Four Essays on Causal Inference in Health Economics: Dissertation

Abstract

This cumulative dissertation consists of four papers. Each paper makes an independent contribution to the literature by investigating the impact of one particular factor on health. The first paper shows that job loss increases body weight (slightly) and the risk of starting smoking (considerably). The related second paper reveals that unemployment decreases the mental health of spouses almost as much as for the directly affected individuals, and that, in general, the decreases in mental health are larger when the husband enters unemployment. The third paper finds that adolescents tend to smoke more often, do less sport, and are more frequently overweight, the lower their mother's education. The paper provides evidence that at least some of these health-related differences can be causally attributed to the mother's education. Furthermore, the paper demonstrates that social differences are already reflected in the health of newborn babies. The fourth paper shows that a ban on late-night off-premise alcohol sales in the German federal state of Baden-Württemberg decreased alcohol-related hospitalizations among adolescents and young adults by about 9 %. In order to establish causal relationships between the various factors and the health outcomes, this dissertation applies a wide range of different microeconometric evaluation tools (e.g., propensity score matching, difference-in-difference, instrumental variables estimation, entropy balancing, fixed effect panel estimation) and employs different data sets (e.g., survey data from the German Socio-Economic Panel (SOEP) and administrative data from hospital diagnosis statistics). Taken together, the findings of this dissertation help to understand which non-genetic factors actually have an impact on individuals' health and why some people are healthier than others.