Brookings Paper Proposes $60 Billion a Year Plan to Cushion AI Job Losses

Three complementary programs

Litan's paper, published by Brookings, lays out three proposals: wage insurance, retraining assistance and entry-level tax credits. According to the author, each addresses a different dimension of the problem, so they work as complementary measures that together could serve as insurance for the economy and its workers against the downside of AI-driven job and income losses.

Wage insurance would reimburse displaced workers for part of their income loss for a limited period after they take a new but lower-paying job. Retraining help could come through government lifetime retraining accounts for individuals, tax credits for employees or employers who provide training, or both. The third idea is to subsidize the hiring of entry-level workers, who would not qualify for either wage insurance or retraining assistance.

Litan says these ideas have so far been advocated only in general terms, with minimal details and no cost estimates. The report aims to fill those gaps and to propose ways to pay for the programs.

Why public insurance, and why not UBI

The author notes that public concern about AI remains high. He cites a 2026 Gallup poll in which 71% of Americans opposed local AI data center construction, and says that in 2025 nearly equal shares of Republicans (50%) and Democrats (51%) were more concerned than excited about AI in daily life. In July 2026, over 1,000 economists signed an online petition organized by Stanford's Erik Brynjolfsson urging action on possible labor market disruption from AI.

Litan argues that a government role is justified because private insurers are unlikely to cover this risk, for two reasons. First, there is no actuarial basis for setting premiums, since AI-induced displacement could differ sharply from past technological change. Second, insurers avoid highly correlated risks, and AI job losses would likely be too correlated across people and regions.

He also rejects universal basic income as too expensive. A UBI with a poverty-level floor has been estimated at a net annual cost of approximately $784 billion after accounting for taxes clawed back, as of 2025. Such a program, he writes, would also do little to ease the anxieties of higher-income white-collar workers whom AI may displace.

Building on earlier wage insurance work

The report traces wage insurance to a 1984 Brookings chapter in "Economic Choices" co-authored by Robert Lawrence. In a 1986 book, "Saving Free Trade," Litan and Lawrence proposed payments of up to 50% of income loss for trade-displaced workers once they found a new job. It was meant as a supplement to unemployment insurance, encouraging faster reemployment.

Fifteen years later, Litan and Lori Kletzer proposed a universal version for all workers involuntarily unemployed for reasons other than cause, with at least two years of tenure. It would cover 50% of the income loss on taking a new full-time job, up to $10,000 a year. They estimated that serving approximately 800,000 eligible workers with an average 22% wage cut would cost approximately $3 billion at a national unemployment rate of 4.9%.

The excerpt available ends as Litan begins describing a later collaboration with Lael Brainard, so the updated cost details are not covered here.

What it means

Litan frames the package as a precaution rather than a prediction. Even if worries about AI job losses prove unfounded, as he says they so far appear to be, he argues it makes sense to have policies ready to ease a transition. The roughly $60 billion annual figure gives policymakers a concrete benchmark, though it rests on what the author calls a reasonable set of assumptions about a sustained 6% unemployment rate.