Expectations, Learning and Discretionary Policymaking

Abstract

When people's policy forecasts are generated by exact replication of the policymaker's decision problem, present and past policy has no effect on expectations of future policy. However, when people's policy forecasts are based on the policymaker's present and past actions, for example through learning and estimation of variables or parameters, policy affects expectations of future policy, and the time-consistent solution improves. Therefore, a successful policymaker needs to know how present policy affects people's policy expectations.

Keywords: Commitment, discretion, learning, expectations

JEL Classification: E52, E58, E61