A Provost Lab white paper, 2026
Beyond the Static Feasibility Study
Evaluating established programs and emerging fields when the categories that anchor traditional analysis can no longer keep pace.
Most program decisions still run on a feasibility study built for a slower, more legible market. This paper argues that the stakes have risen, that the standard study rests on federal classification systems that update roughly once a decade, and that the same opportunity can read as strong or weak depending entirely on which institutions you compare against — a judgment the institution, not the report writer, is best placed to make.
Executive summary
A report is the wrong unit of analysis.
Academic program decisions carry more financial weight than they did a decade ago, and most institutions still make them with instruments designed for a calmer market: a one-time feasibility study, a national market scan, a generic competitor list. Those instruments are not wrong. They answer a narrower question than the one academic leaders are asking — should we do this program, here, with our students, our costs, and our competitors — and they are answering it about a market that no longer holds still.
Two shifts have arrived together. The market has become harder to read: the traditional-age pipeline peaks this year and declines for the next decade and a half, while net revenue per student is falling even at institutions whose enrollment is growing. And a rising share of the most promising opportunities — interdisciplinary fields, new professional roles, stackable credentials — no longer fit cleanly into the federal classification systems that feasibility studies rely on.
The stakes have risen
Fewer graduates, deeper tuition discounts, and falling net revenue per student mean a wrong program decision costs more and is harder to unwind.
The categories lag
Feasibility rests on CIP and SOC codes that revise about once a decade. Emerging fields are structurally invisible until well after they form.
The frame decides
Change the comparator set and the verdict changes. Who you measure against is a strategic choice — and the institution knows its world best.
Section 1: The environment
The decision got harder, and more expensive to get wrong.
For most of the last generation, a new program decision followed a familiar arc. A school identified an opportunity, commissioned or assembled a feasibility study, reviewed labor-market demand and competitor programs, and decided whether to proceed. The arc still works. What has changed is the ground beneath it.
The traditional-age pipeline is contracting on a schedule that is now well documented. The Western Interstate Commission for Higher Education projects that the number of U.S. high school graduates peaks in 2025 and then declines steadily through 2041, ending roughly thirteen percent below that peak at about 3.4 million graduates.1 The authors are careful to call this a slope rather than a cliff — but a slope that runs for fifteen years still reshapes every enrollment assumption a new program rests on.
The revenue picture is the part that should concentrate attention, because it complicates the intuitive fix of simply enrolling more students. Among private nonprofit institutions, the average tuition discount rate reached a record 56.3 percent for first-time, full-time undergraduates in 2024–25 — meaning that for every published tuition dollar, institutions returned about 56 cents in institutional aid.2 On the public side, the pattern is starker still: full-time-equivalent enrollment rose 3.6 percent in fiscal 2025, yet net tuition and fee revenue per student fell 3.5 percent — the second-largest single-year decline in the four-decade history of the dataset — even as appropriations per student slipped one percent.3
Exhibit 1 More students, less revenue per student
None of this argues that the value of a credential is collapsing. Georgetown's Center on Education and the Workforce projects that by 2031, 72 percent of U.S. jobs will require postsecondary education or training, with an average of 18.5 million openings a year.4 Demand for what institutions produce remains real. The point is narrower and more demanding: when the pipeline is shrinking and net revenue per student is falling, the margin for a mispriced or mis-positioned program is thinner than it has been in decades. A program that merely looks attractive in a national scan is no longer a safe bet.
And the institutions making these decisions have less room to absorb a miss than they once did. A 2026 analysis by Huron Consulting Group estimated that 442 of the roughly 1,700 private nonprofit four-year colleges in the United States — more than a quarter, enrolling some 670,000 students — are at risk of closing or merging within the next decade.5 Nearly a third of private nonprofits posted an operating deficit in 2024.6 In that environment, a new program is also a multi-year claim on faculty lines, facilities, and cash the institution may never get back if the bet is wrong. Which changes the question worth asking. Not whether an opportunity looks attractive — most do — but what a misjudgment would cost an institution with little margin to spare.
Section 2: The instrument
What the feasibility study measures well — and what it cannot see.
The traditional feasibility study is good at what it was built to do. It can summarize degree completions in a field, map occupational demand, identify comparable programs, and describe market growth. For an established program — one with a settled name, a clear curriculum, and a decade of completions behind it — that evidence is genuinely useful, and a good study assembles it well.
