Don’t draw the wrong lesson from the stall on direct funding

For two decades, major funders have promised the same thing: money going directly to the civil society organisations closest to the people they serve. Donors met in Paris in 2005 to fix how aid is delivered — too fragmented, too donor-driven, too little owned by the countries receiving it — and promised to put recipient countries and communities in the lead. In 2008, Accra recognised Southern civil society as a development actor. In 2011, Busan made them full partners. The Grand Bargain, agreed at the World Humanitarian Summit in Istanbul in 2016, went furthest, aiming to allocate at least a quarter of humanitarian funding to local and national organisations by 2020. The deadline passed, and the target was missed.

Now, the Humanitarian Reset names what the earlier commitments only talked around: power. It sets a higher ambition than the Grand Bargain, though still a limited one. But the direction has never been clearer. A few funders have moved boldly; most have moved incrementally, but power has moved the least of all.

The evidence bears this out. In its 2025 study of five major donors — Australia, Canada, the Netherlands, the UK and the US — Publish What You Fund found the Netherlands to be the best performer, directing 6.9 per cent of its aid directly to local organisations. But all the donors routinely fund organisations based in their own countries at much higher levels than local partners. USAID went furthest in intent, aiming to provide a quarter of its funding to local organisations by 2025, but this target was eclipsed by the agency’s closure.

Even the front-runners could channel only a small share directly, not for want of capable partners but because of the machinery around the money.

Syria showed the blockage plainly. When international agencies could not operate in opposition-held areas from the mid-2010s, Syrian organisations carried the response — yet the money could not reach them directly. It went through UN pooled funds instead. The risk of funds going astray in territory held by a designated terrorist group was real, and it was right to manage this. But the design was wrong. Donors’ rules met that risk by refusing to contract locally at all, when a body close enough to check those organisations could have managed the same risk and still funded them. The obstacle was never local competence — it was the machinery around the money. And machinery can be rebuilt.

It has been hard, and the wrong lesson is already spreading that direct funding is simply too hard, and we should go back to the way things were. Sweden and Norway have made some of the boldest attempts yet — moving to fund civil society in the Global South directly, rather than through international intermediaries. Recent reporting in Development Today has set out the practical difficulties. What makes the wrong lesson tempting is that established intermediaries in donors’ own countries already have compliance and risk systems in place, making funding through them seem easier and safer. But the danger was never that direct funding is difficult. It is that we quit and let localisation die as a result of a false diagnosis.

Some say we should simply fund local organisations directly. They are right about the goal, only half right about how to get there. Direct funding fails not because the ambition is wrong, but because it asks staff in donor capitals to judge hundreds of organisations they have never met, without the relationships, knowledge or authority to do so properly. That is no reflection on the people who tried. It is a verdict on the tool they were given. Remove the intermediary, and the work of screening and carrying risk does not vanish. It lands on the desk least equipped to hold it.

Some of this weight we pile on ourselves. The rules were written for billion-dollar agencies, and then they were applied to village organisations. BRAC is 50 years old and a leading responder to the Rohingya crisis. Yet even for us, registering for a single UN fund can take the better part of two months. Now, picture a small community group facing those levels of red tape. The pressure is rising as funders move to larger envelopes that cut their own admin costs but raise the bar for who can apply. While those concerns are real, they do not justify a single set of rules for every grant.

We do not need an intermediary for this part. We need rules that fit the size of the grant. Change those, and a small grant does its job: giving an organisation the track record it needs before it can handle larger funds directly.

But the rest of the risk is real, and someone has to hold it — to screen applicants, take the exposure, and answer when things go wrong. The only question is who. Give that job to a body sitting above the money, and you have rebuilt the intermediary under a new name. Give it to those closest to the ground, and you build something new: capacity that grows, and less money lost to overhead.

