How Cost-Effective Is Global Health Aid?

We’re excited to share our latest CGD working paper, “The Cost-Effectiveness of Health Aid: An Exploratory Quantitative Analysis” along with Karen Grepin, David Watkins, and Joe Dieleman. We delve into a long-standing debate in global health: Is health aid good value for money?

We realize we are wading into not only a contentious question but even the framing itself may be presumed as biased. We want to emphasize that our study is descriptive, not normative. The study does make a claim that health aid should always be cost-effective. But before we get ahead of ourselves, what exactly did we find?

Most health aid is cost-effective

Our study analyzes data from the OECD Creditor Reporting System between 2019 and 2021, cross-referencing it with a list of 218 highly cost-effective interventions identified by the Disease Control Priorities Project (DCP3). On one hand, 61 percent of health aid projects appeared to include at least one intervention that is considered cost-effective. The study also found that, on average, 11.9 highly cost-effective interventions were mapped to each project, indicating that most projects are bundled.

Of the interventions identified, 33 percent fell into the category of  “most cost-effective,” with the remaining interventions as still cost-effective. Amidst calls that health aid should not only be cost-effective, but cost-effective at the margin, our paper indicates that donors may already be doing just that.

How do donors differ?

One key finding is that there is little evidence to suggest that donors are tailoring the cost-effectiveness of the interventions based on the income levels of recipient countries. The distribution of cost-effective interventions did not significantly differ across low- or middle-income countries. This finding indicates that donors may practice a one-size-fits-all approach rather than tailoring interventions to specific local needs.

The study also found no correlation between the number of projects from a specific donor and the percentage of a donor’s projects which had cost-effective interventions, suggesting that having more projects does not translate to a greater share of cost-effective interventions.

We also have a juicy table (see table 2) that compares donors by the cost-effectiveness of their project portfolios—see below. The headline message is that Global Affairs Canada, Spanish Municipalities, USAID, and the Gates Foundation are the donors with the most projects that have identified more cost-effective interventions (as ranked by the list of interventions with more keywords/phrases associated with the list of cost-effective interventions, i.e. a threshold of 0.8).

Table 1. Donor agencies ranked by number of projects with a match score ≥ 0.8

Our paper shouldn’t be cannon fodder in cost-effectiveness debates

Although our paper is descriptive, we are mindful that the study risks becoming cannon fodder in one or more ideological debates: “Aid should be cost-effective!” or “Aid should be cost-effective at the margin!” We actually don’t have a horse in those races.

In fact, we sympathize with arguments that aid should be country owned, implying that it should be countries rather than donors that decide what gets covered, not the other way around. Karen Grépin has examined how vertical approaches can crowd out other equally essential services—for example, HIV services have been found to crowd out other services, such as vaccination. Joe Dieleman has done extensive work on aid fungibility, examining whether increases in health aid lead to decreases in government health spending. The implications of aid fungibility mean that donor investments in cost-effective intervention do not guarantee sustainability nor do they come freely—there could be unintended consequences. Thus, we caution readers—and cost-effectiveness believers—to think carefully through the consequences rather than merely land on the simplistic conclusion that “health aid should be more cost-effective.”

Future directions

At a minimum, we call for better data and methods to validate these findings, with closer collaboration in partner countries about how they make priorities and decisions about what to implement. Ultimately, the findings underscore the need for a paradigm shift in the conceptualization of how global health aid is tracked.

We think that the cost-effectiveness of an intervention is an important, but not all-determining, criterion for what countries prioritize—and perhaps more importantly, what they may permit donor countries to offer.

With thanks to several reviewers and colleagues named in the paper itself