Political Order and The Organization of Election Violence in Nigeria
While more countries than ever hold regular multi-party elections, about a quarter globally are marred by violence. Beyond the loss of human life, election violence reduces turnout, undermines accountable and representative government, hurts local economies, and can be a precursor for civil war. Over the past decade, research on election violence has rapidly expanded to shed light on national-level risk factors, including economic malaise, majoritarian electoral systems, weak political parties, competitive national elections, limited institutional constraints on executives, exclusionary identity politics, and the presence of international election observers. These findings have advanced our understanding of important cross-sectional differences between countries with peaceful elections and those with elections that turn violent. Yet election violence, like other forms of political violence, varies significantly within countries; it is not spread evenly across a country, but is concentrated and reoccurs in specific localities. Because existing research overwhelmingly focuses on national-level variables that are constant across subnational units, it cannot tell us how and why election violence is organized and varies within countries.
My book project tackles the puzzle of within-country variation in election violence by studying the conditions under which subnational elites and non-state actors jointly organize and perpetrate pre-election violence. Evidence from Africa, Latin America, South Asia, and the United States shows that political elites rarely work alone in organizing violence (election-related or otherwise) but instead, regularly turn to a wide array of non-state actors, including paramilitaries, gangs, drug-trafficking organizations, ethnic militias, and even ordinary voters. Building off of these observations, I ask, under what conditions do subnational elites recruit non-state groups for pre-election violence? What are the different types of non-state actors that elites outsource pre-election violence to, and when and why do such groups agree or refuse to carry out election violence?
The book uses most-similar and most-different cases of gubernatorial elections from 1999 to 2015 in federal Nigeria to answer the above questions. I compare elections across time in each state to create most-similar cases that control for alternative explanations, specifically whether local ruling party elites and non-state groups share an indigenous identity, whether the political goals of non-state groups reject the Nigerian Constitution’s commitment to secular, multi-ethnic democracy, and whether local ruling party elites belong to the national ruling party. The comparisons of gubernatorial elections across states serve as most-different cases, illustrating the power of the argument in starkly different contexts. I study the Ijaw Youth Council and the Niger Delta Vigilantes in Rivers State, the National Union of Road Transport Workers in Oyo State, the Bakassi Boys in Anambra State, the O’odua People’s Congress in Lagos State, and Boko Haram (pre-insurgency, 2002-2009) in Borno State.
In addition to advancing scholarship on election violence, the book project draws attention to intra-party elite competition for understanding election violence whereas the existing literature has largely focused on inter-party competition. The theory and evidence also speak to broader questions about the relationship between political order and violence. The politics of personal rule creates both the supply (the non-state groups studied in the book) and elite demand for election violence. Personal rule generates mass frustration with corruption and state violence, laying fertile ground for different types of non-state groups that serve as agents of local political order in the absence of strong public goods provision. These groups, however, also provide the infrastructure for pre-election violence that subnational elites try to tap into when they seek to organize election violence.