Civil Dissent and Repression: An Agency-Centric Perspective

Abstract

Do governments make a strategic choice in deciding what type of security agent to use for repression? Research acknowledges the role of auxiliary groups such as militias in repression, yet surprisingly little attention is given to the state's formal domestic security agents, such as the police. We show that formal security organizations and auxiliary groups enhance the government's ability to repress by acting as strategic complements. As the better-regulated force, formal agents are often employed against violent riots, when regimes worry more about the ability to control the agents and their behavior more than about being visibly linked to the violence. In contrast, auxiliaries are often used to repress nonviolent campaigns, when the government seeks to benefit from agency loss in order not to be associated with the violence, which can be costly in these contexts. We empirically verify these linkages on country-month data for Africa using panel vector-autoregression (pVAR), which accounts for endogeneity, not only between the dependent and independent variables, but also the dependent variables. We complement these statistical results with case-based evidence and descriptive original data from non-African countries.