The Safety-Net Index 2026

By Skyler Bissell · Updated August 2026 · 178 cities in 70 countries

The Safety-Net Index ranks 178 cities by how much of family life the state pays for. Its measure is the safety-net gap: what a family of four still pays out of its own paycheck after the state has done its part, over and above a single earner on the same income. The cityparity engine computes it from official 2026 tax schedules, childcare fees, healthcare costs, and benefit rules in 70 countries, the same engine behind the calculator. This is what pricing the safety net means as a dollar figure.

A small gap means a strong net. Vienna's gap is $6,922 a year; New York's is $73,077, 26.7% of household gross. Measured at each city's own top-decile income, the gap runs from 5.6% in Busan to 31.1% in Perth across the advanced economies, and the table below starts where the net is strongest.

The ranking: the safety-net gap in the advanced economies

The 116 cities in the IMF's advanced-economy grouping first, smallest safety-net gap on top, soft income anchors flagged †. The full 178-city table is in the appendix. Each row: the gap, then how it builds: extra tax on the second income, childcare, extra healthcare, and the child benefit paid back.

The prose on this page describes the default view: the gap at the high (p90) anchor. Switching to the median re-ranks both tables and the text stays put; the median comes straight from each distribution's stored value, so it is exact the same way p90 is. The gap always uses both households, because it is their difference; the single earner's own wedge, at both anchors, ships as columns in the CSV.

Once the single earner's wedge is subtracted, high-tax and low-tax systems finish side by side. Sweden's top rates and Korea's moderate ones both land under 7% of gross, because both fund childcare and pay cash benefits. The bottom eight places belong to the US, UK, Australia, Switzerland, and Ireland, where a second income buys market-rate care with money that was taxed on the way in. Every one of the 178 cities is ranked the same way in the full table at the foot of this page.

What each household hands over

Each row prices one top-decile income twice. Amber marks what a single earner hands over, blue what the family of four hands over, each as a share of its own gross. Across most of Europe the blue dot sits left of the amber one, because the system takes a smaller share from the family. In every US, UK, Swiss, and Australian city it sits right, by 5 to 21 points of gross. The two dots divide by different incomes, so the distance between them measures something other than the safety-net gap. Vienna’s family carries the lighter share here and still pays $6,922 more.

Each dot is that household's safety-net wedge as a share of its own gross income at the city's p90 anchor; the family's income is 1.6x the single earner's, so the two shares have different denominators. † thin income anchor. Sixteen of 178 cities shown; every city, both households, and both anchors ship in the CSV.

New York's family wedge is the largest in the advanced world at 49.7% of gross, with London and Perth at 47.8%, Ljubljana at 46.2%, and Zurich and San Francisco at 46.1%. Three different systems arrive at the same place: London and Perth get there through childcare alone ($33,962 and $29,634 a year at the anchor income), Zurich through the dataset's most expensive childcare, $49,071, stacked on five-figure health premiums, and New York and San Francisco by charging on all three lines at once.

The ranking compares systems. Your decision has a salary and two cities in it already, so run your own numbers through the calculator.

What the table shows

Vienna and New York tax their top earners within 1.3 points of each other, and their safety-net gaps sit $66,155 apart. At Vienna's top-decile income the family of four pays $11,204 of tax on the second income and $3,792 for childcare, the kids cost nothing at the doctor, and Familienbeihilfe pays back $8,074 in cash: a $6,922 gap. At New York's, $28,739 of second-income tax, $38,400 of childcare, and a $10,338 family health bill stand against a $4,400 Child Tax Credit: $73,077. Vienna's income anchor carries a thin flag, so hold its exact rank loosely; the gap stays above $50,000 at either anchor, and at the median the pair reads $214 against $51,997. Berlin sits between them at $14,696 a year, nearly all of it the second earner's own tax: childcare costs $1,384 and Kindergeld pays back $7,170.

Across the advanced-economy cut, Korea and Japan price a family lowest. Busan's gap is $4,167 a year (5.6% of gross), Incheon's $4,250, Fukuoka's $4,195, with Vienna, Stockholm, Malmö, Oslo, and Tallinn inside the top ten. At the other end sit the English-speaking childcare markets: Perth's gap is 31.1% of gross and London's 30.6%, driven by childcare bills of $29,634 and $33,962 a year, ahead of Basel, Zurich, and New York, where five-figure childcare stacks on five-figure extra tax.

