It is easy for funders to agree with the language of localisation. It is much harder to change the habits, incentives and relationships through which research funding actually moves. That was, for me, the value of the final learning session of the African Education Research Funding Consortium. It was not a self-congratulatory conversation. Nor was it another discussion in which everyone repeated the right words about long-term support, local ownership and systems change. Instead, Dana Schmidt of Echidna Giving, Rupert Corbishley of the Aga Khan Foundation East Africa, and later Sarah Ruto of HERI Africa, spoke with unusual clarity about why these goals matter, why they remain difficult, and what may be needed to move from reflection to action.
Over the last few years, the consortium brought together funders around five broad recommendations:
- Support a long-term vision;
- Fund African organisations directly;
- Invest not only in research production but also in communication and use;
- Expand opportunities for women and other underrepresented groups; and
- Learn actively from practice.
None of these ideas is controversial. Most people working in the sector would endorse them. The more interesting question is what happens when funders try to put them into practice.
A conversation that began with the evidence
Dana Schmidt’s account of how the consortium began was revealing. It started with a close look at her own organisation’s portfolio. Echidna Giving reviewed where its money was going and found a clear contrast. Organisations delivering programmes were often based in the countries where they worked. But when it came to research, the pattern was very different: much of the funding went to organisations based elsewhere – in the Global North.
That mattered to her for a simple reason. Researchers who are proximate to the contexts they study are often better placed to identify the right questions, understand the politics and institutions around them, engage with policymakers and practitioners, and stay with the issue over time. So the problem was not only one of fairness or representation. It was also one of effectiveness.
Dana took this to be a field-building challenge. If funders wanted to change the flow of resources in any meaningful way, they would need to learn from one another, compare practices, and ask what it would take to interrupt a cycle in which Global North institutions continue to receive the largest grants, continue to build capacity, and therefore continue to appear as the safest organisations to fund.
That point should not be underestimated. Funding does not just support a research ecosystem; it shapes it. It influences who becomes visible, who grows stronger, who is trusted, and who gets to define what quality looks like.
Why learning is easier than change
Rupert Corbishley’s reflections helped explain why this kind of change remains difficult even when the diagnosis is widely shared. One of the great strengths of the consortium, he argued, was that it created a trusted space for open and honest conversations. That may sound like a modest achievement, but it is not. Too many collective spaces remain polite, vague or overly managed. What this consortium seems to have offered, instead, was a place where funders could speak more candidly about the barriers they face within their own institutions.
Rupert described those barriers in practical terms.
- Organisations tend to preserve the status quo.
- People struggle to find time.
- Portfolios are already committed.
- Learning can happen more easily than action.
He spoke, quite memorably, about the effort required to resist what he described through the language of entropy: the natural tendency of institutions to continue as they are unless significant energy is invested in changing them.
Dana made a similar point from another angle. Many funders talk about long-term change, but they often operate through short strategy cycles, fixed budgets and limited room to shift direction quickly. A funder may genuinely want to support stronger local institutions, but still find that most of its money is already tied up, or that change can only happen gradually as new opportunities emerge.
Local agendas, risk and the problem of control
The discussion then turned to local agendas. This is one of those ideas that sounds straightforward until one asks what it actually requires.
Dana’s answer was refreshingly direct: funders have agendas too. That may seem obvious, but it is not always acknowledged with enough honesty. No funder arrives without priorities, goals, or some idea of the change they want to see in the world. The challenge, then, is not to pretend that funders are neutral. It is to recognise that their agendas can easily dominate the relationship unless they are handled with care.
What she suggested, implicitly, was a more realistic and more useful standard. Funders should identify where their priorities overlap with locally defined needs, then support that work without trying to reshape it too heavily. The problem arises when the money comes with such strong direction that local actors are pushed away from what they themselves see as most relevant or necessary.
Rupert picked up this thread by focusing on control. Supporting local agendas, he suggested, requires funders to let go of some of that control. That is hard because control is tied to perceptions of risk, accountability, and quality. His example of more citizen-science approaches to research was useful here. It pointed to the possibility of generating knowledge in ways that are broader, more participatory and more grounded in what communities themselves consider important — but it also showed how quickly questions of trust and data quality come to the surface.
Both speakers, in different ways, were describing the same tension. Funders may support localisation in principle, but in practice, they still tend to rely on familiar institutions, methods, and oversight.
That is why risk matters so much in this conversation. As Dana noted, funders often perceive greater risk in backing institutions that may not have the same track record or visibility as those they have funded for years. But one of the important suggestions in her remarks was that risk could be understood differently. There is also risk in continuing to fund in ways that limit impact, weaken local ownership and reproduce dependence.
A changing context — and a different opportunity
The speakers also reflected on how much the wider context has changed in the last three years. Dana noted that government ownership has become even more important. In a period of shrinking aid and greater uncertainty, it is no longer plausible to assume that a strong idea will simply be picked up and scaled by some multilateral or bilateral actor. If educational change is to happen at scale, governments have to be part of the story.
Local agendas are not only about listening better to researchers or civil society. They are also about recognising that research must be connected to public systems if it is to matter in a deeper and more durable way.
This means philanthropic funders need to learn how best to work with governments and other partisan political actors.
Rupert acknowledged the real damage done by funding cuts and capital flight from the social sector, including education. But he also argued that this moment may force a more serious conversation about domestic responsibility and domestic resources. Referring to debates in Kenya, he suggested that there is growing recognition that countries have resources and that the question is how to mobilise and organise them more effectively.
The session did not deny the seriousness of the current crisis. But it did suggest that the weakening of one model of international support could create space for something more locally led and more institutionally grounded.
HERI Africa and the move from principles to institutions
That broader possibility came into focus when Sarah Ruto joined the discussion to speak about HERI Africa. If Dana and Rupert helped explain the problem, Sarah offered a glimpse of what a response might look like.
HERI Africa — Harnessing Education Research for Impact in Africa — follows the consortium and tries to put many of its recommendations into practice. As Sarah explained, the initiative has been deliberately ambitious. It brings together government institutions, universities and non-state actors. It is not simply asking how to fund more research. It is asking how to organise research leadership, agenda-setting, and institutional support in ways that are nationally owned and lasting.
The launch of HERI Africa is characterised by two key interventions: the creation of a country-driven research agenda and the institutionalisation of research chairs.
They point to a broader vision in which research is not treated as a collection of disconnected projects but as part of a longer-term system.
Sarah also stressed that government leadership matters because it can unlock domestic resources. In other words, this is not only about attracting more funding from outside Africa. It is also about creating mechanisms that enable public systems themselves to support and use research more seriously.
What this conversation leaves us with
Dana reminded us that the question of locally led research is not abstract. It emerges from real funding patterns that can be seen, measured and, if there is enough will, changed. Rupert reminded us that change depends not only on better ideas but on trusted relationships and on institutions willing to work against their own inertia. Sarah showed that the next step is not simply to keep discussing recommendations or principles, but to build structures that can carry them forward.
Taken together, their remarks point to something important. The problem is not that funders do not know what better practice looks like. It is that better practice asks more of them than the language of reform often admits. It asks them to rethink risk, loosen control, support institutions rather than just projects, and recognise that local leadership cannot be funded on purely external terms.
It is a different way of understanding what funding is for.