Responses to Bensch, Rose, Neubauer, Ankel-Peters, and Brodeur

Bensch et al. (2025) successfully reproduce all results of our article "Partisan effects of information campaigns in competitive authoritarian elections: Evidence from Bangladesh" (Ahmed et al., 2024), but they raise some issues "that warrant further clarification." 1. They document that the village classification and the sampling strategy described in the published article are different and more refined than those described in the pre-analysis plan. This is correct and not uncommon in large field experiments. The relevant information has long been in the public domain (including during the refereeing process), and there is no reason to believe that these refinements could bias our results. 2. They conduct a novel empirical analysis, which is very different from our analyses, and see a pattern they find surprising. We highlight that their analysis is conceptually problematic given our sampling strategy and the level of randomization. We also explain why the resulting pattern is unsurprising. 3. They provide credible evidence that the map displaying the villages in our sample contains mistakes. We apologize for these mistakes but highlight that these mapping inaccuracies do not affect our results or our interpretation thereof. We conclude that none of the issues raised by Bensch et al. (2025) undermines the validity of the experiment, the results, or the interpretation presented in Ahmed et al. (2024).