They forgot what they were. We stopped asking.

The instruction to run the university like a business is now so common that it passes without examination. It should not. It is a category error, and the error is specific enough to name.

A business is defined by a single structural fact: someone can refuse to pay, and then the business ends. Every quality admired in commercial organizations is downstream of that fact, including the decisiveness, the focus, and the willingness to kill a project that is not working. The discipline is not a temperament. It is a response to a loss condition. Remove the loss condition and the discipline does not survive on its own.

So the question worth asking of any institution claiming the label is simply what its loss condition is. Four answers are required: a customer who can leave, a competitor who can end you, capital that can be destroyed, and a price set by someone entitled to refuse it.

The customer is eighteen or nineteen, financing the purchase with debt that cannot be discharged in bankruptcy, unable to evaluate the product before consuming it and unable to return it afterward. Whatever that is, it is not a market relationship.

The revenue is substantially public: appropriations, federally backed tuition, and grants. A grant is an achievement. It is not a sale. A sale occurs when someone free to keep their money decides otherwise.

The competitor historically could not end you. Public campuses have closed rarely, because closure is a political decision rather than a market outcome. That is changing, and the manner of the change is the point.

The price of internal standing is denominated in a currency the institution issues itself. Academic reputation is scored on placements in journals, ranked on lists maintained by the same community that publishes in them. The unit of account in a department is the outlet, not the finding. That is what it looks like when a market has lost contact with the thing it was meant to price.

The third item deserves more than a line, because the closures have started. Pennsylvania consolidated six universities into two, effective 2022. Vermont merged three state colleges into one institution in 2023. Wisconsin has closed or moved online eight of its thirteen two-year branch campuses since 2023. In May 2025 the Penn State board voted twenty-five to eight to close seven commonwealth campuses, winding down after spring 2027.

But look at the mechanism. Not one of those campuses lost a market test. Each was ended by a board or a system office, on a budget line, by vote. The loss condition is arriving roughly thirty years after the vocabulary did, and it still is not a market.

Look also at which campuses these are. Wisconsin's closures fell predominantly on branches serving rural students. Penn State's seven are its rural commonwealth campuses, where enrollment had dropped forty-three percent over a decade. These are the two-year feeders and regional locations carrying the students furthest from a flagship, which is to say the parts of the system still doing what it was built to do. The flagships that adopted the commercial costume most enthusiastically are not the ones closing. Applied to a public system, business logic is retiring the mission and preserving the performance of it.

None of this is an accusation of idleness. The work is real and often excellent, and administering it is a genuine skill that academics who disdain all administration are usually being subsidized by someone else to avoid. The objection is narrower. It is that an institution with no loss condition, using the vocabulary of institutions defined by theirs, has adopted a costume, and that the costume displaced something.

What the institution sells

Ask what a university sells and the answers arrive immediately and mean nothing. Excellence. Transformation. Outcomes. A strategic plan with four pillars.

This is not evasion by individuals. It is the designed output of an apparatus built over a century to make the question unnecessary. Rankings, placement rates, time-to-degree, endowment growth, alumni giving: each is a real number, and not one of them requires anybody to hold a defensible position on what an educated person is. That is not a defect in the metrics. It is their function. A purpose can be contested. A number cannot.

There was an answer once, and it was not sentimental. The Morrill Act of 1862, the founding document of practical American higher education and the one routinely cited as proof that college was always vocational, funded institutions to promote the liberal and practical education of the industrial classes. Liberal first. The people who built the agricultural colleges did not wonder whether farmers should read. They held that a republic of farmers required farmers capable of governing themselves, and that this was not severable from competence at anything else.

The distinction underneath the whole tradition is older still, and it is not between hand and mind. It is between a person whose ends are their own and a person whose ends are set for them.

That was the product. The institution existed to produce people capable of refusal.

The exception that proves it

There is one activity in which universities behave with complete commercial competence, and it is worth looking at closely, because it is the same institution using the same faculties.

Online delivery is priced at parity with in-person instruction, and frequently above it once distance-learning, technology, and proctoring fees are added. Enrollment caps rise, because there is no room to fill. Instruction is assigned disproportionately to contingent faculty at a fraction of the cost. In a significant number of cases the institution has contracted a revenue share with a private online program manager, which means a public university is dividing federally backed tuition with a marketing firm at an agreed percentage.

The access argument for online instruction is real and should be conceded without qualification. A working adult with a shift schedule cannot attend a seminar at two o'clock on a Tuesday, and for a great many students asynchronous delivery is the difference between a degree and none. But the access argument makes a prediction about price. If access were the reason, the price would follow the delivery.

It does not. And the cost falls on the students with the least margin for a degree that does not pay off: working adults, military students, rural and first-generation students. They are also the ones with the greatest need for the part that cannot be transmitted asynchronously, which is the argument, the objection, the person across the table who thinks you are wrong and will not let it go.

