What This Paper Covers

This paper makes the case for a sweeping reform of the National Institutes of Health (NIH), the world’s leading biomedical research funder, with a nearly $50 billion annual budget. Outdated organizational structures, inefficient grant processes, and weak data transparency hinder NIH’s role in advancing medical innovation and undermine public trust. The paper lays out structural and procedural changes necessary to modernize NIH for the 21st century.

What We Found

NIH in its current form reflects an antiquated structure and set of internal policies that threaten the future of scientific research and U.S. leadership in the sciences. NIH’s 27-institute structure reflects obsolete disease-specific silos, duplicating work and stifling interdisciplinary breakthroughs. The grant process is mired in bureaucracy, with many researchers spending nearly half their time on compliance. Additionally, the current peer review system favors risk-averse, incremental research that studies what is already known, thereby discouraging transformative innovation.

Indirect costs—those not directly focused on the research—take up an increasing share of grants. Many studies cannot be replicated, and there is excessive “soft money” salary dependence in academia: Researchers are at risk of losing their salaries unless they raise sufficient grant money, thus incentivizing them to study whatever marginal topic will get funding. NIH also fails to sufficiently fund emerging organizational models or enforce transparency in null results and internal data. This undercuts scientific integrity, limits commercialization success, and perpetuates inefficiencies across academia and government.

Why It Matters

While NIH’s contributions to scientific progress over the decades have been considerable, structural inefficiencies, outdated organizational models, and bureaucratic constraints are hampering NIH’s role in breakthrough discoveries and translating research into improved health outcomes. In the wake of COVID-era policies and NIH’s involvement with gain-of-function research, it is time to rebuild public trust in our biomedical research institutions.

Health and Human Services (HHS) Secretary Robert Kennedy’s “Make America Healthy Again” initiative involves significant reforms to public health institutions. NIH reform is imperative, as the United States spends far more per capita on health care than do other OECD nations while achieving poorer health outcomes. As the primary driver of basic scientific research in America, NIH should evolve to better advance scientific and medical understanding.

What We Recommend

This report offers a series of recommendations, almost all of which are actionable via existing administrative authority:

• Structural realignment: Consolidate NIH institutes around shared disease mechanisms and systems biology rather than outdated organ- or disease-specific lines.
• Administrative modernization: Reduce indirect costs (which requires congressional assistance), streamline bureaucratic requirements, and reduce researchers’ compliance burden to 20 percent of their time.
• Grant system overhaul: Reduce reliance on preliminary data, impose term limits for institute directors, and foster innovation through randomized experimentation with peer review.
• Research integrity: Mandate data transparency, fund replication studies and fraud detection, and incentivize the publication of null results.
• Support for scientists: Curb exploitative postdoctoral employment practices, reduce soft money salary models, and fund more independent, non-university-based research entities.

Expand the Common Fund: Triple the Common Fund budget to $2 billion to support cross-cutting research that does not align neatly with existing institutes.

Collectively, these reforms aim to reinvigorate NIH’s core mission—to advance meaningful scientific discovery, improve public health outcomes, and restore trust in federally funded research.

The National Institutes of Health (NIH) stands as the world’s premier biomedical research funder, commanding an annual budget of nearly $50 billion and serving as the backbone of basic scientific research. Over the past few years, including problems related to its handling of the COVID pandemic, it has become increasingly apparent to both policymakers and the public that NIH needs reform. President Trump and Health and Human Services (HHS) Secretary Robert Kennedy are pursuing a “Make America Health Again” agenda, calling for major reforms to public health agencies. These changes are motivated, in part, by the reality that the United States spends far more per capita than any other advanced OECD country on health care¹ yet has one of the lowest life expectancies.³

Indeed, in the past two years, Congress has shown a greater appetite for engaging with NIH and pursuing reform than at any time since 2006. For example, Representatives Cathy McMorris Rodgers (the chair of the House Energy and Commerce Committee in 2023 and 2024, which oversees NIH) and Robert Aderholt (chairman of the Labor, Health and Human Services, and Education Appropriations Subcommittee of the Energy and Commerce Committee in 2023 and 2024) argued for NIH “reform and restructuring”⁴ and released a white paper with specific reforms.⁵ Similarly, the Health, Education, Labor and Pensions (HELP) Committee—which has jurisdiction over NIH—released a white paper with many productive metascience ideas for how to improve NIH.⁶ (These papers are referred to as the “House white paper” and the “HELP white paper” going forward.)

