A 2026 State Agenda for Higher Education Reform

Key Points

• States are a major actor in American higher education, yet most reform attention is aimed at Washington. State governments fund and regulate the institutions that enroll most American undergraduates, but rarely face accountability for the outcomes those institutions produce.
• The Texas State Technical College Returned Value Formula, which ties 100 percent of TSTC’s state appropriation to its graduates’ wages, is the most successful earnings-based funding model in American higher education.
• The One Big Beautiful Bill Act of 2025 raises the federal accountability floor and gives governors new certification authority over Workforce Pell programs. States that do not match the federal standard will fall behind.
• A five-pillar state reform agenda—earnings-based funding, rigorous Workforce Pell certification, ROI-based tuition subsidies, authorization reform, and credential inflation reform—would align state higher-education spending with labor-market outcomes.

I. Why the State Agenda Matters

Most of the public debate about higher education policy in the United States is aimed at Washington. Congress sets the rules for federal loans and grants; the Department of Education (ED) writes the regulations; federal reform bills attract the attention of the national press. And because Washington writes the big checks—$125 billion a year under Title IV of the Higher Education Act of 1965—it is understandable that the reform community spends most of its energy on federal policy.

But the federal checkbook is not the only important lever in American higher education. States spend $127 billion a year on operating support for public colleges and universities, separately from the federal Title IV flow. They set tuition at the public institutions that enroll roughly three-quarters of American undergraduates. They charter and regulate both public and private degree-granting institutions. They administer state grant programs—Cal Grants, HOPE, Bright Futures, Tennessee Promise, and dozens of others—that in many states exceed federal Pell Grant spending within their borders. (Pell Grants are need-based federal financial aid.) States set occupational licensing rules that determine which credentials are required for employment in everything from barbering to nursing. And they run the workforce development systems that, beginning in the 2026–27 award year, will play a decisive role in certifying the new Workforce Pell programs under the OBBBA.

Take a student deciding whether to enroll at a state university. The sticker price she pays is set by a state board of regents. The tuition discount she receives is funded partly by state grant dollars. The program she chooses is approved by a state accrediting or authorizing body. The degree she earns is required, or not required, for her eventual job by a state licensing board. The portion of her loan she might ultimately default on is a federal dollar, but almost every other variable determining whether she ends up in a good financial position has been decided at the state capital. Federal reform matters, but most of the machinery of American higher education sits in Albany and Austin and Tallahassee and Sacramento, not in Washington.

That is the premise that we explore in this paper. The federal agenda that we laid out in a companion paper on federal higher education reform—risk-adjusted institutional accountability, program-level loan caps, a federal return-on-investment database—is necessary but not sufficient. States can do things the federal government cannot.

States can rewire the funding formula of their public institutions so that state appropriations follow the earnings of graduates rather than the headcount of enrollees. They can use the governor’s certification authority under OBBBA’s Workforce Pell to establish a short-term credential ecosystem that genuinely delivers on wage gains. They can subsidize in-state tuition at programs where the return on investment is positive and decline to subsidize programs where it is negative. They can lower the regulatory barriers that prevent serious new institutions from launching. And they can stop requiring four-year college degrees for jobs that do not actually need them—the credential inflation problem that has quietly reshaped the American labor market over the last forty years.

This paper proposes a reform agenda for the states. It has five pillars:

1. Earnings-based state funding, modeled on the Texas State Technical College Returned Value Formula. State operating appropriations for public institutions should be tied to the wage outcomes of their graduates, not to their enrollment or their graduation rates.
2. Governor-led Workforce Pell implementation, with meaningful quality standards. OBBBA Section 83002 gives governors the statutory certification authority for Workforce Pell programs. State workforce boards should build rigorous, data-driven approval processes that use the state’s existing unemployment insurance wage records.
3. ROI-based in-state tuition subsidies. States should direct their tuition-discount dollars toward programs that generate positive lifetime returns on investment, and away from programs that do not.
4. Regulatory reform to ease the entry of new institutions. The University of Austin experience has shown that state authorization and accreditation barriers are higher than they need to be; state legislatures should lower the barriers without compromising consumer protection.
5. Credential inflation reform in state government hiring. States should systematically review job descriptions across state payrolls, identify roles where a four-year degree is required but not needed, and remove the requirement.

As with our federal reform paper, each pillar has a legislative path and an executive path. The legislative path is usually cleaner, but governors and state boards of education have substantial authority to act on several of these items without new legislation. Where the executive path is viable, we describe it.

II. The Texas State Technical College Model

How TSTC was reformed

In 2013, the Texas General Appropriations Act implemented a new operating formula for the Texas State Technical College System known as the Returned Value model. The legislature had mandated development of the formula two years earlier, and had directed the Texas Higher Education Coordinating Board to study its feasibility in a bill enacted in 2009. The methodology was published in July 2013 by the Coordinating Board.

Before 2013, TSTC—the state’s public technical college system, with campuses in Waco, Harlingen, Marshall, Sweetwater, Fort Bend, Abilene, and several other locations—was funded according to the same enrollment-based formula that the Texas Higher Education Coordinating Board applies to community colleges and four-year universities. Contact hours generated appropriations. Students who enrolled in October and disappeared in November produced almost the same revenue for the institution as students who completed programs and went to work. The incentive structure rewarded getting bodies in the door, not getting them back out.

