June 8, 2006

BIRMINGHAM, Ala. – Erik Angner, Ph.D., has dared to ask the question: Are economists the victims of overconfidence?

Angner, an assistant professor of philosophy at the University of Alabama at Birmingham (UAB), published a paper, “Economists as Experts: Overconfidence in Theory and Practice,” in a recent edition of the Journal of Economic Methodology in which he says there are reasons to think that overconfidence is endemic among economists who are called upon by public policy makers for their expertise. And economists’ overconfidence, said Angner, can have dramatic effects on both public policy and investors.

“Overconfidence in economists as experts can lead to misguided policy decisions and undermine public trust in economics as a tool in rational decision-making,” Angner said.

To support his thesis that economists are likely to be victims of overconfidence, Angner’s article points to a number of empirical studies on the psychology of judgment and decision-making. Many of those studies have documented evidence of overconfidence among those called upon to make expert judgments, including economists, doctors, physicists and even CIA analysts.

In his article, Angner examines the work of Swedish economist Anders Åslund, who in the early 1990s was an adviser to the Russian government and a major proponent of a policy of rapid deregulation and privatization known as “shock therapy.” Åslund predicted shock therapy would work quickly and have many beneficial effects. His predictions, however, proved well off the mark, Angner said.

In his article, Angner says overconfidence has been noted in other professions. In one study doctors were asked to diagnose possible pneumonia. In the experiment, when doctors claimed to be 80 percent sure that a patient had pneumonia, X-rays confirmed the existence of the disease in only 20 percent of the cases.

In a study of social scientists, subjects were asked to judge the likelihood that several political, economic and military events would occur within a specified period of time. The expert social scientists in the study displayed high confidence that they were right, but those who assigned a confidence rating of about 80 percent or higher were correct only about 45 percent of the time.

Past studies, however, show that professional bridge players and meteorologists forecasting rain demonstrate a high rate of accuracy, which is attributed to having to make highly repetitive judgments and the receipt of immediate, unambiguous feedback.

In contrast, Angner said economists are asked to make judgments on issues that are complex, difficult and challenging and to deal with problems that offer little or no outcome feedback. Moreover, economists lack the institutional restraints to moderate their confidence.

Another reason to believe economist experts are overconfident is that confident economists tend to make themselves available when expert advice is needed, Angner said. They are more likely to appear on television and to be known outside of professional or academic circles. Confident economists are more often selected as experts and therefore, are more likely to be overconfident than the average economist.

“When expert forecasts fail, repeatedly and dramatically, it is understandable if economist expert advice ultimately tends to be ignored,” Angner said. “In short, overconfident economists give economics a bad name. In contrast, by realizing the limitations of expert knowledge, and by making less overconfident judgments, that trust can be preserved or restored.”