How the world's governments fund development — and what broke in 2025.
The world-scale companion to Public vs Private, which compares US federal spending to US private giving. This page follows government money across borders: foreign aid, the UN system, the Global Fund — and where philanthropy fits.
Official development assistance (ODA) — the aid that rich-world governments send to poorer countries — nearly tripled in real terms over two decades: from ~$83B in 2000 to a peak of ~$232B in 2023 (constant 2024 dollars; ~$227B in then-current dollars, the basis the data explorer uses). But the last leg of that climb flatters the trend: the COVID response added ~15% (2019–21), then Ukraine support plus refugee costs — money donor countries spend hosting refugees at home counts as ODA — added another ~20% (2021–23). Then it broke. Aid fell 6% in 2024, and in 2025 it collapsed 23% in real terms, to ~$163B — what the OECD calls the largest annual contraction ever recorded, taking aid back to roughly its 2015 level, with a further ~6% decline projected for 2026.
For scale: even at its peak, all rich-world aid combined was about a quarter of US non-defense discretionary spending — and roughly 13× all institutional philanthropy for development. Aid is a government instrument first; everything else operates in its shadow.
The contraction wasn't evenly shared: 26 of 34 donor countries cut, but the five largest accounted for 96% of the fall, and the United States alone drove three-quarters of it. US aid fell 57% in real terms to ~$29B (0.09% of national income), dropping the US from ~31% of all rich-world aid in 2024 to ~17% — and making Germany the world's largest donor for the first time, by a whisker, while also cutting 17% itself.
Behind the US line is an institutional story, not just a budget cut. A January 2025 executive order froze nearly all foreign assistance; by March, 83% of USAID's programs had been canceled; on July 1, 2025, USAID ceased operations and its remaining programs were absorbed into the State Department. Total US foreign-assistance obligations fell from $85.8B in FY2024 to ~$46B in FY2025 — roughly $40B withdrawn in a single year (obligations data still partially reported and may revise up). That's the gap the Public vs Private page invokes: several times the annual payout of the entire Gates Foundation, gone in twelve months.
Flip the ledger to the receiving side and the same pool looks different. Across all developing countries, net official aid came to ~$256B in 2024 (counting both bilateral donors and the multilateral agencies' own outflows) — but $70B of it (27%) never lands in any one country's column: in-donor refugee costs, administration, and regional programs the OECD can't allocate country-by-country. Of the money that does land, Ukraine is the largest recipient by far — $35.8B in 2024, ~$947 per person, 16× its pre-war level — while the deepest dependence is elsewhere: aid is equivalent to 28% of national income in Burundi, 21% in Somalia, 19% in the Central African Republic.
Take Malawi — the kind of country the 2025 collapse hits hardest. It received $1.6B in 2024: $74 per person, 15% of national income. Its funder list is this page in miniature: IDA ($506M) and the United States ($399M — a quarter of everything Malawi receives) tower over the rest; the Global Fund ($226M) and EU institutions ($109M) follow; and every OECD-tracked private foundation combined — Gates included at $23M — adds up to $38M, about 2.4%. One caveat on what the OECD can see: direct-delivery NGOs funded by individual donors don't report to it at all — GiveDirectly alone has delivered $140M+ in cash transfers in Malawi since 2019 (its own cumulative figure), a pace comparable to Gates' annual line there — but even doubling the private column doesn't change the shape. When a donor the size of the US walks away, there is no private line on the list remotely sized to replace it.
A large share of government aid doesn't go country-to-country — it routes through multilaterals: the UN system ($68.3B in total revenue across 48 entities in 2024, still below its $74B 2022 peak), the World Bank's concessional arm IDA (~$28B/yr disbursed), and vertical funds like the Global Fund (~$4.8B/yr) and Gavi (~$3B/yr). The crucial point: these are channels, not new money. They are funded by the same donor governments — so when bilateral aid collapses, the multilateral layer starves on a lag, as pledges come up short and arrears pile up.
Two things jump out of the flows. First, it's still overwhelmingly government money — direct government contributions plus EU institutions are ~73% of the system's revenue, so the 2025 bilateral collapse propagates straight through. Second, philanthropy & private money is just $4.5B (~7%) across the entire UN system — and even that is concentrated: UNICEF's National Committee network raises $1.9B of it, and the Gates Foundation's $528M makes it the system's largest private institutional funder — $325M to WHO, $116M to UNFPA (whose US grants were about to be terminated), ~$19M each to UNDP and WFP, and $14M to FAO. (Gates gives more to the vertical funds outside the UN proper: ~$300M/yr to the Global Fund and ~$320M/yr to Gavi via its replenishment pledges.) Note FAO's profile: $2.3B, over half of it from assessed dues and EU/vertical-fund transfers — the quiet, treaty-based layer of the system that rarely makes headlines but runs on the same shrinking government base.
