Income Subsidy Pilot for Workers in Transition

When a job disappears, the bills keep coming. Yet the EITC arrives once a year, and the federal program that paid trade-displaced workers while they retrained stopped certifying new groups in 2022. Congress should clear the benefit cliffs that make cash aid backfire, build the data to judge it honestly, and test a monthly transition payment before deciding anything larger.

The Problem

Sixty-four percent of Americans expect AI to lead to fewer jobs over the next 20 years.¹ No one knows how many jobs AI will displace, and this brief assumes no number. But work supplies more than income: structure, skills, and a place in a community. When a job disappears, a family needs time to retrain, move, or start over.

The help that exists was built for other problems. The EITC sent about $70 billion to 24 million workers and families for tax year 2024,² but only through the annual tax return.

What cash does is better studied than the debate suggests. In one U.S. randomized study, 1,000 low-income adults received $1,000 a month for three years. They spent at least $300 a month more, mostly on housing, food, and cars,³ and worked 1 to 2 fewer hours a week than a comparison group.⁴ Alaska has paid every resident an annual dividend since 1982, $1,200 in 2026,⁵ and economists found no drop in employment.⁶ Neither was designed for workers in the middle of a job loss. Three gaps leave those workers exposed:

1. Cash can cancel benefits. Supplemental Security Income subtracts countable income from the monthly check,⁷ so a poorly designed payment can simply replace help a family already has.
2. Help comes once a year. By law, the IRS cannot issue EITC refunds before mid-February each year.⁸ Rent comes due twelve times a year; the EITC comes once.
3. Transition aid has lapsed. Trade Adjustment Assistance paid income support to workers who lost jobs to foreign competition while they retrained. It has certified no new groups since July 1, 2022,⁹ and a 2026 House bill to revive it has not moved.¹⁰

Why legislation: Only Congress can fund new federal payments and decide how SNAP, Medicaid, SSI, and housing programs count them. It has done this before: federal programs must ignore tax refunds, including the EITC, for 12 months after receipt.¹¹ California already bars its state programs from counting payments from its guaranteed-income pilots.¹² Only a federal test, with federal data, can answer the national question. A free market asks workers to take risks and change course; the country should know what helps them do it before a downturn forces a guess.

The Solution

A three-step staircase: each step stands alone, and each step up adds federal money and commitment. Eligibility turns on income and a documented job loss or earnings drop; no one must prove that AI caused it. The tax treatment of labor and capital, and the measurement of AI's effects on work, are addressed separately.

Step 1 — Clear the cliffs. Direct federal benefit programs to disregard payments from approved income-support studies, public or philanthropic, for the life of the study, as they already disregard tax refunds. Preserve unemployment and disability eligibility, secure state coordination before enrollment, and offer paper and assisted applications. This costs almost nothing and lets the country learn from studies it does not pay for.

Step 2 — Build the evidence. Authorize secure, privacy-protected linking of tax and wage records to study participants, so any approved study can measure hardship, health, work, training, and new businesses without relying on surveys alone. Require pre-registered outcomes and publication of every result, favorable or not.

Step 3 — Run a national transition test. Enroll up to 20,000 adults, one per household, with income below 200% of the poverty line and an involuntary job loss or an earnings drop of at least 20% in the past year, including contract and self-employed workers, across labor markets with different levels of AI adoption. Pay $500 a month for 24 months, reduced by 25 cents per dollar of income above the entry threshold, and assign places by lottery with a consenting comparison group. No vouchers or approved vendors: cash trusts adults to know their own bills. Cost: $300 million, up to $240 million for payments and $60 million for administration and evaluation.

Where to start: Step 1 is the floor; it costs almost nothing. Step 3 is the heart: a real test with a comparison group, before anyone bets the federal budget on a guess.

Administration and enforcement: HHS leads, with Treasury making payments and Labor recruiting; enrollment begins within 18 months of appropriation. Participants get written explanations and appeals; overpayments are recovered proportionately, fraud is penalized, and independent evaluators publish interim and final reports. Any permanent program needs its own financing decision.

Risks and Mitigations

• Less work: Recipients in the randomized study did work 1 to 2 fewer hours a week.⁴ The payment is therefore temporary, phases down with income, and is measured against a comparison group. A 20,000-person test cannot reveal nationwide effects on wages or prices, and the evaluation must say so.
• Pilots become programs: Temporary benefits tend to outlive their purpose. Payments end after 24 months, any expansion waits for the final report, and anything larger needs separate financing. The political pressure remains.
• Benefit losses and errors: A payment that quietly cuts SSI or SNAP would hurt the people it means to help. Step 1 settles those interactions in statute before enrollment, because administrative promises cannot override another law.

Similar Bills

Fit measures similarity to this proposal's mechanisms: High = direct precedent; Partial = useful component with material differences; Related = adjacent approach.

Federal

State

What this adds: H.R. 5776 and California's pilot show how to pay and evaluate guaranteed income without cutting other benefits. This proposal aims the test at workers in transition, defines eligibility without requiring proof of AI displacement, and clears the benefit cliffs for every approved study, public or philanthropic.