The Returned Value formula replaced the contact-hour measure with an earnings-based measure. Under the formula, the Texas Higher Education Coordinating Board calculates the annual wage earnings of TSTC’s former students—drawn from Texas Workforce Commission unemployment insurance records—for the five years following their departure from the institution. From each former student’s earnings, the board subtracts a base-wage amount equal to full-time work at the minimum wage. The difference is the student’s incremental wage, attributable to their TSTC education. The formula multiplies this incremental wage by seven percent to approximate the state sales, excise, and franchise tax revenue the student’s higher wages generate for the state treasury. It then multiplies by 1.5—an economic multiplier drawn from a Bureau of Economic Analysis study of secondary labor-market effects—to approximate the indirect tax revenue the student’s earnings generate through consumption and downstream employment. The resulting number is the total “returned value” the TSTC system has generated for the state.

A fixed percentage of the returned value—in the legislation, the number is approximately 36 percent, though the legislature has sometimes appropriated less—is then paid to TSTC as its operating appropriation. The formula is applied to each TSTC campus separately, so each campus’s operating budget is determined by its own graduates’ earnings, not by the system-wide average. An individual TSTC campus that increases its graduates’ wages receives more state funding the next biennium; a campus that does not, receives less.

The intellectual elegance of the formula is that it turns the state legislature’s accountability question on its head. Rather than asking whether TSTC has produced some number of degrees or spent its appropriation on allowable activities, the legislature asks whether TSTC has produced measurable labor-market value for the state. The legislature does not have to audit the institution; it only has to look at the earnings records. And because the Texas Workforce Commission already maintains unemployment insurance wage records for nearly every employed Texan, the data infrastructure already exists.

What TSTC has achieved

The results at TSTC over the past decade are striking. The Returned Value formula improved former students’ wages by 45 percent over eight years, a rate of growth that exceeded the Texas economy-wide wage growth rate by a substantial margin. As of 2021, TSTC reported median first-year wages for associate degree graduates of nearly $49,000 in 2023 dollars, a figure that surpassed the median first-year earnings of Texas two-year college graduates.

The institution’s program mix shifted in response to the incentives. TSTC closed programs whose graduates did not generate measurable earnings gains and expanded programs in fields where the wage data indicated strong demand: welding, instrumentation technology, process operator training, robotics, and emergency services. The institution also invested in placement services, employer relationships, and apprenticeship partnerships, which it had under-resourced under the old enrollment-based formula because none of those activities generated contact hours. TSTC began actively managing its own labor-market outcomes, not because a federal accountability rule required it, but because the state funding formula had created the incentive to do so.

The Texas Higher Education Coordinating Board, using Unemployment Insurance wage records supplied by the Texas Workforce Commission, has published the system’s performance against the formula each biennium since the 2014–15 cycle. The data are available publicly, the methodology is documented, and the calculations are reproducible. That transparency is itself a consequence of the formula design: because the legislature wanted to verify what it was paying for, the formula forced the Coordinating Board to publish the wage-outcomes data that drive it.

Why TSTC has not been replicated

If the Returned Value formula works so well, why does Texas still operate it for only its technical college system and not its community colleges or its four-year universities? And why has no other state adopted a formula of the same structure in the twelve years since TSTC’s launch?

The answer is partly political and partly substantive, and the two are entangled. The Returned Value formula reallocates funding toward institutions whose graduates earn more, which means any extension produces predictable losers. Institutions whose graduates do not generate strong wage outcomes—including most regional comprehensive universities, many community colleges serving large transfer or adult-learner populations, and four-year programs concentrated in humanities and social-service fields—have every incentive to oppose the model. In states where flagship universities and community college systems dominate the higher-education politics, that opposition is usually sufficient to keep the formula off the table altogether.

The Texas community college case illustrates the dynamic in its more sophisticated form. When the Texas Commission on Community College Finance deliberated in 2021 and 2022, its commissioners—appointed in part by the Texas Association of Community Colleges (TACC) and the Community College Association of Texas Trustees, and advised throughout by TACC leadership—designed a formula that pointedly avoided a single-metric wage approach. They opted instead for a multi-metric structure built around credential completions, transfers, dual-credit completions, and weighted incentives for adult learners and economically and educationally disadvantaged students, on the rationale that community colleges serve large populations of part-time students, adult learners, transfer students, and others whose outcomes are hard to capture in short-run state Unemployment Insurance (UI) wage data. The commissioners explicitly deferred wage and job-placement metrics for possible later inclusion.

The community colleges’ argument has some merit, but its accommodation made the realignment of incentives more difficult. The commission’s response, codified in House Bill 8 of 2023, was to retain an earnings logic for short-cycle credentials—the “credentials of value” designation requires a demonstrated wage premium—but to surround it with weighted credits for transfers, dual-credit completions, and credentials earned by adult and economically disadvantaged students. The result is a more complex, less powerful instrument than TSTC’s. Community colleges accepted it, and indeed publicly endorsed it, precisely because the multi-metric design diluted the redistributive force of a pure wage formula. The pure version never reached a floor vote.

Simple formulas are better. Earnings-based funding works cleanly where the institution’s programs are tightly coupled to labor-market outcomes and its graduates are relatively easy to track in the state’s wage records. Such an approach is harder to implement, though not impossible, at institutions whose students enter part-time, stop out and return, transfer to other states, or pursue credentials that are not occupation-specific. And it faces the most opposition at four-year universities, where graduation rates are lower than at TSTC, variance across programs within an institution is high, and the formula’s institutional-level outputs would produce uncomfortable rankings of programs that flagship administrators have historically been measured on inputs rather than outputs. Texas itself still funds its general academic institutions through a weighted-semester-credit-hour formula that bears no resemblance to the Returned Value model, and no serious legislative effort has been made to change that since TSTC’s formula was enacted in 2013.

The right response is not to abandon the model but to adapt it. The next section of this paper describes how.

III. Extending Earnings-Based Funding Beyond TSTC

General design principles

A state earnings-based funding formula has four parameters. The designers of any given state’s formula need to choose each of them deliberately, because different choices produce very different institutional incentives.

