Unanimous Implementation: A Case for Approval Mechanisms

Abstract

We design the class of (indirect) Approval mechanisms which allow the players' strategies to coincide with the subsets of the outcome space. By focusing on the single-peaked domain, we prove that: a) Each of these mechanisms is characterized by a unique equilibrium outcome, and b) for almost every implementable welfare optimum (outcome of a social choice rule), including the Condorcet winner alternative, there exists an Approval mechanism that unanimously implements it. That is, Approval mechanisms help a society achieve every feasible welfare goal, and, perhaps more importantly, they promote social coherence: The implemented outcome is approved by everyone.