Implementing the Nash Program in Stochastic Games
35 Pages Posted: 20 Jul 2013
Date Written: June 23, 2013
Abstract
Nash' noncooperative and cooperative foundations for "bargaining with threats" are reinterpreted to achieve equilibrium selection in infinitely repeated two player games. The analysis is then extended to stochastic games, where players' choices affect the state transition matrix. Sufficient conditions on the exogenous structure of the game are provided that ensure a unique division of surplus in the stochastic game, supported by both an axiomatic and a noncooperative analysis. Some comparative dynamics results for simple classes of games illustrate the dynamic programming principles governing how bargaining power in a subgame is transferred to the preceding period, and affects behavior in that earlier period. An example illustrates the surprising potential for a bargainer to extort resources from an apparently stronger competitor.
Suggested Citation: Suggested Citation