Will the New Loan Caps Actually Lower Tuition? What the Research Says

The argument behind the new federal borrowing caps is simple enough to fit in a sentence: colleges raise prices because the federal government keeps lending students more money, so cap the lending and prices come down.

The caps are real and they are in effect. Since July 1, 2026, graduate students can borrow $20,500 a year against a $100,000 aggregate limit, students in the eleven fields the Department of Education classifies as professional can borrow $50,000 a year against $200,000, and the lifetime federal cap across all levels is $257,500. We covered the mechanics in our full guide to the OBBBA student loan changes.

What is not simple is whether the theory behind them holds up. We went through the actual research. It is genuinely split, the disagreement is more interesting than either side's summary of it, and there is a problem with the whole literature that almost nobody mentions.

What Bennett actually said in 1987

The idea has a name and a date. It comes from a February 18, 1987 New York Times op-ed by Education Secretary William Bennett titled "Our Greedy Colleges," and the sentence everyone quotes is that increases in financial aid "have enabled colleges and universities blithely to raise their tuitions, confident that Federal loan subsidies would help cushion the increase."

The same op-ed contains a much more careful line that rarely travels with it: federal student aid policies "do not cause college price inflation, but there is little doubt that they help make it possible."

That distinction matters for reading everything below. Bennett claimed aid was an enabling condition, not an engine. The strong version people argue about now, where aid dollars convert to tuition dollars close to one for one, is a later invention.

The strongest evidence that aid raises prices

The most-cited recent finding comes from a study of Texas graduate programs by Sandra Black, Lesley Turner, and Jeffrey Denning, published as NBER Working Paper 31291. Looking at the 2006 introduction of Grad PLUS, they found that a $1 increase in federal loans produced a $1.10 increase in a program's list price and a $0.64 increase in net price, with institutional grant aid absorbing about a third of the sticker increase.

That is a large effect, and it is the single best piece of evidence for the case behind the caps. Four things about it are worth knowing before you lean on it.

It is graduate-only. It is Texas, drawn from state administrative records, and it excludes for-profit institutions entirely while capturing only a subset of private nonprofit programs. It is not a simple before-and-after but a comparison between programs where more students were already borrowing at the pre-2006 cap and programs where fewer were, so it measures relative movement rather than a market-wide jump. And as of now it is still a working paper under revision at a journal, not published research.

On the undergraduate side, the strongest published finding is from David Lucca, Taylor Nadauld, and Karen Shen in the Review of Financial Studies (2019), who studied the 2007 and 2008 increases in federal loan limits and estimated pass-through to tuition of about 60 cents on the dollar for subsidized loans and about 20 cents for unsubsidized ones, concentrated in more expensive programs and at for-profit and two-year institutions.

The strongest evidence against

Robert Kelchen has tested the question directly in the places where you would most expect to find an effect, and did not find much.

In Research in Higher Education (2020), using program-level data from 2001 to 2016, he reported "little consistent evidence" for the Bennett hypothesis in either medical or business schools. In Economics of Education Review (2019), studying 85 public and 114 private nonprofit law schools from 2001 to 2015, he found "rather modest relationships."

Those are the exact phrases, and they are worth keeping intact. They are weaker than "no effect," but they are a long way from dollar-for-dollar pass-through in three of the most expensive graduate fields in the country.

There is also a finding that scrambles the usual story about who captures aid. Lesley Turner's work on Pell Grants estimated that for-profit institutions capture about 9 cents of every Pell dollar while selective nonprofit institutions capture about 93 cents. That is close to the reverse of the common assumption that aid capture is a for-profit problem. It measures grant aid rather than loans, so it does not contradict the loan studies, but it does mean the tidy version, where the Bennett hypothesis only holds at for-profit colleges, is not supportable.

The for-profit evidence that does hold up is narrower than it sounds. Stephanie Cellini and Claudia Goldin found in the American Economic Journal: Economic Policy (2014) that aid-eligible institutions charge about 78% more than comparable programs at institutions outside the federal aid system. That is a real and large effect, and it is specifically about sub-baccalaureate certificate programs, not bachelor's degrees.

The fact that complicates both stories

Here is something we did not expect to find, and it cuts against the simplest version of the pro-cap argument.

If Grad PLUS unleashed graduate tuition in 2006, you would expect graduate tuition growth to accelerate afterward. It did the opposite. Using the NCES Digest of Education Statistics series on graduate tuition, average annual growth across all institutions ran about 6.6% in the five years before Grad PLUS and about 5.2% in the five years after. The slowdown holds in public and private nonprofit institutions separately, and it holds after adjusting for inflation.

This does not refute the Texas study. That study was carefully designed to find relative differences between more- and less-exposed programs precisely because the aggregate series is noisy and driven by many things at once. But it does mean that anyone pointing at the post-2006 climb in graduate tuition as proof that Grad PLUS caused it is misreading the data. Tuition was climbing at least as fast before.

The problem with the entire literature

Now the part that should change how much weight you put on any of this.

Every study above measures what happened when federal aid went up. OBBBA cuts it. We could not find a single study estimating what happens to prices when federal loan availability is reduced.

The two closest real-world experiments both looked at something else. When Parent PLUS credit standards were tightened in November 2011, researchers studied the effect on enrollment at historically Black colleges, where PLUS borrowing fell sharply and was not replaced. When for-profit colleges lost federal aid eligibility through sanctions, researchers studied where the students went. Both measured enrollment. Neither measured price.

