On the Welfare Costs of Imperfect Information for Monetary Policy

This paper considers a policymaker who is uncertain about the state of the economy but observes a potentially large set of noisy indicators. We evaluate the welfare implications of exploiting different information sets in conducting policy. We propose a simple, unified representation of the model equilibrium, whether the policymaker follows an arbitrary policy rule or commits to optimal policy, and whether he perfectly observes the state, optimally estimates it, or responds naively to observed indicators. Using a stylized quantitative model, we argue that systematically exploiting all available information to assess the economy's state is likely to yield substantial welfare gains.