With the energy and motivation of the Trump administration to disrupt the status quo and implement reforms, the time is ripe for addressing problems with the current NIH organization and considering a wide range of reform proposals, including everything from radical restructuring of NIH to engaging in bold experimentation with new models of peer review.

One of the chief problems with NIH is that it is divided up into ICs that are almost all organized around diseases or body parts, such as cancer or eyes or heart disease. This sort of structure does not reflect current understandings of disease processes and human biology, and it can result in wasted resources and duplicative efforts. According to senior NIH insiders, the agency’s organizational structure is so convoluted that it is often jokingly attributed to congressional design: “The structure of NIH is so crazy that only Congress could have come up with it.”⁷ As shown below, however, Francis Collins, who served as NIH director from 2009 through 2021, mostly ignored or even blocked attempts to address this problem.

As a former NIH director said, “Institutes do not follow changes in science, and this is a systemic problem for all of science. Similarly, university departments do not necessarily align with how you treat diseases. For example, if asthma is an immune problem, does it belong with pulmonary or immunology?”⁸

NIH’s outdated structure and organization arguably keep the United States from pushing the boundaries of scientific innovation. For example, NIH contains both a National Institute of Drug Abuse (NIDA) and a separate National Institute of Alcoholism and Alcohol Abuse (NIAAA). With widespread agreement on shared comorbidities, lifestyle choices, and outcomes, why are these institutes separate? Politics and history, not science. There was a major effort to combine those two institutes in the early 2010s, but the plan was shot down by NIH Director Collins, evidently due to political reasons.⁹

More generally, the idea of reorganizing NIH is exactly what multiple NIH directors have wanted—or attempted and failed to effectively do—since the 1990s if not earlier. Harold Varmus, who served as NIH director from 1993 to 1999, wrote in 2001 about the “proliferation” of institutes at NIH,¹⁰ by which he meant the fact that NIH ICs were added and rearranged for political reasons for decades without regard for how they all fit together. As he later wrote in his memoir, he argued that “unless proliferation was stopped, the NIH would become an unmanageable collection of over fifty units within a decade or two at its current rate of growth. I then took the argument a step further, and proposed that the current organization of the NIH be reconsidered and redesigned, creating six centers of approximately equal size, including one for the director, giving him or her increased authority over coordinated programs, new initiatives, and budgets.”¹¹

As discussed in a 2003 National Academies report, Varmus’s 2001 proposal was to create five main institutes: the National Cancer Institute, the National Brain Institute, the National Institute for Internal Medicine Research, the National Institute for Human Development, and the National Institute for Microbial and Environmental Medicine. Each of these proposed institutes would be organized around broad categories of diseases or human biology, rather than body parts such as the heart or the eye, and would conduct both extramural and intramural research. Varmus would have created a sixth entity, NIH Central, headed by the NIH director. In the National Academies’ words, “NIH Central would have responsibility for policies across NIH (e.g., on intellectual property, personnel management, or training programs), the peer-review process, scientific infrastructure (e.g., information technology, buildings and facilities, including the intramural Clinical Research Center), and thematic coordination (through links to the offices in each of the five institutes).”

Varmus’s memoir added this about his motivations: “Anyone who looks at an organizational chart of the NIH will be immediately struck by its complexity, especially by the multitude of institutes and centers, now twenty-seven, then twenty-four, each with its own authorities, leaders, and (nearly always) appropriated funds.” Varmus continued that the “resulting proliferation of institutes and centers has not been healthy in every way” in that it has “created administrative redundancies” and “has placed some topics at a disadvantage, because they are overseen by relatively small institutes that may lack the capacity to conduct large-scale efforts, like major clinical trials.” Finally, he added, the most obvious harm was that this “diversity of institutes complicates the central management of the NIH by the Director.”¹²

