An Experimental Investigation of Buyer Determined Procurement Auctions

Abstract

ABSTRACT. Popular procurement auction formats vary along two main dimensions: (1) whether the winner is determined by the lowest price or by the buyer, and (2) whether auctions are sealed bid or dynamic. The attractiveness of these formats critically depends on information available to bidders about each other's qualities. We find that dynamic bidding with the buyer determined format in full information settings is detrimental to buyers. This is because bidders stop bidding short of the theoretical prediction, and this behavior has the greatest impact in the full information dynamic buyer-determined environment. This finding provides evidence that buyers should prefer sealed bid over dynamic auctions, especially if there is reason to believe that suppliers possess information about one another's quality. Buyers also should avoid revealing quality information to bidders (through individual reserves, for example). A methodological contribution of our approach is that it allows us to arrive at useful bounds for formats that are not well understood theoretically.

Keywords: Bidding, Procurement, Reverse Auctions, Multi-Attribute Auctions, Behavioral Game Theory, Experimental Economics

JEL Classification: D44, C91