Conservatism in Inflation Forecasts

Abstract: This paper examines the extent to which professional forecasters exhibit conservatism in their announced inflation forecasts. Conservatism occurs as a result of a secondary objective captured in the forecaster’s loss function: they would like to avoid having large forecast errors, but also would like to avoid making large revisions to previously announced forecasts. As a result of forecasters’reluctance to make large revisions, announced inflation forecasts change less than true inflation expectations. The Anderson-Rubin approach is used to determine the degree of conservatism in U.S. inflation forecasts, and evidence suggests it may be quite substantial. The findings in this paper have implications for both research and monetary policy that uses inflation forecast data. In the case of a central bank that is interested in understanding the extent to which inflation expectations are well-anchored, the presence of conservatism could lead to a series of inflation announcements that simply adjust too slowly or by too little to provide a useful guide for monetary policy. This paper presents an estimate of the degree of conservatism in inflation announcements as well as a method to undo the conservatism bias and therefore recover a series of forecaster’s "true" or conservatismadjusted inflation expectations. This true inflation expectations series is more volatile than the announced inflation forecasts observed in the data for most forecasters, suggesting that inflation expectations may not be as well-anchored as inflation announcements may suggest.