| This dissertation represents an attempt to reconcile the growing rift between the normative neoclassical approach to modeling human behavior and the descriptive approach of behavioral economics. I focus here on a number of special cases of the consumer's decision problem, all having to do with dietary choice: I ask how people decide what and how much to eat.; This narrow focus facilitates the use of a wealth of scientific knowledge on the subject of dietary choice. Much is known, for example, about the genetics, molecular biology, and endocrinology of obesity; and there exists a rich literature describing the determinants of dietary choice among children and other young animals in laboratory settings.; I find that this wealth of information is consistent with an evolutionary interpretation of choice behavior. In particular, I argue that the behavior in question is best seen as being guided by an underlying preference relation over actions, and that natural selection has acted upon this preference relation over the course of human evolutionary history. A reasonable consideration of the dynamics of the evolutionary process, however, suggests that on time scales of interest to economics (i.e., years or decades) the human preference relation is for all practical purposes immutable. This in turn implies that when technology alters certain aspects of the decision environment, the subjective probabilities implicit in the preference relation will be biased, resulting in an (apparent) dynamic inconsistency of behavior.; The result of this analysis is a parsimonious theory that generates detailed predictions, for example, with respect to the incidence of obesity and to many features of the modern fast food industry. Some implications for the theory and practice of welfare economics are also discussed. |
