I hate the term “administrative bloat.”
It’s become a favorite explanation, both within and outside higher education, for everything from rising tuition to institutional deficits. The implication is that colleges and universities are filled with unnecessary administrators, bureaucrats, and layers of management that contribute little to their educational missions.
Few characterizations are more dismissive of the people who actually make our institutions work. I’ve spent much of my career working alongside dedicated higher education professionals who have committed their lives to serving students, supporting faculty, advancing research, and strengthening their institutions. Many work extraordinarily long hours, often for compensation well below what their skills and responsibilities would command elsewhere. I fundamentally believe that these administrators are not separate from fulfilling an institution’s mission, but rather essential to its long-term success.
But defending the importance of administrators doesn’t mean defending every administrative process, organizational structure, or way of working that higher education has inherited. And increasingly, I worry that our current operating model is unsustainable.
The Growing Financial Pressures on Higher Education
Consider the three major revenue engines that sustain much of American higher education.
Tuition revenue is under pressure from demographic shifts, growing price sensitivity, and intensifying competition for students. Philanthropic support, while critically important, is inherently uneven and influenced by economic conditions and financial markets (yes, Harvard, Stanford, and WashU posted eye watering returns this month, but they’re part of a very, very small club.) And federal research funding faces significant uncertainty, with implications that extend well beyond research-intensive universities.
At the same time, the costs of operating our institutions continue to rise, as do expectations for the services and support we provide to students. The consequences are becoming impossible to ignore.
Last week, Higher Ed Dive reported that the University of Oregon is considering a plan to address a projected $65 million structural deficit, including the elimination of more than 100 positions and 18 graduate programs. The proposal identifies $17.6 million in administrative services savings and $25.8 million in academic cost reductions.
Meanwhile, Syracuse University recently announced a comprehensive review of its executive and administrative functions following enrollment challenges and a projected operating deficit. Its chancellor framed the objective as ensuring that the university is organized and resourced for the institution it is becoming, rather than the one it has historically been.
These are two very different institutions confronting a shared reality. And they are hardly alone.
The traditional response to financial pressure has been painfully familiar: hiring freezes, across-the-board reductions, deferred investments, reorganizations, and, eventually, layoffs. These measures may be necessary in some circumstances, but they rarely address the underlying question of how the work itself gets done.
Too often, we simply ask fewer people to do the same amount of work. What if, instead, we fundamentally redesigned the work?
Rethinking Administrative Work Through AI and Automation
The emergence of AI, automation, and increasingly sophisticated data tools creates an extraordinary opportunity to rethink how colleges and universities operate. But the real opportunity isn’t simply deploying a chatbot or purchasing another enterprise platform. It’s redesigning institutional operations around a basic question: What work should technology do so that our people can spend more time doing what only people can do?
Think about the work that fills a typical administrator’s day: preparing reports, reconciling spreadsheets, scheduling meetings, processing forms, updating records, routing approvals, responding to repetitive inquiries, and transferring information between systems that don’t communicate.
Much of this work is necessary. But does all of it require a person to perform every step?
Imagine an academic advisor who spends less time documenting interactions and searching for student information, and more time helping students navigate difficult academic and personal decisions. Or a financial aid professional who spends less time processing routine documentation and more time helping families understand their options.
Imagine an advancement professional who spends less time assembling donor briefings and more time building meaningful relationships. Or a department administrator who can spend less time coordinating schedules and reconciling budgets and more time supporting faculty, students, and academic innovation.
These are not simply opportunities to save money. They are opportunities to improve the quality of institutional work and the experiences of the people our institutions serve.
At Carnegie, our work with institutions exploring AI-enabled operational transformation is reinforcing a central lesson: the greatest gains come not from automating individual tasks in isolation, but from rethinking the processes, roles, and systems around them.
We’ve developed a five-step administrative optimization approach to help institutions move from the promise of efficiency to measurable operational improvement.