Elias Zerhouni, who served as NIH director from 2002 to 2008, wanted to create a “Scientific Management Review Board” that would consider these organizational issues on a regular basis, without being driven by political considerations (such as which patient group’s lobby was the strongest).¹³ That board was indeed created in the NIH Reform Act of 2006, although it lay dormant from 2015 onward as it was not a priority for Francis Collins. It has recently been resurrected at the behest of congressional appropriators.¹⁴

The 2024 House white paper (along with an early draft of a House Appropriations Committee bill) proposed to reduce the number of NIH institutes and centers to 15 (from the current 27), mostly through combination of existing ones. Among several examples:

The National Heart, Lung, and Blood Institute, the National Institute of Diabetes, Digestive and Kidney Diseases, and the National Institute of Arthritis and Musculoskeletal and Skin Diseases would be combined into a new National Institute on Body Systems Research.

The National Institute of Dental and Craniofacial Research, the National Eye Institute, and the National Institute of Neurological Disorders and Stroke would be combined into a new National Institute on Neuroscience and Brain Research.

The theoretical frameworks of scholars such as William Niskanen and Gordon Tullock are helpful in understanding NIH’s organizational behavior. For example, Niskanen’s theory of bureaucracy suggests that government agencies such as NIH naturally prioritize maximizing their budgets over societal welfare or organizational efficiency (even though their public-facing rhetoric usually focus on attacking diseases). Tullock’s analysis would suggest that the decentralized structure of NIH, with its 27 separate ICs, creates what he called “bureaucratic slack” or inefficiencies that persist because they benefit the internal stakeholders, who lack competitive pressure.

According to Tullock’s theory, established ICs engage in defensive rent-seeking, or protecting their territorial interests by resisting consolidation. Program officers whose authority derives from controlling specific resources have strong incentives to play up their unique contributions by overpromising what can be accomplished in the future. For example, the NIH announced a national plan on Alzheimer’s disease in 2012, which promised to “prevent and effectively treat Alzheimer’s disease by 2025.”¹⁵

Any reform efforts might be more successful if they directly addressed these incentive problems, perhaps by creating performance metrics that reward cross-IC collaboration and resource-sharing rather than individual IC growth.

Improving the NIH Structure

Completely revamping all of NIH’s structure is beyond the scope of this paper. But as a broad principle, Congress and the new administration should consider restructuring NIH around basic science of the causes of disease,¹⁶ biomarkers to measure disease, and broad pathways to prevent or reverse disease.

The Advanced Research Projects Agency for Health

In early 2020, Michael Stebbins (a former official in the White House’s Office of Science and Technology Policy) and Geoffrey Ling (a former biotech leader at the Defense Advanced Research Projects Agency, or DARPA) proposed a health agency modeled on DARPA, which is famous for having funded the original internet.¹⁷ The rationale was that NIH was falling short: “NIH is not designed to develop marketable disease treatments or cures, or to develop new platform technologies that are intended to revolutionize medicine.”

The ultimate result was ARPA-H, which Congress created in FY2023 as an independent agency within NIH. This structure was intended to provide the “flexibility and independent decision making” needed for innovative health research, unfettered by the typical NIH bureaucracy, while leveraging “NIH capabilities to accelerate ARPA-H’s immediate business operations.”¹⁸

A significant congressional debate was about how independent ARPA-H should be from NIH (indeed, some in Congress wanted ARPA-H to be entirely separate from NIH). In the end, Congress decided to ban ARPA-H from employing anyone who worked for NIH in the prior three years and from locating on the NIH campus (instead, ARPA-H had to create offices in “not less than 3 geographic areas”¹⁹).

ARPA-H is still too new to evaluate, but policymakers should let ARPA-H remain independent from NIH for the foreseeable future. Indeed, ARPA-H should be promoted to a fully independent operating division of HHS with no association with NIH. The whole reason for creating a new health agency was that policymakers thought NIH was not sufficiently supporting innovation and efforts to bring new health treatments to the market.

Aging and Dementia

There is a growing sense among many biologists that most chronic diseases are pathologies of aging and can be usefully addressed as such.²⁰ NIH already has a National Institute on Aging (NIA), but more than 80 percent of its $4.5 billion budget (FY2025) is directed to a single pathology: Alzheimer’s disease. The House white paper proposes to relabel NIA as a National Institute on Dementia. That would largely be truth-in-labeling: as can be seen in Figure 1, NIA has been overwhelmed with Alzheimer’s research in the past several years at the expense of a wide range of conditions that affect the elderly.

