Central bank economists and academic economists conducting research on the design of monetary policy have made significant advances in recent years. This work has led to a clearer understanding of the desirable properties of interest rate rules, the role of announcements and communication, and the consequences of inflation targeting for both inflation and the real economy. Dynamic stochastic general equilibrium (DSGE) models have been extended from the small-scale, often calibrated versions initially employed to address policy issues to much larger models that are estimated using Bayesian techniques. Many central banks now use these models for policy evaluation.1 Much of this work neglects one of the key issues that policymaker face, however: the pervasive role of uncertainty.