The world trade network: Country centrality and the intra-Africa trade

Abstract

Global trade is based on a group of multifaceted interactions between nations that can be modeled as an incredibly dense network of intertwined agents. On the one hand, this network might favor the trade performance of countries, but on the other, it can also discourage international trade. In this article, we investigate whether and how much the structure of the trade network may explain for the performances of intra-African trade among certain African nations. We calculated the centrality indexes for the nations and applied them to regression analysis. We then employ a negative binomial regression framework with these indicators as target regressors. In doing so, we also compare the effects of different measures of centrality- specifically, the degree centrality measures and the clustering coefficient. Our findings suggest that, albeit boosting the degree centrality index tends to improve the trade flows inside Africa, on average, the intra-African trade flow was shown to be negatively impacted by the clustering coefficient, which is congruent with theory and our predictions.