fb-pixel Skip to main content

If a college economics class does its job, students will soon realize that even their professors don’t understand why their schools are so expensive. Over the past three decades, college tuition has increased at more than double the rate of inflation. Outstanding student loan debt in the United States now exceeds $1 trillion, a national burden even greater than that of credit cards.

Yet no one agrees on what makes college tuitions so high. Plush state-of-the-art gyms? Classroom technology? The cost of health care, the glut of administrators, too many professors focusing on research at the expense of teaching?

Then there is another theory, one that for 25 years has remained as controversial as it is counterintuitive: that the culprit is federal financial aid. Schools know that students have access to tens of billions of dollars in grants and loans, the thinking goes, and they raise tuition because the aid lets them do it.

This idea — essentially, that federal aid enables college administrators to get greedy — is known as the Bennett Hypothesis, after William Bennett, the conservative thinker who was President Reagan’s secretary of education. Since he first floated it in the 1980s, the Bennett Hypothesis has been debated in economic journals, congressional reports, and popular books. Studies have affirmed it, affirmed it in part, refuted it entirely, kind of refuted it, and many gradations in between.


It is, to say the least, a tough thing to test. And its unruliness stems in part from the colleges themselves. All traditional colleges accept federal aid, and they spend their money in idiosyncratic ways. Is a college public or private? Has its home state just slashed higher-ed funding? How does it fare in the U.S News & World Report rankings? Are there other competing universities nearby? These differences tend to shield the whole issue in opacity.


But a recent study has looked at the effects of financial aid in an unexpected place: for-profit colleges. The for-profit sector — national chains like The University of Phoenix as well as smaller technical institutes — has grown tremendously in recent years, and it offers new and little-examined data on the price of an education. For an economist, the beauty of the for-profit sector is that such schools can choose to either receive federal aid or not, allowing for a cleaner comparison than with traditional schools.

The recent paper, by economists at Harvard and George Washington University, compared more than 2,650 programs within for-profit schools in three states over multiple years, and found that the schools receiving federal grants and loans set their tuition roughly 75 percent higher than those institutions that go without government support. This discrepancy in costs has everything to do, the authors write in the study, with the aid the schools receive — “lending credence to the ‘Bennett hypothesis’ that aid-eligible institutions raise tuition to maximize aid.”

The National Bureau of Economic Research published the paper in February, and since then the argument over the Bennett Hypothesis has returned to the front of the debate over college costs. Conservatives crow that the study proves what they’ve known all along. Even liberals admit the hypothesis seems legitimate. Six days after NBER unveiled the paper, Vice President Joe Biden told an audience at Florida State University that “government subsidies have impacted upon rising tuition costs.”

The authors of the study caution they researched the for-profit sector only, and haven’t tried to test the hypothesis among traditional four-year colleges. Even if their finding holds and is applicable to conventional schools, it still leaves a very tough question: What to do next? Conservatives have already proposed cutting aid as a weapon against tuition increases. But it’s unknown whether that would stall the escalating cost of college or just limit the number of low- and middle-class students who might attend. If college is already too expensive, and aid only exacerbates the problem, it raises the unsettling prospect that making college cheaper might actually exclude some poorer students.



COLLEGE TUITION WASN’T always a national crisis. As recently as the 1970s, the price of tuition actually declined by 17 percent at public universities, and 13 percent at private ones, according to data from the American Council on Education, a leading higher-ed lobby.

By the 1980s, however, things had changed. The price of tuition started to climb: By the end of the decade, it was up 47 percent at public universities and 54 percent at private schools, according to the council. What was different? Student populations were changing; college degrees were becoming more important. And Congress had mandated more federal assistance. In 1978, President Carter inaugurated the modern era of financial aid by signing the Middle Income Student Assistance Act, which had the effect of offering federally subsidized student loans to anyone who qualified for college, regardless of income. During the Carter and Reagan administrations, the government further expanded federal aid, including Pell grants and Perkins loans for needy students. From 1978 to 1981, total available aid grew by 70 percent, to a total of $14.7 billion, according to the Congressional Budget Office.


As Bennett tells the story today, he formed his theory while serving as Reagan’s secretary of education from 1985 to 1988, touring roughly 60 universities and meeting with provosts and presidents. For a Republican like Bennett, that meant asking them why college tuition was so high. Why was it growing two to three times the rate of inflation? Bennett says university officials told him the answer had to do with the federal assistance the schools received. When parents or students balked at higher prices, college administrators could tell them not to worry; federal aid meant the true cost of college would be much lower. The argument had a cyclical perversity to it: As tuition climbed, so did the number of students who relied on aid to offset it. The very loans meant to help students afford college were making college more unaffordable.

Bennett took to The New York Times’ opinion page, and on Feb. 18, 1987 wrote: “[I]ncreases in financial aid in recent years have enabled colleges and universities blithely to raise their tuitions, confident that Federal loan subsidies would help cushion the increase.” Though federal aid policies may not directly cause tuition to rise, Bennett argued that “there is little doubt that they help make it possible.”

