Undermining Collusion in Lowball Lottery Auctions.

Abstract

This paper expands on the lowball lottery auction mechanism proposed in Optimal Procurement with Quality Concerns by exploring how the buyer can reduce the likelihood of collusive equilibria in repeated procurement auctions with exactly two bidders. In this paper, I introduce a novel approach to policing collusion in LoLAs: a reward-for-reporting mechanism. Under this mechanism, both bidders are incentivized to report any attempts at collusion in exchange for a monetary reward, which is funded by amounts they are both required to set aside at the start of the auction. The paper compares this approach to a more traditional method in which the auctioneer independently decides when to investigate potential collusion, with some probability of uncovering it, and to a baseline scenario with no anti-collusion enforcement. I then analyze the conditions under which collusion can be ruled out as an equilibrium and show that, in theory, the reward-for-reporting mechanism offers benefits similar to traditional investigations but with the added advantage of imposing no cost on either the bidders or the auctioneer in equilibrium.