Overall the submissions from negotiators demonstrate the extent to which the discussions sessions are focused on two separate conversations that are touching on the same global set of issues, but not the same narrower topics. On one side are proposals from the representatives of for-profit colleges—particularly those from Marc Jerome from Monroe College in New York and Raymond Testa from Empire Education Group, a chain of beauty schools. These submissions argue for tweaks to the rule that would result in a proposal that is a more forgiving version of what the Department issued in 2011. On the other are proposals from consumer advocates that are focused on three main areas: (1) job placement rates, (2) an approval process for new or existing programs, and (3) relief for student borrowers. The first and second issues represent additional forms of accountability, while the third gets at an additional form of borrower protection.
Party Like It's 2011The lawsuit brought by for-profit colleges that ultimately struck down the first gainful employment regulation challenged whether the Department had the legal authority to promulgate the rule. The subsequent judge's verdict affirmed that the Department had the authority to write a rule in this area, but invalidated the final one published it based upon the justification for the threshold chosen for acceptable levels of student loan repayment.
Having the Department's legal authority upheld has changed the dynamic among the for-profit negotiators. Rather than arguing the Department should do nothing to hold career programs accountable, they are instead pushing for the Department to establish a rule that is close to the 2011 rule instead of the stronger 2013 proposal. For example, the 2013 rule proposed to add an additional performance category for programs where graduates' ratio of student debt payments to income was below the failing level of 12 percent of annual or 30 percent of discretionary income, but was higher than the suggested passing threshold of 8 percent for annual and 20 percent for discretionary income. The proposal from Testa calls for the elimination of this performance level entirely; the one from Jerome would keep it, but increase the thresholds on it so that programs would pass if their debt-to-earnings was at or below 12 percent for annual or 30 percent for discretionary; it would be in the middle if the ratio was between 12 percent and 15 percent for annual or 30 percent and 35 percent for discretionary. The table below breaks down what these different proposals look like and how they compare to the Departments' 2011 rule and 2013 proposal.
[table id=35 /]
As the table above shows, Testa's proposal would duplicate what the Department did before, while Jerome's would duplicate it but with an even more lenient threshold for failing. Here's how the proposals would work out based upon actual earnings and debt data for 2008 and 2009 graduates.
[table id=37 /]
The results are not that dissimilar across proposals. The biggest difference is the Department's 2013 proposal would shift about 13 percent of programs from passing to a middle status, while Jerome wants to make even fewer programs at risk of failing.
The return to 2011 is not limited to thresholds. Both Jerome and Testa also called for having the Department calculate the average monthly loan debt payments using different estimates for how long borrowers would repay based upon the type of credential. In other words, payments for bachelor's degree programs would be based on a 15-year repayment timeframe, while those in a certificate program would be based on 10 years. This is exactly what the Department had in its 2011 rule, which was added as a concession because lengthening the repayment time will reduce annual debt payments, thus allowing programs whose graduates have lower average incomes to pass. While Testa and Jerome each argue that this change better reflects actual time to repay, statistics from the Department note that the vast majority of borrowers are on the 10-year repayment plan and the average overall time to repay is 12 years.
Finally, the Jerome and Testa proposals also argue for allowing programs to cap the debts of students at the amount charged for tuition and fees, another concession by the Department included in the 2011 rule that would lower student debt payments and allow programs with lower average graduate earnings to pass. They argue for this change on the grounds that students may borrow for things like living expenses, which the institution cannot directly control. The notion of overborrowing or taking on debt in excess of institutional charges is a common argument coming from schools, though there's no concrete data to back up the frequency or extent to which this actually occurs. Nor is there any information around the extent to which borrowing for living expenses contributes to completion or not.
