For years, advertisements for ITT Tech filled American airwaves. The company operated a nationwide chain of for-profit career schools. This fall, it was forced to close in the face of new regulations from the Department of Education. But new information shows that one tool used to justify requirements for for-profit schools relied upon false statistics.

In 2014, the Department of Education released a website called the College Scorecard, which aimed to help high school students compare education options by providing information on cost and post-graduation career prospects for thousands of schools.

At the time, the DOE billed the College Scorecard as “the clearest, most accessible, and reliable national data on college cost, graduation, debt, and post-college earnings.” The site compiled data on more than 7,000 schools over a period of 18 years and supplemented information reported by the various schools with tax records and Treasury department statistics to help users distinguish schools on the economic outcomes of graduates.

“Americans will now have access to reliable data on every institution of higher education,” said President Obama as he introduced the scorecard during his weekly radio address.

“You’ll be able to see how much each school’s graduates earn, how much debt they graduate with, and what percentage of a school’s students can pay back their loans – which will help all of us see which schools do the best job of preparing America for success,” he said.

When the site was launched, the DOE believed that much of the value of the Scorecard was the dataset itself. By “giving developers access to [the data], even more customized tools will be created, providing students more options than ever before to find the right school for them,” wrote Lisa Gelobter, then the DOE’s chief U.S. digital service officer, in September, 2015.

One of the main focuses of the site was student loan repayment, used as a proxy for graduates’ financial success. The decision to highlight the student loan repayment rates as a measure of student success was novel.

“At the time, many observers believed that the new regulation was an attempt to hamstring for-profit education option and to push students towards community colleges,” wrote Michael Stratford soon after the site launched. He noted that for-profit colleges and historically black schools fared particularly poorly under this metric.

However, this data-driven focus in the end hurt schools when it was discovered nearly two years later that a “coding error” had sunk the repayment rates of many for-profit schools, making them appear to be comparatively worse options than community colleges and four-year schools.

“An error in the original college scorecard coding to calculate repayment rates led to the undercounting of some borrowers who had not reduced their loan balances by at least one dollar, and therefore inflated repayment rates for most institutions,” said Lynn Mahaffie, then acting assistant secretary of postsecondary education on January 13, 2017.

Even though the site had been displaying incorrect information since its launch, the DOE tried to downplay the scope of the error by calling the relative difference “modest.”

“Over 90 percent of institutions on the College Scorecard tool did not change categories (i.e., above, about, or below average) from the previously published rates,” said Mahaffie.

However, the release of the data shows that the results of the error were anything but small. After recalculating loan repayment rates using the correct dataset, “three-year repayment rates fell from 61 percent to 41 percent, five-year repayment rates fell from 61 percent to 47 percent, and seven-year repayment rates fell from 66 percent to 57 percent.” More significantly, the DOE uses this data to categorize schools as above, at, or below average in terms of repayment rates. Schools with low repayment rates are subject to additional Federal scrutiny.

In October 2016, a new regulation from the DOE required schools with loan repayment rates of below 50 percent to post warnings for potential students, cautioning them that most former students are not paying down their debt. The rule also changed regulations to allow borrowers who claim their school defrauded them to discharge their debt.

The “coding error” and resulting shift in statistics was sufficient to force many for-profit schools to post such a warning. However, the updated data shows that many two-year colleges and historically black schools also should have needed to post warnings, but were not required to.

“The department said it would not be fair to ‘burden’ public and nonprofit colleges with a regulation that would apply to so few,” wrote the Wall Street Journal. “Yet based on the updated data, 60 percent of two-year public colleges and nearly all historically black institutions have repayment rates below 50 percent.”

While the Scorecard now displays accurate statistics, for months it provided students with an incomplete picture of the financial position of graduates. Several large chains of for-profit schools, including ITT Tech, closed in 2016 under the weight of additional regulations, including requirements based off of the College Scorecard data. The quiet alteration of the data is already making some call for a new look at federal student loans.

“This change is likely to get a lot of discussion in coming days,” said Robert Kelchin, an assistant professor of higher education at Seton Hall University, “particularly as the new Congress and the incoming Trump administration get ready to consider potential changes to the federal student loan system.”

Follow Erin on Twitter.