UC Berkeley

Communal violence remains an enduring experience across many multi-ethnic societies. In these diverse communities, small-scale clashes can quickly escalate into severe cycles of violence. An interesting pattern emerges across these communities: tensions are nearly ubiquitous---yet, the outbreak of violence is not. What accounts for these pockets of peace within conflict zones? I advance a theory of violence that hinges on the role of local leaders in dispute resolution, which in turn will influence violent outbreaks. Specifically, I argue that the efficacy of local leaders---and citizen responses---in dispute resolution is shaped by the strength of ties leaders possess both within and across communities. Within-group ties (bonding capital) allows leaders the ability to induce compliance and enforce agreements, while out-group ties (bridging capital) is necessary for credibility and cooperation across group lines. In 88 communities across three states in Nigeria, I conducted a randomized control trial and surveys with local leaders and citizens to test my core theoretical predictions. First, I find that leaders with cross-group ties perceive out-group leaders with more credibility, report more successful dispute resolution and, ultimately, less violent events. I also demonstrate that cross-group ties between leaders affect citizen compliance, satisfaction with arbitration, citizen perceptions of insecurity and reported violence. Additionally, I find tentative evidence that traditional chiefs are more likely to possess extreme forms of in-group ties that may be detrimental to dispute arbitration. On the other hand, alternative leaders---specifically, youth and civil society leaders---often have stronger connections with the out-group, which I posit leads to more effective and credible inter-group mediation.