Affiliation:
1. Department of Economics, University of Warwick and CEPR, Coventry, UK
2. Department of Politics, Princeton University, Princeton, NJ, USA
Abstract
The best empirical research in political science clearly defines substantive parameters of interest, presents a set of assumptions that guarantees their identification, and uses an appropriate estimator. We argue for the importance of explicitly integrating rigorous theory into this process and focus on the advantages of doing so. By integrating a theoretical structure into one’s empirical strategy, researchers can quantify the effects of competing mechanisms, consider the ex-ante effects of new policies, extrapolate findings to new environments, estimate model-specific theoretical parameters, evaluate the fit of a theoretical model, and test competing models that aim to explain the same phenomena. As a guide to such a methodology, we provide an overview of structural estimation, including formal definitions, implementation suggestions, examples, and comparisons to other methods.