Repayment RatesThe repayment rate threshold ultimately undid the last rule, but this measure can serve an important accountability purpose by judging programs on the outcomes for both their graduates and non-graduates. And so a work group came up with a proposal to add repayment rates back in. They did not, however, set a threshold for what these rates should be, but instead called for an expert panel to come up with a high, moderate, and low repayment rate. In the work group's formulation, programs with high repayment rates would not be subject to further accountability. Programs with moderate rates would have to pass a debt-to-earnings test, have their job placement rates audited, and have new programs seek upfront approval. Low repayment rates would be subject to the same requirements as a program with a moderate repayment rate, but would lose eligibility after two years of low rates. It's a concrete idea that would add the rate back in, but still struggles with the need to define a threshold.
This proposal is quite different from the one released earlier this week by New America in "Improving Gainful Improvement." In that paper, I argued that repayment should be based upon the principal reduction of the cumulative amount borrowed for a program. Though the idea of a high, middle, and low repayment rate could be applied to this pooled repayment rate.
- A high repayment rate could be one where the total amount of outstanding principal at the end of four years in repayment is equal to or less than what the total amount of principal outstanding would be if borrowers were retiring their debts as scheduled on the standard 10-year repayment plan.
- A middle rate could be if the principal reduction is at least equal to the amount that would be remaining if the borrowers were retiring their debts on a 12-year repayment schedule--the average length of time it takes a borrower to repay.
- A low repayment rate could be if the principal reduction is less than the amount of principal that would be outstanding after 20 years--meaning borrowers would retire their debts before being eligible for loan forgiveness through the newer forms of Income-Based Repayment.
[table id=40 /]
Mark it ZeroSeveral proposals talked about what to do if a program ended up having average or median loan debt of $0. Since these programs would automatically have a debt-to-earnings ratio of 0, they would all pass, but negotiators representing community colleges in particular wanted to make this clear. Richard Heath from Anne Arundel Community College and Kevin Jensen from the College of Western Idaho argued that programs that can attest to having no average debt should just be exempted from further eligibility reporting and determinations. This idea seems more designed to allay concerns about schools than represent an actual change in results for who is held accountable, but that could change if other measures of some sort were added.
The Waiting is the Hardest PartNegotiators representing for-profit colleges were particularly concerned about when penalties should start kicking in for colleges. Testa called for a hold harmless of a year or two. He also called for programs that require state licensing or a board examination should be measured on graduates after five and six years in repayment, not three and four. Jerome, meanwhile, asked for potentially much longer. His proposal entails one year of rates that are only for informational purposes and have no consequences. Then, colleges would get a transition period before penalties could occur that would be equal to 100 percent of the typical time to complete a program--i.e., one year for a one-year certificate, four years for a bachelor's degree. That means delays of anywhere from two to five years before sanctions could occur. Its not clear if programs would not start having their failures counted until after the transition period, which could potentially lengthen the time even further. Either way, a bachelor's degree program would end up avoiding consequences until the re-election campaign of the next President.
Placement RatesThe negotiators from states' attorneys generals offices are particularly interested in placement rates because it is a common legal hook for them to pursue action against schools. And the lack of a clear definition clearly makes these efforts more difficult, even in the face of obvious fraud. In two documents, the group working on placement rates argued about the need for a definition and laid out their suggestion. Under their proposal, a successfully placed student is one who within 180 days of graduation (or after passing a necessary licensing test) has worked for at least 13 weeks in a full-time position in the relevant field. This is the same definition that currently applies to very short-term programs. Individuals working part time could be counted as placed only with a signed attestation that they intentionally pursued part-time work. The definition also provides greater clarity of what a relevant field is, by saying it has to be a job that the U.S. Department of Labor notes is connected to the program or is a position that requires at least some postsecondary education and makes routine use of the skills and knowledge taught in the program. It also makes allowances for counting someone as placed if they stay at their same employer or at least work in the applicable industry at a salary at or above the 25th percentile as determined by the Bureau of Labor Statistics.
Beyond the definition, there's also the question of what threshold to choose for a placement rate. The suggestion appears to be 70 percent, perhaps paired with a 70 percent completion rate, which is what short-term programs require. But the group did include this interesting chart that breaks down what other states or accreditors require.
