Publisher:
Northwestern University, Center for the Study of Industrial Organization (CSIO), Evanston, IL
Abstract:
Cooperative concepts of renegotiation in repeated games have typically assumed that Pareto-ranked equilibria could not coexist within the same renegotiation-proof set. With explicit renegotiation, however, a proposal to move to a Pareto-superior equilibrium can be deterred by a different continuation equilibrium which harms the proposer and rewards the refuser. This paper introduces a simple protocol of renegotiation for repeated games and defines the stability of social norms and renegotiation-proof outcomes in terms of a simple equilibrium refinement. We provide distinct necessary and sufficient conditions for renegotiation-proofness, which converge to each other as renegotiation frictions become negligible. Renegotiation-proof outcomes always exist and can be all included within a single, most permissive social norm that is straightforward to characterize graphically. The analysis suggests a novel mechanism explaining inefficient equilibria, such as mis-coordination, inertia, and failure to upset an oppressive status quo even when information is complete, communication is frictionless, and players can credibly agree on efficient outcomes.