The prediction of conflict constitutes a challenge to social scientists. This article explores whether the incorporation of geography can help us make our forecasts of political violence more accurate. The authors describe a spatially and temporally autoregressive discrete regression model, following the framework of Geyer and Thompson. This model is applied to geo-located data on attributes and conflict events in Bosnia over the period from March 1992 to October 1995. Results show that there is a strong spatial as well as temporal dimension to the outbreak of violence in Bosnia. The authors then explore the use of this model for predicting future conflict. Using a simulation approach, the predictive accuracy of the spatial–temporal model is compared to a standard regression model that only includes time lags. The results show that even in a difficult out-of-sample prediction task, the incorporation of space improves our forecasts of future conflict
Weidmann, Nils & Michael D. Ward (2010) Predicting Conflict in Space and Time, Journal of Conflict Resolution 54 (6): 883–901.