Comparing Fed Forecasts with the Private Sector’s

To ascertain the state of the U.S. economy, the Federal Reserve produces forecasts of macroeconomic variables, such as gross domestic product (GDP) growth, inflation and unemployment. How accurate are these forecasts relative to those of private forecasters?

In a recent Regional Economist article, Michael T. Owyang, Amy Y. Guisinger and Hannah G. ShellOwyang is an assistant vice president and economist with the St. Louis Fed, Guisinger is an assistant professor of economics at Lafayette College, and Shell is a former senior research associate with the St. Louis Fed. explored how Fed forecasts fared compared with forecasts made by the private sector.

“Academics seem to have spent nearly as many resources studying the Fed forecasts as the Fed has spent creating them,” the authors wrote. “One robust finding is that the Fed forecasts have been, on average, more accurate than many private sector forecasts.”The authors explained that accuracy is often measured by computing the average of the squared forecast error (the difference between the forecast and the realization) over a period of time, with a lower value indicating a higher level of accuracy.

They pointed out three key reasons economists have cited for why the Fed might have an advantage over private sector forecasters:

• The Fed simply devotes more resources to forecasting or has better forecasters.
• The Fed has more timely access to economic data.
• The Fed has knowledge about the path of future monetary policy.

Better Forecasts?

To examine the accuracy of the Fed forecasts, the authors focused on forecasts of macroeconomic variables provided in the Greenbook (now called the TealbookThe authors noted that the Board of Governors staff used to produce both the Greenbook and the Bluebook in preparation for FOMC meetings. These were merged in June 2010 to form the Tealbook. However, this blog post uses the term Greenbook throughout.), which is distributed to the Federal Open Market Committee (FOMC) before meetings at which monetary policy is set. The forecasts are produced by the Board of Governors staff, who use both econometric models and subjective assessments. The forecasts are made publicly available five years after the FOMC meeting for which they were produced.

The authors pointed to a 2000 study by economists Christina Romer and David Romer.Romer, Christina D.; and Romer, David H. “Federal Reserve Information and the Behavior of Interest Rates.” American Economic Review, June 2000, Vol. 90, No. 3, pp. 429-57. The study compared forecasts from 1968 to 1991 for the gross national product (GNP) deflator—which is one measure of inflation—using:

• The Greenbook forecasts
• The consensus forecasts of the Survey of Professional Forecasters (SPF), a collection of individual private sector forecasts

“The Romers found that the Fed, generally, was more accurate forecasting inflation than the private sector,” Owyang, Guisinger and Shell wrote.

The three authors then examined the Greenbook forecasts for the GNP deflator and the GDP deflator from the fourth quarter of 1968 through the fourth quarter of 2012. (In 1991, the Greenbook switched from forecasting the GNP deflator to forecasting the GDP deflator.)

They found that “the Fed's forecasts are more accurate than the SPF's, with the mean squared error of the Greenbook forecast being 3.40 compared with 4.77 for the SPF forecast.”

(For more on the authors’ methodology, see the Regional Economist article “Economic Forecasting: Comparing the Fed with the Private Sector.”)

Why Does the Fed Have an Edge?

The Romers posited that the Fed may have a forecasting advantage simply because it devotes more time and resources than the private sector does. However, Owyang, Guisinger and Shell noted that the Romers’ reasoning doesn’t explain more recent findings by other economists that the forecasting advantage seems to have declined over time.

The Regional Economist article authors then looked at whether the advantage is due to the Fed having more timely access to data, with one example being the current value of industrial production that the Fed constructs.

“This informational advantage, however, lasts only a few days and is often dissipated by the time the private sector forecasts are collected,” they wrote.

The authors then examined a third possible reason for the Fed’s advantage: that Fed forecasts are made conditional on a path for future policy. If the Fed knows future monetary policy, but the private sector doesn’t, this could give the Fed an informational advantage, they noted.

“This explanation could also account for the erosion of the Fed's advantage: The Fed has taken steps to increase transparency in recent years, thus giving the private sector more information about the likely path for policy,” they wrote.

Another Possible Reason for the Declining Advantage

The authors pointed to a 2009 study by economists Edward N. Gamber and Julie K. Smith, who argued that the nature of forecasting has changed over time.Gamber, Edward N.; and Smith, Julie K. “Are the Fed’s Inflation Forecasts Still Superior to the Private Sector’s?” Journal of Macroeconomics, June 2009, Vol. 31, No. 2, pp. 240-51.

The volatility of many U.S. macroeconomic variables has declined since 1984, while economists have found it has become harder to beat even the simplest forecasting models, Owyang, Guisinger and Shell noted.

“In this sense, if the Fed's forecasting advantage stemmed from its devotion of more resources to forecasting than the private sector, but sophisticated models are no longer advantageous, one would suspect the Fed's advantage to decline,” the authors concluded.

Notes and References

¹ Owyang is an assistant vice president and economist with the St. Louis Fed, Guisinger is an assistant professor of economics at Lafayette College, and Shell is a former senior research associate with the St. Louis Fed.

² The authors explained that accuracy is often measured by computing the average of the squared forecast error (the difference between the forecast and the realization) over a period of time, with a lower value indicating a higher level of accuracy.

³ The authors noted that the Board of Governors staff used to produce both the Greenbook and the Bluebook in preparation for FOMC meetings. These were merged in June 2010 to form the Tealbook. However, this blog post uses the term Greenbook throughout.

⁴ Romer, Christina D.; and Romer, David H. “Federal Reserve Information and the Behavior of Interest Rates.” American Economic Review, June 2000, Vol. 90, No. 3, pp. 429-57.

⁵ Gamber, Edward N.; and Smith, Julie K. “Are the Fed’s Inflation Forecasts Still Superior to the Private Sector’s?” Journal of Macroeconomics, June 2009, Vol. 31, No. 2, pp. 240-51.

Additional Resources

• Regional Economist: Economic Forecasting: Comparing the Fed with the Private Sector
• On the Economy: Who Wants to Bet against the Yield Curve: Is This Time Different?
• On the Economy: The Effects of Uncertainty on Economic Outcomes