ANALYSISApril 13, 2026
Should the United States adopt a universal basic income?
The debate over whether the United States should adopt a Universal Basic Income (UBI) — unconditional regular cash payments to all residents regardless of employment or income — is intensifying in 2025, driven by AI-related job displacement anxiety, growing inequality, and expanding pilot program data. As of 2025, no country has implemented a full UBI system, and the U.S. has never adopted one at the federal level, though citizens in 18 states and Washington D.C. have participated in some form of basic income study. A proposed Guaranteed Income Pilot Program Act of 2025 would establish a three-year nationwide pilot program administered by the Department of Health and Human Services.
If the government sent every American a monthly check — no strings attached — would it liberate millions from poverty and the tyranny of bad jobs, or would it hollow out the work ethic, balloon the deficit, and quietly gut the safety net programs that actually target the neediest? The case for and against UBI isn't just about money — it's about what the government owes you and what you owe society.
Not familiar with this story? Get context →
Martinelli finding: who UBI harms
C
The University of Bath modeling found proportionally larger losses concentrated in the bottom three income quintiles under every UBI scenario tested. That is not a minor footnote — it is a structural indictment of the policy's core logic. A program sold as a poverty cure that, under rigorous modeling, most harms the people it claims to save has a design problem, not a messaging problem.
L
We actually agree on what the Martinelli finding means — which is why serious UBI advocacy must insist on a supplemental architecture, not a replacement one. A UBI that strips housing vouchers from disabled renters to write checks to hedge fund managers isn't progressive policy, it's redistribution running the wrong direction. The distributional danger is real, but it's an argument about design, not a death sentence for the concept.
C
The supplemental design you're defending to escape the Martinelli trap is precisely the design that makes the fiscal math indefensible — you can't simultaneously avoid harming the poor and avoid the $4 trillion price tag, because the replacement architecture is what makes costs manageable in every real-world proposal.
L
That tension is real, and I won't pretend otherwise — but it's an argument for designing the program carefully, not abandoning it. The same fiscal constraint applies to the $1.9 trillion in tax expenditures that predominantly benefit households above the median income, and we rarely hear equivalent urgency about that price tag.
Conflicting pilot results on employment
C
The OpenAI-backed study — 1,000 participants in Texas and Illinois, one of the largest experiments ever conducted — found recipients were 2 percentage points less likely to be employed and worked 1.3 to 1.4 fewer hours per week. Proponents cite Stockton as a counter-example, but Stockton had 125 people, private funding, and a time-limited design. Those are not equivalent experiments, and the conservative reads the pattern correctly: small, short, locally-administered pilots produce optimistic results because the income is genuinely supplemental and recipients know it is temporary.
L
You're right that a 125-person, privately-funded, time-limited pilot can't bear the weight placed on it — but notice what that methodological critique actually proves. If population size, funding source, and program permanence all plausibly explain why Stockton diverged from Texas and Illinois, then the labor-supply question is genuinely unresolved rather than settled. You're not discrediting Stockton; you're describing exactly why we need the rigorous federal pilot program rather than premature closure in either direction.
C
Calling the question 'unresolved' is doing a lot of work when the largest U.S. experiment points toward reduced employment — the burden of proof should fall on demonstrating a program works before committing to it at federal scale, not on running experiments until we find the result we want.
L
Waiting until AI displacement 'materializes' before testing interventions means designing policy during a crisis rather than before one — and the conservative position of inaction is itself a choice with consequences, not a neutral default.
Targeted reform versus universal payment
C
Milton Friedman's negative income tax — the intellectual ancestor UBI proponents love to invoke — was deliberately targeted at low-income households, not universal. That distinction is the whole argument. A surgical instrument aimed at poverty-trap dysfunction is not the same as a $4 trillion firehose spraying cash at everyone including people who don't need it. Conservatives who respect Friedman's insight should recognize he explicitly rejected universality as fiscally indefensible.
L
Friedman's targeted design is a fair point, but it proves less than you think — the means-testing machinery required to implement it is precisely what creates the 100%-plus effective marginal tax rates both of us agree are dysfunctional. Every dollar you save through targeting, you risk spending on the bureaucratic apparatus that punishes people for earning more. The negative income tax's elegance was always partly in its simplicity, and complex targeting undermines that.
C
Earned income tax credit expansion and wage subsidies both preserve work incentives without the universality problem — the choice isn't between broken means-testing and a $4 trillion universal payment, and framing it that way lets UBI off the hook for its cost.
