Uncertainty pervades the economy. Analysts have developed a number of tools for understanding choice under uncertainty. These tools are used by industry and government. The traditional expected utility paradigm is used in economics, psychology, decision theory and operations research to model choice under uncertainty. But expected utility theory does not explain the results of many empirical tests. This project is part of an interdisciplinary effort to develop an alternative to expected utility theory that better explains observed behavior. This research project addresses a number of interrelated questions in the theory of choice under uncertainty when preferences do not correspond to the expected utility model. The axiomatic work in the project focuses on a number of preferential attributes, namely, risk aversion, randomization preference, induced preference over assets, preference for the timing of resolution of uncertainty, and intertemporal substitution. These attributes concern both atemporal and temporal aspects of decision making. The project studies each of them individually and also examines their interrelationships. The primary application is to explore the implications of non- expected utility specifications for equilibrium asset prices. This work exploits the flexibility of non-expected utility specifications with regard to the separation of intertemporal substitution and risk aversion. A general equilibrium, Lucas- style model of asset prices is developed and used to clarify the link between preference theory and asset pricing theory. Other applications relate the axiomatic work to behavior in economic contexts. New perspectives, both atemporal and temporal, are applied to phenomena such as insurance demand and gambling. This is an exciting project. It could produce a new paradigm for studying choice under uncertainty. This paradigm would be more general than the traditional expected utility paradigm. It would be more consistent with observed behavior. The decision making tools derived from this paradigm could be used by industry and government to make better decisions.