Parameter 1: Which students count? A formula that measures the wages of all former students—completers and non-completers alike—captures the full labor-market impact of the institution. A formula that measures only the wages of completers creates an incentive for the institution to award credentials to borderline students, which is itself a form of gaming. TSTC’s formula looks at all former students for five years after their last enrollment, which is a reasonable compromise. We recommend that compromise as the default for other states: measure earnings for all former students, including non-completers, for the five years after their last enrollment, and weight the measurement by the number of credit hours the student completed while enrolled.

Parameter 2: What counts as the earnings measure? The TSTC formula uses wages minus the full-time minimum-wage benchmark. The minimum-wage benchmark is a convenient choice but has two problems. First, it does not adjust for the counterfactual earnings a student could have expected without enrolling in the program. A student who would have earned $40,000 without enrolling but earns $45,000 with a credential has a $5,000 incremental wage, which is worth less than it appears because the counterfactual was already far above the minimum wage. Second, the minimum wage is a political variable; raising the state minimum wage, which is unrelated to higher-education policy, mechanically reduces the formula’s output. A better approach, which we recommend for new state implementations, is to use as the counterfactual benchmark the median earnings of working adults aged 25–34 in the state with only a high school diploma—the same benchmark that OBBBA Section 84001 now uses at the federal level. This choice matches the federal accountability framework, uses the same Census data, and is robust to changes in the state minimum wage.

Parameter 3: What fraction of the calculated value goes to the institution? TSTC’s roughly 36 percent rate implies that the state retains about two-thirds of the tax revenue generated by TSTC graduates and pays back one-third as the institution’s operating appropriation. A higher fraction gives the institution stronger incentives, but also exposes state budgets to larger swings in response to economic cycles. A lower fraction softens the incentive. Our recommended starting point for new state adopters is 25 percet, with a floor-and-ceiling band of plus or minus 20 percent against a three-year average, so that no institution sees its appropriation swing by more than 20 percent in any biennium as a result of the formula.

Parameter 4: What is the institutional unit of outcomes measurement? The TSTC formula is applied campus by campus. A stronger formula would be applied program by program within each campus, so that a campus with some strong programs and some weak ones cannot cross-subsidize the weak ones from the strong ones. Program-level measurement is harder because smaller cohorts generate noisier earnings data, but the problem can be managed by the same aggregation rules OBBBA uses for Section 84001: programs with fewer than 30 students in the cohort have their data aggregated with related programs or across years until a viable cohort is reached. We recommend program-level measurement for systems that have the data infrastructure to support it, with campus-level measurement as a fallback for systems that do not.

A community college variant

The objection to applying the TSTC formula to community colleges is that community colleges serve populations whose labor-market outcomes are harder to measure: part-time students, adult learners, workforce re-entry students, immigrants pursuing English as a second language, and students transferring to four-year institutions. Each of those populations raises a legitimate measurement question.

None of these issues are insurmountable. For part-time students, the formula should count only the earnings of former students who have completed at least 12 credit hours, which excludes the most transient enrollees. For transfer students, the formula should exclude students who have subsequently re-enrolled in any institution of higher education, which is exactly how OBBBA’s Section 84001 handles the problem. For workforce re-entry students, the formula should compare the student’s earnings before and after enrollment, using the unemployment insurance records that state workforce agencies already maintain. And for immigrant learners in ESL or foundational programs, the formula should treat those programs as a separate category with its own benchmark; perhaps one that measures progression into credit-bearing coursework rather than direct wage outcomes.

Each of these adjustments is a complication in the formula, but not a reason to abandon it. A community college earnings-based formula that incorporates all four adjustments is more complex than the TSTC model, but it is still simpler than the performance-funding formulas many states have tried and abandoned over the last 20 years.

The performance-funding cautionary tale

About half of U.S. states have adopted performance-based funding formulas for higher education at some point over the last two decades. The research literature on these policies is voluminous. The most comprehensive synthesis was published in 2020 by researchers Justin Ortagus, Robert Kelchen, Kelly Rosinger, and Nicholas Voorhees. Ortagus and colleagues reviewed dozens of empirical studies of state performance-funding policies and found that the policies produced small or null effects on the intended outcomes—retention, graduation, and credential completion—and substantial unintended effects on access, gaming, and equity.

The literature identifies three recurring failure modes. First, most performance-funding formulas measure intermediate outcomes like retention and graduation rather than ultimate outcomes like earnings, and institutions respond by pursuing the intermediate measure in ways that do not improve the ultimate goal. A university that is paid for six-year graduation rates can raise the graduation rate by admitting students who are more likely to graduate, which disadvantages the Pell-eligible first-generation students whose graduation prospects are lower—even if the institution could have served those students well.

Second, performance-funding formulas that measure a dozen metrics simultaneously create incoherent incentives. If an institution is paid a few dollars for each of graduation rate, retention rate, Pell graduation rate, STEM graduation rate, minority graduation rate, research expenditure, and degrees conferred per faculty member; administrators cannot meaningfully optimize any one of them, and the marginal dollar allocated to performance has almost no behavioral effect.

Third, most performance-funding formulas put a small share of state appropriations at risk—often 5 percent or 10 percent—which means that an institution that loses its performance-funding bonus still collects the vast majority of its appropriation. Weak incentives produce weak responses. The Ortagus synthesis finds that the formulas with measurable behavioral effects are the ones that put a substantial share of state funding on the line, and the TSTC formula is the extreme case: it puts 100 percent of TSTC’s instructional and administrative funding at risk, which is why TSTC responded so strongly to it.

