Between 2017 and 2025, at least 122 guaranteed basic income pilots were conducted across 33 U.S. states and the District of Columbia, according to a comprehensive new AEI working paper by Kevin Corinth and Hannah Mayhew. Combined, these pilots distributed roughly $481.4 million in cash transfers to 40,921 recipients, with the average recipient receiving about $11,765 total over an average pilot duration of 18.4 months, at an average monthly payment of $616.
It's worth being precise about data quality here: of those 122 pilots, only 52 have published outcomes, only 35 used randomized controlled trial designs (the gold standard for isolating a program's actual causal effect), and only 30 specifically reported employment outcomes. This means the evidence base, while genuinely substantial, is considerably smaller and more uneven than the raw "122 pilots" headline number might suggest on its own.
Among the 30 randomized-trial pilots with published employment data, the AEI analysis finds guaranteed basic income produces a mean effect of just a 0.8 percentage point increase in the employment rate, essentially negligible in practical terms. But that headline average obscures a more striking finding specific to pilot size: among the four largest pilots, those with treatment groups of 500 or more participants, which together represent 55% of all treatment group participants across every study, the mean effect on employment was actually negative, a 3.2 percentage point decrease.
The researchers note this negative effect is consistent with broader academic literature on income and labor supply (an income elasticity of -0.18), and specifically flag that smaller pilots may produce less reliable or even biased results precisely because of their limited sample sizes, meaning the larger, more statistically robust pilots may be the more trustworthy signal, not the noisier smaller ones that make up the bulk of the 122-pilot count.
Recent commentary in outlets including Newsweek, the LSE Business Review, and Fortune has argued that AI-driven job losses may eventually require something like a universal basic income to offset large-scale displacement. The Daily Economy's analysis pushes back directly on this framing, arguing that the actual pilot evidence gathered so far, particularly the negative employment effect found in the largest, most rigorous studies, doesn't support treating UBI as a ready-made solution to an AI-driven labor disruption that hasn't fully materialized yet, framing this as applying an old, empirically mixed policy idea to a new, still-uncertain problem.
UBI supporters generally argue guaranteed income provides genuine economic security and flexibility, citing pilot findings on improved mental health, educational engagement, and reduced poverty-related stress, and argue that modest or even negative employment effects are an acceptable, even expected, tradeoff for a program explicitly designed to reduce economic pressure to work more hours. Critics generally point to the AEI analysis's core finding, a real negative employment effect specifically in the largest, most statistically reliable pilots, as evidence the policy may create the kind of work disincentive effect economists have long worried about, and argue the broader base of 122 pilots is too methodologically uneven to serve as strong evidence for wide-scale, permanent adoption. Both sides broadly agree the evidence base has grown substantially in the past decade, the genuine disagreement is over which specific findings, the modest overall average or the larger, more negative effect in the biggest pilots, should carry more weight in shaping future policy.
Want the core arguments from both sides, side by side?
See the Left vs. Right Breakdown on Universal Basic Income →