Repayment Rates vs. CDRs

Last week, the U.S. Department of Education released the latest set of cohort default rates. As expected, the overall average default rate continued its upward trend, with 7 percent of the students who entered repayment in the 2008 fiscal year defaulting on their federal loans within two years.

Though the average default rate is at its highest point since 1998, it’s still pretty low–93 percent of loans don’t default. But other data released by the department demonstrate that the actual repayment picture is far bleaker than these numbers suggest. Based on those numbers, about 54 to 56 percent of loan dollars borrowed by students at public or nonprofit colleges are repaid versus just 36 percent of the dollars disbursed to borrowers at for-profit colleges. From this it’s pretty clear that just because over 90 percent of borrowers supposedly don’t default, that doesn’t mean they are in active repayment either.

The discrepancy makes some sense. Cohort default rates are a measure of default–a process that takes a long time and is usually under counted thanks to the short measurement window. But just because a borrower avoids default doesn’t mean they are actively repaying their loans. In some cases that might be because borrowers use a deferment or forbearance that allows them to stop making payments without going into default. Others might be on an income-based repayment plan that reduces their monthly payment to zero because they owe so much money relative to their income.

The differences between the non-default rate and repayment rates still offer important lessons. For one, it underscores the fact that just trusting cohort default rates is pretty meaningless for judging actual student outcomes. As the large increases seen in the trial three-year rates suggest, schools work to actively manage these figures and keep them low. That’s not inherently a bad thing, but it also means we can’t assume that students will be OK if they just avoid default for two years. Second, it helps us identify at which schools the ability to avoid default appears to be a result of strategies that don’t involve repayment versus those where large numbers of borrowers are able to handle their debt payments.

To test the difference, I took the repayment rate information provided by the department  and matched it with the cumulative cohort default rate by school for the last four cohorts (2005 through 2008). I also paired this information with sector and state directory information from the department’s Integrated Postsecondary Education Data System. Finally, I dropped all schools that had fewer than 30 borrowers in the default rate cohort or in the repayment rate denominator.

The table below shows the results of this comparison by institution type (the non-default rate is 100 minus the cohort default rate):

For those with good eyesight, here’s the same chart broken down further by sector:

A few observations. For-profit colleges have a much larger difference between their non-default rate and their repayment rate than public or nonprofit institutions. This is because the difference between repayment rates at for-profit colleges relative to the other sectors is much greater than the difference in default rates between for-profits and other sectors. For-profits have a non-default rate that is between 93 and 95 percent the rate for the other sectors, but their repayment rate is between 58 and 70 percent of the figures for publics and nonprofits.

At a sector level, 2-year publics and all types of for-profits all had differences of above 50 percentage points. But the difference at community colleges was still below that of even four-year for-profit institutions. That said, even the sector with the smallest gap, public institutions of 4-years or above, still had a difference of 41 percentage points.

Out of curiosity, I also ran the same analysis at the state level. North Dakota had the smallest gap at 23 percentage points, followed by Vermont (25 percentage points) and Wyoming (28 percentage points). At the other end, Mississippi had the biggest gap at 57 percentage points, followed by Louisiana (54 percentage points), and Alabama (52 percentage points). At the national level, the gap is 43 percentage points.

Here are the 10 schools with the largest gaps:

Had I used Carnegie classification codes, the theological seminary and the medical schools would have likely been excluded because graduates of medical and divinity schools have to go through extended internship/apprenticeship programs before they start earning actual incomes. But the beauty schools would remain.

Here is the opposite of that list, the schools with the smallest percentage point difference:

What’s interesting here is that some medical and divinity schools are present, so clearly just being in that type of program is not an automatic reason for lower repayment. This list also contains two schools that the general public are more likely to be familiar with: Cal Tech and Harvey Mudd.

It would be foolish to expect the non-default rate of students to match the student loan repayment rates. That said, the large differences between these two figures underscore the need to take a closer look at not just if a student avoids default, but also whether they are actually in a position to pay back their loans. Sometimes that might mean a temporary deferment and forbearance, which is understandable and completely reasonable. But if that is occurring in large numbers and with little sign of students ever being able to repay their debt, then it is certainly worth taking a closer look.