Elections are a core democratic procedure intended to peacefully transfer power between political parties. However, in many parts of the world, elections are associated with the strategic use of violence to affect results and future power dynamics. This study tests a current theory on how the state apparatus affects challengers’ and incumbents’ perpetration of electoral violence. It hypothesizes that incumbents are the main perpetrator of electoral violence because they have a comparative advantage through the coercive state apparatus. This makes resorting to election violence more rational for incumbents. The hypothesis is empirically tested by conducting a logistic regression on global data from the DECO and QoG datasets. The results are inconclusive due to statistical insignificance and low confidence in the accuracy of the expected relationship. Moreover, the inconclusive result is assumably affected by scarce and incompatible data. While the result is inconclusive, the paper contributes to research by providing descriptive statistics on the main perpetrator of electoral violence. Furthermore, the research design can also be a reference for future research on how state capacity affects electoral violence.