A Five-Step Approach to Administrative Optimization in Higher Education
1. Start with the work, not the organizational chart.
Traditional administrative reviews often begin with headcount, reporting relationships, and departmental budgets. Those are important considerations, but they don’t tell us enough about what people actually do. We start by understanding how work happens across an institution: reviewing job descriptions, interviewing staff, mapping workflows, and identifying the activities that consume time and resources.
In our work, we’ve explored how structured, AI-enabled interviews can help employees describe their responsibilities, recurring tasks, bottlenecks, and opportunities for improvement. This approach frequently surfaces challenges that are invisible in an organizational chart: duplicated reporting, unnecessary approvals, manual data transfers, outdated processes, and work that has accumulated over years without anyone questioning whether it still adds value. It also creates an opportunity to engage employees as partners in redesigning their work rather than treating them as the targets of a cost-reduction exercise (during my latest project, I can’t tell you how many employees shared how “seen” they felt by being asked about their work and ideas for making their department more efficient.)
The first question shouldn’t be, “How many people do we need?” It should be, “What work actually needs to happen, and why?”
2. Identify what to eliminate, automate, augment, and protect.
Not every administrative responsibility is equally suited for automation. We find it useful to distinguish among four categories of work:
- Eliminate or simplify: Processes, reports, approvals, and activities that no longer create meaningful value.
- Automate: Repetitive, rules-based tasks such as data entry, document routing, scheduling, and routine reporting.
- Augment: Activities where AI can help professionals work faster or make better-informed decisions, such as synthesizing information, preparing analyses, drafting communications, or identifying patterns.
- Protect: Responsibilities requiring human judgment, empathy, ethical deliberation, relationship-building, or consequential decision-making.
For practical examples of how AI can support different functions across higher education, explore Carnegie’s Practical Map of AI Use Cases in Higher Education.
This distinction is essential. The objective is not to automate everything that can be automated. It is to deploy technology where it improves efficiency without compromising institutional quality, trust, or the student experience. In fact, some of the most important outcomes of this exercise are decisions about where not to use AI.
3. Redesign workflows, roles, and organizational structures.
Automating an inefficient process doesn’t necessarily make it a good process. It may simply allow an institution to perform unnecessary work faster. That’s why operational transformation must go beyond individual productivity tools.
Consider a routine student service request that passes through several offices, requires multiple approvals, and involves repeated entry of the same information. An AI tool might help each employee process their portion more quickly. But the bigger opportunity may be eliminating unnecessary handoffs, integrating the underlying data, and redesigning the entire experience around the student.
The same principle applies to organizational design.
As workflows change, institutions should revisit job descriptions, staffing models, shared services, and the distribution of responsibilities across departments. Some roles may become more strategic. Others may combine responsibilities that historically required separate positions. Still others may shift toward higher-value student-facing or analytical work.
The goal should be to build an operating model around the institution’s future needs, not simply preserve the structures of its past.
4. Build the technology, data, and governance foundation.
One of the most common obstacles to operational efficiency isn’t the absence of technology. It’s the fragmentation of technology and data that institutions already possess. Student information systems, CRMs, financial platforms, HR systems, and departmental tools often operate in disconnected environments.
The result is familiar: staff spend enormous amounts of time moving information between systems, reconciling conflicting records, and creating manual workarounds. Institutions need to assess where existing technologies can be better integrated, where automation can bridge gaps, and where new capabilities are genuinely necessary. Just as importantly, they need governance that addresses data security, privacy, accuracy, accountability, and human oversight.
An institution shouldn’t need to purchase an entirely new technology stack to begin realizing efficiency gains. Often, the first opportunities come from using existing investments more intelligently.
5. Translate productivity gains into measurable institutional value.
This final step is arguably the most important. Saving an employee five hours a week does not automatically produce five hours of institutional value. Nor does it automatically translate into budget savings.
