At a glance
Tenure-track pay has fallen 11.7% behind inflation since 2019-20, one in four university employees is still job-hunting, and nearly half of them would stay for an internal move. What workforce planning in higher education should actually measure.
When a vice-chancellor, a provost or a finance committee asks for the budget justification behind a pay decision, the question underneath the question is narrower than the paper in front of them: who are we about to lose, and what would it cost to replace them? The 2025-26 salary data answers the first half of that whether or not anyone in the room wants to hear it.
CUPA-HR's annual workforce analysis, which compares only institutions reporting salaries in both the 2024-25 and 2025-26 academic years, found median increases of 3.0% for staff, 2.9% for administrators, 2.8% for professionals, 2.0% for non-tenure-track teaching faculty and 1.8% for tenure-track faculty. Inflation over the matching window, measured by the Consumer Price Index in the year to November 2025, ran at 2.7%. Every non-faculty group beat it. Faculty did not.
That has now held for ten consecutive years. Across the decade covered by the survey, tenure-track raises have not once outpaced inflation, and this is the fourth year running they have been the smallest increase of any employee category. In real terms, tenure-track salaries sit a median 11.7% below their 2019-20 level. Non-tenure-track teaching faculty are 6.8% behind. Non-exempt staff, the group that has done best, are still 0.9% down. Nobody on campus has recovered the purchasing power they had before the pandemic.
Jacqueline Bichsel, associate vice president of research at CUPA-HR, told Inside Higher Ed that for most faculty the only realistic route to a meaningful rise is promotion, and that she is now hearing of cases where even promotion delivers less than the candidate expected. Her verdict on a decade of flat real pay was direct: it shows an employee their work is not valuable enough to merit more.
What a defensible benchmark actually contains
Benchmarking gets filed under reporting. Done properly it is a negotiating document, and the difference lives almost entirely in how tightly the comparison group is drawn. An institution that benchmarks all faculty against a national median will usually conclude it is at or near market. The same institution benchmarking assistant professors in accounting, nursing or computer science against the dozen schools it actually loses people to will conclude something quite different, and that second number is the one a dean can defend to a budget committee.
| Employee category | Median increase, 2025-26 | Against inflation (2.7%) | Real salary vs 2019-20 |
|---|---|---|---|
| Staff (mostly non-exempt hourly) | 3.0% | Above | 0.9% lower |
| Administrators | 2.9% | Above | Not restored |
| Professionals | 2.8% | Above | Not restored |
| Non-tenure-track teaching faculty | 2.0% | Below | 6.8% lower |
| Tenure-track faculty | 1.8% | Below | 11.7% lower |
The data underneath those figures is deep enough to slice properly. The 2025-26 CUPA-HR Faculty in Higher Education Survey collected salary and demographic data on 334,666 faculty at 795 institutions, including individual-level pay for 52,520 adjuncts reported from 288 institutions. That is a large enough base to compare by discipline, rank, tenure status and the highest degree an institution awards, rather than by one institutional average that flattens every difference that matters.
Most institutions still benchmark at the wrong altitude. A single campus-wide percentile tells a board nothing about whether the nursing school can hire, whether the IT team is losing second-line engineers to the local health system, or whether the custodial supervisor on night shift is paid below the supermarket down the road. Those are separate labour markets with separate competitors, and a credible workforce plan treats them separately.
Photo by Morgan Housel on Unsplash
The number is a floor, not the argument
Here is where budget papers usually go wrong. They present a peer benchmark, show the institution sitting at the median, and treat that as the end of the analysis. CUPA-HR's own guidance on using survey data for budget planning makes the opposite case: the value lies in comparing trends in actual pay rather than stated intentions, and in splitting the institution into its different employee groups before setting a single increase rate. A median is a floor below which you have a demonstrable problem. It is not evidence you have solved retention.
The 2025 Higher Education Employee Retention Survey, which drew on 3,791 employees across US institutions, found that pay is the reason people most often give for looking elsewhere, but that it is not the strongest predictor of whether they actually go. Job satisfaction and wellbeing predict retention more powerfully than pay does. The sense of belonging, of being valued, of being engaged and of doing work with purpose all carry more weight than a modest raise. Confidence in leadership ethics and values jumped from seventh place to second in predictive strength in two years, which is a striking move for a variable most institutions don't measure.
Read those two findings together and the budget conversation changes shape. Money is the language in which dissatisfaction gets expressed. It is rarely the whole of what is wrong. An institution that fixes pay to the median and leaves workload, flexibility and managerial competence untouched has bought itself a year, not a workforce.
The flexibility gap is the clearest example of a cheap fix nobody is making. In the same survey, 61% of employees said they preferred hybrid or remote work and only 28% had it, while 62% said most of their duties could be done remotely. That is a mismatch running at more than thirty points, and closing it costs almost nothing at a time when almost nothing is affordable.
The turnover risk is concentrated, and the exit is often internal
Voluntary turnover has fallen from its 2022 peak. One in four employees still told CUPA-HR they were likely or very likely to look for another job in the coming year, down from one in three in 2023. Of those considering a move, 56% had already submitted at least one application. This is not latent restlessness; it is a workforce with tabs open.