But the study's strengths come from a foundation that is rarely named out loud. Almost every market scan is anchored to two federal taxonomies: the Classification of Instructional Programs (CIP), which organizes what institutions teach, and the Standard Occupational Classification (SOC), which organizes what graduates do for a living. Completions data, competitor lists, and labor-market alignment all flow through these codes. They are the grammar of the entire genre.
That grammar is precise, comparable, and — by design — slow. And it produces four predictable blind spots.
Exhibit 2 Everything flows through two federal codes
| The limit | Why it happens |
|---|---|
| It is static | A report answers one version of the question at one moment. Change the comparator set, the geography, or the enrollment assumption and the analysis must be rebuilt rather than re-run. |
| It is generic | National data describes a national average that may bear little resemblance to the institution's actual competitive world — its region, its peer set, the handful of institutions its president actually watches. |
| It separates feasibility from viability | A market scan can show that demand exists without connecting it to the tuition, staffing, and contribution-margin assumptions that decide whether the program can sustain itself. |
| It implies false certainty | A study often ends in a recommendation. But program decisions are assumption-sensitive: viable at 25 students per cohort, not at 12; workable online, not residentially. A single verdict hides that sensitivity. |
The deepest limit, though, is the one the codes themselves create — and it is significant enough to take its own section.
Section 3: The structural gap
Some opportunities have no code yet.
The federal classification systems update on a roughly decennial rhythm. The CIP has been revised five times since it was created in 1980 — in 1985, 1990, 2000, 2010, and 2020 — with the next revision expected around 2030.7 The SOC follows the same cadence: the current edition dates to 2018, contains 867 detailed occupations, and is next due for revision in 2028.8 Between revisions, the categories hold still while the world does not.
This matters because of how new categories get added. The CIP adds a code only after a field is already established — its guidelines look for evidence that programs are being offered across multiple institutions before a new code is created.7 The effect is a built-in delay: a field must already exist at scale before the instrument that measures feasibility can register it at all.
The data-science example is the clean illustration. The role had been forming across statistics, computer science, and analytics for years before the 2018 SOC revision finally introduced a distinct "Data Scientists" occupation.8 For most of the decade in which the field was actually emerging, a labor-market scan keyed to occupational codes would have returned a fragmented, undercounted picture — precisely when an institution deciding whether to launch a program would most have wanted clarity.
Exhibit 3 The lag between when a field forms and when it gets a code
So there are really two lanes of program analysis, not one. The first is traditional feasibility: the program already exists in the data world, with a CIP code, a body of completions, and a settled occupational match. Here the standard study is the right tool. The second is exploratory analysis: the opportunity is interdisciplinary, professionally novel, or built around an institutional strength, and it does not yet map cleanly onto the codes. Here a code-anchored scan returns noise, and the institution needs a different way to look — one that examines the learning environment around the idea rather than retrieving a category that does not yet exist.
| Traditional feasibility charted | Exploratory analysis uncharted |
|---|---|
| Begins with a known CIP / SOC mapping | Begins with an emerging field, role, problem, or learning environment |
| Best for established programs | Best for interdisciplinary or forming opportunities |
| Relies on existing completions and occupations | Looks across adjacent programs, skills, credentials, employers, and peer activity |
| Answers: how does this program perform in existing data? | Answers: what is forming here, and how might we participate? |
| Produces a market scan | Produces a strategic opportunity map |
Most opportunities are not purely one or the other; they sit on a continuum, and the right analytical mode shifts as an idea matures.
| Stage | What leaders already know | Best mode |
|---|---|---|
| Known program | CIP and SOC are clear; completions exist | Feasibility scan |
| Adjacent program | Similar programs exist, but naming varies | Feasibility + comparator testing |
| Emerging field | No clean code; signals are mixed | Exploratory mode |
| Strategic concept | Built around mission, region, or institutional strength | Exploratory + value lens |
| Launch candidate | The model is clear enough to price and staff | Viability testing |
Section 4: The frame
Change who you measure against, and the answer changes.
Even for an established program, one choice in a feasibility study shapes the result more than almost any other, and it usually goes unexamined: which institutions count as comparators. When it is left to a default, the comparison writes the conclusion before anyone has weighed in. Chosen deliberately, it becomes one of the sharpest levers a leadership team has.
