The Bifurcation (of Higher Education) Continues Apace

What the speedrun degree, AIM's rulemaking, and the WICHE cliff are all telling us at once

Christie Williams finished her bachelor’s degree in three months. Then she came back and did a master’s in five weeks. The two degrees together cost her just over four thousand dollars. She’s a human resources executive in North Carolina, she did the coursework at night after her day job, and she now runs a side business coaching roughly a thousand other students on how to do the same thing. She charges them $99 to $199 apiece.

When the Washington Post’s Todd Wallack reported on his conversation with her this weekend, her line was: “Why wouldn’t you do that? It’s kind of a no-brainer if you know about it.”

Williams is kinda right, eh? And I think that should worry us more than the speedrun itself.

Degree hacking is higher ed’s bifurcation made legible. For years the sector has been telling itself that the institution does real work with these students’ four years: forms them, adjudicates their knowledge, trains their judgment. What Williams and the thousand people she coaches have figured out is that, for a large slice of American higher education, none of that is actually bundled into the credential anymore. You can unbundle it yourself, one eight-week session at a time, for about $1,800 apiece. At the end of that, you walk across a stage in Maine with a diploma that looks exactly like one that took a kid four years and $120,000 to earn. The transcript is real. The credential is accredited. An employer is going to process it a similar way—and yes, there is a prestige flag that (perhaps) you’re losing, but this is happening.

And if you’re still telling yourself this couldn’t be happening, consider: Larry Schall, the president of the New England Commission of Higher Education (which accredits the University of Maine system), told the Post reporter that he had never heard of students completing a bachelor’s degree in a few months. Then he got off the phone and called Presque Isle to ask them to look into it.

The regulator learned from the press. I mean, wow.

Let me head off the first objection right now, because it’s fair. Degree speedruns are not the modal BA. Out of roughly two million bachelor’s degrees awarded in the US each year, a few thousand get finished in under a year. NCES doesn’t even break out sub-year completion rates because the volume has been too small to measure. So treating this phenomenon as diagnostic of a sector with 1,556 four-year institutions can sound like overreach on a viral anecdote.

Ok, fair. What makes the speedrun worth paying attention to has little to do with volume. A product the sector has been selling for fifty thousand dollars a year is getting pulled off the shelf and rebuilt for two thousand dollars a module, with an accredited diploma at the end and employers who don’t flinch. That kind of gap between sticker and function is what people used to call a stress point in a category. The hackers are a small segment. The segment that could plausibly hack, if they wanted to, is a lot bigger…and growing.

---

So let me walk through how this actually works, because it matters that it’s legal and mostly cheap. The University of Maine at Presque Isle runs a competency-based program called YourPace. Tuition is $1,800 per eight-week session, unlimited classes. If you can prove you’ve learned the material, typically by passing tests or turning in papers, you finish the course. No class meetings. No group discussions. No weekly pacing. Of the roughly three hundred students who earned bachelor’s degrees from YourPace in fall 2024, most finished in under a year. More than a quarter finished the entire degree in a single eight-week session.

Western Governors University is running the same playbook at massive scale: 192,613 students enrolled as of last summer, 395,000 alumni, 445,000 degrees awarded. Their own literature says students finish bachelor’s degrees in an average of 2.5 years, “many significantly faster.” Serenity James, a twenty-five-year-old mom in Atlanta, knocked out sixteen courses on an online learning platform in twenty-two days, then finished her BA at WGU in two months and an MBA in two-and-a-half. All in, under nine thousand dollars, covered by scholarship, employer tuition benefit, and a federal Pell Grant. Her health-insurance employer promoted her into a higher-paying role when she was done. “It’s the best thing that’s ever happened to me,” she told the Post.

James did exactly what the system told her to do. She needed a credential. She found one she could afford. She learned the material, passed the assessments, and handed the paper to an employer who processed it without blinking. The fact that it took her three months instead of three years is, from a student’s point of view, pure upside. What part of that is her fault?

