Tulane students in class. Used under Creative Commons license. Original photo by Flickr user Tulane Public relations.

Today begins the second negotiation session around employment at the U.S. Department of Education offices on K Street. Negotiators will meet for at least three full days from Monday to Wednesday. Live updates will appear below.

9 a.m. Opening Remarks
Jeff Appel, the deputy undersecretary at the U.S. Department of Education is delivering some opening remarks. He notes that the new draft is more comprehensive and reflects the Department’s goals. But he also says “we made tremendous progress but want to emphasize this is not a finished project.” He calls for concrete ideas and indicates that the Department is open to discussing all parts of the rule and is interested in achieving consensus.

Appel walks through some of the changes made to the rule in response to concerns heard by negotiators. The most noteworthy ones include the addition of a loan portfolio repayment rate and a program cohort default rate, as well as new provisions that require programs that are failing to provide some relief for borrowers. While he notes that the two new measures do not have impact estimates yet, he says that they are working on running data and providing estimates as soon as high-quality data are available.

Appel also notes that ideas around a placement rate were not accepted by the Department, but notes those can be discussed.

If you want to read a summary of what the Department is proposing to do, click here. A summary of what the negotiators had proposed is here.

9:10 a.m. Draft Agenda
There’s a discussion of what consensus means. Consensus is assumed by absence of dissent. But there’s nothing binding until the last day. The goal is to reach consensus on the package, but in the past the Department has tried to reach tentative consensus on sub issues.

Marc Jerome from Monroe College objects to the proposed agenda, which would have started with the existing program certification or new program approval on the grounds that he would prefer to talk about the accountability metrics first. Brian Jones from Strayer University also agrees to moving up the accountability metrics first. No one objects.

9:25 a.m. Proposed Text Overview
John Kolotos, one of the Department’s negotiators talks through the new proposal. He notes that the program-level cohort default rate has the same thresholds as what’s in statute–above 30 percent for three years or 40 percent in one year. For the portfolio loan repayment rate, a program fails if the principal balance owed at the start of the year is more than it was at the start of the year.

Kolotos also notes that the draft has new provisions on getting new gainful employment programs approved. He says the goal is to limit the review to programs where there were previous failures. There’s also a new piece around borrower relief. What this would do is not absolve borrowers of the debt, but bring their debt levels down to acceptable levels.

Kolotos turns to a discussion of the materials submitted. He thanks the negotiators for their hard work and high-quality of submissions. He indicates that he’s worried that some ideas lack the support needed to be included in the rule. He says there were some concepts that were new or too theoretical that the Department does not think it could implement at this time.
We’re going to start the morning with a discussion of the accountability metrics, then move onto other issues. There’s nothing new in the debt-to-earnings measure from the last negotiation session.

Jerome from New York-based Monroe College asks how the Department came up with the repayment rate calculation idea. Kolotos says it was tough for them to take the repayment rate idea submitted by the negotiators, so thought about how to get something that was repayment-rate like and came up internally with the idea of the repayment rate.

Richard Heath from Anne Arundel Community College raises concerns about judging programs based upon all loan debt, not just that for tuition and fees.

Jones from Strayer wants to back up and discuss how this proposal deviates from the prior rule, which was finalized in 2011 and struck down in 2012. He notes that rule had different means of compliance, meaning that a program only had to pass one of three measures, whereas now it has to pass everything. He asks for a rationale of the policy shift and how these measures are intended to work together. Kolotos says that thresholds on the metrics would affect only the worst-performing programs, which is not different from how it approached this rule before. For the program cohort default rate, he notes that the Department is moving the statutory framework for the institutional cohort default rate and so that means taking the statutory consequences. For the repayment rate, Kolotos says having people not make enough payments to reduce the interest owed is an extreme case.

Jones continues pushing on the question, asking why the Department now feels like all tests need to be passed versus how it approached the rule in 2011. Kolotos notes that part of it is the Department is no longer targeting the worst-performing programs. It is targeting those that are doing a subpar job at providing gainful employment. He also notes that the tests are less about judging the quality of a program and more about the performance of loan debt. [Edit: Removed struck language because of concerns about clarity of the point, which was that these measures are supposed to be considerations of providing gainful employment, not measures of quality.]

