net stipends are means-tested for the extreme poor) → modeled 0 for an ICP earner", "last_updated": "2026-05-30", "confidence": "medium" }, "non_cash._summary": { "source": "Bangladesh Labour Act 2006: earned/annual leave ~17 days (1 day per 18 worked), ~14 public/government holidays, maternity 16 weeks at 100% pay (sec. 46), paid sick leave 14 days/yr at full wages (sec. 116; casual leave 10 days separate); private cover is the ICP norm so universal_healthcare: false", "last_updated": "2026-05-30", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Dhaka: BDT 1,024,386, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "BDT 1,024,386 / 123.42 = USD 8,300; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.0332, R5-rounded down to 0.03. REPLACES 0.09 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Small but real local sample, about 1.4 to 1.7x the domestic median. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.09; below Karachi (0.11) and Lagos (0.12) — likely the lowest in the set. USD-paid remote/export engineers are a separate higher-earning cohort not captured here." }, "non_cash.parental_leave_paid_pct": { "source": "Bangladesh Labour Act 2006 ss.46-48: 16 weeks maternity (8+8) at full average wage, employer-paid cash. Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (Bangladesh) and docs/superpowers/research/2026-08-06-pl-caps-south-asia.md", "last_updated": "2026-08-08", "confidence": "high", "note": "100% confirmed genuinely UNCAPPED on the 2026-08-06 cap sweep: employer-liability wage continuation, no insured-wage ceiling exists (no social-insurance maternity scheme to attach one to). No statutory paternity (tech/MNC offerings are company policy, excluded). Eligibility note: the entitlement applies to the first two children, the general case stored. Re-verified 2026-08-08: the 2025 Ordinance / 2026 Act changed only the duration; s.48 full average wage, employer-paid cash, still uncapped (no SI maternity scheme exists to attach a ceiling to)." }, "non_cash.parental_leave_weeks": { "source": "Bangladesh Labour Act 2006 s.46 as amended by the Labour (Amendment) Ordinance 2025 (promulgated 2025-11-17), converted to an Act by parliament 2026-04-09 (Extraordinary Gazette): maternity 112 -> 120 days (60 pre + 60 post). Convergent post-enactment sources: businessinbangladesh.com.bd, Global Law Experts employers' guide, bdlawdigest.org, Jural Acuity. bdlaws.minlaw.gov.bd ECONNREFUSED (dead fetch logged). Research: docs/superpowers/research/2026-08-08-pl-b6-south-asia.md", "last_updated": "2026-08-08", "confidence": "high (5 convergent post-enactment sources; primary gazette host unreachable)", "note": "17.14 wk = 120/7. Was 16 (112 days), the pre-Ordinance figure. The 2023 Amendment Bill carrying the same change was RETURNED by the President under Art. 80(3) and never became law - the operative instrument is the 2025 Ordinance / 2026 Act (refutation recorded so the 2023 timeline is not re-derived). Gates SURVIVE the amendment: benefit requires 6 months with the CURRENT employer immediately preceding delivery (the ~9-12-month mover passes even conceiving immediately; resets on a job switch mid-pregnancy) and is void with 2+ surviving children. No statutory paternity through 2026 (tech/MNC offerings are company policy). both_parents not set." }, "non_cash.vacation_days_statutory": { "source": "Bangladesh Labour Act 2006 (Act XLII of 2006) ss. 103, 117, 118 https://www.ilo.org/media/43266/download", "last_updated": "2026-08-08", "confidence": "medium", "note": "CORRECTED 2026-08-08, 17 to 15. s.117(1) is an accrual formula, not a flat grant: one day of leave per eighteen days of work performed. The annual figure therefore depends on s.103's weekly holiday, which differs by establishment type. A commercial establishment, the category covering our office-worker ICP under s.2(xli), gets 1.5 weekly holidays, so 365 less 78 less 11 festival days is 276 days worked and 276/18 is 15. A FACTORY gets one weekly holiday, giving 302 days worked and 17, which reproduces our old stored value exactly and is strong evidence it was computed on the wrong category. Practitioner guides quote 18 to 20 by skipping the holiday subtraction; that is the most favourable reading and was not adopted. A defensible alternative that leaves festival days in the denominator gives 16." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained as a conservative upper bound, and flagged for rework. The RSTP/DHUTS household survey publishes a MEDIAN DAILY all-trips figure of 60 min, which implies roughly 30 one way, and an IGC 2012 study measured 47 min one way for car commuters. Both anchors sit below 60, so this value is probably high." }, "non_cash.sick_leave": { "source": "Bangladesh Labour Act, 2006 (Act No. 42 of 2006, as amended through 2018), Chapter; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-F.md", "source_url": "https://www.ilo.org/media/43266/download", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Income Tax Act 2023, Sixth Schedule Part 3 ; NBR Income Tax at a Glance https://nbr.gov.bd/uploads/publications/107.pdf ; PwC Worldwide Tax Summaries, Bangladesh (rev. 31 Jul 2026) https://taxsummaries.pwc.com/bangladesh/individual/deductions", "last_updated": "2026-08-11", "confidence": "medium", "note": "CONFLICTED, recorded rather than resolved. The Sixth Schedule Part 3 lists self and employer RPF contributions, DPS (capped BDT 120,000) and life premiums as ALLOWABLE INVESTMENT, and the relief is a REBATE, the lowest of 3% of taxable income, 10% of allowable investment, or BDT 750,000. A rebate cuts tax payable, not the base. But PwC's deductions page says verbatim \"Taxpayer's and employer's contribution to a recognised provident fund is deductible\", which reads as loose phrasing for allowable investment since the same page sets out the rebate formula, and it could not be PROVEN because the NBR PDF would not extract. Conservative rule applied." } }, "colombo": { "tax.filing_basis": { "source": "PwC WWTS - Sri Lanka; Inland Revenue Act: the taxable income of each person is determined separately (APIT per person); no joint filing.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Sri Lanka APIT (Advance Personal Income Tax) AY 2025/26 per the Inland Revenue (Amendment) Act No. 2 of 2025: first LKR 1,800,000 tax-free, then 6% (next 1,000k) / 18% (next 500k) / 24% (next 500k) / 30% (next 500k) and a 36% top band above LKR 4,300,000 of taxable income. Inland Revenue Department Sri Lanka / PwC Worldwide Tax Summaries; Sri Lanka (Individual taxes)", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "Employee EPF (Employees' Provident Fund) contribution 8% of gross, deductible before APIT and uncapped (no wage ceiling); ETF 3% is employer-only and excluded. EPF Act No. 15 of 1958 / ETF Act No. 46 of 1980; PwC Sri Lanka", "last_updated": "2026-05-30", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Sri Lanka standard VAT 18% (raised to 18% effective 1 Jan 2024, VAT Act / Inland Revenue Department). Modeled 18% as the headline consumption rate", "last_updated": "2026-05-30", "confidence": "medium" }, "tax.property_tax_rate_pct": { "source": "Sri Lanka has no recurrent national property tax; local-authority rates are assessed on annual rental value, not capital value: modeled ~0.1% as a nominal placeholder", "last_updated": "2026-05-30", "confidence": "low" }, "housing._summary": { "source": "Numbeo Colombo 2026 + lankapropertyweb.com listings, professional Colombo 3 / 5 / 7 tier, LKR/mo", "last_updated": "2026-05-30", "confidence": "low" }, "housing.rent_percentiles": { "source": "ikman.lk listing-count CDF, 2026-08-18 (apartments for rent, Colombo district, enum.bedrooms and money.price.maximum verified binding - the bare ?bedrooms=2 form perturbs the query without filtering and was avoided; paginationData total; all prices 'Rs N /month' LKR, no per-day or USD contamination in the low-tail spot check; featured ads ignore filters but counts do not): 2BR P25/med/P75 = 208,523/320,000/509,574 (n=978) -> 0.652/1.593; 3BR 0.764/1.475 (n=1,222, secondary; 1BR n=104 too thin).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW at the primary slice on both edges (2BR -0.168/+0.353; 3BR corroborates the top at +0.235). Applied 0.65/1.59 (nearest; sits just inside the 1.6 global guard, no exception needed). POOL COMPOSITION recorded: Colombo-2 luxury towers alone are 27% of the district 2BR pool (city core colombo-2..8 = 57.5%); the mass Sri Lankan rental market (annexes, unfurnished walk-ups) mostly never reaches the portal - the tower belt fattens the top for composition reasons while the truncated bottom makes the lower edge a lower bound; both edges are out by multiples of tolerance under every reading. Geography: Colombo DISTRICT scope (incl. Rajagiriya, Dehiwala), recorded. 2BR anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-south-asia.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (G). NOTE the stored 2BR median (200,000) sits well below the tower-heavy pool median (320,000) - pool-basis divergence, a LEVEL question, flagged, out of scope here." }, "healthcare._summary": { "source": "Private/employer health cover family plans: the ICP reality (public care free but thin for professionals, universal_healthcare: false); insurer 2026 lists + OOP/dental, LKR", "last_updated": "2026-05-30", "confidence": "medium" }, "childcare._summary": { "source": "Colombo professional-tier private daycare/preschool/after-school 2026 (NOT the elite international-school tier); no public subsidy", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Colombo 2026 grocery + restaurant indices, professional basket, LKR", "last_updated": "2026-05-30", "confidence": "low" }, "transit._summary": { "source": "Colombo bus/rail monthly spend + private car all-in 2026 (professionals drive), LKR", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Colombo leisure + domestic vacation (Kandy/Galle/south-coast beaches), LKR", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "CMB→Western hub round-trip 2026 (Colombo–Dubai/London economy, Skyscanner/airline) as the long-haul benchmark, LKR", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Sri Lanka has no general statutory child cash allowance (Samurdhi/Aswesuma welfare is means-tested for low-income households) → modeled 0 for an ICP earner", "last_updated": "2026-05-30", "confidence": "medium" }, "non_cash._summary": { "source": "Shop and Office Employees Act: 14 days annual leave (from the 2nd year of service), ~18 public/bank/mercantile holidays, maternity 12 weeks at 100% (first two confinements), 7 days casual leave that also covers illness (no separate statutory sick-leave entitlement); private cover is the ICP norm so universal_healthcare: false", "last_updated": "2026-05-30", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Colombo: LKR 4,065,590, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "LKR 4,065,590 / 335.42 = USD 12,121; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.0485, R5-rounded down to 0.04. REPLACES 0.1 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Global-company fallback shown (Meta and Apple, neither present in Sri Lanka) and about 2.0x the domestic median. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.10; just above Dhaka (0.09), below Karachi (0.11) — Sri Lanka's higher GDP/capita and established IT-export sector lift it above Bangladesh, but the post-2022 LKR collapse compressed USD-denominated local pay. USD-paid remote/export engineers are a separate higher-earning cohort not captured here." }, "non_cash.universal_healthcare": { "source": "National Health Policy 2020-2025 (announced 2019-12-16); NCBI Health-For-All case study", "last_updated": "2026-08-05", "confidence": "high", "note": "FALSE under the ICP-facing definition, with a nuance the flat phrasing loses: Sri Lankan public care is free for citizens (no premium, no point-of-use fees), but the entitlement is restricted by nationality and does not reach a relocating foreign professional. Any reader-facing copy should say \"universal free care for citizens, not extended to foreign nationals\" rather than \"no universal healthcare\". Adjudicated 2026-08-04, flag:01785706582691-7a970f1f." }, "non_cash.parental_leave_paid_pct": { "source": "Shop and Office Employees Act (Cap. 129) s.18B as amended by Act 14 of 2018 (primary gazette PDF read: 70 working days live-birth leave, the old 3rd-child reduction REPEALED by the same Act); srilankalaw.lk s.18B title. Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (Sri Lanka) and docs/superpowers/research/2026-08-06-pl-caps-south-asia.md", "last_updated": "2026-08-06", "confidence": "medium-high", "note": "100% confirmed uncapped on the 2026-08-06 cap sweep: employer-liability remuneration continuation under the ICP-relevant Shop and Office Act, no ceiling found (the 6/7-rate structure belongs to the separate factory-worker Maternity Benefits Ordinance, not this statute). Bonus refutation: the suspected third-child reduction was repealed by Act 14/2018; do not re-import it. Paternity is public-sector only, excluded. Confidence medium-high: the exact s.18B(1) pay-rate clause is paywalled, corroborated by secondaries." }, "non_cash.parental_leave_weeks": { "source": "Shop and Office Employees Act (Cap. 129) Part IA per the 2024 consolidation (lankalaw.net PDF, current through Act 28 of 2024; Part IA last touched by Act 14 of 2018) + Amendment Act 14 of 2018 read verbatim (eohfs.health.gov.lk government-hosted PDF): s.18B(1) up to 14 paid pre-confinement days + s.18B(2) 70 days from confinement (all live births; child-count limit repealed 2018) + s.18B(3) carryover of worked pre-days; s.18C full remuneration both phases; s.18H leave is in addition to weekly/Poya/statutory holidays. Research: docs/superpowers/research/2026-08-08-pl-b7-africa-south-asia.md", "last_updated": "2026-08-08", "confidence": "high (84-working-day total) / medium-high (/5 conversion)", "note": "16.8 wk = 84 working days / 5 (house convention). Was 12, a calendar gloss of the same 84; the sweep's '70 working days' was only the s.18B(2) post-confinement phase. 'Working days' is not literally in s.18B - the working-day computation flows from s.18H (leave additional to all holidays) and administration is unanimous (84 working days excluding weekly, Poya and statutory holidays; De Saram 2026-05-25, Ceylon Public Affairs 2026-02). Floor alternatives recorded in the cluster file: 16.0 (mixed), 14.0 (post-phase only), 12.0 (pure calendar, the stale store). No tenure gate anywhere in Part IA; no private-sector paternity (public-sector 3 days excluded). The draft Employment Act consolidation is NOT enacted (the government pivoted to four separate bills, still in consultation) - WATCH." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. The JICA CoMTrans household survey exists but its published headline is travel SPEED (13.7 km/h), not duration." }, "non_cash.sick_leave": { "source": "Shop and Office Employees (Regulation of Employment and Remuneration) Act, No. 19; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-F.md", "source_url": "https://www.labourdept.gov.lk/downloads/labour_code/20.pdf", "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Inland Revenue Act No. 24 of 2017 s.52 and the Fifth Schedule https://www.ird.gov.lk/en/publications/Acts_Income%20Tax_2017/IR_Act_No_24_2017_E.pdf ; Amendment Act No. 2 of 2025", "last_updated": "2026-08-11", "confidence": "medium", "note": "No qualifying payment covers voluntary retirement saving. An individual deducts only personal relief (raised to LKR 1,800,000/yr from 1 Apr 2025) and qualifying payments, and qualifying payments for individuals are cash donations to approved charities capped at the lower of a third of taxable income or LKR 75,000. EPF/ETF are mandatory. MEDIUM because the IRD PDFs would not extract cleanly and PwC's Sri Lanka deductions URL 404s. SEPARATE OBSERVATION, not this field: several Sri Lankan payroll sources compute APIT after deducting the employee 8% EPF, which the engine does not do." } }, "doha": { "tax.federal_brackets": { "source": "Qatar levies NO personal income tax on employment income (nationals or expats); no municipal/sub-national income tax. Law No. 24 of 2018 taxes only business/commercial profit, not wages", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "Qatar social insurance (Law No. 1 of 2022) applies to Qatari nationals only; non-Qatari expats (the ICP) have 0% employee social-insurance contribution", "last_updated": "2026-05-30", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Qatar has announced but NOT yet implemented VAT: no consumption tax is in force as of 2026 (unlike UAE 5% / Saudi 15%) → modeled 0%", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "No annual/recurrent property tax in Qatar on owner-occupied residential; registration fee applies at purchase only → annual rate 0", "last_updated": "2026-05-30", "confidence": "high" }, "housing._summary": { "source": "Numbeo Doha 2026 + Property Finder QA (West Bay / The Pearl / Lusail professional districts), QAR/mo", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Mandatory private health insurance for residence permits (expats not in the public Hamad system for full primary cover); OOP/dental Pacific Prime + Numbeo Doha 2025-26", "last_updated": "2026-05-30", "confidence": "medium" }, "childcare._summary": { "source": "Doha private nursery / international preschool 2025-26 (Numbeo + international-school DB, mid-tier); no public subsidy for expats", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Doha 2026 (upper-middle basket) + Wise, QAR", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "Doha Metro monthly pass + private car (low fuel cost) 2025-26, QAR", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Doha leisure + regional/domestic vacation", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "DOH->US/Europe economy long-haul round-trip 2025-26 (Qatar Airways hub), QAR", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No cash child allowance for expats (family/social allowances are Qatari-nationals only) → 0", "last_updated": "2026-05-30", "confidence": "high" }, "non_cash._summary": { "source": "Qatar Labour Law No. 14 of 2004: 21-day statutory minimum paid annual leave (rising to 28 after 5 yrs), ~9 paid public holidays, ~7 weeks maternity at 100%, sick leave Art. 82 (2 weeks full / 4 weeks half / further unpaid, max 12 weeks/yr). No public universal healthcare for expats", "last_updated": "2026-05-30", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Doha: QAR 218,580, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "QAR 218,580 / 3.64 = USD 60,049; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2402, R5-rounded down to 0.24. REPLACES 0.62 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. No income tax. Allowance inclusion unknown, and the p25 is exactly one third of the median, a small-n artefact. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.62; 0% income tax + QAR-USD peg lift the USD-equivalent index; package band set below Riyadh (0.70) and Dubai (0.85), reflecting Doha's smaller tech employer base." }, "non_cash.parental_leave_weeks": { "source": "Qatar Labour Law 14/2004 art. 96: 50 calendar days maternity -> 50/7 = 7.14 -> 7.1 (house 1dp; the stored 7 predated the convention). Batch-8 basis audit 2026-08-08: no private-sector Labour Law amendment 2025-2026 (Law 25/2025 amends the PUBLIC-sector Civil HR Law only, excluded); no statutory private-sector paternity (confirmed); nursing hour (art. 97) = working-time reduction, excluded. Research: docs/superpowers/research/2026-08-08-pl-b8-gulf.md", "last_updated": "2026-08-08", "confidence": "high", "note": "Art. 96 verbatim: a worker 'who has spent a whole year in service with the Employer' is entitled to the 50 paid days. Priced CAVEAT-ONLY for the work-to-term mover (~12 months at delivery, HK/Israel precedent); LOUD caveat: the statute does not fix the measurement instant, and a conceive-on-arrival mover starting leave up to 15 days pre-natal sits at ~11.5 months - under 1 year the entitlement is zero days, not a reduced rate." }, "non_cash.parental_leave_paid_pct": { "source": "Qatar Labour Law 14/2004 arts. 96-98: 50 days maternity at full pay, employer-paid (almeezan.qa TLS-failed again 2026-08-08; cercli.com fetched directly + 5 convergent secondaries at the sweep). Batch-8 audit 2026-08-08 re-confirmed 100 uncapped and current. Research: docs/superpowers/research/2026-08-08-pl-b8-gulf.md and docs/superpowers/research/2026-08-06-pl-caps-gulf.md", "last_updated": "2026-08-08", "confidence": "high", "note": "100% confirmed genuinely UNCAPPED on the 2026-08-06 cap sweep: direct employer wage continuation, no social-insurance maternity scheme reaches an expat. Two caveats recorded: (1) a BINARY tenure cliff, not a wage cap: under 1 year of service the statutory entitlement is zero days, not a reduced rate; (2) no statutory private-sector paternity (employer discretion only, excluded). The 2025 Civil HR Law maternity upgrade is PUBLIC-sector only and does not touch Labour Law 14/2004." }, "non_cash.vacation_days_statutory": { "source": "Labour Law No. 14 of 2004 art. 79", "last_updated": "2026-08-08", "confidence": "medium-high", "note": "CORRECTED 2026-08-08, 21 to 15. Art. 79 denominates the entitlement in WEEKS, not days: not less than three weeks under five years' service and four weeks at five years or more. Three weeks is 15 five-day weekdays. The stored 21 came from reading three weeks as 21 days, which is the calendar span rather than the working days inside it. Cleaner than the Saudi and Kuwaiti cases because the statute never states a bare day count. Public holidays sit in a separate article and stay additive." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Gulf mandatory employer-provided health cover, with the region's typical published co-payment and annual co-payment cap structure", "last_updated": "2026-08-08", "confidence": "medium", "note": "CHECKED for the premium-as-out-of-pocket error and CLEAR: the value is co-pay-sized. Dependant coverage is a premium and is not booked here. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Gulf mandatory employer-provided health cover, with the region's typical published co-payment and annual co-payment cap structure", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Not reached within the sweep. Lead: the Qatar Transportation Master Plan." }, "non_cash.sick_leave": { "source": "Law No. 14 of 2004 (Labour Law), the sick-leave articles.; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-H.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "PwC Worldwide Tax Summaries, Qatar (reviewed 26 Feb 2026): \"Income tax is not imposed on employed individuals' salaries, wages, and allowances\"", "last_updated": "2026-08-11", "confidence": "high", "note": "Recorded for explicitness only; it changes no output by construction. Qatar dispatches to the same zero-return function as the UAE, which never reads taxable_income." } }, "kuwait-city": { "tax.federal_brackets": { "source": "Kuwait levies NO personal income tax on employment income (nationals or expats); salaries are received gross = net. PwC Worldwide Tax Summaries: Kuwait Individual", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "PIFSS social insurance applies to Kuwaiti (and partially GCC) nationals only; non-GCC expats (the ICP) make no mandatory contribution (they receive an end-of-service gratuity instead) → employee rate 0. PwC; Rivermate", "last_updated": "2026-05-30", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Kuwait signed the 2017 GCC Unified VAT Agreement (framework 5%) but has NOT implemented VAT; government has ruled it out near-term → no broad consumption tax in force, modeled 0% (vs UAE 5% / Saudi 15%). vatcalc.com; fiscalsolutions.co.uk", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Kuwait has no annual/recurrent property tax (and no transfer/stamp duty on residential) → annual rate 0. RSM Kuwait Tax Guide 2025; PwC", "last_updated": "2026-05-30", "confidence": "high" }, "housing._summary": { "source": "Numbeo Kuwait City 2026 (Salmiya / Sharq / Salwa professional districts); per-m² city-centre ~KWD 1,633 for purchase reference (expats generally cannot own freehold)", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Mandatory expat health scheme (Afya, raised to KWD 100/person/yr in Dec 2025) bundled with a private comprehensive top-up by family size; OOP/dental Numbeo + KhaleejAdvisor 2025-26", "last_updated": "2026-05-30", "confidence": "medium" }, "childcare._summary": { "source": "Kuwait City private nursery / preschool / after-school 2025-26 (Numbeo + Edarabia mid-tier); no public childcare subsidy for expats", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Kuwait City 2026 grocery + restaurant indices (professional basket), KWD", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "Kuwait is heavily car-dependent (minimal public transit); fuel is extremely cheap (~KWD 0.085-0.11/L) so the car line is dominated by depreciation/insurance/maintenance. Numbeo 2025-26", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Kuwait City leisure + regional vacation (limited domestic tourism → short Gulf trips)", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "KWI->US/Europe economy long-haul round-trip 2025-26 (Kuwait Airways/Gulf carriers), KWD", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No cash child/family allowance for expats (family allowances are Kuwaiti-nationals only; expats pay added dependent residency/health fees) → 0", "last_updated": "2026-05-30", "confidence": "high" }, "non_cash._summary": { "source": "Kuwait Private-Sector Labour Law No. 6 of 2010: 30-day statutory annual leave (calendar days; ~22 working days; stored 30 to match the UAE/Dubai 30-day-statute convention), ~13 paid public holidays, 70-day maternity at 100% (~10 weeks), sick leave Art. 69 (15d full / 10d 75% / 10d 50% / 10d 25% / 30d unpaid, max 75d). No public universal healthcare for expats", "last_updated": "2026-05-30", "confidence": "high" }, "compensation_index": { "source": "Derived, not observed, and not measured at country level either: no usable all-levels figure exists anywhere in Kuwait. Prior source: Levels.fyi/Glassdoor Kuwait City tech 2025-26 (senior IC ~KWD 30k-48k/yr); thin tech employer base (banks/telcos/oil-IT, few product firms)", "last_updated": "2026-08-11", "confidence": "low", "note": "0.65 x 0.6000 = 0.3900, R5-rounded down to 0.39. 0.6000 is the median of (re-derived / previous) across the 147 cities worldwide that DO carry an all-levels measurement, i.e. the measured size of the move from senior-over-196k-to-220k onto all-levels-over-250k. THIS IS THE WEAKEST CLASS OF VALUE IN THE COLUMN: it asserts only that Kuwait City now sits on the same basis as everything else, and says nothing about whether its magnitude was ever right. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.65; 0% income tax + the strong KWD peg lift the USD-equivalent index above its nominal-salary rank, but the thin tech demand side keeps it just above Doha (0.62) and below Riyadh (0.70)/Dubai (0.85)." }, "non_cash.parental_leave_paid_pct": { "source": "Kuwait Labour Law 6/2010 art. 24: 70 days maternity (30 pre + 40 post), ALL at full pay, employer-paid (commoner-law.com fetched directly: \"there is no split between full and half pay\"; manpower.gov.kw unreachable). Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (Kuwait) and docs/superpowers/research/2026-08-06-pl-caps-gulf.md", "last_updated": "2026-08-06", "confidence": "high", "note": "100% confirmed genuinely UNCAPPED on the 2026-08-06 cap sweep, and the suspected full/half-pay split REFUTED: all 70 days pay full wage (the tiering pattern belongs to Kuwaiti sick leave, art. 69, not maternity). No tenure gate, no statutory paternity (public-sector 3 days is civil-service policy, out of scope), PIFSS does not cover non-GCC expats. The 4 months of extended childcare leave are unpaid, excluded." }, "non_cash.parental_leave_weeks": { "source": "Kuwait Labour Law 6/2010 art. 24 (70 days paid maternity, gloss 30 pre + 40 post, birth within the period), convergently quoted by 2026-verified guides (Commoner Law, verified 2026-07-18; HLB HAMT, updated 2026-06-26); manpower.gov.kw unreachable 2026-08-08 and three official-translation mirrors dead (logged). Research: docs/superpowers/research/2026-08-08-pl-b7-east-sea-gulf.md", "last_updated": "2026-08-08", "confidence": "high", "note": "10.0 wk = 70 consecutive calendar days /7, all at full employer-paid wage. CONFIRMED on the household basis (batch 7): no statutory private-sector paternity (explicitly reconfirmed 2026), the up-to-4-month childcare extension is unpaid (excluded), and there is NO tenure gate - truein's 'one year of service' claim REJECTED (no statutory basis; art. 24 conditions only on delivery within the period, and both 2026 guides carry no service condition). No 2024-2026 amendment to art. 24; the only live labour draft raises ANNUAL leave 30 -> 35 days (first reading only, unrelated) - watch item." }, "non_cash.vacation_days_statutory": { "source": "Private Sector Labour Law No. 6 of 2010 art. 70, Public Authority for Manpower https://www.manpower.gov.kw/docs/LaborLaw/Labor_Law_Eng.pdf", "last_updated": "2026-08-08", "confidence": "medium", "note": "CORRECTED 2026-08-08, 30 to 22. Art. 70 grants 30 days and names only two exclusions from the count, official holidays and sick leave, which leaves the ordinary weekly rest day absorbed and makes the 30 a CALENDAR figure. Conversion convention documented in docs/superpowers/research/2026-08-08-vacation-days-non-us-sweep.md: a calendar-day entitlement taken as a contiguous block converts to the Mon-Fri days inside its span, so 30 calendar days is 22 weekdays and 21 calendar days is 15. The same rule already governs our Spain, Brazil, Peru, Argentina, Ecuador, Estonia and Azerbaijan values. Converted identically to the UAE because the statutory phrase is the same; splitting the two would be an artefact of who researched them. Watch item: a first-reading draft amendment to raise art. 70 from 30 to 35 days was still unenacted at 2026-08-08, and if it passes the value becomes 25." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Gulf mandatory employer-provided health cover, peer-consistent with the sourced UAE and Saudi co-payment bands", "last_updated": "2026-08-08", "confidence": "low", "note": "Peer-consistent with the sourced Gulf co-payment logic but with no independent Kuwait-specific primary source of its own. Flagged for a future look. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Gulf mandatory employer-provided health cover, peer-consistent with the sourced UAE and Saudi co-payment bands", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Not reached within the sweep. Lead: Kuwait Master Plan transport studies." }, "non_cash.sick_leave": { "source": "Private Sector Labour Law No. 6 of 2010. Article number is disputed between sources; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-H.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "PwC Worldwide Tax Summaries, Kuwait (reviewed 22 Jul 2026): \"There is no personal income tax (PIT) imposed on individuals in Kuwait\"", "last_updated": "2026-08-11", "confidence": "high", "note": "Recorded for explicitness only; it changes no output by construction. Kuwait dispatches to the same zero-return function as the UAE, which never reads taxable_income." } }, "almaty": { "tax.federal_brackets": { "source": "Kazakhstan 2026 Tax Code (in force 1 Jan 2026): individual income tax (ИПН) is progressive; 10% on annual income up to 8,500 MCI (KZT 36,762,500 at MCI 4,325) and 15% above. Verified against live Russian-language payroll calculators (calculator.kz / mybuh.kz) over the EY/PwC 2026 summaries, per the Sri Lanka calculators-beat-blogs lesson. PwC Worldwide Tax Summaries; Kazakhstan", "last_updated": "2026-05-30", "confidence": "medium" }, "tax.social_security_rate": { "source": "Employee deductions that reduce the IIT base: OPV mandatory pension 10% + VOSMS mandatory medical 2% = 12% combined, both under the 50-MZP/mo contribution ceiling (KZT 51,000,000/yr at MZP 85,000). Modeled as one 12% deductible rate under the shared cap, reproducing calculators.kz exactly (1.5M/mo gross → 1,200,975/mo net). PwC; Leinonen Kazakhstan 2026 payroll guide", "last_updated": "2026-05-30", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "VAT (НДС) standard rate set to 16% under the 2026 Tax Code (raised from 12%). PwC; vatcalc.com Kazakhstan 2026", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Individual immovable-property tax is a progressive schedule on cadastral value (~0.05-0.5%), effectively ~0.1% for a typical apartment; modeled 0.1%. State Revenue Committee; PwC", "last_updated": "2026-05-30", "confidence": "low" }, "housing._summary": { "source": "Numbeo Almaty 2026 (central Medeu / Bostandyk districts) apartment rent + purchase price/m²; krisha.kz listings cross-check, KZT", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "OSMS (compulsory social health insurance) provides the public base; professionals carry a private top-up: premiums by family size from Almaty insurers (Nomad / Freedom / Halyk) 2025-26; OOP/dental Numbeo, KZT", "last_updated": "2026-05-30", "confidence": "low" }, "childcare._summary": { "source": "Almaty private nursery / детсад + after-school 2025-26 (Numbeo + private-garden tariffs); state kindergartens are subsidized but waitlisted, so private assumed and no cash subsidy modeled", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Almaty 2026 grocery + restaurant indices (professional basket), KZT", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "Almaty Onay! monthly pass + metro/bus fares; car line (fuel cheap, ~KZT 250-280/L) dominated by depreciation/insurance. Numbeo 2025-26", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Almaty leisure + domestic vacation (Shymbulak / Kolsai / Issyk) 2025-26, KZT", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "ALA->US/Europe economy round-trip 2025-26 (Air Astana / FlyArystan + connections), KZT", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Kazakhstan's child/family allowances are means-tested (targeted social assistance + large-family payments) → ~0 for an ICP-income household. Ministry of Labour and Social Protection", "last_updated": "2026-05-30", "confidence": "medium" }, "non_cash._summary": { "source": "Kazakhstan Labour Code: 24 calendar-day minimum paid annual leave, ~13 public holidays, maternity 126 calendar days (~18 weeks) at 100% via the State Social Insurance Fund, employer-paid sick leave from day 1 (cap ~25 MCI/day). OSMS = universal public-healthcare base. ILO; Labour Code arts. 88, 99", "last_updated": "2026-05-30", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Almaty: KZT 15,188,585, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "KZT 15,188,585 / 474.31 = USD 32,022; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1281, R5-rounded down to 0.12. REPLACES 0.23 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Describes the post-2022 relocated and foreign-employed cohort more than the domestic market. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.23; sits just above Bangalore (0.20) — low nominal pay, but a low cost base and the flat-ish 10% IIT keep real terms slightly higher." }, "tax.filing_basis": { "source": "State Revenue Committee - individual IIT; no joint filing", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "non_cash.parental_leave_weeks": { "source": "Labour Code + Social Code (2023): maternity 126 calendar days (18 wk); childcare allowance for insured parents from the end of maternity to child age 1.5 (548 - 126 = 422 days = 60.3 wk), a shared pot counted once. Batch-8 audit 2026-08-08: structure confirmed unchanged (Astana Times Jan 2026 + mybuh.kz). Research: docs/superpowers/research/2026-08-08-pl-b8-caucasus-central-asia.md", "last_updated": "2026-08-08", "confidence": "high", "note": "78.3 wk = 18 maternity + 60.3 childcare-to-age-1.5, CONFIRMED by the batch-8 basis audit. Paternity settled black-letter: Labour Code art. 97(3) (the sweep's 'art. 99' cite corrected), up to 5 CALENDAR days explicitly UNPAID -> excluded; no paid-paternity amendment 2025-2026. WATCH: Law 334-VIII (7 Jul 2026) adds multi-employer cap aggregation (anti-abuse, no single-employer effect); Law 277 consolidated-text check pending adilet.zan.kz reachability (TLS failure logged twice)." }, "non_cash.parental_leave_paid_pct": { "source": "Social Code: maternity ~100% of capped average income (one-time payment for 126 days); childcare allowance 40% of the same capped base to age 1.5. Batch-8 re-confirmed 2026-08-08", "last_updated": "2026-08-08", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 78.3 weeks into non_cash.parental_leave_capped (the two tiers flattened), so no percentage-paid weeks remain on this field. Was 100, the maternity-tier headline with the 7x-minimum-wage cap ignored and the 40% tier missing." }, "non_cash.parental_leave_capped": { "source": "Social Code: average monthly income for both tiers capped at 7x the minimum wage; 2026 minimum wage KZT 85,000/month CONFIRMED in force from 1 Jan 2026 (unchanged since 2024) -> base cap KZT 595,000/month; childcare coefficient 0.4. Batch-8 corroboration 2026-08-08: the Astana Times ~2.25M max maternity lump independently reproduces the cap arithmetic (595,000 x 4.2 x 0.9 pension withholding). Research: docs/superpowers/research/2026-08-08-pl-b8-caucasus-central-asia.md", "last_updated": "2026-08-08", "confidence": "high", "note": "Two tiers flattened to one duration-weighted record: (18 wk x 595,000 payout + 60.3 wk x 238,000 payout) / 78.3 = KZT 320,069/month, stored rounded down to 320,000; pct 53.8 is the same weighting of ~100% and 40%. At ICP salaries both tier caps bind, so the weighted form reproduces the statutory total payout; the 10% OPV deduction applies to wages and benefit alike and cancels in the replacement ratio. The non-working-parent flat MRP benefit is excluded (token payment). The cap drifts with the minimum wage each January." }, "non_cash.vacation_days_statutory": { "source": "Labour Code of the Republic of Kazakhstan art. 88 https://adilet.zan.kz/rus/docs/K1500000414", "last_updated": "2026-08-08", "confidence": "high", "note": "CORRECTED 2026-08-08, 24 to 17. Art. 88 grants paid annual leave of 'ne menee dvadtsati chetyrekh KALENDARNYKH dney', 24 CALENDAR days, not working days. We had stored the calendar figure unconverted. 24 calendar days taken as a block spans 3 weeks plus 3 days, which is 17 five-day weekdays. Same units defect as the Gulf 30-day entitlements; the conversion convention is documented in docs/superpowers/research/2026-08-08-vacation-days-non-us-sweep.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Not reached within the sweep. Leads: the Almaty SUMP and the EBRD Green City Action Plan." }, "non_cash.sick_leave": { "source": "Social Code of the Republic of Kazakhstan (2023); ceiling of 25 MCI per month; Shard I research, docs/superpowers/research/2026-08-08-sick-leave-shard-I.md", "source_url": "https://www.gov.kz/memleket/entities/enbek", "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "ENPF (Unified Accumulative Pension Fund) https://www.enpf.kz/en/press-center/news/281217/ : \"the abolition of tax deductions on voluntary pension contributions made in one's own favor from January 1, 2026\"", "last_updated": "2026-08-11", "confidence": "high", "note": "ABOLISHED. The new Tax Code (signed 18 Jul 2025, in force 1 Jan 2026) removed the PIT deduction for voluntary pension contributions in one's own favour. The quid pro quo is that all UAPF payouts became PIT-exempt for residents, flipping the regime from EET to TEE; the payout side is not modeled by this engine." } }, "tbilisi": { "tax.federal_brackets": { "source": "Georgia (country) Tax Code: personal income tax is a flat 20% on employment income, withheld at source, with NO basic personal allowance for residents. Modeled as a single flat bracket on full gross (standard_deduction = 0). Verified against live Georgian payroll calculators (ibn.ge, docify.ge) per the calculators-beat-blogs lesson: GEL 10,000/mo gross → 7,800/mo net. PwC Worldwide Tax Summaries; Georgia; Revenue Service of Georgia (rs.ge)", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "No general social-security tax; the only employee payroll deduction is the mandatory funded pension (Pension Agency) at 2% of gross. Employer adds 2% and the state 0-2% (income-tapered), but only the employee 2% reduces take-home. Not deductible from the IIT base. Modeled 2% with no binding cap (the income tapering affects only the state's matched share, not the employee 2%), reproducing net = gross × 0.78 on the live calculators. PwC; pensions.ge", "last_updated": "2026-05-30", "confidence": "high" }, "tax.sales_tax_rate": { "source": "VAT (დღგ) standard rate 18%. PwC; Revenue Service of Georgia", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Property tax on individuals is income-banded up to 1% of market value and only applies above a GEL 40,000 household-income threshold; effectively ~0.1% for a typical apartment, modeled 0.1%. PwC; Tax Code of Georgia art. 202", "last_updated": "2026-05-30", "confidence": "low" }, "housing._summary": { "source": "Numbeo Tbilisi 2026 (central Vake / Saburtalo districts) apartment rent + purchase price/m²; myhome.ge / ss.ge listings cross-check, GEL", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Universal Healthcare Programme (UHP) provides the public base; professionals carry a private top-up: premiums by family size from Tbilisi insurers (GPI / Aldagi / Imedi L) 2025-26; OOP/dental Numbeo, GEL", "last_updated": "2026-05-30", "confidence": "low" }, "childcare._summary": { "source": "Tbilisi municipal kindergartens are free for residents but waitlisted/oversubscribed, so a private nursery + after-school basket is assumed (Numbeo + private-garden tariffs 2025-26); no cash subsidy modeled", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Tbilisi 2026 grocery + restaurant indices (professional basket), GEL", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "Tbilisi metro/bus unified monthly pass (Metromoney) + fares; car line (fuel ~GEL 2.6-2.9/L) dominated by depreciation/insurance. Numbeo 2025-26", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Tbilisi leisure + domestic vacation (Kazbegi / Batumi / Bakuriani) 2025-26, GEL", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "TBS->US/Europe economy round-trip 2025-26 (Georgian Airways / Wizz / Turkish + connections), GEL", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Georgia has no universal child benefit; family support is means-tested (targeted social assistance, subsistence allowance) → ~0 for an ICP-income household. Ministry of IDPs, Labour, Health and Social Affairs", "last_updated": "2026-05-30", "confidence": "medium" }, "non_cash._summary": { "source": "Labour Code of Georgia: 24 working-day minimum paid annual leave; ~17 public holidays; maternity/childcare leave of 730 calendar days total, of which ~183 days (~26 weeks) are paid; headline modeled at 100% but the state allowance is capped (historically GEL 1,000, raised since), so high earners' replacement is far below 100%. Employer-paid sick leave during certified illness (job protection up to 40 consecutive / 60 total days per 6-month period). UHP = universal public-healthcare base. ILO; Labour Code arts. 36, 37", "last_updated": "2026-05-30", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Tbilisi: GEL 132,500, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-europe.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "GEL 132,500 / 2.62 = USD 50,573; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2023, R5-rounded down to 0.2. REPLACES 0.27 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Round-numbered percentiles and a 1.12 senior ratio: thin and senior-weighted. levels.fyi taken over ERI SalaryExpert's roughly 30% higher modeled figure, which is the conservative choice. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.27; sits just above Almaty (0.23) — low nominal pay but a low cost base and the flat 20% IIT keep real terms competitive." }, "non_cash.parental_leave_weeks": { "source": "Labour Code of Georgia (matsne.gov.ge black-letter, batch-8 2026-08-08): art. 37(1) maternity 126 calendar days, ALL paid + art. 37(3) parental leave with 57 PAID calendar days, usable by mother or father (art. 37(4)) = one shared pot counted once. 183 paid days /7 = 26.14 -> 26.1. Research: docs/superpowers/research/2026-08-08-pl-b8-caucasus-central-asia.md", "last_updated": "2026-08-08", "confidence": "high", "note": "Batch-8 restate 18 -> 26.1: the 2026-08-06 duration contest (126 vs 183 paid days) is settled black-letter at 183. The remaining 547 parental days are UNPAID, excluded. Complicated-birth/twins variant (200 paid days = 143+57) recorded, base case stored. The Jul-2026 civil-servant-father 90-day rule is PUBLIC-sector only, excluded; watch private-sector spillover. No separate private-sector paternity." }, "non_cash.parental_leave_paid_pct": { "source": "Georgian maternity/childcare state assistance: up to 100% of salary but hard-capped in TOTAL for the whole leave (interpressnews.ge 2023-01-23 reform report + netgazeti.ge, both contemporaneous; batch-8 re-confirmed 2026-08-08)", "last_updated": "2026-08-08", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all paid weeks into non_cash.parental_leave_capped with a per-leave lump ceiling, so no percentage-paid weeks remain on this field. The old naked 100% was the single most misleading figure in the region: the real effective rate at a Tbilisi professional salary is ~1%." }, "non_cash.parental_leave_capped": { "source": "Order No. 33 of 24 Jan 2023 (the reform interpressnews.ge/netgazeti.ge reported contemporaneously): the state maternity assistance rose from GEL 1,000 to GEL 2,000 JAMURAD ('jamurad araumetes 2000' = in total, cumulatively), private-sector employees; unamended through 2026-08-08 (batch-8 re-check). Research: docs/superpowers/research/2026-08-08-pl-b8-caucasus-central-asia.md", "last_updated": "2026-08-08", "confidence": "high", "note": "cap_period \"leave\": GEL 2,000 is the TOTAL state payout for the whole paid window - since the batch-8 weeks restate that window is 26.1 wk (183 days), so the engine spreads it at ~GEL 77/wk (~1.0% at the rankings basis, was ~1.5% over 18 wk). Same money, longer window; the ISET-PI \"monthly\" framing stays refuted. Public-sector employees are fully compensated with no cap, but our ICP is private-sector, the correct and conservative pick. This effective rate sits ON the token line; the count-the-lump verdict is the 2026-08-06 sweep carry-over, kept deliberately (it is the country's entire paid-leave scheme, not a side allowance)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Not reached within the sweep. Leads: the Tbilisi SUMP and the EBRD Green City Action Plan." }, "non_cash.sick_leave": { "source": "Labour Code of Georgia art. 46(2)(i) and art. 46(4); Shard I research, docs/superpowers/research/2026-08-08-sick-leave-shard-I.md", "source_url": "https://matsne.gov.ge/en/document/view/1155567", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "PwC Worldwide Tax Summaries, Georgia, Individual deductions (rev. 21 Jan 2026) https://taxsummaries.pwc.com/georgia/individual/deductions : \"Employees are not entitled to claim deductions\"", "last_updated": "2026-08-11", "confidence": "medium", "note": "Flat PIT with no employee deductions. The funded pension is MANDATORY (employee 2%, employer 2%, state 1-2% tapering out above GEL 60,000) and belongs to payroll, not here. No voluntary top-up with tax relief was found. MEDIUM because the finding rests on an aggregator rather than a Georgian primary source." } }, "yerevan": { "tax.federal_brackets": { "source": "Armenia personal income tax is a FLAT 20% on full gross employment income, withheld at source, with NO basic personal allowance. The multi-year phase-down (23→22→21→20%) stopped at 20%: no further cut enacted for 2025/2026. (A special 10% applies only to certified IT-company staff: out of scope for the ICP.) Verified against the live b24.am net-salary calculator per the calculators-beat-blogs lesson. PwC Worldwide Tax Summaries: Armenia; State Revenue Committee", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "Combined mandatory employee deductions, modeled via the engine's cap×rate form (cap 13,596,000 × 10% = 1,359,600/yr) because all three components are flat/capped for an ICP earner (>~1.13M/mo gross): (1) funded pension; 5% up to AMD 500k/mo, then 10%×gross−25,000, CAPPED at AMD 87,500/mo = 1,050,000/yr; (2) military stamp duty; fixed AMD 15,000/mo (>1M/mo tier) = 180,000/yr; (3) mandatory health contribution (new, RA law in force 25 Dec 2025); fixed AMD 10,800/mo (>500k tier) = 129,600/yr. Sum 1,359,600/yr reproduces the corrected b24.am anchor (2,000,000/mo gross → 1,486,700/mo net). The 13,596,000 'cap' is an engine artifact, not a statutory figure. b24.am; PwC; pension.am", "last_updated": "2026-05-30", "confidence": "high" }, "tax.sales_tax_rate": { "source": "VAT (ԱԱՀ) standard rate 20%. PwC; State Revenue Committee", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Individual real-estate tax is a progressive cadastral schedule (~0.05-1.5%); the prior exemption threshold was removed in the 2021 reform and is phasing to 100% of the assessed amount by 2026. Modeled ~0.15% for a typical mid apartment. PwC; Tax Code of Armenia", "last_updated": "2026-05-30", "confidence": "low" }, "housing._summary": { "source": "Yerevan central (Kentron / Arabkir) apartment rent + purchase price 2025-26; the post-2022 relocation wave pushed central rents above Numbeo's lagging index, so figures set from list.am listings + local market reporting, AMD", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "The new universal health insurance base (in force 25 Dec 2025) covers ~30% of services at launch, so professionals still carry private cover: premiums by family size from Yerevan insurers (Rosgosstrakh-Armenia / SIL / Nairi) 2025-26; OOP/dental Numbeo, AMD", "last_updated": "2026-05-30", "confidence": "low" }, "childcare._summary": { "source": "Yerevan private nursery / preschool + after-school 2025-26 (Numbeo + private-garden tariffs); the city's 160+ municipal kindergartens are heavily subsidized/near-free but oversubscribed, so private assumed and no cash subsidy modeled", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Yerevan 2026 grocery + restaurant indices (professional basket), AMD", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "Yerevan metro/bus fares + monthly pass; car line (fuel ~AMD 500/L) dominated by depreciation/insurance. Numbeo 2025-26", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Yerevan leisure + domestic vacation (Sevan / Dilijan / Tsaghkadzor) 2025-26, AMD", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "EVN->US/Europe economy round-trip 2025-26 (Wizz / Aegean / FlyOne + connections), AMD", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Armenia's family support is means-tested (low-income family benefit + one-off birth grants) → ~0 for an ICP-income household. Ministry of Labour and Social Affairs", "last_updated": "2026-05-30", "confidence": "medium" }, "non_cash._summary": { "source": "Labour Code of Armenia: 20 working-day minimum paid annual leave; ~16 public holidays; maternity 140 calendar days (~20 weeks) at 100% of average earnings; sick leave employer-paid first 5 days at 80% then State Social Insurance at 80%. The 2025 Universal Health Insurance Law = a genuine public base (true). ILO; Labour Code arts. 159-174", "last_updated": "2026-05-30", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Yerevan: AMD 19,607,565, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-europe.