The unbundling was already in motion. The hackers priced it. The market for a stacked, modular, fast-tracked BA exists because the sector has spent a decade signaling that this trajectory is fine. For the first time, as of January, a majority of US job postings on Indeed don’t require a college degree. The National Student Clearinghouse counts 43.1 million Americans with college credit but no credential — James’s demographic, and Williams’s. 37% of college presidents told Inside Higher Ed they plan to add or expand apprenticeship pathways in the next three years. Registered apprenticeships more than doubled between 2014 and 2024. In the last twelve months, nearly sixty institutions have launched or piloted three-year bachelor’s degrees. Utah’s whole public system is moving that way; Ensign College is converting all ten of its bachelor’s programs to three-year formats this spring.

Universities are already running the hack themselves. They’re just slower at it and don’t market it on YouTube. (Yet.)

(And, then, well, I think you have to remember, as I have to remind myself multiple times a day, “Yo, get it through your head that AI will never be worse than it is right this moment.”)

---

OK. Does the hack work everywhere?

No, and this is where the university map I’ve been building does some of the analytical heavy lifting. Here’s what falls out when you place 1,556 US four-year institutions into a two-dimensional frame on institutional resilience (endowment, revenue diversification, enrollment trajectory, selectivity) and post-college market position (completion, earnings-to-debt, AI exposure, demographic trajectory). Eighty-two percent of R1 universities land in the High Capacity quadrant. Fifty percent of baccalaureate institutions land in High Stress. The master’s tiers (both large-medium and small) sit in between, with 43 to 46 percent in High Stress themselves. And most strikingly: 581 institutions, thirty-seven percent of the mapped universe, are simultaneously losing enrollment and scoring below median on institutional resilience.

Those 581 are the population the hackers can eat first.

I argued a couple of weeks ago that the sector’s structural divide cuts across Carnegie tiers rather than tracking the old prestige ladder. Tier alone explains about a quarter of the variance in where institutions end up in the map. Within-tier variation beats between-tier variation in the middle of the distribution. The map is a story about which institutions have buffers left and which have already run through them, which have room to adapt and which are running out of runway.

Now lay the degree-hacking phenomenon on top of that.

The map is shaped as a gradient. The paper’s sensitivity analysis shows only about a third of institutions hold their quadrant across every alternative specification; the rest are boundary cases. The move that stands up is the tails: they’re pulling apart, becoming different kinds of products sold to different markets at different price points.

The credential at High Capacity and the credential at High Stress share a name. They share less and less substance.

At High Capacity — the 82% of R1s, the brand-name selective privates, the well-endowed flagships — the degree is still bundled. It’s a transcript, it’s prestige, yes, and it’s also a four-year scene with some remaining knowledge coordination, a network that will open your emails in ten years, and a signal employers read as shorthand for this person cleared an admissions moat and spent four years around other people who did too. Be honest: a chunk of that signal is admissions sorting, not pedagogy. Dale and Krueger’s work suggests the elite earnings premium shrinks a lot when you control for selection. The unhackable thing at High Capacity may be less about what the institution does and more about who it let in. Either way, you can’t hack it. You can stack Sophia and StraighterLine courses until your eyes bleed and you still won’t have cleared the sorting or been in the room.

At High Stress (the 581), the credential has thinned to mostly transcript. The institution has shed much of the scene under fiscal and enrollment pressure. It’s running online-heavy programs to stay solvent. It can’t pay adjuncts enough to hold a coherent curricular culture. What’s left is the credential. And when what’s left is the credential, the credential is hackable. The hackers are buying exactly what’s being sold.

“Bifurcation” is the coarse version of what’s actually happening. If you look at what students and employers are transacting over right now, the credential category is fragmenting into multiple distinguishable products: elite mentorship bundles at the top, flagship hybrid models (Arizona State and a handful of the larger publics) that run scale online alongside R1 research, regional traditional bundles under mounting fiscal pressure, competency-based credential factories like WGU and YourPace, modular self-assembled speedruns like the ones Williams and James did, and — outside the degree category entirely — apprenticeships and employer-training pathways that skip the BA. These used to be variations on one product. They’re becoming distinct products, differently priced, differently delivered, differently handled in the labor market.

Marjorie Hass, who runs the Council of Independent Colleges (the association representing more than six hundred small liberal-arts institutions), said it to the Post more sharply than I would have dared: “I would prefer to have some of these degrees called something other than a bachelor’s.” That’s the liberal-arts lobby, not a doomer blogger, saying the product category has fractured.