An earnings-based formula on the TSTC model avoids each of the three failure modes. It measures the ultimate outcome, not the intermediate one; it uses one measure, not a dozen; and it puts a substantial share of state funding at risk. It is simpler than typical performance-funding designs and produces stronger behavioral responses because the incentives are more legible to institutional leaders.

Legislative model language

State legislatures considering an earnings-based funding formula can draw directly from the TSTC Returned Value formula, and from the Texas Higher Education Coordinating Board’s published methodology and biennial formula funding recommendations. The model statute we recommend has the following structure.

First, a declaration of policy: legislatures should tie state operating appropriations for public institutions of higher education to the labor-market outcomes of the institutions’ former students, because those outcomes are the most direct measure of the return state taxpayers earn on their investment in higher education.

Second, a formula definition that specifies the annual wage-earnings measurement window (five years from last enrollment), the counterfactual benchmark (state median earnings for working adults aged 25 to 34 with only a high school diploma, drawn from the American Community Survey), the value-added calculation (wages above benchmark, capped at a generous per-student maximum to prevent outliers from distorting the formula), the state-retained share (75 percent of the calculated value, reflecting the 25 percent institutional share we recommend), and the unit of measurement (program-level, with aggregation rules for small cohorts).

Third, a data-sharing authorization that permits the state workforce agency to share unemployment insurance wage records with the state higher-education agency for the purposes of the formula. Most states already have this authorization for purposes of accreditation and program approval; the model statute extends it to funding calculations.

Fourth, a transition period. The formula should phase in over three or four years, beginning with a 25 percent weight in year one and rising to 100 percent by year four or five. Institutions need time to adjust, and a gradual transition reduces the risk that an unforeseen data problem disrupts the formula’s first implementation.

Fifth, an appeals process. Institutions must have a mechanism to challenge the formula’s outputs, including by providing alternative data on their former students’ earnings or by arguing that a particular program’s cohort should be measured differently. The appeals process should be adjudicated by an independent body, not by the higher-education agency that runs the formula.

Sixth, an annual report requirement. The higher-education agency must publish the formula’s inputs and outputs annually, along with each institution’s risk-adjusted performance. Public reporting is the mechanism by which the formula becomes self-correcting, because it exposes both the institutions that are performing well and the institutions that are not.

IV. Workforce Pell at the State Level

Governors’ new authority

Section 83002 of the One Big Beautiful Bill Act, enacted in 2025, creates the Workforce Pell Grant program and places the governor of each state in the certification chain for eligible programs. Beginning July 1, a short-term postsecondary program—between 150 and 599 clock hours, lasting between eight and fifteen weeks—can receive federal Pell Grant dollars only if the governor of the state, after consultation with the state workforce board, determines that the program meets four criteria:

1. It aligns with a high-skill, high-wage, or in-demand occupational sector as identified by the state under the Strengthening Career and Technical Education for the 21st Century Act of 2018, commonly known as Perkins V.
2. It meets the hiring requirements of potential employers in the identified sector.
3. It leads to a stackable, portable recognized postsecondary credential, or it prepares the student for an occupation with a single recognized credential delivered upon completion.
4. It prepares students to continue into a related certificate or degree program at an institution of higher education, with a guarantee that academic credit transfers.

After the governor’s certification, the U.S. Secretary of Education conducts a secondary verification: the program must have been offered by the institution for at least one year, must demonstrate a verified 70 percent completion rate within 150 percent of normal time, must demonstrate a verified 70 percent job placement rate measured 180 days after completion, and must cap its published tuition at the “value-added earnings” of the program’s graduates, defined as the median earnings of Title IV-recipient completers three years after graduation, adjusted for regional price parities, minus 150 percent of the federal poverty line for a single individual.

The governor’s role is the most important discretionary decision in the chain. The secretary’s verification is largely ministerial; the completion and placement rates are either documented or they are not, and the earnings cap is a mechanical calculation. But the governor’s certification is substantive. A governor can read the statute loosely and certify any program that a state workforce agency has historically approved; a governor can read it strictly and certify only programs that genuinely meet the high-skill, high-wage threshold. The difference in outcomes over the next decade will be enormous, and states that build rigorous certification processes will produce much better returns on their Workforce Pell investment than states that do not.

What a rigorous state process looks like

A state Workforce Pell certification process should have six elements.

Element 1: A published list of qualifying occupational sectors. Perkins V Section 122 already requires states to identify high-skill, high-wage, and in-demand occupations as part of their career and technical education planning. OBBBA Section 83002 incorporates that identification by reference. States should publish their qualifying-occupation list annually, tied to specific Standard Occupational Classification codes and supported by labor-market data from the Bureau of Labor Statistics and the state’s own workforce agency. The list should be narrow enough to mean something—not every occupation is high-skill or high-wage—and it should be updated annually as labor-market conditions change.

Element 2: Data-driven wage verification. The state workforce agency already holds unemployment insurance wage records for most of the state’s workforce, and those records can be used to verify claimed earnings outcomes. A Workforce Pell applicant should be required to provide the state workforce agency with a list of the Social Security numbers of its former students; the agency should match that list to its UI wage records and verify the claimed median earnings, the job placement rate, and the occupational alignment of the placements. This matching is already done, in various forms, by state agencies in at least twenty states under the Workforce Innovation and Opportunity Act; the infrastructure exists and can be repurposed for Workforce Pell.

Element 3: Transferable credit verification. OBBBA Section 83002 requires that Workforce Pell programs guarantee academic credit transfer to related certificate or degree programs. A Governor’s certification process should require that the applying institution provide a signed articulation agreement with at least one Title IV-eligible institution in the state specifying the number of credit hours the Workforce Pell program will transfer into a related associate or bachelor’s program. Without a written articulation agreement, the transferable-credit requirement is unenforceable.

