From efficiency to illness: do highly automatable jobs take a toll on health in Germany?

Automation transforms work at a rapid pace, with gradually increasing shares of the workforce being at risk of replacement by machines. However, little is known about how this risk is affecting workers. In this study, we investigate the impact of exposure to a high risk of automation at work on the subjective (self-reported health, anxiety, and health satisfaction) and objective (healthcare use and sickness absence) health outcomes of workers in Germany. We base our analysis on survey data from the German Socio-Economic Panel (SOEP) and administrative data from the Occupational Panel for Germany (2013-2018). Employing panel regression, we demonstrate that for workers, exposure to a high risk of automation at the occupational level is associated with lower self-reported health and health satisfaction, increased sickness absence, and, depending on how the risk is measured, anxiety. No effect on healthcare use is found. Our heterogeneity analysis provides evidence that none of the analyzed demographic and occupational groups is disproportionally affected by high automation risk. We also conduct several robustness checks (i.e., alternative model specifications and risk measures with different thresholds), with the results remaining largely consistent with our main findings.