A Simpler Mechanism That Stops Agents from Cheating

This note considers a principal–multi-agent model of a firm subject to adverse selection. With just the usual optimal (incentive-constrained) contracts being offered, there exist multiple (Bayes–Nash) equilibria in the agents' subgame. Moreover, from the agents' perspective, there exists an equilibrium that Pareto-dominates the equilibrium desired by the principal. By exploiting the structure of the model, this note develops a new approach for eliminating unwanted equilibria (while retaining the desired equilibrium). The approach, when compared to existing approaches, employs a simpler mechanism (one with a smaller message space) and makes weaker assumptions about the agents' behavior.