Эффективность помощи зависит от способности местных институтов сохранять проекты после ухода доноров

Why scaling what works is not enough if the systems that deliver it cannot sustain it There is a lot to agree with in the growing argument that the aid ecosystem needs to become simpler. As aid budgets tighten, governments face competing domestic pressures and humanitarian needs continue to rise, it is entirely reasonable to ask whether every dollar is being used as effectively as it could be. The argument that assistance should focus on a smaller number of interventions that have been rigorously shown to work, and then deliver them at scale, is therefore an important contribution to the debate. I agree with the problem being identified. I am less convinced by the solution. My concern is not with evidence. It is not with prioritisation. And it is certainly not with scale. The question I keep coming back to is much more fundamental: who owns and delivers the solution when the external funding is gone? That matters because the aid ecosystem has become very good at measuring the scale of delivery, but much less good at measuring the durability of what that delivery has helped create. An intervention can reach one million people and still fail to create a sustainable system. Another can reach 100,000 people, become embedded in a country’s institutions and ultimately have a far greater impact because it continues after the external funding disappears. That is the distinction we need to make. Scale is not the same as sustainability The aid ecosystem has become very comfortable talking about scale. We scale programmes. We scale pilots. We scale interventions. We count beneficiaries. But there is a problem with the way we often define scale. If an international NGO receives $50 million to deliver an evidence-based intervention to a million people, we may call that scale. But suppose the programme ends when the grant ends. What exactly has been scaled? The delivery mechanism of an NGO, perhaps. The intervention itself, perhaps. But not necessarily the country’s capacity to continue delivering it. The question should therefore not simply be how many people we can reach. It should be how we help the institutions responsible for reaching those people do it better, sustainably and at greater scale. I know some will immediately ask what happens if the system itself is the problem. How do you work with a system that is failing? I have heard this question several times. My answer is that we need to stop talking about “the system” as though it is a single, monolithic government. In many of the places where humanitarian and development organisations operate, governance is devolved or decentralised. And even where formal government structures are highly centralised, the actual delivery of services rarely is. Take health. In Nigeria, for example, responsibility is distributed across different levels of government. The federal government is responsible for tertiary institutions, including teaching hospitals. States have responsibility for much of the secondary health system, including general hospitals, while local governments have important responsibilities for primary health care. And that is before we consider private sector providers, faith-based institutions, community organisations and other actors who form part of the actual health delivery ecosystem. There is no single “health system” actor. There are multiple institutions, mandates, incentives and delivery channels. Education provides another example. In Kenya, education governance involves national and county structures, including county education directorates and relationships that require navigating both central and devolved government responsibilities. The formal lines of authority do not always tell the whole story about where decisions are made, where resources flow or where implementation actually happens. This matters because understanding the governance architecture is itself a development capability. If we do not understand who has the mandate, who controls the resources, who influences decisions, who actually delivers the service and who is accountable to citizens, then we can design the most evidence-based intervention in the world and still put it in the wrong place. The real question is not simply whether to “work with government”. It is understanding which part of the governance and delivery ecosystem needs to change, and how we engage it effectively. Sometimes that will be the national government. Sometimes a state or county government. Sometimes a local government institution. Sometimes a public hospital, a school system, a private provider, a faith-based organisation or a professional association. Often it will be several of these working together. Understanding the system therefore has to go beyond describing it as complex. We need to know who has authority, who has influence without formal authority, where the incentives sit, where money flows, where the bottlenecks are, where accountability sits and which institutions are capable of change. That is systems thinking in practice. When evidence becomes part of the fundraising proposition There is another issue we should be willing to discuss. Evidence is essential. But evidence has also become a currency in the competition for increasingly scarce aid funding. An organisation that can demonstrate a rigorous evidence base has a stronger proposition to a donor. An organisation that can demonstrate millions of beneficiaries have a stronger story. An organisation that can show cost per beneficiary, impressive evaluation results and a credible pathway to scale is easier to fund. None of that is inherently wrong. The danger comes when the measures that make an intervention attractive to a funder gradually become confused with the measures that tell us whether it is creating lasting change. Evidence can become a fundraising asset. Scale can become a fundraising asset. Even the language of sustainability can become a fundraising asset. Once that happens, there is a risk that the aid ecosystem begins optimising for what is easiest to demonstrate in a funding proposal rather than what is hardest and most important to build. This is not an argument against evidence. Quite the opposite. It is an argument for being much more honest about what the evidence is actually telling us. Evidence that an intervention works is not the same as evidence that