Element 4: Employer engagement. The statutory requirement that the program “meet the hiring requirements of potential employers” is easy to fake. A rigorous certification process should require the applying institution to document, with letters of commitment from at least three employers in the identified sector, that the program’s curriculum has been reviewed and endorsed by employers who plan to hire its graduates. This is the Perkins V “employer-led” standard, translated into the Workforce Pell certification context.

Element 5: Annual recertification. Programs that pass initial certification should be recertified annually against their actual performance. A program that maintains its 70 percent completion and placement rates, and whose median graduate earnings continue to justify the tuition cap, should receive automatic recertification. A program that falls below the thresholds should receive a year of warning and then be decertified.

Element 6: Public transparency. Every Workforce Pell certification decision should be posted on a state-managed website with the underlying data. Prospective students, journalists, and researchers should be able to see which programs passed, which failed, and why. The transparency requirement is the single most effective tool for preventing the certification process from degrading into a rubber-stamp exercise for politically connected institutions.

What to avoid

The history of state workforce training policy is littered with programs that paid for credentials that did not deliver. The clearest cautionary tale is that of the for-profit Corinthian Colleges chain, which at its 2010 peak operated 105 campuses enrolling more than 110,000 students under brands including Everest, Heald, and WyoTech. Through the early 2010s, state Eligible Training Provider Lists across the country approved Corinthian programs for WIOA-funded training dollars on the basis of the institution’s self-reported completion and job placement rates. Those self-reports turned out to be systematically falsified. In October 2013, then-California Attorney General Kamala Harris sued Corinthian for misrepresenting placement outcomes, ultimately winning a $1.1 billion default judgment in 2016. A 2015 Department of Education investigation found, for instance, that one of Corinthian’s campuses had advertised an 85 percent placement rate for its medical-assistant program when the actual rate was zero; a WyoTech automotive program in Long Beach had claimed 80 percent against an actual 26 percent. Corinthian collapsed in April 2015, stranding approximately 16,000 students mid-program.

The states that have had better experience—Texas, North Carolina, and Indiana—have built certification processes that are driven by labor-market data rather than by institutional self-reporting. The lesson for Workforce Pell is that the governor’s certification should lean heavily on the state workforce agency’s existing wage-records infrastructure and resist any temptation to grant automatic approval to programs that have been certified for Perkins V or WIOA funding under looser standards.

The other pitfall to avoid is occupational licensing capture. Some states have a long history of using the workforce development process to funnel public dollars into programs that prepare students for jobs requiring state-issued occupational licenses, even when the underlying license is itself excessive. A Workforce Pell program that trains students for an occupation where licensing has been shown to restrict entry without improving consumer outcomes—cosmetology, interior design, many branches of construction, some forms of retail food sales—is not delivering what the statute calls a “recognized postsecondary credential” in any meaningful sense. States should distinguish between programs that prepare students for genuinely in-demand occupations and programs that merely prepare students to satisfy licensing boards with captured interests.

V. ROI-based In-State Tuition Subsidies

What states currently do

Every state operates some form of in-state tuition subsidy. The most common form is the tuition differential between in-state and out-of-state rates at public colleges and universities: a subsidy that is large but entirely uniform across programs within an institution. A student who enrolls in a computer science program at a state flagship pays the same in-state tuition rate as a student who enrolls in a program whose graduates earn less than they would have without the credential, even though the state treasury’s return on the two subsidies is radically different.

A smaller number of states operate merit-based tuition programs—Georgia’s HOPE scholarship, Florida’s Bright Futures, Tennessee’s Promise, West Virginia’s PROMISE—that cover a substantial share of in-state tuition for students who meet academic criteria. These programs are also uniform across programs: a HOPE scholarship is worth the same amount whether the student enrolls in nursing or studio art. And many states operate direct grant programs—New York’s TAP, California’s Cal Grant, Massachusetts’s MASSGrant—that subsidize need-based financial aid without any regard to program ROI.

The uniform structure of these subsidies reflects an implicit assumption: that all undergraduate programs are equally valuable to the student and the state, and that the state’s policy interest is in promoting enrollment generally rather than enrollment in particular programs. That assumption was always questionable, and after a decade of College Scorecard data it is no longer defensible. FREOPP’s ROI database reports that approximately 28 percent of bachelor’s programs in the United States produce negative lifetime returns on investment for the typical student, adjusted for the risk of non-completion. States that subsidize in-state tuition at those programs are using public dollars to finance financial harm to their own residents.

A tiered subsidy based on ROI

Our proposed reform is to tier in-state tuition subsidies by program ROI. The structure we recommend has three tiers.

Tier 1: Full subsidy for programs with positive lifetime ROI. Programs that generate positive lifetime net present value for the typical student—using the FREOPP methodology, with state-specific earnings data—should receive the full in-state tuition differential, the full merit scholarship, and the full need-based grant aid that the state currently offers.

Tier 2: Partial subsidy for programs with marginal ROI. Programs whose lifetime ROI is slightly negative—within a defined band—should receive reduced subsidies. In-state tuition should remain discounted but at a smaller rate; merit scholarships should be reduced; and need-based grants should remain available but at a lower maximum. The partial subsidy preserves some access to the program while putting a price signal on its marginal value to the state.

Tier 3: No state subsidy for programs with deeply negative ROI. Programs whose lifetime ROI is well below zero—again within a defined band—should receive no state tuition differential, no merit scholarship, and no need-based grant aid. Students who wish to enroll in these programs may do so at out-of-state tuition rates, and they may use federal Pell Grants and Stafford loans if they qualify, but the state treasury should decline to participate.