So the case that caps will lower tuition rests on assuming the mechanism runs symmetrically in reverse. That is a theoretical extrapolation, not a finding, and there is a specific reason to doubt it: prices are generally stickier going down than going up. Institutions facing a revenue squeeze can cut costs, shrink programs, admit different students, or raise institutional aid instead of cutting the published price.

There is one international precedent worth sitting with. When England tripled its tuition fee cap to 9,000 pounds in 2012, the government expected most universities to charge around 6,000 and reserve the maximum for exceptional cases. Nearly all of them went to or near the cap. Caps do not just constrain prices. They can also become the number everyone converges on.

What the economists themselves predict

The people who study this are far more hedged in their own words than the debate around them suggests.

Kelchen expects at most a small decrease in tuition, on the theory that students may become somewhat more price-sensitive. Sandy Baum, who spent years co-authoring the College Board's pricing reports, reads the literature as mostly not supporting the hypothesis outside for-profit institutions, and put the likely outcome as prices not plummeting but perhaps rising more slowly. Denning, whose own study produced the strongest pro-cap number in this article, said it is possible but that he does not have a crystal ball.

The most confident claim any of them made was about something else entirely. Dominique Baker's point is that the evidence on reducing financial aid is actually robust, and what it shows is that students stop enrolling. That is the effect with the strongest research behind it, and it is not a price effect.

Two of the names quoted most in coverage of this question, Baum and Phillip Levine, are offering expert readings of the literature rather than their own findings on it. That is worth knowing when a headline presents "economists disagree" as a clash of studies. Some of it is a clash of interpretations.

What tuition is actually doing right now

Against all of that, here is the baseline the caps are supposed to bend, as of the most recent College Board Trends in College Pricing report, published November 2025 and covering the 2025-26 year.

Published tuition and fees ran $11,950 at public four-year institutions for in-state students, $31,880 for out-of-state students, $45,000 at private nonprofit four-year institutions, and $4,150 at public two-year institutions. In current dollars those were increases of 2.9%, 3.4%, 4.0%, and 2.7%.

Adjusted for inflation, they barely moved: 0.3% for in-state public four-year, 1.4% for private nonprofit, and flat for public two-year. Over the past decade, real published prices have actually fallen at public institutions, down about 7% in-state at four-year schools and about 10% at two-year schools.

Net price, what students actually pay after grant aid, has fallen much further. Estimated net tuition and fees for in-state students at public four-year schools is about $2,300 in 2025 dollars, down from a peak near $4,450 in 2012-13. At private nonprofits it is about $16,910, down from roughly $19,810 in 2006-07. Those are College Board projections rather than observed figures for the most recent years, and they cover first-time full-time students only, but the direction has been consistent for over a decade.

The picture that emerges is not a sector raising real prices unchecked. It is one where sticker prices have flattened and discounting has grown.

The other force in the room

There is a reason to expect pricing pressure over the next fifteen years that has nothing to do with loan policy.

According to WICHE's Knocking at the College Door projections, the number of US high school graduates peaked with the class of 2025 at about 3.86 million and is projected to fall to about 3.37 million by 2041, a decline of roughly 12.5%. The decline is concentrated: the West is projected down about 20% and the Northeast about 17%, while the South is roughly flat or slightly up.

If tuition growth does slow over the next decade, disentangling the loan caps from a shrinking pool of applicants will be genuinely difficult. Both push the same direction. Our view is that anyone who confidently attributes a future slowdown to OBBBA alone will be guessing.

What this means for your decision this fall

None of this resolves into a clean answer, so here is the practical read.

Do not choose a school on the assumption that its price will fall. Nothing in the research supports planning around that, and the economists closest to the question are forecasting a small effect at most.

Do take the caps seriously as a financing constraint rather than a pricing forecast. That part is not speculative. If a graduate or professional program costs more than the new annual and aggregate limits allow you to borrow, the gap is real starting now, and it gets filled with private loans, family money, or institutional aid. Our graduate loan cap budget plan walks through that arithmetic, and the Parent PLUS cap analysis covers the undergraduate side.

Watch the professional-field classification if you are heading into health care. Nursing, physician assistant, social work, physical therapy, and occupational therapy are not among the eleven fields treated as professional, so they fall under the lower graduate limits despite the cost of those programs. That classification is being contested in court, and it is the single provision most likely to change.

And compare on net price, not sticker. It is where the actual movement in college pricing has been for a decade, it varies far more between schools than published tuition does, and it is the number your family will really pay. You can run it for specific schools with our ROI calculator and put two options side by side with the comparison tool.

Sources: William Bennett, "Our Greedy Colleges," New York Times, February 18, 1987. Black, Turner, and Denning, NBER Working Paper 31291. Lucca, Nadauld, and Shen, Review of Financial Studies 32(2), 2019. Kelchen, Research in Higher Education 61(3), 2020, and Economics of Education Review 73, 2019. Cellini and Goldin, American Economic Journal: Economic Policy 6(4), 2014. Turner, "The Economic Incidence of Federal Student Grant Aid," 2017. NCES Digest of Education Statistics Table 330.50. College Board, Trends in College Pricing and Student Aid 2025 (published November 2025). WICHE, Knocking at the College Door, 11th edition, December 2024. Economist quotes as reported by NPR, June 28, 2026. Borrowing limits effective July 1, 2026; pricing figures as of the 2025-26 academic year.