Policy Evidence Brief
Costs, trials, and research — summarized, quoted, and linked — showing why a no-strings check for everyone is the wrong answer, and why the idea has never actually been tested.
A payment big enough to live on, paid to everyone, is not a line item — it is a restructuring of the entire federal budget.
Hoynes and Rothstein priced the canonical version: $12,000 a year to every U.S. adult. That is roughly three-quarters of all federal spending, before a single existing program is touched.
“A universal payment of $12,000 per year to each adult U.S. resident over age 18 would cost roughly $3 trillion per year. This is about 75 percent of current total federal expenditures, including all on- and off-budget items, in 2017.”
Even wiping out the entire transfer system — welfare, food assistance, housing, the EITC — barely dents the bill. So the money has to come from somewhere new.
“implementing this UBI without cuts to other programs would require nearly doubling federal tax revenue; even eliminating all existing transfer programs – about half of federal expenditures – would make only a dent in the cost.”
The one thing the trials do agree on: free money is spent, and some of it comes out of the paycheck.
OpenResearch gave 1,000 low-income adults $1,000 a month for three years (2020–2023), against 2,000 controls. Recipients worked about 1.3 fewer hours a week, were 4.2 percentage points less likely to be employed, and their household income excluding the transfer fell by about $4,200 a year.
“For every dollar received, total household income excluding the transfers fell by about 28 cents, and total individual income fell by around 16 cents.”
The government's own negative income tax experiments — Seattle/Denver, Gary, New Jersey — measured real declines in hours worked, concentrated among secondary earners and young people. That is decades of evidence pointing one direction.
Where a UBI-like program has actually been run, the promised employment and health gains mostly never showed up.
Finland's nationwide 2017–2018 trial paid 2,000 unemployed people €560 a month. The official evaluation found small employment effects — roughly six extra days of work over the reference period — and the government did not continue the program. The economists who reviewed it were blunter about what the trial could prove.
“The evaluation design was severely compromised with some early policy choices; there were no effects on employment, nor on unemployment.”
The same U.S. trial measured stress, food security, hospital use, bloodwork and dental care. The benefits were real but brief, and there was no lasting health improvement to justify a permanent program.
“it did not improve self-reported physical health, clinical biomarkers, or mental health after the first year, and we can rule out small effects on many outcomes.”
In June 2016 the Swiss were asked directly whether to write an unconditional basic income into their constitution. 76.94% voted no (1,897,528 against, 568,660 for), on a 47% turnout.
Universal is the whole point of UBI — and also its central flaw. The money has to go everywhere, including to people who don't need it.
Because it ignores need, age, disability and children, a UBI moves money toward households the current system already serves least well — and toward the middle of the income distribution.
“A UBI would direct much larger shares of transfers to childless, non-elderly, non-disabled households than existing programs, and much more to middle-income rather than poor households. A UBI large enough to increase transfers to low-income families would be enormously expensive.”
The serious critique from the center-right isn’t “leave people to fend for themselves.” It’s that the same money, aimed at low wages and real need, buys more than a check sent to everyone.
“a UBI would be extremely expensive and yet do very little to reduce inequality or advance opportunity and social mobility … the federal government should pursue a pro-work strategy of income support, paying wage subsidies to low-wage workers along with targeted transfer benefits consisting of both cash and near-cash types of support paid to the most needy individuals and households.”
Trillions of new dollars, new taxes, and a rewritten benefit system would ripple through rents, wages, and prices. The pilots were never built to catch those ripples.
A few thousand recipients can't move national rents or wages. But a program funding itself into a tight housing market can — and the cost-of-living gains could quietly evaporate.
“There are systemwide questions—around financing, inflation, linkages to pensions, relationship to minimum wages, and the political economy—that pilots cannot fully answer.”
This is the part the slogan leaves out. UBI is not a policy that has been tried and is now awaiting a decision. It is an idea that has never been tested at the scale, the amount, or the duration it is being proposed.
Not one. Two countries — Mongolia and Iran — ran a national version briefly, and both were partial or short-lived. Everything else is a pilot: small, temporary, and usually aimed at a slice of the population.
“Currently, no country has a UBI in place, although there have been (and still are) several small-scale pilots and a few larger-scale experiences. Only two countries—Mongolia and the Islamic Republic of Iran—had a national UBI in place for a short period of time.”
This is the honest version of the evidence base. The experiments we have were deliberately made cheap and short so they could be afforded, which means they tested something other than the policy being advocated.
“UBIs meeting the definition we laid out above – large enough to live on, without phase-out or other eligibility restrictions – have never been implemented in a rich country on a large scale or even in a pilot experiment.”
Only 13 countries have ever run a UBI experiment: 125 people in Stockton, California; about 1,000 in Namibia; 2,000 in Finland; 4,000 in Canada. Nothing at the scale of a national program, over the time horizon a national program lasts.
“There have been no large-scale pilots that allow one to evaluate the longer-term consequences of these programs.”