The tier boundaries should be set by the state’s higher-education agency, based on its own analysis of the College Scorecard data and its own judgment about how generous the state wishes to be. A conservative implementation might set the tier three boundary at minus 20 percent lifetime ROI, so that only the most deeply negative programs are disqualified. A more aggressive implementation might set the boundary at minus 10 percent or at zero.

The effect on institutional behavior is the important part. A flagship state university that has allowed a negative-ROI program to persist because the program has a long history and an engaged alumni base will face a direct revenue cost if the state’s subsidy is withdrawn. This is not because the state has imposed a penalty, but because the students who would have enrolled at subsidized tuition will instead enroll elsewhere, and the institution’s own revenue will fall. The universities can respond in any of several ways: raise the program’s labor-market outcomes through curricular reform, lower the program’s tuition to match the lower state subsidy, close the program entirely, or accept a smaller revenue base. Each of those responses is a better outcome than the current equilibrium.

Equity considerations

Some may worry that a tiered subsidy structure will disadvantage low-income students who want to pursue programs in the lower tiers. First, students who are interested in the humanities, the arts, and some social sciences will find their enrollment subsidized at lower rates due to low ROI, which could reduce their access to those fields. Second, low-income students who cannot afford to self-finance education at programs in the lower tiers will be steered toward higher-ROI programs regardless of their interests.

Both parts of the objection deserve serious engagement. The first is the harder one. States have historically treated the humanities and the arts as public goods that deserve subsidy regardless of labor-market outcomes, and there is a real argument that a society without state support for literature, music, and philosophy is diminished. But that doesn’t mean policymakers need to persist with an undifferentiated subsidy. A state that wishes to support the humanities and the arts should fund them directly: through targeted appropriations to humanities programs, through state arts agencies, or through fellowship programs for graduate students in those fields. A targeted subsidy is transparent about what it is doing and can be sized to the value the state actually places on the activity. An untargeted subsidy hidden inside a general tuition discount is not obviously more generous and is certainly less efficient.

The second part of the objection is even easier to address. Low-income students who are steered toward higher-ROI programs by the state subsidy structure are, on average, better off than low-income students who enroll in lower-ROI programs, because the graduates of higher-ROI programs earn more, service their debts more reliably, and accumulate more wealth over their lifetimes. The College Scorecard data are unambiguous about this, and the FREOPP ROI database documents it in detail. A tiered subsidy that steers low-income students toward better outcomes is not paternalistic; it is respectful of the fact that public subsidies come from public funds and should serve the interests of the public, which includes the interests of the students themselves. If a particular student is strongly committed to a program in the lower tiers and is willing to pay the unsubsidized price to attend, nothing in the tiered structure prevents that student from doing so.

VI. Removing Regulatory Barriers to New Institutions

It is almost impossible to start a new college

The University of Austin (UATX)—a new, classically oriented liberal arts institution launched in 2021 by a group of scholars led by Pano Kanelos—attempted to become an accredited, degree-granting institution in the state of Texas between 2021 and 2024. The experience illustrated how difficult it is to start a new college in the United States even when the institution is well-funded, backed by credible academics, and operating within a state whose political leadership was favorable. In effect, the certification and accreditation process is Kafkaesque, with colleges forced to prove that they can successfully educate students before they have enrolled any.

The institution had to obtain authorization from the Texas Higher Education Coordinating Board before it could operate as a degree-granting institution at all. The authorization process took approximately two years and required the institution to demonstrate financial viability, curricular coherence, faculty qualifications, facilities adequacy, and a substantial list of other standards. The institution then had to apply for candidacy for accreditation from a recognized accrediting agency—in UATX’s case, the Middle States Commission on Higher Education—which imposed its own separate set of standards and a longer timeline. Candidacy for accreditation is a preliminary status; full accreditation takes several additional years of compliance with the accreditor’s standards and periodic reviews. UATX opened applications for its first class of students in 2023 and undergraduates started classes in 2024; full accreditation is expected between 2028 and 2031.

Throughout this process, UATX could not participate in federal Title IV student aid, because Title IV eligibility requires accreditation. The institution had to finance its students’ tuition through its own scholarship program, drawn from philanthropic donations. A less well-funded institution would not have been able to survive the multi-year gap between opening its doors and obtaining Title IV eligibility.

The regulatory costs UATX incurred may have been well-intentioned in their origins. State authorization standards protect consumers from fraudulent operators. Accreditation standards protect students from institutions that cannot deliver on their curricular promises. Title IV eligibility requirements protect the federal taxpayer from subsidizing institutions that cannot survive. Each of the individual standards has a plausible justification. The cumulative effect of the standards, however, is that it is nearly impossible to launch a new college in the United States within a reasonable timeframe. The entry barriers fall most heavily on innovative institutions whose curricular or pedagogical models are not familiar to the accrediting agencies, which is exactly the category of institution the higher-education sector most needs.

What states can do

The states cannot unilaterally lower the federal Title IV eligibility barriers or the accreditation barriers. Those are set by the Department of Education and by the recognized accrediting agencies, and the only lever the states have is to lobby for federal reform. But the states can lower their own authorization barriers, and they can do so without compromising consumer protection.

The reform has four components.

First, a time-bounded authorization process. State higher-education agencies should commit to a maximum timeline for authorization decisions—we recommend 180 days from submission of a complete application—with automatic approval if the agency fails to act. The deadline forces the agency to allocate resources to the review and prevents the indefinite delay that drives up entry costs. Some state agencies already operate under such deadlines for other regulatory activities; extending them to higher-education authorization is a straightforward legislative fix.

Second, a reciprocity framework for institutions authorized in other states. An institution that has already obtained authorization in one state and has operated successfully for a defined period should be eligible for expedited authorization in other states, without having to repeat the full original process. The model here is the State Authorization Reciprocity Agreement (SARA), which governs cross-state online education, but the reciprocity framework should be expanded to brick-and-mortar institutions as well.

