The 25% Promise, and the 3.8% Reality

In 2016, donor governments and the world's largest aid organisations signed the Grand Bargain, a set of commitments meant to reshape how humanitarian aid is delivered. Commitment 2 was specific: at least 25% of humanitarian funding would go to local and national responders, as directly as possible, by 2020.

That deadline has been and gone. According to ODI's most recent analysis of international humanitarian funding, published in 2026 using 2024 data, local and national actors received just under 10% of total humanitarian funding that year, combining direct and indirect channels. Direct funding alone, money reaching local organisations without passing through an international intermediary, was just 3.8%.

Perhaps more troubling than the shortfall itself is the trend. Funding to local and national actors shrank faster in 2024 than the overall humanitarian funding pool did. The gap between commitment and practice is not closing. It is widening.

Why does this keep happening, nearly a decade after the commitment was made? Not, in our experience, from a lack of goodwill. It happens because commitment on its own does not build trust, and trust is what actually unlocks funding. Donors extend flexible, direct funding to organisations whose governance, financial systems, and compliance capacity give them confidence to do so. Many local and national organisations are never given the support to build those systems in the first place, so the case for direct funding never gets made as strongly as it could be.

This is precisely the gap ImpactRoot works to close, not through another pledge, but through the practical work of building institutional capacity, advising the donors and institutions who hold the purse strings, and connecting the people on both sides of this equation who too rarely meet directly.

The numbers are sobering. They are also, in a sense, the clearest case for why this work matters right now.

Source: ODI, "The state of international humanitarian funding to local and national actors" (2026), using 2024 data.

Next
Next

Trust in Localisation, Redefined