At a glance
Faculty pay rose 2.3 percent while inflation ran at 2.7 percent. A practical look at academic salary benchmarking, peer groups, market supplements and pay transparency.
Before a shortlist exists, someone in reward opens a spreadsheet. Down one side are your department's pay bands. Across the top is a peer group. Somewhere in the middle sits the number that decides whether the offer you're about to make gets accepted, or gets read out to a candidate as proof that another university values them more.
Academic salary benchmarking is the work of building that number: matching job to job, adding up the allowances, adjusting for pension and teaching load, and deciding what competitive means at your institution this year. It's unglamorous. It's also where most recruitment problems begin, because a university that benchmarks badly loses the candidate it wanted and never learns why.
This year the inputs are pulling in opposite directions. Nominal pay is rising almost everywhere. Real pay is not. And since June 2026, employers across the European Union have to show candidates the range before the interview, which drags an old back-room exercise into the advert itself.
What the latest numbers actually say
The American Association of University Professors (AAUP) recorded a 2.3 percent rise in average full-time faculty pay from fall 2024 to fall 2025. Consumer prices rose 2.7 percent over the same stretch. Real pay slipped about 0.4 percent, and average real salaries now sit roughly 9.5 percent below their fall 2019 level and 5.8 percent below fall 2008.
That follows two better years. The 2024-25 survey found nominal increases of 3.8 percent, the second consecutive year wage growth beat inflation, and still left real pay 6.2 percent under the pre-pandemic mark.
CUPA-HR's workforce data tells the same story with sharper edges. In 2025-26, tenure-track faculty received a median 1.8 percent increase. Staff got 3 percent. Administrators got 2.9 percent. Tenure-track faculty have now had the smallest raise of any employee group for four consecutive years, and across the ten years CUPA-HR tracks, their increases have never once exceeded inflation.
| Employee group | Median increase, 2025-26 | Real position against 2019-20 |
|---|---|---|
| Staff | 3.0% | 0.9% lower |
| Administrators | 2.9% | Below pre-pandemic |
| Professionals | 2.8% | Below pre-pandemic |
| Non-tenure-track faculty | 2.0% | 6.8% lower |
| Tenure-track faculty | 1.8% | 11.7% lower |
In the United Kingdom the settlement was smaller again. The Universities and Colleges Employers Association (UCEA) made a full and final offer of 1.4 percent from 1 August 2025, its lowest uplift since 2020, alongside removal of the bottom two points on the national pay spine. Around half of staff were also eligible for an increment worth up to 3 percent on top. The five higher education unions had claimed RPI plus 3.5 percent or a flat £2,500, rejected the offer, and moved to industrial action ballots that closed in December 2025 without reaching the 50 percent turnout threshold.
Put the three datasets side by side and the benchmarking problem is obvious. A 1.4 percent award, a 1.8 percent faculty increment and a 2.3 percent average increase all look like movement. Against 2.7 percent inflation they are standstill, and any peer comparison that ignores the real-terms line is measuring the wrong thing.
A pay spine is a floor, not a market
Minimum rates exist to stop exploitation. They were never built to describe what a lecturer actually costs. Australia shows the gap plainly. The Higher Education Industry—Academic Staff—Award 2020 set Level B step 1 at $85,329 a year from 1 July 2025, as recorded in the annual wage review variation. Griffith University's enterprise agreement put its own Level B step 1 at $113,226 from March 2025, with Level A step 1 at $79,985. Same rank, same country, roughly a third more money.
An institution that benchmarks against award rates will believe it pays above market while paying under it. The comparison that matters runs against the places you genuinely lose staff to, at the rank and discipline you're hiring into, refreshed every year because enterprise agreements roll over on their own cycles.
Discipline adds a second distortion. Business has been the highest-paid field in CUPA-HR's discipline tables for nine consecutive years and posted the largest median salary increase of any discipline over two decades, up 66.2 percent since 2003-04. English language and literature has ranked among the ten lowest-paid disciplines every year since 2003-04. A single institution-wide midpoint hides both facts, and universities keep setting one anyway. The twenty-year discipline analysis is the document to hand anyone who argues otherwise.
Peer groups get chosen, not discovered
Ask who's in the peer group and you'll get a defensible answer that is also a strategic one. Winthrop University's pay study used three lists built on CUPA data: national, regional and athletics. The national group was defined by seven indicators, including total expenses, retention rate, tuition and fees as a share of core revenue, headcount, staff numbers, faculty numbers and graduation rate. The exercise produced a five-year plan costing about $2.5 million, funded in steps of $400,000 and then $500,000 a year.
That's benchmarking done properly. It's also rare. Research on UK vice-chancellors covering 154 institutions across a decade found pay was set against a point in the peer distribution, and that the movement was asymmetric. Institutions below average caught up. Institutions above average did not come down. Every round of comparison pushed the whole distribution higher.
The same ratchet works on faculty bands. If everyone benchmarks to the median of a group chosen partly for its prestige, nobody stays at the median for long. The honest version sets a stated position, say the median of regional peers and the upper quartile for shortage disciplines, then reports the gap instead of quietly resetting the target.
When the new hire out-earns the associate professor
Two words to learn before the next salary conversation. Compression is when junior pay crowds the pay of more experienced colleagues. Inversion is when a new hire earns more than someone already in the rank above. Both are built into academic labour markets, because most hiring happens at entry level and entry-level pay is set by the external market while everyone else's pay is set by an internal budget.