Third, a lighter-touch review for institutions of a defined scale. A new institution that opens with a small initial enrollment—say, fewer than 500 students—and a limited curricular offering should face a less burdensome authorization process than a large institution that opens with thousands of students and dozens of programs. The risk to consumers is proportional to the institution’s scale, and a scaled regulatory burden is more efficient than a uniform burden.

Fourth, a legislative presumption in favor of non-traditional pedagogical models. Several states have authorization standards that implicitly require new institutions to look like traditional research universities, with particular faculty-student ratios, particular governance structures, and particular curricular requirements. These standards disadvantage institutions that have adopted alternative pedagogical models, including classical liberal arts institutions like UATX, technical institutions that focus on applied learning, and apprenticeship-based institutions. A legislative reform that clarifies the authorization standards as outcome-focused rather than process-focused, and that explicitly permits non-traditional models, would lower the entry barriers for the institutions most likely to generate genuine innovation.

The broader case

The deeper reason to lower state authorization barriers is that American higher education is short on competition. The total number of degree-granting institutions in the United States has actually declined over the past decade, as closures and mergers have outpaced new openings. The institutions that dominate the sector are, in many cases, the same institutions that have dominated it for a century, and the operational and pedagogical innovations that should come from new entrants are instead coming from inside existing institutions, where it is always harder for genuine change to occur.

Lowering the barriers to entry is not the same as abolishing consumer protection. A state regulatory regime can be both more permissive of new institutions and more demanding of existing ones. The right mix of those two instruments varies state by state, but the direction of change is the same: the barriers to entry are currently too high, and the disciplinary mechanisms applied to incumbents are currently too weak. Section II and III of this paper addressed the second problem through earnings-based funding. This section addresses the first through authorization reform.

VII. Credential Inflation in State Government Hiring

The problem

In 1973, approximately 28 percent of American jobs required some form of postsecondary credential. By 2020, the figure was over 60 percent, and Anthony Carnevale of Georgetown University projects that it will exceed 70 percent by 2031. Most of that growth reflects real changes in the skills required for the work, but a substantial share—probably a quarter or more, depending on the methodology—reflected credential inflation: the upward drift of credential requirements for jobs that did not actually need them. The clearest documentation of credential inflation is the 2017 study Dismissed by Degrees by Joseph Fuller and Manjari Raman of Harvard Business School, which reviewed 26 million online job postings and found that 6.2 million middle-skills jobs are at risk of degree inflation. For example, for the occupation of supervisors of production workers, 67 percent of job postings in 2015 required a degree, even though only 16 percent of such workers actually had one.

Credential inflation is a problem for several reasons. It raises the effective cost of entry to the labor market for workers who do not have the credentials, even when they have the skills. It generates excess demand for higher-education credentials, which feeds back into institutional tuition pricing. It discriminates in effect against non-college-graduates, who are disproportionately first-generation, minority, and lower-income workers. And it creates employment inefficiency, because employers are paying for credentials that do not contribute to productivity.

State governments are unusually good places to attack credential inflation, because state governments employ millions of workers and the requirements for state jobs are set by state policy rather than by private employers. A state that eliminates the four-year degree requirement for positions that do not need one immediately expands its own hiring pool and sets a precedent that private employers often follow.

What states have already done

Several governors have issued executive orders along these lines over the past five years. In 2022, Maryland Governor Larry Hogan led an effort to review job descriptions and eliminate degree requirements for positions where the degree was not genuinely necessary. Pennsylvania Governor Josh Shapiro issued a similar order in 2023. Utah Governor Spencer Cox issued one in 2022. New Jersey, Colorado, Virginia, Alaska, and North Carolina have followed. The cumulative effect is that, by 2024, approximately half the states had formal policies in place to review and eliminate unnecessary degree requirements from state government hiring.

The results have been meaningful but limited. Maryland’s implementation of the Hogan order led to the review of roughly 38,000 state positions and a 41 percent increase in non-degreed hires from May to August of 2022. Pennsylvania’s review removed degree requirements from 92 percent of state executive branch job listings. Utah’s review covered 98 percent of classified state jobs, and produced similar results. In each case, the percentage of state hires without a four-year degree increased in the year following the order, though not by as much as the formal policy change would suggest because hiring managers continued to rely on degree preferences in practice even when the formal requirement had been removed.

What the next wave of reform should look like

States that have not yet adopted credential inflation reform should adopt it. States that have adopted it should sharpen the enforcement. The agenda for the next wave of reform has four elements.

Element 1: A statutory presumption against degree requirements. Credential inflation reform by executive order is reversible by the next governor, and several states have seen partial or full reversals when administrations have changed. A statutory presumption—written into the state’s merit system or civil service code—would lock-in the reform and create a default against degree requirements for positions where they are not demonstrably necessary.

Element 2: A rigorous review process for exceptions. A state agency that wishes to retain a degree requirement for a particular position should have to justify the retention to an independent review body, using evidence about the actual duties of the position. The review body should publish its decisions so that the pattern of exceptions is visible and so that subsequent administrations can evaluate whether the review process is working.

Element 3: Skills-based hiring infrastructure. Eliminating degree requirements is only the first step; replacing them with something better requires building skills-based hiring processes. The state should define, for each category of positions, the specific skills and competencies that are genuinely required, and should use assessments, work samples, and structured interviews to evaluate applicants against those criteria. A number of states have partnered with nonprofit organizations—Opportunity@Work, Jobs for the Future, and others—to develop skills-based hiring protocols, and those partnerships have produced usable templates that other states can adopt.