an institution can sustain it. Evidence that an NGO can deliver something to one million people is not evidence that the government can finance or institutionalise it. Evidence of impact during a grant period is not evidence of impact ten years after the grant ends. We need to distinguish between evidence of effectiveness and evidence of durability. The first tells us that something can work. The second tells us whether it can endure. We need both. We are discovering systems thinking. Good. But let’s be careful. There is another trend in the aid ecosystem that I welcome but also approach with some caution. Donors and implementing organisations are increasingly talking about systems thinking. That is a good thing. Development problems are rarely linear. Jobs, education, health, governance, displacement, climate resilience and economic opportunity are connected to institutions, incentives, markets, policies, politics and behaviour. You cannot sustainably change one part of such a system without understanding how it interacts with the others. But there is a danger here too. Systems thinking can become the new language of fundraising. A proposal can describe system transformation, systems change, ecosystem building and adaptive management while the underlying programme remains essentially a conventional project with a five-year funding cycle, predetermined outputs and an implementing organisation at the centre. Changing the vocabulary does not necessarily change the operating model. And genuine systems work requires capabilities that the aid ecosystem has not always prioritised. It requires systems mapping and analysis, political economy analysis, institutional analysis, stakeholder and power analysis, facilitation, coalition building, adaptive management, strategic foresight, negotiation, organisational development, market analysis, policy engagement and financial structuring. It requires people who can move between policy, institutions, markets, finance, politics and communities and understand how they interact. It also requires something that is harder to put into a logframe: the ability to understand how incentives change over time and how different actors respond to one another. These are not simply programme management skills. They are capabilities for navigating complexity. Yet much of the aid ecosystem has historically rewarded people for being excellent programme managers, proposal writers, technical specialists and grant managers. Those skills are important. But they are not necessarily the same skills required to change systems. If systems thinking is genuinely the next evolution of aid practice, donors need to invest in the people who can actually do systems work. Otherwise, we risk taking a useful idea, turning it into a funding category and producing the language without building the capability. The intermediary has become the default This brings us back to the central question of who actually owns the solution. There is a structural problem in the way much external assistance is organised. Money often travels from donor to intermediary to implementing partner to community. Each layer exists for understandable reasons. International organisations have expertise. NGOs also have expertise and proximity to communities. Implementers can often move faster than government systems. But over time, the intermediary can become the system. The donor funds the organisation. The organisation employs the technical experts. It establishes the systems, builds the relationships and reports the results. And several years later, the donor asks whether the programme can be sustained. At that point we discover that sustainability was never really built into the architecture of the intervention. I realise and acknowledge that some implementing partners reading this may instinctively react defensively. After all, many organisations have spent years working in extremely difficult environments, often where governments lack capacity, political space is constrained and communities need services immediately. Their staff have delivered programmes in places where others could not. That contribution should not be dismissed. But I would encourage reflection rather than a defensive reaction. This is not an argument that INGOs and national NGOs are unnecessary. Nor is it an argument that governments should simply be handed responsibility for programmes they are not yet capable of delivering. It is a question about what success ultimately looks like. If an organisation is exceptionally good at delivering a programme, should its success be measured only by how efficiently it can deliver that programme at greater scale? Or should part of its success also be measured by whether it has helped support the creation of the institutional, financial and human capacity for that programme to continue without it? Those are not the same thing. And this is where I think we need to be fair to the implementing partners. They respond rationally to the incentives created by the aid architecture. If donors reward INGOs for the number of people reached, the size of programmes managed, the speed of delivery and the volume of funding absorbed, organisations will naturally build the capabilities required to perform those functions well. That is not necessarily a failure of commitment. It is a consequence of design. The same applies to local NGOs. The local NGO is not simply the smaller intermediary Local organisations are often presented as the answer to this problem, and there is good reason for that. They understand their context. They have relationships that international organisations may take years to establish. They are often closer to communities and national institutions. Many have deep knowledge of local political, social and cultural dynamics. But there is a danger in simply replacing an international intermediary with a local intermediary while leaving the financing architecture unchanged. A local NGO can become the delivery arm of an international organisation just as easily as an INGO can become the delivery arm of a donor. The organisation may be local, but the incentives remain external. The priorities may still be determined elsewhere. The funding may still be short term. The reporting requirements may still reward outputs rather than institutional capability. And the local organisation may still have little opportunity to build unrestricted capital, organisational