1,165 IRS Form 990 filings · 107 schools disclosing aid · $942.4M in tuition assistance
A big endowment does not make a school generous. It makes its generosity survivable.
Five Jewish day schools hold nearly 40% of all reported endowment wealth. Everyone assumes that wealth is what lets them help families. Their own tax filings say the aid is close to identical everywhere. What the wealth buys is something else entirely.
39.1%of endowment held by five schools
0.93points separating their aid from everyone else's
43schools running aid with no endowment
$17.9Mof aid at one school with none
Scroll
The field is treating two claims as one.
Everyone in Jewish education knows the wealth is lopsided. A few flagship schools sit on endowments. Most live on tuition and this year's appeal. That part is not in dispute.
What people take from it feels equally obvious. If a handful of schools hold the money, those are the schools that can afford to say yes. Everyone else must be turning families away.
It is a reasonable inference. It is also a jump. Concentration is a fact about balance sheets. Generosity is a fact about behaviour. The first does not deliver the second.
Look at what schools actually hand over and the two-tier story falls apart. Which leaves a harder question standing where it stood. If the money is not buying generosity, what is it buying?
The same money, sent two different ways.
Two neighbours each wire three hundred dollars to a parent overseas on the first of the month. The parent receives identical amounts on identical days. On any ledger you could build, they are the same person.
One sends it from the interest on money saved a decade ago. The other sends it from that month's wages, and has not missed a month in years.
Then a plant closes. One neighbour keeps wiring. The other writes a letter.
Nothing about them looked different until the year went wrong. Generosity was never the variable. Durability was.
A funded commitment
Paid out of a reserve
Backed by money that was already there. A weak year moves the balance, not the transfer.
An unfunded commitment
Paid out of this month
Identical at the receiving end. Entirely dependent on the next paycheque arriving on time.
All of it is on the tax returns.
Both halves of this are public. Schools report endowment balances on Form 990. They report what they hand out in tuition assistance in the grant schedules of the same form. Across 1,165 filings, 533 aid records at 107 schools were pulled out (531 from structured fields rather than narrative text).
Two terms carry most of the weight in what follows.
Median
The middle value, not the average. With wealth this lopsided an average gets dragged upward by a few enormous funds and describes no actual school. Every figure here is a median unless it says otherwise.
Endowment payout
The slice of an endowment a school may spend in a given year, typically a few percent. It is what converts accumulated wealth into an operating cushion. A school with no endowment has no payout, and nothing standing between its aid budget and its fundraising results.
Start with the claim that survives.
Reported endowment, 90 schoolsConcentration
The claim that holds
The wealth really is that lopsided.
Of $484.7 million in reported endowment across 90 schools, five hold 39.1%. Ten hold 54.4% — more than the other eighty put together.
Nobody in the field will be surprised. Everything that follows asks what this picture does not cause.
The named few
One school holds a tenth of it.
Abraham Joshua Heschel reports $53.0M. Then Golda Och at $44.4M, Fuchs Mizrachi at $35.1M, Charles E. Smith at $35.0M, Perelman at $21.9M.
These are public filings. No school here is being accused of anything. They are the ones that reported.
The long tail
The middle of the field holds $1.8 million.
That is the median endowed school. Already the fortunate half, since it has an endowment at all. The largest fund is about 29 times it.
Not a gradient from rich to poor. A few tall bars and a long flat tail. The tail is the field.
Twelve Orthodox/Chareidi schools report any endowment activity at all. The rest report none. Not a small one. None.
Claim two
So the endowed schools must be the ones who can afford to say yes. Test it.
Concentration on its own is an accounting fact. It matters to a family only if it produces two tiers — schools that can discount deeply, and schools that turn people away. That is a claim about behaviour, and behaviour is measurable here.
The testable claim
Schools with the largest endowments should devote a larger share of revenue to tuition assistance than everyone else.
Tuition assistance as a share of revenueThe test
The result
They don't.
The ten most-endowed schools commit a median 18.47% of revenue to tuition assistance. Every other school commits 19.40%.
Under a percentage point apart, and pointing the wrong way. Whatever $484.7 million of endowment is doing in this field, it is not buying deeper discounts.
Break it out
The group with almost no endowment gives the most.
Orthodox/Chareidi schools hold 3.8% of the field's endowment and post the highest aid intensity of any group, at 24.69% of revenue.
Communal schools, holding 77.5% of the wealth, report 19.08%. Modern Orthodox, 18.84%.
