Rethinking models of civil war settlement

The current study is a replication and expansion of an earlier piece by Hartzell, Hoddie, and Rothchild (2001) exploring civil war settlement characteristics and stability. Their research focused on the importance of territorial autonomy provisions and the role of third party guarantors in predicting settlement success and stability. They employed a model that controlled for such conflict characteristics as international system structure, nature of the previous regime, conflict duration, conflict issue, and conflict intensity. Our study replicates Hartzell et al. using the Regan (2001) civil war dataset, employing a broader definition of conflict and a more stringent definition of when a conflict has ended. The results presented here differ from the previous study to have find that the importance of territorial autonomy provisions as a predictor of settlement stability is greatly diluted when one examines only those conflict settlements that have lasted for six months or longer without reciprocated violence. Third-party guarantors, however, remain a strong factor determining conflict settlement stability in the revised dataset. We also expand the previous work by including cases where the conflict ended through military victories, as well as by adding a dimension to the negotiated settlement variable that separates settlements that were coerced due to external military presence from those that were not coerced.