On the Informed Principal Model with Common Values
20 Pages Posted: 29 Sep 2016 Last revised: 13 Dec 2017
Date Written: December 13, 2017
This paper reconsiders the general informed principal model with unilateral private information and common values and makes the following contributions. First, it provides an alternative, perhaps simpler, proof of Theorem 1 (i.e., the main result) of Maskin and Tirole (1992). Second, it shows that if the principal is restricted to offering mechanisms in which only she makes announcements (e.g., direct revelation mechanisms), then the Rothschild-Stiglitz-Wilson allocation and every allocation that weakly dominates it relative to the prior beliefs (if such an allocation exists) constitutes an equilibrium allocation of the three-stage game even in environments in which this is not interim efficient relative to any non-degenerate beliefs. Finally, it provides a sufficient condition that allows for the complete characterisation of the set of equilibrium allocations even in environments in which sorting is not satisfied.
Keywords: Mechanism design, informed principal, Rothschild-Stiglitz-Wilson allocation, perfect Bayesian equilibrium
JEL Classification: C72, D82
Suggested Citation: Suggested Citation