We are Here to Help! How Third Party Military Interventions Impact Civil Conflict at the Sub-National Level. Evidence from Afghanistan

Abstract

Early 21st century warfare is dominated by conflicts that occur within the boundaries of a state between a government and rebel groups. Nearly 40% of these civil conflicts are joined by an international third party who seek to aid one side or the other. How do these biased deployments of third-party military troops in civil conflicts affect violence at the local level? I argue that the effect of third-party troops is conditional on whether the side they support controls the local territory. Territorial control, or its absence, shapes the way the population interacts with the opposing group and shapes their option for the use of violence. This dissertation supports this argument with local-level data from the ISAF Mission to Afghanistan. I find that the deployment of NATO troops at the battalion level or higher in areas controlled by the Afghan Government is associated with decreased rebel-initiated violence, while the presence of NATO troops in areas controlled by rebels is associated with an increase in rebel-initiated violence. However, this result does not hold for troop deployments below the battalion level. Using novel data from Paktia and Khost provinces, I find that any pacifying effects NATO troops have on violence do not outweigh their presence alone. These results support my argument and are robust to model specifications, fixed effects, and lagging the dependent variable. This study emphasizes the importance of territorial control as an explanation for the sub-national variances in violent outcomes where third party troops intervene.