This paper investigates international macroeconomic fluctuations in light of NOEM (New Open Economy Macroeconomics) models. A model with four major economic disturbances (technology shocks, labor supply shocks, preference shocks, and nominal shocks) is analytically solved to derive theoretical long-run identification restrictions. These restrictions are used to estimate a structural VAR model for the three largest economies (the U.S., the Euro Area, and Japan) over the post Bretton Woods

more »

... The main findings are: (1) the signs of the dynamic responses are mostly consistent with theoretical predictions; (2) supply-side shocks (technology and labor supply shocks) explain most of the fluctuations in cross-country output deviations; (3) preference shocks are the dominant source of real exchange rate fluctuations; and (4) productivity shocks played a prominent role in the recent global imbalances (large U.S. external deficit), while the current account has usually been influenced by all four shocks, with no single shock dominant in all periods. 1 Introduction The decade circa Obstfeld and Rogoff (1995) has been one of the most productive for the field of open economy macroeconomics. Numerous papers followed the lead of Obstfeld and Rogoff, who promulgated the research program of building open economy macroeconomic (monetary) models that embed both sluggish price adjustment and optimizing agents. The "New Open Economy Macroeconomics" (NOEM) has become the conceptual anchor of research in the field and has replaced, for many researchers, the celebrated Mundell-Fleming-Dornbusch model as the central tenet of open-macro models. Brisk theoretical progress notwithstanding, the empirical literature has been relatively scant, especially for the VAR (Vector Auto-Regression) approach which imposes less a priori structure on data than the likelihood-based approaches-including Bayesian approaches-that are identified by estimating the full theoretical structure (e.g. Bergin, 2003Bergin, , 2006 Schorfheide 2005a, 2005b). Moreover, existing VAR papers have rarely taken a comprehensive look at the sources of international macroeconomic fluctuations. Many existing studies that embraced NOEM models in the VAR approach have focused on a particular type of shock (e.g. monetary or technology shocks) at the cost of leaving other shocks not well identified-examples include Kim (2001), Corsetti and Muller (2006), Kim and Roubini (2008) , and Lee and Chinn (2006). In addition, the identifying assumptions in some of these studies have not been shown to be fully consistent with the NOEM models. This paper provides a comprehensive VAR analysis of open economy business cycles on the basis of a NOEM framework. We estimate a structural VAR model, identified by long-run restrictions on the effects of structural shocks that are readily interpreted within NOEM models. The long-run restrictions themselves are consistent with the approximate long-run properties of NOEM models, including the one presented in this paper. Moreover, these long-run restrictions are also consistent with many types of DSGE (Dynamic Stochastic General Equilibrium) models, comprising models with both flexible and sluggish price adjustments. In this sense, our econometric approach is more general than the particular model that we study to derive long-run identification restrictions. We have two main objectives. First, we investigate the sources of international macroeconomic fluctuations, that is, the role of each structural shock in explaining cross-country movements in aggregate variables over business cycles. For this purpose, we focus on four key variables in open economies: relative labor productivity and output across countries, the real exchange rate, and the current account. We simultaneously investigate the movement of all four variables in an open-economy context, in contrast to many previous studies that have explored the sources of fluctuations in the real exchange rate or current account separately. 1 Second, we examine the transmission of structural shocks, interpreting the empirical evidence on the effects of each structural shock in light of theory. We analytically solve a 1 For studies on real exchange rate fluctuations, see Clarida and Gali 2 basic NOEM model subject to four structural shocks, and derive our identifying restrictions from the long-run implication of the model. The model and analytical solution-which bring out some shocks that have been put in the back burner in other analytically solved NOEM models-enable us to discuss the transmission of structural shocks more directly than in the usual simulation exercise with calibrated DSGE (NOEM) models. The four types of structural shocks are: technology shocks, labor supply shocks, taste shocks (shifts in preference for domestic versus foreign goods), and nominal (or monetary) shocks. The technology shock has been regarded as one of the most important sources of business cycles from the inception of the real business cycle studies and has also been at the center of various open economy macro models such as the intertemporal approach to the current account (Sachs, 1981; Glick and Rogoff, 1995), the equilibrium approach to the exchange rate (Stockman, 1980), and various international business cycle models including NOEM models. We consider labor supply shocks as another source of supply shocks that can have long-run effects on output, motivated by the closed economy literature (e.g., Gali, 1999) which found technology shocks to play a small role in explaining business cycles. The importance of labor supply shocks as a source of output fluctuations was also documented in earlier studies (e.g. Shapiro and Watson, 1988 , for the closed economy and Ahmed et al., 1993, for the open economy). The discussion on the role of shifts in preference for domestic versus foreign goods has been an important strand in international business cycle studies (Stockman and Tesar, 1991; Bergin, 2006). These taste shocks can also comprise the fiscal shocks (e.g. government spending shocks) owing to a shared longrun property-both taste and fiscal shocks change the relative demand for domestic vis-à-vis foreign goods. Finally, nominal shocks (e.g. monetary shocks) have been regarded as an important source of economic fluctuations in macroeconomics, though with time-varying consensus, and they have been a key topic of NOEM studies, such as Obstfeld and Rogoff (1995) and Betts and Devereux (2000). To preview some of the results, technology shocks are found to play a prominent role in accounting for Ahmed,