Abstract: Is the risk of civil conflict related to the quality of government? This dissertation contributes to the quantitative research on this topic. First, it provides a more nuanced account of the role of the government in influencing the risk of civil conflict. In doing so, the dissertation bridges a gap between the quantitative literature, which primarily focuses on types of regimes, and the qualitative literature, which emphasizes variations in how political authority is exercised within these institutions. Second, the dissertation introduces novel measures of the quality of government, and tests their association with civil peace across countries, over time. The dissertation consists of an introductory chapter and four separate essays. Essay I examines the risk of conflict across different types of authoritarian regimes. The statistical results suggest that single-party regimes have a lower risk of civil conflict than military and multi-party authoritarian regimes. The finding is attributed to the high capacity for coercion and co-optation within single-party institutions. Essay II studies whether cross-national variations in the occurrence of civil conflict are due to differences in the quality of government. The essay finds that governments that are not able to carry through such basic governing tasks as protecting property rights and providing public goods, render themselves vulnerable to civil conflict. The focus of Essay III is on patronage politics, meaning that rulers rely on the distribution of private goods to retain the support necessary to stay in power. The statistical results suggest that patronage politics per se increase the risk of conflict. The conflict-inducing effect is mediated by large oil-wealth, however, because the government can use the wealth strategically to buy off opposition. Essay IV argues that patronage politics can also lead to violent conflict between groups. The results from a statistical analysis, based on unique sub-national data on inter-group conflict in Nigeria, are consistent with this argument. Taken together, the findings of this dissertation suggest that both the form and degree of government have a significant influence on the risk of civil conflict.