Assessing the impact of the president’s emergency plan for AIDS relief on all-cause mortality

Abstract

This study estimated the impacts of PEPFAR on all-cause mortality (ACM) rates (deaths per 1,000 population) across PEPFAR recipient countries from 2004–2018. As PEPFAR moves into its 3rd decade, this study supplements the existing literature on PEPFAR ‘s overall effectiveness in saving lives by focusing impact estimates on the important subgroups of countries that received different intensities of aid, and provides estimates of impact for different phases of this 15-year period study. The study uses a country-level panel data set of 157 low- and middle-income countries (LMICs) from 1990–2018, including 90 PEPFAR recipient countries receiving bilateral aid from the U.S. government, employing difference-in-differences (DID) econometric models with several model specifications, including models with differing baseline covariates, and models with yearly covariates including other donor spending and domestic health spending. Using five different model specifications, a 10–21% decline in ACM rates from 2004 to 2018 is attributed to PEPFAR presence in the group of 90 recipient countries. Declines are somewhat larger (15–25%) in those countries that are subject to PEPFAR’s country operational planning (COP) process, and where PEPFAR per capita aid amounts are largest (17–27%). Across the 90 recipient countries we study, the average impact across models is estimated to be a 7.6% reduction in ACM in the first 5-year period (2004–2008), somewhat smaller in the second 5-year period (5.5%) and in the third 5-year period (4.7%). In COP countries the impacts show decreases in ACM of 7.4% in the first period attributed to PEPFAR, 7.7% reductions in the second, and 6.6% reductions in the third. PEPFAR presence is correlated with large declines in the ACM rate, and the overall life-saving results persisted over time. The effects of PEFAR on ACM have been large, suggesting the possibility of spillover life-saving impacts of PEPFAR programming beyond HIV disease alone.