And the institutions that built the hack are already fighting with each other about which version of it they want to sell. Purdue Global pulled back on its unlimited-course ExcelTrack program in January, citing “academic integrity and the value of a Purdue Global degree” as the reason. That phrase is the tell. WGU, meanwhile, defends the model and declined to give the Post any data on completion speeds at all. That’s bifurcation inside the credential-factory tier, between institutions trying to hold a legible brand on top of the hack product and institutions willing to let the hack product be the brand. The University of Maine itself is running a two-product strategy under one name: YourPace is restricted to students twenty and older specifically so it won’t cannibalize the traditional on-campus experience they’re still selling to eighteen-year-olds. One institution, two credentials. They drew the line in pencil.

The map shows you where each institution sits. The hackers are showing you what the map means in practice.

If you want to see where your own institution lands, the interactive map lets you filter by tier, region, and component scores. The working paper, replication materials, and the full 1,556-institution dataset are there too.

---

Now look up.

While the hackers are pricing the bifurcation from below, the federal and state governments are hardening it from above, on a timeline that’s going to matter a lot for anyone trying to adapt.

The Department of Education’s AIM Committee (Accreditation, Innovation, and Modernization) held its first session of negotiated rulemaking April 13–17 and votes on regulations May 18–22. The charter includes refocusing accreditation on “data-driven student outcomes” (ROI, completion, placement, licensing) and lowering barriers for new accreditors. Inside Higher Ed flagged concerns about academic freedom, and those concerns are real, but focus on what the rewrite is for: it’s the federal government moving to codify the hackers’ ROI logic into the accreditation standard. AIM would formalize in rule what Williams already figured out on her own.

At the same time, the new endowment tax is squeezing the supposed unhackable tier. Harvard owes $368 million. Yale, $280 million. Princeton, $217 million. Real money even for Harvard. Yale has already cut PhD enrollment thirteen percent in the arts and humanities and five percent in the sciences. MIT is looking at roughly $300 million a year in hits, closing library service desks, ending office leases, freezing raises above $85K. The cuts at these places hit the reproduction of the discipline: graduate students who would become the next faculty, library and research infrastructure that supports the scholarship. The credential holds. The pipeline doesn’t. The unhackable tier is having its seed corn taken even as its degree remains unhackable. Different damage, same direction.

And then the governance side. Kentucky’s HB 490 sits on Andy Beshear’s desk; it gives governing boards “for cause” dismissal authority over tenured faculty. Oklahoma is ending tenure for new faculty hires across every public institution in the state except the two research flagships (OU and Oklahoma State) and their health-sciences campuses, with new contracts capped at five years — which is to say, tenure will survive at the top of the Oklahoma system and disappear everywhere else. Alabama and Tennessee have similar bills moving. Whatever you think of tenure as an institution, it is the mechanism that made long-arc scholarship possible at public research universities, and it is being politically restructured in multiple red states right now, on a compressed timeline.

Even the inside-the-walls diagnosis has arrived. Yale’s Committee on Trust in Higher Education dropped a fifty-eight-page report on April 10 (the same week the Post ran the degree-hacking piece) with twenty recommendations and a Yale president explicitly acknowledging that universities helped erode the trust they now lack. Gallup has public confidence in higher ed at 42 percent, up from a historic low of 36 percent, but with a forty-point partisan gap: 66% of Democrats, 26% of Republicans.

The four pressures have different political valences and different intellectual premises. AIM is a second-term Trump rulemaking agenda. The endowment tax is a populist revenue move. The red-state tenure bills are culture-war products. Yale’s committee is left-of-center elite institutional self-critique about ideological narrowness. These actors don’t agree with each other about much. What they converge on, functionally, is a set of effects: more outcomes-based accountability, fewer resources for research reproduction, weaker faculty autonomy, more public accountability for what institutions are selling. Different actors pushing for different reasons end up pointing the sector in a common direction. That’s structural convergence, not ideological alignment.

From below, the hackers. From above, AIM and the endowment tax and the tenure bills. From inside, Yale acknowledging it helped torch the trust that used to insulate the whole enterprise. Four different vectors, one common direction. The structural divide the map documents is moving from forecast to observation in real time.

