Title: Explaining Cohesion, Fragmentation, and Control in Insurgent and Paramilitary Groups
Name: Paul Staniland

Assistant Professor
Department of Political Science
University of Chicago

Year: 2003
Type: Dissertation Fellowship
Summary: This research project investigates the origins of cohesion and fragmentation in armed non- state actors, primarily insurgent groups. The cohesion of armed organizations plays an important role in determining the length and intensity of civil wars, patterns of violence against civilians, possibilities for negotiation and de-mobilization, and forms of local political order amidst war. Existing explanations tend to prioritize the ideology of organizations (Communist, Islamist, etc), the structure of state power and political opportunities, or the resource flows upon which groups rely, such as lootable minerals and state sponsorship. This project argues that these explanations have trouble explaining the much more subtle range of variation that we observe.

Instead, this project focused on the structure of the social networks upon which organizations are formed and on the evolution of these networks during violent conflict. New insurgent groups tend to appropriate prior social networks and institutions for the purposes of war, and these new organizations reflect the patterns of social ties upon which they draw. For instance, certain Sunni insurgent groups in Iraq were heavily drawn from particular veterans networks, and the Taliban in Afghanistan has relied on overlapping networks of clerical and tribal relationships. These networks vary in systematic ways, with systematic implications for patterns of cohesion and control. Groups built upon networks that pull together leaders across space with local communities will tend to be the most cohesive. Other types of networks also lead to distinctive patterns of organizational formation. This does not mean that other variables are irrelevant, but it does mean that we need to first look at the underlying social structure of insurgent groups and movements. The effects of resource flows, ideologies, and state counterinsurgent policies are refracted through social networks and organizational structures.

The project develops and tests the argument through two types of comparisons. The first was three sets of within-conflict comparisons  in Kashmir, Northern Ireland, and Sri Lanka - that examine multiple armed groups within the same war. Despite shared structural contexts these groups vary in their organization and cohesion. I use this mix of similarity and difference to explore why this variation occurrs. I find that underlying social bases and particular forms of resource flows (especially external support) play an important role in determining organizational cohesion. I then explore the validity of my argument across three sets of Southeast Asian insurgencies to see whether the theory traveled. I compared communist insurgencies in Malaya, French Indochina, and the Philippines after World War II, separatist movements in Aceh and East Timor in Suhartos New Order Indonesia, and two separatist armed groups in the southern Philippines since the 1970s.

The project also has clear policy implications. Most importantly, it shows that insurgency is not a popularity contest for hearts and minds; the demands of violent collective action and organization are not the same as those of appealing to the mass populace. The research also reveals that resource flows have no single effect on armed groups  fears that access to resource wealth necessarily leads to predatory narco-insurgents or loot- seeking revels do not find support in the comparative research. Instead, insurgent organizations represent a complex melding of social and material factors.