The onset of civil armed conflicts and personalization of power in dictatorships

This article examines the relationship between civil conflict onsets and personalism in dictatorships. It argues that civil conflict onsets are temporary shocks that alter elites’ incentives and capacities, allowing dictators to personalize. The new wartime context reduces their capacity for coordination and their willingness to stage a coup, opening a window of opportunity for dictators to pursue strategies that would be too noticeable to elites in peacetime. Using three counterfactual estimators for panel data (fixed effects counterfactuals, interactive fixed effects counterfactuals, and matrix completion) on a global sample of autocratic regimes (1946–2010) and a latent index of personalism, I find that the onset of civil conflict is followed by higher levels of personalism for the treated leaders, who are relatively long-surviving autocrats in the sample. Sample restriction, standard diagnostic tests, and a sensitivity analysis provide mixed rather than consistently supportive evidence, warranting a cautious interpretation of the causal effects.