A G20 country pays 1 million adults an unconditional income

Where things stand (12 Oct 2026)

I found no programme in any G20 country that meets every test the question sets: payments at least monthly, cash the recipient can use freely, reaching at least 1 million adults, with no income test, no occupation-based exclusion and no age band other than 18+. I also found none in preparation. The programmes that come closest each fail on at least one test:

• Punjab, India – Mawan Dheeyan Satkar Yojana (launched 1 Jul 2026, Rs 1,000–1,500 a month to women 18+, about 33–37 lakh women paid). It comes closest, but it fails three tests:
  • The notification excludes women who paid any income tax, which is an income test.
  • It excludes regular and retired government employees and legislators, which is an occupation test.
  • Money is paid in three-month instalments: a "3-month installment" of Rs 1,262.54 crore went to 36.87 lakh women in October.
  • It also started before the question opened.
  • Sources: Indian Express, 2 Jul 2026; Times of India, 4 Apr 2026; Times of India and Punjab Newsline, Oct 2026.
• West Bengal – Annapurna Bhandar (June 2026, Rs 3,000 a month): women aged 25+, income-tested, government employees excluded (Telegraph India, 13 May 2026; News18, 1 Jun 2026).
• Delhi – Lakshmi Yojana: women aged 21–60, family income cap, many exclusions, and part of the money goes to a restricted digital-currency wallet (India Today, 1 Aug 2026).
• Tamil Nadu: the new TVK government is reworking the income-tested women's grant. Its own pledge was capped at age 60 (The Hindu and Hindustan Times, 15 May 2026; India Today, 16 Apr 2026).
• South Korea – rural basic income: the 2027 budget expands it from 17 to 35 counties at ₩150,000 a month (JoongAng, Sept 2026), which may approach about 1 million residents. But it is paid in local currency, which the question explicitly excludes. A planned basic income for artists depends on occupation, so it fails too.
• United States: Trump's promised $5,000 "dividend" would be a one-off payment and needs Congress (PBS, Sept 2026).
• South Africa: the planned replacement for the temporary relief grant targets poor adults aged 18–59 (explain.co.za, 18 Aug 2026).
• Outside the G20: Taiwan's AI dividend is a one-off; the Marshall Islands' universal payment is quarterly (Manifold comments, Dec 2025).

Base rate

No G20 country has ever run a universal, monthly, cash programme for 1 million or more adults. Metaculus puts the chance of a nationwide UBI in the US, UK, China or any EU state before 2030 at about 2% (search snippet). This question counts more cases than that one: all G20 members, subnational programmes and pilots, a horizon running to 2031, and possibly women-only schemes. That justifies a figure a few times higher, but still low.

Paths to YES

1. An Indian state launches a women's scheme with no exclusions at all (about 2%). Gender is not on the question's list of disqualifying conditions, so read literally such a scheme would count. I put the chance a resolver accepts it at about 65%. Elections in 2027–2031 will produce new schemes, but none of roughly 20 so far has been fully universal. Even the AAP government in Punjab, which campaigned on paying every woman, excluded taxpayers and government employees.
2. A universal basic income or "AI dividend" driven by job losses to AI (about 3%). This is weighted toward 2029–2031: if displacement is severe, US politics after 2028, China and others become possible sources. Even then, dividend designs tend to be one-off, annual or quarterly, or income-tested.
3. Emergency monthly transfers in a crisis (about 1%). India's 2020 payments of Rs 500 a month to women holding basic Jan Dhan accounts are an example of the kind of programme that could qualify.
4. Anything else, such as Korea switching to cash (about 0.5%).

Building the series

I set quarterly hazards of 0.12% in Q4 2026, about 0.2% through 2027, 0.25–0.30% in 2028, 0.35–0.40% in 2029–2030 and 0.42–0.43% in 2031. The cumulative figure is 1 minus the running product of (1 − hazard), computed in Python. It reaches 0.96% at end-2027, 3.4% at end-2029 and 6.5% at end-2031.

Strongest case against my estimate

AI-driven job losses may come faster than expected, and India's competition over welfare promises is pushing coverage toward universal (Punjab says about 97% of women are covered). One clean-design state scheme or a single crisis transfer could resolve the question. Against that, the criteria are strict enough to rule out nearly every design that exists in the real world, which keeps me well below 10%.

