Every year, more than $110 billion in federal student aid flows through American higher education. Almost none of it reaches a student without first passing a checkpoint most Americans have never heard of. To award Pell Grants or federal loans, a college or program must be accredited by an agency that the U.S. Department of Education recognizes. The accreditor decides which institutions may exist within the federal financing system and, in many fields, which graduates may enter a profession at all. It is one of the most consequential gatekeeping powers in American economic life, and for most of the last half-century it has operated with remarkably little outside scrutiny.
Get Liberalism.org in your inbox.
That is beginning to change, and the reason is not ideological enthusiasm, but evidence. My colleague Bryan Cutsinger and I have assembled the financial and governance records that let us ask an old public-choice question with new precision: are accreditors independent quality assurers, or have they been captured by the very interests they are supposed to discipline? The answer, documented in the accreditors’ own tax filings, is that capture is not an occasional failure of the system. In important corners of it, capture is the system.
This essay traces how that came to be, what the records show, and why the structural fusion between accreditors and the professional associations they serve is a problem for anyone who cares about open markets, intellectual pluralism, and the rule of law over the rule of organized interest. It closes with the most encouraging development in this domain in a generation: a recent negotiated AIM (accreditation innovation and modernization) rulemaking at the Department of Education that produced consensus language to begin prying the gatekeepers loose.
The final red-lined AIM draft tightens the federal test for whether an accreditor is genuinely independent of the profession it polices, so that the people who decide accreditation can no longer be selected by, or drawn from the staff of, the trade association whose members they judge, and it requires every accreditor to disclose, prominently and publicly, its relationships with any affiliated association. The new version forbids an accreditor from acting to restrict access to a profession unless it first shows the Secretary of Education, by clear and convincing evidence, that the restriction is necessary to protect the public, that its benefits outweigh the costs of reduced access, and that no less restrictive alternative would do. And it defines restricting access to include the very moves a captured accreditor relies on: raising credentialing standards, increasing the cost or length of required training, and reducing the supply of programs in ways that happen to benefit the affiliated association. For the first time, the burden of justifying a higher gate would rest on the gatekeeper, not on the student or the new entrant shut out by it.
Accreditation in the United States began, sensibly enough, as a voluntary and private affair. In the late nineteenth and early twentieth centuries, regional membership associations of colleges formed to articulate shared standards and to vouch for one another’s seriousness in an era with no common definition of what a college or a high school diploma meant. Professional fields developed their own specialized accreditors for the same reason: medicine, after the Flexner Report of 1910, then law, engineering, nursing, and eventually dozens of others. In its origins, this movement sought a genuinely liberal, bottom-up solution to an information problem. Membership was voluntary, the standards were the profession’s own, and an institution that disliked an accreditor’s judgment bore only reputational consequences.
The decisive change came when the federal government decided to use this private machinery for a public purpose. The Higher Education Act of 1965 anchored a system in which eligibility for federal student aid still runs through accreditation by a federally recognized agency. A voluntary reputational signal has become a compulsory financial gate. The accreditors did not have to seek out their members for review; their members now had to come to them, because the alternative was exclusion from the dominant source of higher-education revenue in the country. What had been a club became a tollbooth.
This is the hinge on which the whole story turns. A voluntary standard-setter that loses its members’ confidence loses its members. A federally deputized gatekeeper does not. Once accreditation became the precondition for federal funds, the ordinary market discipline that would justify private standard-setting evaporated, and the accreditor acquired exactly the feature that public-choice economists warn about: durable power over entry, insulated from the consequences of misusing it.
Over the following decades, two further dynamics hardened the arrangement. First, in profession-specific fields, the accreditor and the trade association representing incumbent practitioners frequently turned out to be the same people, or nearly so, often sharing boards, staff, revenue, and in some cases a single physical address. The body certifying that a program adequately trains new entrants to a profession was operated by the established members of that profession, who have an obvious interest in the supply, prestige, and price of their own credential. Second, the requirements themselves usually ratcheted upward over time, and almost never downward. Each new mandated course, standard, or required credential raised the cost of producing a graduate and of becoming one, with the burden falling on students and on any would-be competitor trying to offer a leaner, cheaper, or simply different path to the same competence.
George Stigler’s classic 1971 account of regulatory capture described precisely this equilibrium: regulation is frequently acquired by an industry and operated for its benefit. What makes accreditation a textbook case is that the regulator and the industry need not even collude across a table. In the captured corners of the system, they are designed to be joined at the hip.
Capture is easy to assert and hard to prove. The research I have been doing with economist Bryan Cutsinger has moved the conversation from anecdote to measurement, using sources that are difficult to wave away because the accreditors and associations produced them themselves.
The starting point is the IRS Form 990, the annual return that tax-exempt organizations must file. Assembled into a panel across years and organizations, these filings let us trace the financial and governance entanglements between accrediting agencies and their affiliated associations: shared officers and directors, payments flowing between the two entities, common management, and overlapping addresses. From these we build what we call a Regulatory Capture Index, which is a composite measure of how structurally fused a given accreditor is with the association whose members it accredits. The point of an index is to make capture comparable: to say not merely that entanglement exists, but to rank where it is severe and where it is mild.
