PDF《RESEARCH DIVISION》

pricing decisions are not always based on current information. Mankiw and Reis show that a rational expectations monetary model featuring sticky information is capable of overcoming the aforementioned shortcomings of the sticky-price models.1 The purpose of this paper is three-fold. First, we ask whether the typical output-in?ation dynamics emphasized by the literature are a world-wide phenomenon. Our empirical investigation shows that the answer is Yes . Just like in the U.S., in?ation in other industrial countries is also highly persistent and systematically lags output by several quarters. More importantly, we show that such in?ation dynamics are highly synchronized across countries. Namely, a high in?ation in the U.S. after an output boom is often associated with a high in?ation in Europe at the same time. This phenomenon of positive cross-country correlation in in?ation is linked to the fact that in?ation systematically lags output in each country and that output is positively correlated among countries. However, the data also indicate that the cross-country correlation of in?ation is much stronger than the cross-country correlation of output. For example, in our sample of

18 OECD countries, the average correlation of in?ation is about 0:6 while that of output is about 0:2. What forces are responsible for the global comovements in in?ation dynamics? Modern mon-

1 State-dependent pricing or hybrid sticky-price models can also overcome these shortcomings. See, e.g., Dotsey and King (2005) and Christiano, Eichenbaum, and Evans (2005).

2 etary theories attribute the persistent short-run in?ation dynamics to exogenous monetary shocks under nominal rigidities (see, e.g., Fuhrer and Moore, 1995;

Christiano et al., 2005;

Dotsey and King, 2005;

Mankiw and Reis, 2002;

and Woodford, 2001, among many others). Yet we …nd that movements in the money stock are not signi…cantly and systematically correlated across countries. This fact alone, however, may not necessarily imply a low correlation in in?ations since business cycles can propagate across country borders even though they may be driven by country-speci…c monetary shocks. For this reason, the second purpose of our paper is to investigate whether stan- dard monetary models with nominal rigidities can simultaneously account for the within-country output-in?ation dynamics and the cross-country output-in?ation correlations based on the cali- brated international covariance of monetary shocks. We …nd the answer to be negative. Under country-speci…c money growth shocks, the models are not able generate strongly positively cor- related in?ation rates across countries while maintaining their ability to account for the domestic output-in?ation relationship. This …nding casts doubt on the view that exogenous monetary shocks are the main driving force of short-run in?ation dynamics. The third purpose of this paper, which can be viewed as an independent contribution to the literature, is to provide a simple solution technique to solve DSGE models with sticky-information. While conceptually simple, a dynamic general equilibrium model with sticky-information is di?cult to solve. The di?culty arises because of the potentially large number of lagged expectation oper- ators, which can create an extraordinarily large state space. In our method, variables with lagged expectations are replaced by their forecast errors with undetermined coe?cients. Hence the prob- lem of a large state space is avoided (see Appendix B for details of our solution method). Given that the sticky-information model is becoming increasingly popular, we think this solution technique is a timely contribution to the literature, especially for those who want to estimate sticky-information models with a large number of lagged expectation operators by traditional econometric methods.2 The remainder of the paper is organized as follows. We …rst present three stylized facts about short-run in?ation dynamics: (a) in?ation is highly persistent and it lags output within each country;