College Scorecard

This post was updated on 3/20/17 to reflect repayment rates revised in January, 2017

In an effort to hold colleges accountable, the U.S. Department of Education measures how many borrowers from a college default on their loans – a “cohort default rate” which measures just the tip of the borrower-distress iceberg. In conjunction with its College Scorecard, the U.S. Department of Education has also begun releasing college repayment rates, which measure the share of borrowers paying down the principal balance on their federal loan debt.  

As the Department states in its College Scorecard documentation, default rates “can be manipulated through the use of allowable nonrepayment options like deferments and forbearances” which serve to temporarily keep default rates down and colleges in compliance with federal limits on defaults. Indeed, some for-profit colleges have admitted to doing just that.

While available default and repayment rates have some differences (most notably, default rates include graduate students and the repayment rates do not), they are comparable enough to identify trends and outliers. Comparing default rates and repayment rates tells you about how many students have avoided default but still aren’t paying down their loans: perhaps they’re delinquent, in forbearance or deferment, or in a repayment plan where their balance is growing rather than shrinking. And in schools where default rate manipulation is occurring, this group of borrowers – the missing middle group of those neither in default nor paying down their loans – will be atypically large. 

Across all schools, this missing middle group makes up about 44 percent of borrowers three years into repayment on average. But at 456 schools, 60 percent or more of borrowers are neither in default nor paying down principal. The vast majority of these schools (78%) is for-profit colleges, including Kaplan University which previously hired private investigators to track down former students to put them in forbearance. It includes CollegeAmerica, sued by the U.S. Department of Justice for violating rules on incentive compensation and by the Colorado Attorney General for fraud. It includes National College, which last August was ordered to pay $157,000 to the state of Kentucky for its years-long refusal to comply with a subpoena related to potential job placement rate falsification. It includes now-shuttered Westwood College, which had faced charges of deceptive marketing and recruiting for years and been sued by the Colorado and Illinois Attorneys General. Computer Systems Institute and Marinello Schools of Beauty, both of which were cut off of federal aid after this year having been found in violation of several federal rules, also made the list. This group of schools also includes some Everest College campuses that remain open for business. While Everest schools have new corporate management, the former CEO had this to say about managing cohort default rates during a 2011 investor call: “Forbearance, as you well know, is a pretty easy, just a question you have to agree to it and you’re on your way.”

​Forbearance abuse for the purpose of evading accountability is well documented in the for-profit college sector, but has not been documented at other types of institutions. The pie chart below shows the distribution of schools where 60 percent or more of borrowers are neither in default nor paying down their loans three years into repayment, and the list of schools is available here

Posted in

| Tagged

This post was updated on 3/20/17 to reflect repayment rates revised in January 2017

New Department of Education data released in conjunction with the College Scorecard clearly show how much student loan repayment rates vary by whether or not a student completes their program and the type of school a student attends. Repayment rates in the College Scorecard measure the percent of undergraduate federal student loan borrowers making any progress on paying down their debt (i.e. the share that have paid down at least $1 of their balance when they entered repayment).

As shown in the table below, borrowers who do not complete their program are less likely to be paying down their debt than those who graduate (34% versus 60%). We also found that, for both students who finish their programs and those who do not, repayment rates vary substantially based on the type of college they attend, with students who attend for-profit colleges being the least likely to be paying down any of their federal student loan balance. 

Posted in

| Tagged

This post was updated on 3/20/17 to reflect repayment rates revised in January 2017

The U.S. Department of Education recently released new data on college costs, outcomes, and debt in conjunction with their College Scorecard – a helpful tool for students and families and a treasure trove of data for analysts. 

We took a look at the data to find some of the schools where debt problems are particularly severe: the schools at which the majority of students borrows, and a minority of borrowers is paying down their debt three years into repayment. We identified nearly 1,500 schools that met these criteria. For-profit colleges make up 70% of the 1,478 schools. Of the remaining 447 colleges that meet these criteria, 247 are nonprofit colleges and 200 are public colleges. These data are detailed in the table below, and a list of the 1,478 schools is available here

Posted in

| Tagged

We’re big advocates of providing students and families with tools and information to help them decide where to go to college and how to pay for it, and the Obama Administration deserves credit for making huge progress in this area. They created the College Scorecard, released new college-level data, and encouraged schools to use its voluntary “Shopping Sheet” format for financial aid award letters. Without the Shopping Sheet, college financial aid packages can be incredibly difficult to decipher and compare. It serves the same purpose as the standard window sticker required on all new cars since the 1950s—to provide key information in a consistent format so consumers can more easily understand their options.

