蝌蚪直播

House Republicans鈥� Proposed Overhaul of Student Aid Would Reshape Federal Higher Education Policy

Clare McCann, Jordan Matsudaira, and Tia Caldwell
May 2025

Earlier this month, House Republicans introduced a that, if enacted, would dramatically the federal government鈥檚 role in helping Americans pay for college. It increases annual student loan borrowing limits for undergraduates by as much as five-fold, but leaves their aggregate limits largely unchanged; reduces borrowing limits for graduate loans; and increases student loan repayment for the lowest-income borrowers. It sets out a complex risk-sharing accountability system to require institutions to pay a fraction of their students鈥� unpaid loan balances, so that colleges bear some of the costs of loans from programs that leave students without good-paying jobs. And it makes significant changes to the Pell Grant program, including allowing very short-term training programs to access Pell Grants for the first time.

The Congressional Budget Office the bill would net more than $350 billion in savings over the next decade (mostly from repealing a Biden-era loan repayment plan currently enjoined by the courts), and the House quickly moved it forward on a party-line vote in the education committee. The bill is just one component of a larger effort to pass legislation through reconciliation 鈥� a process that enables the Senate to pass legislation with a simple majority vote, often used when one party controls the House, Senate, and the White House 鈥� and its fate is tied up in the larger politics of developing and voting on that package.听

Given the big savings and quick vote to approve the bill, though, it seems likely that this legislation will help shape what House and Senate Republicans ultimately adopt. The roots of the House bill are in the , introduced last year by then-chair of the House education committee Virginia Foxx (R-NC). But this new bill 鈥� prepared by new education chair Rep. Tim Walberg (R-MI) and the Student Success and Taxpayer Savings Plan 鈥� includes some additional twists and turns.听

This piece provides an overview of major policies included in the bill, and flags some of the most concerning provisions. Stay tuned for deeper analyses from the PEER Center in the coming weeks.

Changes to Loan Repayment Aim to Reduce Borrowers鈥� Balances, but Weaken Protections for the Lowest-Income Borrowers
The bill would repeal the SAVE plan, a generous income-driven repayment (IDR) plan issued by the Biden Administration and now held up in the courts, and sunset the rest of the current suite of existing IDR plans. The 13.1 million borrowers who are on or have applied to any current IDR plan would be transitioned to a modified version of the Income-Based Repayment (IBR) plan for pre-2014 borrowers. Effectively all current borrowers would have access to the modified plan, on which they would repay 15 percent of their discretionary income for 20 years (if they have only undergraduate debt) or 25 years (if they have any graduate debt). This would spike many borrowers鈥� monthly payments (making it a likely target for future litigation), because under existing plans, most only repay 10 percent of their discretionary income. Under the proposal, current borrowers would also retain access to other, fixed repayment plans, like the standard 10-year, graduated, and extended repayment plans.

For new borrowers (those who take out a loan starting in July 2026), the options under the House鈥檚 proposal would look very different. Upon entering repayment, new borrowers would have the choice of just two plans: a modified standard plan or a single income-driven repayment plan. The modified standard plan resembles the current 鈥淓xtended Payment鈥� loan plan for borrowers with higher debt loads, and would set a payment term based on how much debt the borrower has: 10 years if their principal balance is less than $25,000; 15 years if it is between $25,000 and $50,000; 20 years for up to $100,000; and 25 years for $100,000 or more.听

Relative to the current 10-year standard plan, this would lower monthly payments for high-debt borrowers, and may make the standard plan more attractive relative to income-driven plans. There are cliff effects around the edges, though: As Figure 1 shows, as debt increases from just below to just above the thresholds associated with a longer payment term, annual payments would drop substantially while the lifetime payments on the loan would increase. For example, a borrower with just over $25,000 in debt would pay about $760 less per year than a borrower with just shy of $25,000, but since he would pay for 5 additional years, the added interest payments would cost him about $5,400 more over the life of the loan. These are substantial consequences for very minor differences in borrowing to embed in the default repayment plan. The risks for borrowers of falling on either side of a threshold could be mitigated by smoothing out these cliffs, extending the term of the loan more gradually with increases of debt.

