A national report that accused the University of Minnesota of overspending on administration was withdrawn and modified after its authors admitted using flawed data.

The Institute for Policy Studies, which issued the report Sunday, took it down from its website Monday after the U publicly challenged its findings. The report was reposted late Wednesday, with a new set of numbers.

The report, called "The One Percent at State U," singled out the U as one of the five "worst overall offenders" among public universities. It accused the schools of spending lavishly on executive pay and administrative staff, while cutting student aid and hiring low-paid instructors.

On Monday, the institute learned that some of the data used in the analysis had some errors, said Marjorie Wood, one of the authors. At that point, she said, the group "took the report down temporarily" and began a review.

University of Minnesota President Eric Kaler, who has called the report "dead wrong," said Wednesday that the group simply misread the public data. "If you don't know what you're doing, you can easily get the wrong answer," he said.

The U posted a detailed rebuttal on its website, saying: "We consider this report extremely flawed because it uses inaccurate data and presents a misleading picture of the University."

Kaler

The revised report, released late Wednesday, wasn't likely to appease U officials. It now asserts that between 2007 and 2012, the U's Twin Cities campus "increased spending on nonacademic administration by 44 [percent] while decreasing spending on scholarships by 55 percent." It also claims that the number of "adjunct and contingent faculty" increased by 825 or 105 percent. The original report claimed that the U had cut scholarships by 36 percent, and more than tripled adjunct faculty and administrative staff.

'Got that wrong'

The university disputes all those numbers, and says its assertions are backed up by federal data. It points to statistics collected by the Department of Education, which show, for example, that the U's spending on scholarships grew by 49 percent from 2006 to 2012, from $151 million to $225 million.

"It's not like my opinion; these are facts," Kaler said. "That's just nose-on-your-face plain. They just got that wrong."

He said the authors also jumped to the wrong conclusion about administrative expenses and adjunct instructors by using the wrong data. In some cases, he said, the only thing that changed was how the jobs were reported. "They were already employed by the U," he said. "If you look and say, 'Omigod,' it's because you don't understand the data."

The report touched on a sensitive issue at the U, which has been fending off criticism about administrative bloat for several years. Last fall, Kaler announced a six-year plan to cut administrative costs by $90 million, and has hired consultants to confront the issue.

The institute, a nonprofit research group in Washington, D.C., said it relied in part on data from the American Federation of Teachers (AFT), rather than the federal database, to reach its conclusions. When the U challenged its accuracy Wood said, the AFT double-checked its numbers and concluded "they may have found some issues."

Author: Overall findings OK

Andrew Erwin, another author, said some numbers were attributed to the wrong years, but that he did not believe the errors undercut the report's overall findings.

But in an interview Wednesday, neither he nor Wood would respond to the University's specific charges about inaccuracies in the report. "We're not making statements," said Wood. "The report will speak for itself." This is the first time the institute has issued a report about spending priorities at public universities.

The updated report, which was posted on its website, said that the review "did not alter the report's major findings," but that it resulted in a number of changes to its list of "worst overall offenders": "The University of Minnesota dropped from #3 to #4; the University of Michigan rose from #4 to #3; and the University of Delaware became #5. All data related to president pay and student debt levels were unaffected."

The list was based on an analysis of "excessive executive pay, highest student debt and large increases in low-wage and/or contingent faculty labor."

Kaler said the problems with the data go far deeper than the authors suggest. "It looks to me like much more than dates are confused," he said. "We're not real confident in their ability to manage this data."

The U was not the only one to challenge the report. A spokesman for the University of Michigan told the Detroit News that the report "is nearly 100 percent wrong."

Maura Lerner • 612-673-7384