---

OK, slow down. A piece of this isn’t obvious.

The hack works, where it works, because the institution’s epistemic coordination function has thinned out. I wrote about this in January under the heading “the collapsing value proposition” — the argument that universities are, at their best, places where contested knowledge claims get adjudicated and disciplinary standards get transmitted by proximity to people who know them in their hands, not just on a rubric.

A professor making this argument owes the reader a self-audit. I am a tenured professor. The frame “universities have a precious epistemic function and it’s being hollowed out” is exactly the self-image someone in my position would reach for when facing public distrust. I’ve thought about that. The modal BA was thinner than its price-tag implied in 1975 too, so this isn’t a golden-age complaint. The specific structural conditions that used to carry a non-trivial formation layer — small section sizes, full-time faculty doing the teaching, meaningful office-hour contact, coursework that assumed a shared peer community — have been eroding steadily for thirty years under adjunctification, online conversion, and enrollment-management math. The hollowing has been continuous. What’s new is that the credential was still priced as if the formation layer was intact, and the hackers have noticed.

When a student goes from her living room to a WGU transcript without ever talking to a faculty member about a hard question, she hasn’t failed to get an education. She’s bought exactly what the institution is selling her. The question that raises is whether these hackers are accurately pricing the amount of knowledge work a particular institution is actually still doing. And that is a dark, dark question for most in academe.

Think about what we measure when we evaluate a class or even a degree program. We measure competencies: can you pass this test, can you produce this document. We don’t measure whether the student spent two semesters being wrong about something in a seminar before they got it right. We don’t measure whether they developed a nose for bad arguments. We don’t measure whether they met the person who’ll write their reference letter in seven years. The hack works because our measurement apparatus only checks the things that were always hackable. The rest — the scene — we assumed was in the bundle.

It is. Sometimes. Not always. Increasingly, not much.

A second phenomenon here needs to stay distinct from the first. Degree hackers pass real (if thin) assessments on material they actually know. AI cheating is a different story: students faking assessments on material they don’t. Both stress the same apparatus, which is the certification layer. The Turnitin failure rates are running around 94% undetected for AI-generated work. More than fifty institutions, including Vanderbilt, Johns Hopkins, UCLA, and UC San Diego Extended Studies, have disabled Turnitin’s AI detection because it’s inaccurate and over-flags non-native English speakers. Princeton’s faculty are seriously considering universal proctoring on all examinations, which is approximately what you do when you’ve given up on formative assessment altogether — and that’s Princeton. High Capacity. The assessment crisis is hitting the elite tier too.

Put them together and you have a stress test on the whole apparatus: the credential price overshot the knowledge work, and even where the work did happen, the certification that it did is becoming harder to verify.

---

The sector is going to have to pick. Most institutions can’t credibly run all three pathways at once, and the math isn’t going to let most of them try. I sketched three possibilities last month: the Credential Factory (accept the hack, compete on speed and price, build around competency-based assessment honestly), the Elite Mentorship Institution (sell the formation layer unapologetically, lean into small-cohort teaching and dense network effects, accept the smaller footprint this requires), and the Hybrid Knowledge Platform (keep a core formation experience while modularizing the rest).

The three-category frame simplifies a continuous strategic space, and some institutions do look like they’ve pulled off all three. Arizona State is the obvious case, with 150,000+ online learners alongside R1 research and an honors college. ASU is the exception that proves the rule: it took Michael Crow twenty years of deliberate institution-building, a state that let him do it, and a scale most institutions don’t have access to. For the other 1,500-plus, the pick is real.

The map tells you which picks are realistically available at a given institution’s coordinates. A baccalaureate in a shrinking Midwest county is structurally unlikely to become the next Williams College — it can certainly try, and Berea and Olin and College of the Ozarks show that small distinctive-identity moves remain possible at the margin, but the structural headwinds are steep. A flagship R1 can’t credibly compete with WGU on speed and price; doing so would set the endowment on fire. Coordinates constrain pickability. They don’t determine it. Constraint tightens as resources shrink. That’s also what the map is for.

