Paying for results is not the same as paying for sustainability

One of this year's World Water Week sessions is titled Sustaining Public Drinking Water Services at Scale with Results-Based Financing. Read slowly, the title makes a claim. It joins two things that do not automatically go together, paying for verified results and sustaining a service, and it assumes the first produces the second. That assumption is what the sector has been arguing about for a decade, and it reduces it to two questions. What are we buying when we pay for results? And who pays once we stop?

Results-based funding (RBF) is a real advance. Its critics understate it; some of its advocates overstate it. But the test that decides its value lies where the sector still looks last, in whether the systems it touches come out credible enough to carry the service on their own.

What RBF gets right

For decades the sector financed infrastructure on the assumption that services would follow. RBF replaces that logic by linking payment to verified results. A borehole is not a service, and a connection is only useful if water keeps flowing.

The range of financing instruments is broad. Uptime Catalyst Facility contracts, the World Bank’s Program-for-Results (PforR) operations, output-based aid, performance-linked loans and some impact bonds all work differently, but they share a common logic. Targets are set in advance, a financial incentive is attached to their achievement, and results must be verified before funds are disbursed. The OECD's July 2026 report Financing Water Security notes that the indicators used have been shifting away from physical outputs towards service performance and organisational change.

The difficulty arises when achieving a result during the financing period is taken as evidence that it will be sustained once the financing ends.

Three kinds of sustainability

A water service can be considered sustainable in three ways. First, it must be technically sustainable, meaning that water continues to be delivered at an acceptable quality. Second, it must be operationally sustainable, meaning that service providers can maintain that performance over time. Third, it must be financially sustainable, meaning that the country can continue to meet the real costs of service delivery and asset renewal from domestic resources. A well-designed scheme may perform strongly on the first two dimensions while doing little to address the third.

Consider a service whose true annual cost is 100 units. Before the scheme, tariffs bring 30 and the state and local authorities 10, so the service runs on 40. Results-based funding then adds 55, and the incentive lifts tariff revenue to 35. Resources reach 100, and every indicator is strong. When the payments stop, tariffs and public contributions leave 45 against a cost of 100. Five years of measured performance produced a durable gain of five units. The deficit was not removed. It was financed, on condition of performance. That is better than an unconditional subsidy, but the service remains dependent on external support.

Subsidy itself is not the problem. In sparsely populated areas and low-income neighbourhoods, full cost recovery through tariffs is often neither realistic nor desirable, and the GLAAS 2024/2025 report finds that fewer than a third of reporting countries cover even 80% of operating costs from tariffs and household contributions. The key distinction is between predictable domestic subsidies, which can be sustained over time, and open-ended external subsidies with no pathway to replacement, which risk creating long-term dependency. Money is not the only constraint. Roughly half of reporting countries spend less than 75% of the domestic resources they have already committed, largely because of procurement and disbursement bottlenecks.

What the evidence supports, and what it does not

The OECD 2026 report is clear about the central risk. Where subsidies are paid to providers, RBF may foster dependence on external funding rather than incentivise efficiency and the development of more stable revenue sources such as tariffs. The report also flags verification costs, risks transferred to providers, and the limited evidence on cost-effectiveness. It is also transparent about its own limitations: the chapter draws on eight case studies, and the authors had no access to independent evaluations of the initiatives they examined.

On PforR, the World Bank's Independent Evaluation Group (IEG) published an early-stage assessment in 2016. It found disbursement-linked indicators (DLIs) were not consistently linked to long-term outcomes and highlighted a structural tension in their design. To be effective, a DLI must be achievable enough to provide predictable disbursements and demanding enough to drive behaviour change. IEG found that, in practice, the balance often tilted towards predictability rather than transformation. It also found no evidence at that stage that improvements to systems supported through PforR extended beyond the programmes in which they were introduced. IEG stressed that this assessment was premature as no programme had yet closed.

Ten years later, that report remains the only dedicated independent evaluation of the instrument. IEG's annual results reports, including the latest from July 2026, do not distinguish PforR from investment lending in their published ratings. As a result, an instrument of this scale has not been independently assessed since its early years.

