University Research Questioned

Reprinted from The Michigan Review, October 7, 1992

by Chetly Zarko
Waste and misuse of federal research grants have become explosive issues for nearly every major research university as a result of a wave of federal investigations and audits during the past three years. In response to numerous unsettling findings, the federal government has implemented several new policies in an effort to force universities to control research costs. Numerous universities, however, have resisted these efforts, primarily by adopting modified accounting techniques rather than reforming their allocation and research practices.

The Prelude to Funding Cuts

Serious questions about the propriety of research spending initially arose in mid-1990, when federal auditors discovered nearly half a billion dollars in questionable grant expenditures at Stanford University.

According to Martin Anderson, Senior Fellow at the Hoover Institution, Stanford charged the federal government 74 cents "overhead" for every dollar in real research - one of the highest overhead ratios at the time. The audit uncovered"research" charges for the administrative costs of running a shopping center, sending alumni to Paris, and depreciation on a faculty retreat yacht.

Until the Stanford debacle, very little public pressure to account for expenditures had been exerted on the research systems of most universities. Afterward, however, federal auditors informed nearly 250 universities nationwide that their research charges would be subject to investigation.

Kevin Moley, budget director of the Department of Health and Human Services (HHS) (the largest federal medical grant-giving institution), noted that of the first 14 universities investigated, all were found to have improperly charged research expenses. The University of Michigan was among this group and voluntarily withdrew about $300,000 of a contested $6 million in illegitimate expenditures. These contested costs included a Christmas tree for President Duderstadt's office, expenses relating to Martin Luther King, Jr. Day celebrations, and trips to the Rose Bowl for a University administrator and his wife.

Overhead reimbursement rates were attacked by an irate Congress and executive branch. Public pressure mounted in the wake of discoveries and government policies on reimbursement procedures and research usage were progressively tightened.

Lax Conflict of Interest Policies

This year's flurry of activity concerning questionable research allocations included the release of a General Accounting Office (GAO) study entitled University Research: Controlling Inappropriate Access to Federally Funded Research Results. The report, released at the end of May, was commissioned to determine the extent and nature of linkages between universities and businesses which may have access to technology developed through university research, as well as the potential for conflicts of interest in such situations.

A conflict of interest as defined by the survey may arise when a researcher has vested interests, such as stock ownership or personal holdings in the company being given a license. Such conflicts can tempt a researcher to skew results in favor of his company, or can give privileged companies unfair advantages over others in the market.

The GAO surveyed 35 of the largest recipients of National Science Foundation (NSF) and National Institutes of Health (NIH) research grants and found that most universities have industrial liaison programs designed to sell the rights to university-developed technology in exchange for a share of the resulting profits. According to the GAO,"NIH or NSF funding accounted for $82 million, or 73 percent, of the $113.1 million that the 35 universities received in license income" in 1989 and 1990.

While the GAO study"sought to avoid" uncovering specific examples of conflicts of interest, it did underscore the concern that, nationally speaking, there is tremendous potential for conflict of interest in the determination of which companies receive the copyrights for these advantageous inventions. For example,"the university scientist who developed the technology was associated with the licensee in 61 cases, primarily as a consultant," noted the GAO.

A copy of the University of Michigan's responses to the GAO survey reveal weaknesses in controls on access to research and in the avoidance of potential conflicts of interest.

Although 60 percent of the universities surveyed have strict laws for the annual disclosure of outside interests of faculty and technology licensing personnel, the University of Michigan is one of the remaining 40 percent which generally rely"on faculty and other members of the university community to voluntarily disclose a potential conflict of interest."

The U-M's questionnaire responses also indicate only minimally sufficient compliance with state law. Senior executive officers and regents are not required to submit annual financial disclosure statements, and according to a response given by Assistant to the Vice Provost for Research Marvin Parnes, information on their corporate affiliations is"not readily available."

Senior executive officers and Regents are expected to report conflicts as they arise only if they are"covered by state law," said Parnes."No routine disclosure is required" for technology licensing staff (who have the most contact with both businesses and researchers), and"no such record is kept" of"efforts by university related personnel to influence the licensing process."

Rather than maintaining such records, the U-M trusts its employees to supervise themselves."[S]taff are alert to potential conflicts of interest and will bring these to the attention of their supervisors," said Parnes.

The GAO Study's Findings

The GAO study concluded that universities should implement more exhaustive disclosure rules and that, barring any university initiative, NSF and HHS should mandate greater disclosure requirements on its grants.

At the time of the survey, U-M was not considering any revision of its conflict of interest policies. The GAO study itself recommended that annual disclosure requirements be drafted and that NSF or HHS address the matter if universities failed to do so.

In the four months since the study, NSF and HHS have approved guidelines which would require a case-by-case disclosure policy. According to the U-M's Dr. Alan Steiss (Director, Division of Research Development and Administration), however, the high administrative costs of the"continuous filing" which such a policy would entail would defeat part of its own purpose."The irony of the new NSF regulations is that they could conceivably raise administrative costs. Some professors would be filing new case-by-case disclosures every month."

