On the Record
Our Numbers Good, Your Numbers Bad?
Disagree? You're full of beans...
High Level Discussion at University of Minnesota
Finance and Planning Committee
Senate Committee on Finance and Planning
Tuesday, April 5, 2011
2:00 - 4:00
238A Morrill Hall
1. Inaccurate and Misleading Data
Professor Luepker convened the meeting at 2:00, welcomed Professor Cohen, who is serving while Professor Chambers is on sabbatical, and turned to Dr. Radcliffe to discuss data drawn from the Delta Project and recently published in The Chronicle of Higher Education. Those data have received a lot of attention, he noted, but many who have examined it say it is not correctly done.
The data published in The Chronicle came in an article titled "Education Financing for Major Public Universities: Which Ones Get the Most and the Least," and purported to report the "subsidy for education and related expenses per student in 2008." The figure for each institution is said to be "the share of educational spending not covered by tuition," and "for many but not all flagships, state appropriations finance a significant portion of this subsidy." UNC-Chapel ranked first on the list, at a $26,373 subsidy for education and related expenses per student; the U of Washington was second, at $19,575, UC Berkeley was third, $16,165, and so on; the University of Minnesota-Twin Cities was eighth at $13,616.
Dr. Radcliffe said that he has examined the methodology used by the Delta Project; it is not reported as clearly as one might wish. They use data similar to that reported by colleges and universities to the federal government (Integrated Postsecondary Education Data System, or IPEDS), which is good for high-level aggregate analysis. People use the data, however, to answer questions the data are not organized to answer. The data for any institution, for example, will depend on the mix of programs it offers.
Professor Luepker said he did a rough calculation when he first saw these data: He divided the state appropriation by the number of students on the campus and came up with a number similar to that in The Chronicle. It sounds like a simplistic approach.
Vice President Pfutzenreuter said his comments about the data in the article were less kind. He noted that the data are for "education and related expenses" per student. He also pointed out a second, later article that appeared in The Chronicle that explains why the data in the earlier article are not accurate. He said he believes that the Delta Project has an agenda--and one that is not friendly to higher education.
To get to the $13,616 figure for the Twin Cities campus, Mr. Pfutzenreuter explained, "education and related expenses" are defined as a proxy for the full cost of education, which includes direct expenditures plus an estimate of support. "Education and related expenses" is not a common term in higher education. The Delta Project pulls data from IPEDS and manipulates them (in a way that is not clearly explained) to arrive at their definition of "education and related expenses," take net tuition (which is not clearly defined) and subtract it from "education and related expenses," to get a subsidy figure. The graph in the report is titled "A snapshot of state subsidy patterns for education and related expenses--public research sector," but The Chronicle article points out that the "subsidy figure is not necessarily finance by state appropriations at all flagships." (The author of the second article points out that for the University of Washington, the Delta Project reported a subsidy of $19,575--but the state-funded subsidy per student in 2008 was $9,797 per student, which put Washington at 28th among the 50 flagship institutions.) But as a result of the publication of these data, Mr. Pfutzenreuter concluded, people (including legislators) think the University is well-subsidized. There are people passing the Delta Project numbers around to claim that the University is over-funded.
Professor Martin noted that the number for the University of Michigan is very close to that for Minnesota, $13,309, and anyone who knows anything about the financing of the University of Michigan knows that number is ridiculous if it is claimed it represents state support. Mr. Pfutzenreuter said that the University of Minnesota spends a lot of non-state money to support the educational mission. There is no reason to pay attention to these numbers, he said; they are just bad numbers.
Professor Roe said the University needs to put together its own numbers. If one wants to determine state support for education, Mr. Pfutzenreuter said, one could divide the state support of the University by full-year-equivalent students, but the University's mission is research, public service, and education, so many of the state funds are not for education.
Professor Cohen noted that both SUNY-Buffalo and Nevada at Reno have subsidy amounts greater than Minnesota, and those are not institutions with which the University competes, so it may be hopeless to try to clarify the fallacies in the numbers. If the University puts out good numbers, it looks self-serving. Is the AAU or any other national organization preparing good data? They are not, Dr. Radcliffe said, and it is difficult to assemble comparable data.
Mr. Pfutzenreuter said they struggle with how to respond to the use of bad numbers. The University has not done much about responding to them and he asked whether it should be more aggressive in doing so. There are those who write blogs and commentary, and what they write appears to be the truth, but there is no fact-checking so they define the numbers. Mr. Driscoll asked when Mr. Pfutzenreuter's office might be able to respond; not until after the budget is completed. Mr. Pfutzenreuter responded that his office is working on identifying what pays for research, education, public service, financial aid, and so on, based upon the attribution of both direct and indirect costs, in order to determine the "fully-loaded" cost of instruction and other mission activities.
