Essays on causal inference and applied health economics

This dissertation consists of three self-contained essays concerned with estimating causal relationships with microeconometric methods. The goal of causal inference is to draw conclusions about the causal nexus between distinct but possibly related processes or events, the cause and the effect. If a causal link between the two events exists, the effect originates from the cause, and testing the existence and the strength of this link is the aim of causal inference. Causation analysis differs from association analysis, in that the former aims to make inference regarding probabilities and beliefs under dynamic conditions, i.e. changes, and not only static ones. This means that the identification of a causal relationship typically requires a variation in both the cause and the effect, while all the other confounding factors being controlled for or remaining invariant.