Steiss describes the debate within the administration as one between those who believe that disclosure constitutes an unnecessary burden on faculty, and those who feel that because the new NSF guidelines will become law anyway, they have little choice now but to adopt a policy of annual disclosure. Such a policy would reduce the administrative load compared to case-by-case disclosure.

In a coercive and roundabout way, therefore, the federal government will probably succeed in spurring annual disclosure.

While Steiss may be correct in arguing that the federal government has overreacted with its cumbersome case-by-case disclosure standard, it seems clear that universities like Michigan, which lack regular disclosure requirements, should somehow alter their policies. With an already tarnished reputation resulting from the overhead misuse scandals of 1991, universities will likely begin to recognize that regular financial reporting is in their best interests, since increased public scrutiny creates a disincentive for succumbing to favoritism or abuse.

Fee Caps on Overhead Costs: Real Reform or Mere Number Shuffling?

Along with these early summer developments, the Office of Management and Budget (OMB) instituted a nationwide policy that froze the percentage of available reimbursement in each of the accepted overhead categories.

One category of overhead covers purely administrative costs, such as paying employees and completing paperwork. This category was capped at 26 percent. The other type of overhead covers costs associated with the space required to conduct research, such as fuel, heating and electricity. For the U-M this figure was capped at slightly above 21 percent. Pursuant to executive order, universities would have to absorb any costs above those limits.

Despite its appearance to the contrary, OMB's choice of 26 percent as a fee cap for administrative overhead was not arbitrarily made; the average overhead rate for all major universities nationwide was 26 percent. Those universities with low overhead costs (around 20 percent) will feel no effect, but the University of Michigan, with overhead costs of 34 percent, stands to lose quite a bit of funding under the new guidelines.

Following OMB's imposition of overhead fee ceilings, HHS (which controls NIH) and other agencies began to adopt similar measures to enforce administrative cost-control.

While OMB may have reasoned (quite logically) that by setting a fee ceiling it would encourage universities to trim wasteful costs, its goals went largely unfulfilled as universities began a process known as"cost shifting."

"Different universities place some of the allowable costs in the administrative sector rather than the space sector, even though the particular cost could be justifiably placed in either sector [for example, the heating costs of an administration building can be classified in either way]. At the U-M we have tended to attribute more of those costs to the administrative side and therefore were particularly hard-hit by the administrative cap," explained Steiss.

Steiss maintains that the administrative cap has also been harder for other schools to meet because of this bias. He noted that "some schools, like Johns Hopkins, foresaw this shift and changed their accounting methods" before the new rule was adopted.

Steiss estimated that if the current auditing techniques were used, Michigan's total annual losses might exceed $8 million. He said that the U-M is currently engaged in negotiations with the HHS that would permit it to change auditing techniques to allow some previously labelled administrative costs to be counted as space costs. This strategy appeals to the University because its space overhead rate is currently below the reimbursement rate.

Such a number juggling scheme, however, ignores the prospect of"real" cost cutting as a solution. Furthermore, if OMB or HHS were allowed to set a more rigid fee cap, only then might those universities above it have a real incentive to cut actual costs. Unfortunately, those naturally below the cap have an equally strong incentive to inflate costs to meet the cap exactly.

Asked what the U-M is doing to achieve "real" administrative cost savings, Steiss explained that the U-M suffers from over-decentralization and therefore must seek to"reduce redundancies in the system." According to Steiss, the sub-unit structure of the University creates administrative task duplication across units.

Steiss says that the University is trying to reduce these redundancies, but that"it is a long, hard road toward coordinating research" in a system that has been geared toward decentralization for other legitimate reasons. He added that"improvements were being made to centralize information and data retrieval."

It appears that effective reform, however, may depend not merely upon removing redundancies but actually trimming the U-M's research bureaucracy.

Other questions about universities' claims arise when one examines a study released by the Washington-DC based lobbying association, American Associations of Universities (AAU). The study argues that comparisons among universities should not be made because"rates are calculated very differently" from school to school.

According to The Chronicle of Higher Education, however, the AAU study"contained almost no quantitative analysis to support its conclusions."

Steiss defended the study on the grounds that it pointed out deficiencies in federal policy and federal studies. Setting a flat rate for all Universities based on the implicit assumption that cost differentials can be compared,"is like comparing apples and oranges," he said, adding that "Northern universities have necessarily higher heating costs ... while ... agricultural universities have different research support patterns than medical or engineering schools."

Although federal policy has at times amounted to overreaction, it is rapidly and justifiably moving in the direction of tightening controls on research funds.

Perhaps the new fee limitations will give the U-M and other universities an incentive to boost efficiency while keeping costs low. If a select few universities can keep administrative costs near 20 percent, one would hope that Michigan should at least be able to trim its bureaucracy as well.