What is a realistic value for price adjustment costs in New Keynesian models?

Abstract

Rotemberg's (1982) price adjustment costs framework is a popular sticky price specification; yet, the data provides little information on the magnitude of those costs. This article finds a plausible range of parameterizations for those price adjustment costs. Our results show that the specific size of the price adjustment costs depends on the average markup of price over real marginal cost and the average time firms wait to reoptimize their price. In particular, the price adjustment costs are higher when the average markup is lower and the mean time between price reoptimizations is longer.

Notes

¹ Chari et al . (2000) show that quick price adjustment prevents models from producing persistent output effects.

² Rotemberg and Woodford (1997) present survey evidence supporting this claim.

³ Ireland (;2004) also refers to this relationship between the Rotemberg (1982) and Calvo (1983) pricing specifications but does not show explicitly how price adjustment costs are related to the average markup of price over marginal cost and the average time firms wait to reoptimize their price.

⁴ While the microfoundations of the Rotemberg (1982) and Calvo (1983) pricing specifications are different, the resulting inflation dynamics produced by each model are the same. This identical macroeconomic behaviour allows us to calibrate the Rotemberg (1982) specification by appealing to the microeconomic evidence relevant to the Calvo (1983) specification.

⁵ Ireland (1997), 2001, 2003) and Kim (2000) estimate Rotemberg's (1982) price adjustment costs but have no method of evaluating the plausibility of their estimates.