The evidence is old and consistent. Business school survey data from the 2000s showed newly hired assistant professors arriving at 2 percent above existing assistants and 1 percent above existing associate professors, with new doctorates paid 10 percent more than current assistants and 8 percent more than associates. A Californian fact-finding panel catalogued the same pattern in the state university system, where new associate professors were paid above the average for all associate professors.
What it costs is retention. A professor who finds out the person in the next office, hired last September, earns more for the same work rarely files a grievance. They take the call from the search committee elsewhere. Repairing inversion after the fact is expensive. Preventing it means checking internal relativities whenever you make a market-driven offer, and budgeting for the adjustments that follow.
Allowances are where the real money hides
Base salary is the easy part of a benchmark and the least complete. Market supplements move the total materially in disciplines where the external market runs hot. Edinburgh's policy allows a supplement for posts up to grade UE09, initially for up to three years, reviewed annually and included in equal pay audits. Cardiff's version spells out that supplements are non-pensionable and not subject to national pay awards. Brunel routes requests through a panel of the deputy vice-chancellor, chief operating officer and HR director, and normally aligns the figure to the median benchmark, holding the upper quartile for exceptional cases. Essex sets a standard term of two years with an annual market review.
Read those policies closely and an honest comparison becomes possible. A £10,000 supplement that isn't pensionable is worth less than a £10,000 rise in base. A supplement that disappears after three years is worth less again. If your benchmarking spreadsheet lines up base salaries only, you're comparing your permanent money against someone else's temporary money.
The rest of the package needs the same treatment: employer pension contribution, teaching load, sabbatical entitlement, relocation support, visa and sponsorship costs, research start-up, and whether the contract is continuing or fixed-term. A lighter teaching load is a real number. So is a permanent contract in a sector offering fewer of them. The AAUP counted just 31.8 percent of US faculty on full-time tenured or tenure-track appointments in fall 2023, against 53.1 percent in 1987.
Transparency is coming for the advert
The EU Pay Transparency Directive reached its transposition deadline on 7 June 2026. Under it, job applicants must be told the initial pay or pay range before the interview, and employers can no longer ask about salary history. Employees can request the criteria used to set pay and progression, plus average pay levels for colleagues doing work of equal value. Employers with 100 or more staff report gender pay gaps from 2027 using 2026 data, and an unjustified gap of 5 percent or more in any category triggers a joint pay assessment with worker representatives. Once a prima facie gap is shown, the burden of proof moves to the employer.
Implementation is patchy. Only Italy, Slovakia, Lithuania and Malta had final legislation in force by mid-2026. Germany, France, the Netherlands and Spain missed the deadline, and Sweden has signalled it may not transpose at all, as tracked in this summary of member state positions. The European Commission declined to stop the clock, and public-sector employers in late-transposing states are already exposed to the directive directly.
For a university, the practical consequence doesn't depend on any of that. A candidate weighing two offers in Dublin, Milan or Amsterdam will ask for the range and expect it. Institutions that answer cleanly, with a band that has a midpoint and a stated placement rule, look competent. The ones that improvise look like they're hiding something, which is worse than sitting at the bottom of the band.
Existing rules point the same direction. The UK has required gender pay gap reporting from employers with 250 or more staff since 2017. The AAUP's own figures show why the scrutiny continues: women full-time faculty earned an average of $108,678 in 2025-26 against $130,064 for men, a ratio of 83.6 percent, with the widest gap at full professor. Reporting a gap isn't the same as closing it, but it does change what a committee is willing to sign off on a Friday afternoon.
What to do from the hiring side
Six moves, in the order they save money.
- Name your peer group in writing and justify every member. Regional competitors, aspirational institutions and discipline comparators are different lists, used for different searches.
- Benchmark by rank and discipline. Business, health and law will wreck any institution-wide average you produce.
- Compare total reward, including pension, supplement, teaching load and contract type.
- Run an internal relativity check before every market-driven offer. Cheaper now than a retention counter-offer in eighteen months.
- Put the band in the advert. It filters mismatched candidates early, and in much of Europe it's about to be required.
- Report the real-terms position, not just the headline award. If a 1.4 percent uplift follows a year of 4 percent inflation, say so and explain what happens next.
One more item belongs on that list, and committees skip it: publish promotion criteria beside the bands. For most faculty, promotion is the only route to a significant rise. Jacqueline Bichsel, associate vice president of research at CUPA-HR, has said she is now hearing anecdotal cases where even promotion produces less than expected. A band without a route through it is just a ceiling.
Photo by Ibrahim Rifath on Unsplash
What to do if you're the one applying
Ask for the range before you name a number. If a recruiter won't give one, ask what the midpoint of the grade is and where a typical appointment lands. That's a different question from what's the salary, and it usually gets an answer.
Then price the whole offer. Check the enterprise agreement or pay spine, since most institutions publish one. Compare pension contribution rates rather than the percentage on the contract. Work out what happens when the supplement ends, and ask which continuing roles were filled internally in the past two years if the post is fixed-term.
Keep your own record of what you're worth, too. Academics negotiating between institutions start from their last salary, and last salary is a poor proxy for market value after several flat years. If you want a worked example of how one institution's bands translate into an actual offer, our breakdown of what a university lecturer earns is a reasonable place to start.
The number that wins the candidate
None of this makes benchmarking glamorous. It makes it decisive. The institutions losing good people this year are rarely the ones paying least on paper. They're the ones whose numbers don't hold up when a candidate puts two offers side by side and finds a gap nobody explained.
So do the boring work before the advert goes live. Then put the band in the advert and watch what it does to your applicant pool.
- faculty-compensation
- higher-education-pay
- pay-transparency
- academic-salary-benchmarking
- faculty-pay-bands

