Why do some state-led mass killings end quickly while others endure for over a decade? And why do some states murder millions of constituents during the course of mass killings, whereas other states seem to “retreat from the brink” after killing thousands (Straus 2012)? A large body of work has focused on the important role played by civil society and non-governmental actors in initiating different forms of rescue, evasion, and assistance in the midst of different cases of mass killings, as well as the political pressure they have applied in bringing about the ends of civil conflicts. Despite many inspiring and hopeful cases of collective action under systems of intense repression, other research finds civil society can play a much more malevolent force in the context of mass killings. In this paper, we test some basic mechanisms that emerge from the literature on more general relationships between civil society and mass killings. We find that, in general, a relatively participatory and autonomous civil society is correlated with shorter mass killings. However, we also find that active civil societies are associated with higher rates of lethality, particularly when those civil society sectors are active in highly unequal polities. Because most mass killings are relatively short, our findings suggest that civil societies in states with uneven access to power are more commonly correlated with shorter, deadlier spells of government violence. This conclusion seemingly supports the view of civil society skeptics, at least in contexts where mass killings have already begun.
Chenoweth, Erica, and Evan Perkoski. "A Source of Escalation or a Source of Restraint? An Empirical Investigation of How Civil Society Affects Mass Killings." HKS Faculty Research Working Paper Series RWP19-027, September 2019.