While the private sector generates dozens of macroeconomic forecasts, only three "official" forecasts emanate from government agencies. The oldest derives from the Employment Act of 1946, which created the Council of Economic Advisers and requires the Council to submit an annual report to the U.S. Congress. Starting in the early 1960s, this report provided an explicit numerical forecast of current-dollar or nominal GNP growth for the coming year; initially, the breakdown between the rate of real GNP growth and the rate of inflation, as measured by the implicit GNP deflator, had to be inferred from the text of the report. The Congressional Budget Act of 1974 created the Congressional Budget Office and requires it to present periodic reports on fiscal policy to the congressional budget committees. Significantly, this Act also established the practice of developing and presenting the federal budget, based on an economic projection, over a five-year horizon. Finally, in compliance with the Full Employment and Balanced Growth Act of 1978 (often referred to as the "Humphrey--Hawkins Act" in honor of its primary legislative sponsors), the Chairman of the Federal Reserve Board reports biannually to the Congress, presenting the economic projections of the members of the Federal Open Market Committee.
The Council of Economic Advisers' (CEA's) forecasts have been analyzed and compared with private forecasts many times. (See, for example, Moore 1977 and 1983 and McNees 1977 and 1988.) The Congressional Budget Office (CBO) has analyzed its own forecasts periodically, most recently in Reischauer (1995). This article updates these previous studies of CEA and CBO forecasts through 1994 and presents what may be the first published analysis of the Federal Open Market Committee's (FOMC's) "Humphrey-Hawkins" forecasts. It also incorporates alternative measures of changes in real output and prices recently developed by the Bureau of Economic Analysis (described in Young 1992 and 1993).
The results of the broader and longer data set used in this study alter slightly some of the conclusions of the previous research:
(1) Previous research almost uniformly has shown that the one-year-ahead "official" forecasts are about as accurate as forecasts obtained from surveys of private sector forecasters.(1) This study suggests the CEA's two- and four-year-ahead forecasts of real GNP have been slightly less accurate than, and the CBO's forecasts about as accurate as, the private sector forecasts. The lower accuracy stems from an optimistic bias in expectations of long-term real growth. On the other hand, the FOMC's forecasts issued each July for the following year are shown to have been somewhat more accurate than a standard private-sector forecast.
There are ample reasons to be skeptical that these differences will persist in future forecasts, however. For example, most of the advantage of the FOMC's forecasts derives from a superior performance in the early 1980s; since that time, private forecasts have been about as accurate as the FOMC's forecasts.
(2) Previous research has found both nominal and real GNP forecasts more accurate than simple rules of thumb, but at least one earlier study suggested that the "official" inflation forecasts were...