The gap is also why the Gulf does not lead this page. Abu Dhabi charges no income tax at all, yet a family of four there hands over $22,398 a year more than a single earner (11.1% of gross): full-price private childcare and healthcare, with no benefit paid back. A zero-tax system can still be an expensive place to raise kids, and the gap is the number that shows it. Zero tax on a single earner is invisible to this metric by construction; the total-wedge columns in the CSV still measure it.

Tax share against the full wedge

Put every city on two axes and the argument stops being about tax rates. Across the bottom: what a city charges a single top-decile earner in income tax. Up the side: the safety-net gap, what a family of four still pays once the state has done its part. Across the 116 advanced economies those two numbers are all but unrelated, at a correlation of −0.05. Berlin charges that single earner 40.7% and leaves a family a 9.0% gap. Zurich charges 26.2% and leaves 26.7%, three times Berlin’s, because care for two kids runs $1,384 a year in Berlin against $49,071 in Zurich and Kindergeld pays $7,170 of it back in cash.

Widen the chart to all 178 cities and the Gulf makes the same point from the other end. Abu Dhabi and Dubai charge no income tax at all, and a family there still pays $22,398 and $24,196 a year, because private childcare and private healthcare arrive at full price with nothing paid back. At that width a mild pattern does appear, a correlation of −0.43: the heaviest-taxing systems tend to leave the smallest family bills. Back in the advanced cut, London taxes a top earner 14 points more lightly than Berlin and carries a gap three times the size, on childcare alone. A tax rate on its own predicts almost nothing about what a family pays.

Vacation and parental leave stay visible

The wedge is a cash metric, and two of the largest safety-net benefits pay out in time, so they ride beside the ranking, unranked. The EU Working Time Directive guarantees 20 paid vacation days; US federal law guarantees zero, and the engine models a US norm of 26 paid days off including holidays against Oslo's 37 and Stockholm's 38. Parental leave runs 480 days per child in Sweden and 49 weeks at full pay in Norway, against a US federal floor of 12 unpaid weeks under the Family and Medical Leave Act. Both lines appear on every comparison the calculator runs, and the site ranks each by country: vacation days and parental leave.

Methodology

Two households anchor every row. One is a single earner at the city's all-earner 90th-percentile income, no kids. The other is a family of four on the same paycheck plus a second income at 60% of it, kids aged 3 and 6. Each gets the same wedge computation, and the gap is their difference. Rent, food, transit, and discretionary spending stay out on both sides. That exclusion is deliberate. A cheap city cannot climb this table on low rent, and an expensive one cannot fall on it; the only levers are tax, childcare, healthcare, and the child benefit, the four lines a government sets by policy.

How the gap is computed

Take a city. Compute the safety-net wedge for a single earner at the city's anchor income: tax plus out-of-pocket healthcare. Compute it again for the family of four living on the same anchor paycheck plus a second income at 60%, kids aged 3 and 6: tax on both incomes, childcare, family healthcare, minus the child benefit. The safety-net gap is the second number minus the first, in local currency (in earlier drafts this metric was called the family premium). The ranking divides it by the family's gross household income, so both sides of the fraction sit in the same currency and exchange rates cannot move a city's rank.

The denominator is the family's gross because that is the household paying the gap. Subtracting the two published wedge shares instead would divide by two different incomes (the family earns 1.6x the single) and understate the gap by about a third of the single wedge share. The CSV carries both the dollar gap and the share, so either form can be checked against the components.

The extra-tax line inside the gap is the household's tax at 1.6x the anchor minus the single earner's tax at the anchor: the tax on the partner's earnings net of every couple-and-kids relief in the code, joint brackets, income splitting, child credits. Filing systems show through it. Germany files the couple jointly, so Berlin's second income enters at the household's rate and adds $20,481; Australia files every earner alone, so Sydney's partner starts at the bottom brackets and adds $8,908. Berlin still finishes 91 places ahead of Sydney in the cut, because its childcare costs $1,384 a year against Sydney's $23,127 and Kindergeld hands back $7,170.

A gap can go negative where the state pays back more than the partner's tax and the family bills combined; Kraków does it at the median anchor. And because the gap subtracts the single wedge, the level of tax on a single earner drops out entirely. That is why the zero-tax Gulf sits mid-table on this metric while topping the raw wedge: a system that funds nothing for families is expensive for families at any tax rate.

