Does Health Aid Improve Child Health? Micro Evidence from Uganda

Abstract: In 2017, the Organisation for Economic Co-operation and Development (OECD) estimated that more than $10 billion worth of foreign aid commitments to developing countries were directed at the health sector. A large share of this health aid targets Sub-Saharan African countries that continue to display some of the lowest levels of development worldwide. Yet, the evidence on the effectiveness of health aid in improving health outcomes in developing countries is inconclusive. In this paper, I investigate the effect of development assistance to health on infant mortality, as well as immunisation rates, in Uganda, thus further exploring whether immunisation programs are an important channel through which foreign aid can improve child health. Geocoded aid project data on the district level from the AidData initiative is combined with individual-level data from the Demographic and Health Surveys (DHS) to construct a micro panel where identification is established using within-mother variation. The sub-national approach is informative as small-scaled development interventions are taken into account – a dimension which has been neglected in the cross-country literature examining aggregate aid measures. In addition, the use of mother fixed effects accounts for time(mother)-invariant unobserved variables and minimises any bias arising from confounding factors affecting child health outcomes. The results suggest that health aid is ineffective in improving child health. The impact is, however, heterogenous across sub-groups of the sample and seems to depend on both project frequency and the origin of aid flows.