ABSTRACT
This chapter considers the various research designs and methods used to identify causal effects, highlighting the data requirements and assumptions needed for the different methods. In recent years, applied health economics has developed an increasing focus on the identification of the causal impact of an intervention (e.g. a policy, a treatment) on an outcome of interest. It focuses on ex post evaluation, that is, what was the impact of the policy or treatment on a particular outcome? There may be situations, however, where the likely impact of the introduction of an intervention is required, and in this case, techniques such as microsimulation may be appropriate. The chapter also focuses on reduced-form estimation of key relationships; in many cases, econometricians estimate structural models which allow for greater integration of theory and empirical evidence. Difference-in-difference (DID) is one of the most common methods for evaluating the effect of a policy intervention.