Risk clusters. Employees under 45, men, employees of colour and non-supervisors are consistently more likely to be looking. Supervisors are less likely to leave, but they absorb the load: 72% work beyond full-time hours and 21% work eleven or more extra hours a week. More than a quarter of people considering a move cited job security, and among staff in research and sponsored programmes that figure rose to 34%. Fixed-term contracts and grant-funded posts are doing exactly what you would expect them to do to a person's planning horizon.
Then comes the finding that ought to reshape how institutions think about recruitment spend. Of employees likely to leave, 72% said they were open to staying in higher education, and nearly half were open to a different role at their current institution. Most of the people about to walk out the door would consider staying if the door swung inward instead of out.
The UK data points at the same lever from the other side. Vitae's CEDARS 2025 survey found that among staff with management responsibility, 94% felt confident using inclusive recruitment processes and 93% felt confident supporting career aspirations, but only 47% felt confident providing guidance on institutional redeployment. Managers know how to hire people. They mostly don't know how to move them.
CEDARS also exposed a mid-career problem that is easy to miss because it sits between the two groups institutions watch most closely. Job satisfaction among mid-career academic staff came in at 65%, against 73% for early-career staff and 79% for senior staff. That dip occurred despite 90% of mid-career respondents taking part in formal appraisal and 85% holding permanent contracts. The processes exist, they're being used, and they aren't delivering. Meanwhile 43% of respondents spent between one and four days on professional development in the year, and 24% spent less than one.
Restructuring has become the default, and it is expensive in ways nobody budgets for
The financial pressure driving these questions is not hypothetical. In the UK, the University of Nottingham is seeking to remove more than 600 roles from a workforce of which 2,700 were placed at risk, alongside 42 course closures, following a reported deficit. Exeter has consulted on roughly 150 full-time equivalent academic posts, with the union claiming 523 FTE at risk and 85% of them concentrated in humanities and social sciences. Essex cut 400 FTE against a £24m shortfall and closed its Southend campus. Aberystwyth is targeting £16m of savings by 2027-28 after 150 posts went in the previous round.
In the United States the pattern is the same at a different scale. Deloitte's 2026 higher education trends report cites more than 900 layoffs at the University of Southern California, 363 at Stanford and 424 at Northwestern, about 5% of that university's workforce. The same analysis notes a 17% drop in new international graduate enrolments in autumn 2025, more than half of private universities rated by S&P Global posting operating deficits in 2024, and a projected 13% decline in US college enrolment between 2025 and 2041.
What gets cut first is instructive. The Higher Education Policy Institute's September 2026 report on professional and support staff, produced with AHEP Consulting, puts that group at 255,678 full-time equivalents in 2023, or 56% of the UK higher education workforce. Its survey of 450 respondents found two-thirds reporting that change and uncertainty over five years had affected them negatively, and the burden fell hardest at the bottom: 79% of staff earning under £40,000 reported negative impacts, against 51% of those on £60,000 or more. Because lower-paid roles are disproportionately held by women and ethnic minority staff, the damage lands on the groups already carrying persistent pay gaps.
The report also identifies a blind spot that should worry anyone trying to plan a workforce. Only 125 of the 228 providers returning staff data to HESA in 2023-24 reported on all their non-academic staff, and people employed through subsidiaries or outsourced contracts don't appear in the figures at all. It is difficult to argue for investment in a workforce you cannot count.
Restructuring has its own bill. HEPI found institutions hiring consultancies to lead the work and bringing in change managers, with redesigned structures sometimes adding management layers rather than removing them. Training budgets are cut as an easy saving, then the missing capability is bought in externally at a higher price. The report recommends an independent review of the Higher Education Role Analysis framework, the job evaluation system used by 70% of UK universities and structurally unchanged since 1995. It is hard to argue your pay structure is market-responsive when the tool that grades the jobs predates the commercial internet.
What to do before the next budget paper
None of this argues that institutions should pay more than they can afford. It argues that the analysis behind a pay decision should answer the questions leadership is actually asking, and that several of the cheapest levers are sitting untouched.
- Benchmark by unit and discipline against the institutions and employers you genuinely lose people to, not against a national median that makes everyone look competitive.
- Publish the resulting pay bands internally. CUPA-HR's data shows employees who doubt their pay is fair have grown from 33% to 43% in two years, and opacity is doing some of that work.
- Run an internal mobility register before the external advert goes out. Nearly half of the people considering leaving would look at another role on the same campus, and most institutions have no mechanism to tell them one exists.
- Train managers on redeployment and career conversations, not just recruitment. The confidence gap there is close to fifty points.
- Count the workforce properly, including outsourced and subsidiary staff, so the plan describes the institution you have rather than the one in the HR system.
The test of a workforce plan is not whether it survives the finance committee. It is whether the people you most need are still there in eighteen months. On current evidence, most institutions know what their peers pay and very little about why their own staff stay or go. The first is a spreadsheet. The second is strategy, and it's the one that shows up on the balance sheet a year later, in recruitment fees, agency cover and the salary premium required to replace someone who could have been kept.
- staff-turnover
- faculty-pay
- cupa-hr
- university-hr
- academic-salary-benchmarking

