I fondly retain my WIN button. It symbolizes President Gerald Ford's 1974 campaign against inflation (Whip Inflation Now), which I played a bit part in organizing. It also represents the worst economic forecast of the 1970s.
At the inflation summits of September 1974 no economists warned of a serious recession ahead. Yet, within two months it became apparent that the economy was in free fall. Ultimately, the gross national product suffered what is still the largest peak-to-trough decline since the '30s.
There are eerie similarities between 1974 and 1990. Then as now, the economy had recently experienced an oil shock, although the shock of 1973-74 was far greater than that recently imposed by Saddam Hussein. Then as now, the Federal Reserve adamantly fought inflation until the eve of the recession. Then as now, economists claimed that inventories were not excessive.
Only later did it become apparent that the inventory data was of terrible quality and that economists had been misled. Could it be that economists now -- a majority of whom are forecasting a mild recession -- are also being misled?
Inventory data was improved significantly after the 1974 fiasco. But data-collecting agencies have been squeezed by the budget battles of the 1980s.
In a recent survey conducted by the newsletter Blue Chip Economic Indicators, 76 percent of the economists responding felt the government's statistics had deteriorated over the last decade. Since the quality of data was pretty bad 10 years ago that is truly frightening.
While the composition of the economy has changed significantly, our statistical techniques have not. Important data on payroll employment are particularly suspect. A month ago it was reported that employment in October fell by 68,000. This month that October number was revised to a decline of 178,000. That is a large change, big enough to have altered the attitudes of the Federal Reserve and other policy-makers.
The cost of improving the accuracy of aybe yes, maybe no
such data is trivial compared with the economic losses that can ensue if we do not make the effort. To its credit the Bush administration has advocated greater spending for statistical agencies, but much more has to be done.
I do not mean to imply that economists armed with better data will suddenly become brilliant forecasters. No amount of good data will allow us to forecast a Persian Gulf crisis, which will have a crucial impact on the economy in 1991.
More important, the nuances of human behavior are never easily captured by the statistical equations used by economic forecasters, regardless of the quality of the data that is available. Therefore, good data will not ensure good forecasts. But bad data can cause some bad ones.
Even with good data there is an inherent tendency, as we enter a recession, for forecasts to understate the impending troubles. Paul Samuelson has quipped that economists can at best forecast minus one month. That may be optimistic, because of the problems we have figuring out what happened last month.
Recessions seldom look as serious at the beginning as they do at the end. It is the cumulative changes in human behavior as the recession unfolds that kill you, and they develop differently in each downturn. This recession has the shadow of war hanging over it, and we can't know how that will affect the decisions of consumers and investors.
In November and December of 1974, as it became obvious that the economy was collapsing, economists scrambled to ratchet down their forecasts.
This year, with weekly announcements of new and more ominous data, a similar but less intense scramble seems under way. That implies to me that this recession will not be as severe as in 1974.
But, then, who am I to know? I rely on the same data and techniques as everyone else.
RUDOLPH G. PENNER, director of the Congressional Budget Office from 1983-87, is a senior fellow at the Urban Institute.