Implementation with State Dependent Feasible Sets and Preferences: A Renegotiation Approach

In this paper we present a model of implementation based on the idea that agents renegotiate unfeasible allocations. We characterize the maximal set of Social Choice Correspondences that can be implemented in Nash Equilibrium with a class of renegotiation functions that do not reward agents for unfeasibilities. This result is used to study the possibility of implementing the Walrasian Correspondence in exchange economies and several axiomatic solutions to problems of bargaining and bankruptcy.