Water, sanitation, and child health: Evidence from subnational panel data in 59 countries

Article on water, sanitation, hygiene (WASH), and child health. WASH investments are widely seen as essential for improving health in early childhood. However, the experimental literature on WASH interventions identifies inconsistent impacts on child health outcomes, with relatively robust impacts on diarrhea and other symptoms of infection but weak and varying impacts on child nutrition. In contrast, observational research exploiting cross-sectional variation in water and sanitation access is much more sanguine, finding strong associations with diarrhea prevalence, mortality, and stunting. In practice, both literatures suffer from significant methodological limitations. Experimental WASH evaluations are often subject to poor compliance, rural bias, and short duration of exposure, while cross-sectional observational evidence may be highly vulnerable to omitted variables bias. To overcome some of the limitations of both literatures, a panel of 442 subnational regions in 59 countries with multiple Demographic Health Surveys was constructed. Using this large subnational panel, difference-in-difference regressions were implemented that allowed the examination of whether longer-term changes in water and sanitation at the subnational level predict improvements in child morbidity, mortality, and nutrition.