In my dissertation, I investigate the negative effects of "good" institutions on political violence. I define "good" institutions as seemingly well-intended institutional changes or decisions. Across three stand-alone articles, I show that “good” institutions do not always reduce political violence because the actors who design and implement institutions tailor them to meet their own interests, which are often not reducing political violence.
In the first paper, I examine how granting regional autonomy changes the level of insurgent violence. Because regional autonomy gives more independence to lower tiers of government, it is commonly thought to reduce insurgent violence. I argue that elite-captured regional autonomy enables elites to expropriate citizen wealth, generating citizen grievances that fuel insurgent violence. To test the implications of my theory, I investigate the effect of the creation of two new, autonomous Indian states—Chhattisgarh and Jharkhand—on insurgent violence. Using difference-in-difference and geographical regression discontinuity designs, I find the creation of these states to be associated with statistically and substantively significant increases in Maoist violence. More importantly, the relationship between these new autonomous states and violence is stronger when elites are better able to exploit local resources.
The second paper in my dissertation examines the effect of investigations by the International Criminal Court (ICC) of alleged human rights crises on non-state civil conflicts. To date, scholars interested in the effect of the ICC on civil conflicts have treated the "state" as a unitary actor. In this paper, I argue that the ICC's investigations of government leaders and of rebel leaders have different effects on civil conflict. Because of its biased case selection process, when the ICC investigates government leaders, it constrains governments, provides incentives for the rebel groups to fight, and leads to more non-state conflicts—that is, conflicts between rebels. When the ICC investigates rebel leaders, it does not constrain rebel groups, strengthens the government's control, and reduces non-state conflicts. I evaluate the implications of the theory with difference-in-difference tests using data on non-state conflicts cross-nationally and find support.
My third dissertation paper looks at freedom of information (FOI) laws and protests. While more transparency in government should theoretically increase citizens' satisfaction and make them less likely to participate in protests, I argue that providing citizens with information about the wrongdoings of lower-tier officials without giving them ways to punish those officials leads to more protests. Using a difference-in-difference design and cross-country data, I find that adopting FOI laws is associated with statistically and substantively significant increases in the number of protests. More importantly, I found the relationship between the adoption of FOI law and the number of protests is conditioned by electoral democracy, a measure of ways to punish government officials. When there are ways to punish those officials, FOI laws do not lead to more protests.In full, the three papers in my dissertation contribute to the literature on institutions and political violence by examining the negative effects of seemingly well-intended institutions on political violence and carefully considering the preferences of the actors who designed or implemented these institutions. The dissertation contributes to the discussion of the effectiveness of institutional decisions on reducing political violence— why some institutional decisions failed to reduce political violence? My work shows that different institutional designs can be subjected to this common rule—institutions are products of actor’s preferences. We should not expect these elite actors to design or implement institutions to fulfill ordinary people’s goals rather than their own.
