Counting votes or counting bodies? : A qualitative study on the effect Regime Type has on the nature of Pre-election Violence in autocratic states

Abstract: In a quantitative study in 2007, Davenport found that autocratic military regimes statistically face a higher risk of electoral violence than authoritarian party-backed regimes. This thesis has attempted to link Davenports findings with theories on military belligerence presented by Lai and Slater (2006) as well as Geddes et al. (2014), and thereby contribute to our understanding of the matter by investigating the potential causal mechanisms connecting regime type and electoral violence. The analysis specifically focuses on differences in pre-election violence by comparing the 2008 election in Pakistan and the 2007 election in Uzbekistan. Evidence from the cases suggest that there is some support for a covariation between regime type and levels of pre-election violence, although there are alternative explanations worth considering before one can determine whether or not a causal relationship can be observed. Finally, the findings indicate that military regime belligerence or lack of knowledge on how to use nonviolent political repressive tools in order to sway the elections do not explain the observed variation. Rather, the thesis suggests that levels of pre-election violence is more likely to be affected by other conflicts in the region, the design of the election campaign and whether there are established influential opposition parties present in the country.