ABSTRACT
Coordination and equilibrium selection are central to many questions in economics, from the determination of bargaining outcomes to the design of incentive schemes, the efficacy of implicit contracts, the influence of expectations in macroeconomics, and the nature of competition in markets. 1 Such questions are usually modeled as noncooperative games with multiple Nash equilibria – combinations of strategies such that each player’s strategy maximizes his expected payoff, given the others’ strategies. The analysis of such games must address the issue of equilibrium selection, the determination of which, if any, equilibrium should be taken to represent the model’s implications. This issue is particularly acute in games with multiple strict equilibria – those in which players strictly prefer playing their equilibrium strategies to deviating. In applications involving such games, equilibrium selection is often done by introspection, goodness of fit, custom, or convenience. Yet this begs the questions of how players come to expect a particular equilibrium and what they do when their expectations are less sharply focused, limiting the insight the analysis can give into how the environment influences coordination outcomes.
