Continue reading with an Essential or Premium subscription to Higher Education Leadership Intelligence

Cornell also relocates the trust problem from messaging to relationships. The report states that trust ”is not simply a communication problem, it is a relationship problem.” It anchors the claim in survey data: 83 percent of Americans credit universities with driving economic growth, and 76 to 79 percent see positive local and healthcare impact. Meanwhile, institutional confidence has fallen from 57 percent in 2015 to 38 percent in 2026. The report concludes that trust survives where universities touch people operationally. That is closer to the argument we made in April than anything in Yale's report.

On AI, Cornell reads the labor market Yale ignored. Yale treated AI as a syllabus problem. Cornell cites the early evidence of entry-level knowledge-worker displacement and concludes that the transactional degree is ”uniquely vulnerable to competition from artificial intelligence.”

The most useful sentences in the report are its admissions of financial ignorance. Cornell discloses that 83 percent of its institutional grant aid in 2025–2026 came from the operating budget, with the remainder from the endowment. It also states that a question as basic as whether the university makes or loses money when it enrolls more students is ”currently unanswerable.” That single admission is worth more to a chief financial officer than all 58 pages of Yale's report. It also shows that Cornell's aid economics sit closer to a tuition-dependent institution's than to Yale's. That is exactly why Cornell's choices about what it will not do deserve scrutiny.

The report's own theory of trust indicts its omission. Its best argument is that trust is a relationship problem, and that trust survives where universities touch people operationally: hospitals, extension offices, local economies. For a family, the most important operational relationship with a university is whether it keeps their child safe and acts when she reports harm.

The report covers student life in detail. It addresses student mental health, residential life, peer mentoring, tech-free dorms, the ”ruthless club culture” of selective student organizations, and how students repair ”relational missteps.” It says nothing about fraternities, sororities, or campus sexual violence.

The structural explanation is the same fragmentation the report diagnoses everywhere else. Cornell assigned the subject to a separate Presidential Task Force on Campus Sexual Assault. That task force issued its recommendations in March 2026, six months before the committee finished, and the report never cites it. The report's central argument is that the modern university has split into units pursuing separate missions, with no one accountable for the whole. This 238-page document was produced inside the same pattern. Student safety belonged to another vice presidency and another task force, so it fell outside the ‘future of the university.’

A faculty committee on mission could reasonably have excluded Title IX adjudication from its scope. This one chose to write at length about dorms, clubs, and student wellbeing. Once a report covers student life, leaving out sexual violence becomes an editorial decision. The decision left out the most consequential trust event on campus.

The consequence reaches well beyond Ithaca. Cornell's model for the American university now arrives alongside a governor and a senior senator demanding an independent investigation into how the same institution handled one student's report. At most universities, strategic planning treats Title IX, Greek life, and campus safety as compliance matters owned by student affairs and the general counsel. But families make no such distinction. In our assessment, one mishandled case can do more damage to an institution's enrollment pipeline and donor base than anything in its strategic plan can repair. Cornell's report leaves that ordering of risk out, as do most strategic plans written the same way.

By two different measures, Cornell is among the most expensive places in the country to attend after aid. The report's Section 3.8 cites the College Board: the average net price at private nonprofit four-year colleges was about $32,830 in 2025, essentially flat in real terms for twenty years. The report calls that stability ”a truly impressive accomplishment.” It then reports Cornell's own average net cost of $65,370, or $45,395 for New York residents in the contract colleges.

The two figures may not rest on identical definitions, and the report does not say. The federal College Scorecard removes the ambiguity by measuring net price only for students receiving federal aid. On that measure, Cornell's average is $28,690. That is third highest in the Ivy League, $9 below Penn, and roughly 30 percent above the average of the other seven. Princeton's is $6,128. The median American household earned $83,730 in 2024.

The report then walks through every cost lever and closes each one. It rejects scaling instruction, reducing teaching support staff, increasing automation, and cutting co-curricular programs as incompatible with the human-centered education it proposes. What remains is thin:

The formal recommendation is another task force.

Meanwhile, nearly every other recommendation adds cost. The report calls for:

Appendix G compiles 95 recommendations by our count and attaches a cost estimate to none of them. All of this follows a year in which the Resilient Cornell initiative removed $350 million from the operating budget.

Cornell can defer the price question because its brand holds demand at that price. Yours will not. A tuition-dependent institution that copies Cornell's sequencing will exhaust its reserves on the first half and never reach the second. Cornell's sequence makes the spending commitments first and sends pricing to a task force.

