Will a G20 country implement a national UBI program before 2040?

TL;DR

I estimate a 14% chance that a G20 sovereign country or the EU implements a qualifying permanent national UBI by 2040-12-31. No eligible jurisdiction is close today, because the required benefit is roughly 11% to 21% of GDP even in the easiest rich-country cases and far higher in most emerging G20 members. The live YES path is an AI, resource-dividend, or social-crisis break that makes a very large universal cash entitlement politically normal in at least one rich or resource-heavy G20 state.

Context

The eligible set is the current 19 sovereign G20 countries plus an EU-wide program; the African Union is a G20 member but is excluded by the question, and the current G20 country list is Argentina, Australia, Brazil, Canada, China, France, Germany, India, Indonesia, Italy, Japan, Korea, Mexico, Russia, Saudi Arabia, South Africa, Türkiye, the United Kingdom, and the United States (Australian DFAT). The deadline is 2040-12-31, so the remaining window is about 14.4 years from the 2026-07-22 question date.

The resolution bar is much stricter than normal use of the term UBI. It requires payments to have started, to be nationwide or EU-wide, permanent, recurring, unconditional, non-means-tested, paid directly in cash to substantially all adults, and worth at least PPP US$1,000 per month in 2026 purchasing power. Most current basic-income-adjacent policies are pilots, local programs, targeted guaranteed minimum incomes, child benefits, vouchers, or one-off crisis transfers, so they do not count.

Evidence

The historical base rate is zero for the exact target. The World Bank's 2020 UBI review says no country then had a UBI in place and identifies only Mongolia and Iran as national UBI-like cases for a short period (World Bank). Stanford's Basic Income Lab map says very few experiments or programs capture all UBI design features: unconditional, universal, cash, individual, and regular (Stanford Basic Income Lab). The nearest durable analogue is the Alaska Permanent Fund Dividend, but it is subnational and annual (Alaska Department of Revenue); the nearest high-income-country national vote was Switzerland's 2016 referendum, which failed by 76.9% to 23.1% (Library of Congress).

The fiscal scale is the strongest NO evidence. I calculated the gross annual cost as US$12,000 times population aged 15+ divided by 2025 PPP GDP. The population proxy uses World Bank WDI age groups for ages 15-64 and 65+, so it slightly overstates true 18+ adult cost; GDP uses IMF WEO April 2026 PPP GDP in current international dollars (World Bank WDI population table, IMF WEO April 2026). The table is a gross cost screen, not a net budget score after taxes or benefit replacement.

For comparison, the OECD says public social expenditure is about one-fifth of GDP on average across OECD countries, with SOCX covering 1980 through 2021/2024 depending on series vintage (OECD SOCX). OECD microsimulation work also finds that a universal no-questions-asked transfer is not a cost-effective way to reduce poverty because it is not targeted on those most in need (OECD). This does not make UBI impossible, because taxes can claw back payments from high earners without violating the formal criterion, but it makes the required fiscal bargain much larger than most current proposals.

The current policy frontier is far below the line. South Korea is the strongest democratic signal: the government selected a rural basic-income pilot as a national policy task, and residents of 10 selected counties receive KRW 150,000 per month in local-currency vouchers during 2026-2027 (Korea.net). That is real experimentation, but it is local, temporary, voucher-based, and far below the threshold. Canada has active framework legislation for a guaranteed livable basic income, with Bill S-206 in the 45th Parliament, but it is a framework bill rather than an implemented universal payment scheme (Parliament of Canada). Canada's Parliamentary Budget Officer modeled a guaranteed basic income in 2025 with gross cost of C$112.4 billion under a nuclear-family definition and C$57.0 billion under an economic-family definition, but that design is income-tested and therefore fails this resolution (PBO Canada).

The EU-wide route is structurally weak. The Commission registered a European Citizens' Initiative called Introduction of Unconditional Basic Income throughout the EU on 2026-07-07, but registration only means legal admissibility and the organizers must collect one million valid signatures before the Commission has to respond (European Commission). The EU's own-resources ceiling for annual payment appropriations is 1.40% of EU GNI for 2021-2027, while a qualifying EU-wide adult UBI would be roughly 15.7% of EU PPP GDP on my 2025 screen (European Commission budget ceilings).

