The paper behind this story
The Lean Model
Request the full whitepaper
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 dual-curriculum day is supposed to cost more, not less: a second faculty, a second set of materials, a second block of instructional hours. So we took four years of national benchmarking data on 101 Jewish day schools, every one of them carrying that same double mandate, and compared the Orthodox schools with the rest. The Orthodox schools deliver it for a third less money. Here is where the money isn't spent, and here is what the leanness costs them.
Orthodox day schools collect about a third less tuition and spend about a third less per pupil than their non-Orthodox Jewish peers. Nothing on the giving side makes up the difference.
The comparison is deliberately narrow. Rather than measuring Jewish schools against the wider independent-school market, this study measures Orthodox schools against other Jewish day schools, the only peers carrying the same two-curriculum mandate. Holding that structure constant isolates what is distinctively Orthodox about the economics.
Two suspects could explain a lower collected price. Maybe donors fill the gap. Maybe Orthodox schools post ordinary prices and discount them harder. The data rejects both. Fundraising per pupil is statistically indistinguishable between the two cohorts, and so is the discount rate.
What actually separates them is the price on the door. Orthodox schools post list prices roughly a quarter lower on two independently built measures, staff leaner, run larger classes, and hold about a third the endowment. The model is lean, and it has to stay lean every year to work.
A school that spends less, charges less, and raises no more than its peers is not subsidized. It is lean, and the price of lean is having almost nothing in reserve.
Who exactly is being compared with whom?
Every school in this pool runs a general-studies program and a parallel Judaic-studies program, with separate faculty, materials, and instructional hours. That is the shared burden. The question is what happens inside it, so the study pools four years of national benchmarking data and splits the Jewish day schools into two cohorts.
Each small square below is one school-year of financial data. The green squares at the start are the Orthodox school-years, 53 of them. Everything blue after that is the non-Orthodox Jewish school-years, 253 of them. Notice how small the green block is against the blue.
How big is the gap on the two numbers that matter most?
Start with the money that actually moves: net tuition revenue, which is what families really pay after aid, and operating expenses, which is what the school really spends. The chart below shows both, as one pair of bars each. In every pair the blue bar on top is the non-Orthodox Jewish median and the green bar underneath it is the Orthodox median. Both run rightward from zero dollars, so a longer bar means more money.
Watch the green bar in each pair stop well short of the blue one. These are the two largest and sturdiest results in the study, and the p-values attached to them, .0016 and .0001, mean a gap this size would be very unlikely to show up if the two groups actually collected and spent alike.
This is a different sustainability story than "philanthropy replaces price." It is closer to "both sides of the ledger are simply smaller."
Is the difference just donors filling the gap?
The natural hypothesis is substitution: Orthodox communities charge families less and let donors carry the difference. If that were the mechanism, advancement revenue per pupil would run higher in the Orthodox cohort.
The chart below puts four outcomes on one scale. The darker vertical line near the right edge is zero, meaning no difference between the two cohorts. Every bar runs leftward from that line, because on all four measures the Orthodox number is the lower one, and the further left a bar reaches the larger the gap. A bar drawn solid has cleared the usual threshold for statistical significance. A bar drawn as a dashed outline has not, and should be read as a direction rather than a finding.
Find the bottom row, advancement, and notice that its bar is an outline.
The finding is symmetric and worth stating carefully. Orthodox schools raise no less philanthropic revenue per pupil than their non-Orthodox Jewish peers, and this data gives no evidence they raise more. The lower collected price is financed by neither donors nor endowment draw. It is financed by a smaller cost base.
If donors are not paying for it, is the price itself lower?
A lower collected price can come through two channels: a lower posted price, or a deeper discount off an ordinary one. Take the posted price first. Two measures were built independently of each other, the Grade 1–12 weighted sticker and a full-day kindergarten check, and they move together at a correlation of 0.98, which is about as close to reading the same thing twice as two separate measurements get.
Each measure is one row in the chart below. The blue dot on a row is the non-Orthodox Jewish median posted price. The green dot is the Orthodox median. The bar joining them is the distance between the two prices, and both rows share one dollar scale, so a longer bar means a bigger discount on the door.
Watch the green dot travel left away from the blue one on both rows.
Or is the sticker ordinary and the discount deeper?
NAIS's financial-aid fields sit empty for every peer school in the pool, but the discount is recoverable another way: the gap between the gross tuition a school posts and the net tuition it collects. Constructed that way, it covers more of the pool than any other supplementary measure in the study.
