December 30, 1988
"As Reliable as Weather Forecasts - Economists have Dubious Track Record"
San Jose Mercury News
By Timothy Taylor
<< Back to 1988 menu

EACH new year begins with a flurry of predictions about the course of the economy. Sometimes the predictions are even taken seriously, although I've never really understood why. As the saying goes, economists have predicted nine of the last five recessions.

Should you be tempted to enter into any such arguments, let's review some recent predictions of the National Association of Business Economists. Each quarter, the association samples its membership and releases a consensus forecast.

August 1983: "It's doubtful the steady, calm recovery expected through 1984 can be sustained beyond next year without major budget surgery," said the association spokesman. He was wrong twice: The 1984 economic surge was fast and frenzied, not steady and calm, and the economy kept growing though 1985 without major budget surgery.

May 1984: "The majority of business economists believe that high interest rates will kill the economic expansion before the end of next year," reported the Wall Street Journal. But the economy continued to grow.

September 1984: Reversing their view of four months earlier, the association announces that it now believes the U.S. economy will continue to grow at a 3 to 4 percent annual pace through most of next year.

February 1985: "The vast majority of business economists think that the economy will fall into a recession before the end of 1986," reported the Journal. But postponing the recession didn't make it happen; the economy continued growing in 1986.

December 1985: Only 37 percent of the economists polled by the association continue to expect a recession in 1986.

May 1986: The economists predict 3.2 percent growth for 1986, but 47 percent expect a recession in 1987.

October 1986: The median forecast of the survey predicts 3 percent growth for 1987.

October 1987: The consensus forecast is that the economy will continue to grow in 1988, but will fall into recession in 1989. Two weeks after this forecast, the stock market crashes.

December 1987: Because of the stock market crash, half of those surveyed are now predicting a recession in 1988.

March 1988: Only one-fourth of those surveyed expect a recession by the end of 1988.

September 1988: 47 percent now predict a recession in 1989, while another 42 percent predict a recession in 1990.

Frankly, this record of predictions is embarrassing.

Of course, a few forecasters can probably claim that they were right all the way through. But with enough forecasters making predictions, blind chance will imply that a few are almost always correct. In the same way, if you flipped a million coins 20 times each, one or two of them will come up all heads or all tails, all 20 times. Until you can explain to someone else how to do it, it's luck, not science.

As if all this wasn't enough, Steven Strongin and Paula S. Binkley of the Federal Reserve Bank of Chicago have recently calculated that when economic forecasters update their predictions, with six months of additional information, they are just as likely to worsen the forecast as to improve it.

But enough pessimism. Strangely enough, I don't think these results should be seen as very damaging to the reputation of economists.

After all, an inability to foresee the future is surely not limited to economics. Geologists can't predict earthquakes. Political scientists can't predict who will win presidential elections. As Stanford economics Professor and Nobel laureate Kenneth Arrow once pointed out, biologists aren't too good at predicting what species will evolve next. One can know a lot that is interesting and true in any field without having a crystal ball.

The sad truth is that people want perfect fortune tellers, and they seem to have elected economists to fill the bill. Lots of economists are, like anyone else, perfectly willing to give their opinions.

But the truth is that few economists spend most of their time trying to forecast the future of the whole economy. In fact, economists and most other scientists are if/then thinkers. If such-and-such happens, an economist will say, and nothing else goes wrong, then I can tell you the likely results. For the purposes of analyzing a bill before Congress, or the risk of a drought, or whether a business should market a new product, these kinds of analysis are far, far better than wild guesses.

But forecasting the course of the entire economy is much different than this sort of incremental forecasting.

After all, it's hard to blame economists for not foretelling the rise of OPEC and the fall of the shah of Iran. Or the rise of semiconductor technology. Or that the federal government would break out of all historical patterns and approve unprecedentedly large budget deficits. Or that a drought would rock the United States in the summer of 1988. When the "ifs" are so uncertain, the "thens" become truly shaky. Economists can predict general tendencies, but history tends to unroll in unpredictable ways.

Perhaps the best way to make use of economic forecasts is think of them as the dramatization of some very real insights. For example, economists are obviously not too good at predicting the month and year when the economy will have to start paying the price for its high budget deficits.

But threatening a recession seems to be the only way to get people to face up to the fact that the country will have to pay, sooner or later, in one way or another.

<< Back to 1988 menu