Jerome says he’s concerned that some of the metrics may be dealing with assumptions about how many programs might fail. He suggests he’d like to talk about completion rates.

Barmak Nassirian from the American Association of State Colleges and Universities says he at first didn’t like the repayment rate, but then saw it was at least simple and thinks the concept is workable. But he’s concerned the threshold is so low that cohort after cohort of interest-only payments would be acceptable. He also talks about the two debt-to-earnings tests and asks why the two tests are separated. He argues making programs pass both tests. Editorial note: having looked at the effects of requiring passage of both, you are better off just having the discretionary debt-to-earnings test, since almost no one who passes that test fails the annual debt-to-earnings test.

Jerome challenges Nassirian asking if he knows what would happen if the metrics were applied to his AASCU members. Nassirian notes that there’s a statutory limitation that means they are not considered.

Margaret Reiter, who represents consumer advocacy organizations, asks about the thresholds chosen and their justification on the new measures. She asks whether the program-level cohort default rate would be as open to manipulation as they are at the institutional level. Kolotos says the idea for the program-level cohort default rate is to apply the same logic as what’s applied to the institutional level. He says he does not want to get into the issue of manipulation here. On the negative amortization rate, he acknowledges that it is a low standard. Kolotos says he is open to suggestions for a repayment rate threshold, says he is open to ideas on how to set a threshold because it is hard for them to come up with one.

Sandra Kinney from the Louisiana Community and Technical College System asks if there’s a minimum number of students required in the repayment rate cohort, since she notes that 80 percent of her students don’t borrow. Kolotos notes that the program-level cohort default rate would have the same appeals options as exist in the institutional level cohort default rate. This means that programs could in theory challenge the rate based upon what’s called a participation rate index, which looks at the percentage of borrowers as a share of those enrolled multiplied by the default rate.

Jones from Strayer asks if the Department has done data analysis on the validity of the threshold size for the program-level cohort default rate.

Rory O’Sullivan from Young Invincibles notes that the thresholds catch extremes. He describes how a program that negatively amortizes for a few years could then pay down a small amount and still pass with even more debt than they owed at first.

Heath from Anne Arundel Community College notes that the last time the Department did the rule it said very few programs at community colleges would be impacted. He notes that part of that was because if the schools reported tuition and fee levels then the Department would only look at loan debt for tuition and fees. He’s concerned about not having the same option because schools cannot reduce loan debt on a programmatic basis, only on a case-by-case basis. He says they are being tasked with using a performance measure that they have no control over on the front end. He says community colleges would not be in favor of no longer excluding loan debt for non-tuition and fee debt.

Raymond Testa from Empire Education Group brings up the issue of how the new metrics would only look at those who received federal student aid. He calls for the Department to include anyone who filled out the Free Application for Federal Student Aid (FAFSA), regardless of whether they received student aid or not. Testa notes that excluding non-Title IV recipients makes the cohort tougher, the debt-to-earnings tests got tougher, new metrics got added that were tougher. Testa says it is “ludicrous” to negotiate a metric without knowing the impact. Testa notes that if a few higher-balance borrowers are negatively amortizing, they could lead to a program failing the loan repayment test even if a lot of students are mostly repaying. Testa also questions how the Department is pushing everyone into income-based repayment, but that’s what causes negative amortization so it could lead to programs failing. Note: We are taking a look at this issue at New America, but generally, a borrower on income-based repayment negatively amortizes if their earnings are lower than the low $20,000s, though it depends on family size and debt amounts.

Kolotos notes the court restrictions on the types of students they can look at, said they are open to ideas. He again says that he wants to talk about concrete ideas and for the negotiators to stay on topic.

Della Justice from the Kentucky Attorney General’s office objects to the idea of saying they can’t negotiate on something without data, since the last rule ran into the problem of impact determining the thresholds.

10:20 a.m. Still discussing metrics
Nassirian suggests maybe instead judging programs on whether they are amortizing over 30 years, since at least that gets at the idea that the portfolio would get paid off at some point. But the threshold used right now would allow a program to make interest only payments and still pass. He thinks negative amortization is not the correct landing place.

