Title: Guns, Politics, or Bankruptcy: Disentangling the Determinants of Armed Organizations' Post-war Trajectories
Name: Sarah Zukerman Daly
sd2623@columbia.edu
Year: 2009
Type: Dissertation Fellowship
Summary: This research analyzed what happens to armed organizations after they sign peace accords and disarm. Why do they disintegrate, remilitarize, or transform into non-violent, socio- political entities (political parties or civic associations)? A better understanding of the mechanisms that generate distinct post-war outcomes is a prerequisite for productive theory-building on post-conflict security, state-building, consolidated peace, and transitions to democracy. I argue that post-war outcomes derive from armed groups' recruitment, deployment, and post-war migration patterns.

Building on 15 months of fieldwork in Colombia, I found that two factors explain divergence in organizations' post-war trajectories. First, where armed groups recruit and deploy their fighters. This factor determines the combatants' network structures and post-war migration and thus, whether the organization endures or disintegrates. It also determines the group's embeddedness in the civilian population and its ability to maintain political influence without arms. I conclude that rebel and paramilitary groups, which drew a majority of their recruits from the communities in which they fought, are comprised of combatants who are strongly linked to each other, remain in the zone of deployment post-war, and are embedded in the communities. These local groups thus face a high probability of persisting and forming legitimate political entities after disarming. In contrast, militarized units, which recruit in a dispersed fashion and deploy their soldiers away from their towns of origin, lack dense social ties and post-war physical clustering of their combatants. They are thus prone to disintegration. The second variable captures the nature of the post-war distribution of power. By bankrupting some organizations and preserving others, demobilization has differential effects on armed group collective capacity. Where it weakens a group, it destabilizes the groups territorial bargains with the state and with other armed actors. It also creates proprietary information and uncertainty about the extent to which it has done so. As a result, resumed war becomes likely. If instead, the distribution of power within the system is maintained, the armed group will, over time, fully demilitarize and be brought into the state's legal framework.

To test the proposed causal logic, I engaged in a multi-method approach and relied on rich micro-level data on the Colombian conflict. Exploiting Colombia's unparalleled comparative laboratory for this research, I tested the effects of recruitment, deployment, and post-war migration patterns on organizational outcomes using three, multi-method strategies. First, I conducted a detailed, controlled comparison of armed groups in three regions of Colombia. I employed within-case congruence tests and careful process tracing based on interviews of over 300 ex-combatants, civilians, victims, and experts on the armed conflict. The second strategy combined these qualitative sources with quantitative ones to evaluate the proposed hypotheses on the entire universe of community-armed group dyads in Colombia. For this analysis, I relied on municipal-level violent event data, a database of seven years of news articles, and statistical evidence from a series of surveys. This series included surveys of 31,472 former paramilitaries, 120 imprisoned ex-combatants that remilitarized, 5,000 civilians, and 226 psychologists that treated the former fighter population. Third, I exploited several case studies from Asia and Africa to evaluate the theory's external validity. The empirics provided strong support for the proposed model.

The findings dispute the explanatory power of credible commitment problems in accounting for resumed war and demonstrate that there are shocks to the system that cannot be predicted or prevented. The analysis further undermines resource determinism. The political economy theories predict that resource-rich groups will remilitarize, fragment, and lose their socio-political influence. These expectations are not borne out in the data. Finally, my research challenges the conventional wisdoms of the micro peace literature that equates armed group bankruptcy with individual reintegration success and treats ex- combatants as atomized actors rather than as embedded in dense networks and institutions that structure their decisions. The questions posed in this dissertation, the new empirical data it offers, and its unexpected findings are significant for theory development and have important policy implications for war to peace transitions