Private Enforcement against Collusion in Mechanism Design

This paper brings a new point of view into the theory of collusion-proof mechanism design, which highlights the principle of divide and conquer. We relax the restriction of publicly enforced grand contract in the framework of Laffont-Martimort-Itoh, which allows us to incorporate the approach of private enforcement into the theory. In a setting of moral hazard with mutually observable actions, we develop a multi-stage mechanism integrated with secret reporting and private transferring and show that the first-best allocation can be implememted in spite of collusion, which implies that preventing collusion entails no cost under new approach