Accounting for Aid: Estimating the Impact of United States’ Global Health Investments on Mortality Among Women of Reproductive Age Using Synthetic Control and Bayesian Methods

Accounting for Aid: Estimating the Impact of United States’ Global Health Investments on Mortality Among Women of Reproductive Age Using Synthetic Control and Bayesian Methods
Citation: Ahsan, Karar Zunaid and Angeles, Gustavo and Choudhury, Allysha and Singh, Kavita and Taylor, Tory and Majid, Farhan and Lucas, Rachel and Cohen, Rob and Gawande, Atul and Weiss, William, Accounting for Aid: Estimating the Impact of United States' Global Health Investments on Mortality Among Women of Reproductive Age Using Synthetic Control and Bayesian Methods (August 23, 2024).
Abstract: Background: The United States government (USG) is a key global actor in preventing mortality and supporting lifesaving health services among women and children in low-and middle-income countries (LMICs). Since the Foreign Assistance Act of 1961, USG has funded global health programs targeting specific conditions and strengthening health systems for the delivery of essential services via the United States Agency for International Development (USAID) and other USG agencies. However, directly quantifying and attributing the impact of USG health investments is challenging due to the inability of conducting randomized control trials for such large programs at scale. In this study, we compared two quasi-experimental impact evaluation approaches to estimate the effects of sustained USG health investments on mortality among women of reproductive age (WRA).

Methods: We employed synthetic control analysis and a Bayesian alternative to synthetic control to estimate the impact of USG's global health investments on WRA mortality rate across 16 LMICs that received sustained, multifaceted, and above-average USAID global health funding levels from 2005 to 2019.

Findings: Countries receiving sustained, multifaceted, and above-average USAID global health funding had a reduction in the annual WRA mortality rate by 0.65 per 1,000 WRA throughout the post-treatment period. For the years (i.e., 2009-2019) where the effect estimates are statistically significant (p<0.001), the reduction in WRA mortality rate was 0.80 per 1,000 WRA. This finding was consistent with several sensitivity analyses and the Bayesian model. We conservatively estimated that about 1.0-1.3 million WRA deaths were averted in study countries between 2009 and 2019 as a result of USG health investments.
Author(s): Zunaid Ahsan, Gustavo Angeles, Allysha Choudhury, Kavita Singh, Tory Taylor, Farhan Majid, Rachel Lucas, Rob Cohen, Atul Gawande, William Weiss
Year: 2024
Language: English