Colluding on Participation Decisions

We study the principal’s optimal response to collusion in an adverse selection environment. Building on the framework of Laffont and Martimort (1997, 2000) we advance it into several directions. First, unlike most of the literature, we study a stronger collusion when the agents can coordinate their participation decisions in addition to the joint play of the mechanism. Second, we make precise the economic nature of the transaction costs due to the presence of the asymmetric information in the collusive problem. Finally, while most of the literature focused on the two-type distributions, in our model the private information is distributed continuously. We characterize the set of implementable collusion-proof outcomes of the game, and relate it to the interim incentive-efficiency, which is a standard concept of efficiency under asymmetric information. After that we solve the principal’s problem by optimizing over a restricted set of mechanisms, which turn out to be optimal in a class of cases and allow the principal to fight collusion off at no cost.