STRUCTURAL ANALYSIS OF TULLOCK CONTESTS WITH AN APPLICATION TO US HOUSE OF REPRESENTATIVES ELECTIONS

We study the econometrics of an asymmetric Tullock contest model with incomplete information. Identification results are established for parameters in the asymmetric contest success function and for the quantile functions of players' private costs. We propose the corresponding estimators, derive the asymptotic properties, and demonstrate their good finite sample performances by simulation. We also propose an easy-to-implement parametric estimation method that can incorporate contest-specific covariates. Our method is used to study U.S. House of Representatives elections. We quantify the incumbency advantages and conduct policy experiments to evaluate how imposing a term limit or launching campaign finance reforms affects incumbency advantage.