Sub Saharan African states experience more frequent and enduring civil conflict than countries in any other world region. This dissertation confirms that civil wars are related to local governance patterns, as African governments practice differential policies of repression, underdevelopment, marginalization and exclusion throughout their territories. The spatial and temporal patterns of rebellion within states are therefore shaped by the governance patterns of the targeted regime.
In this dissertation, I advocate a return to the perspective that African regimes govern in a manner that has motivated rebellion across marginalized and excluded groups. Civil wars are shaped by the particularities of targeted government and local characteristics. The dynamics of individual civil wars are primarily designed to take hold of typical military targets, which are also determined by the political, economics, social and physical geography of the state.
The dissertation is designed to address four principal themes put forth to explain the occurrence of civil war. First, state territorial control is presumed to influence the opportunity costs of potential rebels. Second, the coercive potential of a state may affect civil war for two reasons: while low levels of coercive potential allow for the rise of rebel groups, high levels of civilian repression may motivate populations to attack the regime. Third, development levels across a state may engender conflict as the provision of public goods may be based on clientistic relationships with regimes. Finally, in a direct test of neopatrimonal ties between the government and ethnic groups, variation in government representation may provoke conflict as those ethnic groups not allied to regimes may be marginalized from power. Previous studies that have relied on these causal explanations are thoroughly discussed.
The two main conclusions evident from this dissertation are that patterns of governance in African states create uneven zones of control, leading to the marginalization and exclusion of entire groups and regions. In addition, the study of conflict dynamics is greatly advanced by the use of disaggregated data and local level specifications, as it is on this level that the motivations and opportunities of both rebels and governments are most apparent.