The four lines of the wedge

• Tax is the engine's full figure for the household: income tax plus employee social contributions, filed the way each country's law files a two-earner couple (joint, individual, or split per the real rules). Where a country funds healthcare through earmarked payroll, that payroll is in this line. Where payroll is compulsory savings the household keeps, such as Singapore's CPF or Chile's AFP, it is also in this line and the row carries a § marker.
• Childcare is the market price for a 3-year-old in full-time care and a 6-year-old in after-school care, minus any subsidy that applies automatically to everyone: Berlin's free Kita, Vienna's free Kindergarten, Sweden's maxtaxa, Norway's maksimalpris, Canada's CWELCC program. Income-tested credits a family has to apply for and might miss are surfaced as notes and left out of the number.
• Healthcare is what the household pays itself: premiums, expected out-of-pocket costs, and adult dental. US figures follow the KFF Employer Health Benefits Survey and the Peterson-KFF Health System Tracker, with family out-of-pocket scaled per the Milliman Medical Index.
• Child benefit is cash the state pays a family at this income, subtracted from the wedge. Austria's Familienbeihilfe pays a two-child Vienna household about $8,074 a year; the US Child Tax Credit is worth $4,400.

How the anchor income is set

Every city is anchored at its own 90th percentile of all earners: the income a comparably successful professional there earns, resolved through the same income-distribution model as the calculator's attainability feature. A measured metro distribution wins where one exists (28 cities); the national distribution is the floor otherwise (150 cities). 68 cities, mostly markets with large informal sectors where a percentile is a soft read, carry anchors we rate as thin. Their rows stay in the table with a † and stay out of the headline claims.

Why the 90th percentile: the professional weighing a cross-border offer lands in the top decile of all earners in nearly every destination market, and p90 is the deepest point the underlying distributions measure without extrapolating. It also keeps continuity: the US all-earner p90 for a single earner sits near $155,000, so the US rows line up with the fixed-$150,000 portable-income columns in the CSV. The partner earns 60% of the primary, the site's standing two-earner household shape.

Held constant everywhere: two kids aged 3 and 6, retirement contributions at 0%, typical lifestyle, a transit pass, no trips home, age band 35 to 44. Housing, food, and the rest of the basket sit outside the wedge entirely.

The anchor toggle above the gap ranking swaps both tables to the medium (median) anchor. Every distribution in the model stores its median directly, measured percentile or published median, so the p50 anchor is read from the data with no interpolation and is exact the same way p90 is. The gap tables always use both households, because the gap is their difference. Both views run through the identical engine, and the CSV ships every variant this page computes: the gap, the family wedge, and the single earner's wedge, at both the top-decile and median anchors.

Disclosures

• Anchor incomes mix geographies and methods. 28 cities use measured metro distributions and 150 use the national floor, so same-country cities can share an anchor and differ only through local prices and sub-national tax. Most anchors are lognormal fits to published medians and top-decile points, exact at p90 by construction. Geography, method, and source basis are disclosed per row in the CSV.
• The ordering of close neighbors is soft. Cities within 2 to 3 points of each other can swap places under a different anchor percentile. Separations of 10 points and more survive any anchor choice we tested.
• Currency conversions drift. Dollar columns use the exchangerate-api.com snapshot of 2026-08-03. The ranking cannot move with them.
• The gap reflects state support and local market prices together. A city can post a small gap because the state covers the bills, or because market-rate childcare and healthcare are cheap there for reasons of its own. The component columns on this page and in the CSV keep the two visible.

Read the full methodology for how the engine solves each figure. Every numeric value on this page carries a source and date in the site's audit trail: official 2026 schedules from the IRS, HMRC, the Bundesfinanzministerium, Belastingdienst, Skatteverket, Skatteetaten, the ESTV, the CRA, the ATO, and their equivalents across the other countries.

Frequently asked questions

Appendix: all 178 cities

The full gap table, ranked ascending. † thin income anchor. § tax line includes compulsory savings the household keeps as assets. The Gulf's zero-tax cities land mid-table here, Abu Dhabi 31st at 11.1% of gross, and the deep end belongs to cities where market-rate childcare meets an emerging-market anchor income: Jakarta's gap alone exceeds the family's gross.