Dourado and Peng at the Institute for Progress argue that the federal government should invest in “understanding of aging biology at the cell, tissue, and organ levels.”²¹ They advocate for an actual NIA—not one dominated by Alzheimer’s but one that actually addresses a broad range of aging-related questions in biology, including on slowing the pace of cellular aging. They argue that this would improve “outcomes for patients with Alzheimer’s as well as a plethora of other chronic diseases associated with aging—diseases that are the leading cause of mortality and disability.” Indeed, a distinguished Alzheimer’s researcher said that “our field was better when Congress wasn’t throwing so much money at Alzheimer’s, because now every neuroscientist is just putting the word ‘Alzheimer’s’ in their proposal regardless of what the grant is actually about.”²³

Chronic Diseases

Chronic disease should be addressed systemically, with the focus changing from organ-specific pathologies to pre-symptomatic causes and interventions. A National Institute of Chronic Disease (NICD) within the NIH framework could serve as the primary federal agency coordinating research, prevention, and treatment strategies for conditions such as diabetes, hypertension, obesity, and cardiovascular disease. This institute would prioritize addressing these widespread health challenges through an integrated approach that recognizes their interconnected nature and shared risk factors.

The NICD would fund and conduct comprehensive research spanning from basic science to large-scale clinical trials focusing on understanding disease mechanisms, identifying novel therapeutic targets, and developing more effective interventions. This institute would emphasize prevention through public health initiatives, lifestyle interventions, and early detection programs.

The NICD would also focus on improving disease management for those already affected by chronic conditions. This would include developing health care models that address comorbidities, studying the long-term progression of chronic diseases, and supporting patient-centered approaches that improve quality of life. Through strategic partnerships with other NIH institutes, health care organizations, and other federal agencies—such as the Food and Drug Administration (FDA) and the Centers for Medicare and Medicaid Services—the NICD could serve as a catalyst for research into the growing burden of chronic diseases that represent the leading causes of mortality and disability in the United States.

Addiction

HHS should combine NIDA and NIAAA into a National Institute on Addiction.²⁴ Alcohol and drug addiction are variations of the same issue, and there is no reason to have separate ICs for them. Indeed, a National Academies report from 2003 made this recommendation and noted that all of the “arguments against merger appear to be primarily nonscientific,”²⁵ such as political opposition from the alcohol industry. This new IC should respond to the emergence of GLP-1s and similar drugs by funding new lines of research on how such drugs might be useful for tackling addictions of all kinds.

Data Resources

A National Center for Data Resources could help centralize and streamline the fragmented data landscape currently hindering research across NIH. With NIH currently supporting at least 147 different data repositories,²⁶ researchers constantly face significant challenges in data accessibility, interoperability, and cost efficiency. An NCDR could serve as a pilot case for data consolidation, working to integrate diverse datasets related to diabetes, hypertension, and other chronic conditions that are currently siloed across multiple NIH ICs.

This data centralization effort could lead to substantial benefits for chronic disease research and treatment. By creating unified, standardized datasets that link genomic and clinical data sources, the NCDR could help researchers more effectively identify patterns, risk factors, and potential interventions. The NCDR could establish common data elements and standardized collection protocols specifically designed for chronic disease research, ensuring that information gathered across multiple studies becomes more compatible and valuable for meta-analyses. This would address the redundant activities and internal spending preferences that have led to duplicative and disconnected data resources throughout NIH.

Nutrition

Nutrition has historically been neglected at NIH due to its inherently interdisciplinary nature. That is, even though nutrition affects the entire human body and many diseases, there is no natural home for nutrition within the NIH framework of specific diseases and body parts. Thus, nutrition research has been funded in a scattershot approach across many ICs.

