The worry has always been a two-tier system: a few flagship schools with the wealth to discount tuition deeply, everyone else turning families away. So we pulled eight years of tax filings for 107 Jewish day schools and read each school's endowment balance next to the tuition-assistance dollars it hands out.
One bar below holds all $484.7M of reported endowment. The rust block on the left is the five largest funds. The brass block beside it is schools six through ten. The pale slate block filling the rest is the other eighty schools. The dashed line marks where the top ten end. Notice how little of the bar is left after it.
Read the blocks left to right: rust to 39.1%, then a brass block that carries the top ten to 54.4%, then eighty schools sharing the 45.6% that is left. Those eighty are the schools this piece ends up being about.
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 years (FY2017–2024). Every school in the run carries both its Schedule D endowment balance and a fresh extraction of its Schedule I tuition-assistance grants, so the cushion and the aid commitment can be read side by side over time.
Benchmarking peer data appears only as a secondary cross-reference where noted; the findings rest on the assembled record above.
Can a worry the field has argued from experience be checked against the schools' own tax filings?
The worry holds two claims. One survives the filings; the other does not.
The familiar version runs like this: Jewish day school wealth has pooled in a handful of flagship institutions, while most schools operate close to the margin, leaning on tuition and this year's fundraising to cover a financial-aid commitment they cannot cut without turning families away.
That has been argued from surveys and from experience. This study tests it against the schools' own tax filings: endowment balances, and, newly extracted, the dollar figures schools report for tuition assistance in the structured grant fields of Form 990 (533 grant records across 107 schools, 531 of them from structured fields rather than narrative text). Those records total $942.4M in reported tuition assistance.
Read closely, the concern turns out to hold two separate claims. One holds up against the filings; the other gives way, and the reason it gives way is where the story gets interesting.
Is the money really pooled in a handful of schools?
It is. Of $484.7 million in reported endowment across 90 schools, the top five hold 39.1%, and the top ten hold 54.4%, more than the remaining eighty combined.
Abraham Joshua Heschel School reports $53.0M, Golda Och Academy $44.4M, Fuchs Mizrachi $35.1M, Charles E. Smith $35.0M, and Perelman $21.9M. These are public disclosures on public tax filings, which is what makes the analysis possible.
The shape is a handful of large funds sitting above a long tail of modest ones: the median endowed school holds $1.8 million, and the largest runs about 29 times that. By denomination, Communal schools hold 77.5% ($375.9M across 59 schools), Modern Orthodox 18.6% ($90.2M, 19 schools), and Orthodox/Chareidi 3.8% ($18.6M, 12 schools). Only twelve Orthodox/Chareidi schools report any endowment activity, so most report none rather than a small balance.
The chart below has four views, and the buttons above it switch between them. The first is a single bar holding all of the reported endowment, cut into three blocks: rust for the top five schools, brass for the next five, pale slate for the remaining eighty, with each block carrying its share of the total. The second and third views break the bar apart into one row per school, where length is dollars and the colour is the school's denomination, rust for Communal and brass for Modern Orthodox. The third view adds the median endowed school as a slate row underneath a dashed rule, drawn to the same scale. The fourth view gives each denomination one bar, with the percent inside it and the dollar figure alongside.
Watch what happens to the median row in the third view when it is drawn against the same ruler as the five largest funds.
This is the version of the concern that matters for families. Wealth concentration only troubles them if it produces a two-tier system: schools that can afford to discount deeply, and schools that must turn families away because they cannot.
Do the schools sitting on that money give more of it away?
They do not. The ten most-endowed schools devote a median 18.47% of revenue to tuition assistance, meaning about eighteen and a half cents of every dollar the school takes in goes straight back out as aid. Everyone else devotes 19.40%, a gap of 0.93 percentage points. If anything, the schools without the big endowments give a slightly larger share of what they raise.
Split by denomination, the least-endowed group gives the most. Orthodox/Chareidi schools, holding 3.8% of the field's endowment, report the highest aid-to-revenue ratio at 24.69%. Communal schools, holding 77.5%, report 19.08%, and Modern Orthodox report 18.84%.
Across the sample, the middle grant handed to one family is $12,602.
The chart below has three views. The first is two bars, and length is aid intensity: the share of everything a school takes in that it hands back as tuition assistance, so a longer bar means more of the money going out again. The rust bar is the ten most-endowed schools. The pale slate bar under it is every other school. The teal dashed line is planted at the tip of the slate bar so you can see whether the rust bar reaches it. The second view replaces those with one bar per denomination, each labelled underneath with the share of the field's endowment that group holds. The third view is a scatter: endowment share runs left to right along the bottom, aid intensity runs bottom to top, each dot is one denomination, and the dashed slate line drawn through them shows which way the relationship tilts.
Look at whether the rust bar reaches the teal line.
Two schools can commit the same dollars to aid and stand in completely different positions.
