Incentive Contracts for Central Bankers under Uncertainty: Walsh-Svensson Non-Equivalence Revisited

We look at the implications of uncertain monetary policy preferences for the targeting and contracting approach to monetary stability. It turns out that in presence of uncertain preferences a linear incentive contract in the sense of Walsh (1995) performs better in terms of social welfare than an explicit inflation target as proposed by Svensson (1997). The reason is that, although both approaches can get rid of the inflationary bias, the impact of uncertain preferences on the variance of inflation will be considerably higher with an inflation target. We also find that on top of an optimal linear contract or target, a quadratic contract, in the sense of Rogoff's (1985) "weight-conservative" central banker, improves the outcome. In the case of an inflation target, a more conservative banker is needed than with a Walsh contract.