Professor Zaheer suggested it would be useful to have comparable numbers from peer institutions. Mr. Pfutzenreuter said he did not know of any other institution that was calculating a fully-loaded cost of instruction, so it would not be possible to compare data. Professor Martin asked about the CIC. Dr. Radcliffe said that all of the institutions use IPEDS data, which are problematic for these kinds of analyses. The University is able to do the analysis because revenues are coded and because of its budget model, so there is a reasonable methodology for the attribution of indirect costs, and they will do the analyses by college. Professor Roe commented that the University should be doing these kinds of analyses anyway so that it knows more about its cost drivers.
The Delta Project data, Mr. Pfutzenreuter said, represent the usual "garbage in, garbage out" phenomenon. They particularly disadvantage research universities, Dr. Radcliffe said, because they massage IPEDS data to reach the conclusion they wish. If institutions code data consistently, what goes into IPEDS is good, but it is not clear that institutions code consistently. Ms. Patil asked if other institutions have responded. The second Chronicle article was a response, Mr. Pfutzenreuter said. Dr. Radcliffe said he has listed "providing accurate and comparable data on administrative costs" as one of the top concerns for the roundtable discussion at next month's meeting of the AAU Data Exchange.
Professor Luepker observed that "things happen fast," and by the time the University gets numbers out, the damage has been done. And when data like these come out, they are damaging--they achieve currency as reality. The University responds four months later. University Relations is in transition, but the University needs a system to respond. While it is not within this Committee's purview to respond, it can ask questions about how the newly-appointed Chief of Staff proposes to deal with the need for a response. Professor Roe suggested it would be useful for the University to have a website with data that people can look at.
Professor Durfee commented, apropos of the "education and related expenses" developed by the Delta Project, that the University could ask the legislature what numbers would be useful while also responding to data like these. Dr. Radcliffe commented that the best numbers are those closest to direct costs; while there is a need to measure indirect costs, there are problems with doing it accurately.
Mr. Pfutzenreuter turned next to an explanation of "Institutional Support" in the University's budget from 2005-2010. He first noted that there are function codes grouped into various categories: instruction, research, public service, institutional support, academic support, operation and maintenance of plant, student services, auxiliary enterprises, scholarships and fellowships, and agency activity. Function codes "answer the question, 'How does this transaction accomplish the mission of the University?'" and they serve to track expenditures. He said he wished to speak about the "institutional support," which has been the subject of public comment, including "garbage that damages the reputation and integrity of the institution" and that is not checked for facts. Some of what has been published has led the public and the legislature to believe that administrative costs are out of control, but people have not bothered to check the facts.
"Institutional support" function codes identify expenditures for such activities as "(1) central executive-level activities concerned with management and long-range planning, . . . such as the governing board, planning and programming, and legal services; (2) fiscal operations including the investment office; (3) administrative data processing; (4) space management; (5) employee personnel and records; (6) logistical activities that provide procurement, storerooms, safety, security, printing, and transportation services to the institution, (7) support services to faculty and staff that are not operated as auxiliary enterprises; and (8) activities concerned with community and alumni relations, including development and fundraising." It also includes, in addition to central administrative offices, cash management, institutional research, insurance, property accounting, payroll, equal opportunity and affirmative action, the patent office, human resources, legal proceedings, Sponsored Projects Administration, public relations, and so on.
From 2005 to 2010 the institutional-support expenditures rose from $107.8 million to $234.3 million, which has led to statements that the cost of administration has exploded at the University, Mr. Pfutzenreuter said. He explained the major categories in "institutional support" and the increases that occurred over the six years 2005-2010; they have increased, but not for the reasons that have been identified by those who have not checked the facts.
1. Salaries: increased from $88.7 million to $121.0 million (+$32.3 million). This increase reflects the fact that the University installed a new enterprise system (EFS) and salary costs in the Office of Information Technology increased noticeably as a result ($5.1 million). The remainder of the increase, $27.1 million, reflects pay changes and modest growth in number of employees in a few central units.
2. Fringe benefits: increased from $7.8 million to $51.9 million (+$44.1 million). The increase reflects growth in the cost of fringe benefits but also reflects a change in federal rules regarding reporting of fringe benefits. This change is booked to a central account and represents $19.1 million in volatility of reporting according to federal rules, not a growth in administration. There are increases in direct department charges (which correlate with salary expenses) of $12 million, and the University also offered an early-retirement program, the cost of which shows up in this category as $11 million.