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "AMD 19,607,565 / 366.15 = USD 53,551; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2142, R5-rounded down to 0.21. REPLACES 0.26 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Thin market distorted by post-2022 IT relocation. Separately recorded: Glassdoor states Yerevan developers average AMD 754,000 per year, which is about USD 1,950 and impossible; it is a monthly figure under an annual label. Do not ingest it. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.26; sits just below Tbilisi (0.27), above Almaty (0.23)." }, "non_cash.parental_leave_weeks": { "source": "Law on Temporary Incapacity and Maternity Benefits: 140 calendar days maternity (70+70) /7 = 20 wk; Labour Code art. 176.1 paternity 5 WORKING days ('hing ashkhatanqayin or', added by HO-415-N of 16 Sep 2020, black-letter verified on arlis.am act 146281) /5 = 1.0 wk. Total 21. Research: docs/superpowers/research/2026-08-08-pl-b8-caucasus-central-asia.md", "last_updated": "2026-08-08", "confidence": "high", "note": "Batch-8 day-basis fix 2026-08-08: the stored 20.7 priced the paternity days /7; the statute says working days -> /5. The under-2 childcare benefit for a parent on leave (50% of the minimum wage = AMD 37,500/mo, Gov decision 1566-N as restated 19 Feb 2026) is 0.82% of monthly pay at the rankings basis -> TOKEN, excluded (Azerbaijan precedent). Unpaid husband's-leave and unpaid childcare-to-3 excluded." }, "non_cash.parental_leave_paid_pct": { "source": "Labour Code art. 176.1: the 5 paternity working days are employer-paid at average daily wage with NO statutory cap (arlis.am black-letter); maternity 100% of average earnings sits in parental_leave_capped", "last_updated": "2026-08-08", "confidence": "high", "note": "100 describes the 1.0 uncapped paternity week (employer wage continuation is not a state benefit, so the 15x state cap does not reach it - Azerbaijan batch-7 precedent; the conservative cap-it-anyway alternative was considered and rejected as shape-dishonest). The 2024 tenure-scale WATCH ITEM (40/60/80% proposal) is CLOSED: the 2025 benefits amendment is procedural only (automatic assignment via the Unified Social Service since Sep 2025), the flat 100%/15x structure is unchanged." }, "non_cash.parental_leave_capped": { "source": "Reference earnings capped at 15x the minimum monthly wage (SSA/ISSA structure + independent confirmation the multiplier is current and self-updating); 2026 minimum wage AMD 75,000 IN FORCE 2026-08-08 (the announced 85,000 rise not yet enacted) -> cap AMD 1,125,000/month. Research: docs/superpowers/research/2026-08-08-pl-b8-caucasus-central-asia.md", "last_updated": "2026-08-08", "confidence": "high", "note": "Payout ceiling AMD 1,125,000/month on the 20 maternity weeks at 100% (base and payout coincide; the maternity multiplier is 15x versus the sickness benefit's 5x). The paternity week is OUTSIDE this block since the batch-8 restate (employer-paid, uncapped). Self-updating with the minimum wage: when the announced AMD 85,000 minimum enacts, the cap becomes 1,275,000; calendar row added, annual refresh." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Not reached within the sweep. Leads: the Yerevan SUMP and the EBRD Green City Action Plan." }, "non_cash.sick_leave": { "source": "RA Law on Temporary Incapacity and Maternity Benefits; Shard I research, docs/superpowers/research/2026-08-08-sick-leave-shard-I.md", "source_url": "https://www.mlsa.am/", "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "PwC Worldwide Tax Summaries, Armenia (rev. 30 Jun 2026) https://taxsummaries.pwc.com/armenia/individual/other-taxes plus an IMF technical-assistance report on Armenian PIT", "last_updated": "2026-08-11", "confidence": "medium", "note": "Voluntary cumulative (Pillar III) contributions by the individual and/or a third party including the employer are deducted from gross income up to 5%. No monetary cap, so no max key: writing max 0 here would zero out a real deduction. MEDIUM and honest about why: the 5% is consistent everywhere it appears but NO primary Armenian source was read, and the PwC pages that opened confirm the mandatory rates while staying silent on the 5%. The 7.5% figure sometimes seen is the CORPORATE deduction limit, not this one." } }, "baku": { "tax.filing_basis": { "source": "PwC WWTS - Azerbaijan: each individual files a return; joint returns are not permitted. (Immaterial at ICP income: both earners sit in the top band, so stacking is a near no-op.)", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Azerbaijan PIT, non-oil-gas PRIVATE sector employee, 2026 schedule. The 7-year 0%-up-to-8,000/mo + 14%-above exemption (in force 2019-2025) EXPIRED 31 Dec 2025 and was NOT extended; a new phased progressive schedule took effect 1 Jan 2026: 3% up to 2,500/mo (30,000/yr), 10% from 2,500-8,000/mo (to 96,000/yr), 14% above. Modeled as a 3-bracket marginal ladder on full gross (standard_deduction = 0; the 200/mo low-income sub-allowance is immaterial for the ICP). Verified against live 2026 calculators (grossnet.az, muhasibat.az) and the official State Tax Service post per the calculators-beat-blogs lesson: AZN 10,000/mo gross -> ~7,951.50/mo net. State Tax Service taxes.gov.az/az/post/4461; apa.az; ifac.az", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "Combined employee social burden modeled as a blended ~11.5% UNCAPPED rate (social_security_cap = null), reproducing the validated net anchors within ~0.1%. The real structure is a split: DSMF social insurance 3% on the first 200/mo + 10% above (retained for income <=8,000/mo from 2026), unemployment insurance 0.5%, and mandatory medical insurance 2% up to 2,500/mo + 0.5% above. The two kink points (200/mo, 2,500/mo) are tiny relative to ICP incomes so the blend is accurate (true effective ~11.8% at 3k/mo, ~11.2% at 10k/mo). The 0.115 is an engine blend, not a single statutory rate. taxes.gov.az/az/post/4461; taxes.gov.az/az/post/1314; muhasibat.az/2026-emekhaqqidan-tutulmalar", "last_updated": "2026-05-30", "confidence": "medium" }, "tax.sales_tax_rate": { "source": "VAT (ƏDV) standard rate 18%. PwC Worldwide Tax Summaries - Azerbaijan; vatupdate.com 2026", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Individual residential property tax is levied per m2 on area exceeding 30 m2 (Baku 0.4 AZN/m2/yr), NOT on value -> on a typical apartment the effective rate vs market value is ~0.008%; modeled nominal 0.01%. Only homeowners pay (renters in the golden personas pay 0). Taxation in Azerbaijan (Tax Code); PwC", "last_updated": "2026-05-30", "confidence": "low" }, "housing._summary": { "source": "Numbeo Baku 2026 (central Nizami / Yasamal / Sahil) apartment rent + purchase price/m2; bina.az / ev10.az listings cross-check; utilities are heavily subsidized (~AZN 110/mo). AZN", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Compulsory medical insurance (icbari tibbi sigorta, nationwide from 2021) provides the public base; professionals carry a private top-up - premiums by family size estimated from Baku insurers 2025-26 (thin English sourcing, order-of-magnitude); OOP/dental Numbeo. AZN", "last_updated": "2026-05-30", "confidence": "low" }, "childcare._summary": { "source": "Baku private nursery / kindergarten + after-school 2025-26 (Numbeo + private-garden tariffs); no means-tested cash subsidy reaches an ICP-income household -> subsidy 0. AZN", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Baku 2026 grocery + restaurant indices (professional basket). AZN", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "Baku BakiKart per-ride fares (~AZN 0.50/ride) expressed as a monthly-pass equivalent; car line (cheap fuel, used-import heavy) dominated by depreciation/insurance. Numbeo 2025-26", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Baku leisure + domestic vacation (Gabala / Quba / Naftalan) 2025-26. AZN", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "GYD->US/Europe economy round-trip 2025-26 (Azerbaijan Airlines / Turkish / Wizz + connections). AZN", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Azerbaijan family support is means-tested (targeted social assistance, one-off birth grants) -> ~0 for an ICP-income household. Ministry of Labour and Social Protection (DOST)", "last_updated": "2026-05-30", "confidence": "medium" }, "non_cash._summary": { "source": "Labour Code of Azerbaijan: 21 calendar-day minimum paid annual leave (~15 working days on a 5-day-week basis, per our locked statutory-minimum convention); ~19-20 public holidays (Islamic dates shift the count year to year; modeled 20); maternity 126 calendar days (~18 weeks) at 100% via the State Social Protection Fund; employer pays the first 14 days of certified sick leave, then SSPF. Compulsory medical insurance = a genuine public base -> universal_healthcare true. ILO; Labour Code arts. 78, 114-117", "last_updated": "2026-05-30", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Baku: AZN 55,177, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "AZN 55,177 / 1.7 = USD 32,457; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1298, R5-rounded down to 0.12. REPLACES 0.3 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Page renders 'not enough data for top locations'; p25 to p90 spans 7.0x, the two-cohort signature. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.30; set above Tbilisi (0.27)/Yerevan (0.26) on the oil-capital wage premium, toward the low end of the expected 0.28-0.36 band given the small tech base." }, "non_cash.parental_leave_paid_pct": { "source": "Labour Code art. 125 as amended (in force 16 Jan 2026): 14 calendar days paid paternity at the art. 177 average wage of the prior 2 months, employer-paid (e-social.gov.az official news + vergiler.az FAQ)", "last_updated": "2026-08-08", "confidence": "high", "note": "100 covers only the 2.0-week paternity remainder - direct employer wage continuation, genuinely uncapped (not a DSMF muavinat, so the 25x-min-pension cap does not apply). The 18 maternity weeks live in non_cash.parental_leave_capped. Was 0 with paternity excluded as unpaid; the 16 Jan 2026 amendment made it paid, and the under-3 allowance remains excluded as token (see the weeks entry)." }, "non_cash.parental_leave_capped": { "source": "dsmf.gov.az maternity page (primary, fetched 2026-08-06): \"monthly maximum amount cannot exceed 25 times the minimum amount of labor pension\"; minimum labor pension AZN 349/month from Feb 2026 (sosial.gov.az + azernews + caliber + oc-media) -> cap = AZN 8,725/month. Research: docs/superpowers/research/2026-08-06-parental-caps-asia-mideast.md (Azerbaijan) and docs/superpowers/research/2026-08-06-pl-caps-central-asia-caucasus.md", "last_updated": "2026-08-08", "confidence": "high", "note": "Payout ceiling AZN 8,725/month on all 18 weeks at 100% (base and payout coincide). A genuinely high cap for the region: it binds only near the top of the Baku band (~AZN 105k/yr). The 25x multiplier is fixed in law; the base (minimum pension) moves roughly each February, annual refresh. Cleanest country in the Central Asia cluster: direct official fetch. Re-verified 2026-08-08 (batch 7): structure, 25x multiplier and the AZN 349 base all current; maternity carries a >=6-month insurance-record gate (see the weeks entry). The minimum labor pension moves ~each February - tracked in docs/data-refresh-calendar.md." }, "non_cash.parental_leave_weeks": { "source": "dsmf.gov.az maternity page (primary, fetched 2026-08-08): 126 calendar days (70 pre + 56 post) at 100% of average daily earnings ; Labour Code art. 125 as amended, IN FORCE 16 Jan 2026: 14 calendar days PAID paternity, employer-paid at art. 177 average wage (official e-social.gov.az news + vergiler.az FAQ + convergent coverage; e-qanun.az consolidated art. 125 text not fetched - pin next sweep). Research: docs/superpowers/research/2026-08-08-pl-b7-europe-caucasus.md", "last_updated": "2026-08-08", "confidence": "high", "note": "20.0 wk = maternity 18.0 (126/7, the capped block) + paternity 2.0 (14 calendar days /7 at 100, uncapped - employer wage continuation under art. 177, not a DSMF benefit, so the 25x-min-pension cap does not apply). Was 18, and the caps sweep's 'paternity 14 days UNPAID, excluded' verdict is OVERTURNED on new primary evidence (the amendment postdates it). Gate: maternity requires >=6 months of social-insurance record (DSMF primary - refutes the earlier 'from day 1' assumption); leave starts ~70 days pre-birth, so the mover at ~6.7-9.7 months of record at leave start passes (narrowly at the bottom; no foreign-record aggregation) - caveat only. The under-3 childcare allowance (AZN 44/28 per month, re-verified current 2026) stays EXCLUDED as a token social allowance (~0.21% replacement at the basis salary; adding ~78 near-zero weeks would only inflate the weeks column). Conditional extensions (156 d difficult delivery / 180 d multiples) excluded. The '1 May 2026 change' is procedural only (proactive e-assignment via EMAS)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Not reached within the sweep. Leads: the Baku SUMP and the EBRD Green City Action Plan." }, "non_cash.sick_leave": { "source": "Labour Code of Azerbaijan art. 78; note to art. 74; Shard I research, docs/superpowers/research/2026-08-08-sick-leave-shard-I.md", "source_url": "https://e-social.gov.az/", "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Tax Code art. 102.1.8 via Ekvita https://ekvita.com/en/view/news/92 and https://cis-legislation.com/document.fwx?rgn=2601 ; PwC Worldwide Tax Summaries, Azerbaijan (rev. 09 Jan 2026, silent on an employee route)", "last_updated": "2026-08-11", "confidence": "medium", "note": "A FORK, resolved by the house rule. Art. 102.1.8 exempts from the EMPLOYEE's income tax the premiums paid BY THE EMPLOYER on life-savings and pension/annuity insurance with Azerbaijani insurers, up to 50% of taxable income, on 3+ year contracts. Nothing was found granting relief for an employee's own direct voluntary contribution. HIGH that 102.1.8 exists and is employer-routed; MEDIUM that no employee-paid route exists, and the fork swings Azerbaijan between zero deduction and full deduction at a 6-10% contribution. The alternative, if the salary-packaging route is ever wanted, is { pct_of_earnings: 0.50 }." } }, "tashkent": { "tax.federal_brackets": { "source": "Uzbekistan PIT (JShDS) is a FLAT 12% on employment income with no basic allowance (the 2019 tax reform replaced the old progressive scale with a single 12% rate; unchanged for 2025-26). Modeled as a single-band marginal ladder [{up_to:null,rate:0.12}] on full gross (standard_deduction = 0) -> net = gross x 0.88. Verified against live Uzbek payroll calculators (net = 88% of gross) per the calculators-beat-blogs lesson. State Tax Committee soliq.uz; PwC Worldwide Tax Summaries - Uzbekistan; Tax Code of the Republic of Uzbekistan art. 381", "last_updated": "2026-05-30", "confidence": "high" }, "tax.social_security_rate": { "source": "social_security_rate = 0: there is NO additional employee social-insurance deduction on top of the 12% PIT. The individual funded-pension contribution (INPS) is 0.1% but it is CARVED OUT of the 12% PIT (the employer transfers 0.1% of taxable income to the worker's INPS account FROM WITHIN the 12% withheld), NOT an extra levy -> the employee's total burden is exactly 12%. Adding 0.1% on top would overstate it. Employer-side social tax (~12-25%) is not borne by the employee. soliq.uz; PwC Uzbekistan (Other taxes - social)", "last_updated": "2026-05-30", "confidence": "high" }, "tax.sales_tax_rate": { "source": "VAT (QQS) standard rate 12% (reduced from 15% effective 1 Jan 2023). State Tax Committee soliq.uz; PwC Worldwide Tax Summaries - Uzbekistan VAT", "last_updated": "2026-05-30", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Property tax on individuals is levied on the cadastral value of residential real estate at ~0.4% (rates set in the Tax Code, indexed; Tashkent residential ~0.4%); modeled nominal 0.4%. Only homeowners pay (renters in the golden personas pay 0). Tax Code of Uzbekistan; PwC", "last_updated": "2026-05-30", "confidence": "low" }, "housing._summary": { "source": "Numbeo Tashkent 2026 (central Yunusabad / Mirzo-Ulugbek / city-center) apartment rent + purchase price/m2; uybor.uz / olx.uz listings cross-check at ~12,000 UZS/USD. UZS", "last_updated": "2026-05-30", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Uzbekistan has a state health system (free public base, expanding under the State Medical Insurance pilot); professionals carry a private/clinic top-up - premiums by family size estimated from Tashkent private clinics 2025-26 (thin English sourcing, order-of-magnitude); OOP/dental Numbeo. UZS", "last_updated": "2026-05-30", "confidence": "low" }, "childcare._summary": { "source": "Tashkent private nursery / kindergarten + after-school 2025-26 (Numbeo + private-garden tariffs); state kindergartens are cheap/subsidized but professionals use private -> modeled private; no means-tested cash subsidy reaches an ICP-income household -> subsidy 0. UZS", "last_updated": "2026-05-30", "confidence": "low" }, "food._summary": { "source": "Numbeo Tashkent 2026 grocery + restaurant indices (professional basket; dining-out baseline set at ~2,000,000 UZS/mo for internal basket coherence with groceries/entertainment, the source value was an outlier). UZS", "last_updated": "2026-05-30", "confidence": "medium" }, "transit._summary": { "source": "Tashkent metro/bus per-ride fare (~1,700 UZS/ride) expressed as a monthly-pass equivalent; car line (cheap fuel, used-import heavy) dominated by depreciation/insurance. Numbeo 2025-26", "last_updated": "2026-05-30", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Tashkent leisure + domestic vacation (Samarkand / Bukhara / Chimgan mountains) 2025-26. UZS", "last_updated": "2026-05-30", "confidence": "low" }, "travel_home._summary": { "source": "UZS->US/Europe economy round-trip 2025-26 (Uzbekistan Airways / Turkish + connections). UZS", "last_updated": "2026-05-30", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Uzbekistan family allowances (child benefit, low-income family allowance) are means-tested via the Unified Register / Ijtimoiy Himoya -> ~0 for an ICP-income household. Ministry of Employment and Poverty Reduction", "last_updated": "2026-05-30", "confidence": "medium" }, "non_cash._summary": { "source": "Labour Code of the Republic of Uzbekistan (in force 30 Apr 2023): 15 working-day minimum paid annual leave (per our locked statutory-minimum convention); ~10 public holidays (Islamic dates shift year to year); maternity 126 calendar days (~18 weeks) at 100% via state social insurance; temporary-disability (sick) benefit paid from day 1 at 60-100% by length of service. State public health system -> universal_healthcare true. ILO; Labour Code arts. on leave/maternity", "last_updated": "2026-05-30", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Tashkent: UZS 359,096,869, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "UZS 359,096,869 / 11861.8 = USD 30,273; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1211, R5-rounded down to 0.12. REPLACES 0.24 (2026-05-30), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Describes the IT-Park export cohort, not the Uzbek labour market. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.24; set just below Tbilisi (0.27)/Yerevan (0.26)/Baku (0.30) and above the South-Asia metros; USD-paid remote/export engineers are a separate higher cohort not captured here." }, "non_cash.parental_leave_weeks": { "source": "Maternity 126 calendar days /7 = 18 wk, confirmed unchanged by the UP-206/Decree 796 reform (Decree 796 of 17 Dec 2025 via NORMA.UZ/buxgalter.uz quotes; lex.uz full text not directly fetched, flagged). No statutory paid paternity (the 5-day paid draft is still before parliament -> watch row). Research: docs/superpowers/research/2026-08-08-pl-b8-caucasus-central-asia.md", "last_updated": "2026-08-08", "confidence": "high", "note": "The under-2 childcare allowance EXISTS as a distinct paid stream (employer-paid, 60% of MROT = UZS 762,600/month from 1 Jul 2026) but is 0.51% of monthly pay at the rankings basis -> TOKEN, excluded (Azerbaijan precedent; arithmetic recorded). Unpaid leave to age 3 excluded." }, "non_cash.parental_leave_paid_pct": { "source": "Presidential Decree UP-206 reform, in force 1 Jan 2026: maternity benefit moved from employer to the State Social Insurance Fund with a tenure-tiered rate; the ladder is now BLACK-LETTER confirmed by Decree 796 of 17 Dec 2025 (insurance record 10-24 months = 75%, 25-60 = 85%, 61+ = 100%); EY Uzbekistan tax alert 2026-02 + NORMA.UZ + buxgalter.uz", "last_updated": "2026-08-08", "confidence": "high", "note": "We store 0: the caps rebuild moved all 18 weeks into non_cash.parental_leave_capped at the 75% mover tier, so no percentage-paid weeks remain on this field. PRICE-THE-MOVER basis (owner ruling 2026-08-06, re-confirmed batch-8): 75% is the lowest qualifying rung, the arriving hire's entitlement; longer tenure pays 85/100%, upside not stored. The sub-10-month gap is now RESOLVED and LOUD: a >=10-month insurance-record gate applies to ANY maternity benefit, so a mover giving birth under 10 insured months gets ZERO - this straddles the work-to-term window (recorded as caveat, not haircut; the 75% rung IS the mover price). SSIF pays fully from 1 Jan 2026 incl. private sector (proactive assignment). Paternity draft still pending." }, "non_cash.parental_leave_capped": { "source": "Decree 796 / UP-206: average monthly salary capped at 10x the minimum wage effective at assignment. Presidential Decree PF-115, official text https://lex.uz/docs/-8283656, sets MROT at UZS 1,360,000/month from 1 Sep 2026 -> base cap UZS 13,600,000 -> payout at the unchanged 75% mover tier = UZS 10,200,000/month. Verification: docs/superpowers/research/2026-08-31-tashkent-mrot-effective-update.md", "last_updated": "2026-08-31", "confidence": "high", "note": "Effective 2026-09-01, not before: 75% x 10 x UZS 1,360,000 = UZS 10,200,000/month. The preceding vintage was UZS 9,532,500/month (75% x 10 x UZS 1,271,000) through 2026-08-31. The cap moves with the minimum wage at assignment date. No field basis changed: this remains the Decree 796 10--24-month insurance-record mover rung; the >=10-month eligibility caveat and separate under-two childcare allowance remain outside this value." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Not reached within the sweep. Leads: the Tashkent SUMP and ADB transport studies." }, "non_cash.sick_leave": { "source": "Labour Code of the Republic of Uzbekistan; state social insurance rules from 1 July 2026; Shard I research, docs/superpowers/research/2026-08-08-sick-leave-shard-I.md", "source_url": "https://lex.uz/", "last_updated": "2026-08-08", "confidence": "medium", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "PwC Worldwide Tax Summaries, Republic of Uzbekistan, Individual deductions (rev. 16 Jan 2026) https://taxsummaries.pwc.com/republic-of-uzbekistan/individual/deductions : \"There are no personal deductions under Uzbek law\"", "last_updated": "2026-08-11", "confidence": "medium", "note": "12% flat PIT with no personal deductions. Only categorical social deductions exist (veterans, disabled, widow(er)s) at 4x BCU, which are not retirement saving. MEDIUM because the finding rests on an aggregator rather than an Uzbek primary source." } }, "delhi-ncr": { "tax.federal_brackets": { "source": "India FY 2026-27 new regime slabs (5/10/15/20/25/30%) + 4% health-and-education cess, modeled as the Bangalore/India peer ladder. Boundaries corrected 2026-08-01 (see the Bangalore tax.federal_brackets note): the INR 75,000 standard deduction is folded in by shifting EVERY slab and surcharge boundary, not just the first. Income Tax Dept https://www.incometax.gov.in/iec/foportal/help/individual/return-applicable-1 ; PwC Worldwide Tax Summaries - India", "last_updated": "2026-05-31", "confidence": "high" }, "tax.social_security_rate": { "source": "Employee EPF 12% of basic capped at the INR 15,000/mo statutory wage base (modeled social_security_cap 180,000/yr). EPFO; PwC India", "last_updated": "2026-05-31", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Delhi MCD unit-area-system house tax on owners ~0.3% effective (renters in golden personas pay 0). MCD property-tax schedules", "last_updated": "2026-05-31", "confidence": "low" }, "housing._summary": { "source": "Numbeo Delhi/Gurgaon/Noida 2026 (DLF / sector high-rise, professional tier) apartment rent + price/sqft. INR", "last_updated": "2026-05-31", "confidence": "medium" }, "housing.rent_percentiles": { "source": "MagicBricks listing-count CDF, 2026-08-18 (SERVER_PRELOADED_STATE_ resultCount; card price = monthly rent, deposit separate as bookingAmtExact, cap inclusive on the rent field, re-verified this run; apartments only; asking INR/month): New-Delhi (NCT) pool 2 BHK P25/med/P75 = 19,090/29,452/43,253 (n=3,247) -> 0.648/1.469; 1 BHK 0.647/1.611 (n=1,890). Gurgaon pool corroborates: 2 BHK 0.707/1.373 (n=2,370), 1 BHK 0.731/1.516.", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW decisively in both measured NCR pools (New-Delhi 2 BHK -0.172/+0.229; Gurgaon agrees directionally). Applied the New-Delhi pair 0.65/1.47 (nearest). GEOGRAPHY WEDGE recorded: this file is the Delhi-NCR REGION but MagicBricks has NO NCR-wide pool (New-Delhi, Gurgaon, Noida, Ghaziabad, Faridabad are separate portal cities), so no portal pool matches the stored basis exactly; an NCR-wide pool adds between-city dispersion and can only read WIDER than the applied pair - the fix is conservative. The New-Delhi pool spans the full NCT dual market (Narela/Mohan Garden budget belt through Chhatarpur Farms premium). 2 BHK anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-india.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (E). NOTE the stored NCR 2BR median (50,000) sits at ~p81 of the New-Delhi pool and ~p84 of Gurgaon's - an NCR-scale premium-belt level, flagged as a separate level question, out of scope here." }, "healthcare._summary": { "source": "Private health insurance + OOP for a professional family (public base thin); premiums by family size from Indian insurer 2025-26 quotes + Numbeo OOP. INR", "last_updated": "2026-05-31", "confidence": "low" }, "childcare._summary": { "source": "Gurgaon/Noida private daycare + preschool + after-school 2025-26 (professional tier); no means-tested cash subsidy reaches an ICP household -> subsidy 0. INR", "last_updated": "2026-05-31", "confidence": "low" }, "food._summary": { "source": "Numbeo Delhi 2026 grocery + restaurant indices (professional basket). INR", "last_updated": "2026-05-31", "confidence": "medium" }, "transit._summary": { "source": "Delhi Metro monthly-pass equivalent + car (fuel/insurance/depreciation); Numbeo 2025-26. INR", "last_updated": "2026-05-31", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Delhi leisure + domestic vacation baseline 2025-26. INR", "last_updated": "2026-05-31", "confidence": "low" }, "travel_home._summary": { "source": "INR->US/Europe economy round-trip 2025-26 (Air India / Vistara / Gulf-carrier connections). INR", "last_updated": "2026-05-31", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "India has no universal statutory child allowance; family benefits are means-tested/scheme-based -> 0 for an ICP household. Peer-inherited (India national field).", "last_updated": "2026-05-31", "confidence": "high" }, "non_cash._summary": { "source": "India peer (Bangalore) statutory baseline: ~18 earned-leave days, ~15 public holidays (varies by state), 26-week paid maternity (Maternity Benefit Act), employer/ESI sick leave; private insurance not universal -> universal_healthcare false. ILO; Maternity Benefit Act 1961 (amended 2017)", "last_updated": "2026-05-31", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Delhi NCR: INR 2,123,815, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "INR 2,123,815 / 95.48 = USD 22,244; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.0890, R5-rounded down to 0.08. REPLACES 0.21 (2026-05-31), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Greater Delhi Area taken over the higher Gurgaon page (INR 2,577,355), which is the lower and therefore conservative read of the NCR. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.22; set just below Bangalore (0.24) and above Chennai/Hyderabad-tier metros." }, "tax.filing_basis": { "source": "Income-tax Act - individual assessee; no joint/spousal returns", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "tax.surcharge_brackets": { "source": "IN-1: new-regime high-earner surcharge on the income tax; 0 (<50L) / 10% (50L-1cr) / 15% (1cr-2cr) / 25% (>2cr, the new-regime cap), applied as a step on total income with marginal relief; 4% cess then runs on tax+surcharge. PwC IN taxes-on-personal-income; Taxmann surcharge/cess.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.social_security_cap": { "source": "EPFO International Worker framework (EPF Scheme 1952 para 83, 2008 amendment, carried forward into the EPF Scheme, 2026 notified 1 July 2026 under the Code on Social Security), described identically across AscentHR, cgaa.org, gconnect.in's EPFO brochure summary and taxaaram.com; SSA list per Dezan Shira / ExpatOrbit, UK signing 10 Feb 2026 per KPMG Flash Alert; US absence per india-briefing.com (\"pursuing\"). epfindia.gov.in itself was unreachable on two separate attempts (DNS, then connection timeout on re-verification 2026-08-06); the gap is network access from our environment, not thin sourcing. docs/superpowers/research/2026-08-06-forced-savings-foreigner-rules.md", "last_updated": "2026-08-06", "confidence": "medium-high", "note": "The INR 15,000/mo wage ceiling (cap INR 180,000/yr, max employee contribution INR 21,600) is the DOMESTIC basis. A foreign-passport International Worker at an EPF-covered establishment contributes 12% on FULL salary with no ceiling, unless a home-country SSA plus Certificate of Coverage exempts them. Engine seam: computeIndiaTaxes reads state.in_epf_intl_worker, absent or false = this capped basis, true = uncapped 12%. The target side of a comparison defaults to the uncapped IW basis when the SOURCE country holds no SSA with India (INDIA_SSA_COUNTRIES in app.js; the US has none, the UK signed Feb 2026); SSA-source movers keep this capped basis rather than the full COC exemption, which is paperwork-gated (Chile Law 18.156 precedent, conservative rule). At INR 5,000,000 gross the two bases are 21,600 vs 600,000 of employee EPF. Confidence medium-high: the mechanism is uniform across independent sources but the EPFO primary text was not reached; a Karnataka HC challenge to the uncapped rule exists and EPFO has stated the framework stands. TODO (2026-08-06): the EPF Scheme, 2026 applies the INR 15,000 ceiling to international workers in some provisions while others still read \"total wages\"; EPFO called this a typographical and drafting inconsistency, said no proposal exists to change the international-worker framework, and is weighing a corrigendum (Business Standard, 16 July 2026). Re-check this cap once the corrigendum issues. Re-checked 2026-08-07: the corrigendum has still NOT issued (latest reporting remains Business Standard 2026-07-16; EPFO maintains the IW framework is carried forward unchanged, so the uncapped IW default stands), and the DOMESTIC INR 15,000/mo ceiling is also confirmed still in force (the proposed rise to 21k is delayed indefinitely; the SC four-month directive of Jan 2026 has produced no enacted decision). The watch now rides docs/data-refresh-calendar.md rather than this dated note." }, "non_cash.parental_leave_weeks": { "source": "Maternity Benefit Act 1961 s.5(3) as amended 2017: 26 weeks at 100% of average daily wage, employer-paid; Act text via state-government mirror (wb.gov.in PDF; central gov.in domains unreachable at check, flagged)", "last_updated": "2026-08-06", "confidence": "high", "note": "26 = maternity only, and on the household-inclusive basis this IS the household total: India has no statutory paid paternity or parental leave for private-sector employees as of 2026 (CCS Rule 43-A covers government employees only; the 2017 Paternity Benefit Bill lapsed). Caveats: establishments with 10+ employees; 26 weeks for the first two children, 12 for the third and beyond (we store the first-child case). Confirmed on the basis 2026-08-06. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "Maternity Benefit Act 1961 s.5(3) as amended 2017: 26 weeks at 100% of average daily wage, employer-paid; Act text via state-government mirror (wb.gov.in PDF; central gov.in domains unreachable at check, flagged)", "last_updated": "2026-08-06", "confidence": "high", "note": "100% of the average daily wage (3-month average), employer-paid, no monetary cap. Confirmed 2026-08-06." }, "non_cash.vacation_days_statutory": { "source": "Tamil Nadu Shops and Establishments Act 1947 s.25; Delhi Shops and Establishments Act 1954; Telangana Shops and Establishments Act 1988 s.30; Karnataka Shops and Commercial Establishments Act 1961 s.15; Maharashtra Shops and Establishments Act 2017 s.18(3); Factories Act 1948 s.79", "last_updated": "2026-08-08", "confidence": "medium", "note": "CORRECTED 2026-08-08, 18 to 12. India has NO national annual-leave statute for office workers, so this nationally-enforced field has to pick one number for six cities in five states. The old 18 was sourced to Karnataka, and Karnataka's own Shops and Commercial Establishments Act 1961 s.15 does not say 18: it grants one day per twenty days worked, the same accrual formula as Factories Act 1948 s.79, whose administered output is about 15. So 18 was unsupported by the statute it cited. Verified range across the states we actually model: Tamil Nadu (Chennai) s.25 gives a flat 12, Delhi 15, Telangana (Hyderabad) s.30 a flat 15, and Karnataka (Bangalore) plus Maharashtra (Mumbai, Pune) s.18(3) both run the 1-in-20 accrual that lands near 15. We store 12, the lower end, per rule 5 and because Tamil Nadu's is a clean section-specific floor covering one of our own cities rather than an extrapolation. LIVE ALTERNATIVE: 15 is equally defensible and is the modal figure across five of the six cities; flip this if the owner prefers representativeness to conservatism. Also checked: the OSH Code 2020 s.32 is in force from 2025-11-21 with rules from May 2026, but it is scoped to factory-type establishments and does not yet displace the state Shops Acts for commercial or IT employers. No six-day or calendar conversion applies: the flat states state days directly, and the accrual denominator counts actual attendance days." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series", "last_updated": "2026-08-08", "confidence": "low", "note": "Almost all Indian health spending is out of pocket, and employer group cover is a premium that is not booked here. Two city tiers preserved. FLAGGED for a future look: claims and premium data both hint row 1 may be low for a Bangalore or Mumbai private-hospital household. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained, and the weakest Indian row. Geography is the National Capital Region, much larger than Delhi proper. The only quotable figure (MoveInSync 60 min 2019, about 64 in 2023) sits 10 to 14 minutes above the stored value, so this one is most likely genuinely low." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-F.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Income Tax Department, New tax regime vs old tax regime FAQs https://www.incometax.gov.in/iec/foportal/help/new-tax-vs-old-tax-regime-faqs", "last_updated": "2026-08-11", "confidence": "high", "note": "No deduction under the regime the engine models. The NEW regime (s.115BAC) is the statutory default from AY 2024-25 and disallows Chapter-VIA deductions except 80CCD(2), 80CCH and 80JJAA, so 80C (where EPF would sit) and the extra 80CCD(1B) NPS deduction are both unavailable; 80CCD(2) is EMPLOYER NPS money, not the employee's own. computeIndiaTaxes models the new regime (slabs shifted by the INR 75,000 standard deduction) and charges EPF in payroll without removing it from the base, which is correct there. If an old-regime path is ever added, the value there is INR 200,000 and it must net off the EPF already flowing through payroll." } }, "chennai": { "tax.federal_brackets": { "source": "India FY 2026-27 new regime slabs (5/10/15/20/25/30%) + 4% cess, India peer ladder (same as Bangalore). Boundaries corrected 2026-08-01 (see the Bangalore tax.federal_brackets note): the INR 75,000 standard deduction is folded in by shifting EVERY slab and surcharge boundary, not just the first. Income Tax Dept https://www.incometax.gov.in/iec/foportal/help/individual/return-applicable-1 ; PwC India", "last_updated": "2026-05-31", "confidence": "high" }, "tax.social_security_rate": { "source": "Employee EPF 12% capped at INR 15,000/mo wage base (social_security_cap 180,000/yr). EPFO; PwC India", "last_updated": "2026-05-31", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Greater Chennai Corporation property tax on owners ~0.3% effective (renters pay 0). GCC property-tax schedules", "last_updated": "2026-05-31", "confidence": "low" }, "housing._summary": { "source": "Numbeo Chennai 2026 (OMR / Adyar / Anna Nagar professional tier) apartment rent + price/sqft; cheaper than Delhi/Bangalore. INR", "last_updated": "2026-05-31", "confidence": "medium" }, "housing.rent_percentiles": { "source": "MagicBricks listing-count CDF, 2026-08-18 (SERVER_PRELOADED_STATE_ resultCount; card price = monthly rent, deposit separate as bookingAmtExact, cap inclusive on the rent field, instrument re-verified this run; apartments only; asking INR/month): 2 BHK P25/med/P75 -> 0.687/1.347 (n=1,433); 1 BHK 0.602/1.594 (n=337, thin).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW on both edges (2 BHK -0.133/+0.107; the thin 1 BHK slice is noisier but far beyond noise at 0.602/1.594). Applied 0.69/1.35 (nearest). Single portal; informal-tail truncation means the true citywide spread is wider - conservative. 2 BHK anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-india.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (E). NOTE the stored medians sit near the portal pool's ~p70 - a level observation, flagged, out of scope here." }, "healthcare._summary": { "source": "Private health insurance + OOP for a professional family; premiums by family size from Indian insurer 2025-26 quotes + Numbeo OOP. INR", "last_updated": "2026-05-31", "confidence": "low" }, "childcare._summary": { "source": "Chennai private daycare + preschool + after-school 2025-26 (professional tier); no means-tested cash subsidy for an ICP household -> subsidy 0. INR", "last_updated": "2026-05-31", "confidence": "low" }, "food._summary": { "source": "Numbeo Chennai 2026 grocery + restaurant indices (professional basket). INR", "last_updated": "2026-05-31", "confidence": "medium" }, "transit._summary": { "source": "Chennai Metro/MTC monthly-pass equivalent + car (fuel/insurance/depreciation); Numbeo 2025-26. INR", "last_updated": "2026-05-31", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Chennai leisure + domestic vacation baseline 2025-26. INR", "last_updated": "2026-05-31", "confidence": "low" }, "travel_home._summary": { "source": "INR->US/Europe economy round-trip 2025-26 (Air India / Vistara / Gulf-carrier connections). INR", "last_updated": "2026-05-31", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No universal statutory child allowance in India -> 0 for an ICP household. Peer-inherited (India national field).", "last_updated": "2026-05-31", "confidence": "high" }, "non_cash._summary": { "source": "India peer statutory baseline: ~18 earned-leave days, ~15 public holidays (Tamil Nadu calendar), 26-week paid maternity, employer/ESI sick leave; private insurance not universal -> universal_healthcare false. ILO; Maternity Benefit Act", "last_updated": "2026-05-31", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Chennai: INR 1,816,475, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-apac.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "INR 1,816,475 / 95.48 = USD 19,025; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.0761, R5-rounded down to 0.07. REPLACES 0.18 (2026-05-31), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.18; set below Delhi NCR (0.22)/Bangalore (0.24), in line with Hyderabad/Pune tier." }, "tax.filing_basis": { "source": "Income-tax Act - individual assessee; no joint/spousal returns", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "tax.surcharge_brackets": { "source": "IN-1: new-regime high-earner surcharge on the income tax; 0 (<50L) / 10% (50L-1cr) / 15% (1cr-2cr) / 25% (>2cr, the new-regime cap), applied as a step on total income with marginal relief; 4% cess then runs on tax+surcharge. PwC IN taxes-on-personal-income; Taxmann surcharge/cess.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.social_security_cap": { "source": "EPFO International Worker framework (EPF Scheme 1952 para 83, 2008 amendment, carried forward into the EPF Scheme, 2026 notified 1 July 2026 under the Code on Social Security), described identically across AscentHR, cgaa.org, gconnect.in's EPFO brochure summary and taxaaram.com; SSA list per Dezan Shira / ExpatOrbit, UK signing 10 Feb 2026 per KPMG Flash Alert; US absence per india-briefing.com (\"pursuing\"). epfindia.gov.in itself was unreachable on two separate attempts (DNS, then connection timeout on re-verification 2026-08-06); the gap is network access from our environment, not thin sourcing. docs/superpowers/research/2026-08-06-forced-savings-foreigner-rules.md", "last_updated": "2026-08-06", "confidence": "medium-high", "note": "The INR 15,000/mo wage ceiling (cap INR 180,000/yr, max employee contribution INR 21,600) is the DOMESTIC basis. A foreign-passport International Worker at an EPF-covered establishment contributes 12% on FULL salary with no ceiling, unless a home-country SSA plus Certificate of Coverage exempts them. Engine seam: computeIndiaTaxes reads state.in_epf_intl_worker, absent or false = this capped basis, true = uncapped 12%. The target side of a comparison defaults to the uncapped IW basis when the SOURCE country holds no SSA with India (INDIA_SSA_COUNTRIES in app.js; the US has none, the UK signed Feb 2026); SSA-source movers keep this capped basis rather than the full COC exemption, which is paperwork-gated (Chile Law 18.156 precedent, conservative rule). At INR 5,000,000 gross the two bases are 21,600 vs 600,000 of employee EPF. Confidence medium-high: the mechanism is uniform across independent sources but the EPFO primary text was not reached; a Karnataka HC challenge to the uncapped rule exists and EPFO has stated the framework stands. TODO (2026-08-06): the EPF Scheme, 2026 applies the INR 15,000 ceiling to international workers in some provisions while others still read \"total wages\"; EPFO called this a typographical and drafting inconsistency, said no proposal exists to change the international-worker framework, and is weighing a corrigendum (Business Standard, 16 July 2026). Re-check this cap once the corrigendum issues. Re-checked 2026-08-07: the corrigendum has still NOT issued (latest reporting remains Business Standard 2026-07-16; EPFO maintains the IW framework is carried forward unchanged, so the uncapped IW default stands), and the DOMESTIC INR 15,000/mo ceiling is also confirmed still in force (the proposed rise to 21k is delayed indefinitely; the SC four-month directive of Jan 2026 has produced no enacted decision). The watch now rides docs/data-refresh-calendar.md rather than this dated note." }, "non_cash.parental_leave_weeks": { "source": "Maternity Benefit Act 1961 s.5(3) as amended 2017: 26 weeks at 100% of average daily wage, employer-paid; Act text via state-government mirror (wb.gov.in PDF; central gov.in domains unreachable at check, flagged)", "last_updated": "2026-08-06", "confidence": "high", "note": "26 = maternity only, and on the household-inclusive basis this IS the household total: India has no statutory paid paternity or parental leave for private-sector employees as of 2026 (CCS Rule 43-A covers government employees only; the 2017 Paternity Benefit Bill lapsed). Caveats: establishments with 10+ employees; 26 weeks for the first two children, 12 for the third and beyond (we store the first-child case). Confirmed on the basis 2026-08-06. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "Maternity Benefit Act 1961 s.5(3) as amended 2017: 26 weeks at 100% of average daily wage, employer-paid; Act text via state-government mirror (wb.gov.in PDF; central gov.in domains unreachable at check, flagged)", "last_updated": "2026-08-06", "confidence": "high", "note": "100% of the average daily wage (3-month average), employer-paid, no monetary cap. Confirmed 2026-08-06." }, "non_cash.vacation_days_statutory": { "source": "Tamil Nadu Shops and Establishments Act 1947 s.25; Delhi Shops and Establishments Act 1954; Telangana Shops and Establishments Act 1988 s.30; Karnataka Shops and Commercial Establishments Act 1961 s.15; Maharashtra Shops and Establishments Act 2017 s.18(3); Factories Act 1948 s.79", "last_updated": "2026-08-08", "confidence": "medium", "note": "CORRECTED 2026-08-08, 18 to 12. India has NO national annual-leave statute for office workers, so this nationally-enforced field has to pick one number for six cities in five states. The old 18 was sourced to Karnataka, and Karnataka's own Shops and Commercial Establishments Act 1961 s.15 does not say 18: it grants one day per twenty days worked, the same accrual formula as Factories Act 1948 s.79, whose administered output is about 15. So 18 was unsupported by the statute it cited. Verified range across the states we actually model: Tamil Nadu (Chennai) s.25 gives a flat 12, Delhi 15, Telangana (Hyderabad) s.30 a flat 15, and Karnataka (Bangalore) plus Maharashtra (Mumbai, Pune) s.18(3) both run the 1-in-20 accrual that lands near 15. We store 12, the lower end, per rule 5 and because Tamil Nadu's is a clean section-specific floor covering one of our own cities rather than an extrapolation. LIVE ALTERNATIVE: 15 is equally defensible and is the modal figure across five of the six cities; flip this if the owner prefers representativeness to conservatism. Also checked: the OSH Code 2020 s.32 is in force from 2025-11-21 with rules from May 2026, but it is scoped to factory-type establishments and does not yet displace the state Shops Acts for commercial or IT employers. No six-day or calendar conversion applies: the flat states state days directly, and the accrual denominator counts actual attendance days." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series", "last_updated": "2026-08-08", "confidence": "low", "note": "Almost all Indian health spending is out of pocket, and employer group cover is a premium that is not booked here. Two city tiers preserved. FLAGGED for a future look: claims and premium data both hint row 1 may be low for a Bangalore or Mumbai private-hospital household. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Same India-wide gap. MoveInSync telematics gives 50 (2019) and 59 (2023) on a biased subpopulation; not cited." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-F.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Income Tax Department, New tax regime vs old tax regime FAQs https://www.incometax.gov.in/iec/foportal/help/new-tax-vs-old-tax-regime-faqs", "last_updated": "2026-08-11", "confidence": "high", "note": "No deduction under the regime the engine models. The NEW regime (s.115BAC) is the statutory default from AY 2024-25 and disallows Chapter-VIA deductions except 80CCD(2), 80CCH and 80JJAA, so 80C (where EPF would sit) and the extra 80CCD(1B) NPS deduction are both unavailable; 80CCD(2) is EMPLOYER NPS money, not the employee's own. computeIndiaTaxes models the new regime (slabs shifted by the INR 75,000 standard deduction) and charges EPF in payroll without removing it from the base, which is correct there. If an old-regime path is ever added, the value there is INR 200,000 and it must net off the EPF already flowing through payroll." } }, "abu-dhabi": { "tax.federal_brackets": { "source": "UAE has 0% personal income tax (no federal or emirate-level PIT on wages); modeled as a single 0% band like Dubai. UAE Ministry of Finance; PwC Worldwide Tax Summaries - UAE", "last_updated": "2026-05-31", "confidence": "high" }, "tax.social_security_rate": { "source": "social_security_rate 0 for expatriate employees (GPSSA pension applies only to UAE/GCC nationals). PwC UAE", "last_updated": "2026-05-31", "confidence": "high" }, "tax.sales_tax_rate": { "source": "UAE federal VAT 5% (since 1 Jan 2018). Federal Tax Authority; PwC UAE VAT", "last_updated": "2026-05-31", "confidence": "high" }, "housing._summary": { "source": "Numbeo / Bayut Abu Dhabi 2026 (Corniche / Reem Island / Al Reef professional tier) apartment rent + price. AED", "last_updated": "2026-05-31", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Mandatory private health insurance for residents (employer-provided / DOH-regulated); premiums by family size + OOP from Abu Dhabi market 2025-26. AED", "last_updated": "2026-05-31", "confidence": "low" }, "childcare._summary": { "source": "Abu Dhabi private nursery + KG + after-school 2025-26 (expat professional tier); no cash subsidy for expats -> subsidy 0. AED", "last_updated": "2026-05-31", "confidence": "low" }, "food._summary": { "source": "Numbeo Abu Dhabi 2026 grocery + restaurant indices (professional basket). AED", "last_updated": "2026-05-31", "confidence": "medium" }, "transit._summary": { "source": "Abu Dhabi bus monthly pass + car (cheap fuel, car-centric city) depreciation/insurance; Numbeo 2025-26. AED", "last_updated": "2026-05-31", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Abu Dhabi leisure + regional vacation baseline 2025-26. AED", "last_updated": "2026-05-31", "confidence": "low" }, "travel_home._summary": { "source": "AED->US/Europe economy round-trip 2025-26 (Etihad / Emirates). AED", "last_updated": "2026-05-31", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No child cash allowance for expatriate residents -> 0. Peer-inherited (UAE national field).", "last_updated": "2026-05-31", "confidence": "high" }, "non_cash._summary": { "source": "UAE Labour Law (Federal Decree-Law 33/2021): 30 calendar-day annual leave, ~12 public holidays, 60-day maternity (45 full + 15 half), 90-day annual sick-leave schedule; expats rely on private insurance -> universal_healthcare false. UAE MOHRE; ILO", "last_updated": "2026-05-31", "confidence": "medium" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Abu Dhabi. Its basis is converted with the rest of UAE and its rank inside UAE is preserved. Prior source: levels.fyi / relocate.me Abu Dhabi tech 2025-26; sovereign-wealth + govt-tech (G42 / ADNOC digital) employers pay strong tax-free packages, slightly below Dubai's broader market", "last_updated": "2026-08-11", "confidence": "low", "note": "0.8 x 0.4118 = 0.3294, R5-rounded down to 0.32. 