Universal Basic Income: Evidence From Pilots and Natural Experiments

Abstract

Universal basic income (UBI) — unconditional cash transfers to all citizens regardless of employment status — has attracted growing interest as technological unemployment concerns have risen. A wave of randomized experiments and natural experiments, including Finland's 2017–2018 UBI trial, Stockton's guaranteed income pilot, and Alaska's Permanent Fund Dividend, has provided direct evidence on the behavioral effects of unconditional cash. Contrary to concerns that free money reduces work effort, the evidence consistently finds that moderate unconditional cash transfers have small negative effects on employment, if any, while improving mental health, food security, and reported wellbeing.

Background

The idea of a universal basic income — a regular cash payment to every person, unconditional on work, family structure, or behavior — has been proposed by thinkers across the political spectrum, from Thomas Paine in 1797 to Milton Friedman's "negative income tax" proposal in the 1960s. Contemporary interest has been driven partly by concerns that automation will displace large numbers of workers and partly by critiques of means-tested welfare programs as stigmatizing, administratively burdensome, and riddled with poverty traps.

The primary empirical concern about UBI has always been labor supply: if income is provided regardless of work, will people work less? This "labor supply elasticity" question has been studied in controlled experiments since the 1970s.

The Evidence

Finland's Basic Income Experiment (2017–2018)

Finland conducted a two-year randomized controlled trial of a basic income, randomly selecting 2,000 unemployed adults to receive €560/month unconditionally and comparing them with a control group receiving standard unemployment benefits. Participants were followed for two years.

The official evaluation (Kangas et al., 2020, published by the Finnish Social Insurance Institution) found that basic income recipients worked slightly more than the control group — 78 versus 73 days — though the difference was not statistically significant. Basic income recipients reported higher levels of wellbeing, life satisfaction, confidence in the future, and trust in institutions. Mental health symptoms were significantly lower in the basic income group.

This was not a full UBI (it was limited to unemployed people and not universal), and €560/month is below subsistence level in Finland, limiting generalizability. But it provided the first large-scale randomized evidence from a high-income country.

Stockton Economic Empowerment Demonstration (SEED)

SEED provided 125 Stockton, California residents with $500/month for 24 months, beginning in 2019. A control group of 200 residents received no payments. Participants were not required to be unemployed.

West et al. (2021, SEED Evaluation Report) found that the guaranteed income increased full-time employment among recipients: 28% of recipients had full-time employment at 12 months compared with 25% in the control group — the opposite of the predicted work-disincentive effect. Recipients also showed improvements in mental health, reductions in anxiety, and improvements in reported financial stability.

Cash spending was primarily on food, utilities, and retail — not, as critics often predict, on alcohol or gambling.

Alaska Permanent Fund Dividend

Alaska has operated a genuine universal basic income since 1982: the Permanent Fund Dividend (PFD) pays a portion of oil revenues to every Alaska resident. Annual payments have ranged from approximately $300 to $2,000.

Feinerman and Morey (2020) and Jones and Marinescu (2022, American Economic Journal: Economic Policy) studied the labor supply effects of the PFD using variation in payment amounts across years. They found no significant effect on employment but a small shift from full-time to part-time work — which they suggest reflects increased ability to pursue preferred work arrangements rather than withdrawal from the labor force. The PFD was not associated with increases in poverty and was associated with small reductions in inequality.

The 1970s Negative Income Tax Experiments

Four large U.S. and Canadian negative income tax experiments in the 1970s — New Jersey, Rural Income Maintenance, Gary, Seattle-Denver — are the largest controlled experiments on unconditional cash ever conducted, with more than 11,000 families enrolled. The consensus from these experiments was that guaranteed income reduced labor supply by approximately 2–7%, primarily by reducing working hours rather than causing complete withdrawal from employment. Secondary earners (typically wives) and youth were most likely to reduce hours.

These experiments informed the Reagan-era rejection of guaranteed income proposals. However, reanalysis of these experiments (Hum and Simpson, 1993, for the Canadian Mincome experiment) found that the labor supply reductions were concentrated in high school graduation (where youth stayed in school longer) and maternal care — arguably positive effects on human capital rather than negative effects on work ethic.

Counterarguments

UBI skeptics raise several concerns the pilots do not fully resolve. First, pilot effects may not generalize: in a pilot, participants know the transfers are temporary and that they are being observed, which may affect behavior. A permanent, universal program would have different economic equilibrium effects.

Second, the fiscal cost of a genuine universal basic income at subsistence level is enormous — on the order of $3–4 trillion annually for the United States — raising fundamental questions about what would be cut or taxed to fund it. Pilots funded by grants or oil revenues do not address this constraint.

Third, by being universal, UBI provides large transfers to people who do not need income support, reducing the poverty-reduction efficiency relative to targeted programs.

What We Can Conclude

The available experimental evidence does not support the prediction that unconditional cash transfers reduce employment. The most consistent findings are improvements in mental health, food security, and wellbeing, with small or no negative effects on labor supply. In some cases (Stockton), employment actually increased.

The harder question — whether a large-scale, permanent, universal program at subsistence-level payments could be fiscally implemented without major economic disruption — is not answered by the existing pilots, all of which involve modest payment levels and voluntary participation. What the evidence does establish is that the simplest version of the anti-UBI argument (free money makes people stop working) is not supported by controlled experimental data.

References

• Jones, D., & Marinescu, I. (2022). The labor market impacts of universal and permanent cash transfers: Evidence from the Alaska Permanent Fund. American Economic Journal: Economic Policy, 14(2), 315–340. https://doi.org/10.1257/pol.20190299
• Kangas, O., et al. (2020). The basic income experiment 2017–2018 in Finland: Preliminary results. Finnish Ministry of Social Affairs and Health. https://julkaisut.valtioneuvosto.fi/handle/10024/162219
• West, S., et al. (2021). Stockton Economic Empowerment Demonstration: Findings from the First Year. SEED. https://www.stocktondemonstration.org/reports
• Hum, D., & Simpson, W. (1993). Economic response to a guaranteed annual income: Experience from Canada and the United States. Journal of Labor Economics, 11(1), S263–S296. https://doi.org/10.1086/298335
• Widerquist, K. (2005). A failure to communicate: What (if anything) can we learn from the negative income tax experiments? Journal of Socio-Economics, 34(1), 49–81. https://doi.org/10.1016/j.socec.2004.09.050