What the index reveals is not a handful of bad actors, but a modal arrangement. Across thirty-two programmatic accreditor–association pairs, twenty-eight are not meaningfully separate from the association whose programs they accredit. Sixteen are the same tax-filing entity as the association, sharing a single Employer Identification Number; another twelve are unified bodies in which the membership organization and the accreditor are constitutionally identical. Twenty-eight of the thirty-two share a registered street address. Only four pairs in the entire sample maintain genuinely separate legal identities. In the health professions, accreditation is the gateway to a state license, and fifteen of the accreditors we examined sit at the maximum value of the index in every year we observe.
That last point matters more than any single number. We built the index across a seven-year panel precisely to see whether capture ebbs and flows with personnel or politics. It does not. The mean score moves by less than two-hundredths of a point across the entire period, and no accreditor in the sample changes its structural category at all. Capture here is not a behavior that good people can resolve by trying harder. It is welded into charters, tax filings, and shared payrolls, which is exactly why it persists regardless of who sits on a given board in a given year.
An accreditor structurally embedded within the profession it oversees cannot be a reliable independent check on that profession’s standards. This is not because its staff are dishonest, but because the institution is built to serve the incumbents it is supposed to discipline. The conflict is in the architecture, and no individual’s good faith can engineer it away.
The financial picture tells the same story from another angle. For the great majority of accreditor-institution pairs, there is simply no arm’s-length transaction to examine, because the accreditor’s revenues, expenses, and staff are line items inside the association’s own return. Where the two entities do file separately, the linkages surface anyway. In one case, an accreditor’s parent association reported a single-year management fee of more than $3.5 million flowing through the accreditation relationship, which is roughly a tenth of that association’s total revenue. And the associations behind these accreditors are not minor players: the twenty parent organizations we could fully observe command more than $600 million in combined annual revenue and over $1.1 billion in assets. That is the financial weight sitting behind the gate.
The pattern is starkest in legal education and in the licensed health professions, such as dentistry, nursing, dietetics, and physical therapy, where graduation from an accredited program is a precondition for entering the field at all. In the most tightly fused arrangements, the sole national accreditor of an entire profession’s training programs is not even separately incorporated. It operates as a commission or division of the trade association representing that profession’s established practitioners, sharing the association’s tax identity, staff, and headquarters. The body that decides which new programs may train the next generation of practitioners is, in legal and financial substance, the incumbent profession itself. Whatever expertise that arrangement supplies, and it does supply real expertise, it cannot supply independence, because there is no independent entity there to begin with.
A careful reader will ask the obvious next question: does this structure produce worse behavior, that is, fewer sanctions for weak programs, tighter restrictions on new entrants, or higher costs? Our paper is deliberately disciplined on this point. It documents the structural conditions under which capture is likely; it does not yet measure the downstream decisions. Connecting the capture index to adverse-action rates, to workforce supply, and to the price of credentials is the central task of ongoing work. But the structural finding stands on its own, because in regulation we do not normally wait for the damage before we object to the design. We do not let an auditor hold equity in the firm it audits and then reassure ourselves with a conflict-of-interest policy. The burden of proof should run the other way.
Put the legal, financial, and governance evidence together and the conclusion is hard to avoid: across most of the system, the accreditor and the association are not two parties in a relationship. They’re one entity wearing two hats. The independence the federal framework assumes is, for the modal accreditor, simply absent by design.
It would be possible to treat all of this as a narrow question of administrative tidiness. It is not. Accreditor capture restricts freedom along four distinct dimensions, and naming them is the surest way to see why the problem cuts across the usual political alignments.
It is, first, a constraint on economic freedom. An accreditor controlled by licensed practitioners that sets ever-higher educational requirements is often functioning as an occupational-licensing cartel, except with federal recognition, and therefore federal enforcement, standing behind it. The predictable results are the ones that cartels always produce: restricted entry, higher prices, protected incumbents. Students pay more and wait longer to enter a field; employers face an artificially constrained supply of qualified workers; and the entrepreneur who could train competent practitioners faster or more cheaply is shut out before serving a single student.
Standards do sometimes rise because a field genuinely advances. New science, new technology, and entirely new areas of practice can each warrant more training than before, and a requirement that looks like a barrier may in fact be an improvement. The difficulty is that, from the outside, a protective ratchet and a genuine upgrade can look identical, because both raise the cost of entry. To tell them apart, we must ask three questions of structure: Who sets the requirement, and do they profit from the scarcity it creates? Is the structure defended with evidence that graduates or the public are measurably better off, or only with the authority of the profession? And does the standard ever move downward when new methods make a leaner path to the same competence possible? A genuine quality regime can answer these questions. A captured one cannot, because its requirements move in one direction only, are written by the incumbents who benefit from scarcity, and are justified by assertion rather than demonstrated outcomes. This is why the remedy is not for regulators to second-guess every standard, but to shift the burden of proof. Before an accreditor may raise the gate, it should have to show that the change serves students and the public rather than the people already inside. A real improvement can meet that test. A cartel’s ratchet cannot.
