The Money

1. Five years of nonprofit filings, in one table

Minerva University files as the Minerva Institute for Research and Scholarship, EIN 46-2589747, a 501(c)(3) school (tax-exempt since July 2015). Figures below are from IRS Form 990 data via the ProPublica Nonprofit Explorer API and filing aggregators, fiscal years ending June 30. FY2025 is aggregator-reported (Cause IQ) pending full-XML publication — treat it as provisional.

*FY2025 via Cause IQ aggregation; full filing XML pending. Note FY2022’s net-asset dip despite a surplus reflects balance-sheet movements (liabilities including unsecured notes) — we report filings as filed and don’t smooth them.

---

2. Where the money comes from: donors, not tuition

Program-service revenue — overwhelmingly tuition and housing — runs $13–22M a year for roughly 600+ students. That implies low-$20k-per-student revenue after aid, consistent with published tuition ($18,700 in 2023–24 plus ~$13,350 housing, before the 78%-of-students aid Minerva reports). The gap is philanthropy: named gifts include Reed Hastings’ $20M (2023, against a reported $100M ten-year pledge), $50M associated with the Nippon Foundation, and Silicon Valley Community Foundation pass-throughs (e.g., $17M in one grantmaker year, $20M in another).

Surplus years are gift-timing years. When big gifts land, the university posts +$15–23M. When they don’t (FY2021, FY2024), it posts −$7–8.5M. That is the definition of a model that is fragile without stable external funding — not an insult, an accounting description. A university with a large endowment smooths this cycle with investment income. Minerva’s investment income is negligible ($693 in FY2022; ~$300–600k in adjacent years). There is no endowment cushion doing the work.

---

3. Where the money goes: pay, aid, and $0 for the founder

Compensation dominates. Other salaries and wages run $11.6–13.4M a year; officer/director/key-employee compensation runs $0.93–2.1M (3–6.6% of expenses). FY2023 reported 37 employees earning over $100k. Grants paid — overwhelmingly student scholarships and aid — were about $9.35M in 2024. Professional fundraising fees: $0 (gifts come through relationships, not telemarketers).

Two things matter here. First, Ben Nelson draws nothing from the nonprofit — in every filing we reviewed, his reportable and related-organization compensation are both $0. Whatever he earns comes from the for-profit side, which discloses nothing publicly. Second, the “Related” column is $0 for every listed officer — the nonprofit states its executives take no pay from related organizations. Keep that in mind when we get to the overlap section: shared people, but (per the filings) not shared paychecks at the top.

---

4. The other Minerva: venture-backed, revenue-light

Minerva Project, Inc. (Delaware corporation, EIN 45-2862192, 1145 Market Street, San Francisco) is the for-profit that owns the Forum platform, curriculum IP, and partnership business. Its scale per company-data aggregators: ~37 employees and ~$6M in annual revenue — against $179.6M in total funding across rounds led by Benchmark (2012 seed, $25M), TAL/ZhenFund/Yongjin (Series B), and ByteDance (Series C, $57M, 2019). Nelson’s “$250M invested” figure plausibly folds venture capital and philanthropy together.

SEC Form D filings (CIK 0001546672, all signed by Ben Nelson as CEO, revenue field: “Decline to Disclose”) show the private raises behind the headlines:

A $6M-revenue company carrying a nine-figure venture valuation has exactly one asset that can justify it: the Minerva reputation and the contracts that reputation can win — first the nonprofit’s platform fees, now the University System franchises. That is why the nonprofit’s success is, economically, the for-profit’s lifeline. The venture doesn’t need the university to be big. It needs the university to be believed.

---

5. Who owns the name on the diploma — and why donors should care

A common shorthand says “Ben holds the Minerva trademark.” The precise, sourced version is nearly as stark: the MINERVA marks are owned by Minerva Project, Inc. — the venture-backed for-profit — not by the nonprofit university whose students carry the name. Ben Nelson is the founder and CEO of that corporation (and signatory on all its SEC filings); whether he holds the marks personally is the wrong question. Control runs through the company he runs.

