We analyze a sequential decision model in which each decision maker looks at the decisions made by previous decision makers in taking her own decision. This is rational for her because these other decision makers may have some information that is important for her. We then show that the decision rules that are chosen by optimizing individuals will be characterized by herd behavior; i.e., people will be doing what others are doing rather than using their information. We then show that the resulting equilibrium is inefficient.

Laboratory data from bargaining experiments have started a debate about the prospects for various parts of game theory as descriptive theories of observable behavior, and about whether, to what extent, and how a successful descriptive theory must take into account peoples' perceptions of “fairness.” Plausible explanations of the observed bargaining phenomena advanced by different investigators lead to markedly different predictions about what should be observed in three different games. A sharp experimental test is thus possible on this class of games, and the present paper reports the results of such a test.

We present experimental evidence on nonbinding, preplay communication in bilateral coordination games. To evaluate the effect of "cheap talk, " we consider two communication structures (one-way and two-way communication) and two types of coordination games (one with a cooperative strategy and a second in which one strategy is less "risky"). In games with a cooperative strategy, one-way communication increases play of the Pareto-dominant equilibrium relative to the no communication baseline; two-way communication does not always decrease the frequency of coordination failures. In the second type of game, two-way communication always leads to the Pareto-dominant Nash equilibrium, while one-way communication does not.