Academic journal article Federal Reserve Bank of St. Louis Review

Subjective Probabilities: Psychological Theories and Economic Applications

Academic journal article Federal Reserve Bank of St. Louis Review

Subjective Probabilities: Psychological Theories and Economic Applications

Article excerpt

Conventional economic analysis of individual behavior begins with the assumption that consumers maximize expected utility, optimizing their planning for the future. Economists incorporate this assumption in models by endowing consumers in those models with the skills of a good statistician--that is, the ability to make rational (and often complicated) calculations. While not always realistic (perhaps never), this assumption facilitates the use of economic models that may work well in the real world. However, in some cases, these models cannot explain some of the evidence uncovered in psychological experiments. In other words, the traditional statistics-based approach sometimes fails to predict individual behavior and aggregate market outcomes that are consistent with the empirical evidence. For instance, observed stock prices and portfolio choices fail to conform to the implications of well-known frameworks, such as the capital asset pricing model (CAPM). Such cases have encouraged a branch of economics that borrows ideas from psychology to explain these discrepancies. (1)

In this area of study, researchers replace the assumption that individuals use complicated statistical formulas to maximize expected utility with the likelihood that they use simple rules of thumb instead, rules that have been identified by psychological research. Psychologists have found evidence that individuals estimate the probability of future outcomes in a nonstatistical, or subjective, manner. Kahneman and Tversky (1973) and Kahneman, Slovic, and Tversky (1982), among others, have introduced the idea of subjective probability heuristics--rules that people tend to rely on when assessing the likelihood of alternative events. Psychological research has shown that the use of these rules can create different outcomes from what statisticians (and economists) might expect, both in the estimated probabilities and in observed behavioral patterns.

Behavioral theories of decisionmaking therefore ask whether economic phenomena may be explained by models in which

* Some, but not necessarily all, agents either fail to update their probabilistic beliefs by applying the appropriate statistical rules or subsequently fail to maximize a standard expected utility objective.

* The remaining fully rational agents, then, cannot completely exploit and eliminate the biases caused by the actions of agents who are not perfectly rational.

While these heuristics are drawn from psychological studies, they may be supported by economic models with boundedly rational agents (Simon, 1955). In other words, agents do not always have the time or the cognitive ability to process all of the data provided by the economic environment with the necessary accuracy. Instead, people might employ these heuristics to arrive at analyses that are less costly to calculate than optimal decisions (Evans and Ramey, 1992); and, often, the optimal decisions themselves are impossible to calculate for difficult problems. Thus, boundedly rational agents do not maximize expected utility as an economist would generally assume. Instead, they maximize perceived expected utility, a quantity based not on actual probabilities but on their beliefs about those probabilities (Rabin, 1998, 2002).

In this article, we focus on the nature and application of psychological rules for probability formation and the biases from anticipated economic outcomes that can result from their use. (2) We examine three heuristics that have been identified by psychologists: the representativeness heuristic (RH), the availability heuristic (AH), and anchoring and adjustment (AA). We review the psychological evidence supporting the common use of these heuristics in estimating subjective probabilities. Finally, we consider a financial application that uses heuristics to estimate probabilities with potentially important economic implications. We then show the effect of these heuristics on people's probability judgments. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.