A hearing on how the U.S. government defrays the cost of doing federally funded research on college campuses might put most people to sleep. But when budgets are tight, the billions of dollars being spent each year on so-called overhead become an irresistible target for lawmakers.
This past Wednesday, the science committee of the U.S. House of Representatives weighed in on the subject, one that is at the core of the U.S. research enterprise but also exceedingly complicated. The hearing gave Republicans an opportunity to voice support for lowering overhead payments, which cover things like electricity, lab maintenance, regulatory compliance, and administration. Lowering those costs is a key to a proposal by the Trump administration that would affect those getting grants from the National Institutes of Health (NIH). Democrats acknowledged that the current system could be improved but warned that some approaches on the table could have unintended negative consequences.
Chaired by Representative Barbara Comstock (R–VA), the 100-minute hearing was refreshingly free of the extreme partisanship that has hobbled much of the committee’s efforts in recent years. The panel of witnesses included both a longtime advocate for blowing up the current system, Ohio University in Athens economist Richard Vedder, as well as James Luther, a senior financial officer at Duke University in Durham, North Carolina, and board chair of a Washington, D.C.–based organization representing the interests of the mainstream academic community.
Vedder said that wealthy universities profit from the present system, which gives them an incentive to seek ever-more federal funding. The government would save money by declaring a flat reimbursement rate, he asserted, although he didn’t suggest a figure and didn’t address the president’s 2018 proposal for NIH. (It calls for reducing overhead payments from roughly 28% of the agency’s grant spending to 10%; that cut would reduce NIH’s overhead payments to universities from about $6.5 billion to $1.9 billion.) In fact, Vedder said he prefers altering the peer-review system to benefit grant applicants whose institutions are willing to accept lower rates of indirect cost recovery.
Luther’s job was to defend a system that nobody likes but that he and his academic colleagues believe is essential to ensuring that basic research continues to be an engine of U.S. innovation. He offered a homey analogy: “If direct costs are the gas for the research engine,” he said in his written testimony, “then reimbursements [overhead] represent the oil. The research engine requires both.”
Some members seemed to welcome the chance to hear those well-worn arguments before making up their minds. “I’m being educated on this issue by university officials in my district,” confessed Representative Roger Marshall (R–KS), a rookie legislator. Marshall may be politically inclined to view overhead as an unnecessary government expense, but his district includes Kansas State University and a national biodefense laboratory under construction.
For Comstock, who leads the committee’s research panel, the hearing was an opportunity to put her mark on an issue that had suddenly taken on a higher profile with Trump’s budget proposal. But it was an imperfect fit.
For starters, the committee has no jurisdiction over NIH. So members were left to explore the practices of the National Science Foundation (NSF), another major funder of academic research that is part of the committee’s portfolio. In addition, NSF doesn’t set overhead rates for the nation’s colleges and universities. Rather, it applies the rates negotiated between universities and one of two other government agencies.
Finally, the committee was so eager to jump on the overhead bandwagon that it took the unusual step of asking the General Accountability Office (GAO), a nonpartisan congressional watchdog agency, to present preliminary results from a study of NSF overhead payments that won’t be completed until the fall (see graph, below/above). GAO studies, which usually find fault with an agency’s accounting practices, aren’t usually released until they are complete. But this time, when Republicans jumped on a preliminary finding that universities received a bigger chunk of indirect costs to go with awards NSF made last year than did companies or other federal agencies, GAO’s John Neumann pleaded with lawmakers to withhold judgment. Neumann said he had yet to examine what factors could be causing the discrepancy, including the nature of the research being conducted and whether the sectors follow different accounting procedures.
What is overhead?
Indirect cost recovery rates are not for the faint of heart. So here’s a primer. [For the record, the NSF witness, William (Dale) Bell, asserted gleefully more than once that indirect costs are indeed “sexy.”]
Every federal grant comes with an additional component for the “indirect” cost of doing research. That word is a catchall for a cornucopia of expenses that can be grouped under either administrative costs or facilities. The first category is what universities need to spend to comply with myriad federal rules governing research activities, from disposing of hazardous waste to ensuring that research subjects—both animal and human—are treated properly. The second covers all the physical costs—bricks and mortar, equipment and instruments, utilities, and grounds—of doing research.
Universities could calculate those costs for each grant. But it would be a nightmare. So instead, every 3 to 6 years, each university and the government negotiate an overall rate to cover every research project, subject to extensive rules on what costs can and cannot be counted. In general, university overhead rates average about 50% (though they vary widely), meaning the government would give the university $50,000 to cover the overhead on a $100,000 grant. (Calculating it another way, one-third of the grant goes to overhead.) Many university officials have long complained that the government formula shortchanges their institution, leaving them to pick up a significant chunk of those expenses.
Why you should care
The idea that the government should reimburse universities for the cost of supporting federally funded research arose shortly after World War II, fueled by science’s contribution to the war effort. Indirect costs also feed into the broader principle of cost-sharing, that is, that requiring institutions to have some “skin in the game” will make federal dollars go further.
But policymakers realized that wealthy institutions with large endowments might have a competitive advantage because of their ability to “buy” more research. Indirect cost recovery was seen as a way to level the playing field by refilling the coffers of poorer institutions.
With the hearing coming 1 day after President Trump proposed cutting NSF’s budget by 11%, it’s no surprise that every member paid homage to the importance of containing costs. However, Republicans gravitated toward the assumption that the current system needs to be flensed, if not stripped bare. “Are taxpayers paying for these costs in an efficient and transparent manner, or are we unnecessarily subsidizing excess?” wondered Representative Lamar Smith (R–TX), chairman of the full committee.
In contrast, Democrats tended to worry that putting too much emphasis on saving money could hurt the research enterprise by eroding quality. “Let’s make sure we are not initiating a race to the bottom,” warned Representative Don Beyer (D–VA), vice-chair for the minority, “with prizes to the lowest bidder doing the least valuable research.”
Without pending legislation and with the GAO analysis unfinished, Comstock is under no pressure to reconcile these competing interests. Last year’s reauthorization of NSF programs created an interagency working group within the White House budget office to examine ways to ease administrative burden on universities. But it has yet to be constituted, and indirect costs will be only one item on its agenda. So this perennial issue is likely to remain in the news, awaiting resolution.