Consider an illustrative case. Riverside University — a composite institution we use to demonstrate the point — is weighing a Doctor of Nursing Practice. The DNP is firmly in the charted lane: it has a code, ample completions, and clear occupational alignment. A single national scan returns a crowded field and a cautious verdict. But Riverside does not recruit nationally, and the national average is not its battlefield.
Exhibit 4 One program, three frames, three verdicts: Riverside DNP
National
All providers
Market saturation: high
Verdict: crowded
State or region
Actual recruiting radius
Market saturation: moderate
Verdict: room to enter
Aspirant peers
Institutions we watch
Peer adoption: rising
Verdict: strategic signal
Same program. Same data sources. The comparator frame — not the program — moved the recommendation.
No external analyst knows that context better than the institution's own leaders. A provost may know that a nominal competitor serves an entirely different student population. A dean may know who actually controls clinical placements in the region. A president may care about an aspirant peer that no standard regional scan would surface. That knowledge should not sit outside the analysis as a caveat — it should shape which analysis gets run. The ability to ask what changes if we compare only to our region, our Carnegie peers, our aspirants, or a list we choose ourselves is the difference between receiving an answer and participating in the decision.
Section 5: From signal to sustainability
A program can be feasible and still not be viable.
Feasibility asks whether a plausible market opportunity exists. Viability asks whether the institution can deliver the program in a way that is sustainable in practice. The two questions are related, but they are not the same question, and the gap between them is where avoidable mistakes live.
A market scan may show strong labor demand, growing completions, and limited competition — and the program may still fail under realistic assumptions, if it requires expensive faculty lines, specialized facilities, low student-to-faculty ratios, or clinical capacity the institution cannot secure. The reverse is just as true. A program with modest market signal can be the right decision because it strengthens an existing cluster, supports mission, uses faculty capacity already on the books, or balances a portfolio that is over-concentrated elsewhere.
Exhibit 5 Feasible is not the same as viable
Viability: can we sustain it?
Often wrongly rejected
Strategic case
Modest market signal, strong fit: mission, existing capacity, portfolio balance.
Strong on both
Proceed
Real demand and a model that sustains itself.
Weak on both
Decline
Little market signal, no sustainable model. Little reason to commit.
Often wrongly launched
Caution
Attractive in the market, unsustainable in the budget.
Feasibility: is there a market opportunity? Low to high
That is why a program decision should not stop at demand. It should connect the market signal to the things that actually determine sustainability — required investment, revenue potential under realistic tuition and discount assumptions, expense structure, contribution margin, and fit with the rest of the portfolio — and it should do so before institutional enthusiasm hardens around a launch date. Treating feasibility and viability as connected-but-distinct layers is what lets leaders avoid the two symmetrical errors: launching programs that look attractive but cannot sustain themselves, and rejecting programs that would work under a model no one took the time to test.
The cost of skipping that step usually shows up later, as a correction — and corrections of this kind are public. When a portfolio drifts out of alignment with what an institution can sustain, the fix can look like the one West Virginia University's board approved in 2023: the elimination of 28 academic majors and 143 faculty positions to close a $45 million shortfall, carried out over faculty no-confidence votes and weeks of protest.9 A faculty line is a multi-decade commitment, and a launched program accumulates students, expectations, and obligations that make it slow and costly to unwind. Viability analysis at the front end is, in part, insurance against having to run that process at the back end.
Section 6: The value question
"Value" is not one number.
It is tempting to reduce program value to a single earnings figure: graduates earn more, therefore launch. That framing is too thin to survive a serious governance conversation, and it misreads what most institutions are actually optimizing for. Value in higher education is economic, but it is also educational, institutional, and strategic — and a program that scores high on one dimension may score low on another.
A more honest evaluation reads an opportunity through several lenses at once and lets the tensions among them surface, rather than collapsing them into a verdict.
| Lens | The question it asks |
|---|---|
| Student value | Does this create meaningful educational, career, or mobility value for the students it enrolls? |
| Workforce value | Does it respond to a real employer, industry, regional, or professional need? |
| Institutional value | Does it strengthen mission, reputation, differentiation, or portfolio balance? |
| Financial value | Can it operate sustainably under realistic enrollment, tuition, and cost assumptions? |
| Strategic value | Does it position the institution for where the market and the learning environment are heading? |
Laid side by side across several candidate programs, the lenses do real work: they show that the strongest market opportunity is not always the strongest institutional fit, and that a program with modest financials can still earn its place on strategic grounds. The judgment stays with leadership — but it becomes a judgment made in full view of the trade-offs.
