A New DFID-Global Fund Performance Agreement: 10 Benchmarks to Achieve Maximum Impact

September 22, 2016

This past weekend in Montreal, the Global Fund to Fight AIDS, TB and Malaria matched and exceeded its last three-year replenishment cycle with contributions of nearly $13 billion USD for its work, making the agency one of the world’s largest external funders for health in low-income countries. The UK was the second largest donor to the replenishment with a 1.1 billion GBP contribution and an innovation—a publicly-posted performance agreement including 10 benchmarks against which 10 percent of the total contribution is held. While 100 million GBP ($130 million USD) at risk in an overall $13 billion USD replenishment may not be significant financially, the real news here is that a donor government is thinking about new ways to enhance the performance accountability of multilateral organizations—certainly a departure from business as usual.

Here are the benchmarks (my summary):

  1. Develop a value for money framework to guide the design and implementation of grants, and report back on implementation progress and impact
  2. Deploy at least 15 percent of grant monies using payment by results approaches
  3. Generate savings of at least $250 million in procurement and supply chain efficiencies
  4. Deliver on anti-corruption commitments already made
  5. Direct at least 85 percent of resources to low-income and lower middle-income countries
  6. Leverage greater recipient government funding for every pound of grant funding (ratio of 3:1 for low-income countries)
  7. Increase private sector funding; develop and deploy innovative models
  8. Reduce HIV incidence in adolescent girls and young women in hardest hit areas of sub-Saharan Africa by 40 percent. Note that this is USG PEPFAR’s DREAMS goal
  9. Develop clear and measurable indicators of health system strengthening, and track progress against these indicators
  10. Strengthen independent advice and scrutiny of the Global Fund to ensure that it is following best practice in seeking value for money

Progress made, challenges remain

Given our work on value for money and results-based funding at the Global Fund, I was thrilled to see the issues show up in the agreement. While progress against the three diseases advances and much has improved at the Global Fund since the 2013 replenishment, challenges remain: a relatively stagnant funding outlook, the challenges of low-quality data and verification of results, and the need to focus human and financial resources on the most effective interventions, in the most at-risk populations, in the most affected countries.

New priorities are also on the agenda: the quest for sustainability and transition away from donor funds, the corresponding imperative for domestic resource mobilization, and expanding coverage goals that offer the promise of an AIDS-free generation—but which, for now, remain aspirational within a context of limited resources and massive dropout along the HIV testing and treatment cascade.

The UK-Global Fund performance agreement sets out an ambitious agenda to meet some of these challenges.

Value for money & payment by results: performance verification needed

Explicitly considering country-based cost-effectiveness evidence (#1) to inform the allocation of grant resources among interventions, populations and products should increase bang for the buck. To get from evidence to resource allocation, however, some thought will need to be given to how these new technical considerations feed into the ad hoc processes now employed by many country coordinating mechanisms (CCM) to determine who gets what in grant proposals to the Global Fund.

Payment by results (#2) changes the basis of grant disbursements from expenses (inputs) to outputs, outcomes, or impact, a move that can create virtuous incentives for better performance, sometimes with measurable effects on health outcomes if well designed. Indeed, the DREAMS goal (#8) may be a good candidate for payment by results given its clear geographic focus and outcome measure, and hopefully a baseline incidence survey already in place.

Both actions (#1 and #2) will require better performance verification than is currently the common practice; a technically sound and robust approach, unannounced visits, representative coverage of reference population and indicators, and use of an independent third party to carry out verification are among our earlier recommendations that we hope will be considered and funded.

Our earlier reports (here and here) go in depth on recommendations for implementation of both measures.

A simple target: directing at least 85 percent of resources to LICs and LMICs

As for more money to lower-income countries (#5), given the inclusion of the lower middle-incomes as a category within the definition of low-income, this seems not at all challenging as a benchmark. According to its allocation paper, between 2014 and 2016, the Global Fund spent less than 9 percent[i] of its resources upper middle-income countries, far from the 25 percent suggested by the benchmark.

And a slightly more challenging one: leveraging greater recipient government funding

Greater co-financing from low-income countries (#6) may be among the most challenging benchmarks, given that low income country co-financing amounted to less than 1 percent of the current grant portfolio in those countries.[ii] The question is whether the Global Fund will switch to a Gavi-style direct co-payment for commodities or whether it will try to more rigorously track and monetize recipients’ in-kind contributions to programs, or both. Incipient work on disease-specific modules of the National Health Accounts (NHA) based at WHO has proceeded slowly given that most countries don’t budget this way, and—as a result—NHA still can’t be used for accountability in real time. Country budget monitoring is another option, but receives only small-scale support at the moment. In addition, it’s not always desirable to shift government spending towards HIV/AIDS programs per se—given tight budget constraints, country governments may well be funding other essential health interventions that may be crowded out via a very rigorous co-financing requirement.

It is tough to figure out how hard #6 will be given that there are a number of interpretations that might be given to the language included in the benchmark. One way to interpret a 3:1 ratio is that 75 percent of a low-income country program’s cost is to be funded through domestic resources; I don’t think this was the intention of the parties. Another interpretation, given the specific language that applies to each pound of UK taxpayer money rather than Global Fund grant funds more broadly, is that 25 percent of the total program would need to be financed by countries themselves. Still hard, but not impossible though this will depend on how the accounting for co-financing will work. In any case, #6 needs more thought on the specific measure that would be used to track progress; this will have implications for its ambition and feasibility.

Private sector funding: nowhere to go but up

More private money with more innovation (#7) seems like a good idea, bolstered by the doubling of private donor contributions that came as part of the replenishment itself. The only way is up—just 5 percent of Global Fund money came from the private sector in the 2014-2016 period. And there’s a development impact bond on malaria that still needs an outcome funder in Mozambique . . .

Finally, I love benchmark #10 but must disclose my conflict of interest as we at CGD are already providing solicited and unsolicited independent advice and scrutiny on this very topic! But since it must be strengthened to meet the benchmark, I had better figure out the baseline value.

Thanks to Rebecca Forman for the calculations of the shares that appear in this blog, and to global health colleagues for review.



CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.