This paper studies the causal effect of involuntary job loss on health using individual social security
data matched to health insurance records from Austria. We address the important issue of reverse
causality - deteriorating health leading to job loss - using worker displacement due to plan closure as
an instrument. The idea is that plant closure induces a sharp and strong change in the employment
career of the worker. However, deteriorating health status of workers is an unlikely cause of plant
closure. Second, we measure health status using data on actual usage of drugs instead of subjective
health status. Estimates suggest that job loss increases drug use strongly. Moreover, there are important
differences in the health effects of unemployment with respect to gender, age, stability of employment,
industry, and skill level. |