Many researchers and policymakers have hypothesized that funding models tying grant payments to achieved and verified results — next generation financing models — offer an opportunity for global health funders to push forward their strategic interests and accelerate the impact of their investments. This brief, summarizing the conclusions of a CGD working group on the topic, outlines concrete steps global health funders can take to change the basis of payment of their grants from expenses (inputs) to outputs, outcomes, or impact. While the report focuses primarily on how the Global Fund to Fight AIDS, Tuberculosis and Malaria can make this shift, this brief offers insights for other global health funders looking to address their strategic objectives, increase the efficiency and effectiveness of their investments, and increase their health impact for the populations they serve.

What Is Next Generation Financing?

Next generation financing models create an explicit and enforceable contract between a health funder such as a government, health aid donor, or philanthropist and recipients of that funding, such as other governments, health-provider networks, or nongovernmental organizations, in which both parties agree that some or all of the funding will be tied directly to achieving mutually important, realistic, and measurable gains in the provision of health services or population health.

Similar to results-based financing and pay-for-performance programs, next generation financing models focus on outputs or outcomes rather than inputs or processes that characterize most existing relationships between funders and recipients in global health.

How Are Next Generation Financing Models Different from Traditional Models?

Next generation grants establish a collaborative and mutually accountable relationship between funder and recipient. Recipients are liberated from paternalistic micromanagement, parallel documentation, and unilateral funder control over disbursement decisions. Meanwhile, the funder gains a clear mechanism with which to hold recipients accountable: it will only pay for results that have been achieved.

These agreements can also align incentives for both parties toward their shared goal of health improvement. The funder’s principal objectives are prioritized; however, if the recipient effectively achieves the funder’s principal objectives, it can then apply any cost savings to pursue its own secondary priorities.

In some cases, next generation grants will not achieve results and will not culminate in any or full disbursement. Although a situation in which people do not receive the health services they need is indeed a failure, unsuccessful grants might still represent effective grant management because no donor funds would be wasted on activities or partners that fail to deliver results. As a result, the funder can redeploy those funds with better strategies or through a different recipient. With a traditional model, in contrast, donor resources are spent even though the recipient has failed to delivered services.

Why Shift to Next Generation Financing Models?

Although use of these models is still emerging, initial experiences suggest incentives matter for results. The best evidence for results-based payments to health facilities in low- and middle-income countries has emerged from the World Bank’s Health Results Innovation Trust Fund. In Argentina, Rwanda, Zambia, and Zimbabwe, rigorous evaluations found that better incentives greatly improved health system performance, sometimes with measurable effects on health outcomes. Evidence also comes from higher-income countries such as New Zealand, where the introduction of incentives for coverage of preventive services coincided with a 30 percentage point increase in childhood vaccination and a 20 percent increase in screening for cardiovascular disease.

The ubiquity of suboptimal incentives in global health financing is compromising our collective ability to maximize impact and contain costs. Even if we do not yet know exactly what will work best — and even though the risks are real and considerable — getting incentives right may be the single most effective strategy for stretching the impact of scarce global health dollars.

What Does It Mean to Change the Basis of Payment?

Most health funders use expenses as the basis of payment for grant agreements. Agreements specify inputs that grant recipients can purchase with donor funds, and funders focus on preapproving and verifying eligible expenses incurred.

But expenses are not the only option. The basis of payment can rest anywhere along the chain of results; it can be grouped into either input financing (traditional basis of payment) or performance-based payments (outputs, outcomes, and impact). The latter group constitutes next generation grant agreements.

Importantly, the decision to change the basis of payment need not be all or nothing; grant design can lie anywhere between 100 percent input financing and 100 percent payment for results. The degree and extent to which payment for results is appropriate will vary according to the country context, disease control objectives, and the grant recipient.

Why, When, and to What Extent Is Changing the Basis of Payment a Good Idea?

Input financing models may fail to provide the grant recipients with sufficient incentive to improve efficiency by reducing costs or increasing outputs or outcomes. By shifting to performance-based payments, the funder can better align these incentives with its strategic objectives. In so doing, it can drive greater impact, faster progress, and more efficient activities.

There are four important considerations for when and to what extent a funder can change the basis of payment:

  1. Are the goals of an agreement contractible?

    Changing the basis of payment is only feasible where program goals can be captured by at least one contractible indicator. A contractible indicator is influenced by or under control of the grant recipient or health system; able to change within the grant’s timeframe; measurable; independently verifiable; and a direct proxy for or on the direct causal chain to a meaningful health outcome. It also must not incentivize violations of human rights.

  2. Can the basis of payment be measured and verified with sufficient precision to inform payment?

    There are two basic options for conducting measurement: a funder can rely on the grant recipient’s self-reports and then verify those reports using an independent verification agent, or it can contract independent measurement or use ongoing independent measurement exercises (e.g., Demographic and Health Surveys) that do not rely on grant recipient self-reports.

  3. Is the risk of nonpayment acceptable to the funder?

    A next generation financing model requires the funder to withhold payments when the grant recipient does not achieve the specified results. In some instances, this risk of nonpayment may be ethically unacceptable or fail to serve the funder’s strategic interests. For example, a funder would not want to cut off access to essential commodities due to poor performance because it would prevent patients from accessing lifesaving medicines.

