A Case Study on Insecticide-Treated Bednets

Both GiveWell and Open Philanthropy (OP) have been longtime supporters of charities that distribute long-lasting insecticide-treated bednets (LLINs), such as the Against Malaria Foundation. Open Philanthropy continues to believe that these are some of the best charitable giving opportunities in the world — on the order of ~10x more cost effective than cash transfers.[1]

During the early 2000s, LLINs were relatively uncommon (see figure below). However, coverage exploded over the next two decades. One paper estimates that the scaleup of LLINs from 2001 to 2010 prevented some 850,000 child deaths.

Figure 4

Since we believe the large scale-up of LLINs has been extremely positive, we commissioned a case study from the Urban Institute on how the scale-up happened, to determine if there are any lessons one might take for future global health scale-ups. In this blog post, we give a summary of OP’s high-level takeaways from this case study, though we encourage readers to read the full report (or its summary) to get a more detailed understanding of the drivers of LLIN scale-up.

The scaleup of LLINs was primarily driven by the availability of additional funding

The scaleup of bednet coverage was largely driven by increased availability of funding to tackle malaria. Between 2000 and 2003, all health ODA for malaria was ~$100M a year (pg. 34). By 2008, it had surpassed $1B a year (pg. 35). Universal bednet coverage — and even free bednets — were simply not on the table when there was less than 10% of current funding available.

Figure 5

Without the expansion of funding — significantly but not exclusively via the Global Fund — there would have been “no bednet explosion” (pg. 42).

Malaria funding appears to have increased partially as a consequence of increased HIV funding

The Global Fund was initially conceived as a funding vehicle for AIDS treatment; the founding executive director of the Global Fund notes that “the creation of the Global Fund was driven primarily by the HIV pandemic. Malaria was serious, and worsening in Africa, but the crisis was HIV.” (pg. 37). Tuberculosis and malaria were only added later in the process.

The United States first committed $100M to the Global Fund in 2001, and this founding donation would be a significant part of the budget through 2003.

Former U.S. President George W. Bush’s first major foray into aid in Africa was PEPFAR (started in 2003). It started with a much larger sum than the Global Fund, with a commitment of $15B over four years. PEPFAR’s success “contributed to the unprecedented growth in global health financing in this period”, and it was only after PEPFAR that it became clear to the Bush administration that they should launch a separate malaria initiative. The case study notes that:

“Staff in the Bush administration, such as Bush’s senior policy aide Michael Gerson, 147 started calling attention in the White House to the effects of malaria on economic and education outcomes in PEPFAR countries. ” (pg. 50).

The President’s Malaria Initiative (PMI) was launched seemingly in response to these concerns in 2005.

… but there were also other drivers

Coordination between buyers and manufacturers mattered

Until 2004, there was only a single manufacturer of LLINs — and thus, it was difficult to scale up distribution. A second manufacturer was approved by the WHO Pesticide Evaluation Scheme in 2004; a third in 2007.

However, this scaleup of technical capacity was secondary to an increase in funding; “most experts indicated that they thought the Global Fund’s market-signaling to net manufacturers helped enable net scaleup. ” (pg. 43).

Manufacturers needed assurance that there was a market before they scaled and the Global Fund needed to know there were enough manufacturers available to meet demand.

High case burdens mattered to inclusion in the Global Fund

Malaria case levels had reached their highest level ever at the beginning of the 2000s. In 2000, nearly 900,000 people died from malaria, only 30% fewer than died of HIV/AIDS that year. This incredibly high case burden was top of mind for many African policymakers and contributed to prioritization of malaria to be included in the Global Fund.

Cohen and Dupas 2010 does not seem to have been singularly catalytic in moving funders towards free bednets (but it helped sustain momentum)

By 2005, there was consensus that nets did work (see table 1 on pg. 8 for a summary of the evidence generation timeline), but questions about distribution remained. There was considerable debate over the optimal price for ITNs (a term for insecticide-treated bednets that includes nets which aren’t as long-lasting). The US and the UK had previously favored subsidized but not free nets. Would people value nets less and be less likely to use them for malaria prevention if they were free?

Prior to the case study, we knew that the discourse shifted from cost-sharing towards free bednets in the mid-2000s. We were interested in knowing how much of a role academic research played in this change. We thought Cohen and Dupas 2010 — an RCT showing that free bednets were far more cost-effective than subsidized bednets, first published as a working paper in January 2008 — was a particularly good candidate for being catalytic. If it was, this would suggest that strategic, well-placed philanthropic efforts to support research could have huge payoffs.

However, the story appears to be somewhat more complicated. We think that Cohen and Dupas was slightly too late to drive the change from subsidized to free nets but did help support the continued distribution of free nets.

Specifically, the WHO announced a shift in its policy recommendations endorsing free mass distributions in August 2007 — before the working paper was published. Thus, the paper could not have caused the change itself.

However, “the evidence generated by Dupas and Cohen on net access [was] a vital ingredient in resolving this resistance [to providing bednets for free]” (pg. 68). The study also helped to bring finance ministers on board by providing an economic justification they could trust: according to John McArthur, “EconLit-focused people resisted free distribution, but once it had economists’ blessing, they said this [was] a good idea, too.”

What can OP conclude from this case study?

Scaleup of ITNs was not a mono-causal story; there was not a single event or paper or decision that led to the massive expansion of bednet distribution. Rather, there were several driving factors that all had to come together for ITNs to become near-ubiquitous as a health technology.

The main lesson for a philanthropist may then be the importance of consensus building across different institutions, particularly governments and international organizations.