Method note: ownership verified via USPTO TESS-derived mirrors (Furm) cross-checked across serials; registrant name and address (San Francisco, CA 94103) match the for-profit’s SEC address in all three Form Ds. Minerva’s own site footer confirms it: “The Minerva name, logo, and trade dress are trademarks of Minerva Project.” “Ben holds” is shorthand we correct to “his company holds, under his control as founder/CEO” — the economic substance is the same, the legal precision matters.

---

6. The revolving door: documented, by name

Here is what public records do prove about personnel overlap — and it is substantial. Minerva Project’s own team page introduces its leaders with the same sentence pattern, over and over: “Before joining Minerva Project, [X] was [Professor/Associate Professor/Head of College] at Minerva University.” LinkedIn histories confirm it:

Beyond the one-way moves, several people hold concurrent dual appointments right now — e.g., Assistant Professor (University) + Academic Program Manager (Project) pairs, and a Professor of Psychology (University) + Director of Partner Programs (Project) pair, per current LinkedIn titles. The 990 headcount context: the nonprofit reported 651 employees + 12 volunteers in 2024 (likely including part-time/global instructional staff) versus 200–300 on company profiles — while the for-profit runs ~37 people. The Project’s senior academic bench is, to a large degree, the University’s former faculty.

Why this matters more than gossip: the people who grade the model (former professors turned partner-program directors) are the people who sell it. Faculty who left the classroom for the vendor have every professional incentive to report that the curriculum works — their current employer sells exactly that conclusion.

---

7. The equity layer: what our sources say, and what the paperwork shows

Our editorial position: the revolving door is proven; the equity breadth is alleged by sources and undisclosed on paper. Both point the same direction — the nonprofit and the venture are not at arm’s length — but we will not collapse the distinction, because precision is what makes the provable parts stick. The three disclosures that would settle the equity question for good:

If Minerva publishes any of the three, we will update this page with the documents inline and correct anything they contradict. If you have primary documents — offer letters with equity terms, cap-table excerpts, board minutes — contact defundminerva@proton.me. We review everything; we publish only what we can verify.

---

8. The price tag: from $10k promise to $56k reality

In September 2013, Minerva announced undergraduate tuition of $10,000 per year — “approximately 25 percent the cost of tuition at other highly selective, private American universities” — with a total cost of attendance of $28,850 including room, board, books, and insurance. The founding class paid nothing at all. The pitch was effective altruism in institutional form: elite education at a price the global middle class could reach, with need-based scholarships and low-interest loans (including for international students shut out of domestic loan programs) closing the rest of the gap.

Tuition has roughly tripled since launch. Minerva’s own financial-aid page lists the Class of 2031 at $31,300 tuition with a $59,600 total cost of attendance — and discloses 5–6% yearly increases on top, before travel, visas, insurance, and taxes. Average grant aid ($14,925, IPEDS 2023–24) brings the typical net to roughly $30,455 — still above the average net cost at public universities (~$20,800).

Set that against what most American families actually pay. Per the College Board’s Trends in College Pricing 2024: average published in-state tuition at public four-years is $11,610, with a total on-campus budget of $29,910 — and inflation-adjusted public tuition has fallen 4–10% over the last decade, while Minerva’s roughly doubled in nominal terms. Minerva’s single-year 26% hike compares with a 2.7% average increase across public four-years. Minerva now costs about 1.5× the average in-state public budget and sits just under the average out-of-state public sticker ($49,080) — without a campus, without research infrastructure, without tenure-line faculty, and without accepting Pell Grants or federal aid, the very programs that make publics affordable (net in-state tuition after grants averages ~$2,300).

So is $23,630 tuition “still expensive”? Put it next to famous names, same academic year (2024–25, IPEDS via CollegeTuitionCompare unless noted):

*Stanford full-pay total per Feb 2024 trustees announcement (tuition $65,127 there vs. $65,910 IPEDS — different fee inclusions; either way ~3× Minerva tuition). †UCLA continuing-student budget, financial-aid office. Averages: College Board 2024.

