Winning the Visa Lottery

Michael Clemens and Ethan Lewis exploit a randomized processing lottery for H-2B visas - a program that admits temporary low-skill seasonal workers to fill landscaping, seafood processing, forestry, and hospitality jobs - to measure how limits on foreign low-skill labor affect firms and US workers. After a 2019 surge, the Department of Labor processed employer petitions in random order; firms drawing early letters typically hired the workers they requested while later-letter firms could not. The authors surveyed 472 firms that entered the 2021-22 lotteries, registered a pre-analysis plan, and treated lottery assignment like a randomized trial to estimate causal effects.

Winning the lottery roughly halved H-2B shortfalls and raised firm revenue (elasticity ≈ 0.20 for a doubling of H-2B employment) and sharply increased investment (elasticity 1.5-2.1). Losing firms did not replace foreign workers with US low-skill hires; across all firms the effect on American employment was zero or positive, and in a rural subsample a 1% increase in H-2B workers raised US employment by 0.61%. Estimated substitution elasticities between H-2B and US workers (0.8-2.2) show poor substitutability, meaning denied visas cause firms to contract rather than substitute labor. Raising the H-2B cap would therefore expand output, investment, and profits without reducing American employment, and chronic visa uncertainty likely causes further underinvestment and shutdowns.