Institutions must decide what they intend to do with the capacity they create. Will they redirect staff time toward student support? Reduce backlogs and improve service levels? Absorb additional work without increasing headcount? Consolidate responsibilities as vacancies arise? Or selectively reduce operating costs?
Each represents a legitimate strategic choice, but each requires intentional leadership decisions. We encourage institutions to establish clear baselines and measure outcomes such as time saved, processing speed, service quality, error rates, employee experience, and actual financial savings. The critical distinction is between productivity gains, capacity that can be redeployed, and recurring budget savings.
All three matter. They are not interchangeable.
How Significant Could the Savings Be?
Our operational efficiency modeling suggests that the opportunity is meaningful. In selected administrative functions, our preliminary assessments have identified potential workload reductions of approximately 15–25% through workflow simplification, automation, and role redesign. (These estimates vary significantly by function and should not be interpreted as institution-wide savings.)
For an illustrative institution with $40 million in annual administrative compensation, our scenario modeling has explored the potential to free the equivalent of approximately 10–12 full-time positions in capacity while reducing administrative compensation costs by roughly 4–5% over time, primarily through attrition, selective nonreplacement, and restructuring.
More aggressive scenarios can produce larger savings, but also carry greater organizational and service-delivery risks. These are modeled opportunities, not guaranteed outcomes or realized client savings. Actual results depend on institutional scale, workforce composition, technology readiness, implementation costs, and leadership decisions.
And the financial return tells only part of the story.
Consider what an institution could accomplish if its advisors, student success professionals, department administrators, and other staff recovered even a few hours each week for more meaningful work. What would that mean for student persistence? Employee satisfaction? Faculty productivity? The responsiveness of institutional services? Those benefits may be just as important as the financial savings, even when they are harder to capture on a balance sheet.
This Is a Leadership Challenge, Not a Technology Project
None of this will happen simply because institutions make AI tools available to employees. Meaningful transformation requires presidents, provosts, CFOs, CHROs, CIOs, and other leaders to work together across traditional organizational boundaries. It requires a willingness to question longstanding practices, reconsider organizational structures, invest in employee development, and make deliberate choices about how efficiency gains will be used.
These leadership and governance considerations are also explored in our article on moving from AI experimentation to institutional readiness.
And it requires a commitment to transparency.
Staff must understand that the objective is not to label their work as unnecessary, but to identify which parts of that work can be performed differently so that their expertise can be directed toward the institution’s highest priorities. That doesn’t mean pretending difficult workforce decisions will never be necessary. Some institutions will need to reduce costs, and some redesigns will change staffing requirements. Leaders owe their employees honesty about those possibilities, alongside meaningful opportunities for reskilling, redeployment, and participation in the transformation process.
This work must also be continuous. Administrative optimization is not a one-time exercise conducted during a budget crisis. It should become part of how institutions routinely assess and improve their operations.
The Future of Higher Education Requires a Different Conversation
I understand why the term “administrative bloat” resonates. Colleges and universities have become complex organizations, and that complexity has real costs. But I reject the implication that the people supporting our institutions are somehow less essential to their missions.
The challenge isn’t that higher education employs too many people who care about students, faculty, and institutional success. It’s that too many of those people are trapped in outdated processes, disconnected systems, and administrative routines that prevent them from contributing at their highest level.
At a time when our revenue models are under pressure and expectations of higher education continue to rise, we cannot afford to preserve inefficient ways of working simply because they are familiar.
Nor can we afford to approach efficiency as an exercise in indiscriminate cuts.
The question isn’t how we can get by with fewer people doing more work. It’s how we can use technology, data, and thoughtful organizational redesign to help our people do more meaningful work.
That’s the administrative revolution higher education needs.
And it is long overdue.
Rethink What’s Possible for Your Institution
Improving institutional efficiency requires more than reducing costs. Carnegie helps higher education leaders rethink operations, identify opportunities for improvement, and align people, processes, and technology with their institution’s priorities.
Explore our Academic & Operational Efficiency solutions or start a conversation to discuss what’s possible for your institution.