Jerome says just because a loan portfolio negatively amortizes in one year doesn’t mean it negatively amortizes over the length of repayment. He argues that the loan repayment performance will reflect the demographics of the students, not the quality of the program. Nassirian responds by asking how many years after finishing should students continue to borrow from the government, which is what he says negative amortization is, before it should be expected to be retired?

One negotiator (I believe Margaret Reiter) asks whether the Department could let schools voluntarily choose to report information on non-Title IV students if they desired, either as an appeal or as part of an upfront process. She also asks whether earnings below the poverty level should be a sufficient measure of gainful employment. She says that deals with the discretionary income measure, since you know if earnings fall below the poverty level it would fail. For the repayment rate, she suggests looking at the amortization schedule on a 10-year plan. Says if it would normally bring it down, say 8 percent, then they should pick some amount it has to come down between 0 percent (the negative amortization test) and 8 percent (the 10-year amortization test).

Jones from Strayer returns to his earlier point about concerns around the Department’s new justification for a rule. He notes that the tone and objectives appear to have changed, which makes him very concerned. He notes a study from the National Center for Education Statistics study that showed a large number of students have debt burdens over 12 percent. He also raises the question of how would this rule work if it was applied to all institutions, not just gainful employment ones? He says he sees a “fundamental shift” in how this rule is being put together and does not see a justification yet for why this change occurred.

Steve Finley, an attorney at the Department, responds to Jones. He says the first proposal used repayment rate as a way to say it was good enough to maintain eligible. But there were changes made to remove things that the Department thought distorted results. It set a higher standard. He says what is here is two floors. What it means you can go through the door and walk on the floor. That’s why he says it’s not an alternative test. It’s just a measure of it is good enough to just be in the door. He says the same thing is true with the program-level cohort default rate.

11 a.m. Back from the Break
Kolotos responds to Barmak about the debt-to-earnings rates idea of holding programs accountable for both measures. He says ideally ED would have programs be judged under the discretionary rate because it’s closer to what’s used in Income-Based Repayment programs, but actually a bit more lenient. But the reason for the annual rate is because if earnings are less than 150 percent of discretionary earnings you can’t calculate the rate. He says there is some merit for a program to have low earnings but also low debt. He says if you want an absolute floor on earnings that needs to get discussed.

Nassirian says that there must be some level of income so low that any level of debt service is not acceptable. He suggests testing programs on both measures, but modulate the percentages to pass. Though he does not say what those percentages should be.

Jerome says that his culinary graduates earn between $17,000 or so a year, but fails discretionary because grads have $7,000 to $8,000 in debt. He thinks the idea of judging programs on either debt-to-earnings measure but not both. Reiter asks if graduates are still earning low salaries three years out. Jerome raises a concern that since the measures look at people three years after entering repayment, that could mean the earnings are measured starting as soon as 18 months after leaving school. Kolotos confirms that the income of students three years after finishing represents earnings starting after 18 months (and measured through month 30) and the others start at two-and-a-half years out.

The question of the ability to limit loan debt comes up again. Kolotos notes again that ED lacks the ability to change that statutory requirement on loan availability.

11:15 a.m. Minimum Program sizes
Rhonda Mohr from the California Community College Chancellor’s Office asks about the minimum number of students for the measures. Kolotos notes that the way the cohort default rates work a college has to have at least 30 borrowers over a three-year period for the rates to have accountability consequences.

Mohr also asks if there is a minimum program number of borrowers for the student loan repayment rate. Kolotos says ED has not proposed one yet.

Jenny Ricakard from the University of Puget Sound asks about the proposal the repayment rate group put forward, which would have absolved programs that had high repayment rates. Kolotos says that idea was not workable because of difficulties with setting thresholds, though he did say it had merit and they would be open to thresholds.

Testa from the Empire Education Group objects to the idea that a training program that takes someone who was unemployed or unemployable and puts them into a position where they earn $10 an hour and move toward earning a business. He notes that most students at his schools are female and seek part-time work at the beginning of their career. He suggests that if they can’t define gainful employment here maybe that’s something Congress needs to do.