L
EITC expansions are worth doing, but they don't reach people outside the formal labor market — caregivers, the disabled, workers in the informal economy — which is exactly the population most trapped by the current system and least served by a work-conditional credit.
The $4 trillion cost framing
C
Four trillion dollars per year — from the UBI Center's own estimates — is larger than the entire federal discretionary budget. That is not a rounding error or a conservative talking point; it is what UBI proponents' own modeling produces. Committing to that expenditure before displacement even materializes would be among the most consequential and irreversible fiscal decisions in American history.
L
The $4 trillion figure assumes zero consolidation of existing programs, and you're treating it as settled when economists genuinely contest the net fiscal picture under a well-designed supplemental UBI. More importantly, you're selectively applying fiscal urgency — Congress let the expanded Child Tax Credit expire after it cut child poverty by an estimated 30%, while the 'unaffordable' debate continued. The burden of proof runs both directions.
C
The Child Tax Credit expiration is a genuine policy failure, but it actually illustrates the conservative point — even a partial, temporary, means-tested program proved fiscally unsustainable under real budget constraints, which is an argument against scaling up, not for it.
L
It expired because Congress chose not to renew it, not because it failed — those are different problems, and conflating fiscal will with fiscal capacity is exactly how status quo bias gets laundered as fiscal responsibility.
Acting before AI displacement arrives
C
The urgency manufactured around AI job destruction is itself disputed by serious economists who note that automation historically creates jobs alongside the ones it eliminates. Acting now by locking in a multi-trillion dollar universal entitlement — before that displacement even materializes — would be building permanent institutions around a speculative future. Ordered liberty means building institutions that can adapt, not pre-committing to irreversible expenditures on a crisis that may not arrive in the predicted form.
L
The historically-creates-jobs argument assumes displacement and creation happen at the same pace and for the same workers — it didn't work out that way for manufacturing communities in Ohio and Michigan, and treating that track record as reassurance requires ignoring the people who fell through the gap. Testing interventions before the crisis is not overreaction; it is the only moment when we have the institutional capacity to design them carefully.
C
The right lesson from deindustrialization is targeted, place-based investment and retraining — not a universal payment that would have done nothing to rebuild the social infrastructure those communities actually lost.
L
We tried targeted retraining programs for decades and they consistently underdelivered — at some point, 'targeted reform' becomes the argument you make when you want to appear sympathetic without committing resources sufficient to the problem.
Conservative's hardest question
The Stockton SEED pilot's finding of increased full-time employment among recipients is genuinely difficult to dismiss, and it cuts against the labor-supply reduction narrative. If even one well-designed iteration of cash transfers increases employment rather than reducing it, the labor disincentive argument — central to the conservative case — is more contested than a simple reading of the OpenAI study suggests.
Liberal's hardest question
The Martinelli finding is not an outlier — multiple economists have modeled UBI designs that inadvertently harm the poorest by replacing targeted benefits with universal payments that deliver less to high-need individuals. A liberal case for UBI depends entirely on insisting the program supplement rather than replace existing benefits, but that design constraint dramatically increases the cost and makes the $4 trillion price tag even harder to offset, undermining the fiscal argument simultaneously.
Both sides agree: Both sides agree that the existing means-tested welfare system is structurally broken, specifically that effective marginal tax rates exceeding 100% for low-income workers create a documented poverty trap that punishes upward mobility.
The real conflict: A direct values conflict over timing and reversibility: the conservative argues that committing to a multi-trillion dollar entitlement before displacement materializes is fiscally irresponsible, while the liberal argues that designing policy during a crisis rather than before one is the greater risk — a disagreement about which kind of error is more costly.
What nobody has answered: If a supplemental UBI costs $4 trillion annually on top of existing programs and a replacement UBI harms the poorest households most, is there any fiscally plausible UBI design that is both progressive in its distributional effects and politically sustainable — or does the entire debate rest on an architecture that cannot survive contact with actual budget constraints?
Sources
- Stanford University Basic Income Lab tracker (2025)
- UBI Center cost estimates
- OpenAI/Sam Altman-backed UBI experiment results (Texas and Illinois, 2022–2025)
- Stockton SEED pilot program follow-up study
- Luke Martinelli, University of Bath, UBI modeling research
- Juliana Bidadanure, Stanford Basic Income Lab, public statements
- Scott Santens, Economic Security Project, poverty elimination claims
- Brookings Institution analysis of UBI vs. pro-work alternatives
- Guaranteed Income Pilot Program Act of 2025 (legislative text/summary)
- 2025 New York City mayoral race reporting (Adrienne Adams proposal)