Why economists can’t forecast
|
Version 0 of 1. You knew it all along: Economists can’t forecast the economy worth a hoot. And now we have a scholarly study that confirms it. Better yet, the corroboration comes from an impeccable source: the Federal Reserve. The study compared predictions of important economic indicators — unemployment, inflation, interest rates, gross domestic product — with the actual outcomes. There were widespread errors. The study concluded that “considerable uncertainty surrounds all macroeconomic projections.” Just how large were the mistakes? The report, though written mostly in technical jargon, gives a straightforward example: “Suppose . . . the unemployment rate was projected to remain near 5 percent over the next few years, accompanied by 2 percent inflation. Given the size of past errors, we should not be surprised to see the unemployment rate climb to 7 percent or fall to 3 percent. . . . Similarly, it would not be at all surprising to see inflation as high as 3 percent or as low as 1 percent.” These are huge margins of error. Clearly, much economic forecasting is guesswork. Worse, the gap between prediction and reality may be widening. The study — done by David Reifschneider of the Federal Reserve and Peter Tulip of the Reserve Bank of Australia — found that forecasting mistakes had worsened since the 2008-09 financial crisis. An interesting question (which the study did not ask) is whether economic forecasting has improved in the past century. In the 1920s, with no computers, forecasters relied on random statistics: freight car loadings; grain harvests and prices; bank deposits. Today, forecasters employ elaborate computer models that scan dozens of statistical series describing the economy. Yet the predictions seem no better. The implications are profound. If forecasts are inevitably flawed, there are bound to be recessions. Government officials — not only at the Federal Reserve but also in Congress and the White House — are condemned to make mistakes. Their vision of the future is blurred; therefore, their policies may blunder. The same is true of the private sector: Consumers and companies, misreading the future, may act to bring the economy down. They may borrow too much or spend too little. The study compared forecasts from 1996 to 2015 not only from the Fed but also from the Congressional Budget Office, the Blue Chip Economic Indicators and the Survey of Professional Forecasters — these last two representing mainly private economists. Crowd behavior dominated; forecasts bunched together. “Differences in accuracy across forecasters are small,” write Reifschneider and Tulip. Naturally, the further forecasts probed the future, the worse their reliability. The bedrock lesson here is as old as time: The future — not all of it, but much of it — is too complex to be predicted. There are too many moving parts; too much is unknown; people wrongly think the future will resemble the past; they don’t foresee political, economic and technological change. Recent upsets to the economy confirm this ignorance. The biggest was the financial crisis and the Great Recession. But there were others: the fall of long-term interest rates on bonds and mortgages; the unexpectedly slow growth of the economic recovery accompanied by an unexpectedly rapid (and inconsistent) decline in unemployment; and the collapse of productivity. Almost none of this was anticipated. It has been said that the past is a foreign country. So is the future. Read more from Robert Samuelson’s archive. |