Where Does the Money Go?

The pie keeps growing

Americans gave a record $617 billion in 2025, up 5.7% on the year. By the headline number, U.S. philanthropy has never been bigger.

But relative to the economy it has barely moved: charitable giving has hovered around 2% of GDP for four decades. The pie grows roughly with the economy — steadily, but not dramatically. The interesting story isn't the size of the pie. It's who's holding the knife.

Who gives is changing

For most of the 20th century, American giving was diffuse: tens of millions of households writing checks. That's still the bulk of it — individuals are ~64% of giving. But the composition is shifting toward concentrated, institutional vehicles.

Foundation grantmaking has grown nearly 5× since 2000 ($24B → $117B). Donor-advised funds went from a rounding error to $65B in grants and $90B in new contributions a year — a parallel philanthropic system with no mandatory payout. Together, foundations and DAFs now route close to a third of every charitable dollar, up from roughly a tenth in 2000.

A new generation of major funders

So who are these institutions? A generation ago, big philanthropy meant Ford, Rockefeller, Carnegie. Today's largest funders are newer fortunes — and they keep arriving in waves.

Tech wealth built the modern giants: the Gates Foundation, Good Ventures (Dustin Moskovitz and Cari Tuna — the donor behind Coefficient Giving), and the Chan Zuckerberg Initiative. Finance and media added Bloomberg Philanthropies and George Soros's Open Society. And now a pharma wave is reshaping the top of the list: the GLP-1 drug boom — Ozempic, Wegovy, Mounjaro — has minted vast fortunes and inflated the foundations that hold the stock. Lilly Endowment, lifted by Eli Lilly's GLP-1-driven surge, overtook Gates in 2024 to become the largest U.S. foundation (~$106B in assets); the Novo Nordisk Foundation, behind Ozempic and Wegovy, is now among the largest charitable foundations in the world.

Alongside the institutions, a new kind of individual mega-donor has arrived. MacKenzie Scott has given away more than $25 billion since 2020 — and in 2025 her roughly $7B in gifts outpaced the entire Gates Foundation, with almost no institution behind her.

New money moves differently

The newcomers don't just have more money — many give it away on different terms. The traditional model is an endowment in perpetuity: pay out the ~5% legal minimum each year, invest the rest, last forever. MacKenzie Scott is the opposite — trust-based, unrestricted, and fast: huge lump sums, no application process, minimal staff, recipients often simply notified. She is deliberately giving the fortune away within her lifetime, not building a dynasty.

And she's not alone in the rush to spend. In 2025 Bill Gates said the Gates Foundation will give away virtually all its wealth and close by 2045 rather than run forever; Good Ventures has long planned to spend down in its founders' lifetimes. The terms turn out to matter as much as the totals: a dollar given freely and quickly does very different work than a dollar parked in a perpetual endowment at a 5% drip — the tension the rest of this vertical digs into.

And the biggest wave is still forming

If GLP-1 money and MacKenzie Scott are already here, the largest wave is still on its way. AI wealth is creating a new tier of mega-philanthropy almost overnight: the OpenAI Foundation (a ~26% stake worth ~$130B+ even at OpenAI's $500B recapitalization valuation — on paper the best-resourced philanthropy ever, larger than Novo or Gates — with $25B committed), Anthropic founders (~$90B pledged), employee donor-advised funds (~$60B), and Good Ventures' own Anthropic stake — together a $280B+ pool that didn't exist a few years ago, and closer to $370B at the Forecast's base-case valuations. But almost none of it has been deployed yet: the OpenAI Foundation has actually granted only tens of millions against its $25B pledge.

Almost all of it will flow through exactly the vehicles already gaining share: private foundations and DAFs. Our Forecast models how fast that new capital reaches the field — the single biggest swing factor is Anthropic's valuation.

A handful of decisions, a third of the money

Put the trends together and the arithmetic is striking. By 2030, major institutional funders — foundations, DAFs, and the new AI mega-philanthropies — are on track to direct on the order of $330 billion of ~$810B in annual U.S. giving — about 40%, with the largest funders growing fastest of all. And the AI wave does something U.S. giving hasn't done in 40 years: it lifts the total above its ~2%-of-GDP ceiling, to ~2.2%. Absent the AI wave, giving stays flat at 2% — the concentration still happens, just within a fixed share of the economy.