Section 3.6, titled ”Roads That Should Not Be Taken,” is the most consequential section of the report for anyone outside the Ivy League. It rejects four options:

Cornell's reasons are defensible for Cornell. A residential Ivy that shortens its degree or moves instruction online gives away the scarcity that justifies its price. For a regional university, a shorter degree is a direct price cut, and online delivery is the only path to scale. The report concedes that its own online business, eCornell, ”barely registered as a topic across the hundreds of conversations behind this report.” The channel through which Cornell competes internationally at scale was nearly invisible to the people designing its future.

Cornell's case against the vocational turn deserves a direct answer, including from us. In April we argued that mid-tier institutions should restructure into vocational credentialing engines in occupations outside the first wave of AI displacement. Cornell objects that vocational training prepares graduates for professions as they currently exist, and ”graduates prepared for current practices become vulnerable to obsolescence as those practices shift.” The objection resonates. Choosing which occupations AI will spare is a forecast, and institutions that stake their identity on a forecast inherit its error bars.

The sharper version of our position is this: The unit of competition is verified employer outcome by program. The occupational mix is a portfolio, rebalanced every cycle as placement and wage data arrive. Nursing e.g., belongs in that portfolio because the demand data is public and durable. Programs whose placement data weakens get cut before the market cuts them.

Cornell's answer to AI is ”sound judgment,” and judgment has value to an employer only when someone can verify it. The report builds its entire value proposition around cultivating judgment. It defines the term carefully and offers no way to measure it. At Cornell, the verification comes bundled: selectivity, the alumni network, and the brand do the certifying. At a regional institution, an unmeasured claim of judgment is indistinguishable from the undifferentiated credential AI is devaluing.

The report's own evidence describes where the value is migrating. Three findings stand out:

Read together, those three findings describe a squeeze. Judgment is built through entry-level work, and AI is automating entry-level work. The institution that can demonstrate its graduates arrive already able to perform second-year tasks will capture the value. That is a mid-tier opportunity. It requires evidence employers accept, and Cornell's report never needed to build it.

Read the recommendations for operational specificity and the institution's revealed priorities become clear. Faculty and governance constituencies receive concrete, implementable commitments:

Families and applicants receive deferrals. Tuition goes to a task force. Admissions gets a commitment to publish the empirical effects of its criteria. On legacy preferences, which three-quarters of Americans say should play no role in admissions, the report commits to communicating clearly and to ”consider whether and how” those preferences reflect institutional values. Cornell's 2025–26 Common Data Set lists alumni relation as ”Considered” in first-year admissions.

Cornell already produces the admissions transparency it recommends, for one audience. Under its federal agreement, Cornell sends the government quarterly admissions data disaggregated by race, GPA, and test scores. That data is the raw material for the empirical analysis the report says Cornell should consider publishing.

The deferral pattern extends across the report's hardest questions. Each one goes to a further process:

The report quotes a stakeholder describing universities as ”over-managed and under-led,” then proposes a new permanent deliberative body as the remedy.

None of this is sinister. Eighteen faculty members wrote the report, and committees recommend what their members understand best and can implement most readily. The stated priority is the price crisis. The revealed priority is faculty working conditions. Trustees at other institutions should expect the same pattern from any committee built the same way.

A 238-page report on public trust, written during a live federal dispute, never names the federal government's counterparty. Federal pressure appears in the passive voice: ”threats to federal funding have been used to pressure universities.” The report also omits Cornell's own agreement with the administration, signed November 7, 2025. The agreement's main terms:

Cornell maintains the payment is neither a fine nor an admission of liability, and the agreement bars the government from dictating speech or curriculum. The terms are defensible, but their absence from a document about transparency is harder to defend.

On viewpoint diversity, the committee could not agree on whether the problem exists. One member cites survey data that roughly 12 percent of faculty identify as conservative and ties the imbalance to declining public trust. Another calls viewpoint diversity ”the fulcrum from which a constellation of organized vicious attacks on higher education have emerged.” The report presents the split as deliberation in action. In practice, the issue driving much of the public's distrust leaves the committee unresolved and passes to a future community conversation.

Accreditation and the employer bypass do not appear at all. Health systems training their own nurses and commercial providers credentialing at scale were central to our April analysis. They are absent from Cornell's, which is consistent with an institution that faces neither threat. Their absence also confirms that the elite register of this debate has little to say about the forces actually reshaping the regional market. The institutional restraint we described in April is still present, now in a more sophisticated register.

The report's most valuable contribution to mid-tier leaders is the set of questions Cornell admits it cannot yet answer, plus one it never asked. Four belong on every cabinet agenda this fall:

If an institution with Cornell's resources cannot answer the first question, most regional cabinets will find that their data systems were built to report enrollment, and nothing in them was built to price it.

Running the numbers is the easy part. The hard judgment comes after: deciding which programs, constituencies, and commitments absorb the loss once the answer is known, and making that call before a lender, an accreditor, or a state board makes it for you.