Unfortunately, the Shopping Sheet and College Scorecard currently display different cumulative debt figures, such that the same institution can appear to be a low debt school on the Shopping Sheet and a high debt school on the College Scorecard.  While the College Scorecard shows the median debt of program completers only, the Shopping Sheet displays the median debt of all students entering repayment, regardless of whether they completed their program.* Including non-completers in the median debt calculation will often produce lower debt figures because non-completers borrow for a shorter period of time, sometimes only a semester or two before leaving a program. This is especially true for colleges where many students borrow but fail to graduate. As a result, the debt figure included on the Shopping Sheet inadvertently makes colleges with low completion rates (and high borrowing rates) look much more affordable than they actually are. This makes it harder for students to determine where they are likely to graduate without burdensome debt.  

The College Shopping Sheet Should Show Median Debt Among Completers, Just Like the College Scorecard

The table below demonstrates just how different the debt figures for the same school can be when you include all students entering repayment versus only program completers, particularly at colleges with low graduation rates.** For example, the median debt of completers at Ashford University is nearly three times higher than the median debt figure of all students who borrowed ($32,813 versus $11,190). This difference can be explained by the fact that 79 percent of students drop out of Ashford before they accumulate as much debt as those who complete their program. A similar difference is also evident at the University of Phoenix-Online, where 80 percent of students do not complete their program.

Differences in Median Debt Calculations


The problem is compounded when you compare schools with low and high graduation rates. To illustrate the potential magnitude of the impact of completion rates on median debt, we looked at median debt figures for Ashford University (where just 21 percent of students graduate) and Stanford University (where 95 percent of students graduate). While the typical Stanford graduate has 63 percent less debt than the typical Ashford graduate ($12,224 versus $32,813), the median debt among all students for the two schools is nearly identical ($11,190 versus $11,500).  Similarly, students at University of Phoenix-Online Campus and Florida State University have nearly identical median debt figures when non-completers are included. But the schools have dramatically different graduation rates (20 percent versus 76 percent), and graduates of the University of Phoenix-Online Campus typically have 67 percent more debt than graduates of Florida State University ($35,500 versus $21,250).

Students shopping for schools based on affordability are aiming to graduate with a degree, and should be informed about the debt they can expect when they accomplish that goal.  We’ve previously called attention to the problems with combining the debt levels of completers and dropouts (e.g., here and here), and we praised the Department last year for putting the median debt at graduation on the College Scorecard. We’re disappointed that the Department did not put median debt at graduation on the Shopping Sheet as well this year. Given the Department’s efforts to highlight colleges that do a good job of graduating students, it is particularly surprising that the Shopping Sheet uses a debt metric that makes colleges with high borrowing rates and low completion rates look more affordable than they are. We hope the next version of the Shopping Sheet will show debt for completers only, so students and families will know how much debt they can expect to have at graduation from different colleges.

* In addition, Shopping Sheet figures reflect one year of data, i.e., median cumulative federal student loan debt for undergraduates entering repayment in 2013-14, while Scorecard figures represents the same metric for those entering repayment in 2012-13 and in 2013-14 (pooled).

** Calculations by TICAS on data from U.S. Department of Education, College Scorecard.  Graduation rates shown here measure the share of first-time, full-time students who started in 2006-07 and 2007-08 who completed their programs within 150% of normal time by 2013-14. Median debt figures represent cumulative federal student loans borrowed for undergraduate education among borrowers who entered repayment in 2012-13 and 2013-14.

Posted in

| Tagged

Beyond the fact that they’re almost all open-admission institutions, there are more differences than similarities between community colleges and for-profit colleges, including when it comes to student debt. New data released by the U.S. Department of Education in conjunction with the updated College Scorecard show just how different. Three comparisons are below.

  1. The vast majority of schools where debt problems are most severe are for-profit collegesInside Higher Ed’s recent analysis looked at schools where the majority of students borrows, and a minority of students is paying down their debt seven years into repayment. Our additional analysis found that, of the 257 schools that meet those criteria, 227 (88%) are for-profit colleges. Only two schools (fewer than 1%) are community colleges.
  1. Community college borrowers are much more likely to be paying down their debt. Three years after entering repayment, federal loan borrowers who attended community colleges were much more likely than for-profit college borrowers to have begun paying down their balance. Most strikingly, the available data on repayment rates by completion status show that borrowers who completed their studies at a for-profit college were about as likely to be paying down their loans as students who withdrew from a community college (53 and 51 percent, respectively). 
  1. Many more students at for-profit colleges are neither paying down their loans nor in default. While available default and repayment rates have some differences (most notably, default rates include graduate students and repayment rates do not), they are comparable enough to identify trends and outliers. Comparing default rates and repayment rates tells you how many students have avoided default but still aren’t faring well: perhaps they’re delinquent, in forbearance or deferment, or in a repayment plan where their balance is growing rather than shrinking. Some for-profit colleges have admitted to what the Department, in its documentation of the Scorecard data, described as “gaming behavior that may push students toward forbearance and deferments, meaning they stay out of default but don’t make progress on repaying their loans and may continue to accrue interest.”

