Pretty much everyone agrees that the federal financial-aid system is broken. But fewer agree on exactly how it is broken, or on what should be done to fix it. In fact, despite the amount of money spent on aid and the number of students who rely on it, experts’ understanding of how the current system affects students remains limited.
But the march to make a better student-aid system goes on. The most recent effort is Reimagining Aid Design and Delivery, a project sponsored by the Bill & Melinda Gates Foundation. In awarding grants for organizations to produce proposals, the foundation aimed to “spark a robust discussion about how financial aid can be used as a lever to increase student success,” it said in a news release. And that hoped-for discussion would begin just as higher education turns its attention to the reauthorization of its major guiding legislation, a process expected to start in 2014.
Just this week, the project’s final grant recipient released a report with its recommendations. The Gates foundation expects to announce a second round of grants in mid-May. It won’t yet say how much money is involved, or who was invited to compete. The goal, Debbie Robinson, a spokeswoman, wrote in an e-mail, is to move toward “convergence and depth on policy options.”
There are now proposals from 16 associations, advocacy groups, and other organizations offering a host of ideas for changing the major aid programs, the information students are given, and the measures on which colleges are held accountable. Some specific recommendations are shared by several of the reports, while others are contradictory.
Differing Ideas
The lack of consensus on how to fix the aid system stems from different ideas of what it is supposed to do in the first place. Many would describe the system’s role simply as providing access, by making college more affordable. That’s how Title IV of the Higher Education Act, which authorized most of the federal programs, explains it.
But in its announcement of the RADD project, the Gates foundation describes aid as a tool to promote college completion. That definition is not universally embraced: Some are wary of any move to tie aid to graduation. After all, one of the easiest ways to enhance graduation rates is to box out anyone unlikely to succeed.
The Gates proposals advocate approaches to change a system no one likes, but uncertainty remains not only about which strategies would be best, but whether they would even work.
The financial-aid landscape has changed considerably since the Higher Education Act was signed in 1965. Today, more colleges receive aid to educate far more students, who come from wildly disparate backgrounds. The system that once served a minority of college students now provides money to two-thirds of undergraduates. That means conversations about the system’s effectiveness have become part of the broader discussion of how to improve higher education, says Judith Scott-Clayton, assistant professor of economics and education at Columbia University’s Teachers College.
In other words, the aid system now touches enough students to be seen as a way of steering college outcomes.
Still, our understanding of the aid system’s effects on student behavior remains hazy. For all the change in college-going patterns, much of the data the government collects about students are still based on a first-time, full-time freshman world. And there are plenty of things researchers and the public might want to know that are not being tracked at all.
The U.S. Education Department “doesn’t have a lot of data that any business would have readily available,” says Mark Kantrowitz, the aid expert behind FinAid and FastWeb, two well-regarded consumer Web sites. When private lenders started exiting the federal loan programs during the credit crunch, he says, the department didn’t provide good information on which ones were still participating.
Since the government is not allowed to create a national student “unit record” system, tracking how aid affects individual students is challenging, says Gigi G. Jones, director of research at the National Association of Student Financial Aid Administrators, one of the grant recipients. Researchers are left with government data sets that include only a sample of students.
The Education Department does keep a lot of transactional data on student loans, says Jason Delisle, who directs the Federal Education Budget Project at the New America Foundation. But it doesn’t ask borrowers about the choices they make, says Mr. Delisle, who helped write two of the project’s reports, from New America and HCM Strategists, an education and health consulting company. That means we have information about what happens to aid recipients but not about why. So when some experts claim that moving all borrowers to an income-based repayment plan would eliminate defaults, Mr. Delisle is not persuaded. That conclusion assumes borrowers don’t make their student-loan payments because they are too large relative to their income, he says. Maybe the borrower has a decent income but is carrying lots of other debt.
How can policy makers possibly craft student-aid programs with so little evidence available? Well, they’re not scientists, Mr. Kantrowitz says. Policy is driven by politics, not research. Making policy, he says, is about reacting to anecdotal examples and keeping programs that appeal to the middle- and upper-income families who make their voices heard.
No Control Group
But it’s hard to say how well the system is doing, even on the most basic question of whether it provides access. Take the Pell Grant program, the most essential government aid for needy students. Most researchers agree that there’s not enough evidence to prove Pell works the way its advocates want it to, says Sandy Baum, a senior fellow at George Washington University’s School of Education and Human Development and an independent higher-education-policy analyst.
That’s not to say that Pell doesn’t promote access. “There are all kinds of reasons why even if it is having a big impact, it’s hard to measure,” says Ms. Baum, who was part of the coalition behind one of HCM Strategists’ reports. After all, there’s no control group of needy college students who are not given Pell Grants, and the program’s incremental changes over time make it difficult to pinpoint examples of cause and effect.
To be sure, researchers can tease out causal relationships when it comes to aid in certain situations. The nonprofit research group MDRC has run a number of experiments testing performance-based grants in various settings. And research also happens at the institutional level. The University of North Carolina at Chapel Hill uses a comparison group of students who would have qualified for its Carolina Covenant aid program but enrolled before the program began, as a way to measure its effectiveness.
But while such investigations can pinpoint real results, it’s not always clear what they mean for the wider world. Smaller studies “give us clues at best,” says Shirley Ort, associate provost and director of scholarships and student aid at North Carolina. The question, then, is “what among those clues might we want to try?” says Ms. Ort, who helped with the National Association of Student Financial Aid Administrators’ report.
It would be nice to be able to test out new ideas, Ms. Ort says. Some of the grant recipients pointed out places where their proposals would need to be studied more before being adopted. And one group, HCM Strategists, suggested that the government devote more money to student-aid research and development to answer questions like how to more cheaply and effectively prepare students to succeed in college-level work.
There are certainly players who want to move forward with their agenda for the aid system whether they have evidence or not, Ms. Baum says. But at the same time, the country can’t afford to wait for perfect proof that an idea will work before trying it out. Moving forward on an untested policy change could have unintended consequences. But doing nothing is also a choice that could end badly. So for now, policy makers must forge ahead, she says, using logic to draw conclusions from existing research.