As of 2020, the Office of the Director has used its small Office of Nutrition Research to promote the first-ever NIH-wide nutrition strategy.²⁷ This is a good start, but NIH could vastly expand the Office of Nutrition Research to, among other things:

• prepare a report (with the Environmental Protection Agency) on microplastics and endocrine disruptors in drinking water;
• research ways to remove microplastics from the human body;
• research whether there are effects from the use of oxidized industrial seed oils;
• expand the intramural metabolic research unit and support more tightly controlled research into the effects of nutrition, exercise, sleep, and stress reduction techniques on obesity, chronic disease, and aging and direct the Office of Disease Prevention and Health Promotion to promptly translate the results into health guidance;
• sponsor research into new or traditional food preservatives and sweeteners that do not disrupt the endocrine system, child development, or health; and
• launch a review with the FDA of the safety of dietary additives.²⁸

Strengthening the Common Fund

The so-called Common Fund was established by the NIH Reform Act of 2006 and consists of funding for the Office of the Director to use for research across NIH, including interdisciplinary research.²⁹ Given the other priorities discussed in this policy brief and the aim of Secretary Kennedy’s Make America Healthy Again initiative, more NIH resources should be directed to the Common Fund, which has hovered around $600-$700 million a year for about the past decade.³⁰

The Common Fund should receive at least $2 billion, as many of the above issues (such as nutrition) do not fit neatly within any of the traditional ICs, and there is no reorganization plan that would have an IC for every possible cross-NIH issue. An expanded Common Fund would enable the NIH director to advance cross-cutting research priorities and enhance replication efforts.

Term Limits

Policymakers have had increased interest in imposing term limits on NIH IC directors. In early 2021, HHS enacted five-year terms for certain policy-level directors or their equivalents across all operating divisions, although those terms could be renewed with no limits. The House white paper makes a further recommendation: “limit every IC Director to a five-year term, with the ability to serve two, consecutive terms, if approved by the NIH Director.”³¹ The rationale for term limits is that no one person should dominate an entire field for decades. Instead, as with DARPA, new people with new ideas can improve innovation and scientific advancement. Indeed, the same rationale applies to the NIH director position.

The most obvious example of a problem in this area is that Anthony Fauci served as the director of the National Institute of Allergy and Infectious Diseases from 1984 to 2022—nearly 40 years. Even assuming Fauci was ideally suited to the role, scientific fields benefit from regularly integrating fresh perspectives and leadership. NIH oversees billions of dollars that research institutions receive each year. Fiefdoms and power centers around the perspectives or desires of one individual can actually hinder scientific and medical advancement. At a minimum, individual researchers within those entities might be particularly hesitant to raise medical or scientific questions that might run counter to a long-time IC director.

At the same time, there is a need to balance new ideas with the value of long-standing expertise. A rigid rule with regard to term limits might be too severe, but at a minimum, NIH should implement a more rigorous and regular review of each IC director’s performance. Such a review would allow NIH to exercise more influence over how an IC director performs.

As many people have pointed out, there are significant problems with how NIH distributes grants. First, researchers, by federal requirement, spend excessive time on proposals and paperwork. Second, the grant process is inherently biased against innovative ideas. Third, the entire process takes entirely too long. This section is intended to explore reforms.

Bureaucratic Burden on Researchers and Universities

Numerous federal surveys over the past two decades have shown that federally funded scientists spend too much time on bureaucratic compliance rather than actual science.³² For two of many examples, researchers commonly complain that it takes way too much time to apply for grants in the first place—there are too many forms and certifications needed—as well as to submit progress reports that are essentially regurgitations of their published articles.

A 2018 survey of 11,167 principal investigators on federal research grants found that “almost half of their available research time for federal projects had to be allocated to fulfilling requirements instead of focusing on the content of their research projects.” That number is just an average: Anecdotally, some scientists say they spent up to 70 percent of their time on paperwork and bureaucracy and that there is “no time to do science.” Even worse, and as Figure 2 shows, “the trend seems to be that time taken from research by requirements is increasing, not decreasing.”³³

Scientists have to spend a significant amount of time preparing a proposal for the grant itself. NIH’s guide to preparing a proposal is 157 pages long.³⁴

Then, if researchers win a grant, they must consistently spend a great deal of time on all the federal requirements as to budgets and finances, personnel, monitoring sub-awards, managing data, and reporting on how they spent their time. For researchers who work with animals or human subjects, it can be quite an additional ordeal to navigate the demands of the ethical review boards.

