The project first required reaching a state-of-art knowledge of the conflict literature and investigating the qualitative literature dedicated to case studies. Then, we built a theoretical model in which two groups contest a resource in the country of origin, while one of them is related to an exongeous diaspora abroad. We first find that above a certain size, the diaspora financially contributes to the war effort of its group, because migrants, who are interested in the conflict’s outcome, do not directly bear its costs. Second, we assume that, given the expected conflict outcome, the two groups of residents can negotiate peace if this makes them both better off than war. Although a larger diaspora translates into a larger contribution to the war effort, this does not systematically means more chances of war. The induced increased relative strike force of the group of origin can deter its rival group to go to war, and make negotiation more likely. A diaspora can thus play either a peace-building or a peace-wrecking role. We finally identify country- and diaspora-level characteristics which determine whether a diaspora is more likely to act as a peace-building or –wrecking force.
We tested the robustness of the results to different assumptions, and developed two extensions of the model to account for (i) the endogeneity of migration and (ii) the possibility of two co-existing diasporas. Finally, we use this theoretical framework to go back to the case studies and enlighten why some diasporas had a peace-wrecking role, while others did not manage to trigger war, and others pushed favored peace. To date, this work led to the publication of discussion papers in three different series and to numerous participations in international conferences.
The second axis of the project empirically analyzes the role of refugees on conflict at home. We adopted a group-level approach, employing recent data on refugees (Vogt et al., 2015) and on violent events (Sundberg and Melander, 2013). We merged these datasets by associating violent events involving ethnicity-based insurgent groups with the outflows of refugees of the same ethnic background. This allows us to investigate the relationship between refugee flows by ethnic group and the intensity of violence involving their ethnic group in the homeland, controlling for a myriad of potential confounders. The estimated correlation turns out to be robustly negative, pointing out the peace-building potential of refugees.
Consistent with the theory, we find a heterogeneous correlation between violence and refugee flows, which are more likely to be favor a de-escalation of violence in poor, small countries, with an important recent history of violence. We also explore the sensitivity of our results to alternative samples of study, notably accounting for insurgent groups with multiple ethnic attachments. This research has yielded a working paper which will soon be diffused through discussion paper series and conference presentations.
