This award supports collaborative research (with SBR-9809208) which deals with the evolution of preferences in strategic environments. Our approach to the evolution of preferences is characterized by the following features: (i) a focus on preference interdependence rather than on other aspects of preferences such as attitudes toward time and risk; (ii) a focus on strategic rather than decision-theoretic environments; (iii) the use of a general analytical framework which encompasses a variety of spatial interaction and matching models as special cases; (iv) consideration of the simultaneous presence of distinct kinds of interdependent preferences in the population, and (v) the pursuit of general results rather than the provision of suggestive examples. An individual is said to have interdependent preferences if her (subjective) well-being is either enhanced or diminished by changes in the (objective) material well-being of others. Preference interdependence may be positive, as is the case with altruism, or negative, as is the case with envy or spite. Interdependent preferences have often been used to explain empirical and experimental findings that are anomalies from the perspective of the more standard postulate of independent preferences. Few attempts have been made to account for interdependent preferences in the economics literature, and those that have tend to focus on simple stylized environments. It is now well understood that the type of interdependence that emerges in such models is inextricably tied to the particular environment that is assumed, with different environments giving rise to different forms of interdependence. This makes it difficult to assess the robustness of any particular model of preference interdependence. What is missing in the literature is a comprehensive analysis of the general properties of environments that give rise to particular forms of preference interdependence. It is this gap in the literature that we attempt to fill. We develop an analytical framework for the analysis of preference evolution which subsumes as special cases the three main classes of evolutionary environments that have been considered in the literature to date: random matching in pairs or larger subgroups, local interaction with geographic or social neighbors, and interactions involving the entire population. Using this framework, we examine a number of questions and conjectures pertaining to the properties of the strategic environment that correspond to the evolution of particular forms of preference interdependence. For instance, we generalize our previous results on preference interdependence, which suggest that in interactions involving the entire population, negatively interdependent or envious preferences emerge in a variety of strategic settings. In contrast, positive interdependence can evolve and persist in certain random matching and local interaction environments, and we investigate the precise set of conditions under which this can occur.
