Endowment is the invested pot a school builds up and spends only the earnings from, and in the Jewish day school world it sits in very few hands. So we read eight years of tax filings for 107 schools, took the endowment balance off one schedule and the tuition-assistance dollars off another, and lined the two up school by school.
One bar below carries all $484.7m of reported endowment. The bright brass block running from the left is the ten largest schools. The pale block continuing to the right edge is every other school reporting one. Length is money, so the wider a block, the more of the sector's accumulated wealth sits inside it.
An original dataset assembled by hand from public IRS Form 990 filings, historical and archived filings, Internet Archive / Wayback captures, and school and public web sources, reconstructed across eight fiscal years. Every endowment figure comes from Schedule D and every aid figure from Schedule I, read filing by filing so that $942.4M in disclosed assistance and $484m of endowment across 90 schools line up school by school over time.
A peer set of comparable independent schools appears only as a secondary cross-reference where noted; the findings rest on the assembled 990 record above.
If ten schools hold most of the wealth, do their families get most of the aid?
Concentration of wealth is stark: five schools hold nearly two-fifths of all reported endowment. The intuitive consequence is a two-tier sector, where well-endowed schools can afford generosity and others cannot. Tested against disclosed tuition assistance, that consequence does not appear.
The two cells below carry the same measurement for two groups of schools: the median share of a school's revenue that goes back out as tuition assistance, so out of every dollar coming in, this is how many cents leave again as aid. The rust cell is the ten schools with the largest endowments. The plain cell is everyone else.
Look at how little separates the two numbers.
The next chart puts the concentration itself on one scale. Each row is a group of schools, and the bar running rightward from the left-hand labels is that group's share of the $484.7m sector total, marked off in quarters along the bottom. The rust bar is the row to read first; the slate bars are context. The buttons above the chart swap the grouping from rank to school type.
Notice that the two rows in the first view overlap, because the top ten includes the top five.
How many schools are running an aid programme with nothing behind it?
Among the 107 schools disclosing tuition assistance, 43 report no endowment at all, and several of them still run aid programmes that rival well-endowed peers. Every dollar of that aid is raised in the same year it is spent, which leaves those programmes exposed to any shortfall in annual giving.
The chart below reads the same way as the last one, with one row per group, but the scale changes: length is now aid intensity, the share of a school's revenue handed back as tuition assistance, and the ticks along the bottom run to 30%. Longer means a school is giving away more of what it takes in. The rust row is the group holding the least endowment wealth. The first button groups the schools by community type, the second by how much endowment they hold.
Watch which row is longest, then check how much endowment that group holds.
If the amounts given are so alike, what actually separates these schools?
The distinction runs along how aid is paid for rather than how much is given. Some schools cover comparable assistance from endowment income, while others raise the same money every year from appeals. Two schools can look identical on generosity in a single year and face completely different odds of surviving a bad one.
A commitment funded from capital and a commitment funded from this year's appeal look the same on paper, right up until fundraising falls short.Endowment & Aid · Season 1
They diverge the moment a fundraising target is missed. One draws on accumulated capital to hold steady; the other has to find the shortfall or cut the aid.
Could those 43 schools simply have left the line blank?
A blank endowment field on a filing cannot by itself distinguish a school with no endowment from one that simply left the line empty. If the second case were common, the fragility finding would be inflated. To test that, a secondary cross-reference against a peer set of comparable independent schools helps, because that record asks two separate questions: whether a school has an endowment, and what it is worth. Across 1,576 school-years answering yes, 96.8% also reported a value. A school that holds an endowment yet stays silent on the amount turns out to be rare where the two answers can be checked side by side, so a blank field in the 990 record mostly means genuinely none.
The last chart splits those 1,576 school-years into two rows, again on a scale running from 0 to 100%. Every school-year counted here has already said yes to holding an endowment. The rust row is the share that then went on to report a dollar figure. The slate row is the share that confirmed the endowment and then stayed silent on how much.
Follow the slate row and see how quickly it stops.
So: ten schools hold more than half the endowment. Does the aid follow the money?
No, and that is the finding. The ten most-endowed schools give 18.5% of revenue back as tuition assistance and everyone else gives 19.4%, while the group holding the least of the $484.7m, Orthodox and Chareidi schools, gives the most at 24.7%. Aid is spread across the sector at close to one rate, and the endowment is not.
What the money changes is what happens next year. A school covering aid from endowment income can miss a campaign and keep the aid intact. For the 43 schools with no endowment at all, this year's appeal is the aid budget, and a shortfall has to come out of the families.
Tuition assistance in parseable form appears for 107 schools. The figures are a lower bound on actual sector aid.
The disclosure test compares rates across two different samples. It is not a school-by-school verification.
Schedule I detail exists only for schools filing the full Form 990, biasing toward larger institutions.
Classification comes from naming conventions and verification rather than self-report.
An original dataset assembled by hand from public IRS Form 990 filings, historical and archived filings, Internet Archive / Wayback captures, and school and public web sources, reconstructed across fiscal years 2017–2024. Endowment was read from Schedule D and tuition assistance from Schedule I, Part III, across 1,165 raw e-filings, with grant-type labels filtered to isolate family-facing aid and exclude staff remission.
Of 533 extracted aid records, 531 came from the structured Schedule I field, spanning 107 schools and $942.4 million in cumulative disclosed assistance. The disclosure test compares rates across two different samples rather than verifying schools one by one, and Schedule I detail exists only for schools filing the full Form 990, which biases toward larger institutions. To gauge whether a blank endowment field means genuinely none, the 990 record was cross-referenced against a peer set of comparable independent schools, used only as a secondary check.
From Endowment Concentration in American Jewish Day Schools, Gavriel Brown, PhD, research note, July 2026.
The paper behind this story
Tell me who you are and where you work. I’ll send the complete paper.
Sent to the author directly. No list, no forwarding, no third parties.
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.
Ahead: Season 2 — Household Affordability · Season 3 — Federation & Charitable Giving
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.