Judgment and Decision Making: Extrapolations and Applications

People who make or implement public policy must often estimate probabilities, predict outcomes, and make decisions that affect the welfare, values, and lives of many others. Until recently many of the disciplines that study policy employed a model of individuals and organizations as rational agents whose predictions conform to the prescriptions of probability theory and whose actions maximize their expected gains in conformity with classical decision theory. Such theories lead a double life. They are sometimes viewed as normative models that tell us what we should do in order to be rational (even if we rarely manage to pull it off). Construed this way, they offer advice: we should have logically consistent beliefs, coherent probability assignments, consistent preferences, and maximize expected utilities. But these same theories have also been viewed as descriptive models; construed this way, they are meant to provide an approximate characterization of the behavior of real people. It is this interpretation that has played a central role in economics, management science, and parts of political science, sociology, and the law. Since the early 1970s this descriptive picture of judgment and decision making has come under increasing attack from scientists working in behavioral decision theory, the field concerned with the ways in which people actually judge, predict, and decide. Much of the criticism derives from the work of Tversky, Kahneman, and others working in the heuristics and biases tradition. Scientists in this tradition argue that people often