That means an unprecedented share of charitable capital will be steered by a small number of institutions and decision-makers — not tens of millions of households, but dozens of program teams. Concentration of giving is following concentration of wealth: the top 1% now hold ~$50T, and their philanthropy increasingly runs through professionalized vehicles.

Concentration isn't the same as deployment

Here's what the headline trend hides: a large share of the money flowing into these institutions doesn't go back out — it accumulates. US private foundations need only pay out about 5% of assets a year, and payout practice is a genuine mix: a few giants spend well above the floor, a large cluster sits within a point of it, and some sit below. What's nearly universal is that endowment returns outrun the payout — so the money compounds either way.

The starkest cases are the largest. The Novo Nordisk Foundation — the world's biggest at ~$148B — pays out barely ~1% a year (it holds a controlling Novo Nordisk stake and reinvests), and Lilly Endowment ~3.6% as its Eli-Lilly-stock surge outran its grantmaking. The new AI tier takes it to the limit: the OpenAI Foundation controls ~$130B+ of equity yet has granted only tens of millions so far. Concentrated capital is piling up faster than it's deployed.

It's not just who gives — it's what's funded

The same diffuse-to-institutional shift shows up in where the money lands. Over 25 years, giving to religion — the most everyday, small-donor cause there is — fell from ~38% of all U.S. giving to under a quarter. It's still the single largest category, but in steady structural decline.

The gains went to more secular, professionalized causes: human services, international (the fastest-growing at ~7%/yr), and public-society benefit — the bucket that contains DAFs, United Way, and other intermediaries. So the rise of donor vehicles registers twice: once as a source of giving, and again here on the receiving end.

Big in philanthropy — small next to government

For all the concentration, this is a modest pool next to public spending — but you have to compare the right slice. Measuring giving against total federal outlays (~23% of GDP) is unfair: most of that is defense, Social Security, Medicare and interest, which philanthropy never funds. The honest yardstick is non-defense discretionary spending — the research, health, education and aid budgets philanthropy actually overlaps.

The balance flips field by field: government owns what only it can do at scale (basic science, space, health research — NIH alone outweighs all US foundation science giving by roughly 10×), philanthropy owns religion, the arts, and higher education, and is a junior partner everywhere the two overlap. So philanthropy's leverage isn't scale — it's the margin: funding what's politically hard, unproven, or newly abandoned, like the ~$40B hole the 2025 dismantling of USAID left behind. The full field-by-field comparison — who funds what, dollar for dollar — is on Public vs Private.

And the world's aid system just broke

The starkest version of that government-dependence story is playing out internationally, right now. Foreign aid from rich-world governments nearly tripled over two decades — then broke: down 6% in 2024, and in 2025 down 23% in real terms, the largest annual contraction ever recorded, with the United States cutting 57% and driving three-quarters of the global fall.

For the fields this site cares most about — global health and development — philanthropy is even more junior on the world stage than at home: all OECD-tracked institutional giving for development, from every foundation on Earth, is ~$17B a year against the ~$50B of government money that just left. And the multilateral system behind it — the UN agencies, the Global Fund, Gavi — is funded by those same retreating governments, so it starves on a lag. Who funds whom, which countries depend most, what happened to the UN system, and what philanthropy realistically can and cannot backfill: Global Aid takes the full tour.

Concentration raises the stakes on effectiveness

Now put the two halves of this story together. Because philanthropy is small next to government, its impact lives at the margin — funding what's neglected, unproven, or newly abandoned. Because it's concentrating, those marginal calls increasingly belong to a handful of institutions. Where those dollars land becomes one of the most consequential questions in philanthropy — and there's almost no systematic infrastructure for answering it: less than 0.2% of annual giving is allocated through systematic cost-effectiveness evaluation.

The charitable tax deduction costs the U.S. ~$70B/year in foregone revenue — a public subsidy for private giving, with almost no measurement of whether it's well-spent. As capital concentrates, that measurement vacuum gets more expensive. This vertical exists to start filling it.