The direction
More wealth, marginally less aid.
Plot the groups against each other and the line runs backwards. The two-tier story does not merely fail to appear. It appears inverted. (median award per recipient: $12,602)
So the intuition was wrong about what concentration produces. Not wrong that concentration matters. Wrong about what it changes.
The reframe
The divide is not generous against stingy. It is cushioned against exposed.
Two schools can commit the same share of revenue to tuition assistance and be in completely different positions. One covers part of it from endowment payout and can absorb a flat campaign without touching a single family's award. The other funds every dollar from this year's tuition and this year's donors. If either comes in short, the aid comes in short.
In a generosity comparison those two schools are identical. That is why the comparison keeps returning a null result, and why the field keeps being surprised by it. The instrument is pointed at the wrong quantity.
The question is not how much a school gives. It is what happens to that giving in a year that goes wrong.
Schools running aid with no endowmentThe split
The split
Two in five aid-giving schools have no endowment at all.
Of the 107 schools disclosing tuition assistance, 64 report some endowment. 43 report none.
The difference in aid delivered is real and small (median cumulative aid $7.4M against $4.0M). The difference in what stands behind that aid is absolute.
The size of it
Some of the unfunded programs are enormous.
Among schools reporting zero endowment, several disclose cumulative aid of $15–21 million over five years. That is the scale of what well-endowed schools in this dataset deliver.
Same commitment. Nothing underneath it.
The starkest case
Politz Hebrew Academy: $17.9M in aid, $7.2M in annual revenue.
Roughly half of annual revenue committed to tuition assistance in some years. Funded from current tuition and current fundraising, with no payout to smooth a bad year.
Read that as a school failing to prioritise affordability and you have it exactly backwards. This is a school prioritising it about as hard as an institution can. What it lacks is a bad year to spare.
Whose problem this is
Nobody's in particular. Which is to say everybody's.
Of the 43 no-endowment schools: 20 Communal, 11 Orthodox/Chareidi, 10 Modern Orthodox, 2 other.
Proportionally it falls hardest on Orthodox/Chareidi schools, since so few of them appear among endowed institutions. But this is not one denomination's exposure. It runs across the field.
What follows
An endowment campaign is not a generosity campaign. It is a durability campaign.
The field has argued about affordability for decades as though the question were whether schools would commit. Most already have, at strikingly similar intensity, whether they sit on fifty million dollars or on nothing. The commitment is not the variable. Its survival is.
That changes what an endowment is for. Not a licence to be generous — schools are being generous without one. It is a guarantee that this year's generosity is still there next year.
01
Ask where the money comes from, not how much of it there is
"How much aid do you give?" barely separates these schools. "What happens to that aid if the campaign comes in twenty percent light?" separates them completely.
02
The most exposed schools look the most admirable
A school committing half its revenue to aid reads as a model in any single year of numbers. The fragility never shows up in the aid line. It shows up when you ask what funds it.
03
Pooled funds may fit better than school-by-school campaigns
Even the largest endowment here throws off a payout that is useful but not transformative against a budget in the tens of millions. Pooled community funds reach a scale individual campaigns do not.
Four things this cannot tell you.
None of them reverses the finding. All of them should travel with it.
Limit one
A missing schedule is not proof of a missing endowment
A school with no Schedule D entry may genuinely have none, or may simply not have populated that form that year. The two are indistinguishable here, so the count of 43 is probably an overstatement.
Limit two
Disclosing schools, not the whole field
107 of roughly 377 schools with financial histories in the underlying data report aid in a parseable form. Schedule I detail exists only for full Form 990 filers, which tilts the sample toward larger, more established institutions.
Limit three
Aid is cumulative, revenue is annual
A $17.9M aid figure spans a school's reported years, not one of them. The revenue figure beside it is a single year. Both are reported here as the filings report them, and neither should be read as a clean ratio.
Limit four
Descriptive, not causal
Nothing here demonstrates that an endowment causes a school to protect its aid budget in a downturn. It shows which schools have the mechanism available and which have none. That is a statement about exposure, not about outcomes.
Ḥeshbon — Center on Data and Jewish Life
A data project on how Jewish life is paid for. The work comes in seasons. This page is part of Season 1 — Jewish Education. It states its own sources, sample sizes and limits.
Working prototype. Every figure comes from the underlying research. Where a chart simplifies a published result, the page says so. Nothing here audits an individual school.