Donors of development assistance for health (DAH) typically provide funding for a range of health issues, including HIV/AIDS; tuberculosis; malaria; maternal, neonatal, and child health; non-communicable diseases; and health sector support. What explains the distribution of health aid across these issues and changes over time? We begin by noting that allocations among health issues does not reflect the relative severity of their impact on population health, and specifically to the loss of Disability-Adjusted Life-Years (DALYs) that is attributable to them. This weak link between impact severity and funding suggests that decisions on health aid allocation are not based primarily on rational health-maximizing calculations. To explain this puzzling fact, we focus on the interdependence among donors in heath aid allocation. Specifically, we argue that health aid allocation decisions of donors are influenced by the allocation portfolio of other donors. This interdependence among donors can lead two opposite effects. On the one hand, influence that produces divergence among health aid portfolios could be interpreted as evidence of rational specialization. On the other hand, influence that produces convergence among portfolios would suggest that social processes such as emulation and mimicking pervade the domain of health aid. We will assess these competing diffusion effects by applying spatial regression models to health aid disbursed globally between 1990 and 2014.
