Outsourcing Policy and Worker Outcomes: Causal Evidence from a Mexican Ban

Abstract

Using Mexican economic census data from 1994 to 2019, we document a rising trend in domestic outsourcing, particularly among large firms, and a negative association between outsourcing and labor compensation, including profit sharing and social security. We leverage higher-frequency data from a manufacturing panel survey, matched employer–employee data, and a ban on domestic outsourcing in 2021 to show that the ban reduced outsourcing, increased labor’s share, and reduced markdowns without raising total labor costs or affecting employment, output, or productivity. We propose a theoretical model in which corporate fiscal incentives drive outsourcing and account for the observed empirical patterns.