Book Project

My book project offers a subnational perspective on civil war peacemaking by examining how local peace agreements shape interactions between armed groups and communities during civil wars. While existing work emphasizes national-level settlements, I show that local agreements operate as strategic tools for both civilians and non-state armed groups seeking to consolidate influence, reduce uncertainty, and access resources in contested areas. Using theories of rebel strategy and civilian agency, I combine original data with flexible machine-learning models to forecast when and where agreements emerge, finding that territorial contestation, state capacity, and proximity to economic assets are strong predictors. I then evaluate the consequences of these agreements through panel data, selection models, and differences-in-differences designs, demonstrating that local agreements significantly reduce violence against civilians and reshape spatial patterns of violence beyond the immediate agreement zones. Together, the project shows that local agreements are not peripheral outcomes of national peace processes but central mechanisms that actively structure conflict dynamics on the ground, highlighting the importance of subnational and predictive approaches for understanding wartime governance and civilian protection.