Exhibit 6 The value lens, applied: three Riverside candidates
| Program | Student | Workforce | Institutional | Financial | Strategic |
|---|---|---|---|---|---|
| DNP (charted) | High | Med | High | Med | High |
| Medical Science Liaison cert. (emerging) | Med | High | Low | Med | High |
| MA Communication (mature) | High | Med | High | Low | Med |
Section 7: The response
From a report to a decision environment.
One point sits under all of this: the important program questions are rarely settled on the first pass. A dean proposes a different staffing model. Finance wants a more conservative ramp. The provost asks for regional peers instead of national ones. The president wants to know whether the idea could start as a certificate. In a report-based process, each of these is a delay — a revision request, a new engagement, weeks of waiting. In a better decision process, each one is simply the next question you test.
That shift — from receiving a framed analysis to participating in the framing — is the real change, and it is what a decision environment is for. That environment should let leaders move fluidly across a connected set of questions: Is there a market? Who are we really competing against? Can we deliver it? Will it work financially? How does it fit the portfolio? Which assumptions matter most? And it should handle both lanes — rigorous feasibility when the category is known, and structured exploration when the opportunity is still taking shape.
This is the premise The Provost Lab was built on. It brings feasibility analysis, exploratory mode for emerging fields, flexible comparator selection, viability testing, the value lens, and portfolio review into the same working space — so that the institution's own knowledge enters the analysis directly instead of arriving afterward as a footnote. The institution still owns the question; the environment only gives its judgment a faster, clearer place to answer it.
Conclusion
Better questions, asked faster, in your own context.
Academic program decisions are too consequential — and, in this market, too expensive to get wrong — to be settled by a static report or a generic national scan. Institutions need evidence, but they also need context, financial realism, strategic alignment, and the ability to test assumptions before they commit resources. They need to know not only whether a program has demand, but whether it can work here.
The classification systems that anchor traditional feasibility will keep doing useful work for established programs. But they will keep lagging the fields that have not yet earned a code — which are, increasingly, the fields worth exploring first. What that lag calls for is not a longer report but a different way of working — one that can take the charted and the uncharted equally seriously, and that leaves the institution in control of the question being asked.
The goal was never to replace the wisdom of academic leadership. It is to give that wisdom better instruments — and a faster way to see what changes when the assumptions do.
The Provost Lab
The Provost Lab is built to help colleges and universities evaluate new program opportunities, explore emerging fields, test financial viability, and weigh portfolio fit. It starts with public data, but it does not stop there. The institution can bring its own assumptions, peer sets, financial model, and strategic priorities into the same conversation. Rather than delivering a single static answer, it gives leaders a place to ask better questions — and to see how the answers move when the assumptions change.
Sources
- Lane, P., Falkenstern, C., & Bransberger, P. (2024). Knocking at the College Door: Projections of High School Graduates (11th ed.). Western Interstate Commission for Higher Education. wiche.edu/knocking
- National Association of College and University Business Officers. (2025). 2024 NACUBO Tuition Discounting Study. nacubo.org
- State Higher Education Executive Officers Association. (2026). State Higher Education Finance: FY 2025. shef.sheeo.org
- Carnevale, A. P., Smith, N., et al. (2023). After Everything: Projections of Jobs, Education, and Training Requirements through 2031. Georgetown University Center on Education and the Workforce. cew.georgetown.edu
- Huron Consulting Group. (2026). Analysis of private nonprofit college closure and merger risk. Reported in The Hechinger Report. via governing.com
- Kelchen, R. (2025). Analysis of private nonprofit institutional finances (nearly one-third posted operating deficits in 2024). See also Kelchen, R., Ritter, D., & Webber, D. A. (2024). Predicting College Closures and Financial Distress. Federal Reserve Bank of Philadelphia, Working Paper 24-20.
- National Center for Education Statistics. Classification of Instructional Programs (CIP 2020). U.S. Department of Education. nces.ed.gov/ipeds/cipcode
- U.S. Bureau of Labor Statistics & Office of Management and Budget. 2018 Standard Occupational Classification (SOC) System. bls.gov/soc
- West Virginia University program and faculty reductions (2023). Reported in Inside Higher Ed (Aug. 11, 2023) and The Washington Post (Sept. 15, 2023). insidehighered.com