Donor Concentration, and the Test It Quietly Threatens

Most charities cannot say what share of their money comes from their largest few donors, because the answer is spread across a database nobody queries that way. The tax returns can say it, for a large part of the sector, and the answer is that concentration is normal and that a significant minority are concentrated to the point of risking their own status.

Of 119,516 charities sitting the public support test that asks about donations, 46,263 (38.7%) report support above the 2% line: at least one donor whose five year giving exceeded 2% of everything they raised. For the median one, 16.5% of five year public support sat above that line. For 5,786 of them, more than half did.

The figures are from the IRS annual extract of Form 990 returns, and the analysis script is published with the method.

What the 2% line is, and why it can measure this

A public charity has to prove it is publicly supported. The commonest route, section 170(b)(1)(A)(vi), tests whether at least a third of its total support over five years came from the public. The arithmetic has a deliberate feature: when the charity adds up its public support, it may only count what each individual donor gave up to 2% of total support. Everything above that is stripped out before the test is applied.

That stripped out amount is reported on its own line, Schedule A Part II line 5, and it is the only field in the public data that sees donor concentration at all. It is not a donor count, and it cannot become one. A charity funded equally by a thousand people reports zero. A charity where three people funded half the budget reports a large number. So the field answers how much money sat above the line, never how few people it came from.

Two properties make it better than it sounds. It runs over five years, which is the test’s own window, so a single exceptional gift does not distort it. And it is reported by every organization sitting that test, which is 46.6% of charities filing a full return, so the base is large.

How concentrated the sector is

Two opposite movements, and both matter.

Larger organizations are more likely to have a donor above the line, rising from 27.3% under $250,000 to 49.3% over $25m. That is mostly arithmetic rather than behaviour: 2% of a large support base is a large gift, but large organizations receive large gifts, and they receive more of them.

Smaller organizations are more concentrated when they are concentrated at all. The median share above the line falls from 23.6% in the smallest band to 10.7% in the largest, and the share where more than half of support sits above the line falls from 5.8% to 2.4%. A $200,000 charity with one $60,000 donor is in a materially different position from a $30m charity with a $4m donor, and the table says the first is the commoner shape.

The finding that should change something

Concentration is usually discussed as a fundraising risk. In this data it is also a status risk, and the two arrive together.

The public support percentage in the fourth column is the test itself, and it has to stay above 33.3%. Read down that column: a charity with no donor above the line reports 99.6%, and the number falls in step with concentration until the most concentrated group sits at a median of 34.9%, which is 1.6 points of headroom.

65.9% of those 5,786 charities are within 7 points of failing the test. That is not a coincidence or a reporting artifact, it is the mechanism: the same dollars that make an organization concentrated are the dollars the test refuses to count. Failing it two years running means being reclassified as a private foundation, with an excise tax on investment income, distribution requirements, tighter rules on self dealing, and a deduction limit that is worse for the donors who are funding it. The difference between the categories is set out in charity against foundation.

So for one charity in twenty sitting this test, the largest relationships are simultaneously the budget and the thing most likely to end the organization’s tax status. Both halves of that sentence rest on the same handful of people.

The objection, tested: are concentrated charities just weaker?

The obvious challenge to everything above is that concentration might be a symptom rather than a cause. Perhaps the organizations sitting on one big donor are simply the struggling ones, and the support test is picking up general weakness. The extract can test that directly, by crossing this analysis with the deficit and cash reserve work published earlier on this site.

The answer is no, and it is worth saying clearly because it cuts against the argument this page is making. Concentrated charities are not running more deficits: the rate is flat at about 41% across every band, which is the sector rate reported in the deficit benchmarks. And they hold more cash, not less: the median rises from 4.9 months where nothing sits above the line to 6.7 months where more than half does.

That makes sense once stated. A large gift arrives as money, and some of it is still on the balance sheet. So the risk of concentration is not that these organizations are poorer or thinner. It is narrower and more specific: their public charity status is tied to a few relationships, and the cushion they are holding is itself a product of the gifts creating the exposure.

What this does not say

Three limits, and the first is the one most likely to be misread.

It is not a donor count. A 40% figure does not mean 40% of money came from one person. It means 40% of public support sat above a threshold that several donors may have crossed. The direction is reliable, the granularity is not.

Governments and other public charities are excluded from the line. A grant from a city or a community foundation is not subject to the 2% cap, so an organization funded by three government contracts can report nothing above the line while being extremely concentrated in the ordinary sense. This measure sees concentration among individual and corporate donors only.

Whole categories of charity never sit this test. Churches, schools and hospitals qualify under their own provisions, and 97,067 organizations sit the 509(a)(2) earned income test instead. Among those, 13.2% report amounts from disqualified persons and 8.0% report amounts above their own 1% line, which is the same idea with different arithmetic.

What to do about it

The useful response is not to refuse large gifts. It is to know the number, and to treat the relationships behind it as infrastructure rather than as a happy accident.

• Read line 5 of your own Schedule A and divide it by line 4. That is the figure in the tables above, for your organization, and it takes a minute with last year’s return in front of you. The Form 990 guide covers where it sits.
• Watch the public support percentage on line 6, not just the dollar total. It is on a five year rolling window, so a large gift keeps affecting it for five years and the trouble arrives slowly enough to be managed if anybody is looking.
• Know who those relationships belong to. When the money is concentrated, the knowledge about it usually is too, and it is usually held by one person. What happens when that person leaves is the subject of the companion page on fundraiser turnover.
• Build the second tier deliberately. Concentration falls when the tier below the top gives more, not when the top gives less, and that is a moves management problem with a three year horizon.
• Hold more cash if you are concentrated. A concentrated income line and thin reserves is the combination that ends organizations, and the cash reserve benchmarks show how thin the sector’s are.

Method and limits

The data is the IRS SOI annual extract for processing year 2024, restricted to 501(c)(3) organizations with at least $25,000 of total expenses, which gives 256,539 returns. Of those, 119,516 report figures for the 170(b)(1)(A)(vi) public support test on Schedule A Part II. Support above the 2% line is line 5, the public support subtotal is line 4, the public support percentage is line 6 over line 11, and the 509(a)(2) figures are Part III lines 7a and 7b.

Four limits. Organizations filing Form 990-EZ or 990-N are absent. 84 returns report a line 5 larger than line 4, which is impossible for a subtotal of the same contributions, and they are excluded and counted rather than blended into the medians. Schedule A is a five year window, so the figures describe a period and not a year. And the concentration measured here is among donors subject to the 2% cap, which excludes government and other public charity support entirely.