A Two-Person Repeated Bargaining Game with Long Term Contracts

Abstract

The folk theorem in the theory of supergame states that the average payoffs of Nash equilibrium points in a supergame are the feasible individual rational payoffs in the component game. The chapter discusses the question whether the possibility of negotiations and of binding agreements enables players to attain a pareto-efficient outcome in every equilibrium point of a supergame. For this purpose, there is a two-person repeated bargaining game based on the supergame model in which players can negotiate for their actions at every period before they choose their own actions independently. In negotiations, the players are allowed to reach a long term contract on actions. A contract includes an action pair that they should choose and a number of future periods for which they should take it. Once they reach an agreement, it is assumed to be enforced. The chapter presents theorems showing that players do not attain a pareto-efficient outcome in every equilibrium point of repeated bargaining game with the strong institutional assumption that any agreement on actions is binding. A behavioral assumption that players do not punish each other for their past behavior is necessary for cooperation.