Fight the Power

The Effect of Post-Election Anti-Government Action upon the Probability of Incumbent Replacement

Abstract

This article investigates whether post-election anti-government actions increase the probability that incumbents will be removed from office. The theoretical argument is that the anti-government groups can choose to use violent or non-violent methods to force the government to make concessions, such as replacement of the incumbent. Both violence and non-violence are argued to have a positive effect on the probability of this to happen, with the latter likely to be stronger. Thus, the first hypothesis argues that anti-government violent action increases the incumbent’s probability of being replaced. This is complemented by a second hypothesis, stating that anti-government non-violent action increases the incumbent’s probability of being replaced more than anti-government violent action does. Applying a logistic regression on 550 national elections in states with a democratization process or hybrid regime, compiled from the National Elections across Democracy and Autocracy (NELDA) and Electoral Contention And Violence (ECAV) datasets, the first hypothesis is not supported due to lack of statistical significance across the regression models. The second hypothesis is supported, as the independent variable non-violent action receives statistically significant results, but these do not hold in the robustness test. Considering the suggestions from previous literature, the results indicate that non-violence is the more successful option of action for anti-government actors, but it cannot be concluded with certainty.