Monetary policy and learning

A new strand of macroeconomic literature examines the relationship between learning and monetary policy-how monetary policymakers learn about the economy as they try to achieve their goals, how the public learns about policymakers' objectives, and how the public's learning, in turn, changes the way monetary policy works. An Atlanta Fed conference in March 2003 brought together some of the main contributors to this emerging literature. ; In the conference keynote address, reprinted here, Lars Svensson focused on what constitutes good monetary policy and how it is related to central-bank learning, how private-sector learning benefits from central-bank transparency, and how good monetary policy is best modeled. ; Good central banks, he noted, engage in forecast targeting: setting interest rates so that their projections of inflation and the output gap are consistent with their objectives. In particular, he argued, good central banks must devote considerable resources to monitoring and extracting private-sector expectations from various sources, use these expectations among other inputs in the forecasting process, and respond appropriately when the expectations affect the central bank's projections of inflation and the output gap. ; Svensson also stressed that central-bank transparency can improve private-sector learning and thereby induce better private-sector decisions. ; Finally, Svensson emphasized that good monetary policy is best modeled in the same way the private sector is modeled-not with ad hoc reaction functions, or instrument rules, but as goal-directed, optimizing policy with the help of targeting rules.