Social Order in Civil War

War zones are usually portrayed as chaotic and anarchic. In irregular civil wars, however, they are often ordered. Furthermore, different forms of order often coexist in areas controlled by the same non-state armed group, where the behavior of both civilians and combatants vary substantially. What explains this variation?

This dissertation develops a theory of the creation of order in war zones by theorizing the behavior of non-state armed groups, the responses of local populations, and the effect of their interaction on wartime institutions. My central argument is that disorder emerges when armed groups fight for control with other warring sides or face internal principal-agent problems; under these conditions, combatants have incentives to discount future outcomes heavily, making social contracts with local populations unappealing. When a social contract is established, armed groups may intervene minimally or broadly in civilian affairs; their choice, I argue, depends on the likelihood of organized civilian resistance, which is, in turn, a function of the quality of pre-existing local institutions.

I test the theory with original data on the armed conflict in Colombia. Using data on local communities gathered with surveys and interviews, I find that armed groups' behavior vary not only across organizations as usually assumed, but also within them: both guerrillas and paramilitaries treat local communities quite differently depending on contextual as well as structural factors.

Civilians adapt to the changing dynamics of war in different ways depending on the institutional context. Using an original survey with civilians and ex-combatants, I find that recruitment is most likely in communities ruled by armed groups. Using quantitative and qualitative data, I also find that resistance usually emerges against armed groups' ruling attempts in communities with legitimate and effective institutions. These findings are supported by an in-depth study of a quasi-natural experiment in three neighboring communities.

By invalidating common assumptions about rebel and civilian behavior in the midst of war, and about the evolution of institutions in war zones, the project has implications on our study of several wartime and post-conflict dynamics. It also has implications on the timeless question about how aspiring rulers come to power.