0.4118 is the MEDIAN of (re-derived / previous) across the 1 UAE city that DO carry an all-levels measurement (Dubai 0.85 to 0.35), so Abu Dhabi moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Abu Dhabi's rank inside UAE is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.80; set just below Dubai (0.82-equivalent peer) on the tax-free package premium." }, "non_cash.parental_leave_weeks": { "source": "Federal Decree-Law 33/2021 art. 30 (maternity 60 days: 45 full + 15 half pay) and art. 32 (parental leave 5 working days each parent); u.ae official pages re-fetched 2026-08-07. Research: docs/superpowers/research/2026-08-07-pl-b5-africa-gulf-sea.md, 2026-08-06-parental-caps-asia-mideast.md (UAE), 2026-08-06-pl-caps-gulf.md", "last_updated": "2026-08-07", "confidence": "high", "note": "10.57 wk = 60 calendar days maternity (8.57 wk) + 5 working days parental leave PER PARENT (2 x 1.0 wk): art. 32 grants the 5 days to EACH parent (male or female employee, u.ae official reading re-verified 2026-08-07), so the mother's own 5 days count too, and working days convert /5 per the house convention (the old 9.29 dropped the mother's days AND converted the father's /7). Batch-5 restatement on the household-inclusive basis; research: docs/superpowers/research/2026-08-07-pl-b5-africa-gulf-sea.md. Unpaid extensions (45 illness days) excluded." }, "non_cash.parental_leave_paid_pct": { "source": "Federal Decree-Law 33/2021 art. 30: 45 days full pay + 15 days half pay; art. 32: 5 working days full pay per parent. Employer-paid wage continuation, no social-insurance intermediation for expats (GPSSA is nationals-only pension)", "last_updated": "2026-08-07", "confidence": "high", "note": "89.9 blended = (6.43 wk x 100 + 2.14 wk x 50 + 2.0 wk x 100) / 10.57 (maternity 45d full + 15d half, calendar; parental 5 working days x 2 parents at full pay). Was 88.5 on the 65-calendar-day construction that dropped the mother's art. 32 days. Confirmed genuinely UNCAPPED on the 2026-08-06 cap sweep, re-verified 2026-08-07: direct employer continuation of the actual contractual wage at any salary, no tenure gate. Minor open detail: whether \"full pay\" includes fixed allowances rests on one secondary source; does not affect the cap verdict." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "DHA and HAAD mandatory employer-provided cover with its published co-payment and co-payment-cap structure", "last_updated": "2026-08-08", "confidence": "medium", "note": "Employer-provided cover is legally mandatory for the employee, so the tier-1 premium is 0 and dependant coverage is a PREMIUM; this field carries the co-payment only, verified co-pay-sized rather than premium-sized. Two emirate values preserved. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "DHA and HAAD mandatory employer-provided cover with its published co-payment and co-payment-cap structure", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. Not reached within the sweep; no evidence either way. Leads: Integrated Transport Centre and SCAD." }, "non_cash.sick_leave": { "source": "Federal Decree-Law No. 33 of 2021 (Regulating Labour Relations), Article 31.; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-H.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "PwC Worldwide Tax Summaries, United Arab Emirates (reviewed 12 Mar 2026): \"There is currently no personal income tax in the United Arab Emirates\"", "last_updated": "2026-08-11", "confidence": "high", "note": "Recorded for explicitness only; it changes no output by construction. computeUAETaxes returns { federal: 0, payroll: 0, state: 0, total: 0 } and never reads taxable_income, so no deduction ceiling can move a UAE number. A deduction from a zero-rate base is meaningless." } }, "hanoi": { "tax.filing_basis": { "source": "PwC WWTS - Vietnam; Vietnam.vn: no joint declaration mechanism; each spouse is assessed separately (joint filing is only a legislative proposal).", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Vietnam 2026 reformed 5-bracket PIT (5/15/25/30/35%) on income net of the personal deduction (VND 15.5M/mo self) + dependent deduction, modeled via standard_deduction 186M single / 260.4M married. Same engine as Ho Chi Minh City. PwC Vietnam; General Department of Taxation", "last_updated": "2026-05-31", "confidence": "high" }, "tax.social_security_rate": { "source": "Employee SI 8% + HI 1.5% = 9.5% capped at 20x base salary (cap 561.6M/yr) + UI 1% capped at 20x regional minimum (1,274.4M/yr); all deductible. Vietnam Social Security; PwC Vietnam", "last_updated": "2026-05-31", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Vietnam VAT standard 10% (8% temporary cut on some goods through 2026; modeled standard). PwC Vietnam VAT", "last_updated": "2026-05-31", "confidence": "high" }, "housing._summary": { "source": "Numbeo / batdongsan Hanoi 2026 (Tay Ho / Ba Dinh / Cau Giay professional tier) apartment rent + price. VND", "last_updated": "2026-05-31", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Private insurance + OOP over the BHYT public base for a professional family; premiums by family size + Numbeo OOP, Hanoi 2025-26. VND", "last_updated": "2026-05-31", "confidence": "low" }, "childcare._summary": { "source": "Hanoi private daycare + preschool + after-school 2025-26 (professional tier); no cash subsidy for an ICP household -> subsidy 0. VND", "last_updated": "2026-05-31", "confidence": "low" }, "food._summary": { "source": "Numbeo Hanoi 2026 grocery + restaurant indices (professional basket). VND", "last_updated": "2026-05-31", "confidence": "medium" }, "transit._summary": { "source": "Hanoi metro/bus monthly pass + motorbike/car running costs; Numbeo 2025-26. VND", "last_updated": "2026-05-31", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Hanoi leisure + domestic vacation (Ha Long / Sapa) baseline 2025-26. VND", "last_updated": "2026-05-31", "confidence": "low" }, "travel_home._summary": { "source": "VND->US/Europe economy round-trip 2025-26 (Vietnam Airlines / Gulf-carrier connections). VND", "last_updated": "2026-05-31", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "No universal child cash allowance in Vietnam -> 0. Peer-inherited (Vietnam national field).", "last_updated": "2026-05-31", "confidence": "high" }, "non_cash._summary": { "source": "Vietnam Labour Code 2019 peer (HCMC): 12 annual-leave days, ~11 public holidays, 6-month paid maternity via VSS, 30-40 day sick leave; BHYT is a genuine public base -> universal_healthcare true. ILO; Labour Code 2019", "last_updated": "2026-05-31", "confidence": "medium" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Hanoi. Its basis is converted with the rest of Vietnam and its rank inside Vietnam is preserved. Prior source: levels.fyi / ITviec Hanoi tech 2025-26; the capital's IT-outsourcing + product scene pays a touch below HCMC", "last_updated": "2026-08-11", "confidence": "low", "note": "0.13 x 0.3571 = 0.0464, R5-rounded down to 0.04. 0.3571 is the MEDIAN of (re-derived / previous) across the 1 Vietnam city that DO carry an all-levels measurement (Ho Chi Minh City 0.14 to 0.05), so Hanoi moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Hanoi's rank inside Vietnam is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.13; set just below HCMC peer on the slightly smaller senior market." }, "non_cash.parental_leave_weeks": { "source": "Law on Social Insurance 41/2024 (in force 2025-07-01) art. 53: paternity \"05 ngày làm việc\" (working days), full statute text via the government portal, fetched 2026-08-07; 6 months maternity. Research: docs/superpowers/research/2026-08-07-pl-b5-africa-gulf-sea.md, 2026-08-06-parental-caps-asia-mideast.md (Vietnam), 2026-08-06-pl-caps-southeast-asia.md", "last_updated": "2026-08-07", "confidence": "high", "note": "27.0 wk = 26.0 maternity (mother, 6 months) + 1.0 paternity (5 WORKING days / 5 per the house convention; the prior 0.71 used the calendar /7 conversion, corrected in batch 5 on the statute's own 'ngày làm việc' wording; C-section/twin uplifts excluded per the basis convention). The 6-in-12-months contribution gate passes for an employed-on-arrival mover by the birth date. The one-off 2x-reference-level birth grant (art. 58(4), VND 5.06M) is a lump allowance, not wage replacement, and stays excluded (Japan precedent)." }, "non_cash.parental_leave_paid_pct": { "source": "Law on Social Insurance: benefit = 100% of the average SI-contribution salary of the preceding 6 months", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 26.71 weeks into non_cash.parental_leave_capped (100% of an SI wage capped at 20x base salary), so no percentage-paid weeks remain on this field." }, "non_cash.parental_leave_capped": { "source": "Decree 161/2026/ND-CP: reference level VND 2,530,000/month from 1 July 2026; SI contribution-and-benefit wage ceiling = 20x = VND 50,600,000/month (multi-source corroborated: vinatpt.com, vietnam.vn, laodong.vn; re-verified 2026-08-07). Research: docs/superpowers/research/2026-08-07-pl-b5-africa-gulf-sea.md, 2026-08-06-parental-caps-asia-mideast.md (Vietnam)", "last_updated": "2026-08-07", "confidence": "high", "note": "All 27.0 weeks pay 100% of an SI-contribution wage structurally capped at VND 50,600,000/month (the ceiling flows through because premiums cannot be levied above it; 20x the Decree 161/2026 reference level VND 2,530,000, re-verified 2026-08-07). Binds only above ~VND 607M/yr, the upper half of the Hanoi/HCMC tech band: at VND 900M/yr the effective rate is ~67.5%, below it the 100% is genuine. cap_amount doubles as the payout ceiling because the rate is 100. The ceiling moves with the reference level (last step 1 Jul 2026). Weeks 26.71 -> 27.0 in batch 5 (paternity working-day conversion)." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "BHYT cost-sharing structure, including a 0% co-payment for children under age 6", "last_updated": "2026-08-08", "confidence": "medium", "note": "Child-discount shape applied on a directly-sourced government child exemption. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "BHYT cost-sharing structure, including a 0% co-payment for children under age 6", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. GSO publishes nothing and the HAIDEP household survey was not reached. Moovit was not used because Hanoi is motorcycle-dominated, which makes a public-transport-riders series unrepresentative." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-E.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Decree 253/2026/ND-CP and Circular 87/2026/TT-BTC (issued 30 June 2026, effective 1 July 2026), via Baker McKenzie https://www.bakermckenzie.com/en/insight/publications/2026/07/vietnam-decree-circular-implementing-personal-income-tax-law and corroborated by KPMG, Vialto, Forvis Mazars and Parker Russell", "last_updated": "2026-08-11", "confidence": "high", "note": "Voluntary pension fund contributions, deductible up to VND 3,000,000 per MONTH, annualised x12 here. LEAD REFUTED: the VND 1,000,000/month in Circular 111/2013 is superseded and the ceiling TRIPLED for the 2026 tax year. The 3m/month is shared with life insurance premiums AND with any employer contribution to the same fund; 36m assumes neither. Binds hard at every senior HCMC/Hanoi salary. Decree 253 itself could not be fetched (thuvienphapluat and luatvietnam both 403), hence the law-firm citation chain." } }, "johannesburg": { "tax.filing_basis": { "source": "PwC WWTS - South Africa (Tax administration): spouses are taxed separately and must submit separate returns.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "SARS 2025-26 PIT 7-bracket 18-45%; primary rebate R17,235 applied (tax_credit). Same engine as Cape Town. SARS; PwC South Africa", "last_updated": "2026-05-31", "confidence": "high" }, "tax.social_security_rate": { "source": "Employee UIF 1% capped at R17,712/mo contribution ceiling. Department of Employment and Labour; PwC South Africa", "last_updated": "2026-05-31", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "City of Johannesburg municipal rates on owners (renters in golden personas pay 0); nominal. CoJ rates policy 2025-26", "last_updated": "2026-05-31", "confidence": "low" }, "housing._summary": { "source": "Numbeo / Property24 Johannesburg 2026 (Sandton / Rosebank / Bryanston secure-complex professional tier) rent + price. ZAR", "last_updated": "2026-05-31", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "Private medical scheme (Discovery/Bonitas tier) + OOP for a professional family; the public base is thin for professionals; premiums by family size 2025-26. ZAR", "last_updated": "2026-05-31", "confidence": "low" }, "childcare._summary": { "source": "Johannesburg private creche + pre-primary + aftercare 2025-26 (professional tier); SA child grant is means-tested -> subsidy 0 for an ICP household. ZAR", "last_updated": "2026-05-31", "confidence": "low" }, "food._summary": { "source": "Numbeo Johannesburg 2026 grocery + restaurant indices (professional basket). ZAR", "last_updated": "2026-05-31", "confidence": "medium" }, "transit._summary": { "source": "Gautrain/bus monthly-pass equivalent + car (car near-mandatory); Numbeo 2025-26. ZAR", "last_updated": "2026-05-31", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Johannesburg leisure + domestic vacation baseline 2025-26. ZAR", "last_updated": "2026-05-31", "confidence": "low" }, "travel_home._summary": { "source": "ZAR->US/Europe economy round-trip 2025-26 (SAA / Gulf-carrier / European connections). ZAR", "last_updated": "2026-05-31", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "SA Child Support Grant is means-tested -> 0 for an ICP household. Peer-inherited (South Africa national field).", "last_updated": "2026-05-31", "confidence": "high" }, "non_cash._summary": { "source": "South Africa BCEA peer (Cape Town): 21 consecutive-day (15 working-day) annual leave, 12 public holidays, 4-month maternity via UIF, ~30-day sick leave per 3-yr cycle; NHI not yet operational -> universal_healthcare false. ILO; BCEA", "last_updated": "2026-05-31", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Johannesburg: ZAR 537,438, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-americas-mea.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "ZAR 537,438 / 16.49 = USD 32,592; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.1304, R5-rounded down to 0.13. REPLACES 0.18 (2026-05-31), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Chosen as the South African anchor over Cape Town (ZAR 857,488 on the same day) because it is the lower and more domestic read; a 37% same-day gap between two ZA metros is sample composition, not geography. South African pay is quoted as cost-to-company, which conventionally includes employer pension and medical and is NOT the levels.fyi basis, so this may be inflated. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.18; aligned with Cape Town peer." }, "non_cash.parental_leave_weeks": { "source": "UI Act 63/2001 s.13(3)(a) (1 credit day per 5 employed days, 365-day cap) + s.13(5); BCEA s25A (10 consecutive days parental leave) + UI Amendment Act 10/2016 s.12(3)(c) (parental benefit at the flat 66%); Van Wyk v Minister of Employment and Labour [2025] ZACC 20", "last_updated": "2026-08-07", "confidence": "medium-high", "note": "11.86 wk = 10.43 maternity + 1.43 s25A parental (10 consecutive calendar days, the father's own UIF-paid entitlement), MOVER-PRICED per the standing zero-tenure ruling: a ~12-month mover accrues ~73 UIF credit days (1 per 5 employed days), below the 121-day maternity maximum, so the credits are the binding number. s.13(5) prevents maternity draws from consuming credits for later unemployment; it does not exempt maternity from the accrual limit (statute reading corroborated by Leave Network + 2026 practitioner consensus, hence medium-high). Tenured alternative 18.71 wk recorded in the flat-block note. Was 17, the un-gated maternity maximum with the father's s25A days dropped. Household-inclusive basis; research: docs/superpowers/research/2026-08-07-pl-b5-africa-gulf-sea.md." }, "non_cash.parental_leave_paid_pct": { "source": "Unemployment Insurance Amendment Act 10 of 2016 (maternity benefits at a flat 66% of earnings, off the 38-60% unemployment IRR scale; parental leave benefit at the same flat 66% per s.12(3)(c)); UIF earnings ceiling ZAR 17,712/month, unchanged since 1 June 2021", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild converted South Africa to the flat shape because the UIF ceiling (ZAR 212,544/yr) sits 5-13x below any ICP salary here, so the benefit is always the capped amount. The 38 this replaces was ALSO the wrong statutory rate: it was the unemployment-benefit IRR floor, not the flat 66% maternity rate the 2016 amendment moved maternity onto. Correctness bug recorded in docs/corrections-ledger.md (internal:za-parental-38pct)." }, "non_cash.parental_leave_flat": { "source": "UI Act 63/2001 as amended by Act 10/2016: maternity at 66% of earnings up to the UIF earnings ceiling of ZAR 17,712/month = ZAR 212,544/yr (unchanged since 1 June 2021; equals the stored tax.social_security_cap, an independent internal corroboration). Research: docs/superpowers/research/2026-08-06-parental-caps-americas-africa.md (South Africa) and docs/superpowers/research/2026-08-06-parental-caps-southern-east-africa.md", "last_updated": "2026-08-06", "confidence": "high", "note": "ZAR 2,697.67/week = 66% x (212,544 / 52), covering all 11.86 weeks: 10.43 wk maternity + 1.43 wk BCEA s25A parental leave (10 consecutive days, the other parent's own entitlement, UIF-paid at the same flat 66% per UI Amendment Act 10/2016 s.12(3)(c) — the batch-5 audit stored the previously flagged follow-up). The cap always binds at ICP salaries (~5.7% effective at the $150k basis). Weeks are MOVER-PRICED (2026-08-07): UIF benefits are credit-day limited, s.13(3)(a) accrues 1 benefit day per 5 employed days, and s.13(5) only stops maternity CONSUMING credits, it does not exempt maternity from the accrual limit; a ~12-month mover holds ~73 credit days = 10.43 wk, not the 121-day (17.29 wk) statutory maximum. Tenured alternative (605+ employed days, e.g. a second child in-country): 17.29 + 1.43 = 18.71 wk, recorded not stored. Van Wyk watch item stands: ConCourt [2025] ZACC 20 (3 Oct 2025) confirmed the gendered-structure invalidity, suspended ~36 months to around Oct 2028; current structure remains law in force, re-check before 2028. Research: docs/superpowers/research/2026-08-07-pl-b5-africa-gulf-sea.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita plus medical-scheme co-payment and self-payment-gap structure", "last_updated": "2026-08-08", "confidence": "medium", "note": "Medical-scheme CONTRIBUTIONS are a premium and are not booked here; this is the scheme's co-payments and the self-payment gap. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita plus medical-scheme co-payment and self-payment-gap structure", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Statistics South Africa, National Household Travel Survey 2020, tables 4.22 and 4.3, Gauteng. https://www.statssa.gov.za/publications/P0320/P03202020.pdf", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Total time travelled to place of work", "year": "2020", "note": "Mode-blind mean of 58.5 computed by weighting the published per-mode times by Gauteng's mode shares. One way, door to door, fieldwork pre-COVID. IMPORTANT: Stats SA's widely repeated headline (train 107, bus 84, taxi 63) is PUBLIC-TRANSPORT RIDERS ONLY and would badly overstate this field, since 45% of Gauteng workers travel by car and 12% walk." }, "non_cash.sick_leave": { "source": "Basic Conditions of Employment Act 75 of 1997 (BCEA), s.22 (sick leave entitlement); Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-H.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "SARS, What are the s11F annual allowable deductions https://www.sars.gov.za/faq/faq-what-are-s11f-annual-allowable-deductions/ ; https://www.sars.gov.za/latest-news/retirement-fund-contribution-deductions-section-11f2a/ ; Budget 2026 FAQ https://www.sars.gov.za/about/sars-tax-and-customs-system/budget/budget-2026-frequently-asked-questions/", "last_updated": "2026-08-11", "confidence": "medium", "note": "S11F: the lesser of the monetary cap, 27.5% of the greater of remuneration or taxable income, and taxable income before the deduction. Pension plus provident plus RA combined; excess carries forward (unmodeled). THE CAP MOVED AND THE HOUSE RULE DECIDED IT: SARS's Budget 2026 FAQ says R430,000 from 1 Mar 2026, the first change since 2016, but SARS's own still-live s11F pages say R350,000 and the increase was OMITTED from the Rates and Monetary Amounts Act gazetted 1 Apr 2026, with Treasury implementing under the ministerial-announcement provision. R350,000 is used (lower cap, higher tax). REFRESH ITEM. Employer contributions DO count against the same ceiling, but less_employer_pct is deliberately not encoded because it would require inventing an employer rate and employee 6-10% plus a typical employer 5-10% still sits under 27.5%. Does not bind at realistic South African salaries." } }, "waterloo": { "tax.filing_basis": { "source": "PwC WWTS - Canada (Individual, Tax administration): married taxpayers file separately; joint returns are not allowed.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Canada 2025-26 federal CRA brackets; Ontario provincial brackets + surtax; same engine as Toronto. CRA; Ontario Ministry of Finance; PwC Canada", "last_updated": "2026-05-31", "confidence": "high" }, "tax.social_security_rate": { "source": "CPP base + CPP2 (second ceiling) + EI employee premiums, all capped per 2025-26 CRA limits. CRA", "last_updated": "2026-05-31", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Region of Waterloo + City of Kitchener/Waterloo combined mill rate ~1.13% (high vs Toronto); owners only, renters in golden personas pay 0. Regional 2025-26 tax policy", "last_updated": "2026-05-31", "confidence": "medium" }, "housing._summary": { "source": "Numbeo / rentals.ca Kitchener-Waterloo 2026 (Uptown Waterloo / tech-corridor professional tier) rent + price; cheaper than Toronto. CAD", "last_updated": "2026-05-31", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "OHIP covers core care; private premiums modeled 0 (employer top-up for dental/vision); OOP + dental/vision from Ontario 2025-26. CAD", "last_updated": "2026-05-31", "confidence": "medium" }, "childcare._summary": { "source": "Kitchener-Waterloo licensed daycare under CWELCC ($10/day phase-in) + after-school 2025-26; subsidy reflects CWELCC fee reduction. CAD", "last_updated": "2026-05-31", "confidence": "medium" }, "food._summary": { "source": "Numbeo Kitchener-Waterloo 2026 grocery + restaurant indices (professional basket). CAD", "last_updated": "2026-05-31", "confidence": "medium" }, "transit._summary": { "source": "Grand River Transit / ION LRT monthly pass + car running costs; Numbeo 2025-26. CAD", "last_updated": "2026-05-31", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Kitchener-Waterloo leisure + domestic vacation baseline 2025-26. CAD", "last_updated": "2026-05-31", "confidence": "low" }, "travel_home._summary": { "source": "CAD->US/Europe economy round-trip 2025-26 from Toronto Pearson (1.5h drive). CAD", "last_updated": "2026-05-31", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "canada.ca Canada Child Benefit calculation sheet (July 2025-June 2026)", "last_updated": "2026-07-17", "note": "CCB is means-tested on adjusted family net income: a 2-child mixed-age family reaches $0 by ~CAD 236k AFNI. The published CAD 231k-339k households receive ~$0 (residual <$150/kid only at the very low end), and Ontario/Alberta have no high-income floor. Set to 0 (was a flat 3,000)." }, "non_cash._summary": { "source": "Ontario ESA peer (Toronto): 2-3 weeks vacation, ~9 statutory holidays, federal EI parental leave (provincial top-up varies), ESA sick days; OHIP -> universal_healthcare true. Ontario ESA; Service Canada EI", "last_updated": "2026-05-31", "confidence": "medium" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Waterloo. Its basis is converted with the rest of Canada and its rank inside Canada is preserved. Prior source: Derived, not observed: no all-levels total-comp measurement exists for Waterloo. Basis re-derived with its country anchor Toronto; Waterloo's stored ratio to Toronto is preserved. Prior source: levels.fyi Kitchener-Waterloo tech 2025-26; a dense tech corridor (Shopify/Google/OpenText + UWaterloo pipeline), pay near Toronto", "last_updated": "2026-08-11", "confidence": "low", "note": "0.62 x 0.5000 = 0.3100, R5-rounded down to 0.31. 0.5000 is the MEDIAN of (re-derived / previous) across the 5 Canada cities that DO carry an all-levels measurement (Calgary 0.72 to 0.32; Montreal 0.62 to 0.31; Ottawa 0.71 to 0.33; Toronto 0.75 to 0.4; Vancouver 0.73 to 0.49), so Waterloo moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Waterloo's rank inside Canada is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.62 was 0.8267 of Toronto's 0.75; that ratio x Toronto's re-derived 0.4 = 0.3307, R5-rounded down to 0.33. Toronto's own figure: CAD 142,630 / 1.4 = USD 101,879, / USD 250,000 = 0.4075. This is a BASIS move, not a re-measurement: Waterloo's rank inside Canada is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.62; just below Toronto on a smaller-but-strong market." }, "tax.provincial_brackets": { "confidence": "high", "source": "Ontario 2026 brackets/BPA (taxtips.ca/taxrates/on.htm); verified tax audit 2026-07-14", "last_updated": "2026-07-17" }, "tax.surtax_threshold_1": { "source": "Ontario surtax 20% of basic ON tax over 5,818 (2026); https://www.taxtips.ca/taxrates/on.htm; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "First Ontario surtax threshold (20%)." }, "tax.surtax_threshold_2": { "source": "Ontario surtax +36% of basic ON tax over 7,446 (2026); https://www.taxtips.ca/taxrates/on.htm; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "Second Ontario surtax threshold (36%)." }, "tax.health_premium": { "source": "Ontario Health Premium, stepped income-tested levy capped at 900; https://www.ontario.ca/page/health-premium; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "Gates the stepped OHP (600/750/900 plateaus) in computeCanadaTaxes." }, "tax.bpa_phaseout": { "source": "Federal BPA supplement phases 16,452 -> 14,829 over net income 181,440-258,482 (CRA 2026); https://www.canada.ca/en/revenue-agency/programs/about-canada-revenue-agency-cra/federal-government-budgets/basic-personal-amount.html; verified tax audit 2026-07-14", "last_updated": "2026-07-17", "note": "BPA modeled as a non-refundable credit at the lowest bracket rate with this federal supplement phase-out." }, "childcare.subsidy_per_kid_annual": { "source": "CWELCC (Canada-Wide Early Learning and Child Care) provincial fee schedules; double-subtraction correction 2026-08-02", "last_updated": "2026-08-02", "confidence": "medium", "note": "Set to 0 on 2026-08-02, was CAD 3,600. The old value was the CWELCC fee reduction. CWELCC is delivered as a lower posted fee at a participating centre, not as cash to the household, and childcare.preschool_monthly already carries the reduced rate, so the reduction was netted twice. The modelled CAD 550/month already sits at Ontario’s CWELCC cap. This field holds CASH THAT REACHES THE HOUSEHOLD. Where the monthly rate it nets against is already the price the household actually pays, the value is 0 and the relief is described here instead. See docs/superpowers/research/2026-08-02-childcare-double-subtraction.md. Open follow-up: the Waterloo monthly rate itself is a blend of the capped and uncapped rates and still owes its own sourcing pass." }, "non_cash.parental_leave_weeks": { "source": "Employment Insurance Act ss. 22-23: maternity 15 wk + standard parental up to 35 wk per parent (40 when shared) https://www.canada.ca/en/services/benefits/ei/ei-maternity-parental.html", "last_updated": "2026-08-06", "confidence": "high", "note": "50 = EI maternity 15 (mother) + standard parental 35, a shared pot counted once at the one-parent maximum. The 5 extra weeks that exist only when the second parent also claims are excluded as conditional on both parents' behaviour (same ruling as Portugal's sharing bonus). The extended option (up to 61 wk one parent, at 33%) was rejected: near-identical capped dollars stretched over 26 more weeks would inflate the weeks column. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. Quebec runs QPIP instead, so montreal legitimately diverges (city-varying by design in data/_national_fields.json)." }, "non_cash.parental_leave_paid_pct": { "source": "canada.ca, EI maximum insurable earnings for 2026: CAD 68,900 https://www.canada.ca/en/employment-social-development/programs/ei/ei-list/ei-employers/premium-reduction-program/2026-maximum-insurable-earnings.html", "last_updated": "2026-08-06", "confidence": "high", "note": "55% of insurable earnings for both maternity and standard parental weeks, capped at the 2026 MIE of CAD 68,900, max CAD 729/wk. Against a CAD 130,000 Toronto tech salary the cap makes the effective rate about 29%; the 55 is real only below the MIE. Headline kept per field convention; the capped-percentage shape has no honest flat-block representation, so the arithmetic lives here." }, "housing.homeowners_insurance_pct_of_value": { "source": "Provincial average premium (rates.ca Q2 2026, MyChoice city data, IBC) over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "CIHI national health expenditure household out-of-pocket + provincial drug plans: Quebec RAMQ covers a registered parent's children free, Ontario Trillium and BC Fair PharmaCare are income-deductible-based", "last_updated": "2026-08-08", "confidence": "medium", "note": "Medically necessary physician and hospital care is free everywhere, so the whole of this field is prescription drugs and allied health, which is PROVINCIAL, and the existing per-city spread is kept for that reason. Employer drug plans are a benefit and a premium, not out-of-pocket. Montreal carries a much smaller child increment than the other cities because of the RAMQ free-child rule. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "CIHI national health expenditure household out-of-pocket + provincial drug plans: Quebec RAMQ covers a registered parent's children free, Ontario Trillium and BC Fair PharmaCare are income-deductible-based", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Statistics Canada, average commute times for the 15 largest CMAs, May 2025 (The Daily, 2025-08-26, Table 4). https://www150.statcan.gc.ca/n1/daily-quotidien/250826/t004a-eng.htm", "last_updated": "2026-08-08", "confidence": "medium", "statistic": "Commuting duration, one way", "year": "2025", "note": "Kitchener-Cambridge-Waterloo CMA 24.3, the only Waterloo geography StatCan publishes. Source is high confidence; the point estimate is medium because this CMA's series is the most volatile of the 15 (22.3, 20.6, 21.6, 27.2, 24.3 across 2021-2025), consistent with a small LFS subsample. The previous 18 was wrong by 6 minutes." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "CRA, MP, DB, RRSP, DPSP, ALDA, TFSA limits and the YMPE https://www.canada.ca/en/revenue-agency/services/tax/registered-plans-administrators/pspa/mp-rrsp-dpsp-tfsa-limits-ympe.html ; https://www.canada.ca/en/revenue-agency/services/tax/individuals/topics/rrsps-related-plans/contributing-a-rrsp-prpp/contributions-affect-your-rrsp-prpp-deduction-limit.html", "last_updated": "2026-08-11", "confidence": "high", "note": "RRSP deduction limit (line 20800), 2026: 18% of prior-year earned income to an annual dollar limit. Pinned at 33,810 via the statutory identity RRSP limit year X = money-purchase limit year X-1 (CRA confirms 2025 MP = 33,810). The PENSION ADJUSTMENT is deliberately NOT modeled, so this is the no-workplace-DB case; a DB-plan member has materially less room. Carry-forward of unused room is also unmodeled. Never binds at our incomes; the 18% governs below CAD 187,833." } }, "wroclaw": { "tax.payroll_deductible": { "source": "Batch B base-deduction convention (tax audit 2026-07-14, verified): employee ZUS is deductible from the PIT base, so the brackets run on gross minus ZUS. The NFZ health contribution (9%) is NOT deductible (financed from net income since the 2022 reform). computePolandTaxes, gated by this flag (bracket path only; the ryczalt flat regime forgoes it). Source: PwC PL individual/other-taxes.", "last_updated": "2026-07-26", "confidence": "high" }, "tax.federal_brackets": { "source": "Poland 2025-26 PIT 0/12/32% (PLN 30,000 tax-free; 32% above PLN 120,000); same engine as Warsaw. Ministerstwo Finansow; PwC Poland", "last_updated": "2026-05-31", "confidence": "high" }, "tax.social_security_rate": { "source": "ZUS / getsix: Social Insurance (ZUS) contributions in Poland 2026 (employee emerytalne 9.76% + rentowe 1.50% = 11.26%, capped at the annual 30x limit). https://getsix.eu/human-resources-payroll-in-poland/social-insurance-institution-zus-contributions-in-poland-in-2026-current-bases-limits-and-contribution-amounts-for-entrepreneurs/", "last_updated": "2026-07-26", "note": "PL-1 (closeout 2026-07-26): capped ZUS stream (pension + disability), split from the former blended 13.71%." }, "tax.inbound_regime": { "source": "Ustawa o zryczaltowanym podatku dochodowym, consolidated text Dz.U. 2025 poz. 843 (https://api.sejm.gov.pl/eli/acts/DU/2025/843/text.pdf). Art. 12 ust. 1 pkt 2b sets 12% for software-related services (PKWiU ex 62.01.1, 62.01.2, ex 62.02, 62.03.1, ex 62.09.20.0, 62.02.10.0 and the 58.29 software-publishing groupings); art. 6 ust. 4 pkt 1 lit. a caps prior-year revenue at EUR 2,000,000, converted at the NBP rate of the first business day of October (table 190/A/NBP/2025, EUR 4.2586), so the 2026 limit is PLN 8,517,200. Art. 8 ust. 2 voids the regime for the whole year if the same services are supplied to a current or former employer. Ulga na powrot: art. 21 ust. 1 pkt 152 PIT, Dz.U. 2026 poz. 592.", "last_updated": "2026-07-31", "confidence": "high", "note": "2026-07-31: confidence upgraded medium -> high, and the source moved from a PwC summary to the Sejm ELI statute text. FRAMING FLAG RAISED, NOT RESOLVED HERE: the 12% ryczalt is NOT an inbound regime. It is a general self-employment option available to a Polish local on day one and unavailable to anyone taxed as an employee, and it taxes REVENUE with no cost deduction, no PLN 30,000 allowance and no joint filing. Poland has no inbound income-tax relief for a newly arrived foreign employee at all; the nearest thing is ulga na powrot (PLN 85,528/yr for 4 years, conditions in art. 21 ust. 43; no schema field for tax.ulga_na_powrot exists). Carrying it under an inbound-regime toggle therefore offers our ICP, a relocating EMPLOYEE, a discount the statute does not give them. Recommend either reclassifying the toggle as a B2B-status scenario or dropping it; not done here because it is a product decision, and the brief for this pass said 'upgrade or flag'. The regime note now says all of this in the reader's own words. The old note's claim that the model UNDERSTATES the benefit (real ryczalt gets Maly ZUS Plus and revenue-bracket NFZ) is true only for someone already self-employed; for an employee it overstates by the whole amount. 2026 changes recorded: revenue limit PLN 8,517,200 (down on a weaker EUR); NFZ base PLN 9,228.64 (M.P. 2026 poz. 117), so a 300k+ contractor pays PLN 1,495.04/mo; Maly ZUS Plus capped at 36 months in any 60 from 2026-01-01." }, "housing._summary": { "source": "Numbeo / otodom Wroclaw 2026 (Stare Miasto / Krzyki professional tier) rent + price; cheaper than Warsaw. PLN", "last_updated": "2026-05-31", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "NFZ public base + private subscription (Medicover/LuxMed) top-up for a professional family; OOP + dental from Wroclaw 2025-26. PLN", "last_updated": "2026-05-31", "confidence": "low" }, "childcare._summary": { "source": "Wroclaw public przedszkole (free core hours) + private zlobek for infants + after-school 2025-26; reflects free preschool. PLN", "last_updated": "2026-05-31", "confidence": "medium" }, "food._summary": { "source": "Numbeo Wroclaw 2026 grocery + restaurant indices (professional basket). PLN", "last_updated": "2026-05-31", "confidence": "medium" }, "transit._summary": { "source": "MPK Wroclaw monthly pass + car running costs; Numbeo 2025-26. PLN", "last_updated": "2026-05-31", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Wroclaw leisure + domestic vacation baseline 2025-26. PLN", "last_updated": "2026-05-31", "confidence": "low" }, "travel_home._summary": { "source": "PLN->US/Europe economy round-trip 2025-26 from Wroclaw/Warsaw (LOT / Wizz / Ryanair connections). PLN", "last_updated": "2026-05-31", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Poland 800+ (Rodzina 800+) universal child benefit PLN 800/mo = 9,600/yr per child, not means-tested. Peer-inherited (Poland national field). ZUS 800+", "last_updated": "2026-05-31", "confidence": "high", "note": "Polish 800+ (Rodzina 800 plus, PLN 800/mo): universal, NOT means-tested." }, "non_cash.parental_leave_weeks": { "source": "Ministerstwo Rodziny, gov.pl urlop rodzicielski: https://www.gov.pl/web/rodzina/urlop-rodzicielski ; ZUS: https://www.zus.pl/swiadczenia/zasilki/zasilek-macierzynski/wysokosc . National statute, peer-identical with Warsaw.", "last_updated": "2026-08-06", "confidence": "high", "note": "20 weeks urlop macierzynski plus 41 weeks urlop rodzicielski = 61 weeks for the household. Corrected 2026-08-01 from 20, which was the maternity block alone. See the Warsaw entry. Re-confirmed on the household-inclusive basis 2026-08-06 (the 20+41 construction already sums the per-parent 9-week reserves inside the shared 41; both-parents flag stays off per the Sweden precedent). See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "ZUS zasilek macierzynski, 70% for the urlop rodzicielski period: https://www.zus.pl/swiadczenia/zasilki/zasilek-macierzynski/wysokosc . National statute, peer-identical with Warsaw.", "last_updated": "2026-08-06", "confidence": "high", "note": "Blended 20 weeks at 100% plus 41 at 70% = 79.8%, stored as 80. See the Warsaw entry. Cap check closed 2026-08-06: zasilek macierzynski is UNCAPPED for employees (art. 36 basis is the plain 12-month average wage; the art. 46 100%-of-average-wage cap applies only after employment ends; the 250% figure is the voluntary contribution-base cap for the self-employed), so the stored blend is real for an ICP, one of the few uncapped schemes." }, "non_cash._summary": { "source": "Poland Labour Code peer (Warsaw): 20-26 vacation days, 13 public holidays, 20-week maternity at 100% then 41-week parental at 70%, 33-day employer sick pay then ZUS; NFZ -> universal_healthcare true. ILO; Kodeks pracy", "last_updated": "2026-05-31", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Wrocław: PLN 198,430, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-sat-north.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "PLN 198,430 / 3.74 = USD 53,056; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2122, R5-rounded down to 0.21. REPLACES 0.34 (2026-08-11), and the interim 0.28 set earlier the same day, which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.34 was 0.9444 of Warsaw's 0.36; that ratio x Warsaw's re-derived 0.3 = 0.2833, R5-rounded down to 0.28. Warsaw's own figure: PLN 281,628 / 3.74 = USD 75,302, / USD 250,000 = 0.3012. This is a BASIS move, not a re-measurement: Wrocław's rank inside Poland is unchanged and still rests on its prior evidence, so confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.34; set just below Warsaw, near Krakow tier." }, "tax.social_security_rate_uncapped": { "source": "ZUS / getsix: Poland 2026 employee ZUS: chorobowe (sickness) 2.45% is NOT subject to the 30x annual limit. https://getsix.eu/human-resources-payroll-in-poland/social-insurance-institution-zus-contributions-in-poland-in-2026-current-bases-limits-and-contribution-amounts-for-entrepreneurs/", "last_updated": "2026-07-26", "note": "PL-1 (closeout 2026-07-26): chorobowe 2.45%, UNCAPPED — continues on income above the 30x cap." }, "tax.social_security_cap": { "source": "ZUS: annual 30x limit (trzydziestokrotność) for pension+disability contributions, 2026 = PLN 282,600. https://getsix.eu/human-resources-payroll-in-poland/social-insurance-institution-zus-contributions-in-poland-in-2026-current-bases-limits-and-contribution-amounts-for-entrepreneurs/", "last_updated": "2026-07-26", "note": "PL-1 (closeout 2026-07-26): 2026 pension+disability cap, up from PLN 260,190 (2025)." }, "tax.filing_basis": { "source": "PwC Worldwide Tax Summaries: Poland, Individual: Tax administration (\"Married taxpayers ... may, under some conditions, choose ... jointly or separately\"); separate assessment is the default and joint (wspólne rozliczenie) is an optional election. https://taxsummaries.pwc.com/poland/individual/tax-administration", "last_updated": "2026-07-26", "note": "Closeout follow-on: separate spousal assessment is Poland’s default; joint filing is an election. Fixes the two-earner overstatement the PL-1 ZUS fix left (it had stayed joint-stacked)." }, "healthcare.statutory_employee": { "source": "NFZ / ustawa o swiadczeniach opieki zdrowotnej finansowanych ze srodkow publicznych art. 79 (9% rate) and art. 81 ust. 5 (30x ceiling does not apply to health). ZUS 30x annual limit for pension+disability 2026 = PLN 282,600.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll. Rate applied to gross rather than the statutory gross-minus-ZUS base because the field shape cannot express a net-of-contributions base; overstates by roughly 1.2pp of gross, the conservative direction, and mirrors computePolandTaxes. The 2025 bill cutting entrepreneurs health contributions was VETOED in May 2025 and is not in force." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "GUS National Health Account household out-of-pocket + gov.pl free-drug programme for under-18s, which is list- and indication-specific rather than unconditional + published Polish private clinic visit pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "NFZ care is free at the point of use, so this is prescription co-payments plus the private outpatient care a professional household actually buys. An employer medical package (Medicover, LuxMed) is a PREMIUM and is not booked here. Poland has no statutory annual ceiling. The previously unexplained Krakow discount is removed; the Warsaw tier is kept and is consistent with Warsaw private pricing running above other Polish metros. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "GUS National Health Account household out-of-pocket + gov.pl free-drug programme for under-18s, which is list- and indication-specific rather than unconditional + published Polish private clinic visit pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Kompleksowe Badania Ruchu Wroclaw 2024, Wykres 2.30. https://bip.um.wroc.pl/attachments/download/152438", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Motywacja dom-praca", "year": "2024", "note": "One way, mode-blind, work-only. 26 min for both the city and the surrounding gminas; the evening return is 29 to 30." }, "non_cash.sick_leave": { "source": "Art. 92 Kodeks pracy (Labor Code, employer-paid `wynagrodzenie chorobowe`) + Ustawa z dnia 25.06.1999 o świadczeniach pieniężnych z ubezpieczenia społecznego w razie choroby i macierzyństwa (\"ustawa zasiłkowa\" , ZUS-paid `zasiłek chorobowy`), Art. 4 (qualifying period), Art. 8 (182-day benefit-perio; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-C.md", "source_url": "https://poradnikprzedsiebiorcy.pl/-okres-wyczekiwania-na-prawo-do-zasilku-chorobowego", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Ministerstwo Rodziny, Pracy i Polityki Spolecznej, IKZE limit wplat https://www.gov.pl/web/rodzina/ikze-limit-wplat (Obwieszczenie 10 Nov 2025, M.P. 2025 poz. 1156)", "last_updated": "2026-08-11", "confidence": "high", "note": "IKZE contributions reduce dochod, the PIT base. The deduction limit equals the annual payment limit, 1.2 x the forecast average monthly wage (1.2 x 9,420 = PLN 11,304 for 2026). Employee-only. IKE is not deductible and PPK employee contributions come out of net pay with neither deduction nor credit. PLN 16,956 is the SELF-EMPLOYED limit and must not be used here. The 2025 figure was PLN 10,407.60." } }, "vilnius": { "tax.federal_brackets": { "source": "GPM 2026 three-tier 20/25/32 at 36/60 VDU (KPMG flash alert 2025-202); verified tax audit 2026-07-14", "last_updated": "2026-07-17", "confidence": "high" }, "tax.social_security_rate": { "source": "Employee Sodra VSD 12.52% capped at 60 VDU (~EUR 126,532) - modeled social_security_rate 0.1252, cap 126,532. Sodra; PwC Lithuania", "last_updated": "2026-05-31", "confidence": "high" }, "tax.health_insurance_rate": { "source": "Employee PSD (compulsory health insurance) 6.98% uncapped, withheld alongside VSD. Sodra; PwC Lithuania", "last_updated": "2026-05-31", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Lithuania VAT standard 21%. VMI; PwC Lithuania VAT", "last_updated": "2026-05-31", "confidence": "high" }, "housing._summary": { "source": "Numbeo / Aruodas Vilnius 2026 (Naujamiestis / Zverynas / Senamiestis professional tier) rent + price. EUR", "last_updated": "2026-05-31", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "PSD public base + private clinic top-up for a professional family; OOP + dental from Vilnius 2025-26. EUR", "last_updated": "2026-05-31", "confidence": "low" }, "childcare._summary": { "source": "Vilnius municipal darzelis (subsidized) + private + after-school 2025-26; no extra cash subsidy beyond the universal child money tracked under govt_benefits. EUR", "last_updated": "2026-05-31", "confidence": "medium" }, "food._summary": { "source": "Numbeo Vilnius 2026 grocery + restaurant indices (professional basket). EUR", "last_updated": "2026-05-31", "confidence": "medium" }, "transit._summary": { "source": "Vilnius public-transport monthly pass + car running costs; Numbeo 2025-26. EUR", "last_updated": "2026-05-31", "confidence": "low" }, "discretionary._summary": { "source": "Numbeo Vilnius leisure + domestic/Baltic vacation baseline 2025-26. EUR", "last_updated": "2026-05-31", "confidence": "low" }, "travel_home._summary": { "source": "EUR->US/Europe economy round-trip 2025-26 from Vilnius (airBaltic / Ryanair / Wizz connections). EUR", "last_updated": "2026-05-31", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Lithuania universal child money (vaiko pinigai) ~EUR 122/mo = ~1,500/yr per child, not means-tested (a supplement exists for low-income/large families). Sodra; modeled 1,500.", "last_updated": "2026-05-31", "confidence": "high" }, "non_cash._summary": { "source": "Lithuania Labour Code: 20 working-day annual leave, 16 public holidays, ~62 weeks paid parental leave (Sodra, ~77.58% blended headline), employer pays sick days 1-2 then Sodra from day 3; PSD public system -> universal_healthcare true. ILO; Lithuanian Labour Code; Sodra", "last_updated": "2026-05-31", "confidence": "medium" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Vilnius: EUR 64,245, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-europe.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "EUR 64,245 / 0.867 = USD 74,100; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2964, R5-rounded down to 0.29. REPLACES 0.38 (2026-05-31), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.38; set above Tallinn (Estonia) and the Polish metros on Vilnius's strong fintech market." }, "tax.filing_basis": { "source": "VMI (vmi.lt) - individual GPM; no joint filing", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "tax.social_security_cap": { "source": "Grant Thornton LT: \"Tax law changes affecting payroll as of 1 January 2026\": VSD/Sodra 60-VDU ceiling EUR 138,729 (VDU EUR 2,312.15/mo x 60). https://www.grantthornton.lt/en/insights/tax-laws-changes-affecting-payroll-as-of-January-1-2026/", "last_updated": "2026-07-26", "note": "LT-1 (closeout 2026-07-26): 2026 60-VDU VSD ceiling, up from EUR 126,532; now consistent with the 32% GPM threshold." }, "healthcare.statutory_employee": { "source": "Lietuvos Respublikos valstybinio socialinio draudimo istatymas Nr. I-1336 art. 10(9), which scopes the 60-VDU annual ceiling and carves health out of it by name ('isskyrus privalomojo sveikatos draudimo imokas'); Sodra 2026 contribution rate notice (sodra.lt, rates applying from 2026-01-01): employee PSD 6.98%, employee VSD 12.52%, 2026 VDU EUR 2,312.15 giving a 60-VDU ceiling of EUR 138,729.", "last_updated": "2026-07-31", "confidence": "medium", "note": "DISPLAY ONLY, already inside taxes.payroll. UNCAPPED is the load-bearing claim and it is the conservative reading, but it is corroborated rather than self-verified: e-seimas.lrs.lt is a JavaScript app that will not render to a fetcher and sodra.lt returns 403 site-wide, so art. 10(9) was read verbatim from a legal-text mirror and confirmed independently, with no contradicting source found. Hence medium confidence on the sourcing route, not on the fact. The supplementary-pension drift this note used to flag is CLOSED as of 2026-08-06, and it was a false alarm in both halves. Nothing in the repo ever carried 2.7% for Lithuania (repo-wide grep: no such value, no II-pillar key in data/vilnius.json, Lithuania absent from FORCED_SAVINGS), and 2.7% was in any case only the 2022 step of the 2019 reform's 1.8-to-3 ramp. The current second-pillar employee rate is 3% of gross, on top of VSD and PSD, and the payroll line here deliberately excludes it: the reform in force from 2026-01-01 ends automatic enrolment, lets savers choose or pause their rate, and opens a full exit window through 2027-12-31, so a 2026 arrival in Vilnius is opted out by default. That makes it elective retirement saving, which the calculator already represents through retirement_contrib_pct, rather than a statutory charge. Sources and the decision: docs/superpowers/research/2026-08-06-tax-value-fix-batch.md" }, "non_cash.parental_leave_weeks": { "source": "Ligos ir motinystes socialinio draudimo istatymas via Sodra: motinystes pasalpa 126 calendar days; tevystes pasalpa 30 days; vaiko prieziuros ismoka from maternity end until the child turns 18 months (kol vaikui sueis 18 menesiu), with 2 non-transferable months per parent (bent po 2 menesius, 62 kalendorines dienas) inside the span. Sodra release via LRT syndication (sodra.lt 403s). Research: docs/superpowers/research/2026-08-06-lt-cr-verification.md", "last_updated": "2026-08-06", "confidence": "high", "note": "92 wk = 126 d maternity + 30 d paternity + ~491.9 d childcare stream (birth+126 to 18 months) = 647.9 d = 92.55, stored rounded down. Was 62, resolved 2026-08-06 as a MIS-RECORD: the source phrase for the per-parent non-transferable requirement is 62 kalendorines DIENAS (days), recorded as weeks by an early researcher; no Lithuanian entitlement combination reconstructs to 62 weeks. The childcare stream is one shared pot counted once; both parents cannot use the non-transferable months simultaneously, and forfeiting them shortens the span rather than converting the rate (Sweden precedent: quotas inside the pot, no both_parents flag). 