To be clear, some level of oversight and ethical review is a good thing: The U.S. government should not hand out billions of dollars every year without knowing how it was spent and whether the research was ethical and well-managed. But the burden of administration should not be nearly half of scientists’ time.

Over the past two decades, numerous agencies—including the Government Accountability Office (GAO), the National Science Board, and the National Academies—have argued for reforming grant administration,³⁵ but no one has buckled down to do the difficult work of choosing which bureaucratic requirements to keep and which to discard.

The federal government should do exactly that by ruthlessly scrutinizing all the administrative burdens created by NIH itself, the White House, and Congress. The goal should be to reduce the burden to 20 percent of researchers’ time in the next three years.

Experimentation Across NIH to Improve Innovation and High-Risk, High-Reward Research

The HELP white paper correctly notes that NIH’s peer review process can be biased against truly innovative research, which often seems too risky or too outside the box for peer review committees.³⁶ An additional source of bias against high-risk, high-reward research lies in the fact that the median age of NIH first-time grantees (which is considered the mark of launching a successful biomedical career) is 42 years old.³⁷ Throughout history, many of the greatest breakthrough ideas came from scientists in their early 20s, when they are arguably more creative and less tied to established ideas. For example, James Watson was 25 years old when he co-published the paper on the structure of DNA in 1953, and Einstein’s best year was when he first published on the theory of relativity at age 26.³⁸

The HELP Committee sensibly recommends that in order to improve innovation, “NIH should pilot and evaluate multiple approaches to changing the application and peer review process, including staffing committees with more generalists, streamlining application discussion of methodology and preliminary results, regularly reviewing the focus and members of study sections to promote alignment with current science, and improving training of peer reviewers to make reviews more consistent.”³⁹

There are many open questions about how best to fund scientific research, conduct peer review of grant applications, and entice more young geniuses into biomedicine rather than finance or high tech. At NIH’s size and scale, it could conduct many randomized experiments that would develop the world’s best evidence on these questions. For example, NIH could randomize some of its study sections (peer review groups) to give higher scores to grants where reviewers strongly disagreed with each other than to grants where reviewers all agree and then check the results against the other study sections.

There are many other examples of possible experiments, including randomizing the length and size of grants, awarding grants for scientists to switch fields (e.g., to switch from molecular biology to cancer therapy, which might introduce new perspectives into a different field), and researching the ways money is steered to early-career researchers.

NIH should reform its peer review process and launch more experiments with peer review. Following the National Science Foundation’s example,⁴⁰ NIH’s Center for Scientific Review could launch a series of ambitious experiments on different models of peer review.

As of now, there is no one at NIH focused on a thoughtful experimentation with NIH’s own processes. The NIH Common Fund was intended to support cross-institute collaboration but has a small budget. To promote experimentation and innovation, NIH could establish a dedicated Center for Experimentation and Innovation to pilot new funding models. Alternatively, it could partner with the Office of Evaluation Sciences at the General Services Administration, which brings over a decade of experience conducting government-wide research with top academic experts.

Either way, NIH is well-positioned to engage in a thoughtful process of testing, experimentation, and self-improvement. Congress or the White House should require it to do so.

Indirect Costs

Indirect costs cover general facility and administrative expenses and are paid as a percentage of certain direct costs of awarded grants. In simplified terms, if NIH awards of a grant of $100 to a scientist, and the university’s indirect cost rate is 50 percent, NIH will send $100 to support the scientist’s work directly, and another $50 to support the university’s costs for keeping the lights on, administrative compliance, and more—for a total of $150.

At the same time, it is important to clarify that while the indirect cost rate is a percentage of direct costs, it is by definition a lower proportion of the total spending. In the example above, $50 out of $150 was spent on indirect costs, or 33 percent. As it stands, NIH spends one-fifth of the current $49 billion budget on indirect costs,⁴¹ and out of all the funding sent to external grantees, the percentage going to indirect costs is around 28 percent.⁴².