The honest verdict cuts both ways, and we print both: Minerva is roughly half the tuition of Stanford, Harvard, or USC — the “quarter the cost of elites” pitch has decayed to about half-to-a-third, but it isn’t nothing. Against the schools most families actually attend, it is expensive: double the average public tuition, ~$8–9k more tuition than Berkeley or UCLA in-state, and a total bill ($45,380) that exceeds UCLA’s in-state total ($42,639) and 1.5× the public average — for no campus, no labs, no library, no tenure-line faculty, and no federal aid access. The fair rebuttal Minerva could make: 85% of its students are international and can’t claim in-state tuition anywhere (Berkeley nonresident tuition alone is $47,265; UCLA nonresident total hits ~$77–85k). The fair rejoinder: those same students could attend excellent home-country or European publics for a fraction of Minerva’s price — the “global access” buyer is precisely the one overpaying most.

---

9. The full ride is gone — and the money explains why

Compare two documents. In 2013, Minerva announced $10,000 tuition, a founding class that paid nothing for all four years, and its own loan program so international students — locked out of domestic loans — could still come. Today, Minerva’s financial-aid page says, twice and in bold-adjacent plainness: “Minerva does not offer full-ride financial aid packages.” Every student gets an annual expected contribution. Aid is “limited and awarded based on availability.” Miss the aid deadline in your admissions cycle and there are no extensions or exceptions.

And then came the second reversal — from need-blind to need-aware. In 2017, Business Insider reported it as fact: “Minerva is need blind, and offers full financial aid to those in need.” Today, Minerva’s own FAQ asks itself “Is Minerva University financial aid ‘need based’ or ‘need aware’?” and answers: “Financial Aid at Minerva University is need aware… we do take the level of need into consideration when making final recommendations for admission.” The aid page adds that the availability of aid “can impact admissions decisions” and urges aid applicants to apply early since funds are limited. The Extended Decision cycle offers no financial aid at all. Need-based aid as originally promised is gone; now even the chance of admission itself depends on what you can pay. For a school whose founding pitch was access for the global middle class, that is the mission statement in reverse.

IPEDS via CollegeTuitionCompare/UnivStats; 2017–2020 gaps are reporting gaps, not zero-aid years. Zero Pell Grant and zero federal-loan dollars in every year — Minerva does not participate in Title IV.

Why did the full ride die? Minerva has never published a memo saying “we ran out of money,” so state the inference chain honestly: aid is funded by private philanthropy (their words, via aid guides describing the program); philanthropic contributions swing between 38% and 70% of revenue year to year (section 1); FY2024 posted a −$8.5M loss while tuition jumped 26% (sections 1, 8). A school that must re-raise its aid budget every single year, with no endowment and no federal aid backstop, cannot guarantee full rides — so it stopped promising them, replacing the promise with “shared investment” philosophy language. The philosophy may be sincere. It is also what a balance sheet in distress would say.

The deepest cut: the full-ride promise was the fuel of the selectivity machine (see our selectivity dive) — tens of thousands of aid-chasing applications producing the famous 2% rate. Kill the full ride and the denominator eventually bleeds out, the rate rises, and the prestige loop unwinds. Read the University System pivot (our Express breakdown) in this light: a brand that can no longer mint selectivity the old way must monetize the name directly instead.

---

10. The federal money left on the table — on purpose

Every IPEDS year on record shows the same zeros: $0 in Pell Grants, $0 in federal student loans, $0 in federal institutional aid. Minerva is eligible for Title IV and simply does not participate — confirmed as far back as 2017 (“eligible for federal financial aid programs but does not participate”). This is not an oversight. It is a founding doctrine, and Nelson has defended it on the record for over a decade:

• Compliance costs money: taking federal aid carries “about $1,000 per student” in compliance overhead — “a tenth of Minerva’s tuition” (2014 business-plan coverage).
• Aid corrupts admissions: “financial aid only leads to higher tuition rates and restricts how universities construct their admissions policies” (2013). Federal money is “a drug” — “if you put a drug into a system, the system changes itself to fit the drug… in 20 years we’d be majority American, with substantially higher tuition” (2014). Like Ulysses, Minerva would “tie itself to the mast” with private-sector funding only.
• Reconfirmed at independence: after 2021 accreditation, the university would “continue to forgo federal financial aid.”

