Discussion of “ Has Globalization Changed the Phillips Curve ? ” ∗

This paper by Eugenio Gaiotti is a refutation paper: it tests and rejects the idea popularized by Borio and Filardo (2007) that globalization explains the flattening of the Phillips curve. While several papers have already been devoted to assessing the Borio and Filardo theory, the distinctive feature of the present article is that it uses microeconomic data. Microeconomic data can be helpful in the present context because they help circumvent the lack of variability in macroeconomic data, and also because of the fact that expectations are not observed. The main result and message of the paper is that globalization had no effect on the price/activity relationship. This result is found to be robust to using various measures of globalization. It is worth underlining that the paper uses a very rich panel data set of Italian firms. One feature of the data set is that it contains observations at the firm level on both price and real quantities. By contrast, much of the recent research on price setting (Bils and Klenow 2004, Dhyne et al. 2006) uses data on price quotes that contain little, if any, information on costs or the firm environment.1 The paper has two distinct but related dimensions. First, it provides empirical evidence on price setting at the firm level. Second, it discusses implications of these findings for the Phillips curve, an aggregate macroeconomic relationship. This discussion will tackle the two dimensions in turn, providing some remarks on the econometric specification, then discussing the link between microeconometric evidence and the breakup of the (reduced-form) Phillips curve.