24-month option (45% to the 1st birthday, then 30%) not stored; the 18-month option is the modeled default and the conservative election. CONFIRMED against the household-inclusive basis by the batch-5 audit 2026-08-07 (tenure gates: maternity 12-of-24, paternity 6-of-24; the employed mover passes all but the non-EU conceive-immediately maternity edge, see the capped note). Research: docs/superpowers/research/2026-08-07-pl-b5-baltics-balkans.md." }, "non_cash.parental_leave_paid_pct": { "source": "Sodra: motinystes/tevystes pasalpos and the non-transferable childcare months at 77.58%-78% of kompensuojamasis uzdarbis; transferable childcare remainder at 60% (18-month option)", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 92 weeks into non_cash.parental_leave_capped (two rate tiers flattened), so no percentage-paid weeks remain on this field. Was 77.58, which applied the high tier to a mis-recorded week count." }, "non_cash.parental_leave_capped": { "source": "Sodra Q3-2026 benefit limits via the LRT republication of the Sodra release (fetched 2026-08-07): rights arising Q3-2026 use VDU 2,542.80 -> 2x VDU = EUR 5,085.60/month; published maxima are exact multiples (paternity/maternity 3,945.41 = 77.58%, NT childcare months 3,966.77 = 78%, 18-month track 3,051.36 = 60%). Research: docs/superpowers/research/2026-08-07-pl-b5-baltics-balkans.md, 2026-08-06-lt-cr-verification.md, 2026-08-06-parental-caps-europe.md (Lithuania)", "last_updated": "2026-08-07", "confidence": "high", "note": "Two statutory slices flattened LOSSLESSLY into one block because both share the same 2x-VDU base: 40 wk at 77.58% (maternity + paternity + the 2+2 non-transferable months, whose literal rate 78% is folded DOWN to 77.58, conservative) and 52 wk at 60%. Duration-weighted pct (40 x 77.58 + 52 x 60) / 92 = 67.64; blended payout cap 67.6435% x 5,085.60 = EUR 3,440.08/month, elbow-exact at the shared 2x-VDU base, so the flatten is exact at every salary. Batch 5 refreshed the cap from the Q1-2026 snapshot (3,268 on base 4,831.40) to the in-force Q3-2026 float per the mid-year-step convention (batch-3 CO/OR precedent: store the figure in force at the research date). The base FLOATS QUARTERLY with VDU (~8.5% y/y growth), refresh row in docs/data-refresh-calendar.md. Rate structure now confirmed from two Sodra releases (Q1 + Q3 2026), confidence raised to high. Mover caveat: a NON-EU conceive-immediately mover narrowly misses the 12-of-24 maternity gate (the right arises ~70 days pre-birth); EU movers aggregate insurance periods under Reg 883/2004; caveat only, no shape change." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Lithuanian medicines co-payment basket EUR 70.02 per person per year (2026) + the near-total child exemption: free public primary, specialist, inpatient and rehabilitation care plus 100%-of-base-price drug reimbursement + published private visit pricing", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 nudged up modestly from 400 to 430, built from adult-only medicines out-of-pocket of ~EUR 288 demographic-corrected plus an allowance for the well-documented private-pay specialist habit. Lithuania's child exemption is markedly more complete than Estonia's reduced-fee approach, so the child increment is set lower here than there. The only ceiling-like mechanism is the EUR 70.02 medicines basket. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "Lithuanian medicines co-payment basket EUR 70.02 per person per year (2026) + the near-total child exemption: free public primary, specialist, inpatient and rehabilitation care plus 100%-of-base-price drug reimbursement + published private visit pricing", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "UNSOURCED (2026-08-08 commute-unit sweep): no acceptable primary series located for this city. This entry deliberately claims no provenance; see docs/superpowers/research/2026-08-08-commute-unit-standardization.md and its regional shard files for what was searched.", "last_updated": "2026-08-08", "confidence": "low", "note": "UNSOURCED, value retained. No published Lithuanian mean commute duration was found: checked osp.stat.gov.lt, the Labour Force Survey, data.gov.lt, the Vilnius SUMP and the JUDU survey. Tallinn's figure was deliberately NOT borrowed." }, "non_cash.sick_leave": { "source": "Darbo kodeksas (Labour Code, employer-paid first 2 days) + Ligos ir motinystės socialinio draudimo įstatymas (Law on Sickness and Maternity Social Insurance, Sodra-paid ligos pašalpa from day 3).; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-C.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "VMI publication KD-0001815 https://www.vmi.lt/evmi/documents/20142/391095/KD-0001815+pensij%C5%B3+kaupimo+%C4%AFmokos.pdf ; https://investuotojui.lt/pensija/gpm-lengvata/ (GPMI art. 21; infolex 403s)", "last_updated": "2026-08-11", "confidence": "high", "note": "Closed to a new saver. GPMI art. 21 allows III pillar contributions to be deducted from annual income, bounded by EUR 1,500/yr (combined with life insurance premiums and additional II pillar contributions above 3% of income) and by 25% of annual taxable income. BUT from tax period 2025 the relief applies only to pension accumulation contracts concluded on or before 2024-12-31, and only for contributions paid up to 2034-12-31. Our ICP is someone moving to Vilnius now, who cannot hold a pre-2025 contract. A first research pass mis-summarised the VMI PDF as EUR 2,000 / 15%; that is wrong and was corrected to 1,500 / 25% before this value was set. The legacy-saver alternative would be { max: 1500, pct_of_earnings: 0.25 }." } }, "belgrade": { "tax.federal_brackets": { "source": "Serbia salary tax: flat 10% on gross net of the monthly non-taxable amount (RSD 34,221/mo = 410,652/yr, law amendment effective 2026-01-01; was 28,423/mo = 341,076/yr in 2025), modeled standard_deduction (married = 2x single, doubled-edge convention). Poreska uprava; PwC Serbia; https://unija.com/sr/izmene-zakona-o-porezu-na-dohodak-gradjana-2026/", "last_updated": "2026-08-07", "confidence": "high" }, "tax.social_security_rate": { "source": "Employee social contributions 19.9% (PIO pension 14% + health 5.15% + unemployment 0.75%) capped at the najviša osnovica doprinosa: 2026 monthly cap RSD 732,820 (5 x avg monthly salary 146,564, Oct 2024-Sep 2025), annual cap RSD 8,793,840 (= stored social_security_cap), Official Gazette RS 112/2025, applies from 2026-01-01 (prior label said 'for 2025' - that vintage was wrong, the figure is the 2026 cap). CROSO; PwC Serbia; https://www.prlegal.rs/new-contribution-base-amounts-for-2026/", "last_updated": "2026-08-07", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Serbia VAT standard 20%. Poreska uprava; PwC Serbia VAT", "last_updated": "2026-05-31", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Belgrade property tax on owners up to ~0.4% of assessed value (renters in golden personas pay 0). Belgrade city tax administration", "last_updated": "2026-05-31", "confidence": "low" }, "housing._summary": { "source": "Numbeo / 4zida Belgrade 2026 (Vracar / Stari Grad / Novi Beograd professional tier) rent + price. RSD", "last_updated": "2026-05-31", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "RFZO public base + private clinic top-up for a professional family; premiums by family size + OOP from Belgrade 2025-26. RSD", "last_updated": "2026-05-31", "confidence": "low" }, "childcare._summary": { "source": "Belgrade public vrtic (subsidized) + private + after-school 2025-26 (professional tier); no extra cash subsidy reaches an ICP household -> subsidy 0. RSD", "last_updated": "2026-05-31", "confidence": "low" }, "food._summary": { "source": "Numbeo Belgrade 2026 grocery + restaurant indices (professional basket). RSD", "last_updated": "2026-05-31", "confidence": "medium" }, "transit._summary": { "source": "Belgrade GSP public transit FREE for all riders since 1 Jan 2025 (monthly_pass 0) + car running costs. City of Belgrade; Numbeo 2025-26", "last_updated": "2026-05-31", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo Belgrade leisure + domestic vacation baseline 2025-26. RSD", "last_updated": "2026-05-31", "confidence": "low" }, "travel_home._summary": { "source": "RSD->US/Europe economy round-trip 2025-26 (Air Serbia / Turkish / Wizz connections). RSD", "last_updated": "2026-05-31", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Serbia child allowance (decije dodatak) is means-tested -> 0 for an ICP household; the one-off birth grant is not a recurring per-kid allowance. Ministry of Family Welfare", "last_updated": "2026-05-31", "confidence": "medium" }, "non_cash._summary": { "source": "Serbia Labour Law: 20 working-day minimum annual leave, ~11 public holidays, generous paid maternity/parental leave (~52 weeks, 100% headline within ceilings), employer sick pay first 30 days then RFZO; RFZO public system -> universal_healthcare true. ILO; Serbian Labour Law", "last_updated": "2026-05-31", "confidence": "medium" }, "compensation_index": { "source": "Derived, not observed, and not measured at country level either: no usable all-levels figure exists anywhere in Serbia. Prior source: levels.fyi / HelloWorld.rs Belgrade tech 2025-26; a large outsourcing + product hub (Microsoft Development Center, NCR, Nordeus), the Western-Balkans tech leader", "last_updated": "2026-08-11", "confidence": "low", "note": "0.32 x 0.6000 = 0.1920, R5-rounded down to 0.19. 0.6000 is the median of (re-derived / previous) across the 147 cities worldwide that DO carry an all-levels measurement, i.e. the measured size of the move from senior-over-196k-to-220k onto all-levels-over-250k. THIS IS THE WEAKEST CLASS OF VALUE IN THE COLUMN: it asserts only that Belgrade now sits on the same basis as everything else, and says nothing about whether its magnitude was ever right. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.32; set above the Caucasus metros, below EU members Vilnius (0.38)/Polish metros." }, "tax.filing_basis": { "source": "Poreska uprava - individual taxation; no joint returns", "last_updated": "2026-07-17", "note": "Activates the per-filer split in compute() for two-earner households; verified tax audit 2026-07-14." }, "tax.annual_avg_salary": { "source": "KPMG RS 2026 annual personal income tax alert + PwC Serbia (2025 income filed 2026): prosečna godišnja zarada RSD 1,813,032; 3x = non-taxable threshold 5,439,096; 6x = 10%/15% split 10,878,192; personal deduction taxpayer 40% + dependents 15% each capped at 50% of base; under-40 extra 3x relief. https://kpmg.com/rs/en/insights/tax-alerts/2026/02/annual-personal-income-tax-for-income-generated-in-2025.html . Re-verified 2026-08-07 against the RZS statistical-office publication of 2026-02-25 (prosečna godišnja zarada za 2025 = 1,813,032; https://www.paragraf.rs/dnevne-vesti/020326/020326-vest7.html) - value confirmed exact.", "source_name": "KPMG RS / PwC 2026; RZS 2026-02-25", "last_updated": "2026-08-07", "confidence": "high", "note": "Drives the godisnji porez na dohodak gradana (annual supplementary tax) as a gated post-pass in compute() (correct-first #5). AGE-STRADDLE LIMITATION: under-40 relief keys off last-day-of-year age, which age_bracket cannot express exactly; 18-24/25-34 get the relief, everything else INCLUDING the straddling 35-44 defaults to the conservative 40-plus (no-relief) treatment. Modeled single-household (net = household gross - household tax). Belgrade is calculator-only. Re-verify AVG each year." }, "non_cash.parental_leave_weeks": { "source": "Zakon o radu cl. 94 (porodiljsko odsustvo starts 28-45 days pre-birth; porodiljsko + odsustvo radi nege deteta run 365 days from the START of porodiljsko for the 1st/2nd child) and cl. 94(6) (nega-deteta transferable to the father, one pot); Zakon o radu cl. 77 (5 paid days: a combined annual allowance across life events). Verified 2026-08-07 via paragraf.rs", "last_updated": "2026-08-07", "confidence": "high", "note": "52 wk CONFIRMED on the household-inclusive basis by the batch-5 audit (2026-08-07): the 365-day clock runs from the START of porodiljsko, so the 28-45 pre-birth days are INSIDE it, not additive; the nega-deteta portion is one transferable pot, never both parents at once -> counted once. The cl. 77 '5 paid days' is a combined annual allowance for weddings/births/etc., not a dedicated paternity entitlement -> EXCLUDED (conservative ruling recorded). Mover note: the benefit base averages the last 18 months of Serbian earnings, but the /18 dilution is CAP-MASKED at ICP salaries (12/18 of the $150k-basis salary still exceeds the 5x-average cap). Research: docs/superpowers/research/2026-08-07-pl-b5-baltics-balkans.md." }, "non_cash.parental_leave_paid_pct": { "source": "Zakon o finansijskoj podrsci porodici sa decom (naknada zarade at 100% of the wage base for porodiljsko odsustvo + odsustvo radi nege deteta, 365 days) via paragraf.rs", "last_updated": "2026-08-06", "confidence": "high", "note": "We store 0: the 2026-08-06 caps rebuild moved all 52 weeks into non_cash.parental_leave_capped (100% of a base capped at 5x the national average salary), so no percentage-paid weeks remain on this field. Was 100 with the ceiling unrecorded." }, "non_cash.parental_leave_capped": { "source": "Zakon o finansijskoj podrsci porodici sa decom, art. 13 (base cannot exceed five average monthly salaries for rights acquired from 1 January 2022) and art. 20 (same ceiling across the whole 365-day span), statute text via paragraf.rs ; Republic Statistical Office of Serbia, Average salaries and wages per employee, June 2026 (25 Aug 2026) https://publikacije.stat.gov.rs/G2026/HtmlE/G20261228.html: average gross salary RSD 166,123, so cap = 5 x 166,123 = RSD 830,615/month", "last_updated": "2026-09-01", "confidence": "medium-high", "note": "SNAPSHOT June-2026 (latest RZS publication as of 2026-09-01): the cap FLOATS MONTHLY at 5x the latest published RZS average gross salary (~11% y/y wage growth means later months run higher); refresh row in docs/data-refresh-calendar.md. The 2022 reform RAISED the cap from 3x to 5x, refuting the \"uncapped after 2022\" hypothesis. Base and payout coincide because the rate is 100. The base is GROSS and naknada zarade takes standard payroll treatment, same convention as the salary it replaces. Research: docs/superpowers/research/2026-08-06-parental-caps-europe.md (Serbia), docs/superpowers/research/2026-08-06-parental-caps-europe-batch-E.md, and docs/superpowers/research/2026-09-01-statutory-refresh-pass.md." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Row 1 held at its prior value: the 2026-08-08 sweep found no evidence that it was clearly wrong, and no red flag (no statutory maximum used as an average, no premium booked as out-of-pocket, no order-of-magnitude currency error). Only the household curve was added, on the conservative no-child-discount shape unless a government child exemption was directly sourced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita (2023) cross-checked against the OWID PPP series and converted at build/fx-rates.json, with the household curve derived from the country's child cost-sharing rule", "last_updated": "2026-08-08", "confidence": "low", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "Belgrade Transport Model Final Report, December 2015, Table 2-24. https://www.bgsaobracaj.rs/dokumenti/66_%D0%90%D0%B6%D1%83%D1%80%D0%B8%D1%80%D0%B0%D1%9A%D0%B5%D0%A2%D1%80%D0%B0%D0%BD%D1%81%D0%BF%D0%BE%D1%80%D1%82%D0%BD%D0%BE%D0%B3%D0%BC%D0%BE%D0%B4%D0%B5%D0%BB%D0%B0%D0%91%D0%B5%D0%BE%D0%B3%D1%80%D0%B0%D0%B4%D0%B0.pdf", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Trajanje kretanja po svrhama, odlazak na posao (trip duration by purpose, going to work)", "year": "2015", "note": "Computed mean 34.4 over N=594,537 trips. One way: the return journey is a separate column, so no halving applies. RZS collects no commute time." }, "non_cash.sick_leave": { "source": "Zakon o radu (Labour Law) Art. 115 (employer-paid naknada zarade, first 30 days) + Zakon o zdravstvenom osiguranju (Health Insurance Law) Art. 50 (qualifying insurance record) and Art. 75 (RFZO-paid phase, episode-continuity and medical-commission referral triggers).; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-C.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Dunav Penzije (non-taxable amount raised to RSD 8,677 from 1 Feb 2026) https://dunavpenzije.com/en/income-tax-exempt-amount-of-pension-contribution-increased-to-rsd-8677/ ; NBS https://www.nbs.rs/sr/scripts/showcontent/index.html?id=6265&konverzija=no", "last_updated": "2026-08-11", "confidence": "medium", "note": "Voluntary pension fund contributions are exempt from salary tax AND from mandatory social contributions up to a non-taxable MONTHLY amount, applying to employer contributions and, in the same amount, to employee contributions withheld by administrative ban. DENOMINATION TRAP HANDLED: the source is monthly, RSD 8,677 x 12 = RSD 104,124/yr. Indexed each 1 February. UNRESOLVED, flagged not fixed: whether the exemption should also reduce the base of Serbia's annual surtax (computeSerbiaAnnualTax, keyed to tax.annual_avg_salary) was not settled; today it does not." }, "healthcare.premium_monthly_by_family_size": { "source": "Serbian statutory health insurance (doprinos za zdravstveno osiguranje) is an employee payroll contribution of 5.15% of gross, inside the 19.90% total employee rate (pension 14.00 + health 5.15 + unemployment 0.75), base floor RSD 51,297/mo and ceiling RSD 732,820/mo. Those are the same rates and the same ceiling this file already carries in tax.social_security_rate 0.199 and tax.social_security_cap 8,793,840 (732,820 x 12).", "last_updated": "2026-09-01", "confidence": "high", "note": "ZEROED 2026-09-01 from 4500/8000/10500/13000/15500 RSD per month. The old values DOUBLE-COUNTED statutory health: it is already withheld inside tax.social_security_rate, and it was charged again here as a standalone premium. Diagnosed from a user flag (flag:01786885294966-8ca735b6) whose reporter put Serbian public cover at about EUR 40 per adult per month, which is what our RSD 4,500 row was, so the field was pricing the public contribution rather than anything on top of it. Basis now matches every peer city whose statutory health sits in payroll (berlin, barcelona, madrid, vienna, warsaw, prague, budapest all carry 0), and follows the payroll-contribution versus optional-private distinction adjudicated for Berlin and Barcelona in the 2026-07-31 wave. Optional private clinic cover, which the same reporter put at EUR 80-90 per adult per month, is NOT modelled for any city and is not modelled here." }, "healthcare.dental_vision_monthly": { "source": "Bottom-up from published Belgrade private-practice price lists: dental examination from RSD 2,000, tartar removal with polish from RSD 4,500, one-surface composite filling RSD 3,500 (Panakeia, Banovo Brdo, https://www.panakeia.rs/sr/page/cene/). A generous routine year for one adult (two cleanings, one checkup, one filling, plus an eye exam and amortised glasses) reconstructs to about RSD 20,500, so RSD 2,000/mo is set at the top of that range.", "last_updated": "2026-09-04", "confidence": "medium", "note": "CORRECTED 2026-09-04 from RSD 7,000/mo, which was unsourced (Belgrade is not among the 12 cities carrying a per-city citation in shared_fields) and was the HIGHEST value in the CEE peer set at USD 69/adult/mo, above Warsaw and 2.1x Prague, in a country with a lower price level than either. Suspected origin is a full private-health premium package that includes dental (quoted at RSD 14,000-30,000/mo per person) read into a routine-spend field; recorded as a hypothesis, not proven provenance. New value places Belgrade at USD 19.75/adult/mo, above Sofia and Bucharest and below Prague, Budapest and Warsaw. Set at the TOP of the defensible range per adjudication rule 5, because lowering a cost flatters the country. NOT zeroed: unlike the premium field zeroed 2026-09-01, routine dental and vision sit outside the 19.90% statutory payroll line and remain a real out-of-pocket cost." } }, "curitiba": { "tax.filing_basis": { "source": "PwC WWTS - Brazil (Tax administration): spouses may elect to file jointly or separately; individual is the two-earner default.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Brazil 2026 IRPF: peer-identical with São Paulo/Rio/Belo Horizonte (national; now peer-consistency-enforced)", "last_updated": "2026-07-19", "confidence": "high" }, "tax.sales_tax_rate": { "source": "ICMS/ISS composite modeling convention 0.25 (inherited, uniform across Brazil cities)", "last_updated": "2026-07-19", "confidence": "high" }, "healthcare._summary": { "source": "SUS public + private top-up plan (national baseline, inherited from SP/Rio/BH)", "last_updated": "2026-07-19", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Bolsa Família means-tested, 0 for tech income (national, inherited)", "last_updated": "2026-07-19", "confidence": "high" }, "non_cash._summary": { "source": "Brazilian labour law CLT (national, inherited); 30 calendar days vacation = 22 working days; 120-day maternity (17 wks)", "last_updated": "2026-07-19", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "IPTU Curitiba ~0.6% effective residential (Prefeitura de Curitiba, LC municipal tabela 2025-26)", "last_updated": "2026-07-19", "confidence": "medium" }, "housing._summary": { "source": "FipeZap Curitiba 2025-26 (~R$11.6k/m2 sale) + Secovi-PR/Inpespar rental survey + Numbeo Curitiba; Batel/Água Verde tech-professional band", "last_updated": "2026-07-19", "confidence": "high" }, "housing.rent_percentiles": { "source": "Zap Imoveis glue-API listing-count CDF, 2026-08-18 (apartments for rent; city+state combo silently returns ZERO for Curitiba - resolved via addressLocationId=BR>Parana>NULL>Curitiba, cross-confirmed by addressCity alone at 952 both; bare aluguel semantics proven on Rio this run; asking BRL/month): 2 quartos P25/med/P75 = 2,175/2,773/3,735 (n=952) -> 0.784/1.347, upper edge stable at ~1.34-1.35 across interpolation choices; 1 quarto 0.819/1.185 (n=1,544, ROUGHLY HOLDS - almost exactly the default).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24. SPLIT verdict adjudicated to a fix on the guadalajara/kuala-lumpur precedent: the primary 2BR upper edge is out beyond tolerance (+0.107, i.e. ~0.05 past the 1.30 ceiling, STABLE across interpolation choices) in the flattering direction, the lower edge is inside, and the 1Q slice holds; the breach is robust in direction (informal-tail truncation only widens; duplication could explain part of the top and is recorded). Applied 0.78/1.35 (nearest). Curitiba is the closest of the six Brazilian cities to the default - the mildest fix of the Brazil block. 2 quartos anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-brazil.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (C). Level observations logged out of scope." }, "childcare._summary": { "source": "Private creche/escola infantil Curitiba market surveys 2025-26; no cash subsidy at tech income", "last_updated": "2026-07-19", "confidence": "medium" }, "food._summary": { "source": "Numbeo Curitiba 2026, cross-checked vs BH/Rio peers", "last_updated": "2026-07-19", "confidence": "medium" }, "transit._summary": { "source": "URBS/RIT single fare R$6.00 (2026) x 60 trips/mo peer convention", "last_updated": "2026-07-19", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Curitiba leisure prices, peer-aligned", "last_updated": "2026-07-19", "confidence": "medium" }, "travel_home._summary": { "source": "CWB->US economy round trip via GRU hub, 2025-26 fare scans", "last_updated": "2026-07-19", "confidence": "medium" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Curitiba. Its basis is converted with the rest of Brazil and its rank inside Brazil is preserved. Prior source: Glassdoor/Levels.fyi Curitiba senior SWE 2025-26 (EBANX/Olist/MadeiraMadeira/Contabilizei cluster)", "last_updated": "2026-08-11", "confidence": "low", "note": "0.28 x 0.4688 = 0.1313, R5-rounded down to 0.13. 0.4688 is the MEDIAN of (re-derived / previous) across the 1 Brazil city that DO carry an all-levels measurement (São Paulo 0.32 to 0.15), so Curitiba moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Curitiba's rank inside Brazil is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.28 — at BH level, below SP 0.32/Rio 0.30." }, "non_cash.parental_leave_capped": { "source": "Lei 8.213/91 art. 72 (salario-maternidade = remuneracao integral, paid by the empresa and compensated against INSS contributions, art. 72 par. 1); CF art. 248 (benefits outside the RGPS teto observe the art. 37 XI ceiling, the STF-minister subsidio); STF ADI 1946 (salario-maternidade exempt from the RGPS teto); subsidio = R$ 46,366.19/month from 1 Feb 2025, Lei 14.520/2023 https://www.planalto.gov.br/ccivil_03/_ato2023-2026/2023/lei/l14520.htm (final step; the Dec-2025 judiciary raise covers servidores, not ministers)", "last_updated": "2026-08-17", "confidence": "high", "note": "The 17.143 maternity weeks pay 100% of remuneration up to the constitutional teto of R$ 46,366.19/month (R$ 556,394.28/yr); the 0.857-wk paternity remainder stays on parental_leave_paid_pct 100 (ADCT art. 10 par. 1, employer-paid at full salary, no ceiling). The ceiling that binds a CLT empregada is CF art. 248's STF-minister subsidio, NOT the RGPS teto (R$ 8,475.55 in 2026), per ADI 1946; the RGPS teto does bind the other claimant categories our ICP is not in. Refutation branch checked and rejected: no statute or binding precedent obliges the employer to top up above the teto, and modelling the cap is the conservative call. At the R$ 782,000/yr base scenario the maternity slice pays 71.2% and the blended 18-wk rate is 72.6%; below R$ 556k/yr it stays an honest 100%. UNIT: the subsidio is a MONTHLY figure, revised only by lei (last step 1 Feb 2025). Research: docs/superpowers/research/2026-08-17-parental-leave-caps.md." }, "non_cash.parental_leave_weeks": { "source": "CF art. 7 XVIII (120 dias); CLT art. 392; Lei 8.213/91 arts. 71-73 (salario-maternidade = remuneracao integral); ADCT art. 10 par. 1 (5 days paternity, in force through 2026); Lei 15.371 phase-in: Senado https://www12.senado.leg.br/noticias/materias/2026/04/01/sancionado-o-aumento-gradual-da-licenca-paternidade", "last_updated": "2026-08-06", "confidence": "high", "note": "18 wk = 120 days maternity at 100% (17.143 wk; capped at the constitutional STF-minister teto, BRL 46,366.19/month, modelled in non_cash.parental_leave_capped as of 2026-08-17; the 2026-08-06 judgment that the teto is irrelevant to any ICP was corrected, it binds at the USD-anchored scenarios this site prices, e.g. the R$ 782,000/yr base scenario) + 5 days paternity at full salary, employer-paid (0.71 wk); 17.86 rounds to 18 (the Germany 66.87-to-67 convention). Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. WATCH ITEM: Lei 15.371 (sanctioned 2026-03-31) phases paternity to 10 days from 2027-01-01 (18.6 wk), 15 from 2028 (19.3), 20 from 2029 (20), with an INSS-reimbursed salario-paternidade at full remuneration; re-touch each January. Excluded: Programa Empresa Cidada 180/20-day extensions (employer opt-in via tax incentive) and collective-agreement top-ups." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Brazilian private consultation and co-payment pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "CHECKED for the premium-as-out-of-pocket error and CLEAR: SUS is free at the point of use and these values are verified far below any real plano de saude premium, so they are co-pays and cash consultations rather than the plan itself. Two city tiers preserved. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Brazilian private consultation and co-payment pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "IBGE, Censo Demografico 2022, SIDRA table 10331, Regiao Metropolitana de Curitiba (29 municipios). https://sidra.ibge.gov.br/tabela/10331", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Tempo habitual de deslocamento de casa para o trabalho", "year": "2022", "note": "One way by construction; band-midpoint mean. Previous value was inflated by 21 minutes." }, "non_cash.sick_leave": { "source": "Lei 8.213/1991 (Plano de Benefícios da Previdência Social) , Art. 60 §3º (employer; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": "https://www.gov.br/inss/pt-br/direitos-e-deveres/inscricao-e-contribuicao/tabela-de-contribuicao-mensal", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Receita Federal, PGBL e VGBL https://www.gov.br/receitafederal/pt-br/acesso-a-informacao/perguntas-frequentes/imposto-de-renda/dirpf/declaracao/pgvl-vgbl (Lei 9.532/97 art. 11)", "last_updated": "2026-08-11", "confidence": "high", "note": "PGBL contributions are deductible up to 12% of rendimentos tributaveis, i.e. 12% of gross for an employee. No monetary ceiling, so no max key. VGBL is NOT deductible. The deduction requires the COMPLETE declaration and this value assumes it, which is consistent with the branch: Sao Paulo carries no standard_deduction, so the engine already models a complete/itemised basis, and at BRL 300k the 12% (R$36,000) beats the simplificado cap (R$17,640) so choosing complete is genuinely rational. Lei 15.270/2025, in force 1 Jan 2026, leaves art. 11 untouched. Never binds at 6% or 10%: pure correctness insurance." } }, "porto-alegre": { "tax.filing_basis": { "source": "PwC WWTS - Brazil (Tax administration): spouses may elect to file jointly or separately; individual is the two-earner default.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Brazil 2026 IRPF: peer-identical with São Paulo/Rio/Belo Horizonte (national; now peer-consistency-enforced)", "last_updated": "2026-07-19", "confidence": "high" }, "tax.sales_tax_rate": { "source": "ICMS/ISS composite modeling convention 0.25 (inherited, uniform across Brazil cities)", "last_updated": "2026-07-19", "confidence": "high" }, "healthcare._summary": { "source": "SUS public + private top-up plan (national baseline, inherited from SP/Rio/BH)", "last_updated": "2026-07-19", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Bolsa Família means-tested, 0 for tech income (national, inherited)", "last_updated": "2026-07-19", "confidence": "high" }, "non_cash._summary": { "source": "Brazilian labour law CLT (national, inherited); 30 calendar days vacation = 22 working days; 120-day maternity (17 wks)", "last_updated": "2026-07-19", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "IPTU Porto Alegre ~0.5% effective residential post-LC 963/2022 reform (Prefeitura de Porto Alegre)", "last_updated": "2026-07-19", "confidence": "medium" }, "housing._summary": { "source": "FipeZap Porto Alegre 2025-26 (~R$7.5k/m2 sale) + Secovi-RS + Numbeo; Moinhos de Vento/Petrópolis band. Post-May-2024-flood: rents elevated, sale prices soft (flood-risk discount)", "last_updated": "2026-07-19", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Zap Imoveis glue-API listing-count CDF, 2026-08-18 (apartments for rent, addressCity Porto Alegre / addressState Rio Grande do Sul - works directly here, cross-confirmed vs locationId 4,181 vs 4,183; bare aluguel semantics proven on Rio this run; asking BRL/month): 2 quartos P25/med/P75 = 1,934/2,875/3,969 (n=4,181) -> 0.673/1.381; 1 quarto 0.646/1.343 (n=5,150, studios fold in; 1Q pool larger than 2Q here).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW on both edges at both slices (2Q -0.147/+0.141; 1Q -0.174/+0.103) - milder than Rio/Brasilia, clearly wider than the default. Applied 0.67/1.38 (nearest). Single portal; informal-tail truncation - conservative. 2 quartos anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-brazil.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (C). NOTE stored 2BR median 3,900 vs measured pool median ~2,875 (1BR nearly exact at 2,600 vs ~2,559) - a LEVEL observation, flagged, out of scope here." }, "childcare._summary": { "source": "Private creche/escola infantil POA market surveys 2025-26; no cash subsidy at tech income", "last_updated": "2026-07-19", "confidence": "medium" }, "food._summary": { "source": "Numbeo Porto Alegre 2026, cross-checked vs BH/Rio peers", "last_updated": "2026-07-19", "confidence": "medium" }, "transit._summary": { "source": "EPTC/Carris single fare R$5.30 (2026) x 60 trips/mo peer convention", "last_updated": "2026-07-19", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Porto Alegre leisure prices, peer-aligned", "last_updated": "2026-07-19", "confidence": "medium" }, "travel_home._summary": { "source": "POA->US economy round trip via GRU hub, 2025-26 fare scans", "last_updated": "2026-07-19", "confidence": "medium" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Porto Alegre. Its basis is converted with the rest of Brazil and its rank inside Brazil is preserved. Prior source: Glassdoor/Levels.fyi POA senior SWE 2025-26 (Tecnopuc, SAP Labs São Leopoldo, Dell, CWI)", "last_updated": "2026-08-11", "confidence": "low", "note": "0.28 x 0.4688 = 0.1313, R5-rounded down to 0.13. 0.4688 is the MEDIAN of (re-derived / previous) across the 1 Brazil city that DO carry an all-levels measurement (São Paulo 0.32 to 0.15), so Porto Alegre moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Porto Alegre's rank inside Brazil is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.28 — at BH level." }, "healthcare.premium_monthly_by_family_size": { "source": "ANS coletivo empresarial coverage data (derivation + caveats in docs/superpowers/research/2026-08-05-brazil-healthcare.md)", "last_updated": "2026-08-05", "confidence": "high", "note": "Re-based 2026-08-05: BRL 800 was the retail individual price; ~70% of ICP-class professionals get the employer channel, which is not household-zero (avg ~60% employer subsidy, employee-paid dependants, copays). Halved single tier with family scaling; OOP raised 3,000 to 5,000 to carry the copay reality. Closes flag:01785000328601-1787fcc3 + flag:01785698167718-26da2666." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Brazilian private consultation and co-payment pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.parental_leave_capped": { "source": "Lei 8.213/91 art. 72 (salario-maternidade = remuneracao integral, paid by the empresa and compensated against INSS contributions, art. 72 par. 1); CF art. 248 (benefits outside the RGPS teto observe the art. 37 XI ceiling, the STF-minister subsidio); STF ADI 1946 (salario-maternidade exempt from the RGPS teto); subsidio = R$ 46,366.19/month from 1 Feb 2025, Lei 14.520/2023 https://www.planalto.gov.br/ccivil_03/_ato2023-2026/2023/lei/l14520.htm (final step; the Dec-2025 judiciary raise covers servidores, not ministers)", "last_updated": "2026-08-17", "confidence": "high", "note": "The 17.143 maternity weeks pay 100% of remuneration up to the constitutional teto of R$ 46,366.19/month (R$ 556,394.28/yr); the 0.857-wk paternity remainder stays on parental_leave_paid_pct 100 (ADCT art. 10 par. 1, employer-paid at full salary, no ceiling). The ceiling that binds a CLT empregada is CF art. 248's STF-minister subsidio, NOT the RGPS teto (R$ 8,475.55 in 2026), per ADI 1946; the RGPS teto does bind the other claimant categories our ICP is not in. Refutation branch checked and rejected: no statute or binding precedent obliges the employer to top up above the teto, and modelling the cap is the conservative call. At the R$ 782,000/yr base scenario the maternity slice pays 71.2% and the blended 18-wk rate is 72.6%; below R$ 556k/yr it stays an honest 100%. UNIT: the subsidio is a MONTHLY figure, revised only by lei (last step 1 Feb 2025). Research: docs/superpowers/research/2026-08-17-parental-leave-caps.md." }, "non_cash.parental_leave_weeks": { "source": "CF art. 7 XVIII (120 dias); CLT art. 392; Lei 8.213/91 arts. 71-73 (salario-maternidade = remuneracao integral); ADCT art. 10 par. 1 (5 days paternity, in force through 2026); Lei 15.371 phase-in: Senado https://www12.senado.leg.br/noticias/materias/2026/04/01/sancionado-o-aumento-gradual-da-licenca-paternidade", "last_updated": "2026-08-06", "confidence": "high", "note": "18 wk = 120 days maternity at 100% (17.143 wk; capped at the constitutional STF-minister teto, BRL 46,366.19/month, modelled in non_cash.parental_leave_capped as of 2026-08-17; the 2026-08-06 judgment that the teto is irrelevant to any ICP was corrected, it binds at the USD-anchored scenarios this site prices, e.g. the R$ 782,000/yr base scenario) + 5 days paternity at full salary, employer-paid (0.71 wk); 17.86 rounds to 18 (the Germany 66.87-to-67 convention). Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. WATCH ITEM: Lei 15.371 (sanctioned 2026-03-31) phases paternity to 10 days from 2027-01-01 (18.6 wk), 15 from 2028 (19.3), 20 from 2029 (20), with an INSS-reimbursed salario-paternidade at full remuneration; re-touch each January. Excluded: Programa Empresa Cidada 180/20-day extensions (employer opt-in via tax incentive) and collective-agreement top-ups." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Brazilian private consultation and co-payment pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "CHECKED for the premium-as-out-of-pocket error and CLEAR: SUS is free at the point of use and these values are verified far below any real plano de saude premium, so they are co-pays and cash consultations rather than the plan itself. Two city tiers preserved. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "non_cash.commute_minutes_avg": { "source": "IBGE, Censo Demografico 2022, SIDRA table 10331, Regiao Metropolitana de Porto Alegre (34 municipios). https://sidra.ibge.gov.br/tabela/10331", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Tempo habitual de deslocamento de casa para o trabalho", "year": "2022", "note": "One way by construction; band-midpoint mean." }, "non_cash.sick_leave": { "source": "Lei 8.213/1991 (Plano de Benefícios da Previdência Social) , Art. 60 §3º (employer; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": "https://www.gov.br/inss/pt-br/direitos-e-deveres/inscricao-e-contribuicao/tabela-de-contribuicao-mensal", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Receita Federal, PGBL e VGBL https://www.gov.br/receitafederal/pt-br/acesso-a-informacao/perguntas-frequentes/imposto-de-renda/dirpf/declaracao/pgvl-vgbl (Lei 9.532/97 art. 11)", "last_updated": "2026-08-11", "confidence": "high", "note": "PGBL contributions are deductible up to 12% of rendimentos tributaveis, i.e. 12% of gross for an employee. No monetary ceiling, so no max key. VGBL is NOT deductible. The deduction requires the COMPLETE declaration and this value assumes it, which is consistent with the branch: Sao Paulo carries no standard_deduction, so the engine already models a complete/itemised basis, and at BRL 300k the 12% (R$36,000) beats the simplificado cap (R$17,640) so choosing complete is genuinely rational. Lei 15.270/2025, in force 1 Jan 2026, leaves art. 11 untouched. Never binds at 6% or 10%: pure correctness insurance." } }, "brasilia": { "tax.filing_basis": { "source": "PwC WWTS - Brazil (Tax administration): spouses may elect to file jointly or separately; individual is the two-earner default.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Brazil 2026 IRPF: peer-identical with São Paulo/Rio/Belo Horizonte (national; now peer-consistency-enforced)", "last_updated": "2026-07-19", "confidence": "high" }, "tax.sales_tax_rate": { "source": "ICMS/ISS composite modeling convention 0.25 (inherited, uniform across Brazil cities)", "last_updated": "2026-07-19", "confidence": "high" }, "healthcare._summary": { "source": "SUS public + private top-up plan (national baseline, inherited from SP/Rio/BH)", "last_updated": "2026-07-19", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Bolsa Família means-tested, 0 for tech income (national, inherited)", "last_updated": "2026-07-19", "confidence": "high" }, "non_cash._summary": { "source": "Brazilian labour law CLT (national, inherited); 30 calendar days vacation = 22 working days; 120-day maternity (17 wks)", "last_updated": "2026-07-19", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "IPTU DF residential 0.30% statutory (Secretaria de Economia do DF); GDF-set, no municipality", "last_updated": "2026-07-19", "confidence": "high" }, "housing._summary": { "source": "FipeZap Brasília 2025-26 (top-3 national price/m2) + Secovi-DF + Numbeo; Asa Sul/Asa Norte/Águas Claras band", "last_updated": "2026-07-19", "confidence": "high" }, "housing.rent_percentiles": { "source": "Zap Imoveis glue-API listing-count CDF, 2026-08-18 (apartments for rent; the addressCity filter does NOT match DF listings in any spelling - resolved to addressState=Distrito Federal alone, sample 100% city Brasilia with RAs as neighborhoods, so the pool covers the WHOLE Federal District incl. satellite RAs, the correct citywide scope; bare aluguel semantics proven on Rio this run; asking BRL/month): 2 quartos P25/med/P75 = 1,317/2,120/3,644 (n=322) -> 0.621/1.719; 1 quarto 0.568/1.745 (n=255). QuintoAndar corroboration (independent, duplicate-free, n=50 exact-2): 0.69/1.58; exact-1 0.63/1.45 - direction confirmed on a second pool.", "last_updated": "2026-08-18", "confidence": "medium-low", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW on both edges, both slices, both portals (2Q -0.199/+0.419). Applied the Zap primary 0.62/1.72 (nearest; per-city p75 guard exception; the QA-vs-Zap upper-edge gap 1.58-vs-1.72 is recorded, not averaged - Zap primary for n and granularity, the wider read conservative, mumbai precedent). THIN POOL recorded: Zap is weak in DF (n=322; the local leader DFimoveis was not tested - quantile noise at this n is a few percent, nowhere near the 0.62->0.76 gap). The DF-wide pool spans Plano Piloto premium and satellite-RA cheap stock - structurally a dual market like Rio. 2 quartos anchor per the flagship convention. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-brazil.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (C). NOTE the stored 2BR median (4,100) reads as Plano-Piloto/premium-basis against the measured DF-wide ~2,120 (Zap) / ~2,500 (QA) - applying DF-wide ratios around a premium-basis median is a recorded wedge (karachi-class), and the median level needs its own adjudication row if ever touched; flagged LOUDLY, out of scope here." }, "childcare._summary": { "source": "Private creche/escola infantil DF market surveys 2025-26 (civil-servant demand premium); no cash subsidy at tech income", "last_updated": "2026-07-19", "confidence": "medium" }, "food._summary": { "source": "Numbeo Brasília 2026, cross-checked vs Rio peer", "last_updated": "2026-07-19", "confidence": "medium" }, "transit._summary": { "source": "BRB Mobilidade bus+Metrô-DF blended fare ~R$5 (2026) x 60 trips/mo peer convention", "last_updated": "2026-07-19", "confidence": "medium" }, "discretionary._summary": { "source": "Numbeo Brasília leisure prices, peer-aligned", "last_updated": "2026-07-19", "confidence": "medium" }, "travel_home._summary": { "source": "BSB->US economy round trip (direct LATAM MIA/MCO or via GRU), 2025-26 fare scans", "last_updated": "2026-07-19", "confidence": "medium" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Brasília. Its basis is converted with the rest of Brazil and its rank inside Brazil is preserved. Prior source: Glassdoor/gov-IT salary data Brasília 2025-26 (SERPRO/DATAPREV/BB Tecnologia + contractors)", "last_updated": "2026-08-11", "confidence": "low", "note": "0.29 x 0.4688 = 0.1359, R5-rounded down to 0.13. 0.4688 is the MEDIAN of (re-derived / previous) across the 1 Brazil city that DO carry an all-levels measurement (São Paulo 0.32 to 0.15), so Brasília moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Brasília's rank inside Brazil is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.29 — public-sector salary floor lifts median; thinner private startup market than SP/Campinas." }, "non_cash.parental_leave_capped": { "source": "Lei 8.213/91 art. 72 (salario-maternidade = remuneracao integral, paid by the empresa and compensated against INSS contributions, art. 72 par. 1); CF art. 248 (benefits outside the RGPS teto observe the art. 37 XI ceiling, the STF-minister subsidio); STF ADI 1946 (salario-maternidade exempt from the RGPS teto); subsidio = R$ 46,366.19/month from 1 Feb 2025, Lei 14.520/2023 https://www.planalto.gov.br/ccivil_03/_ato2023-2026/2023/lei/l14520.htm (final step; the Dec-2025 judiciary raise covers servidores, not ministers)", "last_updated": "2026-08-17", "confidence": "high", "note": "The 17.143 maternity weeks pay 100% of remuneration up to the constitutional teto of R$ 46,366.19/month (R$ 556,394.28/yr); the 0.857-wk paternity remainder stays on parental_leave_paid_pct 100 (ADCT art. 10 par. 1, employer-paid at full salary, no ceiling). The ceiling that binds a CLT empregada is CF art. 248's STF-minister subsidio, NOT the RGPS teto (R$ 8,475.55 in 2026), per ADI 1946; the RGPS teto does bind the other claimant categories our ICP is not in. Refutation branch checked and rejected: no statute or binding precedent obliges the employer to top up above the teto, and modelling the cap is the conservative call. At the R$ 782,000/yr base scenario the maternity slice pays 71.2% and the blended 18-wk rate is 72.6%; below R$ 556k/yr it stays an honest 100%. UNIT: the subsidio is a MONTHLY figure, revised only by lei (last step 1 Feb 2025). Research: docs/superpowers/research/2026-08-17-parental-leave-caps.md." }, "non_cash.parental_leave_weeks": { "source": "CF art. 7 XVIII (120 dias); CLT art. 392; Lei 8.213/91 arts. 71-73 (salario-maternidade = remuneracao integral); ADCT art. 10 par. 1 (5 days paternity, in force through 2026); Lei 15.371 phase-in: Senado https://www12.senado.leg.br/noticias/materias/2026/04/01/sancionado-o-aumento-gradual-da-licenca-paternidade", "last_updated": "2026-08-06", "confidence": "high", "note": "18 wk = 120 days maternity at 100% (17.143 wk; capped at the constitutional STF-minister teto, BRL 46,366.19/month, modelled in non_cash.parental_leave_capped as of 2026-08-17; the 2026-08-06 judgment that the teto is irrelevant to any ICP was corrected, it binds at the USD-anchored scenarios this site prices, e.g. the R$ 782,000/yr base scenario) + 5 days paternity at full salary, employer-paid (0.71 wk); 17.86 rounds to 18 (the Germany 66.87-to-67 convention). Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. WATCH ITEM: Lei 15.371 (sanctioned 2026-03-31) phases paternity to 10 days from 2027-01-01 (18.6 wk), 15 from 2028 (19.3), 20 from 2029 (20), with an INSS-reimbursed salario-paternidade at full remuneration; re-touch each January. Excluded: Programa Empresa Cidada 180/20-day extensions (employer opt-in via tax incentive) and collective-agreement top-ups." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Brazilian private consultation and co-payment pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "CHECKED for the premium-as-out-of-pocket error and CLEAR: SUS is free at the point of use and these values are verified far below any real plano de saude premium, so they are co-pays and cash consultations rather than the plan itself. Two city tiers preserved. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Brazilian private consultation and co-payment pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "IBGE, Censo Demografico 2022, SIDRA table 10331, Area Metropolitana de Brasilia (DF plus 12 Entorno municipios in Goias). https://sidra.ibge.gov.br/tabela/10331", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Tempo habitual de deslocamento de casa para o trabalho", "year": "2022", "note": "One way by construction; band-midpoint mean. Metropolitan basis chosen over the DF alone, and it is the longer of the two." }, "non_cash.sick_leave": { "source": "Lei 8.213/1991 (Plano de Benefícios da Previdência Social) , Art. 60 §3º (employer; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": "https://www.gov.br/inss/pt-br/direitos-e-deveres/inscricao-e-contribuicao/tabela-de-contribuicao-mensal", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Receita Federal, PGBL e VGBL https://www.gov.br/receitafederal/pt-br/acesso-a-informacao/perguntas-frequentes/imposto-de-renda/dirpf/declaracao/pgvl-vgbl (Lei 9.532/97 art. 11)", "last_updated": "2026-08-11", "confidence": "high", "note": "PGBL contributions are deductible up to 12% of rendimentos tributaveis, i.e. 12% of gross for an employee. No monetary ceiling, so no max key. VGBL is NOT deductible. The deduction requires the COMPLETE declaration and this value assumes it, which is consistent with the branch: Sao Paulo carries no standard_deduction, so the engine already models a complete/itemised basis, and at BRL 300k the 12% (R$36,000) beats the simplificado cap (R$17,640) so choosing complete is genuinely rational. Lei 15.270/2025, in force 1 Jan 2026, leaves art. 11 untouched. Never binds at 6% or 10%: pure correctness insurance." } }, "campinas": { "tax.filing_basis": { "source": "PwC WWTS - Brazil (Tax administration): spouses may elect to file jointly or separately; individual is the two-earner default.", "last_updated": "2026-07-26", "note": "2026-07-26 cross-country filing-basis sweep: separately-assessed regime; activates the per-filer split in compute() for two-earner households. Single filers are unaffected (the flag only branches when has_partner and partner_gross > 0)." }, "tax.federal_brackets": { "source": "Brazil 2026 IRPF: peer-identical with São Paulo/Rio/Belo Horizonte (national; now peer-consistency-enforced)", "last_updated": "2026-07-19", "confidence": "high" }, "tax.sales_tax_rate": { "source": "ICMS/ISS composite modeling convention 0.25 (inherited, uniform across Brazil cities)", "last_updated": "2026-07-19", "confidence": "high" }, "healthcare._summary": { "source": "SUS public + private top-up plan (national baseline, inherited from SP/Rio/BH)", "last_updated": "2026-07-19", "confidence": "high" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Bolsa Família means-tested, 0 for tech income (national, inherited)", "last_updated": "2026-07-19", "confidence": "high" }, "non_cash._summary": { "source": "Brazilian labour law CLT (national, inherited); 30 calendar days vacation = 22 working days; 120-day maternity (17 wks)", "last_updated": "2026-07-19", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "IPTU Campinas ~1.0% residential (Prefeitura de Campinas tabela 2025-26)", "last_updated": "2026-07-19", "confidence": "medium" }, "housing._summary": { "source": "FipeZap Campinas 2025-26 (~R$7.9k/m2 sale) + Secovi-SP + Numbeo; Cambuí/Taquaral band, SP-metro demand pull on rents", "last_updated": "2026-07-19", "confidence": "high" }, "housing.rent_percentiles": { "source": "Zap Imoveis glue-API listing-count CDF, 2026-08-18 (apartments for rent; the city+state combo silently returns ZERO for Campinas - resolved via addressLocationId=BR>Sao Paulo>NULL>Campinas, cross-confirmed by addressCity alone, 11,644 vs 11,645; bare aluguel semantics proven on Rio this run; asking BRL/month): 2 quartos P25/med/P75 = 2,496/3,598/5,427 (n=3,851) -> 0.694/1.508; 1 quarto 0.774/1.246 (n=3,612, INSIDE tolerance both edges).", "last_updated": "2026-08-18", "confidence": "medium", "note": "REBASED 2026-08-18 (ledger cluster housing-0818) from the calibrated default 0.82/1.24, TOO NARROW at the primary slice (2Q -0.126/+0.268) while the 1Q slice ROUGHLY HOLDS (0.774/1.246) - a genuine slice divergence: the Campinas 1Q pool is a compressed university-city new-studio market (mass at 3,000-5,000) while 2Q spans old cheap stock to Cambui premium. The primary 2BR slice decides, per the flagship convention; the divergence is recorded as a known single-pair-schema residual (lisbon T3 precedent, opposite direction). Applied 0.69/1.51 (nearest). Informal-tail truncation - conservative. Band edge = median_rent_by_bedroom x multiplier, computed at display time. Evidence + raw CDFs: docs/superpowers/research/2026-08-18-housing-dual-market-sweep-brazil.md; decision: docs/superpowers/research/2026-08-18-housing-dual-market-sweep.md (C). Level observations logged out of scope." }, "childcare._summary": { "source": "Private creche/escola infantil Campinas market surveys 2025-26; no cash subsidy at tech income", "last_updated": "2026-07-19", "confidence": "medium" }, "food._summary": { "source": "Numbeo Campinas 2026, cross-checked vs SP/Rio peers", "last_updated": "2026-07-19", "confidence": "medium" }, "transit._summary": { "source": "EMDEC/InterCamp single fare R$6.00 (2026) x 60 trips/mo peer convention", "last_updated": "2026-07-19", "confidence": "high" }, "discretionary._summary": { "source": "Numbeo Campinas leisure prices, peer-aligned", "last_updated": "2026-07-19", "confidence": "medium" }, "travel_home._summary": { "source": "VCP (Viracopos, Azul direct FLL/MCO) or GRU ->US economy round trip, 2025-26 fare scans", "last_updated": "2026-07-19", "confidence": "medium" }, "compensation_index": { "source": "Derived, not observed: no all-levels total-comp measurement exists for Campinas. Its basis is converted with the rest of Brazil and its rank inside Brazil is preserved. Prior source: Glassdoor/Levels.fyi Campinas senior SWE 2025-26 (Unicamp/CPqD/IBM/Samsung R&D/Eldorado/Venturus cluster)", "last_updated": "2026-08-11", "confidence": "low", "note": "0.3 x 0.4688 = 0.1406, R5-rounded down to 0.14. 