Universal Basic Income
My Positions
- I want a floor beneath which nobody falls. That is settled. Whether a universal basic income is the best way to build the floor is an empirical question, the evidence does not yet settle it, and I hold this note more loosely than most.
- The pilots show that giving poor people money unconditionally does not make them stop working, does improve their health, stress and trust in institutions, and costs less to administer than a system built to catch cheats. They do not show what happens when the payment is universal, permanent, and paid for by the taxes that would fund it, because no pilot has been any of those things.
- The strongest case for it is not utopian. It is that means-tested, conditional welfare is cruel to administer, expensive to police, and traps people in poverty by withdrawing benefits faster than wages replace them, at effective marginal rates no one would impose on the rich. An unconditional floor removes the trap and the humiliation at once.
- The strongest case against it is arithmetic. A basic income high enough to live on costs a large fraction of national income; one low enough to afford does not replace the services it is supposed to replace. "An affordable UBI is inadequate and an adequate UBI is unaffordable" is the objection in one line, and I have not seen it answered.
- Where I land, for now: keep the services (health, education, housing, disability support, which are needs, not incomes) and put an unconditional cash floor beneath them, clawed back through the tax system rather than a benefits office. That is closer to a negative income tax than to the full dream, and it is what I would pilot at scale.
Notes
- A long and mixed ancestry. Thomas Paine proposed in Agrarian Justice (1797) that every citizen receive a capital sum at twenty-one and a pension at fifty, funded by a tax on inherited land, on the ground that the earth had been common property and its enclosure owed the dispossessed a rent. Milton Friedman proposed the negative income tax in Capitalism and Freedom (1962) as a way of abolishing the welfare bureaucracy, and Richard Nixon's version passed the House of Representatives in 1970 before dying in the Senate. Friedrich Hayek supported a guaranteed minimum. Martin Luther King called for a guaranteed income in his last book. When the left and the right both want a thing you should check that they mean the same thing, and here they do not: one side wants to abolish poverty and keep the state, the other wants to abolish the state's welfare apparatus and send a cheque instead. The same two words describe two different programmes, and most arguments about "UBI" are arguments between them.
- What the experiments found. Finland paid two thousand unemployed people 560 euros a month for two years (2017 to 2018) with no conditions; employment rose slightly relative to the control group, and well-being, trust and mental health improved markedly. The Canadian town of Dauphin, Manitoba, received a guaranteed income from 1974 to 1979; a 2011 reanalysis by the economist Evelyn Forget found hospital admissions fell by 8.5 per cent, more teenagers finished school, and the only groups that worked less were new mothers and adolescents. Stockton, California, gave 125 residents 500 dollars a month from 2019; full-time employment rose faster in the recipients than in the control group. GiveDirectly's long-term trial in Kenya finds no reduction in work and more small enterprise. The Alaska dividend, paid to every resident since 1982, has no detectable effect on employment. The largest American study, which gave a thousand low-income adults a thousand dollars a month for three years, reported in 2024 that recipients worked about 1.3 hours a week less, spent the money on food, rent and transport, and showed no lasting measured improvement in health, which is an honest negative result and should be read as one. Wales paid care leavers 1,600 pounds a month from 2022. The consistent findings across all of them: people do not stop working, the money is spent sensibly, and stress falls. The consistent limitation: every one was temporary, small, and paid for by someone else, so none tells you what a whole country does when it knows the money is permanent and sees it on its tax bill.
- The trap it would fix. Universal Credit withdraws 55 pence of benefit for every extra pound earned after tax, and when housing support, council tax reduction and free school meals are stacked the effective marginal rate on a low earner taking extra hours can exceed that on a banker. Add the conditions, the sanctions, the five-week wait and the assessments that must be appealed to be won, and the result is a system that is expensive to run, that misses people who are entitled, and that punishes exactly the behaviour it claims to want. Wealth and Luck argues that a decent society stops bad luck at a floor; this is the floor having a trapdoor in it. An unconditional payment has no cliff, no sanction and no form, and the saving in administration is real, though smaller than its advocates claim.
- The arithmetic. A thousand pounds a month to every British adult is roughly 650 billion pounds a year, more than half of all public spending. Most of that is churn, since the same money is taxed back from people who did not need it, and the net cost depends entirely on the tax schedule, which is why the honest form of the proposal is a negative income tax with a published rate. But a payment that could be afforded net would be a few hundred pounds a month, which does not house anyone, does not replace disability support, and would leave the severely disabled worse off if it replaced the benefits they get now. The automation argument, that robots will shortly leave most people without work, does not rescue the sums: labour-market prophecy has been wrong for two centuries and the case for a floor stands or falls on poverty, not on robots.
- Where I come down, and why loosely. The goal follows from positions I hold firmly: equality of opportunity is a fiction without a floor, money is insurance against bad luck and some people are born uninsured, and a state that sees individuals rather than categories should not be sorting the deserving poor from the undeserving at a desk. The instrument is open. I would fund the services first, because a hospital cannot be bought with a monthly cheque; I would convert the cash benefits into an unconditional floor administered through the tax system; and I would run the experiment that has not been run, at scale and for long enough to see what people do when it is real. If it turned out that the floor cost more than it saved and bought less well-being than the same money in services, I would say so and move the money. That is the point of holding a view provisionally rather than as a creed, and this one is not a creed.
Sources
- Kangas, O. et al., Experimenting with Unconditional Basic Income: Lessons from the Finnish BI Experiment 2017–2018 (2021)
- Forget, E. L., "The Town with No Poverty: The Health Effects of a Canadian Guaranteed Annual Income Field Experiment", Canadian Public Policy 37 (2011)
- West, S. et al., Preliminary Analysis: SEED's First Year (Stockton Economic Empowerment Demonstration, 2021)
- Vivalt, E. et al., "The Employment Effects of a Guaranteed Income: Experimental Evidence from Two U.S. States", NBER Working Paper 32719 (2024)
- Banerjee, A. et al., "Universal Basic Income: Short-Term Results from a Long-Term Experiment in Kenya" (2023)
- Jones, D. and Marinescu, I., "The Labor Market Impacts of Universal and Permanent Cash Transfers: Evidence from the Alaska Permanent Fund", NBER Working Paper 24312 (2018)
- Martinelli, L., Assessing the Case for a Universal Basic Income in the UK, Institute for Policy Research, University of Bath (2017)
- Milton Friedman, Capitalism and Freedom (1962), chapter 12
- Thomas Paine, Agrarian Justice (1797)
- Rutger Bregman, Utopia for Realists (2017), for the case at its most optimistic