Point Process Models Predict Insurgency Hotspots in Sub-Saharan Africa

This study introduces an innovative method for predicting conflict zones using point process models. Geographic Conditions as Predictors: Leveraging classic literature and recent research, the authors demonstrate how geographic factors can be used to anticipate violence locations without relying on causal testing of specific theories.

Novel Cross-Validation Design: The paper employs a unique approach showing that quantitative insights into civil conflict micro-foundations are generalizable enough for reliable out-of-sample predictions.

The Sub-Saharan Africa Context: Focusing specifically on ten countries in this region experiencing post-Cold War insurgencies, the findings offer targeted applications for policymakers and researchers seeking to understand insurgency patterns.