Analyzing the economy is a daunting task

Posted: Apr 30, 2002 12:00 AM
The White House was gratified by the announcement last Friday that the nation's real gross domestic product grew 5.8 percent in the first quarter. Unfortunately, it will be a long time before we know for sure how well the economy is really doing. This GDP report was only the first of many official estimates of growth to come. On May 24, we will get another estimate of first quarter GDP based on more complete data from January and February, and a bit from March. It will not be until June 27 that the Commerce Department has sufficient data to produce a final estimate for growth in the first quarter. Since revisions can be significant, this figure could be considerably higher or lower than the one we now have in hand. But the revisions won't stop in June, after the so-called final estimate is released. Commerce Department economists will continue revising the GDP data literally forever. Every few years, they produce new figures, going all the way back to 1929. This results from continuing research that uncovers new information, as well as improvements in measurement and economic theory. These ongoing revisions can often change our knowledge of the past quite dramatically. It can also affect the reputations of economic forecasters. A good example is that of my friend Arthur Laffer, who was chief economist for the Office of Management and Budget under Richard Nixon. In 1970, Laffer prepared a forecast for gross national product in 1971 of $1,065 billion. This was well above the $1,050 billion estimate of most forecasters. Since the economy was much smaller then, differences of $15 billion in GNP estimates were important, with implications for federal revenues and various government programs. The Laffer forecast was widely ridiculed. Economist Herb Stein of the Council of Economic Advisers was especially critical. There were even efforts made to revoke Laffer's tenure at the University of Chicago, from which he was on leave, because of his absurdly optimistic outlook for the economy, which raised questions about his professional competence. Although Laffer went on to have a successful economic consulting career, a few old-timers still remember the "1065" incident as refutation of his work in other areas, such as tax policy. What none of these critics have done is look at the subsequent revisions to GNP, which now show Laffer actually to have been wildly pessimistic. Contemporary estimates for 1971 put GNP at $1,047 billion -- close to the consensus forecast. But many subsequent revisions have steadily raised that figure. It is now recorded as $1,136 billion -- far, far above Laffer's forecast -- in the latest published data. Yet there are still some economists and reporters out there who only remember him being off by a mile on the high side, instead of on the low side. Laffer's experience is exceptional, but not unique. Unfortunately, it does a forecaster little good, professionally, to be proven correct long after the fact. Those who award forecasters for accuracy do so based on recent figures, not fully revised ones. In essence, they are rewarded for predicting what the initial estimates of economic growth will be, not what it actually is. Of course, it is not really possible to know what the true figures are in many cases, because they are continuously revised. On the other hand, those who employ most forecasters don't really care about ultimate truth. They are only interested in making a buck by trading stocks and bonds. It doesn't do them any good to know they were right after the fact, when the markets moved based on the preliminary data. Opportunities for profit come only from getting the early estimates right. The fact of the matter is that economic forecasting is more art than science. One reason is that the information forecasters must work with is often very soft. This is partly because those who provide the raw data about production and investment have no real incentive to be accurate. In this respect, I am reminded of economist Josiah Stamp's observation more than 70 years ago: "The Government are very keen on amassing statistics -- they collect them, add them, raise them to the nth power, take the cube root and prepare wonderful diagrams. But what you must never forget is that every one of those figures comes in the first instance from the village watchman, who just puts down what he damn pleases." Even today, a single company's reporting can affect the quality of our national economic statistics. Intel, the world's largest computer chip maker, recently ceased providing data to the government on its semiconductor orders. It did so because the particular figures the government wanted were irrelevant to its own operations. Intel's action has had a major impact on the government's durable orders estimates, a figure closely watched by financial markets.