Bennett, Biden, and Boondoggles

Richard Vedder

Joe Biden admitted at a talk at Florida State University the other day that, yes, federal student-financial-assistance programs probably did contribute to rising college costs. That was newsworthy, and I suspect a statement that was not vetted by the White House (not the first Biden rhetorical faux pas from the Administration’s perspective.)

The Biden statement flies in the face of the generally perceived wisdom among supporters of the federal student-loan and grant programs, namely that these programs really have little impact on college costs. The Biden admission, however, implicitly agrees with those who argue the gains from these programs to college students are significantly dissipated by the fee-raising behavior of the schools enrolling the affected students.

Biden, in effect was embracing an idea first vocally proclaimed by Bill Bennett, then Secretary of Education, in 1987. If Biden and Bennett are right, why continue to expand these programs if their effectiveness is significantly reduced, maybe even eliminated, by tuition-increasing behavior?

My colleague Andrew Gillen has completed the best single analysis I have read on federal student financial aid and what he calls the Bennett Hypothesis. The study will be released today by the Center for College Affordability and Productivity (CCAP), and brilliantly explains a paradox: Some scholars claim the hypothesis is not true (that the tuition effects of federal-aid programs are minimal) while others claim pretty big effects—these programs raise tuition fees perhaps 50 cents or even more for each one dollar of aid per student dispensed.

I don’t want to steal Andrew’s thunder, but the key to his analysis comes from the fact that there is a lot of complexity to student-aid programs and college behavior. Pell Grants, for example, have a different impact than so-called unsubsidized federal student loans (which, in fact, have an implicit federal subsidy associated with them). Harvard behaves differently than Slippery Rock State University. Some universities face tuition caps. Most importantly, the long-run effects of college almost certainly are different than the short-run effect, and what college A does impacts on college B’s behavior in the long term.

Andrew’s study uses simple but powerful graphical analysis to make a point, and is an absolutely great example of how simple economic theory and some knowledge of college behavior can deliver stunning insights. In the end, Andrew embraces the essential truth of the Bennett Hypotheis, modified a bit (what he calls, “Bennett Hypothesis 2.0″).

And Andrew points out that paying attention to the three “I’s” of reform—information, incentives, and innovation, can lead to a weakening in the operation of the Bennett Hypothesis, as could, of course, my preferred solution, which is for the federal government to get out of the college-financing business, particularly since it cannot responsibly handle its own internal finances.

Andrew’s study comes as two other empirical realities came to my attention that are, I think, broadly consistent with the Bennett Hypothesis. First, these aid programs have always been justified primarily on the grounds of offering equal educational (and thus economic) opportunity. Yet Postsecondary Education Opportunity points out in its January 2012 newsletter that the degree-attainment gap between those in the top and bottom income quartiles has widened dramatically since 1970—roughly the era of major federal student-financial-assistance programs. This trend is exactly what we would expect if aid payments to price-sensitive lower-income students are largely wiped out by aid-induced rising tuition costs. (Because upper-income students are less price sensitive, their ability to pay is not as damaged by tuition hikes in response to increases in federal financial aid.)

The second piece of news relates to continuing research that I am doing with Daniel Bennett, a Ph.D. candidate in economics at Florida State. Our research so far suggests that expanding college attainment is associated with less, not more, income equality. Could it be that the aid programs that were designed ostensibly to lower inequality have the precisely opposite effects? I must hasten that our research is ongoing, and further investigations and refinements could lead us to different conclusions.

Nonetheless, right now, I am betting that Joe Biden and Bill Bennett—the political odd couple if there ever was one—are right. Run, do not walk, to the CCAP Web site today to read the Gillen study.

A version of this article originally appeared in the Chronicle of Higher Education's Innovations blog on February 13, 2012. 

  • Share

Most Commented

October 25, 2022


NAS President Peter Wood Addresses the Pending Racial Preferences Cases

Read NAS president Peter Wood's remarks on the upcoming Supreme Court cases, which he presented at a meeting of "Oasis," an informal group of academics and intellectuals based in......

July 25, 2022


Against Transgenderism

The ideology of transgenderism strives to slam shut any door that offers opposition to its attempts to acquire power and control. This statement explains our opposition to such an ideology i......

October 20, 2022


NAS Statement on Nomination of Ben Sasse for University of Florida President

We believe that Senator Sasse would make an excellent president of the University of Florida, and we urge the Board of Trustees to follow the search committee’s recommendation....

Most Read

May 15, 2015


Where Did We Get the Idea That Only White People Can Be Racist?

A look at the double standard that has arisen regarding racism, illustrated recently by the reaction to a black professor's biased comments on Twitter....

October 12, 2010


Ask a Scholar: What is the True Definition of Latino?

What does it mean to be Latino? Are only Latin American people Latino, or does the term apply to anyone whose language derived from Latin?...

May 12, 2017


Harvard Prepares to Host All Black Graduation

Is Harvard's all black graduation a benign trend or a step backwards? ...