A negative income tax (NIT) integrates income support into the tax system: rather than separating welfare programs from taxation, a single formula determines whether a household owes tax or receives a transfer. Below a defined income threshold, the "tax" becomes negative — the government pays the household a fraction of the gap between their earnings and the threshold.
The concept is most closely associated with Milton Friedman, who proposed it in Capitalism and Freedom (1962) as a way to replace the patchwork of categorical welfare programs with a simpler, more neutral mechanism. Friedman argued it would preserve work incentives better than means-tested benefits with sharp eligibility cliffs, because the phase-out rate could be set explicitly (e.g., 50%), leaving recipients with positive returns to additional earnings.
A typical NIT has three parameters: a guaranteed minimum (the payment when earnings are zero), a phase-out or "benefit reduction" rate, and a break-even income at which payments fall to zero. Above break-even, the household pays positive tax.
The United States ran four large-scale NIT field experiments between 1968 and 1982 — in New Jersey/Pennsylvania, rural Iowa/North Carolina, Gary (Indiana), and Seattle/Denver (SIME/DIME). Findings showed modest reductions in labor supply, particularly among secondary earners, and contested effects on family stability. Although a full NIT was never enacted, the experiments influenced the design of the Earned Income Tax Credit (EITC), signed into law in 1975, which functions as a partial NIT for low-income workers with earnings.
NIT is often contrasted with universal basic income (UBI): UBI pays everyone the same flat amount regardless of income, while NIT targets payments only to those below the threshold. Mathematically the two can be made equivalent through the tax schedule, but they differ politically and administratively — NIT requires income verification, while UBI does not.
Debates today focus on phase-out rates, interaction with existing transfers, and administrative feasibility.