The Perils of Predicting

Viewpoint from the archives

Editor's note: The Mackinac Center has been publishing Viewpoints for 25 years. Writing on a variety of public policy concerns, our scholars articulate the best remedies for what’s ailing government. Since we believe principle-based ideas are timeless, we dusted off one of our first Viewpoints in honor of our 25th Anniversary.

Originally published Jan. 1, 1988

With a new year just around the corner, the air is filled with the prophesies of economic fortune tellers, but the raw materials they use to prophesy are in need of repair.

Stay Engaged

Receive our weekly emails!

Given the record, it sometimes seems we might be no worse off if we let the economists predict the weather and the meteorologists predict the economy. Economists have made notoriously faulty forecasts throughout the history of the discipline. Indeed, if they really knew as much about the future as some of those finely-detailed predictions suggest, then they should be routinely reaping vast fortunes in the stock and commodity markets. The fact that economists are not a wealthy class and that the few forecasters among them who are well-off usually make more money making their predictions than actually investing in them, is a telling point.

Maybe that’s what the financier Bernard Baruch had in mind when he said, “An economist is a guy who thinks he knows more about money than those of us who have it.”

I make these critical remarks as an economist, and one who has made more than a few forecasts which dramatically missed the mark.

The future will always be more uncertain than most economists will admit, but other reasons also explain why forecasting is in disrepute these days. The theories which form the bases for interpreting economic statistics are often faulty, but equally important is the fact that those economic statistics themselves are frequently defective. Here’s a sample.

Much attention is paid each month to the government’s Index of Leading Economic Indicators, but the LEI is out-of-date and inaccurate. It overweights the influence of the stock market and undervalues the role of global economic developments. It does not account for the growth of the service sector in the U.S. or the declining role of manufacturing. The figures which go into the computation of the index do not take inflation into account. Inventory increases are factored positively into the LEI, though they sometimes are negative for the economy. All things considered, the remarks of Paul Samuelson of M.I.T. draw the appropriate conclusion: “You could have the index saying the economy is going one way when the real economy is going another way.”

Export data from the Commerce Department is of routinely poor quality. Even the International Monetary Fund has argued that the U.S. has consistently undercounted its exports by about
13 percent, which would mean as much as $30 billion worth this year alone.

The unemployment rate is meant to be a key indicator of national economic health but it counts striking workers among the jobless. It counts members of the military stationed in the U.S. as employed. And most economists believe it probably does not accurately reflect true joblessness in the inner city, distressed rural areas, or among migrant workers and illegal immigrants. At the same time, the official rate lists a large number of people as out-of-work who are just voluntarily moving from one job to another.

Retail sales reports from the government are regularly based on a woefully inadequate sampling. The Federal Reserve’s figures on factory utilization fail to consider the capacity of firms to increase output by simply using more efficient machinery. The “errors and omissions” category in such widely-watched numbers as the balance of trade is often large enough to seriously jeopardize the total figure. And discrepancies between the government’s figures and private industry’s figures — in one area after another — are often significant.

Finally, revisions in government statistics are common, sometimes massive and usually quite belated. GNP figures for the first quarter of 1988, for instance, showed the nation’s output rising at a 2.3 percent pace. Only a month later, it was revised to 3.9 percent and that’s not likely to be the last revision. Under pressure to come up with numbers quickly, federal agencies habitually jump the gun, issuing numbers that are so premature and haphazard that their value is dubious at best.

Forecasting the economic future will never be a real science, let alone an exact one, especially if the historical numbers we use are defective. The thing to keep in mind as we hear all the year-end forecasts is this: the only thing certain about the future is that it will remain uncertain until it gets here.