And the part we don’t really talk about in faculty meetings: most institutions haven’t picked. They’re still running a 1985 bundle on a 2026 budget. Portland State declared a financial crisis in March: nineteen departments up for reduction or closure, $35M deficit, headcount down 23% since 2019. University Studies, Conflict Resolution, and the Portland Center are on the closure list. Sterling College in Vermont is closing this semester at forty students, down from a peak of 120. A recent Huron Consulting projection puts 442 of the country’s 1,700 private nonprofit four-year institutions at risk of closure or forced merger within ten years, with 120-plus at highest risk.

And the demographic math underneath doesn’t let up. WICHE’s projections have US high school graduates peaking this year at 3.9 million and declining 13 percent through 2041. The decline is regional, not uniform. The state-by-state data has Illinois losing 32% of its high school graduates by 2041, California 29%, New York 27%, West Virginia 26%, Michigan 20%, Pennsylvania 17%, with Hawaii the steepest at 33%. The Midwest aggregate drops 16 percent; the Northeast 17 percent; the West 20 percent. The South gains 3 percent, which is not going to save the Midwest. Overlay that on the university map and the overlap is brutal: the states facing the steepest demographic declines are where the High Stress institutions concentrate. Two structural stresses stacking on the same populations.

The K-12 sector is already living a version of this math. Bellwether’s analysis shows enrollment declines (5.5% projected through FY31, after a 14% birth-rate drop the decade before) hitting per-pupil funding while fixed staffing and facility costs don’t adjust. Districts built to expand are contracting under duress. Eight states are projecting K-12 declines over 10 percent — where the kids who’d be college freshmen in 2036 are coming from, and where most of them aren’t. K-12 is higher ed’s leading indicator. What’s reshaping school-district budgets now is what will reshape campus budgets a decade from now, on top of everything AIM and the endowment tax and the tenure bills and the ratings agencies are already doing.

So, I would argue that the hackers are just early. They’ve priced what AIM is about to ratify, what the WICHE cliff is turning into math, what Portland State is already living. The sector’s restructuring is going to finish whether faculty senates opt in or not. The only variable left is whether the pick gets made with institutional intent or gets made for an institution by exigency.

My read: if you’re a faculty governance body, a provost, a board chair, and you don’t have a strategic answer to “which of the three pathways are we trying to be, given our coordinates” that you could defend in front of your board, you are late. Not panicking late. Not irrecoverably late. Late. And I know the sector has heard warnings like this for decades without much apparent urgency. The structural difference this time is that there are deadlines attached. AIM’s rulemaking concludes May 22. The endowment-tax compliance year starts July 1. The WICHE cliff peaks this year and then gets worse for sixteen straight years. The clocks have started.

---

Christie Williams is walking at the University of Maine at Presque Isle’s graduation next month. She’s never been to the campus. She’ll be walking with her twenty-two-year-old daughter, Makayla Quackenbush, who finished her own bachelor’s at Presque Isle last summer, summa cum laude, in a single eight-week session.

Mother and daughter. Both speedrun BAs. Both credentials real, both accredited, both together costing less than a single year of the private liberal-arts tuition Williams’s generation used to consider the gold standard.

That’s a mother teaching her kid what the sector is actually selling, and a kid taking her mother’s advice. A generational transmission of an incentive response. The most ordinary thing in the world.

The sector I’ve spent my career inside has trained two generations of one family to buy its product. What that product is, and what it was supposed to be, will define the conversation for the next decade: in faculty meetings, in state capitols, in AIM’s rulemaking rooms, and in a thousand Reddit threads where someone is debating whether their WGU transcript should come with an asterisk.

I don’t think we get to duck this one much longer, friends.

So, the bifurcation is indeed underway. Institutions can pick a pathway now, with intent, or have one picked for them later, by exigency. Those are the options.

Pretending the category still holds as one product is no longer available.

---

Sacred Cow BBQ covers higher ed, politics, institutions, and the trust questions underneath. If this was useful, subscribe for more in this vein, and the interactive map is there if you want to explore where your institution lands. If you know a provost, board chair, or faculty governance chair who’s been avoiding the pathway conversation, forward it their way — that’s where the decisions are being made. The more informed conversations we can have in and about higher ed about this, the better. And, remember, AI will never be worse than it is today.