The same retiree, in 14 countries: what the tax system actually costs

Take one retiree — age 65, needing 40,000 a year after tax until 95, with the same portfolio shape — and run them through 14 real tax engines. Not fourteen articles about tax: fourteen projections, by the same engine this site's calculator uses. The capital they need ranges from 21.00× to 23.00× their annual spending, a spread of 9.5%. Then we add each country's own state pension, and the ranking turns upside down.

How to read this, before you read it

Cross-currency comparisons are where this kind of table usually starts lying, so here is the basis, stated once and applied everywhere:

• Same nominal amounts, each in local currency. The American needs $40,000 a year, the Briton £40,000, a Dubai resident AED 40,000. No exchange rate is applied anywhere on this page. That isolates the tax system — the thing the engines actually know — instead of blending it with the cost of living and an FX rate that would be stale within weeks.
• The ranking metric is a multiple, not an amount. Capital required ÷ annual spending. A multiple has no currency in it, so it compares cleanly across all 14, and the "vs US" column is a percentage for the same reason — BHD minus USD is not a number in any currency, so this page never prints that subtraction. The local-currency column is the concrete illustration, labelled with each country's own currency code from the app's country registry.
• Every figure is computed at build time. Nothing on this page is copied from a tax table. Each row is a full year-by-year projection to 95 — income tax, capital gains, wrapper rules, withdrawal ordering — by the engine documented here.

Ranked: capital needed to fund 40,000 a year to 95

Lower is better. The retiree holds 60% in an ordinary taxable account (40% of it unrealised gain), 30% tax-deferred and 10% tax-free — and where a country has no such wrapper, the engine folds that balance into the ordinary account, so the starting pot is identical in every run. No state pension is counted in this table (that comes next).

Three things in that table are worth more than its ordering. First, the whole spread is only 9.5% — a retiree at this spending level pays so little tax in the United States ($39,004 across thirty years, an effective 2.4% of everything withdrawn) that a zero-tax jurisdiction improves on it by just 1.2%. Second, United Kingdom is the outlier, needing £920,000 and paying £169,453 in lifetime tax — 9.7% of withdrawals, four times the US rate — because the personal allowance is frozen in cash terms until 2031 and 75% of every pension withdrawal is taxable income. Third, New Zealand pays $107,329 despite having no capital gains tax at all, which is the most counter-intuitive number here and worth its own section. Hong Kong, meanwhile, lands on exactly the zero-tax group's figure — because for a retiree living off investments it genuinely is one.

Why New Zealand taxes a retiree with no capital gains tax

New Zealand genuinely has no CGT. Every withdrawal in the New Zealand run is untaxed: net equals gross, exactly. And yet the run still pays $107,329 of lifetime tax and needs 6.0% more capital than a zero-tax jurisdiction — 22.25× spending against 21.00×. The reason is that New Zealand taxes the balance rather than the sale. A multi-rate PIE — which is what KiwiSaver funds and mainstream managed funds are — pays tax every year on its investment income, and a fund holding international shares is taxed on a deemed 5% of opening market value under the fair dividend rate whether it distributed anything or not. At the top prescribed investor rate that is a standing drag of about 1.4% of the balance, every year, inside KiwiSaver as well as outside it. Hold NZ or Australian shares directly instead and the annual tax disappears — the engine models that as a switch, and it moves the capital required from $890,000 to $840,000.

This is the general lesson of the table: where a system taxes matters more than whether it calls itself low-tax. New Zealand taxes the balance annually. Hong Kong taxes salaries but not investment income at all, so an ordinary Hong Kong brokerage account behaves like a tax-free wrapper without being one. The United States taxes realised gains, and a retiree with modest income realises them inside the 0% bracket.

Does the ranking hold at other spending levels?

Partly. Progressive tax systems bite harder as spending rises, so the ordering of the five taxed systems shifts while the zero-tax group's result does not move at all — a flat zero is scale-invariant. Capital required (and the multiple beneath it) at four spending levels, in local units:

At 25,000 a year the gap between the cheapest and dearest system is 2.00× of annual spending; at 150,000 a year it is 4.33× — more than double. Tax residence is a lever that grows with your spending, which is why it matters much more to a fat-FIRE plan than to a lean one — and why a single ranking, including this one, is only true at the spending level it was computed for.

The number that inverts everything: the state pension

The table above deliberately counts no public retirement benefit, so that it measures tax and nothing else. That exclusion is not neutral. Some of these countries hand a retiree a large, inflation-linked, lifelong income; others hand them nothing whatsoever. Adding each country's own benefit — at its own statutory rate, from its own engine — produces this:

🇳🇿 New Zealand needs $260,000 — 6.50× spending — because NZ Super alone covers 84% of the target spending, is flat-rate and universal, and requires no contribution record at all. 🇦🇪 United Arab Emirates stays at 21.00×, because there is no benefit to add. Comparing those two multiples — which is legitimate, since both fund the identical 40,000 of annual spending — the zero-tax retiree needs 3.2 times the capital for the same life. Put the two findings side by side: the tax systems differ by 9.5%; the safety nets differ by 223%.