Student Debt ReliefBoth the for-profit college representatives and the more student focused groups presented ideas around student debt relief. But the former argued for it as a measure to help programs pass, while the latter argued harmed borrowers should receive relief. On the student-focused side, the proposal called for using a program's fail or middle zone status to be grounds for not having to make loan payments or deal with collection actions. This would allow borrowers who went through bad programs to not still have to pay the debts they took on to attend them. The proposal also raised the idea of making schools refund the tuition paid at bad programs, sign a letter of credit to cover potential loan cancellations, or form a "victim's fund" to offset costs for harmed borrowers.
On the student side, Jerome proposed that failing programs should be allowed to limit the student debt of new and continuing students to the point where the program could pass either of the debt-to-earnings measures. And a program would be able to use these measures for the number of years that a student typically takes to complete a program. Using the measures would also exempt a school from debt warnings and enrollment caps. The mechanics of how this would work are not completely clear since there would have to be some judgment of possible income levels, and it's not clear if colleges would just not let students borrow for living expenses,which they may need to graduate. In addition, Jerome asked called for a loan forgiveness program for students that do not finish their first term. In both cases, the goal is clearly focused on ways to hit the numerical targets and less about underlying educational improvements.
Program-Level Cohort Default RatesWhen the Department suggested the idea of a program-level cohort default rate as another accountability metric, it said it would operate separately from the debt-to-earnings rates. In other words, programs that failed the cohort default rate test would still fail, regardless of how they performed on the debt-to-earnings figures. In the working group on this issue, college representatives called for making them connected--so a program that failed the cohort default rate could be "saved" by passing the other rate. They also called for using a larger minimum program size than the 10 that the Department suggested using for debt-to-earnings rates. Connecting cohort default rates to other results would weaken the rule, since every additional measure that can save a program just gives programs another way to avoid penalties. The working group did, however, adopt a proposal from Rory O'Sullivan of Young Invincibles that the Department release data on how much cohort default rates may be manipulated using deferment or forbearance.
Upfront Program ApprovalsSeveral negotiators released their own proposals for how to handle the approval process of gainful employment programs, especially for new ones looking to join the federal student aid programs for the first time. Barmak Nassirian, a negotiator from the American Association of State Colleges and Universities, argued for an approval process that would potentially apply to all programs. Programs where instructional expenses made up more than half of total spending or where the average amount borrowed was $0 would be automatically approved. All other programs would have to submit an application that contained basic information about the program as well as projected debt-to-earnings ratios, borrowing amounts, costs, and a number of other factors. Figures would require an attestation of accuracy from institutional officials. Any program that projected it would fail or be in the middle zone of a debt-to-earnings test would automatically be denied. There would be mechanisms for denied programs to still participate that would involve a letter of credit, while approved ones would have greater oversight through a high-risk designation.
A working group on program approvals released a somewhat similar proposal. Among the criteria it laid out for program approvals are: having programmatic accreditation that is needed to obtain relevant certification or licensing; whether the level of education is not higher than what is generally required by employers (e.g., not a bachelor's degree in cosmetology when a certificate suffices); the program is no more than 10 percent longer than what is typically required; projected earnings are above 150 percent of the poverty level for two individuals (about $23,265); it would pass a debt-to-earnings test; and has placement opportunities lined up. The proposal also contains a lot of details on how these figures would be calculated.
Jones, from Strayer University, offered a different take. His proposal for program approval focused on what other regulatory oversight currently exists. He argues programs that have been substantively reviewed by a state or accreditor should not be subject to review, neither should programs that would be reviewed as part of a change to a program participation agreement. Finally, colleges where all programs pass for two years would not have new programs subject to review.
The proposal from Heath and Jensen also focused on the performance of an institution's programs. Their proposal calls for institutions to have their programs evaluated if they have at least one failing program or two in the middle zone in a given year. Any program that is voluntarily closed after not doing well on the tests or loses eligibility must seek approval if it wants to re-enter the aid programs within five years.