International network analysis based on big data: The case of economic cooperation among the G20

Abstract

How the structure of the international economic cooperation network among the G20 countries has evolved over the past quarter century, and what are the roles and power levels of different countries in the network? We collect big data on economic cooperation events among the G20 countries from 2000 to 2024, and divide the 25 years into four periods using structure changepoint detection algorithms to separately construct international economic cooperation networks. We calculate and compare the multifaceted characteristics of the G20 economic cooperation network and trade network based on network analysis. At the macrolevel of network structure, we find that even when trade relations among the G20 countries became closer, economic cooperation among them became more distant. At the mesolevel of group classification, the result of community detection and the disparity of edge weight of the G20 economic cooperation network and trade network show clear but different regional characters. At the microlevel of power distribution, we measure and compare the power levels of the G20 countries in the economic cooperation network and their dynamics over time using weighted centrality indices based on current flow and random walks. By combining network analysis and event big data, we study economic interactions between countries in a unique approach that differs from traditional approaches employing trade volume or institutional data (e.g. free trade agreements).