Negative income tax

A negative income tax (NIT) guarantees a minimum income through the tax system. Everyone is entitled to a basic amount; as earned income rises, the payment is withdrawn at a set rate until it reaches zero at a break-even income, above which people pay ordinary positive taxes. With a guarantee G and a withdrawal rate t, a household with earnings E receives G − tE while that figure is positive.

For example, with a guarantee of 10,000 and a withdrawal rate of 50%, someone with no earnings receives 10,000, someone earning 8,000 receives 6,000 (total income 14,000), and payments end at 20,000 of earnings. Unlike conventional welfare, which often withdraws benefits pound for pound, an NIT is designed to ensure that working always increases total income.

Milton Friedman popularised the idea in Capitalism and Freedom (1962) as a replacement for a patchwork of welfare programmes, but it also attracted support from economists on the left, including James Tobin, who co-authored a detailed 1967 proposal.

The NIT combines three goals that are hard to achieve together: adequacy (a decent floor), work incentives (low effective marginal tax rates) and affordability (limited cost). Any income-support system must choose two of the three parameters, since the break-even income equals G / t: a high guarantee with a low withdrawal rate is generous and pro-work but expensive, because payments extend far up the income distribution.

Relationship to universal basic income. An NIT and a UBI funded by a flat tax can produce identical net incomes for every household: a UBI pays everyone and taxes back from those with earnings, while an NIT pays only the net amount. The differences lie in administration (UBI is paid upfront to everyone; NIT is assessed on income, often after the fact), perception (an NIT looks like targeted welfare; a UBI like a universal right), and cash-flow timing for people with volatile incomes.

Supporters from the right value its simplicity, its replacement of paternalistic in-kind programmes and its preservation of market incentives. Supporters from the left value the unconditional floor and the removal of stigma.

• US income maintenance experiments (1968–1982). Four federally funded experiments tested NIT schemes: New Jersey and Pennsylvania (from 1968), rural Iowa and North Carolina, Gary in Indiana, and the largest, in Seattle and Denver (SIME/DIME, running through the 1970s). They found reductions in work effort that were modest for primary earners and larger for second earners and young people, alongside positive effects on schooling in some sites.
• Mincome, Manitoba (1974–1979). A Canadian experiment included a "saturation site" in the town of Dauphin, where every resident was eligible. Analysis of administrative health records by Evelyn Forget found that the hospitalisation rate of participants fell by about 8.5% relative to matched controls, particularly for accidents, injuries and mental health.
• The Family Assistance Plan. President Nixon proposed in 1969 a federal guaranteed income for families with children, designed by Daniel Patrick Moynihan and others. It passed the House of Representatives on 16 April 1970 but stalled in the Senate Finance Committee.
• Descendants. The US Earned Income Tax Credit (1975) and the UK's tax credits and Universal Credit use income-tested, tax-administered payments with withdrawal rates, but they are tied to work or household circumstances rather than being universal guarantees.
• Ontario Basic Income Pilot (2017–2018). Despite its name, its design was an NIT: payments were reduced by 50 cents for each dollar earned. The provincial government elected in 2018 cancelled it before its planned completion, so no full evaluation was conducted.

• The trilemma bites. Realistic budgets force either a low guarantee or a high withdrawal rate, which recreates disincentives near the bottom of the income distribution.
• Work and family effects. Early analysis of SIME/DIME reported higher rates of marital dissolution among recipients, which damaged political support; a later reanalysis by Cain and Wissoker found no significant effect, but the episode showed how contested experimental results can be in policy debate.
• Communication of results. Widerquist argues that the NIT experiments were widely misunderstood: technical findings about labour-supply elasticities were translated into political claims that NIT "discourages work" without discussion of whether the reductions were large or desirable.
• Administration. Annual income assessment does not match fluctuating monthly needs. Real-time income reporting, as attempted in Universal Credit, introduces its own errors and hardships.
• What it replaces. An NIT that replaces services such as housing support or disability assistance may leave people with high needs worse off.
• Automation debate. As interest grows in income support for a world with less paid work, critics worry that a means-tested NIT entrenches a divide between payers and recipients, while UBI advocates argue universality builds broader political support.

The NIT has been one of the most rigorously tested ideas in social policy: the US experiments were among the first large randomised social experiments. They established that a guaranteed income reduces work effort somewhat but nowhere near the collapse that critics feared, and Mincome suggested health benefits. Politically, the NIT has not been adopted in pure form anywhere, but its logic shapes tax credits and integrated benefit systems in several countries. In contemporary debates on automation and post-labour economies, it remains the main targeted alternative to universal basic income.

• Forget (2011), "The Town with No Poverty", Canadian Public Policy
• Milton Friedman, Capitalism and Freedom (University of Chicago Press, 1962), ch. 12
• Karl Widerquist, "A Failure to Communicate", Journal of Socio-Economics 34(1) (2005)
• Philip K. Robins, "A Comparison of the Labor Supply Findings from the Four Negative Income Tax Experiments", Journal of Human Resources 20(4) (1985)
• Daniel P. Moynihan, The Politics of a Guaranteed Income (Random House, 1973)