| A common feature of modern civil wars is the tendency of rebel organizations to factionalize and fragment. Yet how can we explain patterns of factionalism and fragmentation in civil wars? How do normal, private disagreements within rebel organizations become full-fledged disputes that lead to a complete collapse of cooperation in war? Given the importance of rebel factionalism and fragmentation to the scale, duration, and outcomes of civil wars, such questions warrant careful consideration.;This dissertation uses case studies of seven Ethiopian rebel organizations, nested within a statistical analysis of an original dataset of Ethiopia's civil wars, to develop an argument about the dynamics of factionalism and fragmentation within rebel organizations. The argument is then tested more broadly, using statistical analysis of data on militant organizations in the Middle East.;The project relies on over two years of field work in Ethiopia, Somaliland, and African refugee communities in Europe and North America, in which the author conducted over 80 interviews and extensive archival research. The evidence suggests that rebel factionalism and fragmentation is caused by territorial gains and losses, as battlefield shifts alter the incentives that rebel elites have to cooperate with one another. Counter-intuitively, territorial stalemate tends to promote cohesion, and is the only sustainable basis for organizational stability in war.;The core argument, along with several corollaries, contributes to a small but growing literature on the internal politics of rebel organizations, and offers lessons for policy makers who seek to influence the dynamics of today's civil wars. |