One school draws its aid from endowment payout, the slice of the fund it is permitted to spend each year, and can absorb a weak fundraising year without touching the aid budget. The other funds every dollar from this year's tuition and this year's donors, with nothing behind it if either falls short.
A generosity comparison hides that difference. A resilience comparison is built around it.
If an endowment does not buy more aid, what does it buy?
It buys a cushion under the aid a school already gives, and four in ten of these schools do not have one. Of 107 schools, 64 report some endowment and 43 report none. The gap in aid delivered is real yet modest: median cumulative aid, which is every aid dollar a school has disclosed across its filing years added together, of $7.4M among endowed schools against $4.0M among the rest.
Several of the no-endowment schools run large programs. Jewish Community High School of the Bay discloses $21.1M in cumulative aid, Yeshiva Derech HaTorah $17.9M, Politz Hebrew Academy $17.9M, Atlanta Jewish Academy $15.5M, and Kohelet Yeshiva High School $14.7M, figures in the range some of the best-endowed schools deliver.
Politz Hebrew Academy shows the strain most plainly: $17.9M in aid against $7.2M in annual revenue, roughly half of revenue committed to tuition assistance in some years, funded entirely from current tuition and fundraising with no payout to smooth a bad year. That is a school prioritizing affordability about as hard as it institutionally can, with little margin for error.
The exposure spreads across the field. Of the 43 no-endowment schools, 20 are Communal, 11 Orthodox/Chareidi, 10 Modern Orthodox, and 2 other. It falls proportionally hardest on Orthodox/Chareidi schools, given how few of them appear among endowed institutions, though no group is exempt.
The chart below has four views. The first is a field of 107 dots, one for every school that discloses tuition assistance: the pale slate dots are the 64 with some endowment behind them, the solid rust dots that follow are the 43 with none. The second and third views pull five of those rust schools out and give each one a bar, where length is all the aid that school has disclosed and the short upright tick crossing the bar is a single year of its revenue. A bar running well past its own tick has committed more to aid than it collects in a year. The third view holds Politz Hebrew Academy in full colour and fades the other four. The fourth counts the 43 by denomination, one bar each.
Find the tick on the Politz bar, then see how far the bar keeps going past it.
So: five schools hold nearly 40% of the endowment. Does that buy their families more aid?
No. What it buys is a cushion under an aid budget those schools were already funding at about the same intensity as everyone else, 18.47% of revenue against 19.40%. The 43 schools with no endowment give at that rate too, out of this year's tuition and this year's donors, with nothing underneath.
The field has spent decades framing affordability as a question of whether schools will commit to aid. The data suggests most already have, at strikingly similar intensity regardless of wealth. The open question is which of those commitments can withstand a recession, a lost major donor, or a soft enrollment year.
"How much aid do you give?" separates schools far less than "what happens to that aid if you miss your campaign by 20%?" Only the second question distinguishes a cushioned commitment from an exposed one.
A school delivering aid at half its revenue looks admirably committed in any single year's numbers. The fragility surfaces only when you ask where the money comes from.
Even the largest endowment here generates payout that helps a budget in the tens of millions without remaking it. Pooled communal funds reach a scale individual campaigns mostly cannot.
One limit worth carrying: a missing endowment schedule can mean a school genuinely has none, or that it did not populate that particular form in that year. The two are indistinguishable here, so the no-endowment group is likely somewhat overstated.
Source and method: this dataset was assembled by hand from public IRS Form 990 filings, then extended with historical and archived filings, Internet Archive / Wayback captures, and school and public web sources across FY2017–2024. Endowment balances were read from Schedule D Part V across 90 reporting schools; tuition-assistance dollars were extracted from Schedule I Part III grant fields across 1,165 filings, yielding 533 aid records at 107 schools, 531 of them from structured fields rather than narrative text. The essay accompanies "Endowment Concentration and the Tuition-Affordability Question in American Jewish Day Schools" (July 2026).
Only 107 of roughly 377 schools with financial histories in the underlying dataset disclose tuition assistance in a parseable form, so this describes disclosing schools rather than the full field, and the totals are a lower bound on actual aid. Schedule I detail exists only for schools filing the full Form 990, which biases the sample toward larger and more established institutions.
A missing Schedule D entry is treated here as "no reported endowment," which likely overstates the true no-endowment population. As a secondary cross-check on that classification, the assembled record was compared against a set of comparable independent benchmarking data covering 1,576 school-years, where 96.8% carried a reported endowment value and 3.2% left it blank. That low blank rate supports reading a genuinely absent schedule as no endowment rather than a filing gap. The peer benchmarking enters only for this validation; every headline figure rests on the hand-assembled 990 record. Comparisons throughout are descriptive rather than causal.
Source: the endowment-and-aid companion paper and the Ḥeshbon education-finance database, Season 1, July 2026. Every figure comes from public IRS tax filings. A visual essay.
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.