Monetary Policy with Imperfect Knowledge

This paper studies the formulation of monetary policy in a changing environment when knowledge regarding some aspects of the structure of the economy is imperfect and an adaptive learning technology is available to the policymaker and economic agents. As a benchmark, we develop a simple model of the economy and identify efficient monetary policy rules under the assumptions that knowledge of the economy is perfect and expectations are formed rationally. We then relax the assumption of rational expectations and examine the design of policy when agents and/or the policymaker must rely on adaptive learning to form expectations. We show that policies that are efficient under rational expectations are no longer efficient under imperfect knowledge, and, conversely, that efficient policies under imperfect knowledge would appear inefficient to the outside observer who assumed perfect knowledge. Using these results, we discuss the role of imperfect knowledge on the evolution of monetary policy in the United States over the past twenty years.