Using surveys to measure information in auctions: Evidence for rational behavior, informative prices, and the importance of credibility in eBay computer auctions

This thesis examines the role of information in online auctions, with specific application to eBay auctions for computers. Market data is augmented by a survey-based measure of unobservable item values and information dispersion in the auctions. These measures are an important source of external information for hypothesis testing. They permit joint identification of the information structure (common or private values) and bidding behavior (Nash or naive strategies) in these auctions. They also permit separate identification of the effects of reputation, information, and their interaction (credibility of information) on price. This information can also be used to determine whether prices on eBay converge to the common value.; My estimates indicate that eBay auctions for computers are best described as common value auctions where prices reflect Nash equilibrium bidding behavior. Sellers with good reputations have powerful incentives to reduce uncertainty and promote efficient trade due to the importance of credibility of information. The ability of eBay bidders to account for the winner's curse in these auctions leads to prices that partially aggregate information about the common value as the number of bidders increases.; The success of the survey-based measure described in this thesis contributes to the tools available to empirical researchers. The quantitative results regarding the role of information in auctions yields a better understanding of the importance of information online auctions. These results are unattainable without employing theory, econometric modeling, and external survey data.