Research by Michael A. Clemens and Ethan G. Lewis in the American Economic Journal: Applied Economics used a randomized H-2B visa lottery to study low-skill immigration effects. The H-2B program authorizes temporary foreign workers for seasonal jobs like landscaping and seafood processing, with 98% requiring no high school education. The 1990 Immigration Act capped visas at 66,000 annually, but demand far exceeds supply — 136,555 workers were requested for just 33,000 slots in late 2022.

After a Department of Labor server crash in 2019, petitions were processed in randomized alphabetical order, creating a natural experiment. Surveying 472 businesses from the 2021-2022 lotteries, researchers found losing the lottery cut H-2B employment roughly in half. Winning firms saw revenue rise with a 0.20 elasticity, while investment in equipment and structures responded even more strongly (elasticity 1.5 to 2.1).

Crucially, employment of low-skill American workers at losing firms did not increase. The effect on US employment was zero or positive, and significantly positive for rural firms (elasticity 0.61). The elasticity of substitution between H-2B and US workers was estimated at 0.8 to 2.2, far below typical findings of 4, suggesting they are poor substitutes. As Lewis noted, restricting immigrant workers simply shrinks firms rather than creating American jobs.