Robots, Natives and Immigrants in US local labor markets

Abstract

I analyze the impact of industrial robots on the employment of natives and immigrants in US local labor markets between 1990 and 2014. The proposed mechanism, through which robot adoption affects the employment of natives and immigrants differentially, is based on two facts: first, robots tend to displace workers based on the task content of occupations, and second, natives and immigrants in the US differ in their task specialization. Therefore, robots should affect their employment unequally. Exploiting plausibly exogenous variation in robot exposure across US local labor markets over time, I test this mechanism and find that the effect on immigrants is roughly 1.76 times greater than that observed for natives. Specifically, I find that one more robot per thousand workers reduces the employment-to-population ratio of natives and immigrants by 0.38 and 0.67 percentage points, respectively. I attribute these results to the fact that immigrants specialize in jobs or tasks at risk of being automated.