Structural Analysis of Tullock Contest with an Application to U.S. House of Representatives Elections

We study the econometrics of an asymmetric Tullock contest model with incomplete information and establish nonparametric identification of the distributions of players’ private cost signals. We propose estimators for the quantile functions, establish their consistency, and develop an asymptotic inference procedure. Monte Carlo experiments demonstrate the good finite sample performance of the estimators and confidence intervals. We employ this newly developed method to analyze data from U.S. House of Representatives elections. The structural estimation results provide insights into the technological advances of the past four decades and quantify the incumbency advantage and the partisan advantage.