Rosenzweig’s argument in Voter Backlash and Elite Misperception is simple yet important: Electoral violence does not help politicians win elections. Politicians don’t know this, he argues, and many of them therefore continue to pursue counterproductive violent campaigns. Rosenzweig explains elites’ delusion about the effect of violence on electoral outcomes with the age-old logical fallacy of correlation vs causation: politicians observe that successful ruling parties use violent campaigns, therefore violent campaigns must be effective. Similarly, academics too often assume that since electoral violence continues to exist, it must be an effective and rational strategy. The author combines multiple empirical strategies to prove his point, including survey experiments, observational data, and interviews. While much of this is convincing, I worry about the realism of some of the survey experiments and whether all parts of the argument really travel from the experimental format to real-world elections. The main empirical pattern Rosenzweig highlights is that while violent campaigning could improve support among a certain subset of the population (even this is not clearly supported by his data), backlash from other parts of the electorate produces a net negative effect on the winning chances for politicians. Nevertheless, political elites in Kenya indicate that they believe violence to be beneficial or at least not detrimental to electoral support. While most of the empirical material is from Kenya, Rosenzweig also shows that in countries where the winner of the first election used electoral violence, violent campaigning is more likely to persist, in line with his argument about elite misperception. Identifying effective electoral campaigning strategies is difficult, and politicians, like everyone else, are prone to using mental shortcuts.