Stated plainly: a zero-tax jurisdiction gives you the highest possible savings rate and the weakest possible safety net. Both halves are true, and only one of them shows up in a tax table. The trade is real either way — the Gulf saver keeps every unit they earn and must fund the whole of old age from the portfolio they built; the New Zealander or Canadian pays tax for decades and is handed a floor they cannot outlive. Which is better depends on how much you earn, how long you stay, and whether you are still resident when you retire.

Benefit rates come from each engine's own constants: the New State Pension at 12,548/yr for 35 qualifying NI years, NZ Super at the single-living-alone rate from 65, and maximum CPP at 65 plus OAS (which is clawed back above net income of 95,323). Social Security is the one benefit here that depends on an earnings record rather than a flat rate, so its row is explicitly illustrative — estimate yours rather than borrowing this figure.

Inside a country: the sub-national spread

Three of these countries tax below the national level, and the comparison above picks the low-tax option where one exists — Texas for the United States and England for the United Kingdom, neither of which levies a sub-national income tax on this retiree. Canada has no zero-tax province, so Ontario is used. At 150,000 a year of spending, the internal spread is:

The other 11 countries have no sub-national income tax to choose — and that is "does not exist", not "not modelled": New Zealand, Hong Kong and all nine zero-tax jurisdictions levy no state, provincial, emirate or city income tax whatsoever, so the app does not ask.

What this comparison does not tell you

The ranking above is a ranking of tax systems, computed by running one plan through fourteen engines. It is not a ranking of places to live, and it deliberately holds one thing constant that in real life is not: the amounts are the same nominal figures in each local currency. No exchange rate is applied anywhere on this page.

• No cost-of-living adjustment. Spending 40,000 a year buys wildly different lives in Monaco, Riyadh and Auckland. Import duty makes Cayman expensive; Gulf housing costs swing with the city. This model takes your spending figure as given — it does not tell you what that figure needs to be.
• No exchange rates, and no FX risk. Because no rate is applied, nothing here goes stale — but nothing here converts, either. If you will earn in one currency and spend in another, that mismatch is a real risk this page does not model.
• Currency size affects the taxed systems. Tax bands are set in local units, so the five countries with income tax are being compared at genuinely different real spending levels. That is precisely why the sensitivity table above exists: the zero-tax result holds at every level, the ordering of the taxed systems does not.
• No healthcare, and it is not a detail. The engine's own modelled out-of-pocket medical shock floor ranges from £3,000 where a public system absorbs catastrophic care to $12,000 in the United States — and in the Gulf and the Caribbean an expatriate is in a private market with no public safety net at all.
• No visas, residency rights, property costs or inheritance tax. Several of these jurisdictions require an employer sponsor or a substantial investment to live in at all, and a residence permit is not a retirement plan.
• Residence rules decide where you are taxed — not preference. You do not choose your tax system by reading a table. You become tax-resident somewhere by living there under its rules, and leaving a system is often harder than entering one.

Five caveats that change the answer for real people

• Monaco: French nationals are excluded. Monaco's exemption explicitly does not cover French nationals, who are governed by the Franco-Monegasque Convention of 18 May 1963. A French national who took up residence in Monaco on or after 13 October 1957 remains subject to French income tax as though still resident in France. Two narrow groups escape it — those resident before 13 October 1957, and French nationals born in Monaco who have lived there continuously since birth. If you hold French nationality, the Monaco row does not describe your position.
• Oman is deliberately not in this table. Oman has legislated a personal income tax effective 1 January 2028 — a flat 5% on income above OMR 42,000, reaching residents and expatriates alike. A retirement plan runs thirty to fifty years, so listing Oman as tax-free would be wrong for all but the first two years of yours. It belongs in a progressive-tax engine, not this one.
• Hong Kong has no state pension at all. Not a small one — none. The Old Age Living Allowance is means-tested social assistance with asset limits any retirement plan on this site would fail by an order of magnitude, so the engine does not model it as a retirement benefit. Hong Kong's low tax bill and its absent safety net are the same fact seen twice.
• Scotland sets its own income tax rates inside the UK. At 40,000 a year of spending the choice makes no difference to the capital required, because the retiree's taxable income never reaches the point where Scottish and rUK rates diverge. At 150,000 a year it costs a Scottish taxpayer 50,000 more capital. Capital Gains Tax is reserved to Westminster and is identical either way.
• The nine zero-tax jurisdictions are one row, not nine. They return identical figures because they share one projection engine — and they share one engine because they share one rule: there is no income tax, no capital gains tax, and no sub-national tax in any of them, so gross equals net and there is nothing left to differ about. What genuinely differs between them is not tax: it is the end-of-service entitlement, social insurance, healthcare exposure and cost of living.

Assumptions, in full: age 65 at retirement, horizon 95, single filer, 6% nominal return, 3% inflation, 60%/30%/10% ordinary/deferred/tax-free with 40% unrealised gain in the ordinary account, no employment income, no housing, no healthcare cost loaded, 40,000 local units a year of after-tax spending. "Capital needed" is the smallest starting pot (to the nearest 10,000) that funds that spending every single year to 95. Educational projections, not tax advice — and not a recommendation to move anywhere.

Common questions

Keep exploring