The Curvilinear Relationship between Political Terror and Internal Armed Conflict

I examine the relationship between political terror and the annual incidence of internal armed conflict. Studies have found that other factors associated with violence have a curvilinear relationship to internal conflict, including regime type, GDP, and ethnic fractionalization. I seek to find if political terror has a significant impact on instances of conflict and, more particularly, whether a curvilinear relationship exists between these two variables. I hypothesize that the relationship will be significant and in the shape of an inverse U. If a country uses no political terror, or extensive political terror, then the number of internal armed conflicts will be low. If the level of political terror employed by a regime is somewhere in the middle, I expect to see a greater number of internal armed conflicts. The results of my test do not support my hypothesis. There is a statistically significant relationship between political terror and instances of internal armed conflict, and the relationship is curvilinear. However, the relationship is convex, rather than being concave as I had predicted. This is a relationship I had yet to come across in previous literature. It could add a new dimension of thought for policy makers and scholars, requiring them to consider different roles violence may play in preventing and starting internal conflicts