HRMARS - Despite decades of global investment in malaria control, under-5 mortality in Sub-Saharan Africa (SSA) remains unacceptably high, and the governance conditions under which malaria programmes translate into child survival gains are poorly understood. This study investigates the direct relationship between malaria incidence, institutional quality, and under-5 mortality using a panel of 35 SSA countries from 2000 to 2023 and examines whether institutional quality modifies the malaria-mortality relationship. Applying static Fixed Effects estimation with Driscoll-Kraay standard errors, confirmed through a five-stage diagnostic sequence encompassing cross-sectional dependence testing, second-generation panel unit root tests, Hausman specification testing, dynamic persistence testing, and control-function endogeneity correction, we find that malaria incidence significantly and robustly increases under-5 mortality across all specifications. At mean institutional quality, a one standard deviation increase in malaria incidence is associated with 5.7 additional child deaths per 1,000 live births. Marginal effects analysis indicates a 35% differential in the malaria-mortality relationship between weak and strong governance settings, though the interaction term does not reach conventional statistical significance. Among individual governance dimensions, Government Effectiveness and Regulatory Quality approach significance as moderators, pointing to state operational capacity rather than accountability or stability as the proximate institutional channel. A novel methodological contribution concerns data quality: three transcription errors in a single fragile state account for approximately 23 percentage points of unexplained model variation, illustrating that data quality failures in conflict-affected settings disproportionately influence panel estimates for the entire region. The findings support treating malaria control and institutional capacity-building as complementary rather than competing priorities in SSA health policy.
Olohunlana et al. (Sun,) studied this question.