Civil Resistance in the Streetlight: Replicating and Assessing Evidence on Nonviolent Effectiveness

Abstract

Does civil resistance work? Research emphasizes the effectiveness of nonviolent resistance over violent resistance in achieving campaign goals, with the seminal study Why Civil Resistance Works (WCRW) by Chenoweth and Stephan being the main point of reference to date. I revisit this pivotal finding in three steps. First, I reproduce WCRW’s results on nonviolent effectiveness. Second, I discuss how cases may have been overlooked due to a streetlight effect. Third, I quantify the results’ sensitivity using simulations. I find that WCRW’s main findings on nonviolent effectiveness are highly sensitive to variable selection, under-coverage bias, bootstrapping, and omitted variable bias. As a routine reference in scholarship and the public discourse, assessing the robustness of WCRW’s findings is relevant to practitioners and researchers.