1
Regime Transition and Communal Violence
Joakim Kreutz and Kristine Eck
Uppsala University
Paper prepared for the 52
nd
International Studies Association conference; Montreal, 16-19 March 2011.
Violence between different communities erupts only occasionally; it is more prevalent in countries like
India, Ethiopia, and Nigeria than in Mozambique, the Philippines, or Yemen. What accounts for these
differences? To date, no existing study has investigated why and when communal violence occurs using
systematic cross-national data; we seek to address this glaring research gap. We argue that the role of the
state plays a major role in constraining or facilitating the expression of violence between different
communities. In particular, we view major regime transitions as a window of opportunity for communal
groups to mobilize and exploit the decreased deterrent capacity of the state; in such periods, the state is
less able to contain the problems of communal opportunism that are otherwise restricted by the threat of
punishment from formal institutions. Similarly, we suggest that informal in-group policing is also
dependent on regime stability and will cease to function effectively during major regime transitions. We
then nuance this argument to suggest that the type of regime in the previous period will affect the
incentive structure for communal violence, with unconstrained and exclusionary regimes particularly likely
to experience communal violence during regime transitions. We test these arguments using new global
data on communal violence 1989-2007 and find that our arguments are supported: communal violence is
more likely in the aftermath of major regime transitions; this is particularly so if the regime in the previous
period was an autocracy, was led by an unconstrained executive, or was dominated by a single ethnic
group.
Page 1 of 19 
Convention