From Blue to Steel-Collar Jobs: The Decline in Employment Gaps?

Abstract

I investigate how the introduction of industrial robots is shaping the demographic composition of the US labor force. Exploiting exogenous variation in robot exposure across local labor markets over time, I find that the adoption of robots between 1993 and 2014 decreased employment of men and women by 3.7 and 1.6 percentage points, contributing to the secular decline in the gender employment gap, and that it decreased employment among whites and non-whites by 1.8 and 4.5 percentage points, widening the race/ethnicity employment gap. I show that these effects are due to persistent occupational segregation in the US labor market, as men and non-whites are often employed in blue-collar jobs which require physical skills that can be automated. Although robots are primarily used in manufacturing industries, I find that their adverse effects also spill over to local service industries, in particular for Blacks and Hispanics.

Keywords: industrial robots, employment, gender, race and ethnicity

JEL Classification: J15, J16, J23