Higher-ed overall turnover reached 14% in 2023-24, down from a 16% high the prior year but still above pre-pandemic levels (CUPA-HR, 2024). Despite that modest improvement, nearly 1 in 4 higher-ed employees say they are likely to look for other employment in the coming year — and compensation is the top reason they give (CUPA-HR, 2025). Institutions that cannot win the pay competition need a different playbook. CUPA-HR's own retention model points to belonging, supervisor quality, flexibility, and visible career paths as the predictors that actually hold people.
14%
Higher-ed overall turnover (faculty + staff combined) in 2023-24, down from 16% high in 2022-23; pre-pandemic rate was ~12%
22%
Turnover rate for part-time non-exempt staff — highest of any category — vs. 7% tenure-track and 11% non-tenure-track faculty
nearly 1 in 4
Higher-ed employees likely to look for other employment in the coming year; compensation is the top stated reason
−0.4% real wages
Real average faculty salary change from fall 2024 to fall 2025; cumulative 7.5% pandemic-era decline (fall 2019–fall 2022) not yet recovered
68.2%
Share of faculty who are part-time or contingent (fall 2023); per-course-section pay ranges $3,200–$6,320
strongest predictors of retention
Sense of belonging, feeling valued, and being engaged — the strongest predictors of retention in CUPA-HR's 2025 model, ahead of compensation
45% less likely to turn over
Well-recognized employees after two years, per Gallup/Workhuman 2024 longitudinal study (3,447 employees) — VENDOR-REPORTED; Workhuman is a recognition platform
01
Where higher-ed turnover sits now
Higher-ed overall voluntary turnover was 14% in 2023-24, down from a 16% high in 2022-23 and above the pre-pandemic rate of approximately 12% (CUPA-HR, 2024). That modest decline is the first in three years and reflects a cooling labor market — not a structural fix. The same CUPA-HR data shows that nearly 1 in 4 higher-ed employees report they are likely to look for other employment in the coming year, with compensation as the top stated reason (CUPA-HR, 2025).
The headline rate masks a wide spread by employee category. Part-time non-exempt staff turn over at 22% — the highest rate of any group. Non-tenure-track faculty turn over at 11%; tenure-track faculty at 7% (CUPA-HR, 2024). CUPA-HR's publicly available data provides this exempt/non-exempt and faculty-status breakdown; job-function detail by department (IT, student affairs, facilities) sits behind the CUPA-HR DataOnDemand paywall and is not publicly citable at this time.
The gap between who is most likely to leave and who institutions typically invest in retaining is real. Part-time non-exempt staff — the group churning fastest — often have the least access to development, flexible scheduling, and recognition programs. Addressing that gap is cheaper than the turnover it prevents.
02
Who leaves and why
The population most likely to exit is part-time non-exempt staff at 22% turnover, but the contingent faculty picture is equally acute. More than two-thirds of all faculty — 68.2% as of fall 2023 — hold part-time or contingent appointments (AAUP, 2024-25). Per-course-section pay ranges $3,200–$6,320, with a median starting rate of $3,121 (AAUP, 2024-25). These instructors are effectively paid by the piece, with no job security, and often no visible path to a permanent role — the profile of PERSONA-003's contingent majority.
Among non-faculty staff, CUPA-HR's 2025 retention data identifies younger employees, men, employees of color, and non-supervisors as the groups most likely to leave — particularly when pay compression leaves them earning near long-tenured colleagues or when career ladders are invisible (PLAY-027). These are not separate problems; they reinforce each other. An employee of color who sees no advancement path and whose pay has not kept pace with a new hire's starting offer has two compounding reasons to exit.
Compensation is the top reason employees give for job-hunting, but CUPA-HR's model consistently shows that belonging, feeling valued, and being engaged are the stronger predictors of who actually stays — even when controlling for pay levels (CUPA-HR, 2025). That distinction matters practically: if you can only move the dial on one thing this budget cycle, the data says to invest in belonging and supervisor quality rather than a one-time market adjustment alone.
03
Belonging and feeling valued beat pay as retention predictors
CUPA-HR's 2025 Higher Education Employee Retention Survey (n=3,791) found that "aspects of job satisfaction and wellbeing, particularly a sense of belonging, feeling valued, and being engaged, are the strongest predictors of retention" (CUPA-HR, 2025). Confidence in leadership ethics and values ranked as the second-largest predictor — a striking jump from seventh place two years earlier (PLAY-012). Compensation matters, but it is not where the model's predictive weight sits.
Belonging is not a soft metric. For the faculty and staff who are furthest from institutional visibility — adjunct instructors teaching across multiple campuses, part-time non-exempt staff in facilities or food service — belonging is the experience of being named, known, and counted. The tactics that create it are low-cost: direct-manager recognition of specific contributions, inclusion in unit meetings and communications, and clear signals that an employee's presence is noticed and valued.
"Job satisfaction and wellbeing is the strongest predictor (by far) of retention" — CUPA-HR, 2025 Higher Education Employee Retention Survey
For contingent faculty, belonging also has a structural dimension. Those offered multi-year contracts, governance participation, and basic professional inclusion report meaningfully higher commitment than colleagues on per-semester or per-course arrangements — even when the pay differential is small (PLAY-011). Institutions that move the belonging needle without touching governance or contract security see partial effects; institutions that address both see the full model play out.
04
Support supervisors and chairs to hold their teams
The most actionable lever in CUPA-HR's retention data is also among the least expensive: invest in supervisors and department chairs. CUPA-HR's analysis finds that supervisors who can advocate for staff, implement flexible scheduling, and receive institutional backing and training are significantly less likely to leave themselves — and they keep their direct reports (PLAY-013).
