Data and code for: Robot Hubs and the Use of Robotics in US Manufacturing Establishments

In this paper, we present results on the distribution of robots in US manufacturing by establishment characteristics using new establishment-level data collected by the US Census Bureau’s ASM for reference year 2018. This establishment-level analysis of the use of robots in US manufacturing complements results on “robot hubs” previously reported in Brynjolfsson et al. (2023a). We find that establishments with robots tend to have a larger share of production workers, lower average wages, and lower labor share, controlling for industry, age, and size. We also find that establishments with robots tend to be located in areas with other robot-adopting establishments, controlling for industry, location, size, and age. These results suggest that there may be peer effects in robot adoption, which may stem from learning or other agglomeration effects. Given the cross-sectional nature of the data—the data are currently available for 2018, but additional years will soon be available—the correlations we observe are not necessarily causal. Nevertheless, the patterns in the data provide useful information about the distribution of robots across establishments. The patterns we find raise multiple questions for future research. For example, does the minimum efficient scale for robot adoption vary with establishment characteristics? What factors lead to agglomeration effects; is it the presence of specialized human capital, important third parties such as robotics integrators, or some type of learning across establishments and firms? Our hope is that the patterns in the data that we document in our paper spark further research in this area and are of use to scholars, practitioners, and policymakers.