It is, second, a constraint on intellectual freedom. When an accreditor is governed by an ideologically or methodologically homogeneous professional community, it tends to impose that community’s single framework on every program it touches. A graduate trained to understand her field in only one sanctioned way is not better educated; she is more narrowly educated. Intellectual monoculture enforced through accreditation is both a workforce-readiness failure and an epistemic harm, precisely the kind of harm that competition among approaches is supposed to prevent.
It is, third, a constraint on cultural freedom. The institutions most likely to offer genuinely different models, such as new delivery formats, nontraditional structures, or faith-based or mission-driven curricula, are exactly the ones an entrenched accreditor has the least incentive to welcome and the most structural power to exclude. Pluralism in higher education can die, not by prohibition, but by gatekeeping.
It is, fourth and most fundamentally, a constraint on political freedom, in the older sense of governance under rules rather than under the discretion of organized interests. For decades, the process by which accreditation standards were set offered almost no meaningful check on the agencies’ self-dealing. Requiring an accreditor to offer clear and convincing justification before it restricts access to a profession is not a partisan maneuver. It is the ordinary logic of any system that prefers the rule of law to the rule of whoever currently holds the gate.
None of these are sectarian commitments. They are the structural preconditions of an open society, and the case for addressing accreditor capture is strongest precisely because it does not depend on agreeing about anything else. The classical liberal should see an entry-restricting cartel with federal power behind it. The modern liberal should see a closed guild raising the price of opportunity and narrowing who gets to walk through the door. Both are right, and they are describing the same building.
Diagnosis is cheap; durable reform is not. The encouraging news is that reform has now moved from the seminar to the regulatory record.
This spring I spent the better part of two multi-day sessions in a ground-floor conference room at the U.S. Department of Education, seated with representatives of universities, accrediting agencies, student advocates, workforce organizations, and federal officials. This was the Department’s Accreditation, Innovation, and Modernization negotiated rulemaking, and I served as a representative of workforce needs. The form of the reform matters here as much as its content.
Most federal rules are written through notice-and-comment: an agency drafts a proposal, publishes it, collects written comments, and issues a final rule. It is legitimate, but it is asymmetric, because well-organized interests with sophisticated legal teams can dominate the comment record, which is, of course, exactly how captured arrangements perpetuate themselves. Negotiated rulemaking works differently. The agency seats the affected parties at one table and asks them to negotiate the actual regulatory language together, with the aim of genuine consensus that every participant can accept. When consensus is reached, it carries unusual legitimacy. This is not because everyone is fully satisfied, but because the language reflects real give-and-take among parties with sharply different interests.
After nine days of negotiation, the committee reached consensus on changes to the regulatory framework governing accrediting agencies. The consensus draft addresses structural separation between accreditors and affiliated associations, requires public disclosure of organizational relationships, establishes a public-interest test before an accreditor may restrict access to a profession, protects academic freedom and viewpoint diversity in faculty evaluation, strengthens research integrity requirements, and builds in timely due process for accreditation decisions.
Read against the institutional history, the logic of those provisions is unmistakable. Structural separation and mandatory disclosure attack the fusion at the heart of capture. They are the regulatory equivalent of forbidding the auditor to hold equity in the firm under audit. The public-interest test inserts a check exactly where, for half a century, there was none: at the moment the gatekeeper moves to raise the gate. And the academic-freedom and due-process provisions guard against the monoculture and arbitrariness that fused governance tends to produce. These are not provisions that favor one vision of higher education over another. They reduce the structural advantage of incumbency and reopen the system to competition in quality, in approach, in intellectual diversity, and in price.
I will not pretend the process was frictionless. Negotiated rulemaking demands listening to people you disagree with, accepting language less precise than you would prefer, and trusting that the architecture of a rule matters more than any single clause. But it produces something that neither litigation nor conventional rulemaking reliably generates: a shared, concrete understanding of the stakes among the very parties who must live with the result. A rule built that way is more likely to be workable, more likely to survive legal challenge, and more likely to achieve its purpose in practice.
It is tempting to dismiss accreditation as an administrative backwater. That would be a mistake. A system that began as a voluntary, bottom-up answer to an information problem was, over half a century, converted by federal funding into a compulsory tollbooth. And in its profession-specific corners, that tollbooth was quietly handed to the incumbents who profit from keeping the road narrow.
The records now let us say this with confidence rather than suspicion: the structural conditions for capture are not occasional; they are the norm, and they are a feature of the system’s design rather than an accident of personnel. Whether that structure also produces measurably worse decisions is the question the next phase of this research is built to answer. The reform that emerged this spring will not dissolve these dynamics overnight. But it will establish, for the first time in a generation, that the gatekeepers can be required to justify the gate. For a free society, that is the right place to begin.