Subsequent evidence is narrower in scope but more concrete. In 2022, IEG assessed Vietnam's Results-Based Rural Water Supply and Sanitation programme, the only independent field evaluation of a completed water PforR to date. The operation was rated highly satisfactory, but two findings qualify that positive assessment. First, the monitoring database and beneficiary feedback system were discontinued after the PforR ended because funding was no longer available. This echoed IEG's earlier concern that systems strengthened through PforR might not be sustained beyond the programmes they were designed to support. Second, targeting appeared to favour better-off areas. Among the selected communes, those with the fifty lowest poverty rates were more likely to be chosen than those with the fifty highest. IEG concluded that a design rewarding the efficient achievement of results may create incentives to avoid the hardest-to-reach locations. The programme delivered, and an independent evaluator confirmed as much. What it did not demonstrate was that key systems strengthened under the programme would continue functioning after PforR financing ended, or that incentives to reach the poorest communities had been fully addressed.

Longer-term evidence points in the same direction. IEG's 2018 evaluation of a decade of World Bank water and sanitation lending, worth USD 30.3 billion, found 71% of projects achieving moderately satisfactory outcomes or better, but that 42% of those outcomes faced significant or high risk, largely because of concerns about financial viability. It concluded that attempts to strengthen providers’ financial performance through loan conditions had produced disappointing results. For years, the sector has tried to improve financial performance through contractual requirements; RBF seeks to create incentives for it instead. The proposition is plausible, but the evidence base remains limited. Paul Clist's synthesis of eight payment-by-results projects found no evidence that the mechanism generated greater innovation or autonomy than conventional aid. Meanwhile, the latest evidence map, published in 2025, catalogues 428 studies most of them in health, agriculture and education, with water absent from the sectors highlighted in the review. 

The design question that follows

Uptime is among the most carefully structured applications of RBF in rural water services. In the UDUMA arrangement described by OECD, payments are linked to availability, water volume used and locally generated revenue, with an additional payment equal to 50% of local revenue collected, so as not to discourage user contributions. Uptime explicitly recognises that user payments rarely cover the full costs in remote settings and treats concessional finance as a catalyst and a transitional source of funding. Its March 2026 strategy places greater emphasis on integration with government programmes and public co-financing. The facility’s self-reported results have not yet been independently evaluated. That shift changes the test the model ultimately has to meet. The question is no longer only how many systems maintained availability above 96%, but also what proportion of their costs is ultimately financed by domestic resources.

PforR can address the other half, because a DLI can be attached to a tariff reform, an operating ratio, a budget allocation or an executed transfer. Egypt's Sustainable Rural Sanitation Services PforR, approved in 2015with USD 550 million from the World Bank and later expanded to USD 850 million, alongside USD 300 million from the Asian Infrastructure Investment Bank, combined access expansion with decentralisation of responsibilities to water and sanitation companies and performance-linked fiscal transfers. It financed the conditions that produce results rather than just the results. The programme does not close until the end of 2026, so it is too early to tell whether those transfers will be a regular feature of the Egyptian budget or remain dependent on external financing.

This points to a two-tier design. At the provider tier, a performance contracts reward availability, quality, volumes, response times and revenue. At the system tier, a results-based agreement between government and a development financier focuses on indicators such as budget allocation and execution, regulation, tariff reform, asset management, coverage of the hardest-to-serve, and the growing share of domestic financing. The financier incentivises the state to strengthen or build the system, and the state purchases provider performance. Over time, the first link disappears while the others remain. That disappearance is the proof of success, not a failure of the partnership. The trajectory can be built from the outset, with external and domestic shares gradually shifting from 80/20 to 50/50 to 20/80 before responsibility for the payment mechanism passes to a national institution.

Changing the measure of success

At Riva Africa, our work follows a clear sequence: institutional credibility produces system performance, system performance produces financial credibility, financial credibility unlocks investment, and investment delivers scale. RBF is most valuable when it is understood as an instrument for creating financial credibility. It is not a substitute for credibility, and it is not a permanent revenue source.

If that is the purpose, the dashboard has to change. Alongside uptime and volumes should sit the domestic financing ratio, operating cost coverage including recurrent national subsidies, asset renewal coverage, budget execution rates, and the real cost per unit delivered. Success then reads differently: not uptime 97%, but uptime 97%, domestic financing from 35% to 75%, cost per user down 20%, budget execution 92%, external financing from 55% to 15%.

A scheme that maintains excellent service for as long as a financier pays is a good performance-financing mechanism. A scheme that uses that money temporarily to build an institution, a service economy and a national payment mechanism capable of outliving it becomes an instrument of transformation. Uptime measures whether the water is flowing today. Sustainability measures whether it will still be flowing when the financier has stopped paying. That second measure deserves a place among the criteria by which results-based funding in water is designed and judged.