Abstract

Territorial control is a key aspect shaping the dynamics of civil war. Despite its im- portance, we lack data on territorial control that are fine-grained enough to account for subnational spatio-temporal variation and that cover a large set of conflicts. To resolve this issue, we propose a theoretical model of the relationship between territorial control and tactical choice in civil war and outline how Hidden Markov Models (HMMs) are suitable to capture theoretical intuitions and estimate levels of territorial control. We discuss challenges of using HMMs in this application and mitigation strategies for future work.