Disaggregating Defection: Dissent Campaign Strategies and Security Force Disloyalty

Abstract

Extant research posits that nonviolent campaigns increase security force defection during dissent, thereby improving their success rates over violent campaigns. Yet empirical evidence remains limited and mixed. We argue that this is partly due to overly aggregated measures of disloyalty that mask important variation. To address this, we disaggregate security force disloyalty both theoretically and empirically. We contend that nonviolent and violent strategies impose distinct costs on security forces—related to repression, fighting, and exit—which shape the likelihood of different forms of disloyalty. Our findings reveal that nonviolent campaigns are more likely to trigger large-scale loyalty shifts, but their effects are confined to specific types of disloyalty. Specifically, nonviolent strategies increase the likelihood of shirking, vocal support for opposition campaigns or criticism of the regime, and withdrawal of regime support. By contrast, desertion and side-switching are more likely during violent campaigns.