Phillips curve instability and optimal monetary policy

This paper assesses the implications for optimal discretionary monetary policy if the slope of the Phillips curve changes. The paper first derives a ‘switching’ Phillips curve from the optimal pricing decision of a monopolistic firm that faces a changing cost of price adjustment. Two states exists, a state with a high cost of price adjustment that generates a ‘flat’ Phillips curve and a low-cost state that generates a relatively ‘steep’ curve. The second aspect of the paper constructs a utility-based welfare criterion. A novel feature of this criterion is that it has a relative weight on output gap deviations that is state dependent, so it changes with the cost of price adjustment. Optimal monetary policy is computed subject to the switching-Phillips curve under both ad-hoc and utility-based welfare criteria. The utility-based criterion instructs monetary policy to disregard the slope of the Phillips curve and keep its systematic actions constant across different states. This stands in contrast to the prescription coming under the ad-hoc criterion, which advises monetary policy to change its systematic behavior according to the slope of the Phillips curve.