infrastructure, strategic capability or long-term institutional resilience. That is not localisation in its fullest sense. It is localisation of delivery. There is an important difference. If we genuinely believe national and local actors should own development and humanitarian solutions, then they need to be treated not simply as implementing partners but as institutions with agency, capability and a long-term role in the ecosystem. That means funding organisational capacity, leadership, governance, financial resilience, innovation and institutional development, not only project activities. It also means being prepared to finance the local institution’s ability to say no, to shape priorities and to invest in capabilities that may not immediately produce donor friendly outputs. Otherwise, we risk creating a strange contradiction. We tell local organisations that they are closest to the problem and therefore best placed to respond, while continuing to finance them as though they are merely contractors. Systems financing is about changing the incentives Donors are not standing still. There is already a significant shift towards systems financing, locally led development, adaptive programming, flexible funding and longer-term approaches. That shift is important and should be welcomed. But a change in the language of financing is not enough. The deeper question is whether the incentives embedded in the financing architecture are aligned with the outcomes donors say they want. If a donor says it wants sustainable systems but funds primarily through short project cycles, measures success through annual beneficiary numbers and rewards organisations for the size of their delivery footprint, it is sending two different messages. The stated objective says: build the system. The financial incentive says: deliver the project. The organisation receiving the money will notice the second message very quickly. This is not about blaming donors. In fact, I think donors have one of the most powerful levers available to change the aid ecosystem precisely because they control a significant part of its incentive structure. If donors want sustainability, they must reward the behaviours that produce it. That could mean funding institutional capability alongside programme delivery. It could mean longer time horizons where systems change genuinely requires them. It could mean paying for the work that makes systems function such as political economy analysis, institutional strengthening, coalition building, government engagement, public financial management, local financing mechanisms, data systems, procurement reform, organisational development and market development. It could mean accepting that some of the most important outputs of a programme may not be easily expressed as beneficiary numbers. It could mean measuring the extent to which external technical assistance is being absorbed into national capability. And it could also mean asking a question that is rarely included in a conventional results framework: what will still exist when this programme ends? Imagine if that became a core donor question. Not simply how many people a programme will reach, but which institution will be stronger because of the investment, which capability will remain after the grant, which functions currently financed externally could eventually be financed domestically, and what the government, market or local institution will be able to do five years from now that it cannot do today. Those questions would change the behaviour of the entire ecosystem. What systems financing should actually finance This is where I think the systems financing conversation needs to become more concrete. If systems are the objective, then financing should not only fund activities delivered within systems. It should finance the capabilities that allow systems to function. That will include strengthening public institutions. But it may also involve improving the relationship between government and private providers, strengthening professional associations, improving procurement, building data infrastructure and developing local financial markets. In essence, it entails bringing in new actors into the system. It may also involve creating mechanisms through which domestic resources can eventually replace external funding. It may entail strengthening the institutions that regulate markets rather than funding another project that operates around them. And sometimes it may mean financing an NGO precisely because that NGO has a unique capability to help change the system. The issue is not whether money goes to an INGO, local NGO, government or private sector organisation. The issue is what capability the money is purchasing and where that capability ultimately resides. What if we measured scale differently? Imagine two programmes. Programme A reaches one million people over five years through an international implementing organisation. It is highly efficient, evidence based and well managed. Programme B reaches 200,000 people over the same period, but the funding is used to strengthen government systems, build local financing mechanisms, improve procurement and data systems, train public servants and embed the intervention within existing institutions. At the end of five years, Programme A closes. Programme B continues. Which one scaled? Our current aid architecture might say Programme A. I would argue that Programme B did. This is why we need to move beyond an output definition of scale towards an institutional definition of scale. The real test should be whether an intervention survives the withdrawal of external funding. That should become one of the most important measures of aid effectiveness. When to work around the system, within it or change it Government systems are not perfect, obviously. Some are dysfunctional. Some are captured by political interests. Some are deeply unequal. Some exclude the very communities that development programmes are intended to serve. There will be circumstances in which working around the system is necessary. In a humanitarian emergency, people cannot be told to wait while institutional reform takes place. But there is a difference between working around a system as an emergency measure and building a permanent parallel system because it is easier to deliver through an NGO. The real skill is knowing the difference. Sometimes we need to work around the system. Sometimes we need