This paper uses sector-level data for 30 countries and up to 28 years to provide a quantitative account of the sources of international GDP comovement. We propose an accounting framework to decompose comovement into the components due to correlated shocks, and to the transmission of shocks across countries. We apply this decomposition in a multi-country multi-sector DSGE model. We provide an analytical solution to the global influence matrix that characterizes every country's general

more »

... GDP elasticities with respect to various shocks anywhere in the world. We then provide novel estimates of country-sector-level technology and non-technology shocks to assess their correlation and quantify their contribution to comovement. We find that TFP shocks are virtually uncorrelated across countries, whereas non-technology shocks are positively correlated. These positively correlated shocks account for two thirds of the observed GDP comovement, with international transmission through trade accounting for the remaining one third. However, trade opening does not necessarily increase GDP correlations relative to autarky, because the contribution of trade openness to comovement depends on whether sectors with more or less correlated shocks grow in influence as countries increase input linkages. Finally, while the dynamic model features rich intertemporal propagation of shocks, quantitatively these components contribute little to the overall GDP comovement as impact effects dominate.