Indirect costs were limited to “8% in 1950, then revised to 15% in 1958, and then raised to 20% in 1963,”⁴³ so it is clearly possible for universities to function with indirect costs at that level, and it is possible that the increased availability of indirect costs allowed universities and the NIH itself to become more bureaucratic over time. According to a 1979 GAO study, immediately following the lift of the caps the government experienced a “rapid increase” in indirect costs for health-related research.⁴⁴ A 2016 GAO report revealed that the indirect cost rates set by three primary agencies on NIH’s behalf had an “increased risk” of including “costs that are not allowable, allocable, and reasonable, and may result in wasted federal resources.”⁴⁵

Given the rise in indirect costs over the decades—and the possibility that university bureaucracies are too top-heavy—there have been a number of proposals to cap or otherwise limit indirect costs. The House white paper recommends that the federal government could tie “the indirect cost rate to a specific percentage of the total grant award, either universally or for certain designated entities,” cap indirect costs “at a graduated rate dependent on a recipient’s overall NIH funding,” or provide “incentives or preferences to recipients with established and proven lower indirect costs.”⁴⁶

At least some of the increase in indirect costs over the past 60 years is due to increased federal regulation and bureaucratic requirements. Figure 3, which is from the Council on Government Relations (COGR), shows the combined regulations and policies affecting federal research since 1991.

Federal regulation and bureaucratic requirements have soared—with little return on investment or positive impact on scientific breakthroughs. The cumulative total of regulations and policies affecting federal research increased exponentially in the past 20 years, according to the COGR Study.⁴⁷

One can imagine a win-win scenario where indirect costs would be limited to a lower level, but at the same time, Congress and the Office of Management and Budget would agree to reduce the bureaucratic requirements by at least half. Moreover, all of these agencies could agree to try to lower the burden of bureaucracy to 20 percent of researcher time. Taxpayers would get a better return on their investment if more of the money went to research and science and less to bureaucratic requirements.

On February 7, 2025—less than three weeks into President Trump’s second term—NIH announced a policy that would have capped allowable “indirect costs” at 15 percent of the overall NIH research award, saving an estimated $4 billion annually.⁴⁸ On April 4, 2025, a federal judge issued a permanent injunction blocking the policy from taking effect.⁴⁹ Regardless of the legal parameters, capping indirect costs is a sensible step to increase the efficiency of government funding for medical research and minimize universities’ bureaucratic bloat. However, federal policymakers should couple such a cap with actions to reduce regulatory compliance costs.

Policymakers can debate whether 15 percent is the right target for a necessary and efficient amount of indirect costs to accompany the grant, but without question, NIH should ensure that tax dollars support medical research and few, if any, subsidize bloated college bureaucracies.

Top-Down Mandates

In a self-aware manner, the HELP white paper acknowledges that Congress probably micromanages NIH too much by tying appropriations to top-down mandates in particular areas: “too much direction can warp the portfolio as a whole, dissuading investigators from pursuing other research topics that could have broad benefit and to create a naturally diversified portfolio.”⁵⁰

Top-down mandates from Congress are rarely effective when it comes to science—the universe does not easily bend to congressional initiatives. Look at how much money has been plowed into the NIH’s initiative on Alzheimer’s, whose website claimed as recently as 2024 that it would “prevent and effectively treat Alzheimer’s and related dementias by 2025,”⁵¹ which obviously has not happened.

Most of the greatest breakthroughs come when scientists explore areas at the boundaries between two fields or when they delve into areas that seem irrelevant. Consider Barry Marshall’s Nobel Prize for discovering that ulcers can be caused by bacteria⁵² or Thomas Brock’s vacation in Yellowstone that led to discovering a key bacterial enzyme that was essential to PCR tests, one of the most widely used biological discoveries of all time.⁵³ NIH would probably have a better chance of solving Alzheimer’s by putting money into botanical explorations throughout the world⁵⁴ or by funding studies on naked mole rats⁵⁵ or elephants⁵⁶ (both of which hardly ever get cancer) rather than another “Alzheimer’s Initiative” or “Cancer Moonshot.”

Metascience and Data

The HELP white paper correctly notes that good metascience on NIH’s decision-making is lacking and that outside analysts need better access to NIH’s internal data. There are far too many accomplished researchers who are perennially frustrated at the difficulty of getting access to NIH’s internal data or getting permission to publish their results.

