Dams and Infant Mortality in Africa∗

This paper investigates the impact of dams on infant mortality using 32 waves of DHS data that have GPS locations of households. This allows us to estimate the impact of dams on households that reside upstream, downstream, and within its immediate vicinity. We use a sample of over 400,000 children in 17 countries in Africa. In contrast to earlier research on the impact of dams on agricultural productivity and poverty at the district level in India (Duflo and Pande, 2007), we examine child-level outcomes, measure the impacts of dams on households that are both close and very far to the dam, and exploit variation in floodplain and non-floodplain regions that is more important for agricultural production in Sub-Saharan Africa than India. For non-migrant households we find the following. First, children born in households that reside immediately downstream to a dam experience a significant reduction of 6.19-6.96% in infant mortality. This is because the benefits of irrigation services of the dam are large for downstream households geographically close to the dam. Second, for children born in households that reside further downstream, infant mortality significantly increases by 2.18-2.36%. This is because dams reduce water levels downriver, and households cannot access compensating irrigation services from dams, or benefit from the reduced volatility of water flow that dams provide. The infant mortality increase rises sharply to 7.57% for children born farther downstream in floodplain areas, as the reduced water level causes degradation of the wetland ecosystem which is crucial to household livelihoods in Africa. Children born in the vicinity of the dam experience increased infant mortality of at least 2.27%, due to increased malaria incidence and reduced agricultural productivity near the dam reservoir. This effect increases monotonically to more than 10% with the number of dams in children’s vicinity. JEL classification: I18, J13, O22, O55