O’Sullivan brings up the idea of maybe adjusting the family size for the discretionary earnings measure, which would make it harder to pass that rate, since the amount of income deducted for expenses would be about $4,000 higher.

11:35 a.m. Metrics Formulas
The only major change in the debt-to-earnings rate is to take the lowest interest rate during the four-year period. Jeff Baker from Federal Student Aid notes it would be 3.86 percent because that’s the current rate that’s tied to May’s Treasury Bills.

Reiter asks whether the interest rate for graduate programs should be tied to the interest rate for graduate students, which is higher than what undergraduate students pay. She also asks why it wouldn’t be fairer to take something rather than the lowest level, since many students are not in four-year programs, but are in one-year ones. She says it doesn’t have to be the highest, but could be the average.

Jerome distributes a handout that raises his concern that the median loan debt might not be a fair measure because they automatically pass the debt-to-earnings measure, though he does note in response to questions from other negotiators that a program with borrowers would still have to pass the loan portfolio repayment rate and the program-level cohort default rate.

Testa also endorses the idea of using the mean debt instead of the median debt due to concerns about programs having borrowers but being exempt because fewer than half borrowed. Editorial note: unstated in this discussion is that the community colleges care strongly about the median debt because many may have a few borrowers but not enough to have a median debt of more than $0.

Kolotos says ED will take back this idea to discuss whether it should use only a mean loan debt and whether it should use something instead of the lowest interest rate.

Jones from Strayer brings up the question of how long borrowers should be expected to pay back loans. This is something that was debated in the first session. Changing the amount of time someone is expected to repay matters because a longer time to repay means smaller monthly debt payments and a lower amount of income to pass a debt-to-income test.

Nassirian asks if two programs really are the same if they have the same amount of income but fewer borrowers? He also asks why is the Department not using the actual interest rates for borrowers? Jeff Baker from the Federal Student Aid office says it would be very complex to calculate the individual rates because it would have to calculate the payment amount loan by loan, versus aggregating up by the cohort.

Jones calls for differential amortization rates by different program levels.

Lunch break. Back at 1p.m.

1 p.m. Loan Debt Calculations
Ray Testa from Empire Education Group kicks off the afternoon session with a discussion of how long a student has to be enrolled to be excluded from a cohort in order to not have their earnings counted in a debt-to-earnings rate. He asks for exclusions to take place if someone is enrolled in a program or in military status for 30 days, instead of 60, to be excluded.

Margaret Reiter, who represents consumer advocacy organizations, asks about how ED would enforce figures on private student loan debt. Kolotos from ED says it has provisions around program review to catch things like that, but not sure what else could be done.

A community college negotiator asks about whether someone who gets a certificate and then an associate degree that’s not counted in the rules if the student would show up in the calculations. Kolotos says if the person is still enrolled in the associate degree program at the time the calculation is done, then they would be excluded due to being in school. If they finished the associate degree at the community college, they wouldn’t be counted at all because the program itself is not subject to the rules.

1:25 p.m. Completion rates?

Jerome asks how many programs have fewer than 10 graduates in a year due to low completion rates and thus aren’t subject to the rules. Kolotos says they have to have 10 in order to get data from the Social Security Administration. He notes the prior rule aggregated students over multiple years in order to capture programs that may have had fewer completers in a given year.

Reiter asks if there should be a completion rate in the debt-to-earnings that makes a program automatically fail if they have too low a percentage of Title IV recipients complete. She does not provide a threshold. Kolotos says ED cannot figure out this information because it does not currently have program-level completion rates for gainful employment programs.

Nassirian says he thinks what Jerome is saying is reasonable, suggests saying if it has fewer than 10 graduates it has to have a graduation rate of X. Jerome suggests you could take institutional completion rates and assume they are stable over programs. Reiter notes that ED will receive data to calculate completion rates, but Kolotos says they cannot generate a threshold.

Sandra Mohr from the Louisiana Community and Technical College System notes that holding programs with small numbers of graduates accountable for their completion rates would magnify problems since it would be dealing with small numbers of students.

1:45 p.m. Transition Periods
ED is proposing that for the first four years of debt-to-earnings rates a program could substitute current loan debt instead of the loan debt of the students being evaluated. This is one year longer than what was in the prior suggested text. Kolotos says this is designed to provide opportunities for programs that want to take immediate action to improve.

