The Causal Effect of Job Loss on Health: The Danish Miracle?

Unemployment can affect health both through the income shock and through non-pecuniary channels like the loss of self-esteem or the loss of a structured schedule. I investigate whether there is still a causal effect of job loss on health when the unemployment risk is well-insured by policy through generous unemployment insurance, active labor market policies and public health insurance with universal coverage. In a difference-in-difference design, I compare roughly 25,000 workers who are in an establishment that closed in Denmark between 2001 and 2006 to a control group of workers matched on lagged observables. I find that in such a setting job losses driven by establishment closures do not cause any significant effect on health, whether looking at mental health proxies such as antidepressant purchases, severe physical health outcomes that require inpatient care or mortality. I can rule out effects of the order of 2% for most health outcomes and 15% for mortality. My results taken together with prior literature suggest that it is possible, presumably through an adequate set of policies, to make the causal effect of job loss on health negligible.