The comparison is the authors' own. An airliner gets tens of thousands of test flights before anyone sells tickets. A program that could reshape the tax code gets a few hundred volunteers for two years, and then a political campaign.
“Such robust, large-scale testing is hardly conducted with social policies, despite their massive scale and impact. In contrast to the airline industry, the attendant budget for research and development is tiny. Yet, the adverse impact of an untested social policy could be more damaging than crashing the A380.”
Alaska's oil dividend is the standard example. It pays $1,000–$2,000 a year — not a basic income by any serious definition, and it is funded by a resource windfall, not by taxpayers. Stockton's trial was 125 people. Both get cited as if they settled the question.
“We know of only two examples of universal programs without strict eligibility requirements, though in each case the transfers are too small to qualify as a basic income as we define it. The Alaska Permanent Fund is a demogrant (with varying yearly payments), financed by the state's oil revenues. Payments in recent years range from $1,000 to $2,000 per year.”
Even the sympathetic literature says it plainly: the case for UBI rests mostly on what it means morally, and advocates still can't agree on what it is. A policy whose size, funding, and relationship to existing benefits are all unsettled can't be tested — and shouldn't be enacted.
“a UBI engenders interest as a societal ideal to which to aspire, and not merely a program.”
Every argument for basic income in the Business Insider explainer “Universal basic income pros and cons — according to experts” (Sep 2024), and the answer to each.
Answer: gaps in a targeted safety net are fixed by fixing the targeted programs, not by mailing checks to millionaires. Brookings economists conclude that pro-work wage subsidies and targeted transfers do more for the poor per dollar than a universal payment, which is expensive and does little to reduce inequality.
Answer: the same article admits most of those pilots were short-term guaranteed income for selected low-income groups — not UBI. Denver and Atlanta tested a few hundred people for a limited time. A targeted, temporary grant succeeding says nothing about a permanent payment to 250 million adults.
“There have been no large-scale pilots that allow one to evaluate the longer-term consequences of these programs.”
Answer: the largest U.S. randomized cash trial ($1,000 a month for three years) found no improvement in mental health, physical health, or clinical biomarkers after the first year. Stress and food-security gains were short-lived.
Answer: in that same trial, recipients worked about 1.3 fewer hours a week, and household income excluding the transfer fell by about 28 cents for every dollar received — roughly $4,200 a year. People did not climb into better jobs; they earned less.
Answer: the Los Angeles and Illinois findings come from programs aimed at struggling families. If anything they are an argument for targeted support to at-risk households — exactly what UBI would dilute by spreading the same money across everyone.
Answer: the bill is about $3 trillion a year — roughly three-quarters of all 2017 federal spending — and would nearly double federal revenue. No carbon tax or tax on the rich comes close; a VAT large enough lands on everyday shoppers. The article itself concedes it is “difficult to definitively predict” what Americans would owe.
Answer: spending money taken from other taxpayers moves demand around; it doesn't create it. Measured earnings fell in the biggest trial. The 130% figure is a model of a world that doesn't exist — the kind of system-wide effect that, as the World Bank notes, pilots cannot test.
Answer: paying for UBI by cutting existing programs takes money from the disabled, the sick and the poorest and hands a flat check to everyone, including the comfortable. Keeping those programs instead means paying for UBI on top of them.
Answer: Alaska's check is small and paid from oil money, not taxes. Economists who study UBI say it is not a basic income at all.
“We know of only two examples of universal programs without strict eligibility requirements, though in each case the transfers are too small to qualify as a basic income as we define it. The Alaska Permanent Fund is a demogrant (with varying yearly payments), financed by the state's oil revenues. Payments in recent years range from $1,000 to $2,000 per year.”
Answer: that is a hope, not a finding — no one has ever run a permanent, livable, universal payment to see. The article itself calls true UBI “largely theoretical.”
“UBIs meeting the definition we laid out above – large enough to live on, without phase-out or other eligibility restrictions – have never been implemented in a rich country on a large scale or even in a pilot experiment.”
Answer: that is a reason not to do it. Injecting $3 trillion a year into the economy with unknown effects on rents and prices is a gamble — and the World Bank says pilots cannot settle financing or inflation questions.
Answer: a floor is a goal every safety-net program shares. UBI's appeal is moral, not evidential — and when voters were asked directly, in Switzerland, 76.9% said no.
“a UBI engenders interest as a societal ideal to which to aspire, and not merely a program.”
UBI asks taxpayers for trillions to fund a program that, in the best-designed trials that exist, reduced work and earnings without lasting health gains — while sending money to people who don't need it. And the trials that would settle the argument have never been run: no country has ever had a real UBI, and the pilots we have were too small, too short, and too cheap to answer the questions that matter. Targeted, work-linked support does more with less, and it has actually been tested.
Every quote above is the linked source's exact wording. Links go to the original working papers, the Finnish ministry's published results, the Swiss Federal Chancellery's vote record, and the World Bank and Brookings reports.