S ~ C ~ and Tlymmics of the Global Economy :

This article reports results from a quantitative network analysis of international commodity tradejlows designed to measure the structure of the world economic system and to identify the roles that particular counfries play in the global division of labor. I t improves on previous network-analytic studies of the world-economy in turo ways. First, by using a newly developed measure of regular equivalence, this operationalization of a nation's roles in the international system is methodologically superior to previous work. Second, we have built a dynamic aspect into the analysis by examining international trade networks at more than one point in time (1965,1970, and 1980). This allows us to answer questions about change both in the overall structure of the world-economy and in the positions of particular countries in the system. Ourfindings generally conform to the theoretically expectations of the world-system perspective as well as qualitative descriptions of r e m t changes in the internafional division of labor. What is the structure of the world-economy? How does it change? These are two basic issues in international political economy. The world-system approach suggests that the structure of the global system, and the roles that countries play within the international division of labor, is crucial in understanding a wide array of social, political, and economic changes within particular societies. The basic claim is that international connections, roles, and relationships are important independent variables in any causal analysis purporting to explain various dimensions of development within countries (on the general advantages of this structural approach, see Tilly 1984). While the premise that the global system plays a crucial role in most types of social change is now widely accepted, there is much less consensus on its * This project was funded by a seed grantfrom Public Policy Research Organization at UCZrvine and NSFgrant SES-88098446 to David Smith and a CRAY supercomputer computation grant through the San Diego Supercomputer Center to Douglas White. We wish to thank Peter Bearman, Steven Borgatti, Michael Burton, Christopher Chase-Dunn, G a y Gereffi, John Liu, Tonya Schuster, and two anonymous reviwers for commentary on earlier drafts. Direct correspondence to David Smith, School of Social Sciences, University of California, Zrvine