Charted: Endowment Payout vs Tuition — How Thin Is the Buffer Below the Top?

When people talk about "rich universities," they usually mean market value — Harvard's $56.9 billion, Yale's $44 billion, Princeton's $36 billion. The operating question is narrower and more useful: what share of the annual budget does the payout actually fund, and how does that compare with net tuition? The 2025 NACUBO-Commonfund Study of Endowments (NCSE) answers the first half for 657 institutions. Audited financial reports and IPEDS Finance ratios answer the second for a panel of the largest endowments.

The headline NCSE figure for FY25 is deceptively calm: endowments funded an average of 15.2% of participants' operating budgets, up from 14.0% in FY24 and 10.9% in FY23. The median is only 6.1%. That 9.1 percentage-point mean–median gap is the story. A thin right tail of mega-endowments and payout-heavy liberal-arts colleges pulls the average; most campuses still live closer to the median — and closer to tuition, appropriations, and grants.

Use the dashboard above to move across four panels: Size bands for NCSE cohort support (toggle % of ops vs effective spend rate), Payout vs tuition for the institution scatter and table, Ops-support trend for the FY18–FY25 mean/median climb and institution-type jump, and Decile buffer for how fast endowment support thins below the top decile.

The NCSE mean hides a thin right tail

NCSE asks institutions what share of the operating budget endowment distributions support. Across FY25 respondents, the mean is 15.2% and the median is 6.1%. That is not a rounding error. It means half of participating campuses fund roughly one-sixteenth of the budget from endowment — or less — while a smaller set of payout-heavy schools drag the average into the mid-teens.

Total withdrawals hit $33.4 billion, up nearly 11% from $30.1 billion in FY24. The average effective spending rate ticked up to 4.9% from 4.8%. Purpose allocation still concentrates on aid: 47.4% of spending went to student financial aid, versus 17.7% to academic programs and research. Payout is not a generic "operations" faucet; a large slice is already earmarked for discounting sticker price — which is why comparing endowment share to net tuition (after aid) is the fair operating lens.

Caveat: NCSE is a volunteer survey, not a census. Share-of-budget definitions are institution-reported and need not match GAAP operating-revenue mixes in audited statements. Treat size-band averages as directional cohort signals, not audited peer benchmarks.

Size bands: the peak is not the very top

Segment FY25 support by endowment market value and the peak is the $1 billion–$5 billion cohort at 20.1% of operating budgets — not the Over $5 billion group, which sits at 18.1%. Cohorts under $100 million already average 14.8%. The least-reliant band is $101–$250 million at 12.1%. The gap between the peak size band and that trough is 8.0 percentage points.

That ordering surprises desks that assume "bigger endowment ⇒ higher budget share" is monotonic. Over-$5B universities often run hospitals, sponsored research, and clinical enterprises that inflate the operating base, so a large dollar distribution can still be a mid-teens share of a huge budget. Mid-sized private colleges with concentrated enrollment can post higher percentage dependence even with smaller absolute payouts.

Effective spending rates tell a different story from budget shares. The highest average effective rate in FY25 — 5.5% — was reported by the $51–$100 million cohort. Over-$5B averaged about 5.0%; the $501 million–$1 billion group was closer to 4.3%. Budget dependence and spending-rate intensity are related but not identical levers.

Mega-endowment financial reports: payout dwarfs net tuition

NCSE size averages compress the elite private tail. Published FY25 operating-revenue mixes show how far above the survey mean the largest private universities sit:

Harvard's FY25 distribution was $2.5 billion — 37% of operating revenue — while the university still had to fund nearly two-thirds of a ~$6.7–6.8 billion operating base from research, tuition, and gifts. Yale's endowment spending was 34% of operating revenue; net tuition, room, and board was only 8%. Princeton sits near the extreme: endowment provides about two-thirds of net annual operating revenue. Those three alone explain why "15% average" feels nothing like the Ivy operating model.

Public flagships reverse the sign. Michigan, UT System, and UVA post endowment shares in the high single digits to low teens while net tuition (and, off-table, state appropriations) carry the larger operating pillar. Absolute endowment size is not destiny: UT System's multi-ten-billion pool is diluted across campuses; Amherst's few billions against a small enrollment base produce a payout-dominant mix.