For new borrowers who want to repay on an income-driven plan, the House bill establishes a single IDR plan called the Repayment Assistance Plan (RAP). Under this plan, borrowers would repay a different percentage of their income depending on their total adjusted gross income (not, as under most current plans, their income above 150 percent of the federal poverty level): 1% for an AGI of $10,000-$20,000; 2% for an AGI of $20,000-$30,000; and so on, up to 10% for an AGI of $100,000. Additionally, borrowers would see a $50 per month reduction in their payments for each dependent child. Borrowers would always owe at least $10 per month, even if their AGI is less than $10,000 or their dependent child payment reductions would lower their payment to zero. This minimum payment would be a notable divergence from the current IDR plans, which allow the lowest-income borrowers to repay $0 per month while earning progress toward forgiveness.听

Under RAP, borrowers would repay until they have paid their balances down to $0, or have made the equivalent of 30 years鈥� worth of payments. To help them pay down, borrowers whose payments are too low to cover all of the interest that accrues in a given month would see that excess interest wiped clean, preventing loan balances from growing. Another subsidy would further ensure a borrower鈥檚 principal balance declines each month, with a pay-down matching the borrower鈥檚 scheduled payments of up to $50. A borrower on the standard plan could always switch into RAP; but once a borrower switches into RAP, they could not switch back to the modified standard plan.

In general, the monthly payments owed under this plan would track relatively closely with current plans like the Revised Pay As You Earn (REPAYE) plan implemented during the Obama Administration for those with income up to about $70,000 in income, as shown in Figure 2. The plan requires, by design, higher payments for higher-income borrowers relative to existing IDR plans. For example, for a borrower with $100,000 in income and no dependents, annual payments would be $10,000, compared with $7,652 under REPAYE. (For a borrower with $100,000 in income and two dependents, the figures are $8,800 per year on RAP, compared with $6,002 on REPAYE.)

Broadly, the structure of the new RAP plan has some appeal. In current income-driven repayment plans, a key parameter to protect lower-income borrowers is the income protection threshold used to calculate discretionary income (e.g., 150 percent of the federal poverty line, as is used in REPAYE). Raising that threshold, as the Biden Administration did under the SAVE plan (to 225 percent of the poverty line) lowers payments to $0 for lower-income borrowers with income just above the previous threshold. But it also has the less-desirable feature of reducing payments for all borrowers, even at the highest income levels (and leaves borrowers below 150 percent of the poverty line, whose payments were already $0, unaffected). The structure of the RAP plan breaks that link, which potentially allows for finer targeting of the generosity of the plan to lower-income borrowers. As Figure 1 shows, though, the bill uses that flexibility to increase payments for higher-income borrowers, keeping nearly the same (or higher) payment burdens as REPAYE for lower-income borrowers.听

There are some other pitfalls and quirks of this plan. Most significantly, the lowest-income borrowers would pay more under RAP than they would under REPAYE, and for longer (as long as 30 years, compared to 20 for undergraduates receiving forgiveness through REPAYE). The $10 minimum payment is likely to be unaffordable for someone in such financial distress that they are making less than $10,000 a year 鈥� well below the federal poverty level of $15,650. It is probable that even this low payment amount would cause many of those borrowers to default on their loans. Additionally, the thresholds determining what share of income borrowers must pay would not be adjusted for inflation, which would make the plan less and less generous over time. And as with the modified standard plan, there are cliff effects in RAP, though they are smaller; borrowers on either side of the income brackets would see a small jump in payment rates (from 1 percent, to 2 percent, to 3 percent, up to 10 percent) with a single extra dollar of income.听

New Loan Limits Would Increase Borrowing for Undergraduates and Reduce It for Graduate Students
If the proposed changes to loan repayment seem significant, the changes to borrowing limits would be even more so. Under the plan, both annual and overall (aggregate) student loan limits would change considerably for the first time in over a decade. The bill would measure the median cost of college for a given program of study across the country 鈥� for instance, the median cost of attendance across every institution offering a bachelor鈥檚 degree in biology, or a master鈥檚 degree in business. That median (or, for undergraduate students, that median minus the student鈥檚 Pell Grant award) would serve as a student鈥檚 annual loan limit.

The scale of the change is hard to predict; program-level cost of attendance data do not exist for all programs (though the Financial Value Transparency regulations would gather the information). However, from the 2020 National Postsecondary Student Aid Study suggest the typical cost of attendance in a year for a certificate is nearly $12,000; for an associate degree, about $11,000; and for a bachelor鈥檚 degree, more than twice that at $26,000. In comparison, under a first-year dependent student can borrow just $5,500, and a first-year independent student $9,500. (The House proposal does away with that distinction, so all undergraduates would qualify for loans up to the median costs in their field.) That means the typical undergraduate would see their annual loan eligibility increase dramatically 鈥� potentially leading students to take on far more debt.

