Leveling the NSF Playing Field

National Science Board urges barring "voluntary" sharing of research costs that is widely seen as favoring wealthier universities in grant competitions.
August 31, 2009

Colleges and universities contribute significantly to the cost of federally sponsored research projects, through what they spend on research labs and equipment, faculty start-up packages, and "indirect" costs that aren't reimbursed by the government.

At various points, federal agencies have either required or encouraged them to quite literally "share" the costs of research grants they win, essentially putting up their own funds to match a portion of the grant's value. The practice has been debated, though, with proponents arguing that it shows institutions' commitment by forcing them to put their own "skin in the game," but detractors saying that requiring or encouraging "cost sharing" puts less-wealthy colleges and universities at a disadvantage against wealthier peers, and can lead grant reviewers to favor proposals from institutions that volunteer to contribute even though that isn't supposed to factor into the decision making.

The 2007 American COMPETES Act asked the National Science Board to weigh the pros and cons in reassessing the cost sharing policies of the National Science Foundation, and the board issued a report Friday that calls for ending the practice of "voluntary" cost sharing in all circumstances, while continuing "mandatory" contributions in a small number of industry-focused federal programs as recommended by an interim report the board issued last year.

"The board firmly believes that prohibiting voluntary committed cost sharing, and permitting mandatory cost sharing requirements only in limited and appropriate circumstances, will not reduce institutional commitment and financial contributions to NSF-sponsored projects or negatively impact institutional stewardship of Federal resources," Steven C. Beering, president emeritus of Purdue University and chairman of the science board, wrote in a memo accompanying the report. "Instead, it likely will enhance the ability of institutions to strategically and flexibly plan, invest in, and conduct research projects and programs, and will promote equity among grantee institutions in NSF funding competitions."

The report, "Investing in the Future NSF Cost Sharing Policies for a Robust Federal Research Enterprise," recounts the history of the science foundation's policies on cost sharing, which has shifted repeatedly over 50 years. In 2004, the science board eliminated mandatory cost sharing requirements in all NSF programs, aligning the agency's practices with those of other federal research agencies. The change did not address "voluntary" cost sharing, which continued to be permitted under NSF rules.

Eliminating required matching by institutions was designed to take an institution's ability to contribute financially out of the peer review equation and "remove eligibility barriers to participation in certain NSF programs by institutions unable to provide the required cost sharing," among other beneficial outcomes, the board writes in its new report.

But the 2004 change created some problems, too, most notably that it made it more difficult for colleges and universities to "leverage" federal money to attract corporate research support in NSF programs that are designed to encourage academic-industry collaboration.

When the panel on cost sharing began its work in late 2007, as demanded by the America COMPETES Act, which was designed to strengthen the country's commitment to research and education in the physical and natural sciences, it very quickly reached the conclusion that the NSF should alter its policies on mandatory cost sharing. Its February 2008 report recommended that the agency reinstate mandatory matching in three programs with strong corporate involvement: the Experimental Program to Stimulate Competitive Research, Industry/University Cooperative Research Centers, and Engineering Research Centers programs (which NSF promptly did).

The report issued Friday, based on a broader review and soliciting of opinions on the NSF's cost sharing policies, affirms that the NSF should "allow, but narrowly circumscribe, the application of mandatory cost sharing requirements in NSF programs in which cost sharing is foundational to achieve programmatic goals," which at this point it defines as the programs above plus two programs in which it was dictated by America COMPETES: the Major Research Instrumentation Program and the Robert Noyce Teacher Scholarship Program.

The panel dedicates much more discussion, though, to the thornier issue of "voluntary" cost sharing -- which like many other financial issues gets complicated because it becomes something other than fully "voluntary," because those that don't pony up often come to be seen (fairly or not) as weaker by those making decisions.

"The proposer community generally views offers of voluntary committed cost sharing in proposals in increasing their competitiveness (i.e., likelihood of receiving funding) in NSF funding competitions," the science board writes. "Correspondingly, failing to offer significant voluntary committed cost sharing in proposals is viewed as creating a competitive disadvantage. These views are widespread and strong ... even though NSF instructs program officers, reviewers, and the proposer community that voluntary committed cost sharing is not to be a factor in the merit review and award decision process."

The assessment that voluntary sharing creates inequity among research institutions, combined with a desire to ease institutions' administrative burdens for tracking voluntary cost sharing and to give colleges and universities "maximum flexibility in expending their discretionary resources on research activities," led the science board to recommend that "NSF should prohibit voluntary committed cost sharing in all components of both solicited and unsolicited proposals."

The science board report includes one further recommendation that could bode well for colleges and universities down the road: a suggestion that the federal government review whether to raise the cap (currently at 26 percent) on the percentage of administrative costs for which universities can be reimbursed by the government on research projects. Universities are alone among among contractors that do federal research on which such a cap is imposed, the Council on Governmental Relations notes.

"The board understands the fundamental intent of the administrative rate reimbursement cap -- to ensure that the majority of research funding supports direct research effort, rather than administrative costs -- but also concurs with the general view of the research community that the current 26 percent reimbursement cap requires re-evaluation," the report states.


Back to Top