This research consists of three projects investigating how economic agents use their past experience to guide their current decisions. The first research project studies agents who use very simple rules of thumb in choosing between alternative technologies; these rules will incorporate the idea that agents try to learn from the experiences of their neighbors. The project will investigate the long-run implications of such "social learning," and in particular on when simple rules can lead to efficient long-run outcomes. The second project studies learning in games. Here the focus is on when repeated interaction will lead play to correspond to a Nash equilibrium or to one of its refinements. The third project will develop a model of optimal contracts between a firm's owners and its manager in a setting where the owners learn about the firm's future prospects from its current performance. The model will provide an explanation for the widely-observed practice of "income smoothing," under which the manager takes actions that increase reported income when income is low, and decrease reported income when income is high. This research is important because it tackles and helps provide answers to a few of the most important issues in the economic implications of learning.
