Nineteenth International Conference on New Directions in the Humanities

Abstract

The finding that countries facing interstate rivalries are more likely to experience civil wars is now well established in the literature. The financing of insurgencies or the weakening of state capacity between rivals are the main mechanisms that explain the link between rivalries and civil wars. However, previous works have not taken into account the relations of indirect rivalries, i.e. allies of rivals and rivals of allies. The omission of this type of relationship can lead to a simplification of complex relations between states and induce a bias in the analysis. In this paper, we develop a new set of data on the relations of indirect rivalries between states based on different matrix calculations. We demonstrate the importance of this type of relationship in the explanation of civil wars. More specifically, controlling for spatial distances between rival countries for a sample of 154 countries, we empirically demonstrate over the period 1970-2012 that 1) the presence of direct and indirect rivals does indeed exert a positive and significant effect on the risk of civil war; and 2) decreasing levels of military capacity of the focal state relative to rivals (direct and indirect) also influence the probability of internal conflict. Finally, we confirm the importance of our indicators through the use of the random forest algorithm, a machine learning method using decision trees.