The Microdynamics of Violence and Order: Variations in Community Security Processes

Internal state conflict is the current predominant form of conflict, and most of these conflicts experience much sub-state variation in their severity of violence. This dissertation examines why violence varies in intensity across an intrastate political conflict. I analyze neighboring communities that experienced varying levels of violence within two internal state conflicts: the "Troubles" in Northern Ireland and the 2007 election crisis in Kenya. Responding to limitations in explanations for political violence that do not disaggregate the "state" and groups in a way that can explain variation between communities with similar demographics, I argue that interactions between local state agents and non-state radicals affect the intensity of violence. When contention develops in a location, non-state actors radicalize and some local state agents deviate from their institutional role as providers of law and order to support radicals with whom they identify. This affiliation makes an escalation in violence more likely.;Within each conflict I examine two communities that share similar economic and demographic characteristics but experience different degrees of violence. Through the use of process tracing with data collected from field research including interviews and archival sources, I show the connection between state agent alignment and radical non-state groups for the escalation of violence in each paired comparison. By demonstrating a similar dynamic of violence in two countries where the mode of violence, length of conflict, and political environment are different, I conduct a second level of comparative analysis that further supports by argument. I find that state actors such as local security force members in Northern Ireland and police and bureaucratic officials in Kenya were crucial to the trajectory of violence through their identification with extremists.;This dissertation develops a new microlevel theoretical perspective emphasizing the influence of state and non-state radical interactions while adding to the small body of research examining sub-state variation. The major implication of the study is these interactions are crucial to explaining the intensity of violence in particular communities and the intensity of violence in the overall intrastate conflict.