Democracy, Democratization, and Civil War

Abstract

An influential body of scholarship has associated both democracy and democratization with civil war. Important findings include the so-called inverted U-shaped relationship between democracy-levels and civil war onset and that propensity for democratic openings to spark internal violence. However, most of these findings have been challenged, particularly by scholars pointing to problems with the aggregate nature of the analyses and the data sources used. Against this background, we enlist new, fine-grained data from the Varieties of Democracy (V-Dem) Project. We discuss how the new data can be used to disaggregate regime variables in order to better understand the causal dynamics that link the regime form and regime change to civil war onset, if any. Guided by these considerations, we use the new data to reassess the ‘inverted U-curve’. Our analysis shows that this relationship is driven by ‘liberal’ aspects of democracy such as freedom of assembly and freedom of speech rather than by the ‘electoral core’ of democracy. The relationship between clean elections and civil war onset is approximately linearly decreasing, and at the indicator level of the clean elections attribute we find several different patterns.