Abstract

Recent auction theory suggests that multi-unit uniform-price auctions, as used by the U.S. Treasury for debt sales, produce incentives that may cause bidders to bid less than their true valuations, resulting in inefficient allocations and reduced revenue. In this paper, we present the results of a field experiment in which we auction nearly $10,000 worth of sportscards in two-unit, two-person sealed-bid auctions. We randomize participants into uniform-price and Vickrey auction treatments, and find underbidding in the uniform-price auctions' second-unit bids, as predicted. In contrast with theoretical predictions, however, we find that individual's first-unit bids are significantly higher in the uniform-price than in the Vickrey treatment. The bid differences are large enough to affect the allocation of goods, as split allocations result significantly more often in the uniform-price treatment. We find no significant difference in revenues across auction formats.

Citations