Question. By the last day of each quarter, has any G20 country started a government-funded programme that pays a cash income, at least monthly, to at least 1 million adults regardless of their income, work, occupation, or age?

Definitions.

• G20 country: Argentina, Australia, Brazil, Canada, China, France, Germany, India, Indonesia, Italy, Japan, Mexico, Russia, Saudi Arabia, South Africa, South Korea, Türkiye, the United Kingdom, or the United States, or a country subsequently admitted as an individual G20 member. Membership in the EU or African Union alone does not qualify a country.
• Government-funded programme: a single programme whose cash benefits are financed wholly or partly by a national, regional, state, or local government in that country, including a government-administered public fund. Government endorsement or funding only the evaluation or administration of privately financed benefits is insufficient. A subnational programme or time-limited pilot counts; nationwide coverage and a national statute are not required.
• Cash income: a positive, non-repayable monetary transfer for recipients' unrestricted use, including bank deposits, cashable cheques, and cash-withdrawable payment cards. There is no minimum payment amount or poverty-line requirement. Restricted-use vouchers, local shopping credits that recipients cannot redeem for cash, in-kind benefits, loans, wages, and reimbursements do not count.
• At least monthly: the programme provides a recurring entitlement payable every calendar month or more frequently. Annual or quarterly entitlements do not count merely because different recipients receive them in different months, or because the amount is expressed as a monthly equivalent.
• Adults: people aged 18 or older. The programme may also pay children, but children do not count toward the threshold.
• Regardless of income, work, occupation, or age: access to a positive qualifying payment cannot depend on income or wealth, employment or unemployment, work-search or service requirements, present or former occupation, or any age restriction beyond reaching 18. All adult ages must be eligible, including retirement ages. Ordinary taxation of the benefit does not disqualify it; a programme-specific income test or clawback that eliminates the payment for higher-income recipients does. Citizenship, residence, identity verification, application requirements, and random pilot selection are permissible. An unemployed-only pilot or a pension, farmer, youth, or working-age-only scheme does not qualify, even if payments are unconditional after enrolment.
• 1 million or more: at least 1,000,000 distinct adults must actually receive a qualifying payment for the same monthly entitlement period, or within the same calendar month for more frequent payments. Counts of eligible people, planned enrolment, households, accounts, children, or cumulative recipients across months are insufficient unless official data establish the required adult count. Payments routed through a household account count only for adults with an identifiable individual entitlement. Separate programmes cannot be added together; regional implementations of one common programme may be aggregated within one country.

Resolves YES when qualifying official publications establish both that the programme has begun disbursement under a recurring at-least-monthly schedule and that actual payments have reached the adult-recipient threshold under the eligibility rules above. A pilot need not be permanent, and no connection to AI or automation is required. Expansion or reform of an existing programme counts when it first meets all these conditions. A first monthly payment reaching the threshold is sufficient if an operative, funded schedule provides for subsequent monthly payments.

Resolves NO if no qualifying publication is available by the horizon. Announcements, proposals, enacted authorisations, appropriations, and projected recipient numbers alone do not establish that payments have started at the required scale. A one-off transfer or retrospective division of a lump sum into monthly amounts does not qualify. An already qualifying programme's routine continuation, renaming, or payment increase does not constitute starting a programme.

Resolving source. Official publications of the responsible government's administering ministry or agency, treasury or finance ministry, legislature or official gazette, national statistical office, or public audit institution. Programme regulations, statutes, and budget documents may establish eligibility, funding, and payment frequency; official disbursement reports, administrative statistics, or audit findings must establish actual payment and recipient counts. The Republic of Korea's Ministry of Agriculture, Food and Rural Affairs rural basic income pilot, as described on Korea.net on 6 April 2026, illustrates excluded restricted-use vouchers, not qualifying cash. Sources are narrower than Public confirmation (A2): press reporting and nongovernmental estimates alone do not resolve this question. Apply Resolution date (A3), the Resolution window, and the Observation period convention to published payment statistics. Later official data may establish an earlier qualifying start, but publication sets the resolution date.