Or subscribe now to read with unlimited access for as low as $10/month.
A free account provides you access to a limited number of free articles each month, plus newsletters, job postings, salary data, and exclusive store discounts.
By most standards, the U.S. biomedical-research system is very successful. NIH-funded university research improves and saves lives while also generating large economic dividends. NIH-funded discoveries contributed to 354 of the 356 new drugs the Food and Drug Administration approved between 2010 and 2019. Economic estimates suggest that every $10 million of NIH-funded research helps spark the development of nearly three additional drug-related patents in corporate laboratories.
Conservative estimates put those patents’ value somewhere between $14 million and $28 million of later drug sales. That’s a 140-to-280-percent economic impact before you consider effects on health, the quality of people’s lives, and all the things NIH-supported research generates that aren’t drugs. Grants also provide advanced research training, employ hundreds of thousands of people, and influence the bottom line of businesses across the country. The people who work on them go on to apply what they learn across the economy.
Why, then, does a drastic cut to this portion of federal grants strike the NIH as a route to improving biomedical research? The answer has to do both with how IDCs are used by universities and with growing government and public distrust of academe. Because IDCs are a lump sum attached to grants that universities rarely account for, they are seen as a “Rorschach test of one’s faith in higher education.” While greater efficiency in their disbursement and increased transparency in their use may benefit everyone, the general logic of our current system is sound.
We need an explicit, hard-nosed look at the ‘fragile contract’ between the state and universities that has governed research since World War II.
That logic created some of the most productive features of our national research-and-development infrastructure. Similar arrangements characterize federal grants and contracts with other types of organizations, which often charge higher IDC rates. The crisis is not IDC rates themselves, but the lack of confidence in how the funds they generate are spent.
The solution shouldn’t be a policy that seems designed to destabilize rather than to reform the current system. Instead, we need an explicit, hard-nosed look at the “fragile contract” between the state and universities that has governed research since World War II. The grounds of that bargain have been shifting since at least the end of the Cold War, but any effort to reset the terms must begin with a broader view: It is time to step away from isolated examples and ask how current IDC policy creates and sustains the research capabilities that have served us so well.
At the heart of the matter are a few basic questions: How much does research really cost? Are the costs universities attribute to IDCs appropriate? Who should pay for them and why?
The “direct costs” on a grant are itemized, justified budget lines used to hire the people and buy the stuff needed to achieve a project’s specific research goals. “Administrative overhead” refers to IDCs, which are also called facilities and administration costs. Universities have used their estimates of costs for these two components to negotiate institution-specific rates since 1966.
IDC rates specify a percentage of direct costs that federal agencies pay to support the human, physical, and scientific infrastructure grant-funded research requires without having to do the expensive and time-consuming work of figuring out the precise infrastructure needs of each specific project. Institutions and agencies agree on a rate that averages out the cost of infrastructure needed to conduct research effectively, safely, and in compliance with relevant regulations across all government-funded projects done on a particular campus. The average university rate in 2023, 57 percent of direct costs, would add $57,000 to a $100,000 project budget, bringing the total grant amount to $157,000.
Arguments about IDC policy can touch on academic freedom, deficit spending, theories of innovation, the dangers of central planning, the proper role of expertise in a democracy, and even the maddening lack of pencils in the copy room.
The NIH and the academy play different roles in biomedical research. Universities are entrepreneurial: They raise funds and assume significant risk by investing in physical and scientific infrastructure. At Carnegie Mellon, for instance, a $75-million foundation grant let the university break ground on a $210-million scientific building with new collaborative spaces to complement a $40-million “cloud laboratory” where automated, robotic instruments will increase researchers’ ability to design experiments and collect data. A similar pattern of philanthropic seed funding coupled with significant institutional investment recently allowed Washington University in St. Louis to open the doors of a new $616-million neuroscience center dedicated to integrative work on diseases such as Alzheimer’s and addiction. Under current policies, the costs of those facilities can be factored into rate negotiations. Under the current system, IDCs on future NIH grants to these institutions might adjust to reflect the cost of these new infrastructure investments, but funds would only flow for the projects that the agencies choose to pursue.
Universities also invest in the human infrastructure needed for research compliance and safety. At my institution, the University of Michigan, colleagues working in research ethics and compliance oversee a broad array of issues: human-subjects protection, laboratory-research safety, required training in the ethical conduct of research, research-information security, conflicts of interest, controlled substances and hazardous materials, animal safety, research integrity, and international-research security. Similarly large, complex, and expert groups fulfill federal mandates for a central university unit to administer sponsored research projects and oversee the specifics of post-award financial management to ensure funds are stewarded in accordance with federal and state laws and sponsor requirements. The list could go on.
Thanks to a political firestorm involving accounting errors and a Stanford yacht, the administrative portion of university IDCs was capped in 1991. Federal research regulations have grown significantly since then, but these limits on what universities — and only universities — can recoup have stayed fixed. Perhaps the largest recent expansion of federal regulatory mandates for research involves international-research security, an arena where numerous federal actions have necessitated the creation of new positions to coordinate with law-enforcement agencies, work to revamp proposal submission systems, implement new research training and conflict-of-interest reporting requirements, and roll out new means to oversee campus visitors, foreign travel and collaboration, and project-risk reviews. These types of changes, which federal agencies often reimburse for other types of organizations, are one reason why universities paid nearly $6.8 billion in “unrecovered indirect costs.”
