| Summary/Abstract |
Despite the prevalence of nonviolent uprisings in recent history, no existing scholarship has produced a generalized explanation of when and where such uprisings are most likely to occur. Our primary aim in this article is to evaluate whether different available models—namely, grievance approaches, modernization theory, resource mobilization theory, and political opportunity approaches—are useful in explaining the onset of major nonviolent uprisings. We assemble a reduced list of correlates based on each model and use each model’s out-of-sample area under the curve and logarithmic score to test each theory’s explanatory power. We find that the political opportunity model performs best for both in- and out-of-sample cases, though grievance and resource mobilization approaches also provide some explanatory power. We use a culled model of the predicted probabilities of the strongest-performing variables from all models to forecast major nonviolent uprisings in 2011 and 2012. In this out-of-sample test, all models produce mixed results, suggesting greater emphasis on agency over structure in explaining these episodes.
|