  4. Is the agreement acceptable to the grant recipient such that it agrees to participate?

    Both the funder and grant recipient must agree on the change in basis of payment. A grant recipient may not accept the new basis of payment because, for example, it may have to assume the risk when service delivery costs end up higher than expected, it may not have sufficient liquidity to prefinance service delivery, or it may feel that achieving the indicators is not fully within its control.

    To counter these issues, a funder could make the change in basis of payment more attractive to a grant recipient by allowing the recipient to keep a greater portion of the cost savings, increasing the overall allocation for accepting additional risk, or relaxing other fiduciary requirements to save the recipient from administrative costs.

How Should Grant Recipients Be Remunerated for Their Achievements?

The simplest way to design payment for results is a fixed-price model whereby the grant recipient is paid a fixed price for every unit of verified performance, with possible adjustments for quality. For example, a funder could pay a fixed price for each newly enrolled patient on antiretroviral therapy who has complete and accurate retention records at the end of the program’s first year. The payment could be adjusted in following years to account for newly enrolled patients and retention of patients enrolled in the first year.

Within this model, a payout schedule determines how different levels of achievement relate to different levels of payment. There are four considerations for determining the payout schedule:

  1. What is the purpose of the fixed price? The fixed price can act as a supplement to input financing, as a subsidy of the cost of achievement, or as a substitute for input financing.

  2. What does the fixed price cover? The funder could offer a fixed-price payment for achievement either above the grant recipient’s baseline level or for the grant recipient’s entire output, including baseline levels.

  3. Is the payout schedule continuous or lumped? In many cases, a continuous payout schedule, which rewards the grant recipient for each incremental unit of progress, is preferable to a lump-sum payout schedule, which rewards the grant recipients for each target reached because lump-sum payments are delivered on an all-or-nothing basis.

  4. Does the fixed price change with greater achievement? The simplest, linear payout schedule of one constant price for each unit of output may not always be appropriate. For example, the funder and grant recipient may agree that the cost of service delivery will increase as the grant recipient reaches more patients. In this case, the fixed price may change after the recipient reaches a certain performance threshold.

There are two basic variations to the simple fixed-price model, which insulate the funder and grant recipient from downsides of the model (e.g., risk to
the grant recipient, difficulty in setting the right price). The first is the fixed-price / cost reimbursement menu and the second is the shared surplus or shared savings model. For more on these variations, see the full report.

How Much Should Funders Reward Grant Recipients for Their Achievements?

The criteria for the right price will vary based on the purpose of the fixed-price payment, the setting, and the policy goals. However, there are some practical approaches to determining the initial price and making adjustments over time as new information becomes available or as the context changes. For instance, the funder could change the structure of the fixed price each year as it learns more about the costs incurred by the grant recipient.

What Is Needed to Ensure That the Incentives Do Not Drive Unintended Consequences?

Stakeholders rightly worry that financial incentives could undermine the rights and welfare of citizens of key affected communities. Next generation grants offer an opportunity to considerably improve the health service access and outcomes for members of these communities while protecting their rights via three strategies:

  1. The choice of indicator and payment schedule should be designed to ensure that key affected communities and underserved groups (e.g., women, rural populations, and the very poor) are served. This could be achieved by disaggregating reporting and payment by population type.

  2. Communities should be part of the process for monitoring, providing feedback, and verifying that results were achieved and that they responded to community needs without coercion.

  3. Where service delivery to key populations is a primary goal, a community-based membership organization should be responsible for recruiting new members; offering prevention, testing, and counseling services; acting as expert patients; and measuring the size of the affected community.

How Can a Funder Ensure That the Grant Recipient Is Achieving the Agreed-Upon Results?

At the outset, the funder should make clear that it will verify at least some of the grant recipient’s self-reported performance at stages where payment is linked to results, performance verification visits will be random and unannounced, and evidence of misreporting will be costly. Penalties could include terminating a grant agreement, precluding access to future funding, or imposing fines. The funder and grant recipient should also agree on how imprecise verification findings will inform disbursement and penalties, and how much imprecision will be tolerated.

Performance verification should be implemented by an agency that is independent of the grant recipient and, to maximize the credibility of the exercise, independent of the funder. Where possible, the funder should seek to piggyback on existing or harmonized data collection and verification efforts; however, the funder will need to ensure that the existing tools are rigorous and independent enough to serve that purpose.

How Will a Funder Know Whether the New Financing Model Is Working with Respect to Its Own Objectives?

New grant designs should be assessed with respect to the stated goals for their implementation and hypotheses about how they should lead to change in: the funder’s and grant recipient’s behaviors, the interactions between the funder and grant recipient, and the impact on the health and welfare of the program beneficiaries. This assessment can be conducted at various stages in the grant cycle.

What Steps Should a Funder Take When Considering Shifting to a Next Generation Financing Model?

See the full report for illustrative examples of how next generation grants could be designed and implemented through a multistage framework. Although these examples are geared toward the Global Fund, the framework can be applied to other global health funders.

The Center for Global Development if grateful to the Bill & Melinda Gates Foundation and UK Department for International Development for contributions in support of this work.

Rights & Permissions

You may use and disseminate CGD’s publications under these conditions.