Protectors or security threats: how political orders shape citizens’ fear of militias

Abstract What explains citizens’ perceptions of counterinsurgency militias? Previous research has focused on differences in “objective” measures of militia performance across militia types, conflicts, and country contexts. However, affected citizens may interpret similar militia activities in contrasting ways—either as a menace to communal security or as a bulwark against external threats. We investigate how citizens’ fear of militias can change due to dynamics of patronage networks in which militias operate. We argue that stable networks improve citizens’ ability to make sense of the (violent) behavior of militias—in terms of general rationales, specific motives, and target selection. The resulting perceptions of predictability and agency can make militia behavior seem more tolerable (and avoidable) to citizens. We investigate this argument in north-east Afghanistan. We draw on four waves of standardized household surveys and qualitative interviews gathered between 2010 and 2018. Our research design exploits variation in militias’ patronage networks over political geographies and time. Our results corroborate our hypothesis: analogous militia behaviors can evoke disparate perceptions of militias conditional upon the context of patronage. This finding contributes to our understanding of how political contexts shape citizens’ attitudes towards violent organizations in contexts of conflict.