Nassirian disagrees with this idea, says that the co-mingled of current debts with past ones does not make sense to him. He says that since the thresholds are so low, why not let the chips fall where they may? Kolotos says this is how ED wanted to build in opportunities for improvement, but Nassirian argues it feels manipulated. He says instead, programs should be allowed to make payments to reduce the debt of students to the point where indebtedness goes down.

Note: the Department handled this issue the last time around by allowing programs to be judged based upon the debt of more recent students, but also the earnings of those newly minted graduates as well.

2 p.m. Program Cohort Default Rates
Jones from Strayer asks if ED has done an analysis to see if the minimum borrower size of 30 that is used for institutions still is valid if used for programs. One of ED’s lawyers notes that the cohort default rate process already has appeals in place to deal with volatility, which include things like low borrowing incidence.

2:10 p.m. Loan Portfolio Repayment Rate
This measure would look at the beginning and ending balance to see if students in the year being evaluated made enough payments to reduce the principal balance owed on their loans.

Jerome asks ED to clarify its rationale for why it chose this repayment rate calculation as a sign of program results/quality. He also produces a handout. He argues that the repayment rate is a reflection of just the demographics of borrowers. Copies of his handout aren’t available, but his description indicates that its data from Mark Schneider, which shows the repayment rate by quartile for institutions based upon their percentage of students the are from minority groups.

Helga Greenfield, who is from Spelman College and represents Historically Black Collges Universities, also agrees with the idea that the demographics are an important part of this. Testa from Empire Education Group also agrees.

2:50 p.m. Continued metrics discussion
Belle Wheelan who missed the morning session to appear at New America’s event on competency-based education at community colleges (which you should totally check out), asks why ED did not take her suggestion for having a year of data that’s solely for informational purposes. This is a request that for-profit negotiators had also requested. Kolotos says that ED extended the transition period instead to four years as a way of making that concession.

She also returns to the discussion of race that carried on before the break. This is an issue that received significant attention in the last iteration of the rule, including a regression analysis and scatterpot presentation. Finding those exceeds the bounds of what the Internet here can entail, but you can find them

Heath from Anne Arundel CC returns to the question of what to do about students who already have a lot of loan debt from a prior credential, such as a bachelor’s degree from elsewhere, who take on more loan debt, giving them a total combined balance that puts them at risk of default. And so even though only the debt from the gainful employment program is judged, the total loan balance accumulated from multiple places causes the negative result. Kolotos says Heath is correct and that someone who did default because of that high cumulative amount would count as a defaulted borrower.

3:05 p.m. Loan Portfolio Repayment
Nassirian asks about when interest is capitalized on a loan. This matters because when interest is added to the principal balance of a loan could affect how high the cumulative principal balance becomes.

ED notes there are a few instances. If a student does not pay interest on their unsubsidized loans while in college, then it capitalizes when they enter repayment. Similarly, any interest on an unsubsidized loan that accrues while it is in deferment or forbearance is capitalized when it exits that status.

3:15 p.m. Notifications and Consequences
Kolotos clarifies in response to a question that the enrollment limits for failing programs applies only to students receiving federal student aid, not all students overall.

Mohr asks if programs would get three separate notifications of results or a single one. Kolotos says the debt-to-earnings and repayment rates would be at the same time and the cohort default rate would likely be on a different schedule. The schools do not like the multiple announcements. Institutions would, however, get their program and institutional cohort default rates at the same time.

Kinney raises a common concern from the day–when will data be available on the new ideas. Kolotos reiterates the Department will provide the data when it has it. Jerome asks if the Department would host another negotiation session with data. Finley from ED says there are no plans for a third session, but he thinks the merits of the proposals are on the table for the discussion. Jerome says it’s hard to vote on a regulation without any data. He indicates he wants some data before negotiators are formally asked to vote on something.

Kolotos explains the middle performance zone concept a bit more. He says the idea is that they don’t want to inhibit programs from being able to improve, so the Department would not ask programs in this category to do anything for the first or second year. But it’s once the program is at risk of losing eligibility sooner is when steps are taken. He notes this is similar to how financial responsibility scores work.

