Black college graduates are more likely than their white peers to leave school with student debt and to default on those loans. Those are both well-known, widely covered issues. Now a Brookings Institution report from a pair of researchers at Columbia University points out a troubling new finding: The gap in debt between black graduates and white graduates more than triples just several years after college, a crucial time for saving and laying the groundwork for retirement.

According to Judith Scott-Clayton, an associate professor of economics and education at the Teachers College at Columbia University, and Jing Li, a research associate in the Department of Education Policy and Social Analysis at the college, black graduates have almost $53,000 in student loans four years after graduation, close to double the amount white graduates hold. Just a few years earlier, when students collect their diplomas, the gap is just $7,400. As the authors note, those racial gaps are larger than many thought. They’re larger than they used to be, too, and seem to be tied not only to default rates, but to where black students enroll in graduate school.

“The magnitude of that gap is huge after four years; it’s enormous,” Scott-Clayton said in a phone interview.

Why is this finding just coming to light? One key reason is that the U.S. Education Department doesn’t always keep track of how borrowers are faring by race. The FAFSA does not include information on race, nor does the system that keeps track of outstanding debt. While the department does do some surveys every four years that get at racial gaps, those surveys don’t let researchers follow the same borrowers over a long period of time.

The lack of information has frustrated advocacy groups so much that earlier this summer they sent a letter to the department asking for more information. The department has said it plans to respond, but for now, researchers are still left trying to find creative ways to study the issue.

Scott-Clayton and Li obtained a set of data that is not entirely public from two of the department’s “Baccalaureate and Beyond” surveys, which followed a set of graduates from 1993 to 1997, and another from 2008 to 2012. Using that information, as well as other data from the department and the Census Bureau, they found the disparities. (The pair focused on white-black gaps because they are the widest.) They also found that nearly 50 percent of black graduates owe more money four years after graduation than they did when they finished school, meaning interest and fees have piled up. Overall, white graduates owe 10 percent less than what they borrowed four years after graduation, while black graduates owe six percent more.

It’s hard to tell exactly why the gaps are so large, since researchers don’t have great data by race across a long period of time. Some of it has to do with the fact that black borrowers have higher default rates, which may be partially tied to the fact that black college graduates earn less than white graduates. Separate studies have suggested that hiring managers are sometimes less likely to consider someone with a “black-sounding name,” and that black workers are paid less than white workers for the same job.

But a lot of the gap seems to be tied to graduate school. Scott-Clayton and Li found that 45 percent of the black-white gap is a result of differences in borrowing for post-baccalaureate studies. While 40 percent of black college graduates pick up debt to pay for these courses, only 22 percent of white graduates do the same. One factor is that blacks with bachelor’s degrees are more likely to go to graduate school at all. Almost half of the 2008 cohort of black graduates the researchers studied enrolled in graduate school within four years, compared to just 38 percent of white graduates. But crucially, of the black graduates who continue their studies, 28 percent sign up at a for-profit school, compared to less than 10 percent of their white peers.

A series of for-profits have been criticized—and even taken down—in recent years for preying on undergraduate students who don’t come from wealthy families, using them to access federal aid without providing a solid education (read: job prospects) in return. As the report notes, the Higher Education Reconciliation Act of 2005 upped the amount graduate students can borrow, which for-profits may have taken advantage of in terms of marketing to certain students for graduate programs. While the report acknowledges that enrollment has increased at traditional public and private universities, it notes that the for-profit industry is growing the fastest and is the only sector where growth is different by race. Crucially, these gaps remain even when parental income and education are considered.

This doesn’t necessarily mean graduate school is bad, Scott-Clayton said. A number of studies have pointed out that those with post-baccalaureate degrees make significantly more than people who stop after a bachelor’s degree. But the report does point out what may have been some “blind spots” for researchers to be aware of as they study the issue. As she and Li outline, they cannot tell which earnings are tied to which type of institution, and the rising enrollment in for-profit graduate programs is too new to really know the long-term effects.

The report offer a series of recommendations, among them the collection of data that would let researchers track debt and repayment by race, and more emphasis on understanding these dynamics at the graduate-school level. Ultimately, the authors write, “federal financial-aid policy alone cannot solve these problems—but neither can it ignore the challenges facing students of color who disproportionately bear the burden of student debt.” In other words, there’s no easy way to shrink the gap, but avoiding it will only make it worse.

Hayley Glatter contributed to this piece.