The chart below measures both Jewish cohorts against the wider market. The line on the left is the market rate itself, the discount taken by 660 comparison independent schools in the same national pool. Each bar runs rightward from that line, and its length is how many additional percentage points of the posted price a cohort leaves uncollected. The green bar ends in a paler section: that is the range its estimate covers across specifications, not a second finding.
Look at how far both bars run before they stop.
Put sections 04 and 05 together and the channel question is settled. The Orthodox tuition shortfall is a list-price phenomenon. Orthodox schools post prices roughly a quarter below their Jewish peers, then discount those prices at approximately the same rate everyone else in the sector does.
If the price is genuinely lower, where is the money not being spent?
A price a quarter lower has to be paid for somewhere, and the trace is visible in the classroom. Orthodox schools carry fewer staff per hundred students and put more children in each room, the two most direct levers a school has over its cost base.
In the chart below, every dot is one child in the median classroom. The blue row of dots is the non-Orthodox median class and the green row underneath is the Orthodox one, and each row is simply as long as its class is large. Count the extra green dots.
What does running this lean cost the schools?
If a communal-financing story were hiding somewhere, the last place to look is accumulated wealth. It isn't there either. The chart below is two columns on a dollar scale: the blue one is the non-Orthodox Jewish median endowment per pupil, the green one the Orthodox median. Height is money per child in the bank.
Look at how much of the blue column the green one fails to reach.
On this data, Orthodox schools are no more philanthropy-dependent than their non-Orthodox Jewish peers, so "how dependent is this model on donors" is the wrong worry. The sharper question for school leaders and communal funders is what happens to a model that must sustain a materially lower cost base every year, without an endowment buffer, when costs spike or a fundraising year disappoints.
Which of the three explanations survives?
Three accounts could explain how Orthodox schools deliver the sector's most intensive dual curriculum at its lowest collected price. The study tested each one directly.
Would predict elevated advancement revenue per pupil among Orthodox schools, communal giving substituting for price.
Found: −11.7% with p=.622. Statistically indistinguishable from their Jewish peers in every specification.
Would predict ordinary posted prices with a bigger gap between gross and net tuition on the Orthodox side.
Found: an apparent 7.9-point gap that falls to 4.0 points under enrollment controls (p=.299) and fails permutation testing.
Would predict lower list prices, lower spending, leaner staffing, larger classes, and thin reserves, all at once.
Found: all five. Prices ~25% lower, spending 32% lower, staffing leaner, classes larger, endowment at roughly a third.
So: how do Orthodox schools run the most expensive kind of school day for a third less money?
By spending a third less, and by nothing else. The dual-curriculum premium does not merely fail to appear in the Orthodox cohort. It runs backward, and the reversal is financed by the price on the door and the cost base behind it.
Orthodox day schools collect about a third less tuition, spend about a third less, post prices about a quarter lower, discount like everyone else in the Jewish sector, out-raise no one, and bank almost nothing. That is a coherent design, and for the families it serves it is the difference between affordable and impossible. The open question is how much shock a model with this little slack can absorb, year after year, while the sector around it leans on endowments the lean model never built.
The Orthodox cohort is 18 schools inside a NAIS-benchmarked pool, so it skews toward larger, Northeast- and Canada-concentrated, institutionally established schools. It likely underrepresents the Chassidic and Yeshiva World schools driving most recent Orthodox enrollment growth nationally. The findings describe the benchmarked segment of the sector, and the core estimates carry wider confidence intervals than a larger cohort would allow.
Sticker price, endowment, and class size rest on as few as 5 to 10 reporting Orthodox schools each, and are analyzed cross-sectionally with Mann-Whitney and permutation tests, the appropriate standard for a small treated group. They should be read as directional rather than precise. The kindergarten list-price check draws on a single school year (2025–26) and one grade, and its proxy status can only be verified within the Jewish cohorts.
The discount rate is recovered as the gap between reported gross and net tuition revenue, which bundles need-based aid, merit awards, and tuition remission into one number. And the null between the two Jewish cohorts means no difference is established, never that none exists: with 14 Orthodox schools, a moderate discounting difference would not reliably be detected.
Staffing intensity misses conventional significance in the full sample (−11.7%, p=.099) and clears it decisively once the 3 Canadian Orthodox schools are dropped (−21.9%, p=.0014), which suggests Canadian Orthodox schools staff more like their non-Orthodox peers. Three schools cannot resolve that question; it is flagged as a candidate for a dedicated cross-border comparison.
Nothing here measures educational quality, teacher compensation adequacy, or family experience; the data speak to revenue, spending, staffing ratios, and reserves. Larger classes and fewer staff per pupil describe a cost structure, and whether that structure is a sustainable equilibrium or a deferred bill is exactly the resilience question the paper poses without claiming to settle.
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.