3. Supplies: increased from $21.2 million to $62.4 million (+$41.1 million). One benefit of adopting the new financial system is that they discovered several accounting practices that were wrong, Mr. Pfutzenreuter explained. The implementation of EFS provided the University with an opportunity to review old budgeting and accounting practices. One major adjustment was related to the accounting for royalty distributions from Glaxo for the HIV drug Ziagen; the correction to the accounting practice resulted in a jump in expenditures of $29 million which had previously been netted against revenues in the same account. This category also includes new software license fees at $3.7 million; the University bought a new financial system. This category also includes $3.4 million in endowment fees for the Consolidated Endowment Fund, which began to be accounted differently beginning in 2009, and a change in the accounting practices related to addressing and mailing services, which also resulted in a $2.1 million increase in the recording of expenditures. Mr. Pfutzenreuter explained that the four examples outlined account for $38.2 million of the $41.1 million increase.
4. All Other: increased by $8.9 million and related to a variety of accounting entries and minor changes in expenditures over time.
In total, Mr. Pfutzenreuter concluded, of the $126.5 million increase in "institutional support" expenditures over the period, $80.3 million had nothing to do with a growth in the administration, despite what outside commentators may claim. A majority of the remaining $46.2 million growth in expenditures relates to pay changes and associated fringe benefit costs along with minor changes in supplies and service spending.
Professor Luepker asked that a comment made later in the meeting by Provost Sullivan also be reported at this point in the minutes: When President Bruininks took office in 2002, there were 74 individuals in executive positions (deans, vice presidents, associate vice presidents, etc.). When he leaves office on June 30, 2011, there will be 74 individuals in executive positions. There will be no change in the number even though the institution's budget has increased by about $1 billion, enrollment is up by one-third, the number of degrees granted is up by 40%, and research funding has reached about $823 million.
Unfortunately, it is unlikely anyone will pay attention to these explanations, Professor Martin commented. Ms. Stahre agreed that people do not find it interesting that there was a change in the reporting system for royalty income.
Professor Luepker observed that there are two large categories of apparent cost increases: accounting changes and swings in fringe benefits. Mr. Pfutzenreuter agreed and noted that the Retirement Incentives Option and rules governing reporting fringe benefits account for about $30 million of the $44 million in that category.
Professor Zaheer commented that one can also demonstrate that the University is a larger enterprise than it was six years ago, so these expenses are likely not a larger percentage of the total budget. But much of the increase is simply due to changes in accounting practices, Mr. Pfutzenreuter pointed out. Professor Seashore suggested that one could look at the amount salaries increased, which have grown some and are 65% of the budget, and then look at all the rest, which is accounting procedures.
The core dilemma is whether it is worth the effort to refute erroneous claims, Mr. Pfutzenreuter said. Several Committee members were emphatic in responding that it is. Professor Cohen said this happened before and the University needs to get the numbers out. Professor Zaheer observed that they may have to hire someone to do the work; Professor Olin said that any comment needs to take into account the increase in the number of students.
Professor Seashore said that anything Mr. Pfutzenreuter says will be viewed with suspicion. This Committee, however, is a representative group of faculty, staff, and students, and it needs to say something. Individuals may blog or comment publicly to express opinions about the University's budget, but the Committee members have looked at the data and know that those expressing the opinions are "full of beans."
There is also a need for a discussion about how to fund things, Mr. Driscoll commented, a subject in which students are very interested. But without common facts, it is difficult get through the noise-driven discussion created by the Delta Project and bloggers. It is dignified not to respond but it may hurt the University not to do so. It would behoove University Relations to provide a responsible faculty/staff/student voice.
Professor Martin said there would be value in having the Committee come up with a response. Committee members pay a lot of attention to the budget over a long period of time and they know the numbers much better than outside commentators. Professor Roe agreed, pointed out that the views from the Committee must be genuine, and suggested that a subcommittee look at Mr. Pfutzenreuter's numbers and certify the Committee's support.
Professor Seashore commented that is not just a local phenomenon and noted that Inside Higher Ed had an article about a national effort to paint higher education as an easy target: All universities need to do is cut the number of vice presidents they have. While one would like to think Minnesota would be immune, it will not be. Note the events in Wisconsin and Michigan, she pointed out. The Committee can respond to a national trend of misunderstanding administrative costs. Or purposefully misrepresenting them, Professor Martin added.
Professor Luepker asked for volunteers to serve on a small subcommittee that would also include representation from University Relations. It would not be just about facts but about how to deliver the message, he said; Committee members are not experts in that arena. Professors Martin and Roe agreed to serve, as did Mr. Driscoll. Professor Luepker said he would contact Ms. Phenix, the new Chief of Staff, about other membership, and would also draft a charge to identify what the Committee would like to see at the end of the effort.
Professor Luepker thanked Vice President Pfutzenreuter and Dr. Radcliffe for joining the meeting.