0.4688 is the MEDIAN of (re-derived / previous) across the 1 Brazil city that DO carry an all-levels measurement (São Paulo 0.32 to 0.15), so Campinas moves onto the anchor by the amount its own country measured rather than by a global average. This is a BASIS move, not a re-measurement: Campinas's rank inside Brazil is unchanged and still rests on its prior evidence, which is why confidence is low. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.30 — matches Rio; SP proximity pulls salaries above BH/Curitiba." }, "non_cash.parental_leave_capped": { "source": "Lei 8.213/91 art. 72 (salario-maternidade = remuneracao integral, paid by the empresa and compensated against INSS contributions, art. 72 par. 1); CF art. 248 (benefits outside the RGPS teto observe the art. 37 XI ceiling, the STF-minister subsidio); STF ADI 1946 (salario-maternidade exempt from the RGPS teto); subsidio = R$ 46,366.19/month from 1 Feb 2025, Lei 14.520/2023 https://www.planalto.gov.br/ccivil_03/_ato2023-2026/2023/lei/l14520.htm (final step; the Dec-2025 judiciary raise covers servidores, not ministers)", "last_updated": "2026-08-17", "confidence": "high", "note": "The 17.143 maternity weeks pay 100% of remuneration up to the constitutional teto of R$ 46,366.19/month (R$ 556,394.28/yr); the 0.857-wk paternity remainder stays on parental_leave_paid_pct 100 (ADCT art. 10 par. 1, employer-paid at full salary, no ceiling). The ceiling that binds a CLT empregada is CF art. 248's STF-minister subsidio, NOT the RGPS teto (R$ 8,475.55 in 2026), per ADI 1946; the RGPS teto does bind the other claimant categories our ICP is not in. Refutation branch checked and rejected: no statute or binding precedent obliges the employer to top up above the teto, and modelling the cap is the conservative call. At the R$ 782,000/yr base scenario the maternity slice pays 71.2% and the blended 18-wk rate is 72.6%; below R$ 556k/yr it stays an honest 100%. UNIT: the subsidio is a MONTHLY figure, revised only by lei (last step 1 Feb 2025). Research: docs/superpowers/research/2026-08-17-parental-leave-caps.md." }, "non_cash.parental_leave_weeks": { "source": "CF art. 7 XVIII (120 dias); CLT art. 392; Lei 8.213/91 arts. 71-73 (salario-maternidade = remuneracao integral); ADCT art. 10 par. 1 (5 days paternity, in force through 2026); Lei 15.371 phase-in: Senado https://www12.senado.leg.br/noticias/materias/2026/04/01/sancionado-o-aumento-gradual-da-licenca-paternidade", "last_updated": "2026-08-06", "confidence": "high", "note": "18 wk = 120 days maternity at 100% (17.143 wk; capped at the constitutional STF-minister teto, BRL 46,366.19/month, modelled in non_cash.parental_leave_capped as of 2026-08-17; the 2026-08-06 judgment that the teto is irrelevant to any ICP was corrected, it binds at the USD-anchored scenarios this site prices, e.g. the R$ 782,000/yr base scenario) + 5 days paternity at full salary, employer-paid (0.71 wk); 17.86 rounds to 18 (the Germany 66.87-to-67 convention). Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch2.md. WATCH ITEM: Lei 15.371 (sanctioned 2026-03-31) phases paternity to 10 days from 2027-01-01 (18.6 wk), 15 from 2028 (19.3), 20 from 2029 (20), with an INSS-reimbursed salario-paternidade at full remuneration; re-touch each January. Excluded: Programa Empresa Cidada 180/20-day extensions (employer opt-in via tax incentive) and collective-agreement top-ups." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Brazilian private consultation and co-payment pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "CHECKED for the premium-as-out-of-pocket error and CLEAR: SUS is free at the point of use and these values are verified far below any real plano de saude premium, so they are co-pays and cash consultations rather than the plan itself. Two city tiers preserved. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "healthcare.annual_out_of_pocket_avg": { "source": "World Bank SH.XPD.OOPC.PC.CD household out-of-pocket per capita cross-checked against the OWID PPP series, plus Brazilian private consultation and co-payment pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "IBGE, Censo Demografico 2022, SIDRA table 10331, Regiao Metropolitana de Campinas (20 municipios). https://sidra.ibge.gov.br/tabela/10331", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Tempo habitual de deslocamento de casa para o trabalho", "year": "2022", "note": "One way by construction; band-midpoint mean." }, "non_cash.sick_leave": { "source": "Lei 8.213/1991 (Plano de Benefícios da Previdência Social) , Art. 60 §3º (employer; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-G.md", "source_url": "https://www.gov.br/inss/pt-br/direitos-e-deveres/inscricao-e-contribuicao/tabela-de-contribuicao-mensal", "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "Receita Federal, PGBL e VGBL https://www.gov.br/receitafederal/pt-br/acesso-a-informacao/perguntas-frequentes/imposto-de-renda/dirpf/declaracao/pgvl-vgbl (Lei 9.532/97 art. 11)", "last_updated": "2026-08-11", "confidence": "high", "note": "PGBL contributions are deductible up to 12% of rendimentos tributaveis, i.e. 12% of gross for an employee. No monetary ceiling, so no max key. VGBL is NOT deductible. The deduction requires the COMPLETE declaration and this value assumes it, which is consistent with the branch: Sao Paulo carries no standard_deduction, so the engine already models a complete/itemised basis, and at BRL 300k the 12% (R$36,000) beats the simplificado cap (R$17,640) so choosing complete is genuinely rational. Lei 15.270/2025, in force 1 Jan 2026, leaves art. 11 untouched. Never binds at 6% or 10%: pure correctness insurance." } }, "ljubljana": { "tax.federal_brackets": { "source": "Dohodnina lestvica 16/26/33/39/50 (ZDoh-2); official gazetted 2026 thresholds (9,721.43 / 28,592.44 / 57,184.88 / 82,346.23) per Pravilnik, Uradni list RS 104/2025. FURS 2026 lestvica: https://www.fu.gov.si/fileadmin/Internet/Davki_in_druge_dajatve/Podrocja/Dohodnina/Letna_odmera_dohodnine/Opis/Lestvica_za_leto_2026.docx", "last_updated": "2026-07-19", "confidence": "medium", "note": "DRAFT pending 2026 valorization confirmation from FURS; flagged for Skyler review." }, "tax.social_security_rate": { "source": "Employee contributions 22.10% (PIZ 15.50 + ZZ 6.36 + unemployment 0.14 + parental 0.10), uncapped, PIT-deductible (ZPIZ-2/ZZVZZ; PwC Tax Summaries Slovenia) + 1.0% long-term-care contribution (ZDOsk-1, from 2025-07-01) = 0.231 modeled", "last_updated": "2026-07-31", "confidence": "high", "note": "LTC 1% inclusion and its PIT-deductibility VERIFIED 2026-07-31 against FURS (statutory-health batch 2). It rides inside this 0.231 rate, so the engine already deducts it from the PIT base." }, "tax.ozp_annual": { "source": "OZP (obvezni zdravstveni prispevek) EUR 37.17/mo to Feb 2026, EUR 39.36/mo from 1 Mar 2026 (ZZZS: https://zavezanec.zzzs.si/prispevki-za-obvezno-zdravstveno-zavarovanje/obvezni-zdravstveni-prispevek/; ZDS 2026 notice). 2026 blended: 2x37.17 + 10x39.36 = 467.94 -> 468/yr per INSURED ADULT. Forward run rate 472.32; reindexed each 1 March to prior-year average gross wage growth", "last_updated": "2026-07-31", "confidence": "high", "note": "Replaced abolished complementary insurance 2024-01-01. Charged per insured ADULT, not per household: children are exempt as insured family members, but a non-working spouse is individually liable and gets their own monthly payment slip from ZZZS, so a single-earner couple owes 2x (FURS/ZZZS, verified 2026-07-31). PIT-DEDUCTIBLE: FURS is explicit that OZP reduces the income-tax base, so computeSloveniaTaxes takes the filer's own OZP off the base (a non-earning spouse has no income to set theirs against)." }, "tax.sales_tax_rate": { "source": "Slovenia standard VAT 22% (ZDDV-1; FURS)", "last_updated": "2026-07-19", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "NUSZ (nadomestilo za uporabo stavbnega zemljisca) municipal charge, Ljubljana; no national residential property tax. Effective ~0.1% of value approximation", "last_updated": "2026-07-19", "confidence": "low", "note": "DRAFT approximation modeled like Vilnius; NUSZ is area-based (EUR/m2 points), not ad valorem. Pending verification." }, "tax.standard_deduction_single": { "source": "Splosna olajsava EUR 5,551.93/yr (gazetted 2026 value, Uradni list RS 104/2025; up from 5,260.60 in 2025). Slovenia files individually, so each earner gets the base allowance; dependent-child and dependent-spouse allowances (EUR 2,995.83 / 3,256.77) apply on top when relevant", "last_updated": "2026-07-19", "confidence": "medium", "note": "DRAFT pending 2026 valorization confirmation; flagged for Skyler review." }, "tax.filing_basis": { "source": "ZDoh-2 - strictly individual taxation; no joint assessment in Slovenia", "last_updated": "2026-07-19", "confidence": "high", "note": "Activates the per-filer split in compute(), so a two-earner couple runs two single-filer clones and charges two OZP. A single-earner couple never reaches that path, so computeSloveniaTaxes counts insured adults directly and charges two there as well." }, "housing.median_rent_by_bedroom": { "source": "Numbeo Ljubljana (802 entries, 2026-07: 1BR centre 944/outside 769 -> 850 blend; 3BR 1,730/1,275 -> 1,500) corroborated by nepremicnine.net garsonjera listings EUR 600-800 (https://www.nepremicnine.net/oglasi-oddaja/ljubljana-mesto/stanovanje/garsonjera/) and rentola.si", "last_updated": "2026-07-19", "confidence": "medium", "note": "No official rental medians exist: the GURS 2025 annual report states landlords under-report rents/prices, so portal+Numbeo blend is the best available. Slovenian '2-sobno' = US 1BR; ladder uses US-bedroom convention like other EU cities." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "housing.median_home_price_by_bedroom": { "source": "GURS Letno porocilo o slovenskem nepremicninskem trgu 2025: Ljubljana used-apartment MEDIAN EUR 5,050/m2 (https://www.e-prostor.gov.si/fileadmin/Podrocja/Trg_vrednosti_nep/Trg_nepremicnin/Porocila_o_trgu_nepremicnin/2025/Letno_porocilo_za_leto_2025.pdf) x typical sqm 35/50/70/95. Corroboration: Investropa ~4,750/m2, Numbeo 5,545 centre/4,305 outside", "last_updated": "2026-07-19", "confidence": "high", "note": "Transaction-register based (1,802 sales 2025, +4.8% y/y per cenenepremicnin.com aggregation)." }, "housing.utilities_monthly": { "source": "Numbeo Ljubljana 85m2 basic utilities EUR 271 (range 200-380), trimmed ~12% per the Numbeo-runs-high rule -> 240", "last_updated": "2026-07-19", "confidence": "medium", "note": "No official composite utility statistic; SURS publishes only per-energy-carrier price indices." }, "housing.renters_insurance_monthly": { "source": "Zavago 2026: home-contents insurance ~EUR 120/yr without earthquake cover (https://zavago.si/novica/zavarovanje-hise-in-stanovanja/) -> 10/mo", "last_updated": "2026-07-19", "confidence": "medium" }, "housing.internet_phone_monthly": { "source": "Numbeo Ljubljana: fiber 38.51 + mobile w/data 15.71 = 54, rounded to 50 for local bundle pricing (Telekom/T-2/A1 bundles)", "last_updated": "2026-07-19", "confidence": "medium" }, "healthcare.premium_monthly_by_family_size": { "source": "ZZZS compulsory insurance via payroll; complementary top-up insurance ABOLISHED 2024-01-01, replaced by flat OZP modeled in tax.ozp_annual. No separate premium exists -> 0 all tiers (tax_funded regime)", "last_updated": "2026-07-19", "confidence": "high" }, "healthcare.annual_out_of_pocket_avg": { "source": "WHO/OECD Slovenia Country Health Profile, out-of-pocket 11.7% of health spending in 2024 and among the lowest in the EU https://www.oecd.org/en/publications/health-at-a-glance-2025_15a55280-en/slovenia_766a9bd5-en.html + Ljubljana self-pay specialist pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "healthcare.dental_vision_monthly": { "source": "Ljubljana private dental price lists (vvdental.si/cenik: exam EUR 25-45, scaling ~EUR 60; zobozdravstvo-sekulic.si) + glasses ~EUR 250/2-3yrs. EUR 20-35/mo estimate -> 28", "last_updated": "2026-07-19", "confidence": "medium" }, "childcare.infant_monthly": { "source": "Ljubljana vrtec cena programa EUR 732/mo first age group (Uradni list 2023-01-2834, unchanged through 2025/26); top-income bracket (9th, 77%) minus MOL municipal discount = EUR 419.91/mo actually paid (vrtec-mladirod.si bracket table)", "last_updated": "2026-07-19", "confidence": "high", "note": "MOL-resident figure. Mayor committed to no increase until summer 2026; recheck after Sept 2026." }, "childcare.preschool_monthly": { "source": "Ljubljana vrtec cena programa EUR 531/mo second age group (3-6); top bracket 77% minus MOL discount = EUR 323.42/mo (Uradni list 2023-01-2834; vrtec-mladirod.si)", "last_updated": "2026-07-19", "confidence": "high", "note": "Younger sibling attending simultaneously is FREE (state-funded); 3rd+ child free regardless. Both gratuities are left unpriced (we bill every kid at full rate), so reality favors Slovenia further than we show." }, "childcare.after_school_monthly": { "source": "Podaljsano bivanje (grades 1-5) free to parents (21. clen ZOsn); parents pay meals: snack EUR 1.10 + lunch EUR 2.53-2.79 + pm snack EUR 1.20/day (Ljubljana school price lists 2025/26) ~ EUR 90/mo + modest activities -> 110", "last_updated": "2026-07-19", "confidence": "medium" }, "childcare.subsidy_per_kid_annual": { "source": "Income-based pricing is already applied in the fee itself (ZVrtP bracket scale 0-77%); no separate cash subsidy for top-bracket households -> 0", "last_updated": "2026-07-19", "confidence": "high" }, "food.groceries_monthly_per_adult": { "source": "My Global Slovenia: EUR 250-350/person/mo supermarket spend (https://www.myglobal.si/blog/how-much-does-it-cost-to-live-in-slovenia/); trek.zone daily basket EUR 6.5-13 -> 300", "last_updated": "2026-07-19", "confidence": "medium" }, "food.groceries_monthly_per_kid": { "source": "~60% of adult basket (household-scaling convention used across cities) -> 175", "last_updated": "2026-07-19", "confidence": "medium" }, "food.dining_out_monthly_baseline": { "source": "Numbeo Ljubljana: inexpensive meal EUR 14, mid-range dinner for two EUR 60; ~5 cheap meals + 2 mid-range + cafes -> 180", "last_updated": "2026-07-19", "confidence": "medium" }, "transit.monthly_pass": { "source": "Urbana 'General 1' monthly pass EUR 37.00, official JHL price list (https://www.jhl.si/en/single-city-card-urbana/prices); LPP price list 2025-10-01 (https://www.lpp.si/en/public-transport/price-list)", "last_updated": "2026-07-19", "confidence": "high" }, "transit.annual_car_cost_avg": { "source": "Composite: depreciation ~1,700 + insurance ~450 (Statista SI motor) + fuel ~1,000 @ EUR 1.55/L (fuel-prices.eu) + maintenance ~450 + registration ~150 (2slovenia.eu) + parking ~250 = ~4,000", "last_updated": "2026-07-19", "confidence": "medium" }, "discretionary.entertainment_monthly_baseline": { "source": "Numbeo Ljubljana gym 45.91 + 2x cinema ~9 + streaming ~25 + ~60 events -> 150", "last_updated": "2026-07-19", "confidence": "medium" }, "discretionary.local_vacation_annual_baseline": { "source": "Slovenia ~EUR 119/day (amigosim.com), Croatia coast mid-range EUR 120-180/day (theworldwasherefirst.com); one coast week + Slovene weekends -> 2,000", "last_updated": "2026-07-19", "confidence": "medium" }, "travel_home.avg_round_trip_per_person_to_other_city": { "source": "SEA-LJU economy RT: Kayak from $490, sampled $958-973; FareCompare avg $1,395 (https://www.farecompare.com/flights/Seattle-SEA/Ljubljana-LJU/market.html). Typical ~$1,080 x FX 0.874 = EUR 950", "last_updated": "2026-07-19", "confidence": "medium", "note": "FX rate used: USD_EUR 0.874 (build/fx-rates.json vintage 2026-07)." }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Otroski dodatek is means-tested (8 brackets); entitlement ends above ~EUR 1,328.28/family-member/mo (2026 uskladitev; zastarse.si) -> EUR 0 for the ICP household", "last_updated": "2026-07-19", "confidence": "high", "note": "A EUR 100k household of 4 sits at ~EUR 2,000+/member/mo - well above the top bracket." }, "non_cash.vacation_days_statutory": { "source": "ZDR-1 statutory minimum 4 weeks = 20 days (spot.gov.si/en/info/employees/annual-leave-in-slovenia); +1 day per child under 15, +3 age 55+; collective agreements add more (typical professional 22-26)", "last_updated": "2026-07-19", "confidence": "high", "note": "Statutory-floor convention per _national_fields.json. Mandatory regres (holiday allowance) >= EUR 1,481.88/yr paid by July 1 is left unpriced - a real cash benefit the model omits, so reality favors Slovenia." }, "non_cash.public_holidays": { "source": "15 statutory work-free dates, of which 2 (Easter + Whit Sunday) always fall on Sunday -> 13 year-neutral weekday-capable holidays (prazniki.si)", "last_updated": "2026-07-19", "confidence": "high" }, "non_cash.parental_leave_weeks": { "source": "e-uprava.gov.si: materinski dopust 105 days at 100% (uncapped); ocetovski dopust 15 days; starsevski dopust 160 days per parent, 60 non-transferable / 100 transferable; verified 2026-08-06", "last_updated": "2026-08-06", "confidence": "high", "note": "63 = 440 paid days / 7: 105 materinski + 15 ocetovski + 160 starsevski per parent x 2. The both-parents flag is deliberately NOT set per the Sweden precedent: 100 of each parent's 160 days are transferable, making it functionally a family pot with 60-day reserves. Confirmed on the household-inclusive basis 2026-08-06. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "e-uprava.gov.si: 100% for all blocks; paternity and parental benefit capped at 2.5x the average monthly gross wage, EUR 6,340.08/month in 2026; maternity benefit uncapped", "last_updated": "2026-08-06", "confidence": "high", "note": "CORRECTION 2026-08-06: the previous note claimed the 2.5x-average cap does not bind below about EUR 150k; in fact EUR 6,340.08/month is about EUR 76k/yr, so it binds within the Ljubljana tech band; at EUR 90k the duration-weighted effective rate is about 88%. Materinski (maternity) benefit is uncapped. Recorded for the standing caps-convention item." }, "non_cash.universal_healthcare": { "source": "ZZZS compulsory universal coverage funded by payroll contributions + flat OZP; tax_funded regime in data/_healthcare_regime.json", "last_updated": "2026-07-19", "confidence": "high" }, "non_cash.commute_minutes_avg": { "source": "SURS (Statistical Office of Slovenia), SI-STAT table 2281442S, survey Dnevna mobilnost potnikov, Mestna obcina Ljubljana. https://pxweb.stat.si/SiStatData/pxweb/sl/Data/-/2281442S.PX", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Povprecen cas, porabljen za poti, na posamezno pot (v min), namen poti Delo/sluzba", "year": "2025", "note": "31.37 min. ONE WAY as published: the table unit is literally per individual trip, so no conversion and no halving. Work-only, since education is a separate purpose row at 38.93. Mode-blind. Replaces a Numbeo citation of 26.35 that ran about 5 min low against the official series. Series values: 30.36 (2017), 31.99 (2021), 31.37 (2025). Note Eurostat Urban Audit TT1019V sourced its Ljubljana 2017 cell from this same SURS series. Eurostat LFS 2019 puts Slovenian cities at 21, but that is the known LFS-versus-travel-diary instrument gap, and our declared unit is door-to-door, so the diary source is correct." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Ljubljana: EUR 44,753, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-europe.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "EUR 44,753 / 0.867 = USD 51,618; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.2065, R5-rounded down to 0.2. REPLACES 0.34 (2026-07-19), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Its senior p25 (EUR 35,668) sits BELOW its all-levels median, which means a very thin senior bucket. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): 0.34; naive USD ratio is ~0.27 but the repo convention stores tech-weighted indices above naive (methodology Sec. 2-3): peer-calibrated above Belgrade (0.32, lower gross), below Warsaw (0.36) and Vilnius (0.38). Glassdoor SI figures are monthly gross; regres/13th-month not included." }, "healthcare.statutory_employee": { "source": "FURS contributions brochure (16th ed., May 2026) and FURS employment-income guidance: employee total 23.10% = PIZ 15.50 + ZZ 6.36 + LTC 1.00 + unemployment 0.14 + parental 0.10; ZPSV art. 3 sets no ceiling; ZPIZ-2 art. 145 scopes its 3.5x cap to arts. 15/16/17 (self-employed, partners, farmers), not to art. 14 employees. OZP amounts from the FURS OZP table with gazette citations: EUR 37.17/month to 2026-02-28, EUR 39.36/month from 2026-03-01, reindexed each 1 March to prior-year average gross wage growth.", "last_updated": "2026-07-31", "confidence": "high", "note": "DISPLAY ONLY, already inside taxes.payroll. 6.36% has no internal split: parental protection 0.10% is its own line and work-injury 0.53% is employer-only. LTC 1.00% is included because the regime definition covers long-term care. BOTH PREVIOUSLY-SPECCED ENGINE ITEMS ARE NOW SHIPPED, verified 2026-08-02 by reading computeSloveniaTaxes: (1) OZP is charged per insured ADULT, not once per household. The function derives insured_adults = 1 + (has_partner ? 1 : 0) and bills ozp_annual * insured_adults, so a single-earner couple pays twice, matching FURS and ZZZS. (2) OZP REDUCES the income-tax base. The filer’s own ozp_annual is subtracted alongside social contributions before computeProgressiveTax, so it is not treated as post-tax and Slovenian tax is not overstated. This note previously described both as open defects (a ~EUR 468/yr undercharge on single-earner couples, and a possible post-tax OZP); neither is live. The tax.ozp_annual and tax.filing_basis entries already described the shipped behaviour." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "WHO/OECD Slovenia Country Health Profile, out-of-pocket 11.7% of health spending in 2024 and among the lowest in the EU https://www.oecd.org/en/publications/health-at-a-glance-2025_15a55280-en/slovenia_766a9bd5-en.html + Ljubljana self-pay specialist pricing", "last_updated": "2026-08-08", "confidence": "medium", "note": "CHECKED AND REFUTED, recorded so nobody re-runs it: this is NOT the obvezni zdravstveni prispevek in disguise. The OZP (EUR 37.17/month to 2026-02-28, EUR 39.36/month from 2026-03-01, introduced when complementary insurance was abolished on 2023-12-31) is a contribution, and it is already modelled separately in tax.ozp_annual, billed per insured adult with children exempt, and deducted from the income-tax base. Row 1 is held at 550, the top of its own stated EUR 400-700 range. The refutation was reached twice independently: once here against ZZZS and RTV SLO, and once in the batch-D research against SURS and NIJZ, which found no discontinuity in the household out-of-pocket trend through the 2024 OZP transition, which is what a contribution mislabelled as out-of-pocket would have produced. Basis: the shared declaration for this field in shared_fields, which since 2026-08-08 is the single place it is stated." }, "non_cash.sick_leave": { "source": "ZDR-1 (Zakon o delovnih razmerjih, Employment Relationships Act) for delodajalec-paid nadomestilo plače; ZZVZZ (Zakon o zdravstvenem varstvu in zdravstvenem zavarovanju, Health Care and Health Insurance Act) for ZZZS-paid nadomestilo from day 31.; Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-C.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "ZDoh-2 art. 117 https://zakonodaja.com/zakon/zdoh-2/117-clen-olajsava-za-prostovoljno-dodatno-pokojninsko-zavarovanje ; 2026 indexed amount, Uradni list RS st. 2/2026 (9 Jan 2026)", "last_updated": "2026-08-11", "confidence": "high", "note": "Olajsava za prostovoljno dodatno pokojninsko zavarovanje: the annual tax base is REDUCED by the premium the taxpayer paid for themselves, limited to the LOWEST of 24% of mandatory pension/disability contributions, 5.844% of gross salary, and the indexed annual amount EUR 3,224.18. The 24%-of-contributions and 5.844%-of-gross tests are algebraically identical (24% x 24.35% total PIZ). The EUR 2,390 in the statute text is the UN-INDEXED base amount, republished each year; do not use it. An employer collective-scheme premium consumes the same ceiling." } }, "dallas": { "tax.federal_brackets": { "source": "IRS 2026 inflation-adjusted federal brackets (inherited USA baseline)", "last_updated": "2026-07-29", "confidence": "high" }, "tax.state_brackets": { "source": "Texas has NO state income tax and prohibits local income taxes", "last_updated": "2026-07-29", "confidence": "high" }, "tax.sales_tax_rate": { "source": "TX state 6.25% + Dallas city 1% + DART transit 1% = 8.25% statutory cap (Texas Comptroller / Avalara)", "last_updated": "2026-07-29", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "DFW nominal combined effective rates 2.16-2.31% by city (Dallas 2.18, Fort Worth 2.31); ACS 2024 realized 1.58% Dallas Co / 1.65% Tarrant Co", "last_updated": "2026-07-29", "confidence": "medium", "note": "Stored at 2.1 on the same nominal/recent-buyer basis as Houston 2.0 and Austin 1.9, which correctly ranks DFW above both." }, "housing._summary": { "source": "RentCafe Dallas + Fort Worth 2026 asking rents and DFW multifamily metro blend; Zillow ZHVI DFW metro $358,078", "last_updated": "2026-07-29", "confidence": "high", "note": "HUD FY2026 FMRs not used: they are 40th-percentile GROSS rents including utilities, which are a separate key here." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "KFF 2025 Employer Health Benefits Survey worker contribution ($1,440 single / $6,850 family) + Peterson-KFF out-of-pocket; Mid-A tier, matched to Houston/Nashville/Phoenix", "last_updated": "2026-07-29", "confidence": "high" }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "childcare._summary": { "source": "tooTRiS DFW center-based rates + Care.com Dallas; Texas CCDF eligibility caps near 85% of state median income", "last_updated": "2026-07-29", "confidence": "medium" }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit convention", "last_updated": "2026-07-29", "confidence": "high", "note": "Zeroed: means-tested out at ICP income. The DCFSA is a per-HOUSEHOLD benefit and lives in subsidy_household_annual." }, "childcare.income_tested_childcare": { "source": "https://www.wfsdallas.com/child-care-assistance/", "last_updated": "2026-08-05", "confidence": "high", "note": "Entry ceiling is the statewide Texas 85% SMI line (TWC BCY 2026 income limits, $92,041 for a family of 4, effective 1 Oct 2025 per WD Letter 11-25); Texas publishes no separate exit ceiling; a household under the line may qualify but Workforce Solutions Greater Dallas runs a waitlist (~12,000 families reported)." }, "food._summary": { "source": "BLS Consumer Expenditure Survey, Dallas-Fort Worth-Arlington metro 2023-24 + USDA moderate-cost food plan (TX)", "last_updated": "2026-07-29", "confidence": "medium" }, "transit._summary": { "source": "DART Local 31-Day Pass $96, derived from the published $192 Regional 31-Day under DART Local = half of Regional", "last_updated": "2026-07-29", "confidence": "medium", "note": "DART live fare table would not render; derived rather than read directly." }, "discretionary._summary": { "source": "BLS Consumer Expenditure Survey DFW metro, scaled to the Houston/Austin peer convention", "last_updated": "2026-07-29", "confidence": "medium" }, "travel_home._summary": { "source": "BTS domestic itinerary fares May 2026: DFW average $406.83 vs $369.93 national; Love Field/Southwest provides a cheaper parallel market", "last_updated": "2026-07-29", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Federal CTC $2,000/child (no TX top-up)", "last_updated": "2026-07-29", "confidence": "high", "note": "US Child Tax Credit (~$2,000/kid): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at this ICP household income." }, "non_cash._summary": { "source": "FMLA federal floor only; TX HB 2127 (2023) preempts local paid-sick-leave ordinances and the Dallas ordinance was enjoined; ACS 2024 mean commute 28.8 min", "last_updated": "2026-07-29", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Dallas: USD 133,800, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "USD 133,800 / 1 = USD 133,800; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.5352, R5-rounded down to 0.53. REPLACES 0.9 (2026-07-29), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Set at 0.90 (Atlanta/Portland peer slot) rather than the more flattering OEWS-implied 0.93; sample composition differs (DFW skews enterprise/finance, Austin skews big-tech)." }, "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 64500 USD, Dallas-Fort Worth-Arlington, TX metro area, table B20017 count-weighted)", "last_updated": "2026-07-29", "confidence": "high" }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Dallas County TX (FIPS 48113), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Texas caveat: from 2015 onward the state reports the prior year figures, so the 2022 row is 2021 fieldwork and this value is marginally understated after escalation. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $900/mo and the ages 1-2 rate $960/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,000. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: Texas has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. Texas law only enables voluntary paid-leave insurance products; no entitlement. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Dallas-Fort Worth-Arlington TX Metro Area (31000US19100). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US19100", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 28.8. Stored value was already correct; citation added." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Dallas County TX (FIPS 48113), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Texas caveat: from 2015 onward the state reports the prior year figures, so the 2022 row is 2021 fieldwork and this value is marginally understated after escalation. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,250." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "philadelphia": { "tax.federal_brackets": { "source": "IRS 2026 inflation-adjusted federal brackets (inherited USA baseline)", "last_updated": "2026-07-29", "confidence": "high" }, "tax.state_brackets": { "source": "PA flat personal income tax 3.07%, unchanged since 2004 (PA Dept of Revenue)", "last_updated": "2026-07-29", "confidence": "high" }, "tax.city_income_tax_rate": { "source": "Philadelphia resident Wage/Earnings Tax 3.735% effective July 1, 2026 (City of Philadelphia Revenue Dept); non-resident 3.425%", "last_updated": "2026-07-29", "confidence": "high", "note": "Second step of the 5-year reduction plan (3.75% pre-7/2024, 3.74% from 7/2025). A 2.235% figure circulating online is wrong." }, "tax.sales_tax_rate": { "source": "PA state 6% + Philadelphia County 2% = 8%", "last_updated": "2026-07-29", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Nominal 1.3998% of assessed value (0.6159% city + 0.7839% school); Philadelphia assesses at 100% of market value", "last_updated": "2026-07-29", "confidence": "medium", "note": "Stored at 1.15 on a market-value basis with the $100,000 Homestead Exemption applied, matching how peers are stored (NYC 0.88, Boston 1.04, DC 0.85). ACS-derived 0.82-1.13% is depressed by long-tenured owners, LOOP caps and the 10-year abatement, which do not describe the ICP." }, "housing._summary": { "source": "HUD FY2026 FMR Philadelphia metro + Zumper and Rent.com city medians; Redfin median sale $290K (3 mo ending May 2026)", "last_updated": "2026-07-29", "confidence": "high", "note": "HUD is metro-wide and Zumper is city-only; median taken. 3BR sources spread 1950-3512; luxury-listing figures discarded." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "KFF 2025 Employer Health Benefits Survey worker contribution + Peterson-KFF out-of-pocket; Mid-B tier, matched to Pittsburgh/Chicago", "last_updated": "2026-07-29", "confidence": "medium", "note": "Judgment call: Independence Blue Cross is pricier than Pittsburgh UPMC/Highmark, arguing for High tier, but Philadelphia does not out-cost Chicago. Mid-B chosen; higher local costs carried by rent, childcare and car lines." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "childcare._summary": { "source": "daycarecalc Philadelphia center-based $1,480/mo infant; southeastern PA is the state’s most expensive region; PA CCW caps at 85% of state median income and PHLpreK is capacity-limited", "last_updated": "2026-07-29", "confidence": "medium" }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit convention", "last_updated": "2026-07-29", "confidence": "high", "note": "Zeroed: means-tested out at ICP income. The DCFSA is a per-HOUSEHOLD benefit and lives in subsidy_household_annual." }, "food._summary": { "source": "RentCafe Philadelphia cost-of-living index (groceries ~4% above national, overall ~7% above)", "last_updated": "2026-07-29", "confidence": "medium" }, "transit._summary": { "source": "SEPTA Monthly TransPass+ (Bus & Metro) and Zone 1 TrailPass both $116 in 2026, reflecting the August 2025 fare increase", "last_updated": "2026-07-29", "confidence": "high" }, "discretionary._summary": { "source": "Peer banding at Philadelphia’s cost-of-living level; Jersey Shore and Poconos are cheap short-haul destinations", "last_updated": "2026-07-29", "confidence": "medium" }, "travel_home._summary": { "source": "PHL is an American Airlines hub with persistently above-average fares (BTS domestic average fare series)", "last_updated": "2026-07-29", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Federal CTC $2,000/child (no PA top-up)", "last_updated": "2026-07-29", "confidence": "high", "note": "US Child Tax Credit (~$2,000/kid): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at this ICP household income." }, "non_cash._summary": { "source": "Philadelphia Promoting Healthy Families and Workplaces Ordinance as expanded May 27, 2025 (80 hrs/yr at 50+ staff); PA has no state PFL (SB 906 not enacted as of 2026); ACS 2024 mean commute 33.2 min", "last_updated": "2026-07-29", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Philadelphia: USD 128,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "medium", "note": "USD 128,000 / 1 = USD 128,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.5120, R5-rounded down to 0.51. REPLACES 0.88 (2026-07-29), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): (none)" }, "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 70500 USD, Philadelphia-Camden-Wilmington, PA-NJ-DE-MD metro area, table B20017 count-weighted)", "last_updated": "2026-07-29", "confidence": "high" }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Philadelphia County PA (FIPS 42101), NDCP study year 2020.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2020 to 2026 by CUUR0000SEEB03 (x1.27208; 2026 = Jan-Jul mean 396.748 against 2020 annual average 311.888). Tier C: Pennsylvania has no 2021 or 2022 data; the latest observed year is 2020, escalated over the longer window at x1.27208. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $1,070/mo and the ages 1-2 rate $1,170/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,200. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.income_tested_childcare": { "source": "https://www.pa.gov/en/agencies/dhs/resources/early-learning-child-care/child-care-works.html", "last_updated": "2026-08-05", "confidence": "high", "note": "Eligibility-only (CC2 shape): Child Care Works ceiling at 200% of the Federal Poverty Guidelines, $66,000 for a family of four. FPL-based, not SMI. May qualify, never a promise of a spot." }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: Pennsylvania has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. Pennsylvania grants paid parental leave to state employees only. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Philadelphia-Camden-Wilmington PA-NJ-DE-MD Metro Area (31000US37980). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US37980", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 29.7. The previous 33 tracked Philadelphia CITY (33.2)." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Philadelphia County PA (FIPS 42101), NDCP study year 2020.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2020 to 2026 by CUUR0000SEEB03 (x1.27208; 2026 = Jan-Jul mean 396.748 against 2020 annual average 311.888). Tier C: Pennsylvania has no 2021 or 2022 data; the latest observed year is 2020, escalated over the longer window at x1.27208. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,500." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "tampa": { "tax.federal_brackets": { "source": "IRS 2026 inflation-adjusted federal brackets (inherited USA baseline)", "last_updated": "2026-07-29", "confidence": "high" }, "tax.state_brackets": { "source": "Florida levies no individual income tax (FL Const. Art. VII s.5) and no local income tax", "last_updated": "2026-07-29", "confidence": "high" }, "tax.sales_tax_rate": { "source": "FL state 6% + Hillsborough County 1.5% surtax = 7.5% (FL DOR DR-15DSS 2026)", "last_updated": "2026-07-29", "confidence": "high", "note": "Hillsborough CIT half-cent extended to 2041 by Ordinance 24-3, so no 2026 sunset. Pinellas is 7.0%, so a metro blend would be ~7.3%; Hillsborough used for a city-named metro." }, "tax.property_tax_rate_pct": { "source": "Hillsborough total millage 19.7-19.84 in the City of Tampa, 17.9049 unincorporated; 2026 homestead exemption $51,411", "last_updated": "2026-07-29", "confidence": "medium", "note": "Stored at 1.6 on a RECENT-BUYER basis (assessed value resets at purchase, Save Our Homes cap not yet accrued). NOTE: Miami is stored at 1.02 on a long-tenure realized basis, so the two Florida cities are currently on different bases." }, "housing._summary": { "source": "HUD FY2026 FMR Tampa-St. Petersburg-Clearwater MSA cross-checked against Zumper, Rent.com, Apartment List and RentCafe; Zillow ZHVI metro $351,847", "last_updated": "2026-07-29", "confidence": "medium", "note": "Apartment List and RentCafe differ ~30% on 1BR/2BR; source median taken. Florida HOMEOWNERS insurance (~$2,600-5,900/yr) is a real outlier but renters insurance is near the US norm, so it is not inflated here." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "KFF 2025 Employer Health Benefits Survey worker contribution ($1,440 single / $6,850 family) + Peterson-KFF out-of-pocket; Mid-A tier, matched to Houston/Nashville/Phoenix", "last_updated": "2026-07-29", "confidence": "medium", "note": "Placed one notch under Miami: MEPS-IC shows FL employee contributions above the US average, but Miami-Dade provider prices run materially above Tampa Bay." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "childcare._summary": { "source": "Child Care Aware / Florida DCF state averages trued up to Tampa Bay center-based rates; FL School Readiness is means-tested near 150% FPL", "last_updated": "2026-07-29", "confidence": "medium", "note": "Florida VPK (free, 540 hrs, no means test) would cut a real 4-year-old bill 25-35% during the school year; not netted out, matching how Miami is stored, and the conservative call." }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit convention", "last_updated": "2026-07-29", "confidence": "high", "note": "Zeroed: means-tested out at ICP income. The DCFSA is a per-HOUSEHOLD benefit and lives in subsidy_household_annual." }, "childcare.income_tested_childcare": { "source": "https://www.elchc.org/families/tuition-assistance/", "last_updated": "2026-08-05", "confidence": "high", "note": "Entry at 55% SMI, statewide since HB 1255 ($59,467 for a family of 4, FY2027 sliding fee scale effective 1 Jul 2026), with exit at 85% SMI ($91,904); a household under the entry line may qualify, subject to coalition funding." }, "food._summary": { "source": "BLS CPI South-region food-at-home plus Tampa cost-of-living indices (groceries ~4-6% above US average)", "last_updated": "2026-07-29", "confidence": "medium" }, "transit._summary": { "source": "HART 31-Day Unlimited Local & Limited Express $65.00, also the Flamingo Fares monthly cap", "last_updated": "2026-07-29", "confidence": "high" }, "discretionary._summary": { "source": "BLS Consumer Expenditure Survey South region, scaled to the Tampa price level between Atlanta and Miami", "last_updated": "2026-07-29", "confidence": "medium" }, "travel_home._summary": { "source": "BTS domestic itinerary fares; TPA is a low-cost-carrier stronghold (Southwest, Frontier, Spirit, Breeze, Avelo)", "last_updated": "2026-07-29", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Federal CTC $2,000/child (no FL top-up)", "last_updated": "2026-07-29", "confidence": "high", "note": "US Child Tax Credit (~$2,000/kid): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at this ICP household income." }, "non_cash._summary": { "source": "FL Stat. s.218.077 preempts local paid-leave ordinances and Florida has no state PFML; FMLA 12 weeks unpaid; Tampa Bay mean commute ~28 min", "last_updated": "2026-07-29", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Tampa: USD 120,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "USD 120,000 / 1 = USD 120,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.4800, R5-rounded down to 0.48. REPLACES 0.77 (2026-07-29), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. n=198, thin. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): (none)" }, "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 57300 USD, Tampa-St. Petersburg-Clearwater, FL metro area, table B20017 count-weighted)", "last_updated": "2026-07-29", "confidence": "high" }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Hillsborough County FL (FIPS 12057), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $880/mo and the ages 1-2 rate $1,020/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,050. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: Florida has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. Florida preempts local paid-leave mandates and only enables voluntary insurance products. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Tampa-St. Petersburg-Clearwater FL Metro Area (31000US45300). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US45300", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 29.4. Survey self-report, MSA geography." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Hillsborough County FL (FIPS 12057), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,300." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "charlotte": { "tax.federal_brackets": { "source": "IRS 2026 inflation-adjusted federal brackets (inherited USA baseline)", "last_updated": "2026-07-29", "confidence": "high" }, "tax.state_brackets": { "source": "NC flat individual income tax 3.99% for taxable years after 2025 (NCDOR tax rate schedules); was 4.25% in 2025", "last_updated": "2026-07-29", "confidence": "high", "note": "Trigger-based further cuts (possible 3.49% in 2027) do not apply to tax year 2026. NC prohibits local income taxes. NOTE: NC’s own standard deduction is $12,750 single / $25,500 MFJ vs the federal $16,100 / $32,200 the engine applies, understating NC tax by roughly $134/yr single." }, "tax.sales_tax_rate": { "source": "Mecklenburg County levied an additional 1.00% effective July 1, 2026, moving combined state+local from 7.25% to 8.25% (NCDOR Important Notice)", "last_updated": "2026-07-29", "confidence": "high", "note": "2026 change; Wake County (Raleigh-Durham) remains 7.25%." }, "tax.property_tax_rate_pct": { "source": "Mecklenburg County FY2027 49.27c per $100 (held flat, adopted June 2 2026) + City of Charlotte 29.30c (raised from 27.41c, adopted June 8 2026) = 0.7857%", "last_updated": "2026-07-29", "confidence": "high", "note": "Nominal basis, matching how Raleigh-Durham is stored. ACS 2024 implies 0.71% because Mecklenburg assessments are frozen at Jan 1 2023 values." }, "housing._summary": { "source": "Median across Zumper July 2026 and RentCafe 2026 Charlotte; Zillow ZHVI $397,125 (Mar 2026), Canopy MLS $420,000 city median", "last_updated": "2026-07-29", "confidence": "high", "note": "Apartment List (Aug 2025 vintage) runs 22-29% above and is internally inconsistent with its own median; excluded as an outlier." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "KFF 2025 Employer Health Benefits Survey worker contribution + Peterson-KFF out-of-pocket; Low tier, matched to Raleigh-Durham", "last_updated": "2026-07-29", "confidence": "medium", "note": "Charlotte’s employer base is dominated by very large self-insured employers (Bank of America, Wells Fargo, Truist, Ally, Honeywell, Lowe’s, Duke Energy) with a high employer premium share; NC sits below the national average on employee contribution share." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "childcare._summary": { "source": "Child Care Aware state-by-state; Charlotte center-based infant ~$1,250/mo and roughly 44% above the NC state average; NC Pre-K and NC subsidy are means-tested near 200% FPL", "last_updated": "2026-07-29", "confidence": "medium" }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit convention", "last_updated": "2026-07-29", "confidence": "high", "note": "Zeroed: means-tested out at ICP income. The DCFSA is a per-HOUSEHOLD benefit and lives in subsidy_household_annual." }, "childcare.income_tested_childcare": { "source": "https://ncchildcare.ncdhhs.gov/parent/paying-for-child-care", "last_updated": "2026-08-05", "confidence": "high", "note": "Entry at 200% FPL for children 0-5 ($64,296 for a family of 4, statewide chart effective 1 Jul 2025; ages 6-12 enter at 133% FPL) with recertification capped at 85% SMI, an exit ceiling of $92,976; a household under the entry line may qualify." }, "food._summary": { "source": "BEA regional price parities; Charlotte goods RPP near 97-100 with groceries ~2% above the national average", "last_updated": "2026-07-29", "confidence": "medium" }, "transit._summary": { "source": "CATS Local Monthly Pass $88.00, unlimited local/neighborhood bus plus LYNX Blue Line", "last_updated": "2026-07-29", "confidence": "high", "note": "Well above Raleigh-Durham’s $32 because Charlotte runs light rail." }, "discretionary._summary": { "source": "BLS Consumer Expenditure Survey South region, scaled to Charlotte’s cost-of-living index", "last_updated": "2026-07-29", "confidence": "medium" }, "travel_home._summary": { "source": "BTS average fare data; CLT is an American Airlines fortress hub (~90% AA-operated) carrying a documented hub fare premium above RDU", "last_updated": "2026-07-29", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Federal CTC $2,000/child (no NC top-up)", "last_updated": "2026-07-29", "confidence": "high", "note": "US Child Tax Credit (~$2,000/kid): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at this ICP household income." }, "non_cash._summary": { "source": "NC has no state paid family leave or paid sick leave mandate; FMLA 12 weeks unpaid; ACS mean commute Mecklenburg County 25.3 min", "last_updated": "2026-07-29", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Charlotte: USD 130,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "USD 130,000 / 1 = USD 130,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.5200, R5-rounded down to 0.52. REPLACES 0.82 (2026-07-29), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. n=300. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): The two measures disagree structurally (broad but top-truncated distribution: BLS p90 $172,360 Charlotte vs $209,410 Durham-Chapel Hill). 0.82 is the midpoint of a defensible 0.78-0.85 range." }, "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 63700 USD, Charlotte-Concord-Gastonia, NC-SC metro area, table B20017 count-weighted)", "last_updated": "2026-07-29", "confidence": "high" }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Mecklenburg County NC (FIPS 37119), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $890/mo and the ages 1-2 rate $1,360/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,000. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: North Carolina has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. North Carolina grants paid parental leave to state employees only. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Charlotte-Concord-Gastonia NC-SC Metro Area (31000US16740). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US16740", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 27.5. Survey self-report, MSA geography." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Mecklenburg County NC (FIPS 37119), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,300." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "san-antonio": { "tax.federal_brackets": { "source": "IRS 2026 inflation-adjusted federal brackets (inherited USA baseline)", "last_updated": "2026-07-29", "confidence": "high" }, "tax.state_brackets": { "source": "Texas levies no individual income tax (TX Const. Art. VIII s.24-a) and prohibits local income taxes", "last_updated": "2026-07-29", "confidence": "high" }, "tax.sales_tax_rate": { "source": "TX state 6.25% + City of San Antonio 1.25% + VIA Metropolitan Transit 0.50% + Advanced Transportation District 0.25% = 8.25% statutory cap (Texas Comptroller)", "last_updated": "2026-07-29", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Bexar county-wide entities total $1.2853 per $100 (City 0.541590, County ~0.30, University Health 0.276235, Alamo Colleges, River Authority) plus a school district of $0.9572 to $1.2575; combined nominal $2.24-$2.54", "last_updated": "2026-07-29", "confidence": "medium", "note": "Stored at 2.3 on the same nominal/recent-buyer basis as Houston 2.0, Austin 1.9 and Dallas 2.1, making San Antonio the highest Texas metro. Ownwell reports 1.55% median effective, which is post-homestead-cap and NOT the peer basis." }, "housing._summary": { "source": "RentCafe San Antonio (Bexar County) July 2026 apartment asking rents, set toward the upper end so 3BR reflects single-family rentals; SABOR metro median sale $306,000 (May 2026)", "last_updated": "2026-07-29", "confidence": "medium", "note": "Overall average rent spans RentCafe $1,273 to Zillow ZORI $1,625 (~28%) because ZORI includes single-family homes." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "KFF 2025 Employer Health Benefits Survey worker contribution + Peterson-KFF out-of-pocket; Mid-A tier, matched to Houston/Nashville/Phoenix", "last_updated": "2026-07-29", "confidence": "medium", "note": "Mid-A over Low (Austin): San Antonio employer-plan cost structure tracks Houston, and Mid-A is the less flattering of the two plausible tiers." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "childcare._summary": { "source": "childcarecostfinder and Care.com San Antonio center-based rates; Texas CCS vouchers are income-tested and do not reach this ICP", "last_updated": "2026-07-29", "confidence": "medium", "note": "Licensed center care, matching the peer basis. In-home family daycare (~$833/mo) and nannies (~$2,167/mo) excluded." }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit convention", "last_updated": "2026-07-29", "confidence": "high", "note": "Zeroed: means-tested out at ICP income. The DCFSA is a per-HOUSEHOLD benefit and lives in subsidy_household_annual." }, "childcare.income_tested_childcare": { "source": "https://www.workforcesolutionsalamo.org/childcare-parents/", "last_updated": "2026-08-05", "confidence": "high", "note": "Entry ceiling is the statewide Texas 85% SMI line (TWC BCY 2026 income limits, $92,041 for a family of 4, effective 1 Oct 2025 per WD Letter 11-25); Texas publishes no separate exit ceiling; a household under the line may qualify but boards run waitlists, and the board's own page still prints a stale BCY-labeled $7,311/mo." }, "food._summary": { "source": "Numbeo and livingcost San Antonio grocery basket; the cheapest of the four Texas metros in the dataset", "last_updated": "2026-07-29", "confidence": "medium" }, "transit._summary": { "source": "VIA Metropolitan Transit 31-day pass $38.00 (single ride $1.30, annual $365)", "last_updated": "2026-07-29", "confidence": "high" }, "discretionary._summary": { "source": "Scaled from the Texas peers on San Antonio’s lower restaurant and recreation index; Hill Country and Gulf Coast trips are short-drive", "last_updated": "2026-07-29", "confidence": "medium" }, "travel_home._summary": { "source": "BTS average fare data; SAT is a mid-size origin with fewer nonstops than IAH, DFW or AUS", "last_updated": "2026-07-29", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Federal CTC $2,000/child (no TX state child allowance or credit)", "last_updated": "2026-07-29", "confidence": "high", "note": "US Child Tax Credit (~$2,000/kid): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at this ICP household income." }, "non_cash._summary": { "source": "No Texas PFML or paid sick leave; the 2019 San Antonio PSL ordinance was enjoined and then preempted by HB 2127 (2023); ACS 2024 metro mean commute", "last_updated": "2026-07-29", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, San Antonio: USD 126,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "USD 126,000 / 1 = USD 126,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.5040, R5-rounded down to 0.5. REPLACES 0.73 (2026-07-29), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. n=72, thin. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Set at 0.73, below Phoenix 0.75, weighting employer composition (USAA, Rackspace, H-E-B corporate IT, DoD cyber cluster, no big-tech bands) over the near-identical OEWS median. Conservative reading." }, "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 55800 USD, San Antonio-New Braunfels, TX metro area, table B20017 count-weighted)", "last_updated": "2026-07-29", "confidence": "high" }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Bexar County TX (FIPS 48029), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Texas caveat: from 2015 onward the state reports the prior year figures, so the 2022 row is 2021 fieldwork and this value is marginally understated after escalation. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $810/mo and the ages 1-2 rate $850/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $975. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: Texas has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. Texas law only enables voluntary paid-leave insurance products; no entitlement. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, San Antonio-New Braunfels TX Metro Area (31000US41700). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US41700", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 27.6. Survey self-report, MSA geography." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Bexar County TX (FIPS 48029), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Texas caveat: from 2015 onward the state reports the prior year figures, so the 2022 row is 2021 fieldwork and this value is marginally understated after escalation. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,200." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "sacramento": { "tax.federal_brackets": { "source": "IRS 2026 inflation-adjusted federal brackets (inherited USA baseline)", "last_updated": "2026-07-29", "confidence": "high" }, "tax.state_brackets": { "source": "CA FTB, 2025 California Tax Rate Schedules (Form 540 booklet) https://www.ftb.ca.gov/forms/2025/2025-540-tax-rate-schedules.pdf", "last_updated": "2026-08-01", "confidence": "high", "note": "Schedule X, single or married/RDP filing separately, tax year 2025: 1% to 11,079 then 2/4/6/8% to 72,724, 9.3% to 371,479, 10.3% to 445,771, 11.3% to 742,953, 12.3% above. Refreshed 2026-08-01: the file carried the TY2024 thresholds (10,756 / 25,499 / ... / 721,314) while already citing the TY2025 schedule, so the data disagreed with its own citation. TY2025 is FTB's newest published year; the TY2026 schedules are not out. Corrected 2026-08-06: this note used to say the 1% Behavioral Health Services Tax above $1M taxable (renamed from the Mental Health Services Tax for TY2025 by Prop 1) was absent from the engine. It is not absent. app.js computes it as taxes.state_surtax from tax.state_surtax_rate 0.01 and tax.state_surtax_threshold 1000000, both of which this city carries, and a single filer on $1,500,000 gross pays $4,839 of it (identical in all four California cities). The threshold is per return and does not double for joint filers, which is the statute rather than a gap: RTC 17043(c) disapplies section 17045, the income-splitting rule that doubles California's regular MFJ bands, so no schema field for tax.state_surtax_threshold_married exists in any city JSON. https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=RTC§ionNum=17043" }, "tax.state_disability_insurance_rate": { "source": "California EDD, SDI Contribution Rates and Benefit Amounts https://edd.ca.gov/en/disability/Contribution_Rates_and_Benefit_Amounts/", "last_updated": "2026-07-30", "confidence": "high", "note": "Employee State Disability Insurance / Paid Family Leave contribution, 1.3% for 2026 (up from 1.2% in 2025). EDD lists the taxable wage ceiling and the maximum contribution as N/A: SB 951 eliminated both effective 2026-01-01, so this is levied on every dollar of wages with no cap. Added 2026-07-30; before that the engine omitted it entirely, understating California payroll by 1.3% of gross. Statewide levy, identical across the CA metros." }, "tax.sales_tax_rate": { "source": "6.00% state + 0.25% Sacramento County + 1.00% City of Sacramento + 1.50% special district = 8.75% (City of Sacramento / CDTFA Q1 2026)", "last_updated": "2026-07-29", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Prop 13 recent-buyer basis (assessed value resets at purchase): 1% base plus voter-approved school and city bonds and direct district levies puts City of Sacramento parcels at 1.10-1.25%", "last_updated": "2026-07-29", "confidence": "medium", "note": "Stored at 1.2, the upper half of the city range and the less flattering choice; ranks just above LA 1.18 and SF 1.183 on Sacramento’s heavier school-bond and Mello-Roos load." }, "housing._summary": { "source": "Median of Zumper, RentCafe and Rent.com Sacramento July 2026; Zillow metro typical value $560,000-$575,000 used for the home-price ladder to match the peer metro basis", "last_updated": "2026-07-29", "confidence": "medium", "note": "Studio quotes span $1,399-$1,889 (~35%); Rent.com’s studio exceeds its own 1BR and was treated as an outlier. City ZHVI $482,968 vs metro $560k+ is a ~16% spread; metro basis chosen for peer consistency." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "KFF 2025 Employer Health Benefits Survey worker contribution + Peterson-KFF out-of-pocket; High tier, matched to the other California metros", "last_updated": "2026-07-29", "confidence": "medium", "note": "High rather than Mid-B on its own merits: Covered California Region 3 (Sacramento, Placer, El Dorado, Yolo) is persistently one of the most expensive rating regions in the state due to concentrated hospital-system market power." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "childcare._summary": { "source": "childcarecostfinder Sacramento and tooTRiS California city data; center-based infant $1,400-$1,667/mo against a CA center average of ~$1,725", "last_updated": "2026-07-29", "confidence": "medium", "note": "Licensed center care to match the peer basis; in-home family daycare (~$1,292/mo) excluded. CalWORKs/ELCC vouchers and TK are income-tested or age-4-only and deliver no cash to this ICP." }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit convention", "last_updated": "2026-07-29", "confidence": "high", "note": "Zeroed: means-tested out at ICP income. The DCFSA is a per-HOUSEHOLD benefit and lives in subsidy_household_annual." }, "food._summary": { "source": "RentCafe Sacramento cost-of-living calculator; groceries ~5% above the national average, with Central Valley produce access partly offsetting California retail and labor costs", "last_updated": "2026-07-29", "confidence": "medium" }, "transit._summary": { "source": "SacRT adult monthly pass $100.00 (single ride $2.50, daily $7.00); the ZipPass app was retired July 1, 2026", "last_updated": "2026-07-29", "confidence": "high" }, "discretionary._summary": { "source": "Interpolated between the Sun Belt and coastal California peers; Tahoe, Napa and the Pacific coast are weekend-drive destinations", "last_updated": "2026-07-29", "confidence": "medium" }, "travel_home._summary": { "source": "BTS average fare data; SMF lacks the transcontinental nonstop density of SFO or OAK", "last_updated": "2026-07-29", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Federal CTC $2,000/child; California’s Young Child Tax Credit is income-tested and phases out far below this ICP", "last_updated": "2026-07-29", "confidence": "high", "note": "US Child Tax Credit (~$2,000/kid): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at this ICP household income." }, "non_cash._summary": { "source": "CA SB 616 (40 hrs paid sick leave); CA PFL 8 weeks bonding with CFRA/FMLA extending job-protected time to 12 weeks; 70% wage replacement at ICP earnings under AB 908/SB 951", "last_updated": "2026-07-29", "confidence": "high", "note": "Aligned with san-diego 70. los-angeles 80 and san-francisco 100 reflect LOCAL top-up ordinances; Sacramento municipal code has none. STALE CLAUSE REMOVED 2026-08-02: this note used to say the CA SDI/PFL contribution was \"NOT MODELED\" and that \"there is no schema field for it\", so all four CA metros understated California payroll tax by ~1.3% of gross. There is a field and it is populated. tax.state_disability_insurance_rate carries 0.013 in all four CA metros, and computeUsTaxes applies it UNCAPPED to the FICA wage base and folds it into payroll. California payroll is no longer understated. See the tax.state_disability_insurance_rate entry for sourcing." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Sacramento: USD 152,000, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "USD 152,000 / 1 = USD 152,000; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.6080, R5-rounded down to 0.6. REPLACES 0.89 (2026-07-29), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. n=107, thin. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Set at 0.89, above Salt Lake City 0.88 and below San Diego 0.92, taking the lower end of the plausible 0.89-0.91 window. Employer base is state-government IT, Intel Folsom, health-system IT and partially-adjusted Bay Area remote bands." }, "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 72200 USD, Sacramento-Roseville-Folsom, CA metro area, table B20017 count-weighted)", "last_updated": "2026-07-29", "confidence": "high" }, "tax.state_brackets_married": { "source": "CA FTB, 2025 California Tax Rate Schedules (Form 540 booklet) https://www.ftb.ca.gov/forms/2025/2025-540-tax-rate-schedules.pdf", "last_updated": "2026-08-01", "confidence": "high", "note": "Schedule Y, married/RDP filing jointly or qualifying surviving spouse, tax year 2025. Verified band by band as EXACTLY 2x Schedule X on all 8 finite thresholds, which is what RTC 17041(b) prescribes; the base-tax columns double too. Added 2026-08-01. Before this, couples were run through the single-filer ladder, overstating California tax by a flat $3,457.64 for any household above about $141k of taxable income, and by $4,029 at a $450k household. Head of household is NOT 2x; no schema field for tax.state_brackets_hoh exists." }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Sacramento County CA (FIPS 6067), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $1,360/mo and the ages 1-2 rate $1,570/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,300. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "childcare.income_tested_childcare": { "source": "https://www.cdss.ca.gov/Portals/9/Additional-Resources/Letters-and-Notices/CCBs/2026/CCB_26-17.pdf", "last_updated": "2026-08-05", "confidence": "high", "note": "Eligibility-only (CC2 shape): CDSS ceiling at 85% SMI, $115,637 for a family of four (CCB 26-17, FY 2026-27; same threshold initial and ongoing). Tested on ADJUSTED monthly income per 5 CCR 18078(a), so a gross comparison is approximate; CCDF voucher minus copay, never a percentage. May qualify, never a promise of a spot. Revise each July." }, "non_cash.parental_leave_weeks": { "source": "EDD Paid Family Leave https://edd.ca.gov/en/disability/paid-family-leave/ ; https://edd.ca.gov/en/disability/Calculating_PFL_Benefit_Payment_Amounts/", "last_updated": "2026-08-06", "confidence": "high", "note": "16 = 8 weeks of CA PFL bonding per parent within a 12-month period, each parent's own SDI-funded non-transferable entitlement, summed (both-parents flag set). The birthing parent's SDI pregnancy disability (usually 4 weeks before plus 6 after, 8 for cesarean) is medically certified, EXCLUDED, conservative. Was 12/70: the unpaid FMLA window with a fabricated headline. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "EDD: SB 951 rates for claims from 2025 (90% low-earner band, 70% above USD 20,931.31 highest-quarter earnings); https://edd.ca.gov/en/disability/Calculating_PFL_Benefit_Payment_Amounts/", "last_updated": "2026-08-06", "confidence": "high", "note": "0 percentage-paid weeks: for claims from 2025 under SB 951, CA pays 90% for highest-quarter earnings of 722.50-16,279.90, a flat 1,127/wk transition band to 20,931.30, and 70% above 20,931.31 per quarter, capped at USD 1,765/wk in 2026 (1,681 in 2025), binding from about USD 131,000. CAVEAT recorded: below that threshold CA pays a true 70% for the ICP band, so the flat store overstates for earners under about 131k (the engine clamps at 100%); at the salaries we price it is exact and less favorable than any headline. At USD 150,000 the effective rate is about 61%." }, "non_cash.parental_leave_flat": { "source": "EDD Contribution Rates and Benefit Amounts: 2026 maximum weekly benefit USD 1,765 https://edd.ca.gov/en/disability/Contribution_Rates_and_Benefit_Amounts/", "last_updated": "2026-08-06", "confidence": "high", "note": "USD 1,765/wk, 2026 maximum. 16 x 1,765 = USD 28,240 total." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Sacramento-Roseville-Folsom CA Metro Area (31000US40900). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US40900", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 27.7. Survey self-report, MSA geography." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Sacramento County CA (FIPS 6067), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "medium", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier B: the state’s most recent MRS had no usable county estimates, so NDCP carried the 2018 base forward to 2022 by propensity-score matching (technical report, \"Imputations for States Using Alternative Price Methodology\"). Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,650." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "las-vegas": { "tax.federal_brackets": { "source": "IRS 2026 inflation-adjusted federal brackets (inherited USA baseline)", "last_updated": "2026-07-29", "confidence": "high" }, "tax.state_brackets": { "source": "Nevada levies no individual income tax and no local wage tax (Nevada Constitution Art. 10 s.1(9))", "last_updated": "2026-07-29", "confidence": "high" }, "tax.sales_tax_rate": { "source": "Clark County combined 8.375% = 4.6% state + 3.775% county, unchanged for 2026", "last_updated": "2026-07-29", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Nevada assesses at 35% of taxable value; City of Las Vegas consolidated rates ~$3.20-3.50 per $100 assessed, but the 3% owner-occupied abatement cap follows the PROPERTY, so effective rates on market value are far lower", "last_updated": "2026-07-29", "confidence": "medium", "note": "Clark County median effective is reported at 0.50% and the recent-buyer range at 0.50-0.65%. Stored at 0.65, the top of the recent-buyer range, per conservative bias; this is >20% above the county median figure." }, "housing._summary": { "source": "RentCafe Las Vegas 2026 apartment asking rents with the 3BR taken from Zillow since 3BR stock is mostly single-family rentals; Las Vegas Realtors June 2026 median existing single-family $490,000, condo/townhome $292,000", "last_updated": "2026-07-29", "confidence": "medium" }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "KFF 2025 Employer Health Benefits Survey worker contribution + Peterson-KFF out-of-pocket; Mid-A tier, matched to Houston/Nashville/Phoenix", "last_updated": "2026-07-29", "confidence": "medium", "note": "Nevada sits near the national middle on health costs; matched to the Sun Belt peer set." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "childcare._summary": { "source": "daycarecalc Las Vegas center-based (infant $1,080/mo) against a competing $1,350-1,650 range; median of the range rounded up. Nevada subsidy caps near $7,132/mo household income for a family of 4", "last_updated": "2026-07-29", "confidence": "medium", "note": "Sources conflict by more than 20% on infant care." }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit convention", "last_updated": "2026-07-29", "confidence": "high", "note": "Zeroed: means-tested out at ICP income. The DCFSA is a per-HOUSEHOLD benefit and lives in subsidy_household_annual." }, "childcare.income_tested_childcare": { "source": "https://www.dss.nv.gov/programs/child-care/childcare-parents/", "last_updated": "2026-08-05", "confidence": "high", "note": "Intake ceiling at 41% SMI ($41,280 for a family of 4, NV DSS transmittal CC PT 06-25, effective 1 Oct 2025) with renewal at 49% SMI, an exit ceiling of $49,332; the ARPA-era 85% SMI ceiling was rolled back on 1 Oct 2024; a household under the intake line may qualify." }, "food._summary": { "source": "RentCafe Las Vegas cost-of-living calculator; grocery index ~3% above the national average, dining set above Phoenix for the tourist-priced restaurant market", "last_updated": "2026-07-29", "confidence": "medium" }, "transit._summary": { "source": "RTC Southern Nevada 30-day residential pass $65, unchanged for 15+ years", "last_updated": "2026-07-29", "confidence": "high", "note": "The RTC board has PROPOSED $80 effective Jan 31 2027; still in comment stage as of June 2026, so the adopted 2026 price is stored. Flag for the next data refresh." }, "discretionary._summary": { "source": "RentCafe cost-of-living calculator; entertainment above Phoenix on the local-resident price environment, local vacation matched to the Sun Belt drive-market peers", "last_updated": "2026-07-29", "confidence": "medium" }, "travel_home._summary": { "source": "BTS average fare data; LAS is one of the cheapest large US origins for domestic round trips on Southwest/Spirit/Frontier leisure capacity", "last_updated": "2026-07-29", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Federal CTC $2,000/child (no NV top-up)", "last_updated": "2026-07-29", "confidence": "high", "note": "US Child Tax Credit (~$2,000/kid): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at this ICP household income." }, "non_cash._summary": { "source": "NV SB 312 / NRS 608.0197: 40 hrs/yr paid leave for any reason at employers with 50+ staff, no reason disclosure; Nevada has no state PFML, so parental leave is federal FMLA only; ACS metro mean commute 25 min", "last_updated": "2026-07-29", "confidence": "high" }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Las Vegas: USD 111,111, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "USD 111,111 / 1 = USD 111,111; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.4444, R5-rounded down to 0.44. REPLACES 0.72 (2026-07-29), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. n=61, the thinnest sample in the US set. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): CONFLICT: levels.fyi shows Las Vegas senior SWE median total comp $142,800 ABOVE Phoenix $135,625, which would imply a higher index. That sample is thin and skewed by remote entries, so the structural read was used instead. Worth revisiting." }, "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 54200 USD, Las Vegas-Henderson-North Las Vegas, NV metro area, table B20017 count-weighted)", "last_updated": "2026-07-29", "confidence": "high" }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Clark County NV (FIPS 32003), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $1,180/mo and the ages 1-2 rate $1,300/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,000. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: Nevada has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. NV SB 312 is 40 hrs/yr any-purpose leave at 50+ employers, not parental leave. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Las Vegas-Henderson-North Las Vegas NV Metro Area (31000US29820). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US29820", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 25.9. Survey self-report, MSA geography." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Clark County NV (FIPS 32003), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,250." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "detroit": { "tax.federal_brackets": { "source": "IRS 2026 inflation-adjusted federal brackets (inherited USA baseline)", "last_updated": "2026-07-29", "confidence": "high" }, "tax.state_brackets": { "source": "Michigan flat individual income tax 4.25% for tax year 2026, confirmed by the Michigan Dept of Treasury on April 15 2026 (the Section 51 rate-reduction trigger did not fire)", "last_updated": "2026-07-29", "confidence": "high", "note": "The 4.05% figure on several tax sites is the one-year 2023 rate and is stale." }, "tax.city_income_tax_rate": { "source": "Detroit resident city income tax 2.4% (nonresidents 1.2%), set by state statute as a Detroit-specific exception to the 1% cap other Michigan cities face (City of Detroit OCFO)", "last_updated": "2026-07-29", "confidence": "high", "note": "Deliberately city-of-Detroit basis; suburban Oakland and Macomb residents pay 0%." }, "tax.sales_tax_rate": { "source": "Michigan 6% statewide with no local add-on", "last_updated": "2026-07-29", "confidence": "high" }, "tax.property_tax_rate_pct": { "source": "Lincoln Institute puts the city of Detroit at 3.02% effective on market value, the highest of any major US city; county effective rates are Wayne 2.35%, Oakland 1.68%, Macomb 1.64%", "last_updated": "2026-07-29", "confidence": "medium", "note": "Stored at 2.5 as a recent-buyer METRO rate, because home prices in this file are metro basis and pairing the 3.02% city rate with metro prices would double-count. Michigan Proposal A uncaps taxable value to 50% of market on sale, so a new buyer pays full millage. Above the ~1.97% three-county population-weighted median, per conservative bias." }, "housing._summary": { "source": "METRO basis (Wayne/Oakland/Macomb). RentCafe Detroit 2026 nudged up toward suburban comparables (Royal Oak, Troy); Zillow Detroit-Warren-Dearborn ZHVI $243,918", "last_updated": "2026-07-29", "confidence": "medium", "note": "Zumper city-only rents run >20% lower (1BR $950) because they reflect non-downtown city stock, not the metro. City of Detroit alone has a ~$100K median sale price, which is why the metro basis matters here." }, "housing.rent_percentiles": { "source": "Calibrated default (docs/superpowers/research/2026-07-17-housing-range-research.md §4): London ONS within-bedroom quartiles avg 0.84/1.23, low edge nudged to 0.82 for asking-rent dispersion; HUD FMR40-to-FMR50 +9% step cross-check. Within-listing (within-bedroom) basis.", "last_updated": "2026-08-07", "confidence": "medium", "note": "Band edge = median_rent_by_bedroom x multiplier, computed at display time. Pair is data-driven so a swap to the cross-neighborhood 0.78/1.30 pair is a data edit (research §4). Derivation arithmetic: docs/superpowers/research/2026-08-07-housing-data-apply.md" }, "healthcare._summary": { "source": "KFF 2025 Employer Health Benefits Survey worker contribution + Peterson-KFF out-of-pocket; Mid-B tier, matched to Chicago/Minneapolis/Pittsburgh", "last_updated": "2026-07-29", "confidence": "medium", "note": "Michigan’s 2026 ACA benchmark silver is $523/mo vs $625 national, which would argue for Low, but these fields model the employee share of EMPLOYER coverage where Michigan tracks the national average. Peer consistency with Chicago/Minneapolis/Pittsburgh plus conservative bias put Detroit in Mid-B." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Unchanged by the 2026-08-08 basis unification. The US rows were already realized average spend; this sweep moved every OTHER country onto the US basis. Basis: the shared declaration in shared_fields, INCLUDING its US exception. These rows are covered-medical cost-sharing PLUS a dental and vision out-of-pocket margin (~USD 400 at row 1, ~USD 400-450 at row 4) plus out-of-network and non-covered care, per docs/superpowers/plans/2026-07-17-us-healthcare-recalibration.md lines 129 and 131. The separate dental_vision_monthly is the PREMIUM single share, so the two columns are complementary and app.js bills the US dental line flat. Corrected 2026-08-08 (cluster dental-scaling-0808): this note used to claim the column excluded routine dental and vision, which contradicted both the same city's dental note and the data itself." }, "childcare._summary": { "source": "childcarepath Michigan metro data: Detroit metro is the state’s most expensive (infant $1,000-1,600/mo, preschool $800-1,300) against lower statewide averages; metro range midpoint rounded up", "last_updated": "2026-07-29", "confidence": "medium" }, "childcare.subsidy_per_kid_annual": { "source": "benefits-cliff audit convention", "last_updated": "2026-07-29", "confidence": "high", "note": "Zeroed: means-tested out at ICP income. The DCFSA is a per-HOUSEHOLD benefit and lives in subsidy_household_annual." }, "childcare.income_tested_childcare": { "source": "https://www.michigan.gov/mileap/early-childhood-education/early-learners-and-care/cdc/parents", "last_updated": "2026-08-05", "confidence": "high", "note": "Entry at 200% FPL ($64,296 for a family of 4, MDHHS RFT 270 scale effective 21 Sep 2025) with a higher exit column of $99,492 consistent with 85% SMI; a household under the entry line may qualify." }, "food._summary": { "source": "BLS CPI Detroit metro and Redfin cost-of-living; metro groceries within +/-2% of the national average, set in line with the Midwest peers", "last_updated": "2026-07-29", "confidence": "medium" }, "transit._summary": { "source": "DDOT 31-Day Dart Pass $70 full fare, valid on both DDOT and SMART", "last_updated": "2026-07-29", "confidence": "high" }, "discretionary._summary": { "source": "Set in line with the Rust Belt peers; local vacation reflects the Northern Michigan and Great Lakes drive-vacation pattern", "last_updated": "2026-07-29", "confidence": "medium" }, "travel_home._summary": { "source": "BTS average fare data; DTW is a Delta fortress hub with limited low-cost-carrier competition", "last_updated": "2026-07-29", "confidence": "medium" }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Federal CTC $2,000/child (no MI top-up)", "last_updated": "2026-07-29", "confidence": "high", "note": "US Child Tax Credit (~$2,000/kid): means-tested phase-out begins at $400k MFJ AGI, so it survives in full at this ICP household income." }, "non_cash._summary": { "source": "Michigan Earned Sick Time Act (effective Feb 21 2025, MCL 408.963): 72 hrs/yr paid at 11+ employers, 40 hrs at 10 or fewer; Michigan has no state PFML; ACS Detroit-Warren-Dearborn metro mean commute ~26-27 min", "last_updated": "2026-07-29", "confidence": "high", "note": "A widely-cited 50-minute commute figure comes from a Robert Half self-report survey, not ACS, and was rejected." }, "compensation_index": { "source": "levels.fyi all-levels Software Engineer median TOTAL COMP, Detroit: USD 123,333, fetched 2026-08-11. Percentile ladder and sample-health check recorded in docs/superpowers/research/2026-08-11-alllevels-fetch-usa.md.", "last_updated": "2026-08-11", "confidence": "low", "note": "USD 123,333 / 1 = USD 123,333; / USD 250,000 (the declared anchor, data/_compensation_anchor.json) = 0.4933, R5-rounded down to 0.49. REPLACES 0.78 (2026-07-29), which divided a SENIOR figure by a USD 196,000 to 220,000 senior reference while the anchor is ALL-LEVELS over USD 250,000. n=416. Re-derived 2026-08-11. Working: docs/superpowers/research/2026-08-11-index-rederivation.md. Prior note (basis superseded): Set at 0.78, between Phoenix 0.75 and Pittsburgh/Nashville 0.82. Genuine engineering base (Ford, GM, Stellantis software, Rocket Companies, Ally, defense contractors) but pay lags coastal and Chicago markets." }, "income_distribution.all_earner_distribution": { "source": "US Census Bureau ACS 2024 1-Year (combined full-time year-round median earnings 65100 USD, Detroit-Warren-Dearborn, MI metro area, table B20017 count-weighted)", "last_updated": "2026-07-29", "confidence": "high" }, "non_cash.vacation_days_statutory": { "source": "https://www.dol.gov/general/topic/workhours/vacation_leave", "last_updated": "2026-08-05", "confidence": "high", "note": "STATUTORY FLOOR IS 0: the FLSA does not require payment for time not worked. Field basis is the legal floor per _national_fields.json; the previous 15 was a BLS market average, which broke the cross-country basis. Typical US tech offers 15-20 by contract; the UI label says employers exceed the floor." }, "childcare.preschool_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Wayne County MI (FIPS 26163), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Michigan caveat: the state MRS does not split by program type and COMBINES centre and family child care home prices, so Detroit is not a pure centre rate and is biased LOW against the other cities. Value is the equal-weight blend of the six-month age bands over ENGINE AGES 1-5 (12-71 months), which is what this field bills for US cities (app.js free_tier_start_age defaults to 1). The narrower ages 3-5 rate is $860/mo and the ages 1-2 rate $910/mo; that split was measured and deliberately NOT stored, see docs/superpowers/research/2026-08-13-us-childcare-ndcp.md. Previous value $1,050. Does NOT model free universal pre-K where it exists. PRIOR NOTE: DCFSA moved to the tax model 2026-08-05: the flat $2,000 household subsidy is no longer netted here; computeUsTaxes reduces taxable income by up to $5,000 when a child is in paid care and both spouses earn." }, "non_cash.parental_leave_weeks": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: Michigan has no state paid family or parental leave program for private-sector employees with benefits payable in 2026. Federal FMLA (29 U.S.C. 2601) provides 12 weeks of UNPAID job-protected leave, which the documented basis excludes: paid weeks only. Was 12: the FMLA window. MI ESTA is sick leave, not parental leave. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "BPC state paid-family-leave tracker (upd. 2026-04-23) https://bipartisanpolicy.org/explainer/state-paid-family-leave-laws-across-the-u-s/ ; A Better Balance https://www.abetterbalance.org/family-leave-laws/", "last_updated": "2026-08-06", "confidence": "high", "note": "0: no statutory paid parental leave exists, so there is no rate to store; FMLA is unpaid job protection and is never counted as paid." }, "housing.homeowners_insurance_pct_of_value": { "source": "Higher of Insurance.com 2026 state-average and ValuePenguin June 2026 city-average annual premium over our stored 3BR median price; per-metro row with URLs in docs/superpowers/research/2026-08-07-hoi-north-america.md", "last_updated": "2026-08-07", "confidence": "medium" }, "healthcare.annual_out_of_pocket_avg": { "source": "KFF 2025 EHBS worker premium share + Peterson-KFF Health System Tracker 2023 realized OOP ($869/person, $3,564 family of four) trended ~1.18x to 2026 + Milliman MMI 2025 upper anchor", "last_updated": "2026-08-08", "confidence": "medium", "note": "Single adult. Equals annual_out_of_pocket_by_family_size row 1; larger households scale via that table. Kept as a scalar for back-compat only." }, "non_cash.commute_minutes_avg": { "source": "US Census Bureau, ACS 2024 1-year, B08013/B08012, Detroit-Warren-Dearborn MI Metro Area (31000US19820). https://api.censusreporter.org/1.0/data/show/acs2024_1yr?table_ids=B08013,B08012&geo_ids=31000US19820", "last_updated": "2026-08-08", "confidence": "high", "statistic": "Mean travel time to work (one way)", "year": "2024", "note": "ACS mean 26.2. Survey self-report, MSA geography." }, "non_cash.sick_leave": { "source": "Full sourcing in docs/superpowers/research/2026-08-08-sick-leave-shard-D.md", "source_url": null, "last_updated": "2026-08-08", "confidence": "high", "note": "Statutory floor, first year of service, short-term sickness only. Basis in data/_national_fields.json _definitions." }, "tax.retirement_deduction_cap": { "source": "IRS, \"401(k) limit increases to $24,500 for 2026\" https://www.irs.gov/newsroom/401k-limit-increases-to-24500-for-2026-ira-limit-increases-to-7500 ; Notice 2025-67 https://www.irs.gov/pub/irs-drop/n-25-67.pdf", "last_updated": "2026-08-11", "confidence": "high", "note": "IRC 402(g) elective-deferral limit, 2026 calendar year. Employee-only; the employer match sits under the separate 415(c) ceiling and does not touch this. The 50+ catch-up (USD 8,000) and the 60-63 catch-up (USD 11,250) are NOT modeled: the schema has no age-banded absolute max, and the flat figure is the conservative reading. No-op at every published scenario (the tightest is 225,000 x 10% = 22,500); it binds above 245,000 gross at 10% and 408,333 at 6%." }, "childcare.infant_monthly": { "source": "US DOL Women’s Bureau, National Database of Childcare Prices (NDCP2022) https://catalog.data.gov/dataset/national-database-of-childcare-prices ; technical report https://www.dol.gov/sites/dolgov/files/WB/National-Database-of-Childcare-Prices-Technical-Report.pdf ; escalated with BLS CPI-U \"Day care and preschool\" series CUUR0000SEEB03 https://data.bls.gov/timeseries/CUUR0000SEEB03. County: Wayne County MI (FIPS 26163), NDCP study year 2022.", "last_updated": "2026-08-13", "confidence": "high", "note": "Set 2026-08-13 (cluster childcare-0813). NDCP publishes WEEKLY, full-time MEDIAN prices; this is the CENTRE-BASED median (MC* columns), converted with NDCP’s own 52/12 factor and escalated 2022 to 2026 by CUUR0000SEEB03 (x1.20116; 2026 = Jan-Jul mean 396.748 against 2022 annual average 330.305). Tier A: observed 2022 state market rate survey. Michigan caveat: the state MRS does not split by program type and COMBINES centre and family child care home prices, so Detroit is not a pure centre rate and is biased LOW against the other cities. Value is the equal-weight mean of the 0-5 and 6-11 month bands, i.e. ENGINE AGE 0 only. Deliberately NOT NDCP’s MCINFANT column, which is a 0-THROUGH-23-month band and would import two years into a one-year field. Previous value $1,350." }, "childcare.after_school_monthly": { "note": "NOT sourced from NDCP (2026-08-13, cluster childcare-0813): NDCP’s MCSA column is a FULL-TIME weekly school-age rate (summer, or a child not yet in school), not the before-and-after-school wraparound this field holds, and states report inconsistently into it ($82.60/wk Cook County against $197/wk Travis County). Using it would have moved Austin from $425 to $1,025. This field stays on its prior basis and still needs a wraparound-specific source." } }, "bengaluru": { "non_cash.parental_leave_weeks": { "source": "Maternity Benefit Act 1961 s.5(3) as amended 2017: 26 weeks at 100% of average daily wage, employer-paid; Act text via state-government mirror (wb.gov.in PDF; central gov.in domains unreachable at check, flagged)", "last_updated": "2026-08-06", "confidence": "high", "note": "26 = maternity only, and on the household-inclusive basis this IS the household total: India has no statutory paid paternity or parental leave for private-sector employees as of 2026 (CCS Rule 43-A covers government employees only; the 2017 Paternity Benefit Bill lapsed). Caveats: establishments with 10+ employees; 26 weeks for the first two children, 12 for the third and beyond (we store the first-child case). Confirmed on the basis 2026-08-06. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "Maternity Benefit Act 1961 s.5(3) as amended 2017: 26 weeks at 100% of average daily wage, employer-paid; Act text via state-government mirror (wb.gov.in PDF; central gov.in domains unreachable at check, flagged)", "last_updated": "2026-08-06", "confidence": "high", "note": "100% of the average daily wage (3-month average), employer-paid, no monetary cap. Confirmed 2026-08-06." } }, "delhi": { "non_cash.parental_leave_weeks": { "source": "Maternity Benefit Act 1961 s.5(3) as amended 2017: 26 weeks at 100% of average daily wage, employer-paid; Act text via state-government mirror (wb.gov.in PDF; central gov.in domains unreachable at check, flagged)", "last_updated": "2026-08-06", "confidence": "high", "note": "26 = maternity only, and on the household-inclusive basis this IS the household total: India has no statutory paid paternity or parental leave for private-sector employees as of 2026 (CCS Rule 43-A covers government employees only; the 2017 Paternity Benefit Bill lapsed). Caveats: establishments with 10+ employees; 26 weeks for the first two children, 12 for the third and beyond (we store the first-child case). Confirmed on the basis 2026-08-06. Household-inclusive basis (adopted 2026-08-01): all statutory PAID weeks around one birth, summing each parent non-transferable entitlement and counting shared pots once. See docs/superpowers/research/2026-08-06-parental-leave-batch3.md." }, "non_cash.parental_leave_paid_pct": { "source": "Maternity Benefit Act 1961 s.5(3) as amended 2017: 26 weeks at 100% of average daily wage, employer-paid; Act text via state-government mirror (wb.gov.in PDF; central gov.in domains unreachable at check, flagged)", "last_updated": "2026-08-06", "confidence": "high", "note": "100% of the average daily wage (3-month average), employer-paid, no monetary cap. Confirmed 2026-08-06." } }, "nicosia": { "tax.federal_brackets": { "source": "Cyprus income tax ladder as ENACTED for 2026 by the tax-reform law voted in the House Plenary on 22 December 2025 and published in the Official Gazette on 31 December 2025, in force for income earned from 1 January 2026: 0 to 22,000 at 0%, 22,001 to 32,000 at 20%, 32,001 to 42,000 at 25%, 42,001 to 72,000 at 30%, above 72,000 at 35%. PwC Worldwide Tax Summaries Cyprus, Individual, Taxes on personal income, which prints the 2025 and 2026 tables side by side https://taxsummaries.pwc.com/cyprus/individual/taxes-on-personal-income ; KPMG Cyprus, The Cyprus Tax Reform https://assets.kpmg.com/content/dam/kpmg/cy/pdf/2026/cyprus-tax-reform.pdf", "last_updated": "2026-09-13", "confidence": "high", "note": "DO NOT carry the \"tax-free to 20,500\" ladder. That is the draft bill, which parliament amended; 19,500 is the pre-reform 2025 figure. Both are still in wide circulation. The enacted thresholds reconcile to the euro against PwC own accumulated-tax column (10,000 x 20% = 2,000; +10,000 x 25% = 4,500; +30,000 x 30% = 13,500). Cyprus has no joint assessment, so the married ladder mirrors the single one and exists only to satisfy the schema. The 0% band IS the personal allowance: it lives in the first bracket and standard_deduction_single and _married are 0, because granting it in both places would exempt the first EUR 44,000." }, "tax.filing_basis": { "source": "Income Tax Law N.118(I)/2002: spouses are assessed separately, each on their own income, with no aggregation and no income splitting. PwC Worldwide Tax Summaries Cyprus, Individual, Tax administration (\"Separate returns must be filed by each individual on a calendar-year basis\") https://taxsummaries.pwc.com/cyprus/individual/tax-administration", "last_updated": "2026-09-13", "confidence": "high", "note": "Activates the per-filer split in compute(), so a two-earner couple runs as two single-filer clones. That matters twice over in Cyprus: each earner gets their own EUR 22,000 zero band AND their own social-insurance ceiling, so stacking them through one ladder would overstate a two-earner household badly." }, "tax.social_security_rate": { "source": "Employee social insurance 8.8% of insurable emoluments, matched by the employer. Fixed from 1 January 2024 for five years, so unchanged through 31 December 2028. PwC Worldwide Tax Summaries Cyprus, Individual, Other taxes https://taxsummaries.pwc.com/cyprus/individual/other-taxes", "last_updated": "2026-09-13", "confidence": "high", "note": "The Social Insurance Law sets a five-yearly escalation, so the NEXT rise is 1 January 2029 and the indicative ladder runs to roughly 10.3 to 10.7 percent by 2039, each step subject to the actuarial study. This value expires 2029-01-01. The Social Cohesion Fund (2%), Redundancy Fund (1.2%), HRDA levy (0.5%) and Central Holiday Fund are EMPLOYER ONLY and are deliberately absent here; the Social Cohesion Fund is also uncapped, charged on actual rather than insurable earnings." }, "tax.social_security_cap": { "source": "Social Insurance Services notice on the maximum insurable earnings for 2026, issued 22 December 2025 https://sisweb.mlsi.gov.cy/anotato2025/ : EUR 1,325 per week from 5 January 2026 and EUR 5,742 per month from 1 January 2026. Corroborated by KPMG Cyprus, Amendment to the maximum amount of insurable earnings for 2026 https://kpmg.com/cy/en/insights/tax-alerts/amendment-to-the-maximum-amount-of-insurable-earnings-for-2026.html", "last_updated": "2026-09-13", "confidence": "high", "note": "UNIT TRAP, and it is live in 2026. The statute publishes the ceiling BY PAY FREQUENCY and gives no annual figure, and for 2026 the two no longer reconcile: 5,742 x 12 = 68,904 but 1,325 x 52 = 68,900. In 2025 they agreed exactly (1,281 a week = 5,551 a month = 66,612 a year). We carry the MONTHLY-derived 68,904 for two reasons pointing the same way: our reader is a monthly-paid professional whose binding limit is the monthly one, and 68,904 is the higher of the two, therefore the reading less favourable to Cyprus. Per individual earner, never per household. Do NOT carry EUR 62,868, which is the 2024 figure and still circulating. History: 2024 62,868, 2025 66,612, 2026 68,904. Revised every January, so this expires 2027-01-01." }, "tax.gesy_rate": { "source": "GESY (General Healthcare System) employee contribution 2.65% of emoluments, employer 2.90%. Unchanged since the rate step of 1 March 2020 and unchanged for 2026. PwC Worldwide Tax Summaries Cyprus, Individual, Other taxes https://taxsummaries.pwc.com/cyprus/individual/other-taxes", "last_updated": "2026-09-13", "confidence": "high", "note": "Charged on gross, and deductible against the income-tax base inside the Article 14 one-fifth block. The CHARGE is computed here in computeCyprusTaxes from tax.gesy_rate and tax.gesy_cap; the DISPLAY row is computed from healthcare.statutory_employee. Two places therefore hold one rate and one ceiling, which is deliberate (it keeps Cyprus inside the generic statutory-health guards rather than taking the Germany carve-out), and tests/integration/tax-cyprus.test.mjs asserts the two agree so they cannot drift." }, "tax.gesy_cap": { "source": "GHS Law: the GESY ceiling is EUR 180,000 of TOTAL annual income per individual, across employment, self-employment, pensions, dividends, interest and rent combined, not employment income alone. PwC Worldwide Tax Summaries Cyprus, Individual, Other taxes; Bank of Cyprus, Contributions to the General Healthcare System https://www.bankofcyprus.com/en-gb/Personal/the-bank/documents/contributions-to-the-general-healthcare-system-ghs/", "last_updated": "2026-09-13", "confidence": "high", "note": "A fixed figure in the GHS Law, NOT indexed: EUR 180,000 since the system started, untouched by the 2026 reform. Maximum employee charge EUR 4,770 a year. Salary is taken first when the cap is applied, so at every household income this product plausibly models the cap never binds and 2.65% reads as flat. Carried anyway for correctness." }, "tax.allowance_ceiling_pct": { "source": "Article 14 of the Income Tax Law N.118(I)/2002, as amended by the law gazetted 19 December 2019 which raised the overall ceiling from one sixth to one fifth and brought GHS contributions inside it. Employee social insurance, GESY, pension and provident-fund contributions and life premiums all deduct against that ONE combined ceiling. PwC Worldwide Tax Summaries Cyprus, Individual, Deductions https://taxsummaries.pwc.com/cyprus/individual/deductions", "last_updated": "2026-09-13", "confidence": "high", "note": "WHAT THE FIFTH IS A FRACTION OF was the real modelling fork, and PwC own Cyprus sample calculation settles it arithmetically: 74,320 / 5 = 14,864 exactly, and 74,320 - 14,864 = 59,456 exactly. So the base is income AFTER other deductions but BEFORE this block, which for a salary-only employee is gross. It is NOT a fraction of the final chargeable income, which would be circular. For a pure employee the ceiling is INERT: social insurance plus GESY is 11.45% of gross against a 20% ceiling, and the share falls further above the insurable-earnings cap. It only binds once a provident-fund contribution approaches its own 10% sub-limit, which is exactly why retirement_deduction_cap carries 0.10." }, "tax.retirement_deduction_cap": { "source": "Article 14 of the Income Tax Law N.118(I)/2002: individual pension and provident-fund contributions are deductible up to 10% of remuneration, inside the overall one-fifth ceiling shared with social insurance, GESY and life premiums. PwC Worldwide Tax Summaries Cyprus, Individual, Deductions https://taxsummaries.pwc.com/cyprus/individual/deductions", "last_updated": "2026-09-13", "confidence": "medium", "note": "The ceiling measures the voluntary pension or provident-fund contribution only, at its own 10% of remuneration sub-limit, NOT the 20% overall block. 20% here would have been wrong twice over: it is not this item statutory limit, and the engine already takes social insurance and GESY off the same pot inside computeCyprusTaxes, so a 20% figure would have let a user deduct 31.45% of gross against a statutory 20%. Confidence is medium because the 10% sub-limit is sourced to the Big-4 summary rather than to the article text, which could not be read directly." }, "tax.sales_tax_rate": { "source": "Cyprus standard VAT 19%, unchanged for 2026. PwC Worldwide Tax Summaries Cyprus, Corporate, Other taxes https://taxsummaries.pwc.com/cyprus/corporate/other-taxes", "last_updated": "2026-09-13", "confidence": "high", "note": "Reduced rates in force in 2026 are 9%, 5% and a 3% super-reduced band, plus a temporary 0% rating on a list of essential goods that the Tax Department extended to 31 December 2026. None of that changes the standard rate we carry." }, "tax.inbound_regime": { "source": "Article 8(23A) of the Income Tax Law N.118(I)/2002, in force for first employments commencing on or after 1 January 2022 and amended in 2023: 50% of remuneration exempt from income tax, remuneration threshold EUR 55,000 a year, 17 tax years, 15 consecutive prior years of non-residence. PwC Worldwide Tax Summaries Cyprus, Individual, Income determination https://taxsummaries.pwc.com/cyprus/individual/income-determination ; KPMG GMS Flash Alert 2023-150 https://kpmg.com/xx/en/our-insights/gms-flash-alert/flash-alert-2023-150.html", "last_updated": "2026-09-13", "confidence": "high", "note": "THE THRESHOLD IS TESTED ON GROSS, BEFORE THE EXEMPTION, so a EUR 60,000 salary qualifies and the exempt amount is EUR 30,000; the engine default min_income_basis does exactly that. The prior-non-residence test was tightened from 10 years to 15 around 29 June 2023 and 15 is current law, which is also the harder test and so the conservative one. The exemption is INCOME TAX ONLY: social insurance and GESY stay on full gross, which is the most commonly mis-modelled point about this regime and carries its own test. Changing employer does not forfeit the relief. TWO OTHER RELIEFS EXIST AND ARE NOT MODELLED, because they are mutually exclusive with this one and sit below our reader salary: the older 20% capped at EUR 8,550 under Article 8(21A), and a NEW 25% capped at EUR 25,000 under Article 8(21B), introduced by Law 17(I)/2026 for commencements from 2025 to 2030 with a EUR 30,000 threshold." }, "healthcare.premium_monthly_by_family_size": { "source": "GESY (General Healthcare System), universal for residents since 2019, funded by the earmarked payroll contribution rather than by a premium. There is no insurance product a Cyprus resident must buy, so every premium tier is 0 and Cyprus sits in the statutory_contribution regime in data/_healthcare_regime.json. https://www.gesy.org.cy/", "last_updated": "2026-09-13", "confidence": "high", "note": "Voluntary private top-up cover is common and is a frequent employer benefit, at roughly EUR 80 a month per adult for a local plan (band EUR 40 to 150, more for over-55s and for international plans). It is bought to skip specialist waiting lists into a large private hospital sector, not to fill a coverage gap, so it is deliberately NOT priced here and not in the out-of-pocket column either." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Derived from the official GESY co-payment schedule and its statutory annual ceilings: personal doctor EUR 0, referred specialist EUR 6, self-referred specialist EUR 25, per prescription item EUR 1 (about EUR 15 a prescription maximum), per lab test EUR 1, radiology EUR 10, A&E EUR 10 refunded on admission, inpatient EUR 0; annual co-payment ceiling EUR 150 per adult and EUR 75 per child under 21. https://www.gesy.org.cy/", "last_updated": "2026-09-13", "confidence": "medium", "note": "HIGH confidence on the schedule and the ceilings, which are official; MEDIUM on the adopted utilisation, which is derived because no Cyprus survey of realized household co-payment spend exists. Children are capped separately at HALF an adult ceiling rather than exempt, which is where the sub-linearity comes from: the marginal child adds about EUR 60, not EUR 110. The household absolute worst case is EUR 450 a year, which is the number worth quoting, because Cyprus household medical exposure is capped by statute rather than merely low on average. The EUR 25 self-referral contribution is assumed NOT to count toward the ceiling, the reading less favourable to Cyprus." }, "healthcare.dental_vision_monthly": { "source": "GESY covers preventive dental only, in practice one basic annual check-up; adult fillings, crowns, root canals, orthodontics and implants are private, and prescription glasses and contact lenses are not subsidised at all. Build: private check-up and scale beyond the free preventive visit EUR 80, a filling every other year EUR 35 amortised, eye test EUR 35, glasses every two years at EUR 150 giving EUR 75 amortised. https://www.gesy.org.cy/", "last_updated": "2026-09-13", "confidence": "low", "note": "Per ADULT per month, so the engine bills it by adults. Band EUR 225 to 300 a year, top of band taken per the conservative rule. Confidence is low because Cyprus dental clinics overwhelmingly publish \"price on request\" and the two clinic price lists reached had their tables withdrawn. An eye examination can be reached through a GP referral to an ophthalmologist at the EUR 6 specialist co-payment, which is the one part of this line the public system does carry." }, "healthcare.statutory_employee": { "source": "GESY employee contribution 2.65% of emoluments under a ceiling of EUR 180,000 of total annual income per individual (GHS Law). Employer pays a further 2.90%. https://www.gesy.org.cy/ ; PwC Worldwide Tax Summaries Cyprus, Individual, Other taxes https://taxsummaries.pwc.com/cyprus/individual/other-taxes", "last_updated": "2026-09-13", "confidence": "high", "note": "DISPLAY-ONLY: the money is already inside taxes.payroll, so it must never be added to total_costs. computeCyprusTaxes charges the same amount from tax.gesy_rate and tax.gesy_cap, and tests/integration/tax-cyprus.test.mjs asserts this block and those two tax keys agree, so the figure shown and the figure charged cannot drift apart. Maximum EUR 4,770 a year. Because the ceiling is on TOTAL income rather than employment income, and is not indexed, it never binds on a salary alone at any income we model." }, "childcare.free_tier_start_age": { "source": "Cyprus has NO universal public childcare below pre-primary. The free public pre-primary year is compulsory from 4 years 8 months in 2026/27, stepping down toward 4 years 0 months by 2032/33, and it runs 07:45 to 13:05 Monday to Friday. Eurydice Cyprus, Early childhood education and care https://eurydice.eacea.ec.europa.eu/eurypedia/cyprus/early-childhood-education-and-care", "last_updated": "2026-09-13", "confidence": "medium", "note": "Stored as 5 rather than 4 for two reasons, both conservative: the entitlement starts PART WAY THROUGH age four rather than at the birthday, and it is a HALF DAY ending at 13:05, so a two-earner household still buys afternoon cover and preschool_monthly stays positive rather than dropping to zero. The national under-four fee subsidy covering up to 80% of fees is gated on drawing Cyprus child benefit, which our reader household income fails, so subsidy_per_kid_annual is 0." }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Cyprus child benefit is INCOME-TESTED on the gross annual income of all family members in the previous calendar year. Thresholds: EUR 49,000 for one dependent child, EUR 59,000 for two, plus EUR 5,000 for each further child. European Commission, Your social security rights in Cyprus, 2025, pp.7-8 https://ec.europa.eu/social/BlobServlet?docId=13744&langId=en ; OECD, Tax and benefit policy descriptions for Cyprus 2024, section 4.1 https://www.oecd.org/content/dam/oecd/en/topics/policy-sub-issues/incomes-support-redistribution-and-work-incentives/TaxBEN-Cyprus-latest.pdf", "last_updated": "2026-09-13", "confidence": "high", "note": "ZERO for our reader, and three independent gates each produce that answer. The income test: the household band this product models starts at EUR 60,000, above the two-child threshold, and the official amounts table prints a literal dash in the one-child column from EUR 49,000.01 upward. Residence: five years of legal continuous residence in the government-controlled areas is required before any claim, which a mover fails outright. Assets: over EUR 1.2 million disqualifies. The only non-zero corner inside the modelled band is three children between EUR 60,000 and 64,000, or four between 60,000 and 69,000. A Cabinet-approved reform of 2026-09-10 would raise the three-child threshold to EUR 90,000 and the four-child to EUR 100,000 from January 2027, but it still needs a House vote and is deliberately NOT priced in. Separately, the childbirth grant (EUR 1,059.40 for a first child, rising by birth order) IS contributory rather than income-tested so an insured household does receive it, but it is one-off and belongs in no annual figure." }, "non_cash.vacation_days_statutory": { "source": "Annual Holidays with Pay Law 8/1967 as amended: the statutory minimum is four weeks, expressed as 20 WORKING DAYS on a five-day week and 24 working days on a six-day week, exclusive of public holidays, with the full entitlement conditional on 48 weeks worked in the leave year and pro rata below that.", "last_updated": "2026-09-13", "confidence": "high", "note": "UNIT CHECKED: the 20 and the 24 are the same four weeks counted against a different weekly divisor, so there is no six-day-week inflation here of the kind Norwegian virkedager produce. Public holidays are explicitly EXCLUDED and sit in their own field. The practical private-sector floor is a day higher, because an employer exempted from the Central Holiday Fund must grant leave at least as good as the statutory minimum, but the stored value stays the legal floor of 20 per the field definition." }, "non_cash.public_holidays": { "source": "Central Bank of Cyprus 2026 public-holiday notice read against the Council of Ministers list: 15 days, being 10 fixed dates and 5 Orthodox-Easter-linked days. The rule list and the substitution finding live in data/_holiday_rules.json.", "last_updated": "2026-09-13", "confidence": "high", "note": "NO substitute weekday when a holiday falls at the weekend, proved from the notice itself, which flags 15 August and 26 December as Saturdays in 2026 and declares no replacement for either. 13 of the 15 therefore land on a weekday in 2026. CYPRUS IS THE ONE COUNTRY IN THIS FIELD WHOSE VALUE IS OBSERVED PRACTICE RATHER THAN A STATUTORY FLOOR, and that is deliberate: public holidays are compulsory only for retail workers under the Shops Law closure regime, so for an office employee the legal floor is ZERO and storing zero would be absurd. The declared list binds the public sector and in practice reaches white-collar contracts nearly in full. Easter Tuesday is the day most often absent from private contracts." }, "non_cash.sick_leave": { "source": "Social Insurance Law N. 59(I)/2010, Fourth Schedule (benefit rate table), consolidated text on CyLaw http://www.cylaw.org/nomoi/enop/non-ind/2010_1_59/full.html ; European Commission, Your social security rights in Cyprus, 2025, p.25 https://ec.europa.eu/social/BlobServlet?docId=13744&langId=en ; OECD, Tax and benefit policy descriptions for Cyprus 2024, sections 2.1.3 and 2.1.7, for the six-day benefit week https://www.oecd.org/content/dam/oecd/en/topics/policy-sub-issues/incomes-support-redistribution-and-work-incentives/TaxBEN-Cyprus-latest.pdf ; basic insurable earnings 2026 from the MLSI Social Insurance Services table", "last_updated": "2026-09-13", "confidence": "high", "note": "THE UNIT IS THE FINDING. The statute says 156 days, but the Cyprus benefit week is a SIX-day week, so 156 days is 26 weeks, and the stored duration is written in weeks to stop a reader converting at five or seven days. Reading 156 as calendar days understates by 15%, as five-day working days overstates by 20%. There is NO employer obligation at all in Cyprus law: the benefit comes from the Social Insurance Fund and is not paid while the employer pays the full wage. The supplementary component is capped at one week of basic insurable earnings by Part II of the Fourth Schedule, a sentence deliberately ABSENT from the maternity part, which is why the 2026 sickness ceiling is EUR 353.22 a week against a maternity ceiling of EUR 954.00. Extension from 156 to 312 days is discretionary, so only the 26-week floor is stored." }, "non_cash.parental_leave_weeks": { "source": "Social Insurance Law N. 59(I)/2010, art. 29 (maternity allowance, 22 consecutive weeks for a first or second child, 26 for a third or later, and art. 29(2) defines the week as Monday to Sunday) and art. 29B (parental leave allowance, 8 paid weeks for a first child rising to 10, 12 and 14 by birth order), consolidated text on CyLaw http://www.cylaw.org/nomoi/enop/non-ind/2010_1_59/full.html ; Leave (Paternity, Parental, Carer, Force Majeure) and Flexible Working Arrangements Law N. 216(I)/2022, gazetted 16 December 2022, transposing Directive (EU) 2019/1158", "last_updated": "2026-09-13", "confidence": "medium", "note": "Household-inclusive basis: maternity 22 plus paternity 2 plus 8 PAID parental weeks for each parent, non-transferable, so 40 weeks around a first child. The 10 further parental weeks each parent holds are UNPAID job-protected and are excluded by the field definition. TWO CORRECTIONS TO THE EC 2025 GUIDE, which is stale twice over: the third-child maternity increment is +4 weeks, not +8 (Law 14(I)/2024, in force 1 March 2024, which also raised the base from 18 to 22), and the paid parental entitlement is 8, 10, 12 and 14 weeks by birth order, not a flat 8 (Law 157(I)/2024, in force 31 March 2025). Confidence is medium on one point only: whether the 8 paid parental weeks are per PARENT or per CHILD could not be closed beyond the statute non-transferability language, and a shared pot would make this 32 rather than 40. A HARD THROTTLE of 5 parental weeks per calendar year also means the 8 cannot all be drawn in one year." }, "non_cash.parental_leave_paid_pct": { "source": "Social Insurance Law N. 59(I)/2010, Fourth Schedule: maternity and paternity allowance 72% of the weekly value of the basic insurance points plus 72% of the supplementary; parental leave allowance 72% basic but only 60% supplementary. Basic insurable earnings 2026 EUR 220.76 a week, maximum insurable earnings EUR 1,325 a week, so a high earner holds one basic point and five supplementary.", "last_updated": "2026-09-13", "confidence": "medium", "note": "STORED AS 0, AND THAT IS CORRECT rather than a gap: parental_leave_capped covers all 40 weeks, so no percentage-paid weeks are left for this field to describe, and the schema requires 0 once the slices cover the entitlement. The headline rate lives in parental_leave_capped.pct. For the record, a duration-weighted headline would be 67: 24 weeks of maternity and paternity at 72% and 16 parental weeks at the 60% supplementary rate that actually binds a high earner. THE CEILINGS ARE THE REAL STORY and there are TWO of them: EUR 954.00 a week for maternity and paternity, which equals 0.72 x 1,325 to the cent and is the independent check that the maternity supplementary component carries NO sub-cap, and EUR 821.49 for parental leave. The engine carries one capped slice, so parental_leave_capped applies the LOWER of the two to all 40 weeks, which understates the maternity ceiling and is the conservative direction. A caveat that belongs on any Cyprus page: every allowance is computed on the relevant contribution year, one to two years back, so a household that has just moved to Cyprus receives far less than these ceilings imply." }, "non_cash.parental_leave_capped": { "source": "Social Insurance Law N. 59(I)/2010, Fourth Schedule, priced against the 2026 insurable-earnings parameters: basic insurable earnings EUR 220.76 a week and maximum insurable earnings EUR 1,325 a week (MLSI Social Insurance Services). Maternity and paternity ceiling 0.72 x 220.76 + 0.72 x 1,104.24 = EUR 954.00 a week; parental ceiling 0.72 x 220.76 + 0.60 x 1,104.24 = EUR 821.49 a week.", "last_updated": "2026-09-13", "confidence": "medium", "note": "CYPRUS HAS TWO STATUTORY WEEKLY CEILINGS and the engine carries one capped slice, so this block applies the LOWER of them, the parental-leave ceiling of EUR 821.49, to all 40 weeks. That understates the 24 maternity and paternity weeks, whose own ceiling is EUR 954.00, and understating a benefit is the conservative direction. The alternative shape, a 24-week slice at 954 with the 16 parental weeks left uncapped at a percentage, was rejected because it would pay a EUR 100,000 earner 60% of salary for those 16 weeks against a real ceiling of 42.7%, which flatters Cyprus by far more than this understates it. The EUR 954.00 figure is itself the proof that maternity carries NO supplementary sub-cap: it equals 0.72 x the maximum weekly insurable earnings to the cent, and the capping sentence that produces the much lower sickness ceiling of EUR 353.22 appears in Part II of the Fourth Schedule and is absent from Part III. Realised replacement rates on the maternity ceiling: 72% at EUR 68,904 of gross, 62% at 80,000, 49.6% at 100,000, 41.3% at 120,000. A 5-weeks-per-calendar-year throttle applies to the parental slice, and every allowance is computed on the relevant contribution year one to two years back, so a household that has just moved to Cyprus receives materially less than any of this implies." }, "tax.property_tax_rate_pct": { "source": "Nicosia Municipality immovable property levy 0.35 per thousand and Sewerage Board of Nicosia levy 0.63 per thousand, both on the 1 JANUARY 2021 general valuation by the Department of Lands and Surveys, plus refuse collection of about EUR 168 a year: roughly EUR 330 a year on the modelled EUR 191,000 two-bedroom, or 0.17% of current market value. https://www.nicosia.org.cy/el-GR/news/announcements/2025/21408/ and https://www.sbn.org.cy/el/apoxeteftika-teli", "last_updated": "2026-09-13", "confidence": "medium", "note": "THE \"1980 VALUES\" PREMISE IS STALE and nearly every aggregator still repeats it. That was the NATIONAL Immovable Property Tax, abolished with effect from 1 January 2017. Nicosia surviving municipal and sewerage levies run on the 1 January 2021 general valuation, which is why a nominal per-mille rate applied to current value gives roughly the right answer here rather than being an order of magnitude out. Two independent routes agree within 4%: our 0.173% against a published Nicosia 120 square metre worked example at 0.167%. A claimed 0.30% sewerage rate was REJECTED as producing bills far outside the observed EUR 100 to 300 band. The sewerage USAGE charge per cubic metre is counted once, inside utilities, and deliberately not repeated here." }, "housing.median_rent_by_bedroom": { "source": "INDEX.cy Nicosia apartment rent guide https://index.cy/articles/apartments-for-rent-in-nicosia-guide/ , cross-read against Investropa, Numbeo Nicosia (623 entries from 56 contributors, updated 7 September 2026), Global Property Guide and the RICS-KPMG Cyprus Q1 2026 index. The Council for the Registration of Real Estate Agents 2026 city figures put Nicosia at EUR 600 to 650 for a one-bedroom and EUR 800 to 900 for a two-bedroom, below our adopted values.", "last_updated": "2026-09-13", "confidence": "medium", "note": "Asking rents, unfurnished to lightly furnished; furnished adds EUR 100 to 200 a month and is the norm for studios and one-bedrooms. Adopted values lean to the higher of the credible reads per the conservative rule, except the three-bedroom where the INDEX.cy 1,450 was cut against Global Property Guide 1,150 and Numbeo suburb-weighted 1,290. rentspotcy.com was REJECTED outright: it reports a Nicosia penthouse median of EUR 15,000 and an island-wide studio median of EUR 1,800 off four listings, and taking it on trust would have poisoned the whole field. bazaraki.com, the dominant local portal, returns HTTP 403 to automated fetches, so its data reached this figure only second-hand." }, "housing.rent_percentiles": { "source": "INDEX.cy Nicosia rent-guide district bands against the adopted EUR 1,050 citywide two-bedroom median: Lakatamia and Latsia two-bedrooms from EUR 750 at the low end, Engomi EUR 1,100 to 1,500 and Aglantzia EUR 950 to 1,300 at the high end.", "last_updated": "2026-09-13", "confidence": "medium", "note": "Cross-neighbourhood rather than within-listing basis, which is what the Cyprus evidence supports: no Cyprus source publishes within-bedroom quartiles. Band edge equals median_rent_by_bedroom times the multiplier, computed at display time." }, "housing.median_home_price_by_bedroom": { "source": "Derived from a blended EUR 2,250 per square metre, a suburb-weighted read between the EUR 2,042 outside-centre and EUR 2,763 centre figures, at assumed unit sizes of 42, 58, 85 and 115 square metres. Global Property Guide Cyprus price history https://www.globalpropertyguide.com/europe/cyprus/price-history", "last_updated": "2026-09-13", "confidence": "medium", "note": "DERIVED from euro per square metre, not from observed per-bedroom medians, because none exist for Nicosia. Sanity check: the two-bedroom figure of EUR 191,000 sits just above the EUR 177,000 district-wide average apartment transaction price for 2025, the right relationship given that average includes studios, one-bedrooms and the cheap rural periphery. The Limassol equivalent district average is EUR 403,000, which is the scale of the gap." }, "housing.utilities_monthly": { "source": "EAC domestic tariff 01 build-up (10.34 cents energy, 3.66 distribution, 0.87 transmission, 0.50 RES fund, plus a monthly fuel adjustment, at the 9% reduced VAT extended to 31 March 2027), giving an all-in 25 to 32 cents per kWh; INDEX.cy Cyprus electricity guide https://index.cy/articles/electricity-in-cyprus-guide/ ; expats.cy Cyprus cost of living 2026 for water and municipal charges.", "last_updated": "2026-09-13", "confidence": "medium", "note": "Build: electricity 100 plus water 25 plus tenant-billed refuse and sewerage 15 plus common expenses (koinochrista) 50 equals EUR 190. THE ELECTRICITY LINE IS AN ANNUAL AVERAGE THAT INCLUDES THE SUMMER COOLING LOAD: a Nicosia two-bedroom runs EUR 55 to 85 a month in winter and EUR 100 to 150 in summer, so a figure taken in April would understate by about a third. Nicosia is INLAND and carries the island heaviest air-conditioning load, at 40 degrees plus from May to October with no sea breeze. No separate heating line: the same split units reverse-cycle for heat. Common expenses are a real and unavoidable Cyprus tenant cost that models built on northern-European assumptions routinely miss. A competing aggregator figure of EUR 30 to 60 for electricity was REJECTED as irreconcilable with the tariff unless consumption is under 200 kWh a month, which means no summer cooling at all. Cross-check: Numbeo Nicosia basic utilities for 85 square metres is EUR 185.48." }, "housing.internet_phone_monthly": { "source": "expats.cy Cyprus cost of living 2026, Nicosia block (broadband 200 Mbps EUR 19.99 to 33, unlimited mobile EUR 14.99 to 22.40) and Numbeo Nicosia (internet 60 Mbps and above EUR 29.67, mobile with 10GB and above EUR 18.08). Operators: Cyta, Epic, PrimeTel, Cablenet.", "last_updated": "2026-09-13", "confidence": "medium", "note": "EUR 28 fixed plus EUR 20 mobile. Not read off an operator price list: Cyprus operators discount heavily for 12-month contracts and fixed-plus-mobile bundles, so the list price a newcomer first sees is above this." }, "housing.renters_insurance_monthly": { "source": "digicare-insurance.com Cyprus home insurance guide https://digicare-insurance.com/blog/home-insurance-cyprus-guide : contents-only cover for a rented apartment at about EUR 120 a year; buildings-plus-contents on an owner-occupied apartment at roughly 0.15% of value a year on the Nicosia stock.", "last_updated": "2026-09-13", "confidence": "low", "note": "Lowest-confidence pair in the Nicosia file. No Cyprus insurer rate card was reached and Numbeo carries no insurance line. Cyprus does not mandate household insurance for owner-occupiers; mortgage lenders require buildings cover, which is why almost every mortgaged owner carries it." }, "childcare.preschool_monthly": { "source": "Numbeo Nicosia private full-day preschool or kindergarten EUR 458.44 a month (range 400 to 500); privateschools.cy early-years fee guide EUR 4,500 to 7,000 a year; the national under-four subsidy schedule, whose own ceilings of EUR 350 for under-twos against EUR 250 for over-twos independently confirm that infants are the expensive tier. https://cyprus-mail.com/2026/09/07/applications-open-for-nursery-subsidies", "last_updated": "2026-09-13", "confidence": "low", "note": "REAL PRIVATE OUT-OF-POCKET: Cyprus has no universal public childcare below pre-primary. Almost no Nicosia nursery publishes a fee schedule (privateschools.cy lists 106 Nicosia settings with zero confirmed fees), so the after-school figure of EUR 160 is the weakest number in the file and is derived from the subsidy schedule EUR 50 afternoon-care increment plus the all-day public-kindergarten extra-fee model. Infant at EUR 560 applies the staff-ratio premium to the top of the observed EUR 340 to 600 band. The national subsidy covering up to 80% of fees is gated on drawing Cyprus child benefit, which our reader household fails, so subsidy_per_kid_annual is 0." }, "food._summary": { "source": "expats.cy Cyprus cost of living 2026 Nicosia block (single-person groceries EUR 250 to 300 a month) and Cyprus Mail, Cost of living in Cyprus in 2026 https://cyprus-mail.com/2026/04/10/cost-of-living-in-cyprus-in-2026 (single 250 to 400, family of four 700 to 1,200), with Numbeo Nicosia unit prices and a Numbeo single-person basket excluding rent of EUR 850 a month.", "last_updated": "2026-09-13", "confidence": "medium", "note": "Groceries take the top of the Nicosia-specific band per the conservative rule. The per-child figure is derived at about 0.58 of an adult and reconciles the sources: two adults at 300 plus two children at 175 is EUR 950 a month, inside the Cyprus Mail family band, whereas a competing claim of EUR 400 to 600 for a family of four cannot be reconciled with any credible per-person figure and is rejected. Food and non-alcoholic beverages rose 6.16% year on year to March 2026, so this group ages fast. Groceries are NOT where Nicosia is cheap: they are only 7.1% below Limassol while rent is 42% below. Dining out at EUR 150 is built from Numbeo Nicosia prices (inexpensive meal EUR 15, mid-range cover EUR 30) at six to eight outings a month; the frequency is an assumption." }, "transit._summary": { "source": "Cyprus Public Transport official tariff https://www.publictransport.com.cy/cms/page/cash-tickets , effective 3 August 2026: 30-day pass EUR 66.90, single EUR 2.70. Car cost assembled line by line: depreciation 2,720, finance 300, insurance 700, road tax 90, fuel 1,480 at EUR 1.625 a litre, MOT 25, maintenance and tyres 650, parking 450.", "last_updated": "2026-09-13", "confidence": "medium", "note": "Nicosia bus pass is HIGHER than Limassol (EUR 66.90 against EUR 50) because Cyprus Public Transport applied a contractually indexed increase on 3 August 2026 that the separately operated Limassol and Paphos systems did not follow. A circulating EUR 40 figure is stale and EUR 50 is a different city. Adopted on internal consistency: the CPT 7-day fare of EUR 33.40 is exactly half. Public transport carries about 3% of greater-Nicosia journeys against 67% by private car, which is why car_dependent is set; Cyprus has no rail of any kind and a bus commute runs 50 minutes against 23 by car. No consolidated Cyprus car total-cost-of-ownership study exists, so the annual figure is a build with each line sourced." }, "discretionary._summary": { "source": "Numbeo Nicosia (fitness club EUR 66.11 a month, cinema ticket EUR 10.00) and expats.cy gym pricing, plus published 2026 Cyprus accommodation rates for the domestic-holiday line (mid-range hotels and holiday apartments EUR 60 to 110 a night, Troodos guesthouses EUR 35 to 70 a room, peak July and August plus 30 to 50%).", "last_updated": "2026-09-13", "confidence": "low", "note": "Entertainment build: gym 68 plus cinema 20 plus streaming 20 plus going out 30. Restaurant meals are excluded here because they sit in the dining-out line; do not double-count. The domestic-holiday figure is built from a seven-night coastal or Troodos break at EUR 45 a night per person sharing, food at EUR 40 a day, fuel EUR 100 and activities EUR 150, rounded up. Cyprus is 240 km end to end so a domestic holiday is a short drive rather than a flight. No CYSTAT domestic-tourism expenditure-per-trip figure exists, so both lines are baskets rather than measurements." }, "travel_home.avg_round_trip_per_person_to_other_city": { "source": "Lufthansa Larnaca to Seattle lead-in fare from EUR 1,019 https://www.lufthansa.com/xx/en/flights/flight-larnaca-seattle , with momondo, Expedia and Travelocity quotes for the pair spanning USD 676 to 2,084 on 2026-09-13.", "last_updated": "2026-09-13", "confidence": "low", "note": "No direct service exists; every itinerary is one or two stops through a European hub, about 18 hours 50 minutes each way. Aggregator cheapest-found numbers are lead-in fares on awkward two-stop itineraries booked far out, and Larnaca-originating fares run consistently higher than Seattle-originating ones on the same metal, so the median a household actually books lands near USD 1,100 to 1,300. EUR 1,000 is the conservative middle. Live aggregator quotes on one day, not a fare-basis series." }, "compensation_index": { "source": "levels.fyi Cyprus senior software engineer median EUR 62,781 for Nicosia https://www.levels.fyi/t/software-engineer/levels/senior/locations/cyprus ; Glassdoor Cyprus senior software engineer average EUR 72,000; Jobs Nicosia recruiter survey of 24 hiring managers, March 2026 (senior backend EUR 6,800 a month, senior frontend 5,800, senior DevOps 7,500) https://jobsnicosia.com/cyprus-tech-salaries-2026/ ; CYSTAT average gross monthly earnings EUR 2,601 in Q1 2026, with ICT at EUR 3,197 a month in 2024 https://cyprus-mail.com/2026/07/29/average-monthly-pay-in-cyprus-rises-to-e2601-but-1-in-3-workers-earn-less-than-e1500", "last_updated": "2026-09-13", "confidence": "medium", "note": "Implied Nicosia total comp of about EUR 49,800 on the dataset basis (0.23 times EUR 216,732, being the USD 250,000 anchor at 1.1535). Derivation: the levels.fyi read gives a raw 0.29, and the same source queried the same day at the same seniority for Madrid (EUR 65,015), Barcelona (EUR 60,598) and Athens (EUR 35,158) shows this dataset existing anchors sitting at a median 0.86 of a levels.fyi-derived raw index, giving 0.25. Taken down to 0.23 for two reasons: the declared anchor basis is ALL-LEVELS median total comp while the figure used is a senior one, and 0.25 would rank Nicosia above Madrid when levels.fyi itself puts Madrid slightly above Nicosia. Nicosia is the DISCOUNT within Cyprus rather than the premium: the high-paying forex, fintech and relocated-tech employers are concentrated in Limassol, while Nicosia is banks, the central bank, Big-4 professional services, universities and government-adjacent IT. The Cyprus 80% IP Box exemption on qualifying software income is documented as having pulled engineering pay up across every level, which is why Cyprus sits above Greece." } }, "limassol": { "tax.federal_brackets": { "source": "Cyprus income tax ladder as ENACTED for 2026 by the tax-reform law voted in the House Plenary on 22 December 2025 and published in the Official Gazette on 31 December 2025, in force for income earned from 1 January 2026: 0 to 22,000 at 0%, 22,001 to 32,000 at 20%, 32,001 to 42,000 at 25%, 42,001 to 72,000 at 30%, above 72,000 at 35%. PwC Worldwide Tax Summaries Cyprus, Individual, Taxes on personal income, which prints the 2025 and 2026 tables side by side https://taxsummaries.pwc.com/cyprus/individual/taxes-on-personal-income ; KPMG Cyprus, The Cyprus Tax Reform https://assets.kpmg.com/content/dam/kpmg/cy/pdf/2026/cyprus-tax-reform.pdf", "last_updated": "2026-09-13", "confidence": "high", "note": "DO NOT carry the \"tax-free to 20,500\" ladder. That is the draft bill, which parliament amended; 19,500 is the pre-reform 2025 figure. Both are still in wide circulation. The enacted thresholds reconcile to the euro against PwC own accumulated-tax column (10,000 x 20% = 2,000; +10,000 x 25% = 4,500; +30,000 x 30% = 13,500). Cyprus has no joint assessment, so the married ladder mirrors the single one and exists only to satisfy the schema. The 0% band IS the personal allowance: it lives in the first bracket and standard_deduction_single and _married are 0, because granting it in both places would exempt the first EUR 44,000." }, "tax.filing_basis": { "source": "Income Tax Law N.118(I)/2002: spouses are assessed separately, each on their own income, with no aggregation and no income splitting. PwC Worldwide Tax Summaries Cyprus, Individual, Tax administration (\"Separate returns must be filed by each individual on a calendar-year basis\") https://taxsummaries.pwc.com/cyprus/individual/tax-administration", "last_updated": "2026-09-13", "confidence": "high", "note": "Activates the per-filer split in compute(), so a two-earner couple runs as two single-filer clones. That matters twice over in Cyprus: each earner gets their own EUR 22,000 zero band AND their own social-insurance ceiling, so stacking them through one ladder would overstate a two-earner household badly." }, "tax.social_security_rate": { "source": "Employee social insurance 8.8% of insurable emoluments, matched by the employer. Fixed from 1 January 2024 for five years, so unchanged through 31 December 2028. PwC Worldwide Tax Summaries Cyprus, Individual, Other taxes https://taxsummaries.pwc.com/cyprus/individual/other-taxes", "last_updated": "2026-09-13", "confidence": "high", "note": "The Social Insurance Law sets a five-yearly escalation, so the NEXT rise is 1 January 2029 and the indicative ladder runs to roughly 10.3 to 10.7 percent by 2039, each step subject to the actuarial study. This value expires 2029-01-01. The Social Cohesion Fund (2%), Redundancy Fund (1.2%), HRDA levy (0.5%) and Central Holiday Fund are EMPLOYER ONLY and are deliberately absent here; the Social Cohesion Fund is also uncapped, charged on actual rather than insurable earnings." }, "tax.social_security_cap": { "source": "Social Insurance Services notice on the maximum insurable earnings for 2026, issued 22 December 2025 https://sisweb.mlsi.gov.cy/anotato2025/ : EUR 1,325 per week from 5 January 2026 and EUR 5,742 per month from 1 January 2026. Corroborated by KPMG Cyprus, Amendment to the maximum amount of insurable earnings for 2026 https://kpmg.com/cy/en/insights/tax-alerts/amendment-to-the-maximum-amount-of-insurable-earnings-for-2026.html", "last_updated": "2026-09-13", "confidence": "high", "note": "UNIT TRAP, and it is live in 2026. The statute publishes the ceiling BY PAY FREQUENCY and gives no annual figure, and for 2026 the two no longer reconcile: 5,742 x 12 = 68,904 but 1,325 x 52 = 68,900. In 2025 they agreed exactly (1,281 a week = 5,551 a month = 66,612 a year). We carry the MONTHLY-derived 68,904 for two reasons pointing the same way: our reader is a monthly-paid professional whose binding limit is the monthly one, and 68,904 is the higher of the two, therefore the reading less favourable to Cyprus. Per individual earner, never per household. Do NOT carry EUR 62,868, which is the 2024 figure and still circulating. History: 2024 62,868, 2025 66,612, 2026 68,904. Revised every January, so this expires 2027-01-01." }, "tax.gesy_rate": { "source": "GESY (General Healthcare System) employee contribution 2.65% of emoluments, employer 2.90%. Unchanged since the rate step of 1 March 2020 and unchanged for 2026. PwC Worldwide Tax Summaries Cyprus, Individual, Other taxes https://taxsummaries.pwc.com/cyprus/individual/other-taxes", "last_updated": "2026-09-13", "confidence": "high", "note": "Charged on gross, and deductible against the income-tax base inside the Article 14 one-fifth block. The CHARGE is computed here in computeCyprusTaxes from tax.gesy_rate and tax.gesy_cap; the DISPLAY row is computed from healthcare.statutory_employee. Two places therefore hold one rate and one ceiling, which is deliberate (it keeps Cyprus inside the generic statutory-health guards rather than taking the Germany carve-out), and tests/integration/tax-cyprus.test.mjs asserts the two agree so they cannot drift." }, "tax.gesy_cap": { "source": "GHS Law: the GESY ceiling is EUR 180,000 of TOTAL annual income per individual, across employment, self-employment, pensions, dividends, interest and rent combined, not employment income alone. PwC Worldwide Tax Summaries Cyprus, Individual, Other taxes; Bank of Cyprus, Contributions to the General Healthcare System https://www.bankofcyprus.com/en-gb/Personal/the-bank/documents/contributions-to-the-general-healthcare-system-ghs/", "last_updated": "2026-09-13", "confidence": "high", "note": "A fixed figure in the GHS Law, NOT indexed: EUR 180,000 since the system started, untouched by the 2026 reform. Maximum employee charge EUR 4,770 a year. Salary is taken first when the cap is applied, so at every household income this product plausibly models the cap never binds and 2.65% reads as flat. Carried anyway for correctness." }, "tax.allowance_ceiling_pct": { "source": "Article 14 of the Income Tax Law N.118(I)/2002, as amended by the law gazetted 19 December 2019 which raised the overall ceiling from one sixth to one fifth and brought GHS contributions inside it. Employee social insurance, GESY, pension and provident-fund contributions and life premiums all deduct against that ONE combined ceiling. PwC Worldwide Tax Summaries Cyprus, Individual, Deductions https://taxsummaries.pwc.com/cyprus/individual/deductions", "last_updated": "2026-09-13", "confidence": "high", "note": "WHAT THE FIFTH IS A FRACTION OF was the real modelling fork, and PwC own Cyprus sample calculation settles it arithmetically: 74,320 / 5 = 14,864 exactly, and 74,320 - 14,864 = 59,456 exactly. So the base is income AFTER other deductions but BEFORE this block, which for a salary-only employee is gross. It is NOT a fraction of the final chargeable income, which would be circular. For a pure employee the ceiling is INERT: social insurance plus GESY is 11.45% of gross against a 20% ceiling, and the share falls further above the insurable-earnings cap. It only binds once a provident-fund contribution approaches its own 10% sub-limit, which is exactly why retirement_deduction_cap carries 0.10." }, "tax.retirement_deduction_cap": { "source": "Article 14 of the Income Tax Law N.118(I)/2002: individual pension and provident-fund contributions are deductible up to 10% of remuneration, inside the overall one-fifth ceiling shared with social insurance, GESY and life premiums. PwC Worldwide Tax Summaries Cyprus, Individual, Deductions https://taxsummaries.pwc.com/cyprus/individual/deductions", "last_updated": "2026-09-13", "confidence": "medium", "note": "The ceiling measures the voluntary pension or provident-fund contribution only, at its own 10% of remuneration sub-limit, NOT the 20% overall block. 20% here would have been wrong twice over: it is not this item statutory limit, and the engine already takes social insurance and GESY off the same pot inside computeCyprusTaxes, so a 20% figure would have let a user deduct 31.45% of gross against a statutory 20%. Confidence is medium because the 10% sub-limit is sourced to the Big-4 summary rather than to the article text, which could not be read directly." }, "tax.sales_tax_rate": { "source": "Cyprus standard VAT 19%, unchanged for 2026. PwC Worldwide Tax Summaries Cyprus, Corporate, Other taxes https://taxsummaries.pwc.com/cyprus/corporate/other-taxes", "last_updated": "2026-09-13", "confidence": "high", "note": "Reduced rates in force in 2026 are 9%, 5% and a 3% super-reduced band, plus a temporary 0% rating on a list of essential goods that the Tax Department extended to 31 December 2026. None of that changes the standard rate we carry." }, "tax.inbound_regime": { "source": "Article 8(23A) of the Income Tax Law N.118(I)/2002, in force for first employments commencing on or after 1 January 2022 and amended in 2023: 50% of remuneration exempt from income tax, remuneration threshold EUR 55,000 a year, 17 tax years, 15 consecutive prior years of non-residence. PwC Worldwide Tax Summaries Cyprus, Individual, Income determination https://taxsummaries.pwc.com/cyprus/individual/income-determination ; KPMG GMS Flash Alert 2023-150 https://kpmg.com/xx/en/our-insights/gms-flash-alert/flash-alert-2023-150.html", "last_updated": "2026-09-13", "confidence": "high", "note": "THE THRESHOLD IS TESTED ON GROSS, BEFORE THE EXEMPTION, so a EUR 60,000 salary qualifies and the exempt amount is EUR 30,000; the engine default min_income_basis does exactly that. The prior-non-residence test was tightened from 10 years to 15 around 29 June 2023 and 15 is current law, which is also the harder test and so the conservative one. The exemption is INCOME TAX ONLY: social insurance and GESY stay on full gross, which is the most commonly mis-modelled point about this regime and carries its own test. Changing employer does not forfeit the relief. TWO OTHER RELIEFS EXIST AND ARE NOT MODELLED, because they are mutually exclusive with this one and sit below our reader salary: the older 20% capped at EUR 8,550 under Article 8(21A), and a NEW 25% capped at EUR 25,000 under Article 8(21B), introduced by Law 17(I)/2026 for commencements from 2025 to 2030 with a EUR 30,000 threshold." }, "healthcare.premium_monthly_by_family_size": { "source": "GESY (General Healthcare System), universal for residents since 2019, funded by the earmarked payroll contribution rather than by a premium. There is no insurance product a Cyprus resident must buy, so every premium tier is 0 and Cyprus sits in the statutory_contribution regime in data/_healthcare_regime.json. https://www.gesy.org.cy/", "last_updated": "2026-09-13", "confidence": "high", "note": "Voluntary private top-up cover is common and is a frequent employer benefit, at roughly EUR 80 a month per adult for a local plan (band EUR 40 to 150, more for over-55s and for international plans). It is bought to skip specialist waiting lists into a large private hospital sector, not to fill a coverage gap, so it is deliberately NOT priced here and not in the out-of-pocket column either." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Derived from the official GESY co-payment schedule and its statutory annual ceilings: personal doctor EUR 0, referred specialist EUR 6, self-referred specialist EUR 25, per prescription item EUR 1 (about EUR 15 a prescription maximum), per lab test EUR 1, radiology EUR 10, A&E EUR 10 refunded on admission, inpatient EUR 0; annual co-payment ceiling EUR 150 per adult and EUR 75 per child under 21. https://www.gesy.org.cy/", "last_updated": "2026-09-13", "confidence": "medium", "note": "HIGH confidence on the schedule and the ceilings, which are official; MEDIUM on the adopted utilisation, which is derived because no Cyprus survey of realized household co-payment spend exists. Children are capped separately at HALF an adult ceiling rather than exempt, which is where the sub-linearity comes from: the marginal child adds about EUR 60, not EUR 110. The household absolute worst case is EUR 450 a year, which is the number worth quoting, because Cyprus household medical exposure is capped by statute rather than merely low on average. The EUR 25 self-referral contribution is assumed NOT to count toward the ceiling, the reading less favourable to Cyprus." }, "healthcare.dental_vision_monthly": { "source": "GESY covers preventive dental only, in practice one basic annual check-up; adult fillings, crowns, root canals, orthodontics and implants are private, and prescription glasses and contact lenses are not subsidised at all. Build: private check-up and scale beyond the free preventive visit EUR 80, a filling every other year EUR 35 amortised, eye test EUR 35, glasses every two years at EUR 150 giving EUR 75 amortised. https://www.gesy.org.cy/", "last_updated": "2026-09-13", "confidence": "low", "note": "Per ADULT per month, so the engine bills it by adults. Band EUR 225 to 300 a year, top of band taken per the conservative rule. Confidence is low because Cyprus dental clinics overwhelmingly publish \"price on request\" and the two clinic price lists reached had their tables withdrawn. An eye examination can be reached through a GP referral to an ophthalmologist at the EUR 6 specialist co-payment, which is the one part of this line the public system does carry." }, "healthcare.statutory_employee": { "source": "GESY employee contribution 2.65% of emoluments under a ceiling of EUR 180,000 of total annual income per individual (GHS Law). Employer pays a further 2.90%. https://www.gesy.org.cy/ ; PwC Worldwide Tax Summaries Cyprus, Individual, Other taxes https://taxsummaries.pwc.com/cyprus/individual/other-taxes", "last_updated": "2026-09-13", "confidence": "high", "note": "DISPLAY-ONLY: the money is already inside taxes.payroll, so it must never be added to total_costs. computeCyprusTaxes charges the same amount from tax.gesy_rate and tax.gesy_cap, and tests/integration/tax-cyprus.test.mjs asserts this block and those two tax keys agree, so the figure shown and the figure charged cannot drift apart. Maximum EUR 4,770 a year. Because the ceiling is on TOTAL income rather than employment income, and is not indexed, it never binds on a salary alone at any income we model." }, "childcare.free_tier_start_age": { "source": "Cyprus has NO universal public childcare below pre-primary. The free public pre-primary year is compulsory from 4 years 8 months in 2026/27, stepping down toward 4 years 0 months by 2032/33, and it runs 07:45 to 13:05 Monday to Friday. Eurydice Cyprus, Early childhood education and care https://eurydice.eacea.ec.europa.eu/eurypedia/cyprus/early-childhood-education-and-care", "last_updated": "2026-09-13", "confidence": "medium", "note": "Stored as 5 rather than 4 for two reasons, both conservative: the entitlement starts PART WAY THROUGH age four rather than at the birthday, and it is a HALF DAY ending at 13:05, so a two-earner household still buys afternoon cover and preschool_monthly stays positive rather than dropping to zero. The national under-four fee subsidy covering up to 80% of fees is gated on drawing Cyprus child benefit, which our reader household income fails, so subsidy_per_kid_annual is 0." }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Cyprus child benefit is INCOME-TESTED on the gross annual income of all family members in the previous calendar year. Thresholds: EUR 49,000 for one dependent child, EUR 59,000 for two, plus EUR 5,000 for each further child. European Commission, Your social security rights in Cyprus, 2025, pp.7-8 https://ec.europa.eu/social/BlobServlet?docId=13744&langId=en ; OECD, Tax and benefit policy descriptions for Cyprus 2024, section 4.1 https://www.oecd.org/content/dam/oecd/en/topics/policy-sub-issues/incomes-support-redistribution-and-work-incentives/TaxBEN-Cyprus-latest.pdf", "last_updated": "2026-09-13", "confidence": "high", "note": "ZERO for our reader, and three independent gates each produce that answer. The income test: the household band this product models starts at EUR 60,000, above the two-child threshold, and the official amounts table prints a literal dash in the one-child column from EUR 49,000.01 upward. Residence: five years of legal continuous residence in the government-controlled areas is required before any claim, which a mover fails outright. Assets: over EUR 1.2 million disqualifies. The only non-zero corner inside the modelled band is three children between EUR 60,000 and 64,000, or four between 60,000 and 69,000. A Cabinet-approved reform of 2026-09-10 would raise the three-child threshold to EUR 90,000 and the four-child to EUR 100,000 from January 2027, but it still needs a House vote and is deliberately NOT priced in. Separately, the childbirth grant (EUR 1,059.40 for a first child, rising by birth order) IS contributory rather than income-tested so an insured household does receive it, but it is one-off and belongs in no annual figure." }, "non_cash.vacation_days_statutory": { "source": "Annual Holidays with Pay Law 8/1967 as amended: the statutory minimum is four weeks, expressed as 20 WORKING DAYS on a five-day week and 24 working days on a six-day week, exclusive of public holidays, with the full entitlement conditional on 48 weeks worked in the leave year and pro rata below that.", "last_updated": "2026-09-13", "confidence": "high", "note": "UNIT CHECKED: the 20 and the 24 are the same four weeks counted against a different weekly divisor, so there is no six-day-week inflation here of the kind Norwegian virkedager produce. Public holidays are explicitly EXCLUDED and sit in their own field. The practical private-sector floor is a day higher, because an employer exempted from the Central Holiday Fund must grant leave at least as good as the statutory minimum, but the stored value stays the legal floor of 20 per the field definition." }, "non_cash.public_holidays": { "source": "Central Bank of Cyprus 2026 public-holiday notice read against the Council of Ministers list: 15 days, being 10 fixed dates and 5 Orthodox-Easter-linked days. The rule list and the substitution finding live in data/_holiday_rules.json.", "last_updated": "2026-09-13", "confidence": "high", "note": "NO substitute weekday when a holiday falls at the weekend, proved from the notice itself, which flags 15 August and 26 December as Saturdays in 2026 and declares no replacement for either. 