At a glance
Enrollment is flattening, grant funding is volatile and restructuring arrives by spreadsheet. How universities are forecasting headcount now, what it costs when they get it wrong, and the questions candidates should ask before signing.
The budget paper reaches the faculty before the announcement does. A student-load tab, an establishment tab, and a gap between them that nobody in the room wants to name out loud.
What the load tab says this year isn't collapse. It's something harder to plan around. Preliminary fall 2026 figures from the National Student Clearinghouse Research Center put total US postsecondary enrollment up 0.2 percent year on year, with undergraduates ahead 0.6 percent — a quarter of the growth recorded a year earlier. Community colleges were the only sector adding students in real numbers. Master's enrollment fell 3.7 percent.
"The important change this fall isn't widespread decline," the Clearinghouse's Matthew Holsapple said as the data landed. "The important change is the loss of that broad-based growth we saw last year."
Read the discipline lines and the calm evaporates. Undergraduate computer science at four-year institutions dropped 12.4 percent. Health majors climbed 6.2 percent. A department can be an eighth smaller while its university is fractionally bigger, and a single institutional headcount target will never tell you which of those you're actually running.
The preliminary number is not the number
Fall 2025 is the cautionary case. The early Clearinghouse read showed total enrollment up 2.0 percent. The final count in January landed at 1.0 percent — 19.4 million students, half the growth the preliminary panel implied. Institutions that froze or opened lines in October on the preliminary figure spent February correcting the correction.
Then the growth itself turns out to be borrowed. Enrollment among students 17 and younger rose 3.9 percent this fall, almost all of it dual enrollment. The Community College Research Center puts dual enrollment at roughly 3.1 million students. Strip those out and undergraduate headcount is already in decline. Those teenagers are also a pull-forward effect: they're counted now, and they won't turn up as freshmen later.
The 18-year-old cohort — the one every staffing model is built on — declined 0.7 percent. Nathan Grawe's projection of a 15 percent fall in the college-age population between 2025 and 2029 has been quoted so often it has stopped functioning as a warning. It should still function as one, because the composition of the decline is more disruptive than its size.
Three inputs that never move together
Enrollment moves over years, and unevenly by geography. The same Clearinghouse panel that shows national stability also shows Vermont down 3.4 percent and Louisiana up 7.8 percent in the same semester. No national headline survives contact with a recruitment region.
The regional picture has shifted since the last forecast round. Analysis of updated WICHE projections by EAB found the 18-year-old decline through 2030 is deeper than previously modelled across the West, Midwest and Northeast — in some cases nearly 5 percent lower than the last estimate — while the Southeast and Southwest finish the decade three to four percent higher than expected. A university that planned on the old national curve is now wrong in one direction or the other, and it doesn't get to choose which.
Grant funding moves on a faster clock. A negotiated indirect cost rate, a federal appropriations line, an agency's discretionary call: any of these can change the affordability of a research line between one hiring committee meeting and the next. Restructuring runs on a third clock entirely — council or board papers, statutory consultation windows, a date on which a decision must be minuted. Treat those three as one forecast and the plan will be wrong on at least one axis at all times.
The ratio is the policy now, so publish it
The most honest workforce plans I've read this year don't lead with a number of jobs. They lead with an assumption.
Aberdeen is the clearest example. Its governing court approved a £10 million savings plan and told staff that future staffing budgets would be informed by indicative student:staff ratios of 20:1 for STEM disciplines and 25:1 for SHAPE. The modelling pointed to a reduction of up to 111 full-time equivalent academic posts. But read what the court actually said: "the focus of the decision made by Court was to reduce our costs, not to reduce our workforce by a specific number." That distinction is real, and it's also the reason staff distrust it. A ratio generates a budget. A budget generates a number. Somebody in an office has already seen the number.
Course-level viability thresholds do the same work more brutally. UTS proposed discontinuing 167 courses and cutting about 134 full-time equivalent academic roles for $80 million in savings. Macquarie targeted 42 arts and 33 science and engineering roles for $15 million net. In the United States, several states have taken the threshold out of institutional hands entirely and written it into law, requiring universities to justify or close degrees that fall below set graduation averages.
Whatever formula you choose, the point is that headcount stops being a judgement call and becomes an output. That's defensible. What isn't defensible is keeping the formula internal while the consequences are public.
Research money is the least forecastable line on the sheet
February 2025 is the case study every research-intensive university should still be studying. The National Institutes of Health announced it would cap indirect cost reimbursement at 15 percent, replacing rates that individual institutions had negotiated and audited over decades. Oregon Health & Science University, with a negotiated rate of 56 percent, modelled a revenue loss of roughly $80 million a year. Its total NIH funding in fiscal 2025 was $309 million.
