Testing the convergence hypothesis: a longitudinal and cross-sectional analysis of the world trade web through social network and statistical analyses

Different theoretical perspectives support opposite views on convergence: although the dominant view is that convergence is the inevitable outcome of globalization, divergentists (that is, world-system economists and, potentially, also evolutionary geographic ones) argue that convergence forces could be annihilated by the need to keep power relationships within the international division of labor. Even when limiting the convergence issue to international trade, the debate has so far been inconclusive, because various studies have dealt with different and/or short time series or selected too small and different sets of countries. Moreover, none of these studies have analyzed trade patterns and have instead been limited to the aggregate value. Here, through a social network analysis, we examine the world trade patterns from 1980 to 2016 (1980–1992, 1993–2007 and 2008–2016) of at least 164 countries, which have been divided into import and export patterns and into four groups of countries: from core countries to far periphery ones. We test the convergence hypothesis in two directions: the level and trend of convergence, and its possible determination by means of structural or economic globalization, measured in terms of exchanges density and economic values, respectively. We have found that the convergence hypothesis only seems to be confirmed when considering the pure structural aspect and core countries. Conversely, economic convergence—which also includes the structural dimension—has been found to be high for core countries and to increase over time. Moreover, our analysis shows that economic globalization influences convergence, albeit in a strongly negative way. Therefore, our findings seem to support divergentists and the convergence hypothesis should be rejected.