Working Paper: NBER ID: w13396
Authors: Till Brnighausen; David E. Bloom
Abstract: Without large increases in the number of health workers to treat HIV/AIDS (HAHW), most developing countries will be unable to achieve universal coverage with antiretroviral treatment (ART), leading to large numbers of potentially avoidable deaths among people living with HIV/AIDS. We use Markov Monte Carlo microsimulation to estimate the expected net present value (eNPV) of a scholarship for health care education that is conditional on the recipient entering into a contract to work for a number of years after graduation delivering ART in sub-Saharan Africa. Such a scholarship could increase the number of health workers educated in the region and decrease the probability of HAHW emigration. "Conditional scholarships" for a team of health workers sufficient to provide ART for 500 patients have an eNPV of 1.23 million year-2000 US dollars, assuming that the scholarship recipients are in addition to the health workers who would have been educated without scholarships and that the scholarships reduce annual HAHW emigration probabilities from 15% to 5% for five years. When individual variable values are varied from this base case within plausible bounds suggested by the literature, eNPV of the "conditional scholarships" never falls below 0.5 million year-2000 US dollars.
Keywords: HIV/AIDS; Health Workers; Conditional Scholarships; Sub-Saharan Africa; Antiretroviral Treatment
JEL Codes: I18; I22; J2; J24
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| conditional scholarships (I22) | increase in the number of health workers educated (I25) |
| conditional scholarships (I22) | decrease in the probability of health worker emigration (I25) |
| increase in the number of health workers educated (I25) | increase in overall health worker output (I15) |
| conditional scholarships (I22) | increase in overall health worker output (I15) |
| conditional scholarships (I22) | positive ENPV (H43) |