How thin is the buffer below the top decile?

Map size cohorts into a decile-style profile anchored on NCSE bands and the institution panel. Top-decile mean endowment support sits near ~28% of operations in this construction — pulled by mega-privates and payout-heavy colleges. Decile 9 (roughly the $1B–$5B / upper-public mix) averages about 20%. By decile 8, mean endowment support falls toward the mid-teens while net-tuition shares climb into the high twenties. Mid-pack deciles cluster near the NCSE mid-teens-to-low-teens endowment shares — and tuition takes the lead.

The buffer below D10 is therefore steep in the first step (~8 pp into D9) and then flattens: D7–D4 sit only a few points apart on endowment share, but their tuition dependence keeps rising. That is the operating risk asymmetry. Losing a few points of payout share hurts a Princeton-like budget differently than a tuition-majority public — but the modal NCSE campus already looks more like the mid-pack than like Harvard.

Another buffer metric is mean versus median for the whole survey: 15.2% vs 6.1%. If markets force a synchronized cut in distributions, the median campus has less endowment cushion in percentage terms — yet may also have less absolute dollar exposure. The top decile has the opposite problem: deep percentage dependence on a smoothed spending rule that still assumes long-run returns clear spending + inflation + fees.

Private, public, and foundation paths diverge

By institution type, FY25 was not a quiet year. Public institutions' average endowment support of the operating budget jumped to 13.1% from 8.4% in FY24. Institutionally-related foundations averaged 24.7% (from 19.4%), though IRF medians in prior years have often sat near 1–2%, implying a few large foundations dominate the average. Private nonprofits remain the steadier double-digit story across FY18–FY25.

Longer history reinforces the level shift. Commonfund's reconstruction shows average endowment support of budgets near ~4% in the late 1970s, roughly 10.5% when the question returned in FY08, mid-single to low-double digits through the 2010s, then a climb through FY18–FY25 that pushed the mean past 15%. The median rose from 4.0% in FY18 to 6.1% in FY24–FY25 — slower than the mean, but still a one-half increase.

Smaller endowments closed part of the gap with the largest. From FY18 to FY25, under-$50M support more than doubled from 6.9% to 14.8%, while Over-$5B moved from 16.7% to 18.1%. The size gap shrank from nearly 10 pp to about 3 pp. Convergence from below is not the same as convergence to Ivy operating models; it does mean tuition-and-appropriation campuses are asking more of thinner asset pools.

What the payout–tuition gap implies for risk

Portfolio risk tolerance and budget dependence are linked. Institutions whose payout is a small slice of the budget can, in principle, absorb larger mark-to-market swings without immediate program cuts. Institutions whose payout is a third or two-thirds of revenue need spending rules, reserves, and liquidity that survive equity drawdowns without freezing hiring and aid. FY25's 4.9% average effective spending rate against a 7.7% 10-year average return looks manageable at the survey mean — until HEPI (3.6% in FY25) and fees claim their share of the hurdle.

Tuition dependence is the other side of the hedge. Net tuition shares in the 20–30% range (Columbia, Penn, Michigan, UVA in the panel) expose budgets to enrollment and discounting cycles. Payout-dominant schools hedge enrollment risk with market risk. Tuition-dominant schools do the reverse. Neither mix is "safer" in the abstract; they fail in different years.

Caveats and reading rules

Several measurement traps matter. First, endowment share of budget (NCSE) and endowment share of operating revenue (financial reports) are cousins, not twins — donor restrictions, transfers, and hospital consolidations move the denominator. Second, net tuition already embeds discounting; comparing endowment aid spending to gross tuition double-counts the discount story. Third, system endowments (e.g., Texas) should not be read as single-campus operating dependence. Fourth, FY25 public and IRF averages jumped sharply; one year does not prove a permanent regime. Fifth, IPEDS Finance vintages lag some FY25 audited reports, so panel tuition shares mix adjacent fiscal years where necessary.

Read the dashboard as a map of operating mix, not of prestige. Market value ranks who has capital. Payout and net-tuition shares rank who can miss a tuition cycle — or a market cycle — without rewriting the budget in a single summer.