These high annual loan limits could pose a problem when paired with the proposed aggregate loan limits of $50,000 for undergraduates (higher than the current $31,000 for dependent students and a bit lower than the current $57,500 for independent students). Students maxing out their federal debt loads each year for a bachelor鈥檚 degree program could run out of their loan eligibility in their second or third year of school, forcing them to drop out or to seek financing from the private loan market. While schools would have the ability to set lower loan limits for specific programs of study, many would not. And some students may feel they have no choice but to use up all of their loan eligibility in their first years of college, as loans may be the only way they could afford their tuition bills. Parent PLUS loans would also be newly capped, set to a lifetime limit of $50,000 per parent borrower (across all college-going children, not per child), meaning Parent PLUS loans may no longer be able to fill the financing gap left when undergraduates exhaust their loan eligibility. While current college dropouts typically have small debt loads (because dependent students are not permitted to even accumulate more than $12,000 across their first two years), that may no longer be the case under the newly proposed limits.

For graduate students, the new annual limits would again be set at the median cost of college for the program of study. That the typical master鈥檚 degree program at around $24,000 in borrowing for a single year; the typical law degree at about $58,000 per year; and the typical M.D. program at about $73,000. In addition, new aggregate limits would be set at $100,000 for graduate students and $150,000 for professional students. Together, these limits would reduce graduate students鈥� loan eligibility 鈥� largely because the bill eliminates the Grad PLUS program, which enables students to borrow up to the full cost of attendance. Analysis we published last month shows that, while the typical master鈥檚 degree student in many fields borrows below the annual levels the bill proposes already, that is not the case for all programs 鈥� especially health-related fields, which collectively account for almost half of student loans borrowed by graduate students each year. Oddly, some of the fields that would see their borrowing reduced most by these proposed limits 鈥� like medicine, dentistry, and law 鈥� are also the fields where graduates go on to earn the most, leaving them well-situated to repay their debt.

Risk-Sharing Framework Would Create a Convoluted Set of Penalties for Colleges, but Could Leave Programs with the Worst Outcomes Largely Unscathed
Both loan repayment and loan limits have implications for another cornerstone of the House proposal: risk-sharing. The risk-sharing framework in the reconciliation bill is a complicated scheme that includes multiple components and tests that separately address different types of students and programs. Briefly, here鈥檚 how it would work:

• The system would hinge primarily around unpaid loans: For all the loans made to students in cohorts completing a particular program in a particular year, the framework defines a 鈥榥on-repayment balance,鈥� equal to the sum of annual payments that borrowers owed and didn鈥檛 make, plus the total amount of forgiven interest and principal on their loans (including for subsidies on monthly payments and 30-year forgiveness under the RAP, as described earlier in this piece).
• Schools would be required to repay the Department a percentage of this amount, which varies based on the total tuition required to complete the program relative to the earnings of recent program graduates. The percentage is set to 0 for programs that are low-cost relative to their earnings, and 100 for programs that are high-cost relative to their earnings. The percentage is fixed for the cohort, and the school would pay that fraction of the unpaid loans each year until the entire cohort鈥檚 loans are gone 鈥� repaid, or forgiven. The calculation is repeated for each subsequent cohort, so that schools would ultimately be paying penalties across many cohorts each year.听
• All undergraduate students who don't graduate would be assigned to a single, institutional cohort of non-completers. Graduate students, who generally are admitted to a particular program, would be assigned to program-level cohorts of non-completers. The school would pay on their non-repayment balances (both at the undergraduate and graduate levels) at a rate equivalent to the school鈥檚 dropout rate (e.g., if 52% of undergraduate students don鈥檛 graduate, the school pays a fine of 52% of the undergraduate non-completer cohort鈥檚 unpaid loans).

The motivation behind this scheme is understandable: Tests are designed to penalize colleges whose students fail to repay their loans while (relatively) advantaging colleges with positive outcomes across multiple dimensions: low costs, high returns (in the form of graduates鈥� earnings), high student graduation rates, and high rates of loan repayment. That said, the plan鈥檚 complexity could undermine its efficacy as institutions consider how to respond. Some of the elements of the design seem redundant, or add only new complications, while other aspects are difficult to rationalize.听

