Identifying Demand in Ebay Auctions

Abstract

This paper presents assumptions and identification results for eBay type auctions. These results are for private value auctions covering three major issues; censoring bias, auction heterogeneity and dynamic bidding. The first section of the paper presents two identification results for second price open call auctions with private values and unobserved participation (eBay type auctions). The second section presents identification results for eBay type auctions that have either observed bidder heterogeneity, observed and unobserved item heterogeneity or unobserved auction heterogeneity. In particular it is shown that a traditional demand estimation model is identified. The third section presents identification results when bidders face an infinite sequence of eBay type auctions for a single item.

Keywords: Internet auctions, non-parametric identification, censoring bias, demand estimation

JEL Classification: C14, D44