13 of the 15 therefore land on a weekday in 2026. CYPRUS IS THE ONE COUNTRY IN THIS FIELD WHOSE VALUE IS OBSERVED PRACTICE RATHER THAN A STATUTORY FLOOR, and that is deliberate: public holidays are compulsory only for retail workers under the Shops Law closure regime, so for an office employee the legal floor is ZERO and storing zero would be absurd. The declared list binds the public sector and in practice reaches white-collar contracts nearly in full. Easter Tuesday is the day most often absent from private contracts." }, "non_cash.sick_leave": { "source": "Social Insurance Law N. 59(I)/2010, Fourth Schedule (benefit rate table), consolidated text on CyLaw http://www.cylaw.org/nomoi/enop/non-ind/2010_1_59/full.html ; European Commission, Your social security rights in Cyprus, 2025, p.25 https://ec.europa.eu/social/BlobServlet?docId=13744&langId=en ; OECD, Tax and benefit policy descriptions for Cyprus 2024, sections 2.1.3 and 2.1.7, for the six-day benefit week https://www.oecd.org/content/dam/oecd/en/topics/policy-sub-issues/incomes-support-redistribution-and-work-incentives/TaxBEN-Cyprus-latest.pdf ; basic insurable earnings 2026 from the MLSI Social Insurance Services table", "last_updated": "2026-09-13", "confidence": "high", "note": "THE UNIT IS THE FINDING. The statute says 156 days, but the Cyprus benefit week is a SIX-day week, so 156 days is 26 weeks, and the stored duration is written in weeks to stop a reader converting at five or seven days. Reading 156 as calendar days understates by 15%, as five-day working days overstates by 20%. There is NO employer obligation at all in Cyprus law: the benefit comes from the Social Insurance Fund and is not paid while the employer pays the full wage. The supplementary component is capped at one week of basic insurable earnings by Part II of the Fourth Schedule, a sentence deliberately ABSENT from the maternity part, which is why the 2026 sickness ceiling is EUR 353.22 a week against a maternity ceiling of EUR 954.00. Extension from 156 to 312 days is discretionary, so only the 26-week floor is stored." }, "non_cash.parental_leave_weeks": { "source": "Social Insurance Law N. 59(I)/2010, art. 29 (maternity allowance, 22 consecutive weeks for a first or second child, 26 for a third or later, and art. 29(2) defines the week as Monday to Sunday) and art. 29B (parental leave allowance, 8 paid weeks for a first child rising to 10, 12 and 14 by birth order), consolidated text on CyLaw http://www.cylaw.org/nomoi/enop/non-ind/2010_1_59/full.html ; Leave (Paternity, Parental, Carer, Force Majeure) and Flexible Working Arrangements Law N. 216(I)/2022, gazetted 16 December 2022, transposing Directive (EU) 2019/1158", "last_updated": "2026-09-13", "confidence": "medium", "note": "Household-inclusive basis: maternity 22 plus paternity 2 plus 8 PAID parental weeks for each parent, non-transferable, so 40 weeks around a first child. The 10 further parental weeks each parent holds are UNPAID job-protected and are excluded by the field definition. TWO CORRECTIONS TO THE EC 2025 GUIDE, which is stale twice over: the third-child maternity increment is +4 weeks, not +8 (Law 14(I)/2024, in force 1 March 2024, which also raised the base from 18 to 22), and the paid parental entitlement is 8, 10, 12 and 14 weeks by birth order, not a flat 8 (Law 157(I)/2024, in force 31 March 2025). Confidence is medium on one point only: whether the 8 paid parental weeks are per PARENT or per CHILD could not be closed beyond the statute non-transferability language, and a shared pot would make this 32 rather than 40. A HARD THROTTLE of 5 parental weeks per calendar year also means the 8 cannot all be drawn in one year." }, "non_cash.parental_leave_paid_pct": { "source": "Social Insurance Law N. 59(I)/2010, Fourth Schedule: maternity and paternity allowance 72% of the weekly value of the basic insurance points plus 72% of the supplementary; parental leave allowance 72% basic but only 60% supplementary. Basic insurable earnings 2026 EUR 220.76 a week, maximum insurable earnings EUR 1,325 a week, so a high earner holds one basic point and five supplementary.", "last_updated": "2026-09-13", "confidence": "medium", "note": "STORED AS 0, AND THAT IS CORRECT rather than a gap: parental_leave_capped covers all 40 weeks, so no percentage-paid weeks are left for this field to describe, and the schema requires 0 once the slices cover the entitlement. The headline rate lives in parental_leave_capped.pct. For the record, a duration-weighted headline would be 67: 24 weeks of maternity and paternity at 72% and 16 parental weeks at the 60% supplementary rate that actually binds a high earner. THE CEILINGS ARE THE REAL STORY and there are TWO of them: EUR 954.00 a week for maternity and paternity, which equals 0.72 x 1,325 to the cent and is the independent check that the maternity supplementary component carries NO sub-cap, and EUR 821.49 for parental leave. The engine carries one capped slice, so parental_leave_capped applies the LOWER of the two to all 40 weeks, which understates the maternity ceiling and is the conservative direction. A caveat that belongs on any Cyprus page: every allowance is computed on the relevant contribution year, one to two years back, so a household that has just moved to Cyprus receives far less than these ceilings imply." }, "non_cash.parental_leave_capped": { "source": "Social Insurance Law N. 59(I)/2010, Fourth Schedule, priced against the 2026 insurable-earnings parameters: basic insurable earnings EUR 220.76 a week and maximum insurable earnings EUR 1,325 a week (MLSI Social Insurance Services). Maternity and paternity ceiling 0.72 x 220.76 + 0.72 x 1,104.24 = EUR 954.00 a week; parental ceiling 0.72 x 220.76 + 0.60 x 1,104.24 = EUR 821.49 a week.", "last_updated": "2026-09-13", "confidence": "medium", "note": "CYPRUS HAS TWO STATUTORY WEEKLY CEILINGS and the engine carries one capped slice, so this block applies the LOWER of them, the parental-leave ceiling of EUR 821.49, to all 40 weeks. That understates the 24 maternity and paternity weeks, whose own ceiling is EUR 954.00, and understating a benefit is the conservative direction. The alternative shape, a 24-week slice at 954 with the 16 parental weeks left uncapped at a percentage, was rejected because it would pay a EUR 100,000 earner 60% of salary for those 16 weeks against a real ceiling of 42.7%, which flatters Cyprus by far more than this understates it. The EUR 954.00 figure is itself the proof that maternity carries NO supplementary sub-cap: it equals 0.72 x the maximum weekly insurable earnings to the cent, and the capping sentence that produces the much lower sickness ceiling of EUR 353.22 appears in Part II of the Fourth Schedule and is absent from Part III. Realised replacement rates on the maternity ceiling: 72% at EUR 68,904 of gross, 62% at 80,000, 49.6% at 100,000, 41.3% at 120,000. A 5-weeks-per-calendar-year throttle applies to the parental slice, and every allowance is computed on the relevant contribution year one to two years back, so a household that has just moved to Cyprus receives materially less than any of this implies." }, "tax.property_tax_rate_pct": { "source": "EOA Lemesos (Limassol sewerage board) published fees https://eoalemesos.org.cy/en/fees , assessed on the 1 JANUARY 2021 general valuation, plus the Limassol Municipality immovable property fee and refuse collection. Value-based municipal and sewerage charges alone are about EUR 205 a year on the modelled EUR 340,000 two-bedroom; the all-in owner-occupier total including refuse is about EUR 452 a year, which is 0.13% of market value.", "last_updated": "2026-09-13", "confidence": "low", "note": "The ALL-IN figure is stored, to match the field meaning of what an owner actually pays each year. The Limassol percentage is LOWER than Nicosia (0.13 against 0.17) purely because the denominator is bigger: a broadly similar euro bill sits on a much more valuable property. Cyprus ABOLISHED the national Immovable Property Tax from 1 January 2017, so the widely quoted 1980-value rates are dead; EOA Lemesos assesses on 2021 values. Confidence is low because eoalemesos.org.cy publishes water and sewerage only: the municipal immovable property fee and refuse figures are Cyprus-municipality mid-range rather than the Limassol own schedule, which is likely a Greek-language PDF. The sewerage usage fee per cubic metre is counted once, inside utilities." }, "housing.median_rent_by_bedroom": { "source": "Landbank Analytics asking-rent prints reported by cbn.com.cy (January 2025) and cyprus-mail.com/2025/07/15 (July 2025), brought forward on the RICS-KPMG 2026 Q1 apartment rent print of plus 5.10% year on year and Eurostat Cyprus rents plus 4.2% to March 2026, plus a live 30-listing spitogatos.gr sample. The Council for the Registration of Real Estate Agents 2026 city figures put Limassol at EUR 1,000 to 1,100 for a one-bedroom and EUR 1,400 to 1,500 for a two-bedroom.", "last_updated": "2026-09-13", "confidence": "medium", "note": "Limassol asks 2.4 to 2.5 times Nicosia at every bedroom count and holds 45% of all Cyprus apartment rental listings. The market is genuinely bifurcated: a large relocated forex, fintech, shipping and post-2022 tech workforce plus a high-end seafront and tower segment pull asking rents well above the rest of the island, and the advertised stock skews FURNISHED, which these figures therefore are. Studio is interpolated from the one-bedroom at about 0.82 because no Cyprus source publishes a Limassol studio series. TWO WARNINGS. Rent and price do not reconcile through the yield: the adopted pair implies 7.2% gross against the RICS-KPMG Cyprus apartment yield of 5.44%, which is unresolved, and if achieved-rent data ever appears the rents come down rather than the prices going up. And a content farm has laundered a fabricated EUR 30 per square metre Limassol rent figure into the Cyprus Mail, attributed to Investropa, so a future pass that verifies against national press can re-import it." }, "housing.rent_percentiles": { "source": "Spread of the live 30-listing spitogatos.gr Limassol two-bedroom sample against the Landbank asking-rent prints, taken across the inland-suburb floor and the seafront and new-tower band the relocated workforce rents in.", "last_updated": "2026-09-13", "confidence": "low", "note": "Wider than the Nicosia and Larnaca pairs on purpose: Limassol is the one Cyprus market with a genuine luxury tail, so its within-bedroom dispersion is real rather than noise. Band edge equals median_rent_by_bedroom times the multiplier, computed at display time." }, "housing.median_home_price_by_bedroom": { "source": "Landbank Analytics H1 2026 off-plan study reported by cyprus-mail.com/2026/09/03, whose all-sizes Limassol apartment median is EUR 297,600, priced up to per-bedroom figures at typical Cyprus unit sizes of 45, 60, 95 and 130 square metres at EUR 4,000, 3,800, 3,500 and 3,400 per square metre.", "last_updated": "2026-09-13", "confidence": "medium", "note": "Transaction-based rather than asking. The two-bedroom figure is about 14% above the all-sizes median, the right direction given that median includes studios and one-bedrooms. Worth knowing for any Cyprus copy: Limassol takes 40 cents of every euro spent on Cyprus property and is by far the most expensive place to RENT, yet its MEDIAN off-plan apartment is cheaper than the Paphos one. Limassol expensiveness is a rent and top-of-market story, not a median-purchase one. The Numbeo per-square-metre figures (EUR 4,951 centre, EUR 3,672 outside) sit above the transaction-implied number, a contributor skew toward new coastal stock." }, "housing.utilities_monthly": { "source": "Official EOA Lemesos water and sewerage tariff https://eoalemesos.org.cy/en/fees plus Eurostat electricity price statistics for the Cyprus all-in household rate of about EUR 0.26 per kWh. Build: electricity EUR 110 on 5,000 kWh a year, water EUR 20, sewerage usage EUR 8, tenant-billed refuse EUR 13.", "last_updated": "2026-09-13", "confidence": "medium", "note": "ANNUAL AVERAGE INCLUDING THE SUMMER COOLING LOAD. A Limassol two-bedroom runs air conditioning hard from June to September and the July and August bimonthly bill can be 2 to 2.5 times the shoulder-season one, so a figure taken in April would understate by roughly a third. No separate heating line: there is no gas grid, water is solar-thermal with an electric immersion backup, and space heating is the same reverse-cycle units already inside the electricity line. The EAC own tariff page refused the connection, so electricity is priced off the Eurostat all-in national figure, which is a better basis anyway; the refuse figure is unsourced Cyprus mid-range." }, "housing.internet_phone_monthly": { "source": "Numbeo Limassol, 2 September 2026: fixed broadband 60 Mbps and above unlimited EUR 30.00 (range 25 to 40) and a mobile plan with calls and 10GB and above data EUR 19.36. Operators: Cyta, Epic, PrimeTel.", "last_updated": "2026-09-13", "confidence": "medium", "note": "EUR 33 fibre plus EUR 19 mobile, inside the observed EUR 45 to 60 bundle band. Cyprus advertised consumer prices include the 19% VAT, and line rental and the router are usually inside the headline price. No operator tariff page was fetched, so the market price points are not individually sourced." }, "housing.renters_insurance_monthly": { "source": "Cyprus market bands: contents-only cover for a furnished rented two-bedroom at EUR 20,000 to 30,000 sum insured prices at EUR 90 to 180 a year; buildings-plus-contents on an owner-occupied apartment at roughly 0.10 to 0.14% of REBUILD value, which on MARKET value lands near 0.08 to 0.09% because a large share of a Limassol price is land and location.", "last_updated": "2026-09-13", "confidence": "low", "note": "Both figures are reasoned estimates with no citable source and are the lowest-confidence numbers in the Limassol file: no Cyprus insurer rate card was reached and Numbeo carries no insurance line. The rebuild-versus-market distinction is the trap here, and it is why the Limassol homeowner rate is lower than the Nicosia one despite the same market." }, "childcare.preschool_monthly": { "source": "Numbeo Limassol private full-day preschool or kindergarten EUR 613.96 a month (range 350 to 950), against published Limassol nursery annual bands of EUR 4,500 to 7,000 (Erma Babyland, Happy Kids, Little Wonders, Angel Sun Nest, Dream Wonder). Little Wonders publishes its structure https://www.littlewonders.com.cy/admission . Eurydice Cyprus for the free pre-primary year.", "last_updated": "2026-09-13", "confidence": "medium", "note": "THE HOURS TRAP, and it is why Numbeo sits above the published bands: the Cyprus norm is a 07:30 to 13:30 CORE DAY, with 13:30 to 15:00 afternoon care and lunch charged separately as optional add-ons. The EUR 4,500 to 7,000 annual bands are therefore NOT a full working day, and a household with two people working to 17:00 pays the band plus the afternoon add-on plus lunch and still needs cover to 17:00. Infant at EUR 700 reflects genuinely thin supply below 12 months in Limassol plus a 15 to 20% infant staffing premium; no setting publishes an under-two fee. Recorded for context, not priced: the Limassol relocated workforce supports an international-school-adjacent early-years tier at EUR 9,650 a year (The Island Private School K1, EUR 11,750 in year one with registration and deposit), roughly 1.7 to 2 times a mainstream local nursery, and even that is cheap against Western European or Gulf international-school pricing." }, "food._summary": { "source": "Eurostat 2025 price level indices for Cyprus food, Numbeo Limassol unit prices (inexpensive meal, mid-range cover for two) and a Numbeo single-person basket excluding rent of EUR 965.90 a month at 882 entries from 89 contributors, updated 2 September 2026.", "last_updated": "2026-09-13", "confidence": "medium", "note": "Limassol groceries run about 7.6% above Nicosia and dining out about 23% above, which is the smaller half of the Limassol premium: the large half is rent. Food and non-alcoholic beverages rose 6.16% year on year to March 2026, so this group ages fast. The per-child figure is derived rather than observed; no Cyprus source publishes a per-child grocery line." }, "transit._summary": { "source": "EMEL (Limassol urban bus operator) official fare page https://limassolbuses.com/fare-changes-2/ : personalised 30-day Motion card EUR 45 rising to EUR 50 from 5 November 2026. Car cost assembled from the Cyprus Fuel Retail Price Observatory via cyprus-mail.com/2026/09/07 plus depreciation, insurance, road tax, MOT, maintenance and parking.", "last_updated": "2026-09-13", "confidence": "medium", "note": "The pass carries the announced 5 November 2026 price rather than the EUR 45 that expires inside the data year: conservative and forward-correct. Numbeo independently reports EUR 45 today, which corroborates the current price. Limassol runs its OWN operator, which is why its pass is cheaper than the Nicosia and Larnaca ones despite being the expensive city: Cyprus Public Transport applied an indexed increase on 3 August 2026 to the Nicosia and Larnaca networks that EMEL did not follow. Public transport carries about 5% of Cyprus journeys to work against 67% by private car, and there is no rail anywhere on the island, which is why car_dependent is set." }, "discretionary._summary": { "source": "Numbeo Limassol gym and cinema lines plus a sourced Cyprus social-life band, with published 2026 Cyprus accommodation rates for the domestic-holiday figure.", "last_updated": "2026-09-13", "confidence": "low", "note": "Neither line is measured anywhere: both are baskets calibrated against the Athens row and the Numbeo Limassol components. Restaurant meals sit in the dining-out line and are not counted here." }, "travel_home.avg_round_trip_per_person_to_other_city": { "source": "Momondo and Expedia quotes for Larnaca or Paphos to Seattle on 2026-09-13, one or two stops through a European hub, with the Lufthansa lead-in at EUR 1,019.", "last_updated": "2026-09-13", "confidence": "low", "note": "No direct service exists. Set EUR 100 above the Nicosia figure because a Limassol household drives to Larnaca or Paphos and books the same itineraries at a marginally later average purchase date; the honest read is that this is one number for the island with a small local adjustment, not a separately measured fare. Live aggregator quotes on one day, not a fare-basis series." }, "compensation_index": { "source": "levels.fyi Limassol all-levels software engineer median total comp, the Jobs Nicosia survey of 24 Cyprus hiring managers (March 2026) which measures Limassol paying 4 to 7% more than Nicosia for tech roles, the Numbeo Limassol-versus-Nicosia average net salary gap of 40.1%, and CYSTAT average gross monthly earnings of EUR 2,601 in Q1 2026.", "last_updated": "2026-09-13", "confidence": "low", "note": "Implied Limassol total comp of about EUR 60,700 on the dataset basis (0.28 times EUR 216,732). Limassol is unambiguously the island pay leader: it is the centre of the forex and CFD brokerage industry, shipping, fintech and the post-2022 relocated tech employers, and it holds 45% of Cyprus rental listings. The SPREAD in the evidence is why confidence is low. A mechanical levels.fyi derivation gives 0.30 to 0.315, the tech-specific recruiter survey implies only 4 to 7% over Nicosia (about 0.24), and the Numbeo all-occupation average implies 40% (about 0.32). 0.28 is the reasoned middle: it keeps Limassol clearly first in Cyprus and above Madrid at 0.24, and deliberately BELOW Barcelona and Tallinn at 0.29, both of which have deeper product-engineering markets with big-tech satellites and scale-ups where seniors clear EUR 80,000 to 100,000. levels.fyi city splits inside Cyprus are demonstrably unusable on their own: the same source prints Paphos above Limassol, and prints Limassol both above and below Nicosia depending on whether the query is all-levels or senior. One thing no aggregator captures: Exness gives relocating staff a company car and a private-school or kindergarten allowance, both of which land on lines this calculator models as costs, so 0.28 understates the relocated-senior case." } }, "larnaca": { "tax.federal_brackets": { "source": "Cyprus income tax ladder as ENACTED for 2026 by the tax-reform law voted in the House Plenary on 22 December 2025 and published in the Official Gazette on 31 December 2025, in force for income earned from 1 January 2026: 0 to 22,000 at 0%, 22,001 to 32,000 at 20%, 32,001 to 42,000 at 25%, 42,001 to 72,000 at 30%, above 72,000 at 35%. PwC Worldwide Tax Summaries Cyprus, Individual, Taxes on personal income, which prints the 2025 and 2026 tables side by side https://taxsummaries.pwc.com/cyprus/individual/taxes-on-personal-income ; KPMG Cyprus, The Cyprus Tax Reform https://assets.kpmg.com/content/dam/kpmg/cy/pdf/2026/cyprus-tax-reform.pdf", "last_updated": "2026-09-13", "confidence": "high", "note": "DO NOT carry the \"tax-free to 20,500\" ladder. That is the draft bill, which parliament amended; 19,500 is the pre-reform 2025 figure. Both are still in wide circulation. The enacted thresholds reconcile to the euro against PwC own accumulated-tax column (10,000 x 20% = 2,000; +10,000 x 25% = 4,500; +30,000 x 30% = 13,500). Cyprus has no joint assessment, so the married ladder mirrors the single one and exists only to satisfy the schema. The 0% band IS the personal allowance: it lives in the first bracket and standard_deduction_single and _married are 0, because granting it in both places would exempt the first EUR 44,000." }, "tax.filing_basis": { "source": "Income Tax Law N.118(I)/2002: spouses are assessed separately, each on their own income, with no aggregation and no income splitting. PwC Worldwide Tax Summaries Cyprus, Individual, Tax administration (\"Separate returns must be filed by each individual on a calendar-year basis\") https://taxsummaries.pwc.com/cyprus/individual/tax-administration", "last_updated": "2026-09-13", "confidence": "high", "note": "Activates the per-filer split in compute(), so a two-earner couple runs as two single-filer clones. That matters twice over in Cyprus: each earner gets their own EUR 22,000 zero band AND their own social-insurance ceiling, so stacking them through one ladder would overstate a two-earner household badly." }, "tax.social_security_rate": { "source": "Employee social insurance 8.8% of insurable emoluments, matched by the employer. Fixed from 1 January 2024 for five years, so unchanged through 31 December 2028. PwC Worldwide Tax Summaries Cyprus, Individual, Other taxes https://taxsummaries.pwc.com/cyprus/individual/other-taxes", "last_updated": "2026-09-13", "confidence": "high", "note": "The Social Insurance Law sets a five-yearly escalation, so the NEXT rise is 1 January 2029 and the indicative ladder runs to roughly 10.3 to 10.7 percent by 2039, each step subject to the actuarial study. This value expires 2029-01-01. The Social Cohesion Fund (2%), Redundancy Fund (1.2%), HRDA levy (0.5%) and Central Holiday Fund are EMPLOYER ONLY and are deliberately absent here; the Social Cohesion Fund is also uncapped, charged on actual rather than insurable earnings." }, "tax.social_security_cap": { "source": "Social Insurance Services notice on the maximum insurable earnings for 2026, issued 22 December 2025 https://sisweb.mlsi.gov.cy/anotato2025/ : EUR 1,325 per week from 5 January 2026 and EUR 5,742 per month from 1 January 2026. Corroborated by KPMG Cyprus, Amendment to the maximum amount of insurable earnings for 2026 https://kpmg.com/cy/en/insights/tax-alerts/amendment-to-the-maximum-amount-of-insurable-earnings-for-2026.html", "last_updated": "2026-09-13", "confidence": "high", "note": "UNIT TRAP, and it is live in 2026. The statute publishes the ceiling BY PAY FREQUENCY and gives no annual figure, and for 2026 the two no longer reconcile: 5,742 x 12 = 68,904 but 1,325 x 52 = 68,900. In 2025 they agreed exactly (1,281 a week = 5,551 a month = 66,612 a year). We carry the MONTHLY-derived 68,904 for two reasons pointing the same way: our reader is a monthly-paid professional whose binding limit is the monthly one, and 68,904 is the higher of the two, therefore the reading less favourable to Cyprus. Per individual earner, never per household. Do NOT carry EUR 62,868, which is the 2024 figure and still circulating. History: 2024 62,868, 2025 66,612, 2026 68,904. Revised every January, so this expires 2027-01-01." }, "tax.gesy_rate": { "source": "GESY (General Healthcare System) employee contribution 2.65% of emoluments, employer 2.90%. Unchanged since the rate step of 1 March 2020 and unchanged for 2026. PwC Worldwide Tax Summaries Cyprus, Individual, Other taxes https://taxsummaries.pwc.com/cyprus/individual/other-taxes", "last_updated": "2026-09-13", "confidence": "high", "note": "Charged on gross, and deductible against the income-tax base inside the Article 14 one-fifth block. The CHARGE is computed here in computeCyprusTaxes from tax.gesy_rate and tax.gesy_cap; the DISPLAY row is computed from healthcare.statutory_employee. Two places therefore hold one rate and one ceiling, which is deliberate (it keeps Cyprus inside the generic statutory-health guards rather than taking the Germany carve-out), and tests/integration/tax-cyprus.test.mjs asserts the two agree so they cannot drift." }, "tax.gesy_cap": { "source": "GHS Law: the GESY ceiling is EUR 180,000 of TOTAL annual income per individual, across employment, self-employment, pensions, dividends, interest and rent combined, not employment income alone. PwC Worldwide Tax Summaries Cyprus, Individual, Other taxes; Bank of Cyprus, Contributions to the General Healthcare System https://www.bankofcyprus.com/en-gb/Personal/the-bank/documents/contributions-to-the-general-healthcare-system-ghs/", "last_updated": "2026-09-13", "confidence": "high", "note": "A fixed figure in the GHS Law, NOT indexed: EUR 180,000 since the system started, untouched by the 2026 reform. Maximum employee charge EUR 4,770 a year. Salary is taken first when the cap is applied, so at every household income this product plausibly models the cap never binds and 2.65% reads as flat. Carried anyway for correctness." }, "tax.allowance_ceiling_pct": { "source": "Article 14 of the Income Tax Law N.118(I)/2002, as amended by the law gazetted 19 December 2019 which raised the overall ceiling from one sixth to one fifth and brought GHS contributions inside it. Employee social insurance, GESY, pension and provident-fund contributions and life premiums all deduct against that ONE combined ceiling. PwC Worldwide Tax Summaries Cyprus, Individual, Deductions https://taxsummaries.pwc.com/cyprus/individual/deductions", "last_updated": "2026-09-13", "confidence": "high", "note": "WHAT THE FIFTH IS A FRACTION OF was the real modelling fork, and PwC own Cyprus sample calculation settles it arithmetically: 74,320 / 5 = 14,864 exactly, and 74,320 - 14,864 = 59,456 exactly. So the base is income AFTER other deductions but BEFORE this block, which for a salary-only employee is gross. It is NOT a fraction of the final chargeable income, which would be circular. For a pure employee the ceiling is INERT: social insurance plus GESY is 11.45% of gross against a 20% ceiling, and the share falls further above the insurable-earnings cap. It only binds once a provident-fund contribution approaches its own 10% sub-limit, which is exactly why retirement_deduction_cap carries 0.10." }, "tax.retirement_deduction_cap": { "source": "Article 14 of the Income Tax Law N.118(I)/2002: individual pension and provident-fund contributions are deductible up to 10% of remuneration, inside the overall one-fifth ceiling shared with social insurance, GESY and life premiums. PwC Worldwide Tax Summaries Cyprus, Individual, Deductions https://taxsummaries.pwc.com/cyprus/individual/deductions", "last_updated": "2026-09-13", "confidence": "medium", "note": "The ceiling measures the voluntary pension or provident-fund contribution only, at its own 10% of remuneration sub-limit, NOT the 20% overall block. 20% here would have been wrong twice over: it is not this item statutory limit, and the engine already takes social insurance and GESY off the same pot inside computeCyprusTaxes, so a 20% figure would have let a user deduct 31.45% of gross against a statutory 20%. Confidence is medium because the 10% sub-limit is sourced to the Big-4 summary rather than to the article text, which could not be read directly." }, "tax.sales_tax_rate": { "source": "Cyprus standard VAT 19%, unchanged for 2026. PwC Worldwide Tax Summaries Cyprus, Corporate, Other taxes https://taxsummaries.pwc.com/cyprus/corporate/other-taxes", "last_updated": "2026-09-13", "confidence": "high", "note": "Reduced rates in force in 2026 are 9%, 5% and a 3% super-reduced band, plus a temporary 0% rating on a list of essential goods that the Tax Department extended to 31 December 2026. None of that changes the standard rate we carry." }, "tax.inbound_regime": { "source": "Article 8(23A) of the Income Tax Law N.118(I)/2002, in force for first employments commencing on or after 1 January 2022 and amended in 2023: 50% of remuneration exempt from income tax, remuneration threshold EUR 55,000 a year, 17 tax years, 15 consecutive prior years of non-residence. PwC Worldwide Tax Summaries Cyprus, Individual, Income determination https://taxsummaries.pwc.com/cyprus/individual/income-determination ; KPMG GMS Flash Alert 2023-150 https://kpmg.com/xx/en/our-insights/gms-flash-alert/flash-alert-2023-150.html", "last_updated": "2026-09-13", "confidence": "high", "note": "THE THRESHOLD IS TESTED ON GROSS, BEFORE THE EXEMPTION, so a EUR 60,000 salary qualifies and the exempt amount is EUR 30,000; the engine default min_income_basis does exactly that. The prior-non-residence test was tightened from 10 years to 15 around 29 June 2023 and 15 is current law, which is also the harder test and so the conservative one. The exemption is INCOME TAX ONLY: social insurance and GESY stay on full gross, which is the most commonly mis-modelled point about this regime and carries its own test. Changing employer does not forfeit the relief. TWO OTHER RELIEFS EXIST AND ARE NOT MODELLED, because they are mutually exclusive with this one and sit below our reader salary: the older 20% capped at EUR 8,550 under Article 8(21A), and a NEW 25% capped at EUR 25,000 under Article 8(21B), introduced by Law 17(I)/2026 for commencements from 2025 to 2030 with a EUR 30,000 threshold." }, "healthcare.premium_monthly_by_family_size": { "source": "GESY (General Healthcare System), universal for residents since 2019, funded by the earmarked payroll contribution rather than by a premium. There is no insurance product a Cyprus resident must buy, so every premium tier is 0 and Cyprus sits in the statutory_contribution regime in data/_healthcare_regime.json. https://www.gesy.org.cy/", "last_updated": "2026-09-13", "confidence": "high", "note": "Voluntary private top-up cover is common and is a frequent employer benefit, at roughly EUR 80 a month per adult for a local plan (band EUR 40 to 150, more for over-55s and for international plans). It is bought to skip specialist waiting lists into a large private hospital sector, not to fill a coverage gap, so it is deliberately NOT priced here and not in the out-of-pocket column either." }, "healthcare.annual_out_of_pocket_by_family_size": { "source": "Derived from the official GESY co-payment schedule and its statutory annual ceilings: personal doctor EUR 0, referred specialist EUR 6, self-referred specialist EUR 25, per prescription item EUR 1 (about EUR 15 a prescription maximum), per lab test EUR 1, radiology EUR 10, A&E EUR 10 refunded on admission, inpatient EUR 0; annual co-payment ceiling EUR 150 per adult and EUR 75 per child under 21. https://www.gesy.org.cy/", "last_updated": "2026-09-13", "confidence": "medium", "note": "HIGH confidence on the schedule and the ceilings, which are official; MEDIUM on the adopted utilisation, which is derived because no Cyprus survey of realized household co-payment spend exists. Children are capped separately at HALF an adult ceiling rather than exempt, which is where the sub-linearity comes from: the marginal child adds about EUR 60, not EUR 110. The household absolute worst case is EUR 450 a year, which is the number worth quoting, because Cyprus household medical exposure is capped by statute rather than merely low on average. The EUR 25 self-referral contribution is assumed NOT to count toward the ceiling, the reading less favourable to Cyprus." }, "healthcare.dental_vision_monthly": { "source": "GESY covers preventive dental only, in practice one basic annual check-up; adult fillings, crowns, root canals, orthodontics and implants are private, and prescription glasses and contact lenses are not subsidised at all. Build: private check-up and scale beyond the free preventive visit EUR 80, a filling every other year EUR 35 amortised, eye test EUR 35, glasses every two years at EUR 150 giving EUR 75 amortised. https://www.gesy.org.cy/", "last_updated": "2026-09-13", "confidence": "low", "note": "Per ADULT per month, so the engine bills it by adults. Band EUR 225 to 300 a year, top of band taken per the conservative rule. Confidence is low because Cyprus dental clinics overwhelmingly publish \"price on request\" and the two clinic price lists reached had their tables withdrawn. An eye examination can be reached through a GP referral to an ophthalmologist at the EUR 6 specialist co-payment, which is the one part of this line the public system does carry." }, "healthcare.statutory_employee": { "source": "GESY employee contribution 2.65% of emoluments under a ceiling of EUR 180,000 of total annual income per individual (GHS Law). Employer pays a further 2.90%. https://www.gesy.org.cy/ ; PwC Worldwide Tax Summaries Cyprus, Individual, Other taxes https://taxsummaries.pwc.com/cyprus/individual/other-taxes", "last_updated": "2026-09-13", "confidence": "high", "note": "DISPLAY-ONLY: the money is already inside taxes.payroll, so it must never be added to total_costs. computeCyprusTaxes charges the same amount from tax.gesy_rate and tax.gesy_cap, and tests/integration/tax-cyprus.test.mjs asserts this block and those two tax keys agree, so the figure shown and the figure charged cannot drift apart. Maximum EUR 4,770 a year. Because the ceiling is on TOTAL income rather than employment income, and is not indexed, it never binds on a salary alone at any income we model." }, "childcare.free_tier_start_age": { "source": "Cyprus has NO universal public childcare below pre-primary. The free public pre-primary year is compulsory from 4 years 8 months in 2026/27, stepping down toward 4 years 0 months by 2032/33, and it runs 07:45 to 13:05 Monday to Friday. Eurydice Cyprus, Early childhood education and care https://eurydice.eacea.ec.europa.eu/eurypedia/cyprus/early-childhood-education-and-care", "last_updated": "2026-09-13", "confidence": "medium", "note": "Stored as 5 rather than 4 for two reasons, both conservative: the entitlement starts PART WAY THROUGH age four rather than at the birthday, and it is a HALF DAY ending at 13:05, so a two-earner household still buys afternoon cover and preschool_monthly stays positive rather than dropping to zero. The national under-four fee subsidy covering up to 80% of fees is gated on drawing Cyprus child benefit, which our reader household income fails, so subsidy_per_kid_annual is 0." }, "govt_benefits.child_allowance_per_kid_annual": { "source": "Cyprus child benefit is INCOME-TESTED on the gross annual income of all family members in the previous calendar year. Thresholds: EUR 49,000 for one dependent child, EUR 59,000 for two, plus EUR 5,000 for each further child. European Commission, Your social security rights in Cyprus, 2025, pp.7-8 https://ec.europa.eu/social/BlobServlet?docId=13744&langId=en ; OECD, Tax and benefit policy descriptions for Cyprus 2024, section 4.1 https://www.oecd.org/content/dam/oecd/en/topics/policy-sub-issues/incomes-support-redistribution-and-work-incentives/TaxBEN-Cyprus-latest.pdf", "last_updated": "2026-09-13", "confidence": "high", "note": "ZERO for our reader, and three independent gates each produce that answer. The income test: the household band this product models starts at EUR 60,000, above the two-child threshold, and the official amounts table prints a literal dash in the one-child column from EUR 49,000.01 upward. Residence: five years of legal continuous residence in the government-controlled areas is required before any claim, which a mover fails outright. Assets: over EUR 1.2 million disqualifies. The only non-zero corner inside the modelled band is three children between EUR 60,000 and 64,000, or four between 60,000 and 69,000. A Cabinet-approved reform of 2026-09-10 would raise the three-child threshold to EUR 90,000 and the four-child to EUR 100,000 from January 2027, but it still needs a House vote and is deliberately NOT priced in. Separately, the childbirth grant (EUR 1,059.40 for a first child, rising by birth order) IS contributory rather than income-tested so an insured household does receive it, but it is one-off and belongs in no annual figure." }, "non_cash.vacation_days_statutory": { "source": "Annual Holidays with Pay Law 8/1967 as amended: the statutory minimum is four weeks, expressed as 20 WORKING DAYS on a five-day week and 24 working days on a six-day week, exclusive of public holidays, with the full entitlement conditional on 48 weeks worked in the leave year and pro rata below that.", "last_updated": "2026-09-13", "confidence": "high", "note": "UNIT CHECKED: the 20 and the 24 are the same four weeks counted against a different weekly divisor, so there is no six-day-week inflation here of the kind Norwegian virkedager produce. Public holidays are explicitly EXCLUDED and sit in their own field. The practical private-sector floor is a day higher, because an employer exempted from the Central Holiday Fund must grant leave at least as good as the statutory minimum, but the stored value stays the legal floor of 20 per the field definition." }, "non_cash.public_holidays": { "source": "Central Bank of Cyprus 2026 public-holiday notice read against the Council of Ministers list: 15 days, being 10 fixed dates and 5 Orthodox-Easter-linked days. The rule list and the substitution finding live in data/_holiday_rules.json.", "last_updated": "2026-09-13", "confidence": "high", "note": "NO substitute weekday when a holiday falls at the weekend, proved from the notice itself, which flags 15 August and 26 December as Saturdays in 2026 and declares no replacement for either. 13 of the 15 therefore land on a weekday in 2026. CYPRUS IS THE ONE COUNTRY IN THIS FIELD WHOSE VALUE IS OBSERVED PRACTICE RATHER THAN A STATUTORY FLOOR, and that is deliberate: public holidays are compulsory only for retail workers under the Shops Law closure regime, so for an office employee the legal floor is ZERO and storing zero would be absurd. The declared list binds the public sector and in practice reaches white-collar contracts nearly in full. Easter Tuesday is the day most often absent from private contracts." }, "non_cash.sick_leave": { "source": "Social Insurance Law N. 59(I)/2010, Fourth Schedule (benefit rate table), consolidated text on CyLaw http://www.cylaw.org/nomoi/enop/non-ind/2010_1_59/full.html ; European Commission, Your social security rights in Cyprus, 2025, p.25 https://ec.europa.eu/social/BlobServlet?docId=13744&langId=en ; OECD, Tax and benefit policy descriptions for Cyprus 2024, sections 2.1.3 and 2.1.7, for the six-day benefit week https://www.oecd.org/content/dam/oecd/en/topics/policy-sub-issues/incomes-support-redistribution-and-work-incentives/TaxBEN-Cyprus-latest.pdf ; basic insurable earnings 2026 from the MLSI Social Insurance Services table", "last_updated": "2026-09-13", "confidence": "high", "note": "THE UNIT IS THE FINDING. The statute says 156 days, but the Cyprus benefit week is a SIX-day week, so 156 days is 26 weeks, and the stored duration is written in weeks to stop a reader converting at five or seven days. Reading 156 as calendar days understates by 15%, as five-day working days overstates by 20%. There is NO employer obligation at all in Cyprus law: the benefit comes from the Social Insurance Fund and is not paid while the employer pays the full wage. The supplementary component is capped at one week of basic insurable earnings by Part II of the Fourth Schedule, a sentence deliberately ABSENT from the maternity part, which is why the 2026 sickness ceiling is EUR 353.22 a week against a maternity ceiling of EUR 954.00. Extension from 156 to 312 days is discretionary, so only the 26-week floor is stored." }, "non_cash.parental_leave_weeks": { "source": "Social Insurance Law N. 59(I)/2010, art. 29 (maternity allowance, 22 consecutive weeks for a first or second child, 26 for a third or later, and art. 29(2) defines the week as Monday to Sunday) and art. 29B (parental leave allowance, 8 paid weeks for a first child rising to 10, 12 and 14 by birth order), consolidated text on CyLaw http://www.cylaw.org/nomoi/enop/non-ind/2010_1_59/full.html ; Leave (Paternity, Parental, Carer, Force Majeure) and Flexible Working Arrangements Law N. 216(I)/2022, gazetted 16 December 2022, transposing Directive (EU) 2019/1158", "last_updated": "2026-09-13", "confidence": "medium", "note": "Household-inclusive basis: maternity 22 plus paternity 2 plus 8 PAID parental weeks for each parent, non-transferable, so 40 weeks around a first child. The 10 further parental weeks each parent holds are UNPAID job-protected and are excluded by the field definition. TWO CORRECTIONS TO THE EC 2025 GUIDE, which is stale twice over: the third-child maternity increment is +4 weeks, not +8 (Law 14(I)/2024, in force 1 March 2024, which also raised the base from 18 to 22), and the paid parental entitlement is 8, 10, 12 and 14 weeks by birth order, not a flat 8 (Law 157(I)/2024, in force 31 March 2025). Confidence is medium on one point only: whether the 8 paid parental weeks are per PARENT or per CHILD could not be closed beyond the statute non-transferability language, and a shared pot would make this 32 rather than 40. A HARD THROTTLE of 5 parental weeks per calendar year also means the 8 cannot all be drawn in one year." }, "non_cash.parental_leave_paid_pct": { "source": "Social Insurance Law N. 59(I)/2010, Fourth Schedule: maternity and paternity allowance 72% of the weekly value of the basic insurance points plus 72% of the supplementary; parental leave allowance 72% basic but only 60% supplementary. Basic insurable earnings 2026 EUR 220.76 a week, maximum insurable earnings EUR 1,325 a week, so a high earner holds one basic point and five supplementary.", "last_updated": "2026-09-13", "confidence": "medium", "note": "STORED AS 0, AND THAT IS CORRECT rather than a gap: parental_leave_capped covers all 40 weeks, so no percentage-paid weeks are left for this field to describe, and the schema requires 0 once the slices cover the entitlement. The headline rate lives in parental_leave_capped.pct. For the record, a duration-weighted headline would be 67: 24 weeks of maternity and paternity at 72% and 16 parental weeks at the 60% supplementary rate that actually binds a high earner. THE CEILINGS ARE THE REAL STORY and there are TWO of them: EUR 954.00 a week for maternity and paternity, which equals 0.72 x 1,325 to the cent and is the independent check that the maternity supplementary component carries NO sub-cap, and EUR 821.49 for parental leave. The engine carries one capped slice, so parental_leave_capped applies the LOWER of the two to all 40 weeks, which understates the maternity ceiling and is the conservative direction. A caveat that belongs on any Cyprus page: every allowance is computed on the relevant contribution year, one to two years back, so a household that has just moved to Cyprus receives far less than these ceilings imply." }, "non_cash.parental_leave_capped": { "source": "Social Insurance Law N. 59(I)/2010, Fourth Schedule, priced against the 2026 insurable-earnings parameters: basic insurable earnings EUR 220.76 a week and maximum insurable earnings EUR 1,325 a week (MLSI Social Insurance Services). Maternity and paternity ceiling 0.72 x 220.76 + 0.72 x 1,104.24 = EUR 954.00 a week; parental ceiling 0.72 x 220.76 + 0.60 x 1,104.24 = EUR 821.49 a week.", "last_updated": "2026-09-13", "confidence": "medium", "note": "CYPRUS HAS TWO STATUTORY WEEKLY CEILINGS and the engine carries one capped slice, so this block applies the LOWER of them, the parental-leave ceiling of EUR 821.49, to all 40 weeks. That understates the 24 maternity and paternity weeks, whose own ceiling is EUR 954.00, and understating a benefit is the conservative direction. The alternative shape, a 24-week slice at 954 with the 16 parental weeks left uncapped at a percentage, was rejected because it would pay a EUR 100,000 earner 60% of salary for those 16 weeks against a real ceiling of 42.7%, which flatters Cyprus by far more than this understates it. The EUR 954.00 figure is itself the proof that maternity carries NO supplementary sub-cap: it equals 0.72 x the maximum weekly insurable earnings to the cent, and the capping sentence that produces the much lower sickness ceiling of EUR 353.22 appears in Part II of the Fourth Schedule and is absent from Part III. Realised replacement rates on the maternity ceiling: 72% at EUR 68,904 of gross, 62% at 80,000, 49.6% at 100,000, 41.3% at 120,000. A 5-weeks-per-calendar-year throttle applies to the parental slice, and every allowance is computed on the relevant contribution year one to two years back, so a household that has just moved to Cyprus receives materially less than any of this implies." }, "tax.property_tax_rate_pct": { "source": "DLGOL Larnaca sewerage levy 0.76 per mille (0.60 wastewater plus 0.16 stormwater) of the latest Land Registry general assessment, per the official table at eoal.org.cy, giving about EUR 161 a year on the modelled EUR 235,000 two-bedroom; municipality immovable property levy 1 to 2 per mille of the 1 January 1980 value, about EUR 27; refuse collection EUR 100 to 250 flat, taken at EUR 200. Total about EUR 388 a year, or 0.165% of market value.", "last_updated": "2026-09-13", "confidence": "low", "note": "UNRESOLVED CONFLICT, recorded rather than papered over: a secondary source claims the Larnaca sewerage rate is 0.32% of 2018 values, which is 4.2 times the official table and would put the all-in figure at 0.38%. We ship the OFFICIAL table 0.17% because an official tariff beats a secondary summary, and because 0.38% would make a smaller, cheaper city carry more than double the Nicosia burden, which is implausible. Needs confirming with DLGOL. Cyprus ABOLISHED the national Immovable Property Tax from 1 January 2017, so the 1980-value 0.6 to 1.9% rates still widely quoted are dead; only the small municipality levy still uses 1980 values, and the sewerage board uses the current assessment. The sewerage usage charge per cubic metre is counted once, inside utilities." }, "housing.median_rent_by_bedroom": { "source": "bazaraki.com Larnaca-district apartment rental listings, bedroom-filtered and scraped 2026-09-13 (studio about 9 listings, one-bedroom about 64, two-bedroom about 61, three-bedroom about 62), cross-checked against Numbeo Larnaca (275 entries from 27 contributors, updated 5 September 2026) and the Council for the Registration of Real Estate Agents 2026 city figures of EUR 500 to 600 for a one-bedroom and EUR 700 to 800 for a two-bedroom.", "last_updated": "2026-09-13", "confidence": "medium", "note": "Asking rents, predominantly furnished. A SCRAPE TRAP worth carrying forward: bazaraki accepts a number-of-bedrooms query parameter and SILENTLY IGNORES it, returning the unfiltered page; it was caught because one-, two- and three-bedroom results came back byte-identical at a EUR 1,000 median. The parameter that works is attrs__number-of-bedrooms. Any future Cyprus rent scrape must assert that the distributions actually DIFFER by bedroom count. Our medians sit above the regulator city bands, which is the expected direction for a listing sample against a market read, and the regulator own figures are what prove Larnaca is separately measured rather than a Limassol discount: it prints Larnaca and Paphos with different one-bedroom ranges." }, "housing.rent_percentiles": { "source": "Observed quartiles of the bazaraki.com Larnaca-district two-bedroom asking-rent sample scraped 2026-09-13, about 61 listings.", "last_updated": "2026-09-13", "confidence": "medium", "note": "The one Cyprus city where the percentile pair is OBSERVED rather than derived from district bands, because the listing sample was large enough to take quartiles off directly. Band edge equals median_rent_by_bedroom times the multiplier, computed at display time." }, "housing.median_home_price_by_bedroom": { "source": "bazaraki.com Larnaca-district sale listings, bedroom-filtered, scraped 2026-09-13, against the Landbank Analytics H1 2026 study reported by cbn.com.cy, which prices the Larnaca apartment market on its own base of 673 sales across Aradippou, Larnaca city and Livadia (84.7% of district apartment sales) with apartment prices up 12.2% year on year.", "last_updated": "2026-09-13", "confidence": "medium", "note": "Larnaca has its own transaction base, which is the second of the five independence tests it passed. CYSTAT and the Central Bank publish a house price index per district as well, for Larnaca and Paphos both, so there is an official per-district series behind this. Asking-based per-bedroom medians on a transaction-anchored level." }, "housing.utilities_monthly": { "source": "EAC all-in domestic rate of 25 to 32 cents per kWh giving EUR 900 to 1,400 a year for a two-bedroom, plus the official DLGOL Larnaca water and sewerage tariff. Build: electricity EUR 108, water and sewerage EUR 30, heating allowance EUR 15. The Numbeo Larnaca 85 square metre figure is EUR 162.52.", "last_updated": "2026-09-13", "confidence": "medium", "note": "ANNUAL AVERAGE INCLUDING THE SUMMER COOLING LOAD. Larnaca is coastal, so its cooling load sits below inland Nicosia and its utilities line is the lowest of the three. The EUR 15 heating allowance is the weakest component: no Cyprus apartment heating-spend figure was found, and most heating is the same reverse-cycle air conditioning already inside the electricity line." }, "housing.internet_phone_monthly": { "source": "Numbeo Larnaca fixed broadband and mobile lines, with Cyta, Epic and PrimeTel as the operators of scale.", "last_updated": "2026-09-13", "confidence": "low", "note": "Cyprus telecom pricing is national rather than city-level, so this tracks the Nicosia and Limassol figures. No operator tariff page was fetched." }, "housing.renters_insurance_monthly": { "source": "Cyprus market bands: renters contents cover EUR 80 to 130 a year for a EUR 30,000 sum insured; an apartment buildings policy at EUR 220 to 350 a year over the modelled EUR 235,000 two-bedroom.", "last_updated": "2026-09-13", "confidence": "low", "note": "NOT the 0.25% rule of thumb that circulates, which is on REBUILD value rather than market value. Both figures are reasoned estimates with no insurer rate card behind them." }, "childcare.preschool_monthly": { "source": "Larnaca private nursery and kindergarten band EUR 300 to 600 a month, accepting children from three months (dom.com.cy Larnaca guide); Numbeo Larnaca private full-day preschool EUR 340.67; after-school care EUR 100 to 250 a month. Eurydice Cyprus for the free pre-primary year.", "last_updated": "2026-09-13", "confidence": "low", "note": "Set above the Numbeo midpoint because only about 4 of roughly 30 Larnaca private preschools are English-speaking, so a relocating household real choice set is the pricier end. Public pre-primary is free and compulsory from 4 years 8 months but runs 07:45 to 13:05, a HALF DAY, and children between 3 and the compulsory age pay EUR 40 to 45 a month basic or EUR 70 to 130 extended. No Larnaca setting publishes an infant-specific tariff, so the infant figure applies an age premium to the city band." }, "food._summary": { "source": "Single-person grocery bands of EUR 250 to 400, 300 to 450 and 250 to 350 across index.cy, expats.cy and cyprustaxlife, with Numbeo Larnaca dining prices (inexpensive meal EUR 12, mid-range cover for two EUR 50) and a Numbeo single-person basket excluding rent of EUR 780.20 a month.", "last_updated": "2026-09-13", "confidence": "low", "note": "The per-child figure is derived by netting a couple (EUR 300 to 500) out of a family of three (EUR 800 to 1,200), not observed. Food CPI rose 6.16% year on year to March 2026, so this group ages fast. The Numbeo Larnaca panel is 275 entries from 27 contributors, about a third of the other two Cyprus cities, so it is directional and must not adjudicate a value on its own." }, "transit._summary": { "source": "Cyprus Public Transport current urban tariff https://www.publictransport.com.cy/cms/page/cash-tickets after the indexed increase of 3 August 2026: 30-day pass EUR 93.60. Car cost assembled: depreciation 2,000 on a EUR 29,400 VW Golf at eight years and 50% residual, insurance 550, road tax 220, MOT 35, fuel 1,250 at EUR 1.44 a litre over about 12,000 km, maintenance 500, parking 200.", "last_updated": "2026-09-13", "confidence": "medium", "note": "LARNACA BUS PASS IS THE MOST EXPENSIVE IN CYPRUS, which inverts the intuition that the cheap city is cheap on everything: Larnaca shares Cyprus Public Transport with Nicosia and took the same 3 August 2026 contractual increase, while separately operated Limassol and Paphos did not. Validated because the CPT single and night fares match the post-increase press figures digit for digit and the 7-day fare of EUR 46.80 is exactly half. REJECTED: EUR 40, a stale Zinonas table, and EUR 50, which is the Paphos operator and also the contaminated Numbeo Larnaca entry. The Cyprus sources quoting EUR 1,200 to 2,000 a year for a car exclude depreciation, fuel and parking." }, "discretionary._summary": { "source": "A sourced EUR 170 to 250 a month Cyprus social-life band plus Numbeo Larnaca gym EUR 59.17 and cinema EUR 9.50; the domestic-holiday line built from Troodos guesthouses EUR 35 to 70 a room, mid-range hotels EUR 60 to 110 a night and a mid-range EUR 130 a day spend.", "last_updated": "2026-09-13", "confidence": "low", "note": "The domestic-holiday figure is the weakest value in the Larnaca file: no Cyprus domestic-tourism per-resident spending series exists at all. Set below a mainland peer because the island is day-trippable end to end." }, "travel_home.avg_round_trip_per_person_to_other_city": { "source": "Momondo Seattle to Larnaca dated 2026-09-07: cheapest USD 822, typical range USD 1,152 to 1,295, no direct service, cheapest itineraries one stop via London Heathrow. Midpoint USD 1,223.50 converted at the 0.863 rate in build/fx-rates.json.", "last_updated": "2026-09-13", "confidence": "medium", "note": "The best-evidenced of the three Cyprus airfare figures, because Larnaca is the airport: it uses a published TYPICAL RANGE rather than a lead-in fare. No direct Cyprus to Seattle service exists." }, "compensation_index": { "source": "CyprusWork Larnaca software engineer posting EUR 25,000 to 35,000 gross; a Larnaca AI startup full-stack posting EUR 45,000 to 65,000; Glassdoor Larnaca software engineer EUR 41,674 to 60,139; levels.fyi Larnaca 25th percentile EUR 33,065; HF Markets (HFM), a forex and CFD broker HEADQUARTERED in Larnaca, carrying 46 Glassdoor salary reports across 35 job titles at that office. CYSTAT average gross EUR 2,601 a month in Q1 2026.", "last_updated": "2026-09-13", "confidence": "medium", "note": "Implied Larnaca total comp of about EUR 39,000 on the dataset basis (0.18 times EUR 216,732). CITY-SPECIFIC AND NOT A LIMASSOL DISCOUNT, which was the test this city had to pass to be added at all: the figure is built bottom-up from Larnaca job ads, the Glassdoor Larnaca office page and the levels.fyi Larnaca 25th percentile, converging on an all-levels median of EUR 37,000 to 42,000, with the lower half taken. The Limassol ratio appears only as a corroborating check and lands in the same place. THE ANCHOR RULE APPLIED LITERALLY WAS REJECTED: levels.fyi Larnaca at EUR 55,602 gives 0.258, which would rank Larnaca above Madrid, and the same source Cyprus national median of EUR 72,536 is 2.3 times the CYSTAT actual average gross because its sample is expat-package employers. 0.18 slots between Milan at 0.17 and Porto at 0.19. Re-derive if the peer scale is ever rebased." } } }, "_disclaimer": "Values are approximations gathered from public sources as of last_updated. Re-verify before any production use. Tax brackets and statutory amounts change annually.", "behaviour_claims_use_the_phrasebook": "A note that asserts engine BEHAVIOUR (not just a source citation) is machine-checked by tests/meta-claims.check.mjs (npm run meta:check), and only if it uses one of four phrases: 'no schema field for X' / 'X is not a field' (checked: key absent from every city JSON), 'held at 0' / 'zeroed' / 'we store 0' (checked: stored value is 0), 'NOT MODELED' / 'not applied' (checked, for regime-style claims on non_cash._summary: compute() does not move when the block is toggled; other shapes are catalogued, not asserted), 'specced, not fixed here' / 'TODO' (surfaced as a dated to-do). The vocabulary is closed on purpose: a handful of regexes, not language understanding. A note that describes the same limitation in different words is simply not checked, so write behaviour claims IN these words. Design + every finding from the first run: docs/superpowers/research/2026-08-06-meta-claims-guard.md." }