Here we study the economic effects of a large-scale experiment in the United States: nationwide, firm-level, natural randomization of restrictions on the employment of immigrants for low-skill jobs. The United States has one principal work visa for low-skill labor in the nonfarm economy— the H-2B visa. U.S. employers’ access to that visa is limited by a quota and allocated in part via a randomized lottery conducted by the federal government. This exogenous variation in restrictions on immigrant employment allows unusually transparent, policy-relevant estimates of how U.S. firms and workers adjust. After publicly committing to our hypothesis tests and predicted treatment effects with a pre-analysis plan, we collected data from both winners and losers of the 2021 H-2B visa lottery in a novel firm survey. This allows prespecified tests of basic theoretical predictions about the magnitude and hetergeneity of the effect of low-skill immigration restrictions. It furthermore allows estimation of the firm-level, immigrant-native “combined” elasticity of substitution (Hicks 1936).

We find that exogenous permission to employ immigrants for low-skill work causes the marginal firm to expand production. Put differently, exogenous restrictions on employing the profit-maximizing number of immigrants for low-skill work cause the marginal firm to contract. These restrictions cause a large and statistically significant decrease in revenue and investment. The restrictions cause no increase, or a decrease, in the employment of low-skill native workers and the rate of profit. Losing the lottery reduces firms’ employment of low-skill immigrants by 56%. This decrease causes firms to contract, reducing operations with an elasticity of +0.164 for revenue and +1.03 for investment (statistically distinguishable from zero at conventional levels), and with an elasticity of +0.102 for low-skill U.S. employment, and +0.100 for the profit rate (statistically indistinguishable from zero at conventional levels).

This is from Michael A. Clemens & Ethan G. Lewis, “The Effect of Low-Skill Immigration Restrictions on US Firms and Workers: Evidence from a Randomized Lottery,” NBER Working Paper #30589, October 22, 2022.

The methodology is quite clever. Because the lucky employers are chosen by lottery, there’s no selection bias. That means that Clemens and Lewis can look at changes in employment of employers that won the lottery and changes in employment of employers that lost.

The authors put it more succinctly in their abstract:

Firms exogenously authorized to employ more immigrants significantly increase production (elasticity +0.16) with no decrease or an increase in U.S. employment (elasticity +0.10, statistically imprecise) across several pre-registered subsamples. The results imply very low substitutability of native for foreign labor in the policy-relevant occupations.

In short, low-skilled Americans are not losing their jobs to low-skilled migrants.

HT2 Tyler Cowen.

Update: I was informed that these workers are NOT immigrants. So I changed “Immigrant” to “Migrant” in the title.