THE TEACHING ECONOMIST - William A. McEachern                 

HomeAbout The Teaching Economist Contact the Editor Support

Issue 37, Fall 2009

William A. McEachern, Editor


The headline describing a Wall Street Journal survey of about fifty economists reads: "Economists Say Worst of Housing Bust Is Over." The story notes that by a 2-to-1 majority, economists believe the housing market should start looking up. Welcome news, if not surprising, given all that the housing sector has been through. But what may surprise you is that the headline appeared in November 2006, at a time when real housing troubles were just beginning.

It's easy to cherry-pick a headline to make forecasters look bad, but how did they do last fall, when the entire economy started to look shaky? By early September 2008, the Lehman Brothers bankruptcy was still about ten days away, but a number of troubled financial institutions had already failed or been merged, Bear Stearns had a forced liquidation, and the unemployment rate had crept to 6.2%, up from 4.9% in January of that year. Given this setting, what did the WSJ forecasters see as they looked ahead? In the survey taken in early September 2008, the WSJ panel forecast an average unemployment rate of 6.3% for June 2009 and 6.2% for December 2009 (the lowest forecast for either month was 4.6% and the highest was 7.3%). Thus, on average they were forecasting the rate to remain about where it was at the time. The actual rate in June 2009 turned out to be 9.5%. December 2009 has not yet arrived as I write this, but the rate for September stood at 9.8%. With regard to jobs, the WSJ panel forecast average losses of 18,500 jobs a month between September 2008 and September 2009 (forecasts ranged from a loss of 125,000 jobs a month to a gain of 137,000 jobs a month). The economy actually lost an average of 486,000 jobs a month, or 25 times the survey mean.

The forecasters were pretty far off in looking ahead, but how did they do in looking around? They were also asked in early September 2008 to "forecast" real GDP for that third quarter of 2008, the quarter that was then in its third month. Could they look around and see what was going on? Not really. They "forecast" an average real GDP growth of 1.2% for the quarter (the highest forecast was 3.5% growth and the lowest, a decline of 0.7%, one of only three forecast declines). Real GDP that quarter actually fell by 2.7%. The NBER business-cycle dating committee three months later announced that the recession began in December 2007, so by early September 2008, the recession was already nearly three quarters old, reaching the duration of the previous two recessions. Thus, in the final month of the third quarter of 2008, the forecasters had little inkling that the economy was experiencing what at the time would turn out to be the largest quarterly GDP drop since 1990.

Many academic economists are skeptical of forecasters and forecasting, but forecasting is the public face of economics. It's what the public thinks economists do. The forecasters in the WSJ survey are smart people, most with Ph.D.s in economics (the University of Pennsylvania is a notable source); some are employed by academic institutions. They are also experienced in what they do—about two thirds in the September 2008 poll had been part of the survey for at least five years. They have access to the best models, most sophisticated computers, and most reliable data sources money can buy. These resources are often buttressed with proprietary information from their own firm and industry. And they have an abiding interest in forecast accuracy. Yet with all this going for them, they had a hard time seeing something all around them. They were not alone. A few economists warned of big trouble, but most believed the economy would ride out the popped housing bubble with at most a run-of-the-mill recession. On the other hand, according to polls conducted in early 2008, half of Americans said we were already in a recession, and, by a large majority, they expected the economy to get worse. The wisdom of crowds?

We can conclude that macroeconomic forecasting is hard stuff. As the Nobel prize-winning physicist Murray Gell-Mann noted, "Think how hard physics would be if particles could think." In acknowledging this, we should not forget that there are hundreds of other questions to which economists have more reliable answers.

Top                                                                                                                                                                                          Next