The effects of sudden health reductions on labor market outcomes: Evidence from incidence of stroke

While increasing attention is given to how health reductions affect workers, estimating its their effects is usually challenging. This paper aims to identify the causal effect of health deterioration on labor market outcomes by exploiting the incidence of stroke. Stroke, which often reduces health suddenly and unexpectedly, allows us to exploit the within-person random variation of the timing and isolate the effects of health reduction. By applying the fixed-effects method to a sample of stroke survivors in the University of Michigan Health and Retirement Study data, I find that stroke immediately halves the employment probability as well as hours and weeks worked 1 year after the occurrence and its effects persist for at least 3 years, while earnings reduction is relatively moderate and gradual. I also find the negative effects of stroke are larger among men with severe stroke and women with longer pre-stroke job tenure, while the effects are mitigated for younger women. These results make a stark contrast with the studies on other health events such as cancer diagnosis, which generally find much smaller effects and significant heterogeneity by education and occupation. This analysis shows that the labor market effects largely differ by types of diseases and calls for disease-specific studies in order to understand the social gradient in health and how workers adjust to work limitations.