Recognizing and Communicating Uncertainty in Monetary Policy Projections

It is well recognized that projections of monetary policy objectives and policy actions are uncertain. Inflation-targeting banks, to preserve credibility, try to make clear that their projections are uncertain by providing interval forecasts or fan charts rather than simple point forecasts. Probability models, if they were reliable, could help make those intervals and charts more accurate and therefore less likely to erode central bank credibility. But constructing believable model-based fan charts is difficult. There are often multiple, conflicting, plausible models. There is a history of forecast accuracy that may conflict with the model-based measures. This paper takes up these issues and makes some suggestions about methods.