The Job of Economic Forecasting

Last Monday, President Bush issued his annual Economic Report. It is prepared by the Council of Economic Advisers, a group of three distinguished economists and an equally distinguished staff. Presently, they include N. Gregory Mankiw of Harvard, Harvey Rosen of Princeton and Kristin Forbes of M.I.T.

This year’s report seems to be getting more than the usual amount of notice. Most of the press attention has focused on the economic forecast, especially for jobs. This forecast was actually prepared last year in conjunction with the Treasury Department and the Office of Management and Budget. Its main purpose is to give government agencies something to use in planning their budgets.

Since the forecast was completed on Dec. 2, clearly it does not represent the administration’s current thinking about how well the economy will do this year. In the intervening two and a half months, a great deal of new economic data has become available that probably would significantly alter the forecast were it to be redone today. Nevertheless, the press continues to treat the forecast numbers as if they are writ in stone.

The fact is that economic forecasting is not very good in general, and all administration forecasts tend to be fairly far off. For example, the last forecast made by Bill Clinton’s CEA in January 2001 confidently predicted that there would be no recession. “Let me be clear,” said CEA Chairman Martin Baily, “we don’t think that we’re going into a recession.”

Of course, we now know that at the very moment Baily spoke we were on the brink of a recession and may already have been in one. According to the National Bureau of Economic Research, a private research group that determines the official dates for business cycles, the most recent recession began in March 2001. However, recent data revisions suggest that it may have started in the fall of 2000, and it is possible that the NBER may move back the official start of the recession.

Another point to remember is that when an administration makes a forecast error, it is usually shared by the economics profession as a whole. For example, in May 2001 — well after the recession had already begun — less than a third of the “Blue Chip” economic forecasters were expecting any recession at all. If you can’t forecast a recession when you are already in one, you are not likely to know when one is coming down the pike.

Even the great Alan Greenspan, chairman of the Federal Reserve Board, has been known to make huge forecasting mistakes. According to the minutes of the Federal Open Market Committee meeting on Oct. 2, 1990, Greenspan said, “At the moment it isn’t raining. … The economy has not yet slipped into a recession.” Yet according to the NBER, a recession had begun three months earlier in July. And Greenspan didn’t even have the excuse that he didn’t know what Federal Reserve policy was going to be, a problem that often trips up private forecasters.

Finally, it is worth noting that administration forecasts are sometimes criticized at the time they are made because some aspect appears to be outside the range of conventional wisdom. But, later, it turns out to be correct. In 1981, the Reagan administration forecast a much larger reduction in inflation than most private forecasters thought was possible, and it was widely ridiculed as a “rosy scenario.” In reality, the forecast was not optimistic enough. Inflation fell even more rapidly than the administration predicted. To my knowledge, none of those who criticized the administration’s forecast ever apologized for being so wrong.

This history is worth remembering because right now a number of liberal economists, such as Brad DeLong of the University of California at Berkeley, are saying that the administration’s jobs forecast is implausibly high. It is forecasting a gain of 2.6 million payroll jobs this year, after two years of negative growth.

I don’t know if the administration’s figure is too high or too low. But I do know that the unemployment rate has already fallen faster than most private forecasters thought it would. Last July, the Wall Street Journal surveyed the nation’s 54 best economic forecasters. Virtually all of them thought the unemployment rate in November would be well above 6 percent. In fact, it was 5.9 percent. As recently as Jan. 2 of this year, the same group thought the unemployment rate would not reach 5.7 percent until May. It is already at 5.6 percent.

I think the odds favor an optimistic outlook for jobs. I know it has been a long time in coming, but with every single other economic indicator pointing upward, jobs must also rise at some point. We may even do better than the administration is forecasting by year’s end.