Beware Expert Predictions

SIGNS OF THE TIMES

ou would think that experts are more likely to correctly predict the future (stock market and economic swings, to name a few). Well, think again. In fact, the more specialized the forecasters, the less likely their predictions are to hit the mark.

This counterintuitive assessment is the result of 20 years of research by a University of California, Berkeley, psychologist who selected 284 people who made their living commenting or offering advice on political and economic trends. He asked and received predictions for a total of 82,361 questions, assessing the probability of future events in their area of expertise—and found their forecasts were no more likely to be accurate than if he tallied the dart-throwing scores of monkeys.

The psychologist, Philip Tetlock, assembled all these data for his book, Expert Political Judgment: How Good Is It? (Princeton University Press, 2005).

The results may be a surprise to us ordinary folk—those not in the psychological community. But more than 100 earlier studies agreed that experts’ predictions scored no better than statistical chance.

Specialists sought after for forecasts by media do even worse than their colleagues not in the limelight. Tetlock’s explanation: The more famous the forecasters, the more overblown their forecasts. And these experts suffer the common shortcoming of all humans: We tend to fall in love with our hunches. An even worse drawback suffered by experts is that, because they know so much, they are able to enlist more support of their pet theories.

—Stanley Zarowin