Political Distance in the Network of International Trade

Abstract: This study is concerned with the underexplored link between political distance and international trade. It considers the network of international trade which is investigated using Temporal Exponential Random Graph Models. While related to gravity models in the data used as well as the choice of covariates, this allows for greater insight, particularly into higher order dependencies in the network of international trade, considering not only dyadic country pairs but the network as a whole. The results suggest a negative relationship between political distance, based on a measure of difference in votes at the UN General Assembly, and the probability of a trade tie. Additionally, it finds that this effect is present when political distance is defined using the democratic scale of the Polity V dataset, but for the autocratic scale only at more granular levels of industry. The network model suggests a grouping of countries with low political distance having the strongest effect on the likelihood of a trade tie, giving support to the relationship between democratization and an increase in trade activity. Additionally, it suggests that countries that are distant in the democratic scale are integrated into the international trade network to bind them to the international order, given they are not far apart in the autocratic scale, a basis for weaponized interdependence exclusive to democratic states. The results highlight the importance of political distance in international trade and the role of politics in trade relationships as a factor in the likelihood of ties between countries.