Recent research on the new Phillips curve (NPC) (e.g., Galí, Gertler, and López-Salido 2001a) gives marginal cost an important role in capturing pressures on inflation. In this paper we assess the case for using alternative measures of marginal cost to improve the empirical fit of the NPC. Following Sbordone (2000), we derive the aggregation factors when firms use Cobb-Douglas with overhead labour and constant elasticity of substitution (CES) technologies. We estimate the NPC for Canada, the United States, and the euro area. Our structural results indicate that: (i) ignoring aggregation factors gives implausibly large estimates of the duration of price stickiness—in this respect, the aggregation factors improve the fit of the NPC substantially; (ii) the marginal cost measures based on CES technology improve the fit of the NPC relative to the Cobb-Douglas technology, particularly for Canada and the euro area; (iii) for Canada, the backward-looking component of inflation is quite strong relative to the United States and the euro area; and (iv) the incorporation of open-economy considerations for Canada does not yield better estimates of the NPC.

Also published as:

European Economic Review (0014-2921)
August 2005. Vol.49, Iss.6, pp.1571-1602