Mechanism Design and Direct Communication in Collusion

2I am thankful to my main advisor professor Steven Matthews for very detailed feedback and patient support. I am grateful to my dissertation committee members professor George Mailath and Nicola Persico for incisive comments. I also thank Jing Li, and Ichiro Obara, and audiences at the Economic Theory workshop at University of Pennsylvania. All errors are my own responsibility. This paper studies a principal-supervisor-agent model in which a privately informed supervisor is susceptible to collusion. We model collusion as a side mechanism pro-posed by the supervisor who observes a private signal correlated with the agent’s type. We explore how the supervisory information helps the principal to extract informa-tion rent from the agent when the informational asymmetry between the supervisor and agent results in inefficient collusion on their part. We first analyze an informed principal problem with correlated signals in our set-ting, and show conditions under which the existence of an informed principal problem does not affect the optimal design of the grand mechanism. Under those conditions, the optimal mechanism is the same as when collusion is formed by an uninformed benevolent outsider. The second part of this chapter focuses on the supervisor’s participation decisions once she obtains new information in collusion. The principal can take advantage of the implicit information revealed by the supervisor’s exit decision when the coalition has limited ability to control it. If the coalition cannot write a contract based on the exit decision, or the principal can distinguish the timing of the supervisor’s exit decisions, the principal using a simple mechanism can attain the outcome of the direct supervision benchmark in which she can directly observes the supervisor’s signal