Reputational Constraints on Monetary Policy

Abstract

Recent advances in game theory have made it possible tostudy monetary policy credibility in a structured fashion. Somehave concluded from these models that reputational considerationssubstantially discourage the monetary authorities from everattempting surprise inflations. Hence legal constraints on moneysupply growth are unnecessary and can only be harmful. In thisstudy, I critically assess a number of alternative models ofmonetary policy reputation, including some new variants. Thebulk of the paper is concerned with comparing specific details ofthese models. One general conclusion is that although this firstgeneration of monetary policy reputation models yields asignificant number of important insights, it is premature toargue that time consistency is not a major issue in the design ofmonetary policy institutions. The main problem is that themodels either yield a multiplicity of equilibria, or/and yieldconclusions which are very sensitive to apparently minor changesin the information structure. Whereas an optimal reputationalequilibrium may arise without any explicit cooperation amongatomistic private agents, it is not (yet) clear why we shouldexpect them to coordinate on the most favorable equilibrium.Strategic uncertainty may be an important drawback toinstitutional setups which place few constraints on monetarypolicy.