Behavioral Economics: Past, Present, and Future

In recent years there has been growing interest in the mixture of psychology and economics that has come to be known as “behavioral economics.” As is true with many seemingly overnight success stories, this one has been brewing for quite a while. My first paper on the subject was published in 1980, hot on the heels of Kahneman and Tversky’s (1979) blockbuster on prospect theory, and there were earlier forerunners, most notably Simon (1955, 1957) and Katona (1951, 1953). The rise of behavioral economics is sometimes characterized as a kind of paradigm-shifting revolution within economics, but I think that is a misreading of the history of economic thought. It would be more accurate to say that the methodology of behavioral economics returns economic thinking to the way it began, with Adam Smith, and continued through the time of Irving Fisher and John Maynard Keynes in the 1930s. In spite of this early tradition within the field, the behavioral approach to economics met with considerable resistance within the profession until relatively recently. In this essay I begin by documenting some of the historical precedents for utilizing a psychologically realistic depiction of the representative agent. I then turn to a discussion of the many arguments that have been put forward in favor of retaining the idealized model of Homo economicus even in the face of apparently contradictory evidence. I argue that such arguments have been refuted, both theoretically and empirically, including in the realm where we might expect rationality to abound: the financial markets. As such, it is time to move on to a more constructive approach. On the theory side, the basic problem is that we are relying on one theory to accomplish two rather different goals, namely to characterize optimal behavior and to predict actual behavior. We should not abandon the first type of theories as they are essential building blocks for any kind of economic analysis, but we must augment them with additional descriptive theories that are derived from data rather than axioms. As for empirical work, the behavioral approach offers the opportunity to develop better models of economic behavior by incorporating insights from other social science disciplines. To illustrate this more constructive approach, I focus on one strong