Education Secretary Betsy DeVos on Tuesday announced a second attempt at offering students seeking loan forgiveness the opportunity for partial debt relief, introducing a new statistical method comparing the median earnings of graduates with borrower defense claims to the median earnings of graduates in comparable programs. Some lawmakers and higher education experts, though, were quick to denounce the new policy and highlight what appeared to be faulty math.
This news comes as DeVos is set to testify before the House education committee on Thursday about her management of borrower defense claims — more than 200,000 of which are currently pending.
DeVos was blocked in May 2018 from pursuing her original plans for partial debt relief, which involved comparing students’ current earnings with those of their peers from passing gainful employment (GE) programs. Under that plan, students would receive full relief only if their earnings were currently less than 50% of their counterparts’ from passing GE programs, and students earning at least 50% of what their peers earn from programs that pass GE standards would be compensated proportionally for the difference.
While a judge decided against an argument brought by a group of former Corinthian Colleges borrowers who claimed the partial relief violated their right to due process and that they deserved full relief, she did stop ED from pursuing its debt cancellation policy for violating the Privacy Act.
Specifically, ED was flagged for sending a list of names, dates of birth, and Social Security numbers of borrowers who previously filed attestation forms for Corinthian to the Social Security Administration (SSA). In return, SSA sent ED the mean and median annual earnings of the students "in aggregate form, without any personal identifying information," according to court documents.
DeVos announced Tuesday that while ED appealed the court’s decision, “when it became obvious that the court was not going to move quickly, [ED] began to develop the new methodology released today for adjudicating claims.” DeVos said her new plan instead uses “publicly available data” to compare programs by looking at the median earnings of graduates with borrower defense claims and those in similar programs.
According to the new method, borrower defense applicants would only receive full or partial relief if the median earnings of graduates from their school were shown to be lower than the median earnings of graduates of similar programs across all institutions. Applicants would receive full relief if those earnings were lower than two standard deviations from the mean of a comparable program, and either 25%, 50% or 75% relief if their program’s earnings fell between the median and two standard deviations less.
DeVos added that ED will be awarding no less than 10% of relief to former Cornithian students regardless of their earnings “because of promises made by the prior administration, and damages likely caused to those borrowers by ED’s continuous efforts to use Corinthian Colleges, Inc. institutions as an example.”
“Despite the mess we inherited from the previous administration, we committed from day one to getting this right for students and taxpayers,” DeVos said. “We cannot tolerate fraud in higher education, nor can we tolerate furiously giving away taxpayer money to those who have submitted a false claim or aren’t eligible for relief.”
While DeVos added that her new method “treats students fairly and ensures that taxpayers who did not go to college or who faithfully paid off their student loans do not shoulder student loan costs for those who didn’t suffer harm,” many were quick to denounce the new calculations.
Sen. Patty Murray (D-Wash.), ranking member of the Senate Health, Education, Labor, and Pensions (HELP) Committee, argued that the new method “uses faulty math to justify denying borrowers the relief they are owed” and that borrowers should receive relief in full.
“While this step is deeply concerning, unfortunately, it’s not surprising — because it is one more in a long list of examples of Secretary DeVos skirting her legal responsibilities to protect students and borrowers,” she said.
ED press secretary Angela Morabito told NPR in a statement that full relief “sounds nice, but it really means 'full liability' for taxpayers — and that's not fair in cases where a borrower is not entitled to it.”
“That's why it's imperative that we consider each borrower defense claim individually. To force taxpayers to provide blanket forgiveness would be abandoning our duty to be good stewards of tax dollars,” she told the outlet.
However, not everyone at ED is in support in partial debt relief. NPR also released memos Wednesday from the career staff working with borrower defense claims in which they recommend full relief for Corinthian and ITT Educational Services, Inc. students in 2017, just before DeVos was sworn in.
Others were concerned about ED’s use of statistics.
A chart ED included in its press release showed it was assuming that 95% of the earnings for a program would fall within two standard deviations of the median earnings for that program, with the intention of targeting relief to outliers with lower incomes. However, Ben Miller from the Center for American Progress (CAP) said that income does not fall into a bell curve such as ED suggested, which results in large gaps between standard deviations and only extremely low or negative earnings meeting the threshold for full debt relief.
One obvious flaw with the borrower defense partial relief formula is its completely blind to how the minimum wage puts a lower bound on earnings. Consider a certificate in accounting technology. ED's median comparison earnings are $19,996 and two standard deviations is $17,566.— Ben Miller (@EduBenM) December 11, 2019
“It’s so obviously flawed. I don’t see how you can look at something that says that full relief requires negative earnings and not say you screwed up,” Miller said in an interview. “[ED] mathematically stacked the deck against them.”
Douglas Webber, an economics professor at Temple University, wrote on Twitter that he was concerned with the overall use of standard deviations, arguing that “the use of standard deviations and medians is probably not appropriate” because standard deviations are defined as relative to the mean of a figure, not the median.
“In a right-skewed distribution (which earnings almost always are), there will mechanically be very few cases that 'pass' this test,” he wrote.
Miller said another issue is that ED is using data that “may not even tell us anything about the people who actually filed the claims.” Because ED is looking at earnings of graduates, he said, it does not take into account the many students who filed claims who were forced to stop their education.
Additionally, according to ED’s calculations, even if a borrower seeking loan forgiveness is personally earning below the median of a comparable program, the borrower would not be taken into account if their program’s median earnings were above the median.
ED wrote that this week it will be releasing the results of hundreds of claims — filed mostly by Corinthian Colleges and ITT students — that were decided using this new methodology.
Publication Date: 12/12/2019