    ​Across all schools, this missing middle group – those neither in default nor paying down their loans – composes on average about 21 percent of borrowers three years into repayment. But at 483 schools, 40 percent or more of borrowers are neither in default nor paying down principal. The vast majority of these schools (79%) is for-profit colleges, including Kaplan University and several Kaplan Colleges and Kaplan Career Institutes, which previously shared a parent company that hired private investigators to track down former students to put them in forbearance. It includes several campuses of Education Management Corporation-owned schools Argosy, Brown Mackie, and the Art Institutes. It includes Harris School of Business, Drake College of Business, and Westwood College. This missing middle group also makes up more than half of all borrowers at several Everest College campuses that remain open for business. While Everest schools are now under new corporate management, the former CEO had this to say about managing cohort default rates during a 2011 investor call: “Forbearance, as you well know, is a pretty easy, just a question you have to agree to it and you’re on your way [sic].”

    ​Forbearance abuse for the purpose of evading accountability is well documented in the for-profit college sector, but has not been documented at other types of institutions. Just 6% of the schools where 40 percent or more of borrowers are neither in default nor paying down their loans three years into repayment are community colleges. 

Any college with default and repayment problems of any scale should be focused on better enabling their students to repay their loans. But as the newly available data continue to underscore, for-profit colleges have by far the farthest to go.

Posted in

| Tagged

As promised in last night’s State of the Union Address, today the Administration unveiled its new College Scorecard. By providing information about colleges’ costs and outcomes in a clear and comparable format, it has the potential to be a game-changer for higher education. We have long touted the importance of the Scorecard and other tools designed to help students and families pick a college, and we applaud the Administration for supporting and promoting a higher education agenda that puts students first.

Overall, the data provided on the Scorecard are what students and families need to better understand their college options. Today’s version of the Scorecard includes some marked improvement over earlier drafts: it is more interactive and now links directly to colleges’ own net price calculators. However, two types of data on the Scorecard are less helpful and even downright misleading:

  • Loan default rates are provided without any context about how many students at the college borrow and are therefore at risk of ever defaulting.  For instance, American River College’s default rate of 27.5% is much higher than the national average of 13.4%. But what consumers can’t tell from the Scorecard is that only 8% of American River College students actually borrow federal loans. The default rate only represents the share of borrowers – not students – who default. Certainly American River should work on improving its default rate. But when 92% of its students never borrowed in the first place, implying that the default rate alone is indicative of student outcomes more generally does a disservice to would-be students.
  • The median borrowing figures provided are for all federal loan borrowers who entered repayment, regardless of whether the student entered repayment after graduating or dropping out after a semester or two. This makes colleges with high drop-out rates look like a good deal, compares apples to oranges, and undermines the value of other outcome information. Here’s an example. The Scorecard shows that the median federal debt of borrowers entering repayment at Grand Canyon University and Duke University (which both primarily grant bachelor’s degrees) is very similar: $9,500 at Grand Canyon and $8,840 at Duke.

Meanwhile, other federal data not reflected on the Scorecard show that these figures represent very different student outcomes. Because 41% of entering first-time, full-time students at Grand Canyon didn’t return for a second year, the low median debt for borrowers entering repayment reflects just one year of loans in many cases.  Indeed, the average federal loan amount for undergraduate borrowers at Grand Canyon in 2010-11 – what they borrowed in just that one year -- is a remarkably similar figure: $9,444. In contrast, all but 3% of entering first-time, full-time students at Duke return for a second year. And the average annual federal loan amount for undergraduate borrowers is $3,751, less than half their median debt at repayment. That makes sense considering 94% of their entering students leave with degrees.

In these cases, bad comparisons are worse than no comparisons.  Fortunately, both of these data problems have simple fixes.  It would be easy for the Department of Education to add the share of students borrowing to the loan default rate box and provide the necessary context for interpreting default rates. For the median borrowing figures, the Department is already taking steps to collect cumulative federal debt at graduation -- which would be an apples-to-apples comparison of students at the same point in their academic trajectory. Until those figures are available, the Scorecard should take the same approach to median borrowing as it already does to earnings information: make clear that federal data aren’t yet available and encourage prospective students to ask the school for more information.

For more about how to improve information for students and families and other ways to increase college affordability and completion, see TICAS’ new white paper (released yesterday): Aligning the Means and the Ends: How to Improve Federal Student Aid and Increase College Access and Success.  

Posted in

| Tagged

Subscribe to College Scorecard