Clarification, Critique, Framework Construction, and Research Concerns

Abstract

The study of how human beings make decisions has evolved in the last century, as the classical expected-utility (EU) model (developed by D. Bernoulli, 1967 [1738]) was refined by von Neumann and Morgenstern (1947) and then modified by Savage (1954). Savage’s model of subjective expected-utility (SEU) has since served as the primary theory guiding research on decision making in the social sciences. While the normative strength of EU and SEU is often accepted, the descriptive accuracy of the Bernoullian family of theories has been seriously questioned. The failure of SEU to explain the behavior of significant numbers of experimental subjects and real world decision makers has led to further revisions and the development of generalized EU theories that have the ability to explain some (if not all) of these anomalies. All of these theories attempt to explain, at least in part, decisions under risk—where outcomes are not certain, but the chances of the outcomes occurring follow a known probability distribution. Recently, skepticism regarding the descriptive accuracy of EU and SEU has crossed disciplinary boundaries from psychology and economics to political science, sociology, and business. In particular, researchers in political science studying foreign policy decision making have begun to examine alternatives to the Bernoullian models.