Estimating Consumer Surplus in eBay Computer Monitor Auctions.  Tugba Giray Economics Department Bilkent University Ankara, Turkey Kevin Hasker Economics Department Bilkent University Ankara, Turkey Bibo Jiang Economics Department Rice University Houston, Texas Robin Sickles Economics department Rice University Houston, Texas July 28, 2009 Abstract Our paper utilizes semi-nonparametric and nonparametric methods to directly estimate consumer surplus in eBay computer monitor sales. We also use nonparametric methods to select among alternative paramet- ric models wherein the private values are assumed to have log-normal, gamma, Weibull, half-logistic, and Pareto distributions. We nd sub- stantial comparability between the estimated consumer surplus based on nonparametric methods and the parametric model selected using the non- parametric testing criteria. Although nonparametric estimators have their advantages in terms of robustness to distributional misspecication they require substantial sample sizes in order for standard asymptotic re- sults to hold when observed covariate e/ects must be estimated. This suggests that at least for our eBay data, parametric models chosen in this way may o/er advantages that can overcome their potential lack of distributional robustness. JEl classication : C22, C51 Key words and phrases : eBay auction, semi-nonparametric estimation, nonparametric estimation, consumer surplus, consumer share of surplus  We would like to thank Sandra Campo and participants at the Conference on Auctions and Games, Virginia Polytechnic Institute and State University, Blacksburg, Va., October 12- 14, 2007 as well as Ravi Bapna and participants of the FTC Roundtable: Economics of Internet Auctions, October 27th, 2005, Washington, DC. for insightful and needed criticism on earlier drafts of this paper. The usual caveat applies. 1