As the new year begins, one of the regular features I have come to expect is a new batch of economic forecasts. In the media, those put out by private forecasters will be contrasted with that of the Bush administration, which will be included in the budget message. No doubt, the administration will be accused of being overly optimistic in order to make the budget numbers look better.
The record of the private forecasters, who are assumed to be unbiased and more accurate when such comparisons are made, however, is never reviewed. The truth is that while all administration forecasts tend to be more optimistic than most private ones, there is no evidence that the latter are more accurate. In fact, all forecasters are just about equally wrong. For every time one manages to hit a key number right on the mark, there are many more occasions when they were off by a country mile.
Most of the time, a forecaster will be pretty close to right if he says that the near future will be about the same as the recent past. But this is not very useful information. What companies and policymakers need is forewarning that things are about to change one way or another. On this score, the record of forecasters is especially dismal.
Let us look back a year. On Jan. 2, 2001, The Wall Street Journal published the predictions of 54 well-known economic forecasters. The great majority of these work for private banks. Presumably, therefore, there is a high premium on being right. Those who are successful are very well paid, and those who aren't often lose their jobs.
Yet looking over these forecasts, only five were even close to accurate in predicting a recession -- informally defined as at least two back-to-back quarters of negative real gross domestic product growth. These were the only ones to have even a single quarter of negative growth in their forecasts, all the rest predicting positive growth for the whole year. Only two forecasters, A. Gary Shilling and Tracy Herrick, had the quarters right, showing negative growth in the third and fourth quarters. However, Shilling was predicting negative growth continuously since the third quarter of 2000 and had growth declining much more sharply than was the case.
Thus anyone who based their investments on any of these forecasts probably didn't do too well. Indeed, these forecasters weren't any more accurate six months later, when the Journal asked the same group for an updated forecast in July. Even though the recession had already begun in March, according to the National Bureau of Economic Research, only five predicted a negative third quarter, and only two saw negative growth in the third and fourth quarters. Herrick of Jeffries and Co. appears to have been closest to the mark for the year.
For what it is worth, the record of foreign economists is no better than those of American economists. A recent survey by the International Monetary Fund found that "the record of failure to predict recessions is virtually unblemished." As in the United States, recessions were often not even predicted after they had already started.
One excuse that forecasters always make is that they cannot predict changes in key policies, such as Federal Reserve policy, that can have an enormous impact on economic growth. But even knowing what the Fed is going to do does not guarantee accuracy. According to the published minutes of the Fed's Oct.r 2, 1990 meeting, Fed Chairman Alan Greenspan still saw no recession on the horizon. Yet we now know that the economy was already in recession. According to the NBER, it had started in July.
One explanation for the persistence of inaccurate forecasts is that most forecasters are not really paid to be accurate at all. According to a study by the Federal Reserve Bank of New York, forecasters are compensated more for getting publicity for their companies than for being accurate. Indeed, in court testimony, the head of Bear Stearns, a Wall Street investment bank, once described its chief economist as primarily an "entertainer."
Another explanation is that certain economists stake their reputations on a particular forecasting model and they just never change it. But, like a stopped clock, they are right twice a day. Economist Ravi Batra, for example, has been predicting another Great Depression for at least 10 years. If he keeps at it, perhaps some day he will be right.
Despite the failure of forecasting, it still has to be done. Plans have to be made, and they need to be based on something more rigorous than a finger in the wind. However, those who use economic forecasts need to be aware of their limitations.