The cap never took effect. A federal appeals court upheld an injunction in January 2026, and Congress wrote protection for negotiated rates into the following spending bill. But the fourteen months of uncertainty did the damage anyway. Institutions froze staff hiring, delayed graduate admissions offers while they modelled two futures, and quietly declined to replace departing technicians. A department can lose a decade of technical capability without a single redundancy notice.
This is the part of workforce planning that gets least attention and causes the most whiplash: grant-funded roles don't have a notice period, they have an end date. A five-year plan that treats soft money as permanent establishment is a plan built on someone else's budget cycle.
The CUPA-HR workforce survey shows what happens when institutions respond to that volatility at once. Full-time staff headcount fell 6.6 percent in 2025. Part-time staff fell 23.9 percent. Tenure-track faculty headcount rose 7.0 percent — the largest single-year increase in a decade. Look closer and the rise is concentrated in about one institution in ten, nearly half of them doctoral. At associate's, baccalaureate and master's institutions, tenure-track numbers went down. "Faculty hiring is back" is a true statement about perhaps 90 campuses and a false one about most of the rest.
Attrition is the first lever and it isn't free
Dundee has removed 675 posts since August 2024, most through two rounds of voluntary severance, and still needed roughly £20 million more in annual savings. Sussex lost about 500 staff voluntarily before proposing at least 200 more to close a £35 million gap. Voluntary schemes are the humane option and they are also the option that lets a university choose its number without choosing its people: the staff most likely to leave are often the ones with somewhere else to go.
They're expensive too. Across 118 UK institutions, 16,800 staff received severance payments in 2024-25 on top of 11,300 the year before. The total bill came to £441.9 million, an average of £3.8 million per university and an 83 percent rise on the previous year, according to analysis of sector accounts. Nottingham proposed more than 600 cuts on top of 350 the year before, after an £85 million deficit and a warning that it runs out of money by 2031 without intervention.
The cheapest lever is the one nobody announces. Western Sydney included 120 vacant positions in its restructuring — posts it simply stopped filling. No consultation, no severance, no press release. If you want to know what a university's real headcount plan is, ask which approved positions have gone unfilled for more than two budget cycles.
A cap is a number. A visa refusal rate isn't.
Australia spent two years arguing about international student caps and got, in the end, the least bad version of the problem: a known number. The National Planning Level is frozen at 295,000 commencements for 2027, unchanged from 2026 and roughly 8 percent below the post-COVID peak, per the education minister's announcement. Commencements in 2026 are tracking 8 percent below the same period in 2025 and 13 percent below 2019.
The volatility sits underneath the cap, in approvals. The offshore visa grant rate averaged 71 percent in the first half of 2026 against 88 percent in the first half of 2025. A provider can hold a full allocation and still enrol two-thirds of a plan, which is what happened at several institutions forecasting revenue against a number they were told they could reach. You cannot run a staffing model on an approval rate set by an entirely different department.
Which is the whole discipline in one sentence: forecast the thing you control, scenario the thing you don't, and stop writing plans that assume the second behaves like the first.
What a candidate should ask before signing
If you're applying into this market, treat the interview as a two-way audit. Four questions do most of the work.
- Is this post continuing, fixed-term, or funded from a specific grant, and what is the end date on that funding?
- What student:staff ratio or course-viability threshold does the faculty plan against?
- When did the last structural change conclude, and how many of the vacancies created by it have been filled?
- What happened to the person in this role before me?
The answers tell you more than the strategic plan. A department that can name its ratio, its funding source and its last restructure date in plain language is a department with a plan. One that answers all four with variations on "the sector is challenging" is telling you something too.
On the other side of the desk, the same logic governs whether you get a strong field. Candidates read restructure news before they read your advert, and the institutions that keep attracting good people through a contraction are the ones that show a coherent shape on the far side of it. That's an employer brand problem dressed up as a finance problem, and it's fixable in a job ad.
Write the assumption down
Nearly every dispute I've watched this year — the picket lines, the no-confidence votes, the senate inquiries — has been an argument about a number that existed in a spreadsheet months before anyone outside a leadership team saw it. Institutions that published the assumption before the decision have had a materially easier time, because staff can argue with an assumption. They can't argue with an announcement.
So here's the action, for anyone sitting on one of these plans right now. Put the number and the assumption behind it in the same paragraph, send it to the people whose jobs depend on it, and give them a date by which the assumption will be re-tested against actual load. Headcount planning fails quietly for a year and then loudly in a single afternoon. The gap between those two moments is where all the goodwill gets spent.
Photo by Abhinav Anand on Unsplash
- enrollment-cliff
- research-funding
- university-restructuring
- higher-education-finance
- academic-job-cuts

