Auction–Lottery Hybrid Mechanisms: Structural Model and Empirical Analysis

Abstract

This article studies an auction–lottery hybrid mechanism that is widely adopted in allocating new vehicle licenses in China. We characterize individuals entry and bidding strategies in a symmetric Bayesian Nash equilibrium, structurally estimate individuals value distribution from a data set of a Guangzhou program, and evaluate the performance of the mechanism. Based on the estimated distribution and counterfactual analysis, our study suggests that a hybrid mechanism preserves 83% efficiency and 52% of revenue, while it improves equity by 25 times compared to a pure auction. We show that allowing auction losers to participate in the lottery can further enhance the performance.