to work within it. Sometimes we need to strengthen it. And sometimes we need to help change it. That requires judgement, political economy understanding, institutional analysis, relationships and negotiation. It also requires humility. The people designing an intervention may understand the technical solution extremely well while understanding the institutional environment into which they are introducing it very poorly. That is not a minor implementation problem. It can determine whether the intervention survives. From funding gaps to investable systems But the next question should not always be how much aid is required. It should be what would make this investable and sustainable. A donor can fund a youth employment programme for 100,000 young people. Or it can help build the institutional, financial and market infrastructure that enables millions of young people to find or create productive employment without depending indefinitely on donor funding. The first approach funds an intervention. The second builds an ecosystem. That is the pivot development finance needs to make. The same principle applies to humanitarian action. This is also why I have increasingly argued for a shift from an aid mindset towards what I call ‘Investment governance’. It is not about turning humanitarian organisations into investment funds. It is about asking better questions about capital: where it should go, what risk it is taking, what capability it is creating and what happens after the grant ends. It also means asking whether public resources can eventually finance the intervention, whether private or diaspora capital can participate, whether the intervention can generate economic activity rather than permanent dependence, and whether today’s expenditure is actually reducing tomorrow’s need for external assistance. These questions become particularly important as humanitarian crises become protracted. A refugee may spend 20 years in displacement. A community may live through repeated cycles of drought, conflict or economic shocks. At some point, providing the same emergency intervention year after year becomes less a temporary response and more an alternative operating system. That is when we need to ask a different question. Not simply how we fund the next response, but what it would take to make the next response less necessary. The wider opportunity This shift also changes how we think about Africa’s development challenges. Too often, the conversation starts with needs and ends with funding gaps. Africa needs more jobs. The continent needs better infrastructure. It needs more agricultural productivity. It needs better digital skills and more access to finance. All of that is true. If a refugee has a viable business but lacks working capital, the answer may not always be another training programme. It may be finance. If a community has an agricultural market but lacks infrastructure, the answer may not be another livelihood project. It may be investment in the infrastructure and institutions that make the market work. If a government has identified sectors capable of creating millions of jobs, the question is not simply how much donor money can be allocated to employment programming. The question is what needs to happen between government ambition and investable opportunity. That is where development capital can play a very different role. A different definition of success Perhaps this is ultimately where the debate needs to land. The objective of the aid ecosystem should not be to become better at delivering externally financed projects. It should be to become better at creating the conditions in which countries, institutions, markets and communities can solve more of their own problems. That does not make external assistance less valuable. It makes it more strategic. A donor dollar that delivers a service today has value. A donor dollar that helps an institution deliver that service for the next twenty years has a different kind of value. And a donor dollar that helps create the financial and institutional conditions in which that service eventually no longer requires donor funding has greater value still. That is not an argument for abandoning delivery. It is an argument for thinking beyond it. It also gives us a different way of thinking about scale. Scale is not simply reaching more people. It is increasing the number of people who can be reached sustainably by institutions capable of reaching them. That changes the question we ask of every intervention. Instead of focusing only on how many people were reached, we should also look at what became stronger because the intervention happened, whether institutions are more capable, whether local organisations are more resilient, whether government functions have improved, whether markets are working better and whether the financing mechanisms created can continue without external support. And ultimately, what remains when the programme ends? This also changes what success should mean for INGOs and local NGOs. Their value should not be measured only by how much funding they can mobilise or how many people they can reach. Increasingly, it should also be measured by the capability they leave behind and the extent to which they strengthen the institutions that will ultimately carry the solution. For donors, this means something equally important. If sustainable systems are genuinely the objective, then financing needs to reward the behaviours that create them. This is not about blaming donors. It is about aligning incentives with outcomes. READ ALSO: The aid ecosystem already knows how to deliver. It has become increasingly sophisticated at demonstrating evidence, measuring reach and scaling interventions. The next challenge is harder. It is learning how to build institutions, markets and financing systems capable of carrying those solutions long after the external funding has gone. That requires different skills, different incentives and, in some cases, a different conception of what the role of an aid organisation actually is. The right idea is to find what works and take it further. But taking it further cannot simply mean putting more money behind the same delivery model. The real test of scale is whether the system can carry it. If it can, we have not simply scaled a programme. We have built something that can last. And that, ultimately, is the kind of impact the entire aid ecosystem should be financing.