The practical implication is that supervisor development is a retention investment in disguise. A chair who knows how to run a stay conversation, advocate for a direct report's reclassification, and normalize boundary-setting during high-demand periods holds more people than a generic wellness benefit does. Most institutions promote staff into supervisory roles without equipping them for the relational demands of the job.
Hybrid and schedule-flexibility misalignment is a specific gap CUPA-HR flags: when staff want flexibility they are not getting, the supervisor is often the person who can authorize exceptions — and institutions that give supervisors explicit permission and tools to flex arrangements see faster improvement in intent-to-stay than those that announce top-down policies and leave implementation to chance (PLAY-013).
For PERSONA-003 — tenure-track faculty and adjuncts alike — the relevant supervisor is the department chair. A chair who shields teaching loads, advocates for course-load balance, and acknowledges scholarship publicly plays the same retention role as a skilled manager in a staff context. Chair development is rarely funded at the level that its turnover impact warrants, and that gap is a retention gap.
05
Flexibility and career paths
Compensation is the top stated reason staff plan to job-hunt (CUPA-HR, 2025). But for many employees, pay is the permission slip to leave once a second grievance — usually lack of flexibility or career visibility — has already built up. CUPA-HR's 2025 model points to PLAY-027 for where limited dollars go furthest: visible career ladders for non-faculty staff and pay-compression repairs for long-tenured employees being caught by new-hire starting rates.
Where a market-rate adjustment is not immediately feasible, the substitution that performs best is a combination of schedule flexibility, clear development pathways, and explicit belonging investment. CUPA-HR's data names younger employees and employees of color as the most likely to leave — and both groups consistently cite career visibility ("I can see a path here") and schedule control as factors in stay decisions (PLAY-027).
Build career ladders that are visible, not aspirational. A staff member who knows what the next role looks like, what skills it requires, and that their supervisor is actively tracking their readiness for it has a concrete reason to stay. Pair any lateral moves with skill development and explicit recognition of growth — which also addresses the belonging deficit that CUPA-HR flags as the primary retention predictor in its 2025 model.
06
Recognition and engagement as retention levers
Gallup and Workhuman's 2024 longitudinal study (3,447 employees) found that well-recognized employees are 45% less likely to have turned over after two years, and 65% less likely to be actively job-searching when recognition quality is high (VENDOR-REPORTED — Gallup/Workhuman, 2024; Workhuman is a recognition platform). Even accounting for the vendor context, the directional finding aligns with CUPA-HR's independent belonging and feeling-valued predictors: being seen and acknowledged is a structural retention lever, not a perk.
For faculty, recognition that lands looks different than corporate framing. What tenure-track and contingent faculty value is protected scholarly time, research support, and acknowledgment of their academic contributions in the language of their discipline — not generic "employee of the month" awards (PLAY-011). For adjuncts, being included in department meetings, named in departmental communications, and given access to professional development signals belonging. Basic belonging is a material retention factor at this population's pay level (STAT-025, CUPA-HR, 2025).
For non-faculty staff — non-exempt, facilities, administrative — specific, in-the-moment recognition from direct supervisors is the tool with the highest return on the smallest investment. Peer-to-peer recognition expands the source of appreciation beyond the single-supervisor channel and reaches people whose work is invisible to department leadership. Tools like Actify can deliver peer and manager recognition on mobile devices with no corporate email required — which matters directly for the part-time non-exempt staff churning at 22% (CUPA-HR, 2024). Actify's activity-first engagement, gamification (points, leaderboards, badges), and automatic monthly pulse add an engagement infrastructure on top of the recognition layer, at flat pricing accessible for mid-market institutions. It does not substitute for shared governance, contract security, or faculty scholarly autonomy — those are structural and must be addressed separately.
07
The pay reality — and what software can and cannot fix
Higher-ed real faculty wages fell approximately 0.4% from fall 2024 to fall 2025 after adjusting for inflation, per AAUP's preliminary 2025-26 Faculty Compensation Survey (AAUP, 2025-26). The cumulative pandemic-era decline of approximately 7.5% in real wages between fall 2019 and fall 2022 has not been recovered. For contingent faculty earning $3,200–$6,320 per course with no benefits or job security, any engagement or recognition initiative starts from a credibility deficit — and rightly so (AAUP, 2024-25).
This is the honesty block: engagement software, recognition platforms, and belonging initiatives do not fix sub-living wages, job insecurity, eroding shared governance, or unmanageable workloads. The structural fix must be named first. For contingent faculty, that means multi-year contracts, pay equity, governance inclusion, and pathways to stable employment. For non-exempt staff, it means market-rate pay and visible career ladders. Where institutions cannot deliver those things immediately, they should say so plainly and pair incremental improvements with honest timelines — not substitute a recognition tool for a fair contract (PLAY-023).
The "do it for your vocation" framing that mission-driven sectors weaponize is particularly corrosive in higher ed. Appealing to academic calling as a reason to accept declining real wages or contingent precarity erodes institutional trust faster than any turnover number captures. Leaders who model boundaries, reward sustainable performance, and name structural constraints honestly retain more people than those who ask faculty and staff to absorb financial stress in silence (PLAY-023).
Where pay cannot move significantly this budget cycle, the CUPA-HR data provides a clear priority order: (1) supervisor quality and backing, (2) belonging and inclusion — especially for the populations most at risk, (3) flexibility alignment, and (4) visible career paths. Recognition and engagement tooling can amplify those efforts as an affordable multiplier. It is not a substitute.
