Lines Of Conflict: Ethnicity And Deadly Elections In Africa : A quantitative study on how ethnic composition affects severity of election violence in Africa.

This thesis examines how ethnic diversity affects the severity of electoral violence in Africa, measured as the number of election-related fatalities. While much existing research focuses on the onset of electoral violence, the deadliness of such violence remains underexplored. Drawing on literature on ethnic diversity and polarization, the study initially hypothesizes that lower ethnic diversity would increase electoral stakes and lead to more deadly violence. Using a quantitative research design, the analysis covers 1,955 country–election–year observations from African national elections between 1989 and 2017. Election-related fatalities are measured using the Deadly Electoral Conflict (DECO) dataset, while ethnic diversity is operationalized as the number of politically relevant ethnic groups from the Ethnic Power Relations (EPR) dataset. Ordinary Least Squares regression is applied to a log-transformed fatality measure, with controls for regime type and GDP per capita. The results contradict the theoretical expectation. Higher ethnic diversity is significantly associated with increased electoral fatalities, while GDP per capita is negatively related to deadly electoral violence and regime type shows no consistent effect. These findings suggest that ethnic diversity does not necessarily mitigate electoral violence and highlight the importance of considering how ethnic structure interacts with economic and institutional factors when assessing the severity of electoral violence.