Divided Loyalty: Are Conscript Militaries Less Likely to Repress Protesters?

• In Event: Paramilitaries, Praetorians, Police: Designing & Deploying State Security Forces

Abstract

Are conscript militaries less likely to crack down violently on protests than volunteer militaries, or compared to other security forces like police? Conscript armies are generally thought to be more representative of the population and therefore to identify with protesters as opposed to the regime. While intuitive and appealing, this argument has rarely been tested. We use event-level data from the Integrated Crisis Early Warning System (ICEWS) dataset to identify more than 40,000 protests in 174 countries between 1995 and 2013, which we use to provide the first large-n test of the common assumption that conscript armies are less likely to use violence against protesters. In addition, we account for the reality that conscription does not always produce representative armies. Our findings indicate that conscription is only associated with a decreased use of repression as the size of the military relative to the population increases. In supplemental analyses, we also examine the regime’s strategic choice of which force to deploy, military or police, in response to an antigovernment protest.

Authors

• Paul Lorenzo Johnson, US Air Force

• Max Zeldes Margulies, United States Military Academy