One qualification, because this page corrects its own side’s slogans too: with ~85% international enrollment, foregone Pell dollars are modest — on the order of $0.5–1M a year, not a rescue package. “Low-hanging fruit” is real, but it wouldn’t close an $8.5M hole. The incompetence charge lands harder elsewhere: a university this dependent on philanthropy, running its own loan book, hiding from federal transparency, and hiking tuition 26% — while lecturing the sector on sustainability — is not making hard trade-offs. It’s refusing to make them.

---

11. The math that never closed

The 2014 business plan did the arithmetic in public: grow to 2,500 students per class, and “annual revenue of up to $280 million” follows. The plan even specified the cadence — 200–300 in the next class, “double in size roughly every year for a few years after that.” Run the tape:

Nelson’s own 2020 numbers convict the model without our help: $30,000 all-in cost, ~$16,000 of it tuition — while “80 percent of our students cannot afford $30,000, and more than 70 percent cannot afford $20,000.” Do that multiplication: 650 students × ~$18,000 net tuition ≈ $12M against $30–40M in costs. The gap was never closable from operations. It was always philanthropy-shaped — which is why every forecast error in Minerva’s history points the same direction: enrollment up, revenue up, costs down, all wrong, all flattering the pitch.

Even the famous selectivity number fails its own arithmetic. A “1–2% acceptance rate” beside the founder’s admission that ~96% of applicants “shouldn’t be there” is not a measurement. It’s a denominator someone chose to keep.

So: is it that the people running Minerva “can’t do basic math”? Sharper verdict: they can. They did it in 2014, in public, and missed by 93%. A decade of errors that all flatter the pitch is not innumeracy — it’s promotion. The math closed exactly once: as a fundraising story.

---

12. The SEC paper trail: five offerings, partial fills, one exit

The for-profit’s own SEC filings tell the fundraising story in the company’s handwriting. Minerva Project, Inc. (CIK 0001546672) filed five Form D exempt-offering notices. Every document is mirrored in our SEC filing browser — read them yourself:

Three things stand out — with one honest caveat. The caveat first: a Form D “offering” figure is a ceiling the company sets, not a promise. Issuers routinely authorize more than they sell, file before closing, or split rounds across vehicles — so a low sold percentage alone doesn’t prove a failed raise. It can be intentional. First, even allowing for that, the 2014 remainder never materialized. Press announced a $70M Series B target and the first close at ~$45–50M; no further Form D ever filed the rest, and contemporaneous reporting (GlobalVenturing) concluded the company never raised the remainder. Second, the 2019 notice is a $7.67M fully-sold tranche filed one day after the $57M Series C announcement — proof that a Form D captures one vehicle, not the round, and a warning against treating any single filing as the whole raise. Third, the board seats trace the China arc: TAL Education’s founder joins in 2014, TAL’s CFO replaces him in 2019, ByteDance’s founder joins for the Series C — and by 2022, all three are gone, replaced by operating executives. TAL’s own NYSE 20-F for fiscal 2026 mentions Minerva zero times: whatever the stake once was, it no longer clears a disclosure threshold.

The honest limit: private companies don’t disclose cap tables, so nobody outside the company can state TAL’s or ByteDance’s ownership percentage. Anyone who claims otherwise is guessing. What the filings do prove — amounts sought, amounts sold, who sat on the board and when they left — is documented, linked, and downloadable in the filing browser.

---

Corrections: defundminerva@proton.me. We timestamp every correction inline. Nothing on this page is legal, financial, or academic advice.