Can inflation contract discipline central bankers when agents are learning?

This paper studies how the government should design a linear inflation contract to deal with the time-inconsistency problem arising from incentives for the central bank to exploit the inflation-output tradeoff with an overambitious output-gap objective when private expectations are based on adaptive learning. An intertemporal tradeoff due to learning leads the central bank to accommodate less the effect of inflation expectations and cost-push shocks on inflation. An optimal linear inflation contract is able to achieve many of the benefits, i.e., reducing inflation bias and stabilization bias, resulting from central bank conservatism and inflation targeting rules. The government can impose either a long-term or a short-term contract. The first is equivalent to appointing a hawkish central banker. The second implies that inflation penalty rate should be adjusted for inflation expectations in each period, and could be positive or negative, i.e., the central banker should shift between hawkish and dovish stances depending on inflation expectations and the speed of learning.