Bounded rationality, reciprocity, and their economic consequences

The conventional economic paradigm assumes that individuals make choices to maximize their utilities. However, the process by which they reach optimum decisions is not usually modeled. In my dissertation I take a close look at this process. I replace the assumption of instantaneous optimization by an assumption that individuals use a social learning algorithm. General properties of this algorithm and its economic consequences are studied.; I replace the assumption that an agent makes an optimal choice by an assumption that the agent executes a computational algorithm directed at finding a choice that is good at the time the choice is made. The algorithm utilizes the individual's own information, and information about choices of other members of society. It is shown that under some regularity conditions any such algorithm can be decomposed into generalized gradient dynamics, direct imitation, experimentation, and imitation of scope.; I consider in some detail a special case of the general model in which the individuals' behavior is governed by a noisy gradient adjustment. I show that there exists a unique steady state distribution of individual choices. It is shown that the steady state distribution of choices satisfies the Axiom of Independence of Irrelevant Alternatives only under very special conditions.; I also apply the general learning model to the design of an optimal incentive scheme. I show that it differs from an optimal incentive scheme under full rationality in several important ways. This model helps to shed light both on the under-provision of incentives that is a common empirical finding and on departure from the sufficient statistics result that underlies standard moral hazard theory. I also suggest that in environments in which bounded rationality plays an important role an incentive contract should be replaced by a contract based on reciprocity, provided that there are enough “reciprocal agents” in the population. To tie this to economic fundamentals I also develop an evolutionary model of reciprocity.