Indirect rule and mass threat: Two paths to direct rule

What is the impact of popular discontent on the transition from indirect to direct rule? The current literature suggests contrasting theories, variously arguing that rulers are more likely to govern directly when facing either a particularly high or particularly low probability of mass resistance. I reconcile these views by arguing that the decision to rule indirectly is subject to competing dilemmas. In a formal model, I show that these twin tensions influence the choice to centralize power in opposing manners. Accordingly, there are two distinct political logics driving direct rule: one resulting from a high likelihood of revolt and the other from a low likelihood, with contrasting comparative statics. The model therefore reconciles contrasting views in the literature. I illustrate the model's logic with reference to key cases and provide heuristics for predicting comparative statics in new empirical settings.