This paper compares non-state with state-based armed conflicts. I rely on theoretical arguments by Kaldor, Weinstein and Kalyvas in order to link differences in violent actors (their nature, number and motives) with differences in the character, intensity and duration of applied violence. The subsequent empirical analysis explores differences and changes over time in the incidence and significance of these sub-types of internal fighting. In addition, it tries to uncover whether non-state armed conflict is characterized by a significantly larger number of involved actors, whether it significantly more often happens in a context where conflict resources are known to occur and/or produced and where state structures are particularly weak, whether it causes significantly more battle-related deaths and whether it is of a significantly longer duration as compared with state-based internal fighting. This study not only covers both types of violent events but three levels of analysis: the war-, dyadic-conflict- and conflict-episode-level.