The research done on the hypothesis since then is enough to induce whiplash. In their 1998 book, “The Student Aid Game,” economists Michael S. McPherson and Morton Owen Schapiro said public colleges and universities tended to increase tuition by $50 for every $100 in aid. In 2001, however, in a report for Congress, the National Center for Education Statistics found no evidence to support the Bennett Hypothesis anywhere. A 2003 study by Cornell University economists Michael Rizzo and Ronald Ehrenberg didn’t find evidence among public schools, but a 2007 study from University of Oregon economists Larry Singell and Joe Stone found a nearly dollar-for-dollar match at private institutions.


The professor whose work sparked the latest brouhaha didn’t set out to start one at all. She wasn’t even looking at the Bennett Hypothesis. Stephanie Riegg Cellini was an assistant professor of public policy and public admninistration, and economics at George Washington University and interested in the for-profit education sector, an industry that other academics seemed to be ignoring. She found that even the federal government had no idea how many for-profit schools there were. In California, the government estimated there to be 500 such colleges, but these were only the schools that received federal aid for their students. Cellini counted roughly 2,000 for-profit colleges in the state, most of which did not receive federal aid and were not accountable to anyone but the board that licensed them. (Some schools choose to forgo aid for various reasons, one of them being the accreditation that accompanies it is too expensive.)

Cellini presented her findings at a lecture at Harvard in December 2009. Afterward she met with Claudia Goldin, an economics professor there, who recommended they pair up. “I’m a data digger,” Goldin says. The data from certain states were rich with details. And because some of those colleges offered aid and some didn’t, one thing Goldin and Cellini could test was the Bennett Hypothesis.

They focused on colleges in Wisconsin, Michigan, and Florida, three states in which they could see comparable data on tuition and programs offered. The authors discovered that the schools that received aid charged roughly 75 percent more in tuition than the schools that didn’t.

The economists tried to control for quality. Maybe the aid-receiving schools graduated more students or offered more and better classes. They didn’t. They charged higher tuition, it appeared, because they got aid. “Sound economic theory suggests that if you can get more money to raise tuition, you will,” Cellini says. The National Bureau of Economic Research published their paper last month.


THEIR WORK CAME at a timely moment. The National Association of Consumer Bankruptcy Attorneys issued a statement in February saying they feared student loans will be the next “debt bomb” to ravage the country. The new education plan President Obama outlined in January uses federal aid as a stick, proposing to tie how much schools receive to how low they can keep tuition. And Vice President Biden, out promoting the plan at Florida State, said something must be done because the Bennett Hypothesis seemed to be real: “It’s a conundrum,” he added. (A White House spokesman, Matt Lehrich, tells The Boston Globe: “There is no evidence to suggest that federal aid is a driver of tuition increases, but it’s absolutely true that the formula we’re using to distribute campus-based aid right now has not created the right incentives to bring down costs and promote affordability in higher education.”)

Days after the study was made public, and partly in response to it, Andrew Gillen, an economist at the free-market Center for College Affordability and Productivity, issued a 32-page report that he called “Bennett Hypothesis 2.0,” arguing that to prevent further tuition increases, it was time to start limiting federal student aid to only those who need it, and only at the dollar amounts they need. Richard Vedder, an economist at Ohio University and scholar at the conservative American Enterprise Institute, took Gillen’s proposal a step further, arguing on the higher-ed website Minding the Forum, “The federal government needs to wind down its financial aid commitment. Restrict eligibility for aid to truly low-income students. Impose performance criteria for aid recipients: mediocre students will lose aid. Make the college absorb some of the risk for loan defaults — a lesson we should have learned from the financial crisis.”

But it’s unknown whether cutting aid would actually work, in part because it’s hard to draw a clear causal link between aid and tuition, especially among the not-for-profit colleges that make up most of American higher education. Goldin and Cellini both declined to comment on the applicability of their paper within the traditional, four-year sector.

Higher-education advocates tend to point to state funding as a more important driver of tuition: As states cut support, students need to pay more; when universities have more support, they’ve sometimes actually lowered tuition. In California, for instance, tuition declined in eight out of the 10 years in the 1990s, says Terry Hartle, senior vice president of the American Council on Education, because the dot-com and real estate booms helped the state Legislature fund colleges at adequate levels, easing pressure on the state’s large public higher-ed systems. On the other hand, nationwide college prices climbed from 1984 to 2009 regardless of whether state funding levels were down or up, according to a report from the State Higher Education Executive Officers. Moreover, as Neal McCluskey at the free market Cato Institute pointed out in a recent blog post, cuts in state funding don’t explain “constantly increasing private school costs.”

Because the data contradict themselves, it’s always more politically expedient for members of Congress and a sitting president, regardless of anyone’s political leanings, to expand federal aid. That way no qualified student is denied access to college. To put Bennett’s idea to the test — to cut large swaths of federal funding to hopeful college students — takes a courageous politician, who risks the wrath of not only students and their parents but American universities and their leaders.

This is why federal aid keeps growing. Last year, for the first time in history, students took out $100 billion in loans — money that will end up as revenue for colleges and as debt for students. President Obama wants to lower these costs of tuition by tying the flow of aid to the value schools offer. Gillen thinks one way to do that is to make more college graduates take an exit exam that quantifies what they learned, and that future students could use to decide if they were getting their money’s worth. “It’s not as hard as some people think to begin to reform the system,” Gillen says. But given the huge amounts of money involved, he points out — for the students, the government, and the schools at the heart of it — it won’t be easy either.

Paul Kix is a general editor at ESPN: The Magazine.