A study using the US government’s randomized handling of oversubscribed seasonal-work visa applications found that businesses denied foreign workers became smaller, while employment of lower-skilled American workers did not rise. The research, published in the July 2026 issue of the American Economic Journal: Applied Economics, examines the H-2B program for temporary non-agricultural jobs.
The program covers seasonal work in industries including landscaping, hospitality, forestry and seafood processing. Federal law generally caps H-2B admissions at 66,000 a year, split between two halves of the fiscal year. Demand can substantially exceed that supply: employers sought 136,555 workers for 33,000 places in the second half of 2022, according to the American Economic Association’s account of the study.
Economists Michael A. Clemens and Ethan G. Lewis used a change introduced after a 2019 rush of petitions overwhelmed a Labor Department server. The department began assigning applications to randomly ordered processing groups. Employers near the front were much more likely to obtain the workers they requested than employers assigned later positions. That created a natural comparison resembling a randomized trial.
The researchers surveyed 472 businesses that entered the lotteries in 2021 and 2022. They registered their hypotheses and methods before responses arrived, reducing the risk that results would be selected after the fact. Winning firms hired almost all of the requested H-2B workers, while losing the lottery cut foreign hiring by roughly half.
Access to those workers was associated with business expansion. The study estimated that doubling H-2B employment raised revenue by about one-fifth. Investment in equipment, vehicles and structures responded even more strongly. Yet firms that lost access did not compensate by hiring more lower-skilled Americans. Across the full sample, foreign hiring’s effect on US employment was zero or positive; among rural firms identified in advance for separate analysis, it was significantly positive.
The findings point to a scale effect. At a fixed level of output, an employer might substitute a domestic worker for a foreign employee. In practice, however, businesses facing a labor shortfall reduced operations. A seasonal company might serve fewer customers or operate for a shorter period, leaving less work overall rather than transferring the same amount of work to Americans.
The evidence concerns marginal changes to one temporary-worker program over a short period, not every form of immigration or every labor market. Within that scope, it challenges the assumption that denying seasonal visas automatically creates equivalent jobs for US workers. The authors conclude that additional H-2B access could expand production and investment without reducing American employment.
