Empirical Relevance of Ambiguity in First-Price Auctions
We study the identification and estimation of first-price auctions with independent private values if bidders face ambiguity about the valuation distribution and have maxmin expected utility. We exploit variation in the number of bidders to nonparametrically identify the true valuation distribution and the lower envelope of the bidders’ set of priors. We extend the identification result to allow for constant relative risk aversion and unobserved auction heterogeneity that can be correlated with the number of bidders. We propose a flexible Bayesian estimation method based on Bernstein polynomials, and our Monte Carlo experiments show that the method estimates the parameters precisely and chooses reserve prices with nearly maximal revenues – whether there is ambiguity or not. Incorrectly assuming no ambiguity may induce estimation bias and loss of revenue. We apply our method to a sample of U.S. timber auctions and find evidence of ambiguity in auctions from the Pacific Northwest region.