One question settles the institution's private view of its own product more efficiently than any outcomes literature. Would you place your own child in the online section, if a seat in the seminar were available? The institution prices the two as equivalent, staffs them unequally, and markets them as identical. Those positions cannot all be held honestly at once.

When this is raised, the reply is familiar. The student is an adult. The student chose the modality. The student is a willing market participant.

That sentence is doing the same work as that is simply how the market operates. It launders a decision through an abstraction so that nobody has to sign it. And it is offered on behalf of a nineteen-year-old carrying non-dischargeable debt, advised into the modality by a salaried employee of the party collecting the money.

The institution is not incompetent at business. It is competent at business exactly once, and the occasion requires selling students something its own staff would not choose for their children.

When refusal was required

An institution whose product was the capacity for refusal was, over the last three years, asked to refuse. The record is a matter of public documents.

Texas Senate Bill 18 (2023) did not abolish tenure. It defined it in statute as an entitlement to continue in one's position "unless dismissed by the institution for good cause in accordance with the policies and procedures adopted by the institution," which is to say by reference to a document the employer writes. Section 4 of the same act reads, in its entirety:

"Section 51.942(d), Education Code, is repealed."

Subsection (d) had provided that a faculty member facing termination on the basis of a performance evaluation could be referred to alternative dispute resolution, and had required the governing board to state its reasons in writing. Nine words removed the hearing and the obligation to explain. Nothing was violated. A sentence was deleted, by ordinary legislative process, with a comment period.

Senate Bill 37 (2025) then restricted grievance and discipline decisions to the president, the provost, system administration, and their designees; made faculty senates advisory only; and abolished every existing senate on 1 September 2025 unless a governing board affirmatively voted to retain it. Where senates are reconstituted, only the board may create them, and the president appoints half the membership and all officers. When such a body meets on a vote of no confidence, the names of members in attendance are recorded.

The institutional response, where it was articulated at all, took a consistent form. The research enterprise had to be protected. There was a fiduciary duty. The deal had to be taken. This is the provost's register of that is simply how the market operates: the same passive construction, performing the same function, which is to relocate a decision into a condition so that no person is answerable for it.

Columbia agreed in July 2025 to pay the United States two hundred million dollars in installments plus twenty-one million into a claims fund, while expressly denying liability. The agreement provides that nothing in it shall be construed as authorizing the government to dictate faculty hiring or the content of academic speech. Seven paragraphs later it requires review of the regional studies programs, names six of them, extends to "all aspects of leadership and curriculum," and mandates a standard review process for hiring non-tenured faculty. Neither provision violates the other. One is a rule of construction and the other is an obligation.

Compliance is now reported twice yearly to an outside monitor through an administrator answerable to the president. The April 2026 report runs forty-three pages across twenty-three tracked obligations, and records the payment as "Satisfied to Date." The report's own legend distinguishes the two statuses:

Complete: requirement has been fully executed.

Satisfied to Date: requirements met through the current reporting period; obligation is continuing.

Asked what it was for, the institution produced a compliance dashboard. This is not hypocrisy. It is the only language remaining. After a century of pricing itself by rankings and placement rates and endowment growth, a number was the only kind of claim anyone inside could still agree was real. Everything the institution could defend had one attached.

Harvard did not sign and has not settled, and its president has expressed no regret. That is worth recording, and it is worth being honest that it proves less than it appears to. Harvard negotiated for months, and Harvard has an endowment. Whether the refusal was principle or solvency is not answerable from the documents.

They did not sell the thing. Selling requires knowing what you hold. They forgot.

And we let them

None of this was done to a passive public. Every metric exists because somebody demanded it. Legislatures funded by outcome and defunded by outcome. Employers made the degree a hiring filter and then expressed surprise when it optimized into one. Parents asked about starting salaries. Everyone asked eighteen-year-olds what they were going to do with it.

The reason the question was never pressed is not that it was hard to ask. It is that the honest answer would have cost the people asking.

If universities had said plainly that they exist to produce people whose ends are their own, that claim would have had to be wanted, and funded, without a placement-rate table attached. It would have required accepting that some of the most valuable people such an institution produces are inconvenient, unmarketable, and difficult on purpose, because that is the point rather than a side effect.

That is expensive. So the substitute was accepted: excellence, the four pillars, the strategic plan. The amnesia was mutual and convenient, which is precisely why it went unremarked. Every party to the exchange received something. The only party that lost had not arrived yet.

They forgot what they were. We stopped asking.

One of those is repairable immediately, by anyone, at no cost.

The next time the word excellence appears in a strategic plan, a brochure, a dean's remarks, an institutional homepage, ask what it means. Require the sentence to be finished.