Abstract:
We reassess Rossi and Villar (2020), who study whether large subsidies for malaria prevention in Senegal increased parental investments in child health. Using the authors' curated analysis files, we first establish computational reproducibility of the published tables and figures. We then conduct a focused and robustness replication of the main difference-in-differences result on child health expenditures and health-seeking behavior. We corroborate the original findings that child health expenditures and health-seeking behavior in more malarious areas "catch up" after the subsidy, primarily via the extensive margin. The replication scope was constrained by the absence of raw data and cleaning scripts, as well as sparse documentation of variable construction in the primary dataset.
