We generally think the swing voter is in the best of all worlds: he can choose policy at will, and the policy will fit exactly his desires. What could possibly disrupt this bliss? James Robinson and Ragnar Torvik argue that if the political climate is characterized by some level of repression or violence, the swing voters are going to be the first ones to suffer from it. Indeed, the policy makers in such environments do generally not like to make concessions.
Robinson and Torvik illustrate this with Zimbabwe, which is indeed too easy an example. Would their point also apply in less extreme situations? The critical aspect here is that violence is more effective on swing voters than opposition voters. The study here ignores that violence may radicalize opposition and in particular push swing voters into opposition. The big empirical questions thus becomes: are swing voters intimidated or radicalized?