Heath asks about the lead time between when a program learns it loses eligibility and when that eligibility loss actually happens. He notes that without that, a program could lose eligibility in the middle of a semester, which would cause problems.

Nassirian says there need to be meaningful intermediate mechanisms for programs that reach the zone so that they cannot get a lot worse before losing eligibility. He cites ideas like preventing the award of dividends, cap enrollment, among other ideas. This would also include immediate notice to students with a warning about results. He later adds that maybe having institutions make co-payments to the cohort that failed one of the tests to make the debt service amount work.

Jerome asks what programs would do to improve the loan portfolio repayment rate in a year before it lost eligibility. Kolotos says it would be open to consideration. He notes that last time programs could calculate their results based upon the most recent cohort to enter repayment as a way of testing activities that may help improve results, like better career counseling. Jerome asks for institution-wide loan repayment rates.

There’s been a lot of discussion about a recent National Center for Education Statistics study called “Degrees of Debt.” What it looked at was bachelor’s degree recipients one-year after completion, not students in associate or certificate programs. It also looked only at bachelor’s degree recipients who borrowed, which is a bit different from the gainful employment calculation, which would look at all students who received Title IV aid (i.e. a Pell student who never borrows is included in gainful but not in the NCES study). What it found was that of those students who received a bachelor’s degree and borrowed, and whose loans entered repayment one year later, 31 percent had monthly payments above 12 percent of their income. So it’s different group in a different time period and not directly comparable.

Jerome and Jones both indicate frustration with how things are going, raising a concern heard before about how they have programs nearly identical to those offered at other types of institutions that are subject to this rule whereas their counterparts do not have to meet these standards.

4:05 p.m. Restrictions
The committee moves on to a discussion of how a program that loses eligibility has to wait three years before rejoining the aid programs. Programs that were failing and shut down voluntarily also have to wait three years. Kolotos notes this is the same framework as the 2011 version of the gainful employment rule.

Nassirian raises a concern that a program could shut down once it finds out that it is likely to fail based upon the draft set of rates and before those figures are finalized. He’s worried about this because it might allow a program to shut down and then restart itself without the required waiting period. Kolotos says that ED could think about possibly still finalizing measures in those cases.

Wheelan raises concerns that this regulation will create work for accreditors, because programs that shut down must do a teach out plan and would potentially have to go to the accreditor to restart the program.

Della Justice and Nassirian raise concerns that the enrollment limit still allows for new students to enter, just not so many new students that it doesn’t exceed the total number that were there before.

Jerome asks if Nassirian’s concerns about results would look at the completion rates as well. He points out schools near him that have graduation rates in the low single digits of percentages. Nassirian says his outrage is on dropouts being saddled with debt, not just high rates of non-completion. He thinks a low completion rate does not say as much about its gainful employment performance as does looking at whether institutions leave students worse off or not. But he notes that’s a philosophical answer.

Kolotos asks what should be done to a program that’s a year away from failure since negotiators don’t seem to like what ED put forward. Nassirian says limit eligibility to only currently enrolled students. He says there should be some kind of clawback mechanism to get money from schools. Reiter says she thinks penalties should apply to programs that are one year away from losing eligibility because they are in the zone as well as those that are about to lose eligibility because they are failing.

4:35 p.m. Zero median debt
Community college negotiators ask about a proposal they had put forward that would have exempted a program that had a median loan debt of $0 from being subject to a debt-to-earnings measure. The negotiators say this idea was in the final rule. Mathematically, a program where less than half of students borrow have a median debt of $0, so they couldn’t fail a debt-to-earnigns test. This idea would just exempt them from the measures and presumably ease some concerns about failure. The one issue is that a program with a median debt of $0 would still possibly need to pass the repayment rate or cohort default rate test, which are not based on average amounts. Heath clarifies that the goal is to not scare colleges out of the loan programs and by telling programs upfront they can’t fail would help them avoid those fears.

Jerome asks what about a program with low percentages of borrowers but high numbers of borrowers.

Nassirian says he likes the idea of an exemption for low borrowing.

That’s a wrap. Back tomorrow at 9 a.m.