The geography of regime support and political violence

Abstract

ABSTRACT In this article, we investigate how the political geography of local power and national-subnational alignment influences the distribution of conflict between election contests. We ask how consistently local intermediaries, such as members of parliament (MPs), can sustain their own support and that of their successful national candidate. The best measure for the ability of local elites to generate and keep support is candidate vote margins. These election results recast elites and areas as core, swing or costly regime investments, which in turn influences the streams of patronage and authority between election cycles. Based on the political geographies that result from subnational election data for 13 African countries, we find that in core areas where support for the leader is high, there is limited violence by state forces; state violence is significantly higher in areas costly to the national leader; and violence by non-state armed actors (e.g. militias) is most likely to occur in swing areas where the winning margin of the last legislative election is narrow for the local candidate. It is in these swing regions that competing political elites engage in violence to replace poorly performing local intermediaries.