Essays in applied dynamic microeconometrics

The ideal approach in microeconomics to evaluate the effect of a policy is to conduct a randomized experiment. As its name suggests, this approach randomly assigns some individuals - the treatment group - to some policy measure and others - the control group - to a baseline without the policy. The effect of the policy is given by the difference in some outcome(s) of interest between these two groups at a later point in time.
This thesis focuses on the evaluation of policies when the usual static framework of a randomized experiment cannot or does not apply perfectly. The first two chapters look at microeconometric evaluation methods to analyze dynamic situations when the policy assigns treatment at different times and the measurement of the outcome of interest also depends upon a time dimension. The last two chapters of this thesis are empirical crime studies on tax evasion and street prostitution which use panel data methods.