What the Numbers Say: Institutions Matter

by Erin Dillon on February 24, 2010

in Undergraduate Education

Yesterday I blogged about six HBCU’s in Texas that successfully reduced their student loan default rates without changing the students they enroll. These colleges provide an example of what institutions are capable of when resources and attention are focused on reducing loan defaults and increasing student success. But my colleague Robin Smiles and I looked at more than just the story of these successful institutions in ES’s recent report on student loan default rates, we also looked at the numbers behind the cohort default rates.

We analyzed student demographics and institutional characteristics at over 3,000 four-year and two-year colleges and universities to determine what factors contributed the most to an institution’s default rate. We found that while student characteristics do matter, they are certainly not the only predictor of an institution’s default rate, and that institutional characteristics, particularly a school’s ability to retain and graduate its students, are critically important to reducing loan default rates. Basically, that institutions matter—and they matter a lot.

As the chart below shows (Figure 2 from the report), some institutions do a better job of this than others, even while serving demographically similar students. The graph charts an institution’s actual cohort default rate and an institution’s predicted cohort default rate that we calculated based on the demographics of the students the institution enrolls, its HBCU status, and the percent of students admitted, as a measure of selectivity. The diagonal line indicates the point at which an institution’s predicted and actual default rates match perfectly. Institutions above the line have default rates that are better than expected, and institutions below the line have default rates that are worse than predicted.

Figure2

Note that NO institution has a predicted default rate as high as 20 percent, much less the 25 percent threshold for federal sanctions. Also, note the wide range in actual default rates among institutions with the same predicted default rates. As we note in the report, institutions with a predicted default rate between 10 and 11 percent had actual default rates ranging from 3.3 percent to 13.2 percent.

ES isn’t alone in its finding that institutions matter. As Student Lending Analytics noted yesterday, Sallie Mae highlighted in their recent investors meeting that, “You can see 30% differences in default rates with a borrower with a 700 [credit] score depending on what school that student attends.” (emphasis mine)

Bottom Line: Where a student enrolls can make a big difference in whether or not that student will default on his loan.

Comments on this entry are closed.

{ 1 trackback }

Previous post:

Next post: