An extensive evidence concludes there exists a strong association between waged violence by non-state armed organizations and election results in the midst of civil wars. One usual assumption in such literature is that the level of violence signals the level of electoral influence a non-state armed organization attains. I argument such assumption falls short. The paper examines the effect of paramilitary groups power on election results in Colombia, de fining two categories of power: de facto power (violence), and de jure power (land tenure or territorial control). Using a diff-in-diff strategy, the results show that de facto power is crucial to establish de jure power, i.e. municipalities affected by paramilitary violence report more land expropriations. The paper provides suggestive evidence pointing out how land expropriations, taken as de jure power, influenced elections in order to benefit political parties whose candidates were strongly linked with paramilitary groups. Violence by itself has no effect on election results once I control by de jure power. I argue that before any decision about meddling in elections is made, non-state armed organizations have to achieve a genuine control over territory in order to have the possibility of influencing local politics and, in particular, democratic elections. The strategical use of violence is an important instrument in this regard, but not a definitive one
publication date
May 19, 2021 3:29 PM
Research
keywords
Democracy and civil war in Colombia
Elections in the context of war in Colombia
Influence of paramilitary groups in the electoral process
Political economy
Strategic use of violence for political control
Territorial control and violence as a mechanism of electoral coercion
