Introduction

The classic time-consistency problem results in a counterproductive inflation bias to discretionary monetary policy (Kydland and Prescott, 1977; Barro and Gordon, 1983). The standard formulation of the problem is based on the following time sequence of events: wage setters negotiate nominal wage contracts, setting wage growth equal to expected inflation; the policy maker then sets the inflation rate. Output is stimulated if the inflation rate exceeds the nominal wage growth fixed in the wage setters' contracts. However, such output stimulation comes at the price of increased inflation. The equilibrium inflation rate trades off the policy maker's output stimulation and inflation stabilization goals. Wage setters are rational and thus understand the policy maker's decision rule. Consequently, their inflation expectations do not systematically deviate from actual inflation. It follows that, in equilibrium, the policy maker cannot stimulate output, but her futile attempt to do so leads to an inflation bias. The policy maker would be better off if she could credibly commit herself to refraining from attempting to stimulate output. Since she cannot fulfill her output goal, she could then at least achieve her inflation goal.

The reputational solution

In an infinite-horizon repeated game setting, there exists a reputational solution to the time-consistency problem in monetary policy (Barro and Gordon, 1983); that is, the inflation bias can be reduced via reputational trigger-punishment strategies (Friedman, 1971).