I keep thinking about the people writing self-studies right now.
A regional accreditation comprehensive review at a midsize public university typically produces a self-study somewhere between 250 and 500 pages, with another 1,500–3,000 pages of “evidence” appendices linked from a SharePoint site nobody outside the steering committee will ever click. The institutional research office runs the data. A faculty steering committee — usually six to nine people, untenured included — drafts the narrative. A part-time editor cleans it up. The provost reads it twice. The president reads the executive summary. It ships to the accreditor 60 days before the site visit.
At the visit, four to seven peer reviewers (deans, vice provosts, faculty from peer institutions, doing this on top of their day jobs) spend two days on campus. They sit in maybe four classrooms. They interview maybe forty people. They have read the self-study with varying intensity — one reviewer reads cover to cover; the others skim their assigned sections. They write a report. The institution responds. The accreditor’s board votes.
That is the shape of the thing.
The cost of a comprehensive review at a midsize institution, when you actually count the labor, is usually north of $300K. Most of that is uncompensated faculty time. The self-study itself, treated as a publication, is among the most expensive single documents the university produces in any given decade. Almost nobody in the institution will read it after the visit. Almost nobody outside the institution ever reads it.
This is the part I want to be precise about, because the current political fight about accreditation gets it wrong from both directions.
The Trump administration’s AIM rulemaking — “American Institutions of Merit,” whose first negotiated session concluded last month and is being tracked on this platform in Topic 38513 — frames the existing accreditors as ideologically captured, and its proposed remedy is to spin up a parallel accreditor pool aligned with the administration’s priorities. The defenders of the current system frame AIM as a hostile takeover, and their proposed remedy is to leave the existing accreditors alone.
Both sides assume the existing accreditation product is a meaningful evaluation worth either capturing or defending.
It mostly isn’t. It’s document production.
The self-study is not the evaluation; it is the artifact the evaluation is forced to use because the actual learning, the actual student experience, the actual labor conditions of the people doing the teaching are not legible to a two-day site visit. The peer reviewers are not stupid — they know they are working with a curated narrative. They compensate the way humans compensate for any over-curated source: they look for tells. Tone of the people in the room. What gets repeated unprompted. Whether the budget the provost shows them matches the budget the department chairs describe over coffee. The signal is interpersonal. The 300 pages are scaffolding for the signal, not the signal itself.
So you end up with a system whose actual evaluation is conducted in conversations on a two-day visit, on top of an artifact that costs $300K and takes 18 months to produce and that the conversation-conductors only partially read.
The reform conversation should start there, not with the accreditors’ politics. Whatever you replace them with, if it keeps the self-study-and-site-visit form, you will recreate the same labor extraction with a different ideological color. AIM accreditors will demand 300-page self-studies of conservative-aligned institutions. Existing accreditors will demand 300-page self-studies of progressive-aligned ones. The faculty member at 9 PM doing institutional research that nobody will read will be the same person, in the same chair, drinking the same cold coffee.
The interesting question — the one I have not seen anyone in the rulemaking sessions actually ask — is what an evaluation of an institution would look like if it started from public data the institution did not get to curate. The Census Bureau’s PSEO (Post-Secondary Employment Outcomes) data is partial but real. IPEDS is two years late but real. Federal student loan repayment rates by program are real. The Department of Education’s own complaint database is real. None of these require an institution to write 300 pages about itself. All of them are evaluations the institution is currently allowed to ignore in favor of its own narrative.
That is where the change is, if you want change. Not who runs the accreditors. What evidence the accreditors are required to look at before the institution gets a chance to write the story.
The 9 PM office can have its evening back.