Behind those bars is one dominant fact: the United States stopped paying. WFP — the system's largest agency — lost a third of its funding in a single year as US contributions halved from $4.4B to $2.1B. The US had been UNHCR's top donor at $2.1B; it paid $0.8B in 2025 (−60%). IOM, ~40% US-funded, lost a quarter of its revenue and began cutting programs touching 6,000+ staff. UNFPA had all 48 of its US grants (~$377M) terminated. UNRWA hasn't received a US dollar since January 2024. And the UN core itself: the US paid nothing toward the 2025 regular budget (its ~$820M assessment), pushing total US arrears to ~$4.5B and forcing peacekeeping to repatriate a quarter of its uniformed personnel. Summed across all humanitarian operations, reported funding fell 31% — $32.7B to $22.6B — and the UN's 2025 appeal, cut mid-year from $47B to a "hyper-prioritized" $29B, ultimately received about $12B: the lowest in a decade, reaching 25 million fewer people than 2024.
The forward-looking funds fared little better. The Global Fund's 8th replenishment raised $11.3B at its November 2025 summit against an $18B target. Gavi's 2026–2030 round secured over $9B against an $11.9B ask — with the US announcing it would stop contributing entirely. The WHO lost its historically largest funder when the US withdrawal took effect in January 2026, forcing a 22% cut to its base budget (from a proposed $5.3B to $4.2B for 2026–27) even as member states raised their dues 20% to compensate. Only IDA entered the storm well-capitalized: its record 2024 replenishment turned $23.7B in donor pledges into $100B of financing capacity through mid-2028.
Institutional philanthropy directed toward development runs about $17B a year (OECD's census of 506 foundations, 2020–23) — roughly 10% of government aid over the same window, with ~$13B of it actually crossing borders. It is intensely concentrated: the Gates Foundation alone is 28% of all tracked philanthropy for development — it would rank as the world's 8th-largest donor if it were a country — and it provides over half of all philanthropic health funding. Health takes 40% of the philanthropic total, a far deeper tilt than any government portfolio.
Then 2025 inverted the field. With US global-health spending down two-thirds, IHME's preliminary estimate is that the Gates Foundation surpassed the US government as the single largest funder of global health — the first time in the modern era a private foundation tops the table. The same story at the WHO: long its largest non-state funder, Gates became its largest contributor outright once the US left. Gates pledged $912M to the Global Fund's 2025 round and $1.6B to Gavi's — while announcing it will spend ~$200B over 20 years and close entirely by 2045, with ~70% of its now-$9B annual budget going to global health.
But being the biggest funder in a shrinking room is a grim kind of promotion. Philanthropy's number held constant in 2025 while government funding collapsed around it — its share rose because the denominator fell.
That's the bottom line: even if every philanthropic development dollar on Earth were redirected to the UN's humanitarian appeal — abandoning global health, vaccines, research, everything else — it would cover barely half the gap in that one slice of need. Redirecting Gates entirely would cover a quarter of it. The money that left is government-sized, and only government money comes in that size.
The human terms are harder to bottom-line, but the best-measured slice gives the shape: development assistance for health fell from $49.6B in 2024 to $39.1B in 2025, its lowest level in over 15 years and less than half its pandemic peak, with IHME forecasting a further slide toward ~$36B by 2030 under current policies. The US share of global health funding fell from over 25% — where it had been every year since 1990 — to 12%. One peer-reviewed projection in The Lancet estimates that if the cuts persist, they could mean more than 14 million additional deaths by 2030, over 4.5 million of them children under five — a modelled counterfactual, not an observed toll, but a sense of the stakes.
The through-line from Public vs Private holds at world scale, only sharper: philanthropy is a junior partner in aggregate — ~10% of the pool — and pivotal at the margin, in the specific programs where a billion dollars of flexible private money preserves what a retreating government abandons. In 2025 the margin got very wide, very fast. Choosing those marginal bets well is the whole game — which is the effectiveness question. But before asking whether the dollars work, follow the private money home: before a dollar reaches any field, it usually parks in a foundation or DAF first. That's the next chapter — the Intermediaries.
Chart the aid contraction donor-by-donor in the Explorer, or step back to the US public-vs-private comparison.