The most plausible non-democratic or resource path is Saudi Arabia, but the live program is not close. The Citizen Account Program is a cash-transfer system, but the Saudi national platform says it redirects benefits to eligible groups affected by reforms, and the original SPA policy note says entitlement varies with household income, dependents, and age (Saudi National Platform, Saudi Press Agency). Brazil is legally interesting because Law 10.835 of 2004 created a right to a citizen basic income for all Brazilian residents and long-term foreign residents, regardless of socioeconomic condition, but the executive sets the value subject to fiscal responsibility and no universal high-payment scheme has begun (Brazil Law 10.835/2004). Brazil's Supreme Court response focused implementation on people in poverty and extreme poverty, not a full universal adult benefit (STF).

The AI channel keeps the probability in the teens. IMF staff estimated in January 2024 that almost 40% of global employment is exposed to AI and about 60% of jobs in advanced economies may be affected (IMF). The ILO's 2025 update says one in four jobs worldwide has some generative-AI exposure, but that transformation rather than full redundancy is the most likely outcome (ILO). This points to policy pressure, not automatic mass unemployment. In the United States, the Sanders AI sovereign wealth fund bill introduced in June 2026 framed AI rents as a public dividend, but the cited dividend was more than US$1,000 per year, not per month (Sanders Senate office). The U.S. BOOST Act proposal in the 119th Congress would pay US$250 per month to qualified adults aged 19-67, also below the threshold and only introduced (Congress.gov). A large U.S. randomized cash trial gave 1,000 low-income adults US$1,000 per month for three years and found lower labor-force participation and hours worked, which gives opponents a concrete labor-supply argument even if the result does not settle national policy (NBER).

China is a weak but real intellectual signal. Cai Fang, a prominent Chinese labor economist, discussed conditions for universal basic income in the AI era in 2026, and a Pekingnology translation summarized him as urging broader income support that could reference UBI as AI reshapes work (Sina Finance, Pekingnology). But China's official 2026-2030 human-resources and social-security plan emphasizes job creation, keeping surveyed urban unemployment within 5.5%, basic old-age insurance coverage above 95%, minimum wage mechanisms, and an AI employment-creation plan, not a high universal adult cash payment (State Council / Xinhua).

My final model is scenario-based because country paths are highly correlated. A world where AI is mild, fiscal politics stays normal, and no first mover appears suppresses most countries at once. A world with severe AI displacement or a new social-dividend politics raises many countries at once, but governments may still choose tax credits, targeted benefits, public jobs, wage subsidies, pensions, or services instead of a qualifying UBI.

This gives about 14%. A country-level sanity check points to the same range: South Korea, the United States, Saudi Arabia, and Canada carry most of the visible probability; the EU-wide path is below 1%; China, Brazil, India, Indonesia, Mexico, and South Africa are mostly tail cases because the required benefit is too large relative to current output. The country chances cannot be added mechanically, because the main upside driver is the same AI and fiscal-regime shift across countries.

What's non-obvious

The main filter is the payment level, not the word UBI. A small universal dividend by 2040 is quite plausible in several G20 countries. A permanent standard adult benefit worth PPP US$12,000 per year is a different object: it is a double-digit-GDP gross payment stream in the rich countries and an enormous share of output in middle-income G20 countries.

The strongest YES path is not today's pilots scaling smoothly. It is a funding story. AI sovereign-wealth funds, data dividends, carbon dividends, land rents, commodity rents, or public ownership of strategic capital could turn UBI from welfare policy into social-dividend policy. That frame is visible in the 2026 U.S. AI wealth-fund bill and in Chinese AI-era social-security commentary, but both are still far from a permanent US$1,000-per-month adult payment.

Limitations

The largest uncertainty is technological. Current IMF and ILO evidence measures exposure and task transformation, not actual job destruction, and the difference matters. If AI produces visible, persistent unemployment or a broad collapse in wage bargaining power in rich G20 countries before the mid-2030s, this probability should move up sharply; if AI mostly raises productivity without mass displacement, it should move down.

The cost table is an approximate 2025 fiscal screen. It uses population aged 15+ rather than exact legal adults, current PPP GDP rather than projected 2040 output, and no model of tax-back, benefit replacement, migration, labor-supply response, or PPP revisions. Real GDP per adult growth by 2040 would lower the cost share, while aging, debt service, health spending, and political resistance to higher taxes would push in the other direction.

I did not prove a negative across every non-English legislative database. A qualifying program would be visible once payments start, but early bills, party platforms, and internal policy studies can be missed. The ambiguous edge case I would watch most is a citizens-only dividend in Saudi Arabia or another high-migrant state; the question allows adult citizens or permanent residents, but a resolution authority would still need to decide whether exclusions are minor enough to count.