For example, one of the more baffling aspects of the scheme is that the calculation of the non-repayment balances each year would ignore loans that have gone into default. A borrower placed in the standard plan who never made payments and defaulted on their loan would generate penalties for the first nine months of payments missed -- but thereafter, any missed loan payments would not add to a school鈥檚 penalties. In contrast, a borrower who enrolled in IDR and made payments but in amounts lower than their scheduled payments, or who made full monthly payments that qualified the borrower for an interest or principal subsidy under RAP, would generate penalties for the institution every year, despite likely generating far lower costs to the government. In fact, even RAP borrowers whose incomes start low but who ultimately earn enough to fully repay their loans rather than receiving forgiveness at 30 years would generate risk-sharing penalties for the college in the years when their payments were too low to cover interest. That is perhaps unfair to institutions, since many borrowers who elect an income-driven plan 鈥� even those who will go on to repay their total balances 鈥� would qualify for the interest subsidy in the early years of repayment as a function of lower early-career earnings and interest accrued while in school.听

Other parts of the risk-sharing formula seem redundant, or add complexity with little additional benefit. For example, the amount that an institution would pay on the non-repayment balance is a function of the completion rate of the institution: Penalties would be repaid based on a rate equivalent to the school鈥檚 dropout rate. But the dropout rate of the institution would also determine the size of the non-repayment balance itself -- the more students dropping out with debt, the higher their collective balance will be. That means, in effect, the completion rate would be factored into that piece of the formula twice.

Stepping back, a major concern with this proposal is that, despite the potential enormity of this new system of penalties for colleges, it is nearly impossible to predict its impact. It鈥檚 impossible to know how much schools would be expected to repay, for instance, because many of the key data points 鈥� the amount of unpaid loans (as defined in the proposal) and the price of programs, in particular 鈥� do not currently exist. Estimating the effects of the scheme is made all the more difficult given the wide-ranging changes elsewhere in the bill 鈥� particularly changes to loan limits (both annual and aggregate), which would affect the scale of the unpaid loans on which schools need to repay; and changes to repayment, which would change how much borrowers are required to pay and lead some borrowers into delinquency or default on their loans.

Eligibility Cuts Would Address the Pell Grant Funding Shortfall, but Extend Grants to Very Short-Term Programs
The bill would put some of the savings it generates back into the Pell Grant program 鈥� not a moment too soon, because the program is currently running a that, , will require a $17 billion infusion before the end of fiscal year 2027. However, rather than simply providing enough funding to completely close the shortfall, the bill would also cut students鈥� eligibility, mainly by increasing the required number of credit hours to qualify as full-time from 24 per year (12 per semester) to 30 per year (15 per semester). For students who could move up to 15 credit hours, the change could help make sure they earn enough credits to graduate on time. For those who couldn鈥檛, it would be tantamount to a cut in their awards, which would be prorated down. The bill would also eliminate Pell eligibility for students attending college at less than half-time intensity, even while the new requirements would increase the number of credits needed to reach half-time status. These changes could put many more students in danger of losing eligibility.听

Despite the goal of saving money and putting the Pell program on more stable financial footing, the bill includes a likely-expensive plan to extend Pell Grants to programs as short as 8 weeks, despite that many very-short-term programs have little benefit in the labor market. The dollars could go to noncredit programs as well as credit-bearing ones, a significant change from current federal financial aid rules. The only real eligibility criteria based on the program鈥檚 performance would be to require that the median earnings of program graduates minus 1.5 times the poverty line (just $23,000) is greater than the program鈥檚 total price -- a threshold that is so low that it would permit programs to charge high tuition for short programs despite mediocre earnings. And even more concerning, the bill would extend federal dollars to entities that can鈥檛 (or won鈥檛) earn accreditation, allowing unaccredited training providers to skirt existing federal rules on receiving federal financial aid dollars. The expanded eligibility, which comes just as the Pell program faces considerable financial challenges, will likely serve as an invitation for colleges 鈥� including those otherwise outside the federal financial aid system 鈥� to start up low-value programs with limited benefits for students, potentially diverting those students from higher-earning programs or wasting taxpayer dollars. There鈥檚 no word yet on the costs of this provision, but it would undoubtedly exacerbate the current funding gaps.

Other Provisions
The scope of proposals in the bill is expansive, and we don鈥檛 detail them all here. For instance, the bill would establish a separate program in the bill, called PROMISE Grants, that would redistribute the risk-sharing fines that schools pay to institutions that are serving low-income students well. The legislation would also make relatively modest changes to student loan deferments and forbearances and Public Service Loan Forgiveness; eliminate critical in current law; and limit the Education Department鈥檚 authority to issue new loan regulations that would cost the government.听

We encourage you to read the full bill (or at least the Committee鈥檚 summary of the bill ), and reach out to the PEER Center if there are particular pieces you鈥檇 like to read about in future analysis. In the meantime, stay tuned for additional deep dives into some of these issues.

听