Focal points and bargaining

This is a contribution to the growing experimental literature on how trial-and-error adjustment processes can establish a convention for coordination on an equilibrium in a game. A simple bargaining game introduced by Nash is used for this purpose. Subjects are conditioned in different treatments to use four different bargaining solutions. The stability of the conditioning is then studied as the bargaining game is played over an extended period. The data obtained is unusually sharp. In the long run, the median subject behaves as though optimizing often down to a fraction of a penny. The results are therefore not supportive of the view that strategic considerations in such situations can be neglected in favor of a study of fairness norms. Indeed, the equilibrium actually achieved in a session turns out to be a very good predictor of what the median subject says is fair in the game after play is over.We are grateful to the National Science Foundation for funding the experiments reported in this paper under Grant NSF-SES-8821521. We also gratefully acknowledge funding from the University of Michigan to set up the Michigan Economics Laboratory, where the experiments were conducted. We would also like to thank Richard Stallman and the Free Software Foundation for developing EMACS, Luke Tierney for developing LISPSTAT, and Hal Varian for showing us how to use both.