The last fifteen years have seen the emergence of widespread consensus that optimum monetary policy is designed on the basis of three pillars: a short-term official rate of interest as the sole policy instrument and the placing of that instrument in the hands of a central bank which is (a) independent of government and (b) transparent in its decision-making. We take a critical look at each of these. In the first case, we focus attention on the failure of mainstream economics to recognise the choice of instrument and the implications of its adoption. In the case of independence we argue that he theoretical case for independence has been misunderstood and that it is not an essential requirement for successful policy. We also show that ‘independence’ is not best measured against a checklist of statutory characteristics. As regards ‘transparency’ our argument is slightly different, though we come to a similar conclusion. Unlike independence, ‘transparency’ does address a real problem for central banks. However, the evidence suggests that transparency is not the only, or even the best, solution. A variety of evidence tells us that agents can understand and anticipate the actions of the most secretive institutions.
