This project involves the design and execution of laboratory experiments on cooperation in finitely repeated games. Particular conditions to be examined include nonbinding communication ("cheap talk"). Other experiments are designed to investigate the relationship between cooperation levels and the payoff properties of outcomes that serve as possible punishments or rewards. Game theorists have devoted considerable attention to models of equilibrium cooperation in dynamic, multi-stage games. The conditions thought to be necessary for such cooperation to occur are affected by market and individual characteristics that alter the structure of the game. This project examines the behavioral significance of these characteristics. Such an assessment is important for future theoretical work. In addition, insights gained from laboratory data may aid in predicting when cooperative outcomes will arise in naturally occurring markets. Many policy debates turn on assumptions about the extent to which economic agents behave cooperatively. The desirability of challenging a horizontal merger, for example, depends largely on the likelihood of a post-merger conspiracy. Similarly, when product quality is unobserved prior to purchase, consumer-protection policies may be unnecessary if consumers and producers cooperate.
