Advanced Econometrics II

1.5 Credits · Level: Graduate

A major recent trend in empirical social science has been a concerted effort to approximate the gold standard of scientific inquiry in many natural sciences: the experimental method. Experiments involve a scientist randomly manipulating a “treatment” in a carefully controlled environment in order to isolate its causal effect on dependent variables of interest. Experiments in social sciences and program evaluation have become increasingly common. At the same time, many social phenomena we care about cannot be experimentally manipulated for practical, ethical, and/or financial reasons. As an alternative, researchers have been developing a number of quasi-experimental techniques that attempt to approximate the true experiments. This class will give an overview of the design and analysis of social science experiments, as well as quasi-experimental methods such as matching, instrumental variables, differences-in-differences, and the regression discontinuity design. We will study both the theoretical underpinnings and read applied research employing these methods.