In Bangladesh, we have built that new thing more than once, never as a template, always alongside the donor, each version asking a harder question than the last. The first of these, the Climate Bridge Fund, supported by KfW, asked how to stop the body in the middle from picking its own winners. Later ones — the Pooled Fund for Localisation with Canada and Australia and the Bangladesh-America Maitree programme funded by the US — asked something harder: how does that body carry the risk without competing with the organisations it funds?

Now, Sweden, through the Catalyst for Change project with BRAC, is co-designing an intervention to fund the hardest reach of all: Indigenous communities in climate-vulnerable areas, so far from the system that they are cut off even from the public services they are entitled to. The aim is not to build a parallel system, but to connect people to what already exists — government schools, clinics, clean water — and add specialist support only where the state leaves a gap. What we are after is a community tied to its own government, not to us.

Let me be honest about where I stand, because I stand at both ends of this. BRAC holds the Sida Catalyst grant, at the very bottom of the ladder, building community groups so new that they cannot yet hold money at all. It also runs funds at the top, where local groups have outgrown us and now manage their own. I have climbed both ends, which is why the next distinction matters to me more than to most.

An intermediary is not localisation; a strong body in the middle is the very thing localisation was meant to end. A backstop is different: it is a capable national body that takes on the donor's screening, risk and compliance so local groups can be funded directly. So the question is not whether a backstop works — ours do. What separates it from an ordinary intermediary is not cost, proximity, or good intentions; an intermediary can claim all three. It is who controls whether it survives: an intermediary holds that control and never leaves; a backstop gives it away and is built to make itself unnecessary.

It does this in three ways. First, it does not run projects itself, so it never competes with the organisations it funds, and the money it has can only go to them. Second, where it does compete, it does not decide the outcome: an independent panel of experts and government officials judges the proposals. We cap how much of a fund BRAC can win, and the panel can give us less when another proposal is better.

Third, and hardest: a backstop must be built to shrink until the organisation it supports no longer needs it. I know the endpoint is real because BRAC has all but reached it. We began in 1972 as a small relief effort in one remote corner of newly independent Bangladesh. Fifty years of direct, flexible funding and devolved trust, cited by the OECD as an example of locally led development working, have made us one of the largest NGOs in the world, now covering most of our own costs. There is no ceiling on what a local organisation can build when it is trusted with real money and real decisions.

That is why we do this. The backstops I have described are how we try to pass the same trust on to organisations smaller and closer to the ground than we are, and to keep passing it on until they, too, no longer need us. And it is already happening. In our largest such fund, the Pooled Fund for Localisation, an organisation graduates not when we decide it is ready, but when it has raised and delivered donor money on its own — proof that sits outside our opinion, in another funder's decision to back it. Of the 27 organisations we have worked with, four have left the funds entirely and now hold their funding directly; sixteen more have started winning direct grants but have not yet exited. A small number, and I will not pretend otherwise — but a model that has walked even a few out the door is a different thing from one that only promises to.

Underneath it all is one plain rule. Holding the money and deciding how it is spent are two different jobs, and good design keeps them apart. The backstop can hold the money and carry the risk; it should not hold the choice. That belongs to the community the money is meant to serve.

None of this happens overnight — which is exactly why the design has to keep moving, so that "not overnight" never quietly becomes "not ever". A local backstop is real progress on a distant one: the money, the knowledge and the decisions all sit closer to the crisis. But it progresses only for as long as it shrinks. A backstop that shrinks is a bridge; one that stays is the old intermediary with a local face.

The choice was never between direct funding and intermediaries, or between fast and safe. When a tool fails, you build a better one; you do not throw away the goal. Backstopping is one better tool, and it has already begun to graduate the organisations it serves. It is not the only one — there must be more, in other countries and sectors, waiting to be found. Throwing out direct funding because the first tools were badly built is the wrong lesson. The right one is to find the tools that work.

_____________________________________________________

KAM Morshed is Senior Director leading advocacy, communication, and engagement at BRAC.