Ambiguous Random Variables and Menu Effects

Princeton University, Princeton NJ

Investigators

Abstract

This award funds research in the economic theory of how people make decisions. The first part of the project will focus on how people make decisions in dynamic situations characterized by ambiguity. These situations arise when people have so little information that they cannot formulate ideas of how likely different outcomes are. The second project seeks to develop a model that will predict what psychologists and behavior economists call menu effects. These occur when the choices available to an individual (the so-called 'menu') affects his/her decision in ways that are not consistent with standard economic theory. This project contributes to national interest because understanding individual decision making in these economic contexts will help other researchers develop new insights for management practice and public policy that will benefit the U.S. public. The research focuses on ambiguity and on menu effects. The first part offers a new theory of conditional probability and conditional expectation (called evaluation) suitable for the analysis of dynamic choice problems under ambiguity. It develops a new foundation for belief revision under ambiguity and will characterize this updating rule by imposing restrictions on the evaluation operator. Existing dynamic ambiguity models tend to restrict the set of conditioning events, and this limits applicability to problems that involve costly information acquisition. The new model permits information to arrive in every possible order and identifies a new parameter allows decision makers with identical behavior in static situations to behave differently when facing dynamic problems. The second project provides a framework for analyzing menu effects, in particular attraction, compromise, and endowment effects. The goal is a model demonstrating that a variety of observed deviations from standard consumer theory can be derived from a common underlying source: the agent's inability to tailor his or her choice behavior to the specific problem. The analysis reveals common features (causes) of each of the three menu effects.