How forecasts evolve: the growth forecasts of the Federal Reserve and the Bank of England

Publication date: 29 Jul 2005 | Publication type: National Institute Economic Review | External Author(s): William Allen, City University and Terence Mills, Loughborough University | Journal: National Institute Economic Review No. 193

Summarised from the National Institute Economic Review, number 193, July 2005. To order a the full version of this article or a subscription, please contact Sage Publications by telephone: +44 (0) 20 7324 8701, email: <a href=""></a> or online at <a href=""></a>.

Monetary policy decisions taken now affect the economy over a period of time in the future. Therefore, they have to be based on forecasts of the future behaviour of the economy. Forecasts thus play a crucial role in monetary policy, and central banks devote a great deal of effort to forecasting, and to estimating the economic effects of interest rate changes.

Forecasts of GDP in the coming year are particularly important, because in most economic models, it is susceptible to influence by monetary policy decisions taken now, and it is a major determinant of inflation 1 - 2 years in the future, which is the time horizon over which many central banks try to manage the rate of inflation.

It is therefore of great interest to explore how central bank forecasts of GDP growth evolve through time. Central banks' accounts of their forecasts often describe the forecast by reference to current economic developments, but do not always relate the current forecast to earlier forecasts. The paper investigates how the GDP forecasts of the Federal Reserve and the Bank of England have reacted to emerging estimates of actual GDP growth. There are clear signs that the Fed has adjusted its forecasts in the light of estimated outturns that were different from what it had previously expected. In the case of the Bank of England, the signs are that, if anything, its reaction to slower than expected economic growth was to revise up its forecast for the coming year. However, there are tentative signs that, since 2001, the Bank of England has reacted to emerging estimates in the same kind of way as the Fed.

The paper goes on to explore whether there are readily-detectable systematic errors in the forecasts of either central bank, in the sense that the pattern of forecast errors is statistically related to variables that were known when the forecasts were made. There is no evidence of systematic errors in the Fed forecasts. There is, however, evidence of systematic errors in Bank of England forecasts in the data for the period 1998 - 2005: the data suggest that the forecasts would have been more accurate had they taken less account of the Bank's earlier forecasts and more account of the most recent GDP estimates published by the Office for National Statistics. However, the evidence of systematic errors in Bank of England forecasts is not apparent in the post-2001 data.