Fed Inflation Forecasts

Article excerpt


It's the Federal Reserve's job to set monetary policies that maintain price stability and promote maximum sustainable growth in output. Economic forecasting is central to this task. Among the many economic variables regularly forecast at the Fed, none is more important than inflation. It's also one of the most difficult variables to predict.

Since Fed inflation forecasts directly influence monetary policy, in particular short-term interest rates, it's relevant to inquire how the forecasts are made and how accurate they are.

Prior to each meeting of the Federal Open Market Committee (FOMC), the group that sets the federal funds interest rate, Federal Reserve Board staff economists distribute their forecast to committee members. Individual members may not always agree with the staff forecasts, but they nevertheless carry considerable weight.

The Fed staff forecasts are judgmental - though with the benefit of a variety of economic models, inputs and analyses - and are consistent with an assumed future path for monetary policy. Judgmental predictions are preferred since models based on past data cannot adequately factor in new developments, nor build in qualitative information or intuition.

In remarks to a research forum in Germany on May 20, Federal Reserve Gov. Donald L. Kohn spoke about inflation modeling at the Fed and presented a chart showing board staff forecasts of inflation for 1984 to 2000. (Later data were not given, since staff forecasts are kept confidential for five years.)

For each quarter, the value forecast was the percent change four quarters ahead in the core Consumer Price Index (the index excluding food and energy), a measure of underlying inflation.

A comparison showed two-thirds of inflation forecasts were within a half-percentage point of what actually occurred. In other words, a third of the time the actual price change was more than a half-point higher or lower than predicted.

For 1984-2000 as a whole, the staff forecast was slightly higher than the actual increase, by an average of about 0.2 percentage points yearly. Unsurprisingly, forecast errors were larger when inflation rates were high. Still, the forecast overall was distinctly better than predictions based on extrapolating the inflation trend of the four preceding quarters. …