Universal Basic Income: Would It Work in America?

Andrew Yang made it a campaign issue. Elon Musk has endorsed it as a response to automation. Sam Altman funded research on it. Finland ran a national experiment. Multiple US cities have piloted it.

Universal Basic Income has gone from the fringes of economic policy to a serious conversation among people across the political spectrum.

The question is whether any of the evidence actually supports it at scale — and whether the United States could realistically fund it.

What the Pilot Programs Actually Found

The most rigorous UBI evidence comes from randomized controlled trials — studies where some people received payments and comparable people did not.

Stockton, California (SEED program, 2019-2021): 125 residents received $500/month for 24 months with no conditions. Results: full-time employment increased by 12 percentage points among recipients (compared to 5 points in the control group). Mental health improved significantly. Spending was primarily on food, clothing, and utilities.

Finland (2017-2018): 2,000 unemployed people received €560/month unconditionally. Results: employment outcomes were similar to the control group but well-being, mental health, and trust in institutions improved significantly. Participants reported less stress, better health, and greater confidence in their futures.

Kenya (GiveDirectly, ongoing): The largest long-term UBI experiment in the world — providing 12 years of basic income to villages in rural Kenya. Early results show increased assets, improved food security, higher local economic activity, and no significant reduction in work.

Alaska Permanent Fund (1982-present): Alaska has paid every resident an annual dividend from oil revenues for over 40 years — ranging from $1,000 to $2,000 per year. Alaska has lower poverty rates than comparable states, and studies have found no significant effect on employment or welfare dependency.

The consistent finding across very different contexts: people do not stop working when their basic needs are met. They make better decisions.

The Cost Problem Is Real

The pilots work. The scale is another issue entirely.

A $1,000/month UBI for all 260 million American adults = approximately $3.1 trillion per year.

The entire 2024 federal budget was $6.7 trillion. Defense spending was $886 billion. Social Security was $1.4 trillion. Medicare was $869 billion.

A meaningful UBI would cost more than Social Security and Medicare combined. This is not a fringe concern — it is an arithmetic reality that any honest UBI proposal has to address.

Yang's proposal to fund it with a 10% Value Added Tax would raise roughly $800 billion — a significant gap. Other funding mechanisms — wealth taxes, carbon dividends, consolidating existing welfare programs — each have their own limitations and political obstacles.

The More Realistic Version: Negative Income Tax

The idea that actually polls well across party lines — supported by both Milton Friedman (free market economist) and liberal economists — is the negative income tax.

The concept: instead of giving everyone a check, the government tops up the incomes of everyone below a threshold, tapering off as income rises. This is essentially what the Earned Income Tax Credit already does, just expanded dramatically.

It is less universal, less philosophically pure, and far cheaper than full UBI — but it accomplishes the core goal of eliminating poverty-trap welfare cliffs and providing a genuine floor under low incomes.

Automation and the Long Game

The UBI debate has a time dimension that often gets lost in immediate cost discussions.

If AI and automation eliminate a significant portion of jobs over the next 20 years — estimates range from 10% to 40% of current jobs being automatable — the social insurance mechanisms built around employment will break down. People who lose jobs to automation don't get unemployment insurance indefinitely.

UBI is partly a hedge against a future where the connection between labor and income is weaker than it has ever been. Whether that future arrives on the schedule AI optimists predict is uncertain. But the policy conversation needs to start now, not after the transition is already underway.