Element 4: Data collection and public reporting. A state that eliminates degree requirements should track the percentage of hires without four-year degrees, by agency and by position category, and publish the results annually. Transparent reporting creates accountability for the implementation and lets researchers evaluate which reforms produced the strongest effects.

The spillover to private employers

State government hiring is the most tractable target for credential inflation reform, but the larger payoff comes from the spillover to private employers. When a state governor announces that the state will no longer require four-year degrees for the majority of state positions, large private employers in the state often follow. IBM, Bank of America, Delta Air Lines, Walmart, Accenture, and several other large employers have announced similar policies in the years since the state-level reforms began, and at least some of the corporate changes were explicitly motivated by the state-level precedent. A state-level reform is the most effective private-sector reform, because it forces the change on the largest single employer in the state and generates social proof that other employers use to justify making the same change.

VIII. A Model State Legislative Package

The proposals in this paper can be enacted through a single state higher-education bill, adaptable to the particular circumstances of each state. The model bill should include:

1. An earnings-based funding formula for state public institutions of higher education, modeled on the Texas State Technical College Returned Value Formula, with the parameters described in Section III of this paper. The formula should phase in over four years, beginning with a 25 percent weight and rising to 100 percent by year four.
2. A Workforce Pell certification framework for the governor’s new authority under OBBBA Section 83002, specifying the six-element certification process described in Section IV. The framework should be codified in state statute rather than in executive order, so that it survives administration changes.
3. A tiered in-state tuition subsidy structure, keyed to program-level ROI as measured by the College Scorecard and supplemented by state-specific earnings data from the unemployment insurance wage records. The tiering should apply to both the in-state tuition differential at public institutions and to state grant and scholarship programs.
4. Authorization reform for new degree-granting institutions, including a 180-day maximum decision timeline, a reciprocity framework for institutions authorized in other states, a scaled review process for small institutions, and a statutory presumption in favor of non-traditional pedagogical models.
5. Credential inflation reform in state government hiring, including a statutory presumption against degree requirements, a rigorous exceptions process, skills-based hiring infrastructure, and annual public reporting.

Not every state will adopt every pillar. States that already operate earnings-based funding in some form—Texas, most notably—can focus on the other four. States that have already enacted credential inflation reform can build on that base by sharpening enforcement. States in the middle of their Workforce Pell implementation can start with that pillar and add the others in subsequent sessions.

The sequencing matters. We recommend that states begin with the Workforce Pell framework, because OBBBA Section 83002 makes the governor’s certification authority active as of July 1 and states that do not have a framework in place by that date will default to a loose or ad hoc approval process that will be hard to tighten later. The earnings-based funding formula comes next, because it generates the data infrastructure that the other pillars depend on. The ROI-based tuition subsidy, authorization reform, and credential inflation reform can follow in any order.

IX. A Concluding Argument

State higher-education policy is, in one sense, simpler than federal policy. States do not have to manage the Title IV loan programs. They do not have to regulate the accreditors. They do not have to oversee the data infrastructure that makes federal accountability possible. Their reform agenda is narrower and their legislative bodies are closer to their constituents.

In another sense, though, state higher-education policy is harder than federal. States are directly accountable to the institutions that dominate their political landscapes—the flagship universities, the state colleges, the community college systems—in ways that the federal government is not. A state legislator who votes to cut funding to a flagship university is voting against her constituents’ alma mater; a U.S. senator who votes the same way can point to the same vote as evidence that she is protecting the federal fisc. The political economy of state higher education has produced, in most states, an equilibrium in which funding continues to flow to institutions regardless of outcomes, tuition continues to rise, and accountability mechanisms are designed to be measurable-but-ignorable.

The TSTC Returned Value Formula is the exception that proves the rule. Texas enacted it only because TSTC’s legislative champions made the deliberate choice to subject themselves to an outcome-based test, believing—correctly—that TSTC would perform well under the test. The formula spread to no other Texas institution because no other Texas institution was confident enough in its own outcomes to volunteer for the same treatment. And no other state has adopted a similar formula because, in every state, the institutions that would be measured have had enough influence over the legislative process to prevent the measurement from happening.

The political case for state reform is therefore not a case to be made to the institutions. It is a case to be made to governors, to legislative leaders, and to the voters and taxpayers who ultimately fund the system. It is a case that sounds in three notes: the state is spending a great deal of money on higher education, the return it is getting is not what it could be, and the tools exist to do better. OBBBA has added a fourth note: the federal government has begun to take accountability seriously, and the states that do not follow will find themselves explaining to their voters why the federal standard is more demanding than the state standard for institutions those voters are funding.

The agenda in this paper is not radical. An earnings-based funding formula has been operating in Texas for over a decade. Workforce Pell programs will be operating in every state by the end of 2026. Credential inflation reform has been adopted in about half the states. Authorization reform has precedents in several states’ regulatory modernization initiatives. ROI-based tuition subsidies are the newest of the five pillars, but they are a natural extension of the College Scorecard data every state already consumes. None of the pillars requires a state to invent a new policy tool. All of them require a state to use existing tools with greater rigor and greater political will.

The payoff is not theoretical. The states that lead on these reforms will produce better labor-market outcomes for their residents, reduce their public colleges’ dependence on underperforming programs, expand access to higher education for workers without traditional credentials, and strengthen the fiscal position of their public treasuries. They will also, incidentally, make it easier for the federal government to tighten its own accountability rules, because state-level reform creates the political space for federal reform by demonstrating that the sky does not fall when outcomes are measured.

This is the moment for state action. The federal government has enacted the most significant higher-education reform in a generation. The states that respond with their own reform agendas will shape the next decade of American higher education. The states that do not respond will spend the next decade explaining why.