NIH thus gets to pick among proposals that rely on established capabilities. IDCs give universities an incentive to keep as close to the frontier as they can manage to make their faculty as competitive as possible. The agency treats research universities as resources it can lease on a project-by-project basis, like a national Airbnb for biomedical research.
Research universities bet on infrastructure needs such as automated experimental and data-collection equipment. Faculty develop projects that respond to NIH programs using the tools their institutions provide. Agencies define programs and make tens of thousands of annual funding decisions guided by policymakers’ priorities. Policymakers respond to the concerns of voters. This bottom-up process is driven by conceptions of the public good and some sense of the direction science could or should be heading. The result is a diverse, decentralized discovery system that often anticipates the needs of the next breakthrough, even when we can’t articulate exactly what that breakthrough might be.
While not perfect, this system has built a national research enterprise that is a uniquely powerful engine for economic growth, public health, and national security. Its variation, decentralization, and competitive drive makes it a resource for dealing with complex challenges. Now it is at risk.
The Trump administration’s attempt to change course is based on an oversimplified view of IDCs and their purpose. The NIH’s announcement appeared to pull language from a Heritage Foundation report that argues “the Left” uses these funds to capture universities for political ends. Both documents harmonize with Project 2025, which recommended “market based” IDC reform pegged to the “lowest rate a university accepts from a private organization.”
The agency treats research universities as resources it can lease on a project-by project basis, like a national Airbnb for biomedical research.
The administration’s view that universities are untrustworthy agents working counter to the public interest is galling. But holding aside the partisan flourishes — the accusations of “woke ideology” and debates over transgender mice — favoring the lowest bid for an adequate quality product is entirely appropriate for many contracts. However, it’s a questionable strategy for sourcing frontier research, especially given that the NIH’s portfolio of investments has tended to pay off handsomely. The fact that universities accept a relatively small percentage of research grants at less than full IDCs doesn’t tell us very much about what they need to sustain the research infrastructure NIH will require to keep American biomedical infrastructure the best in the world. What it does tell us is that universities are sometimes willing to take a loss on infrastructure costs for a small portion of their portfolio to support research they take to serve their goals.
According to my calculations using data from the Higher Education Research and Development survey, U.S. higher-education institutions conducted $62.2 billion of life-science research in 2023. About $0.46 per dollar was supported by HHS agencies, including NIH. Other federal agencies funded $0.08 per dollar, slightly more than nonprofits and businesses at about $0.07 each. Universities put $0.24 per dollar into the pot.
University research investments would have to skyrocket to pick up the government’s slack under the new NIH policy. My institution estimates the immediate cost of a 15-percent IDC rate would be $181 million. In the short term, it might be possible for some campuses to sustain current infrastructure at NIH’s new rate. But the money needed to make big investments in new capabilities can only really come from two sources, tuition and philanthropic gifts. Neither are realistic propositions at this point.
I, myself, am in favor of work that may yield more effective vaccines against diseases like Covid, even if the infrastructure is expensive. But that level of facilities and administrative support on a standard NIH research grant is akin to what the federal government pays defense contractors. A Department of Defense overview of IDCs includes sample rates in the 110- to 120-percent range. The NIH and other federal agencies will pay premium prices to support infrastructure for projects they take to be in the public interest. They just won’t pay them to universities.
If we want world-class biomedical research, we need a means to update its infrastructure as science and the nation’s needs change. The NIH seems to expect that direct-cost investments alone will suffice to keep America ahead of the knowledge curve. They won’t. Overweighting direct costs will most likely limit us to research using the infrastructure we have now. Absent drastic change, that’s a recipe for long-term decline.
The system we have works reasonably well. But its continued workings depend on complementary investments made by federal agencies and universities.
The policy logic is disconcerting. If the federal government holds an abiding distrust of universities, why would it leave the nation’s research infrastructure even more completely in their hands? A better option would be to strengthen incentives to attend to NIH priorities by raising the prices it pays universities, perhaps with an expectation of greater transparency about how funds are spent. Both the NIH and the universities that do the bulk of the research it funds pursue their missions in service to a conception of the public good. While those conceptions may sometimes differ, they are more often aligned.
The system we have works reasonably well. But its continued workings depend on complementary investments made by federal agencies and universities. If NIH bows out of real support for the infrastructure it uses, stagnation is the likely outcome, not the world-class research the agency says it wants.
New discoveries are opening exciting possibilities in biomedicine. Half of last year’s Nobel Prize in chemistry went to AI researchers for a 2020 model that promises revolutionary changes in synthetic biology by predicting protein structures. AI and data innovations are transforming opportunities in precision medicine. Neither the questions nor the infrastructure needs are settled in either field. Discovery in both areas may require the kind of decentralized, competitive, and varied research infrastructure that an IDC policy like the one NIH is seeking to eliminate helped create. It’s a terrible time to abandon the strengths of the current system.
Instead, policy makers and academic leaders should recognize that the strength of our system lies in the joint venture to construct and sustain the kind of research infrastructure the nation needs, then negotiate ways to increase its efficiency and transparency. That may require new types of university accounting, new government willingness to reduce regulations or accept their costs, and a recognition that common interests make good partnerships — but seeking complete agreement on how best to serve the public good may do more harm than good.
