Comment

The implications of globalization and increased openness for optimal monetary policy have been the subject of a rapidly growing literature. In this paper, Richard Clarida uses the New Keynesian two‐country model of Clarida, Gali, and Gertler (2002) to partly address such questions. The paper focuses on the degree of trade openness. It does not address the implications of financial globalization, which is arguably an equally important feature of the current wave of globalization. The Clarida et al.model has become a benchmark for studying the interaction of monetary policy in a two‐country world. The model is extremely elegant and allows seeing exactly what is going on. At the same time, it is very stylized, focusing on the terms of trade as the main international transmission channel. International risk sharing is assumed to be perfect. In the second part of the paper, Clarida then performs an empirical analysis of recent Fed behavior. This exercise is in the spirit of Taylor’s (1993) original analysis using “eyeball econometrics.” Themain question asked is whether the Taylor rule can fit recent Fed behavior. Clarida argues that globalizationmay have affected some of the necessary inputs for calibrating a policy rule such as the equilibrium real interest rate and expectations of inflation. He therefore uses expectations data from financial markets to overcome this problem. These expectations measures should be immune to breaks and changes in regime that may plague the use of traditional instrumental generalized method of moment estimation of forward‐looking Taylor rules. The Clarida et al. model gives hints about how openness may affect the optimal reaction coefficients in a Taylor rule. It would therefore be useful to explore some of those hints in the empirical part of the paper.However, in the second part the reaction coefficients in the closed‐economy Taylor rule are basically kept constant. How does openness affect the optimal relative weight on output gap versus inflation stabilization? In the stylized model of Clarida et al.