Do Birds of a Feather Flock Together? Rebel Constituencies and Civil War Alliances

Abstract

What factors drive alliances between rebel organizations? The literature emphasizes military power, local-level rivalries, and external support. We contend that the relative neglect of ethnic and ideological rebel constituencies is unwarranted. Owing to similar agendas and the endorsement of supporters and members, rebel organizations with shared constituencies should tend to form alliances to reap the power-aggregation benefits. However, we argue that this cooperation-inducing effect is at least partially offset by a cooperation-suppressing effect for organizations sharing an ethnic constituency, as they often compete intensely for control of their community. Therefore, we expect common ideological constituencies to have a stronger impact on alliance propensity than shared ethnic constituencies. Moreover, co-ideological rebel organizations should engage in in-depth, structured cooperation (formal alliances), while cooperation among co-ethnic organizations should be transient and superficial (informal alliances). Leveraging a novel data set of all rebel dyads in multiparty civil wars (1946-2015) and using a variety of empirical methods, we find support for our theoretical expectations: rebel organizations sharing a constituency are more likely to ally than those without, but the effect is stronger for co-ideological rebel dyads; further, when co-ethnic rebel organizations form alliances, these tend to be informal, while co-ideological organizations engage in formal alliances.