Zusammenfassung:
Depetris-Chauvin et al. (2020) analyze the effects of football on civil conflict in Africa. We engage into a replication of their analysis. This unveils major inconsistencies between the sample selection criteria described in the article and the composition of the estimation sample (41 errors for 109 country × qualification campaign pairs). Once these inconsistencies are removed, the evidence on the effect of a close qualification to the African Cup of Nations on conflict becomes substantially weaker. Several specifications produce a non-significant coefficient for the treatment. The placebo test proposed by the Authors themselves casts doubts on the validity of the identifying assumption in most specifications with a significant effect of the treatment.