In March 2025 I suggested the often-heralded end of organised aid from the members of the OECD Development Assistance Committee (DAC) was perhaps actually in view. My deliberately conservative forecast at that time had DAC Official Development Assistance (ODA) falling at least 25% by 2027 from its 2023 peak; a worse case was also envisaged, in which between a third and a half of OECD-reported aid would disappear within several years, courtesy of planned reductions by a sub-group of ODA donors (hereafter, the “cutters”).
The OECD’s preliminary 2025 data release of 9 April 2026 confirms the worse case is now upon us. On the OECD’s grant-equivalent headline basis, total DAC ODA (from the 32 standard members plus associate member Romania) fell from a US$229 billion peak in 2023 to US$165 billion in 2025, in constant 2024 prices — a 28% real-terms decline in two years. Within the 50-year net-ODA series in Figure 1, and fact since the advent of ODA, the 2023-2025 fall is by far the largest two-year contraction, and the 2024-2025 single-year fall of US$49.4 billion (23.3%) is also a record. The cutters’ 37% decline which I had previously forecast to take place over 2023-2027 was already 34% by 2025: what I had expected to take four years happened in two.
Some of this is the unwinding of exceptional increases: in-donor refugee costs after February 2022, COVID-19 vaccine donations and bilateral aid to Ukraine. However, the OECD’s own decomposition assigns the largest component of the 2025 fall to core donor-budget contraction rather than to exceptional factors.Since 2024, ten DAC members have announced or legislated multi-year cuts: the US, Germany, the UK, France, the Netherlands, Canada, Belgium, Sweden, Finland and Switzerland. Their combined ODA fell from US$173 billion in 2023 to US$113 billion in 2025 — the 34% real-terms fall mentioned above. The US accounted for roughly three-quarters of the 2024-2025 reduction, dropping from US$65 billion to US$28 billion under the second Trump administration’s aid restructuring. Most European cutters tracked their announced paths (UK -10.8%, Netherlands -4.9, Switzerland -6.9, Finland -5.7, Canada -2); Germany (-17.4), France (-10.9) and Belgium (-22) cut harder than their announced trajectories implied. The non-cutters did not deliver the inflation-linked offset I had forecast: they fell 8%.
Some non-negligible donors are maintaining or growing their aid: Italy is flat at US$6.8 billion on migration-management grounds via the Mattei Plan; Spain is up 10.7% under its VI Plan Director; Norway is up 1.7%, and in 2025 was the only DAC donor still above 1% of gross national income (GNI); Denmark, Luxembourg and Ireland are on modest growth paths.
Figure 2 shows two forecasts, one the OECD’s and the other my own, beyond the DAC actual outcome figures for both 2024 and 2025. Two things stand out in relation to last year’s forecasts. First, my deliberately conservative forecast (the top edge of the pink envelope terminating in 2027), representing the best (least-cut) case, was far too conservative. My central estimate was closer to the mark: US$166.3 billion against an actual of US$165.5 billion. Second, the OECD’s June 2025 Cuts in ODA brief midpoint forecast sat near US$184 billion through 2027, US$19 billion above the 2025 actual; even its higher-cut scenario overshot by US$11 billion. The OECD’s April 2026 outlook, built on the same stated-plans method, is therefore more realistically read as a best case than a central case.As for my 2026-2028 forecasts, for each donor I made low, central and high estimates based on a range of policy documents, media reports and other public sources. The central projection for the 33 country donors is US$150 billion in 2026, US$139 billion in 2027 and US$136 billion in 2028 in constant 2024 prices — falls of about 34, 39 and 40% relative to the 2023 peak. The low case takes 2028 to US$119 billion (-48%); the high holds it at US$155 billion (-32%). In other words, these forecasts firmly point toward what I last year saw as a pessimistic outcome of between one-third and one-half less ODA from OECD sources.
Figure 3 separates realised 2023-2025 falls from projected 2025-2028 falls. The first phase was primarily a US event: the US fell by US$38.0 billion, Germany US$12.0 billion, no other donor more than US$5 billion. The next phase spreads the cutting load. By 2028 Germany falls to US$21.5 billion, the US to US$19.5 billion in the central case, the UK to US$11.5 billion — the lowest in cash terms since 2012 — and the Netherlands to US$4.0 billion as its EUR 2.4 billion structural cut begins in 2027. Spain and Italy, on the other hand, add about US$3 billion between them. With total US falls of about US$47 billion, Germany is likely to be the largest single DAC donor for the next several years, even with its own cuts. Figure 4 shows the contrast between the ten “programmatic” cutters — the donors with announced intentions to reduce aid over the coming years — and the rest of the DAC membership. The cutters provided US$172.5 billion, or 75.3% of DAC country ODA, in 2023; on the central projection they fall to US$83.3 billion by 2028 — a 52% decline. (Japan is shown separately because it has no announced cutting agenda but declines in real terms anyway, from US$18.8 billion to US$13.8 billion.) The remaining DAC countries are broadly flat at around US$38 billion; EU institutions, not shown here, are also broadly stable at around US$27 billion. So there is no net offset within the wider DAC republic.Nineteen non-DAC aid providers also report to the OECD through the Creditor Reporting System, and the OECD gives estimates for China, India, Argentina and South Africa. Reported non-DAC ODA was approximately US$16.5 billion in 2022, up from US$10.9 billion in 2015 — a 50% real increase from a small base (these data are compiled with quite a lag). Even doubling that base would be modest against the realised US$64 billion DAC contraction. In addition, the categories most exposed to the DAC fall — humanitarian aid, low-income country grants and multilateral core funding — are not where non-DAC growth is concentrated.
Uncertainties are mostly on the downside. Germany’s 2027 budget could cut BMZ more deeply than current plans suggest; the US administration could even more aggressively hold back spending where Congress appropriates funds against its wishes; the Netherlands’ 2027 structural cut might be front-loaded. On the upside, continued growth in Spain and Italy, or better-than-expected performance from Japan, would lift the high case. A few specific signals will help to indicate which case is closer: Germany’s 2027 budget vote (if BMZ holds at EUR 9.5 billion or higher, the central is too pessimistic); US fiscal year 2027 appropriations and FY26 disbursement speed; and the Netherlands’ 2027 budget.
Traders in the financial markets talk of “support” and “resistance” levels that tend to constrain the fluctuations of stock and other prices. Given aid’s precipitous fall since 2023, the question now is whether there is any non-zero support level for it. If one simply draws a trend line through long-term aid levels, aid might well have bottomed out in 2025 — that is effectively what the OECD’s most recent forecast says. However, there are many strong indicators that, if there is a natural support level, we have not yet reached it, and it could be as low as the levels that prevailed in the late 1990s or early 2000s. That time was the epilogue of the Cold War; this time looks like the prologue of something like the opposite.
Technical note. Figure 1 uses cash-flow/net ODA for the 1970-2025 long-run comparison; all other figures are on the OECD’s grant-equivalent basis (the headline measure since 2018) and in constant 2024 US dollars, sourced from the OECD SDMX DAC1 dataflow as updated for the 9 April 2026 release. The widely reported US$174.3 billion 2025 headline is expressed in current US dollars. Japan is excluded from the cutters group because OECD attributes its 2025 fall mainly to bilateral program dynamics rather than an aid reduction policy. The method used for 2026-2028 forecasts is based on informed judgement, applied donor by donor; nominal local-currency growth paths are converted to constant 2024 USD using OECD Economic Outlook GDP deflators and 2024 base-year exchange rates.
