Credibility and the effectiveness of inflation targeting regimes

This paper investigates the design of monetary policy in inflation targeting regimes. It is partly motivated by our view that, despite the widespread adoption on inflation targeting regimes in a number of countries including the UK, the policy debate is insufficiently concerned with the central question; how exactly should interest rates be moved to control inflation? We argue that the resulting lack of clarity in official descriptions of the practical operation of monetary policy in the UK may be undermining the credibility and hence the likely effectiveness of the announced regime of inflation targeting. The paper describes a stylised model of the determination of inflation which is used to illustrate some of the technical considerations necessary in the appropriate design of an inflation targeting regime. In particular, it is argued that policy should be defined and announced in terms of a relatively simple feedback rule which relates changes in interest rates to inflation and to other indicators. The advantages of real rather than nominal interest rate rules are demonstrated. Simple output indicator regimes are compared with the fully optimal policy rule. The effect of incomplete credibility where private sector agents need to learn about the true policy regime are illustrated for different rates of learning. By calibrating a stochastic version of the model to UK data, the expected variance in inflation is illustrated for different feedback regimes. The effect of incomplete credibility in a stochastic context is illustrated by showing the effects of assuming that the private sector's perception of the inflation target depends on the track record of the monetary authorities.