There is no reason for NIH to be so stingy about sharing data with outside researchers, including the National Academies. For instance, with one National Academies report commissioned by Congress on NIH’s small business program, NIH refused to share data pursuant to a congressional mandate, citing “confidentiality concerns.”⁵⁷ This is remarkably intransigent. There are many ways to protect confidentiality while still allowing research to occur.

Moreover, there are multiple cases where top scholars were allowed access to NIH’s internal data but were afraid to publish their findings or were actually vetoed by NIH because NIH did not like the results.

NIH should be required to share its internal data with many more outside scholars and should be banned from vetoing publications except in cases of indisputable incompetence in the research.⁵⁸

Null Results and Scientific Honesty

A huge problem with the scientific literature is publication bias—the fact that both scholars and journals are biased toward publishing exciting, positive, and politically correct results.⁵⁹ To be sure, exciting results likely make a bigger splash than do results that are null or failed. But if the latter are rarely published, then the public is likely to overestimate the effects of interventions (such as pharmaceuticals) and underestimate the value of nonintervention.

The HELP white paper notes that science benefits when researchers publish null or negative results, but that does not happen nearly often enough: “NIH should consider strategies to encourage the voluntary sharing of negative results, such as the establishment of a negative results repository within the National Library of Medicine.”⁶⁰

Indeed, NIH should go further and pilot a requirement that any principal investigator on an NIH grant must report any significant failures or null results from the laboratory and must share that document on a public server such as bioRxiv. NIH could also do random audits of researchers who have never reported negative/null results in the past.

A complete absence of null results may indicate selective reporting or misconduct—no researcher, even Einstein himself, is immune to inconclusive findings. Indeed, one regular feature of any exposé on a scientist who was committing fraud for years is that his peers say something like “it did always seem amazing that he could produce such a solid string of positive results.”

Replication and Fraud Detection

Large-scale reproducibility projects have identified significant problems with reproducibility. As one major project in cancer biology found, replication experiments returned results that were fully 85 percent smaller than the original experiments, and only 46 percent of the findings “replicated successfully on more criteria than they failed.”⁶¹

The HELP white paper argued that “Congress and NIH should identify opportunities to leverage [new] technologies, coupled with data sharing policies, to restore public confidence in research claims. Congress should also review the statutory authorities of HHS’ ORI to identify any areas that could be strengthened, clarified, or updated to reflect current research misconduct challenges.”⁶²

Yet NIH has never systematically funded replication experiments. That should change. NIH should proactively fund efforts at both replication and fraud detection and deploy machine learning, artificial intelligence, and other advanced algorithms to determine the replicability of the scientific literature.

Moreover, scientific fraud is significant in some fields, such as Alzheimer’s research. Indeed, NIH’s own director of Alzheimer’s funding, Eliezer Masliah, recently had to be fired because of “falsification and/or fabrication” in his own studies.⁶³ Yet no one systematically and proactively looks for signs of fraud. Academic fraud like Masliah’s is typically discovered only years or decades after the fact, usually by online sleuths and amateurs.

Via the Director’s Office, NIH should be required to spend at least 0.1 percent of its overall yearly budget on systematic replication studies and fraud detection efforts across a variety of fields—for example, cancer, cardiology, Alzheimer’s, and more—whether as special-purpose projects or as part of ongoing initiatives. That would be a mere $49 million or so compared to a $49 billion overall budget. But it would be of immense importance to improve the credibility of NIH-funded research.

Commercialization

In 2023, NIH’s Small Business Innovation Research (SBIR) and Small Business Technology Transfer (STTR) programs funded 914 projects at a total cost of $544 million.⁶⁴ However, a 2022 National Academies evaluation found that these programs are inefficient and recommended that NIH stop treating SBIR/STTR projects as “research” grants that require peer review by the usual panels that lack expertise in biotech, startups, or industry.⁶⁵ The HELP white paper proposed a number of sensible reforms to make SBIR and STTR more successful, such as speeding up review and “hiring individuals with venture capital or other business expertise to serve as program officers.”

To improve NIH’s ability to assist entrepreneurs in the launch of new biotech startups, these programs should be overhauled to (1) include significant outside expertise, (2) make bigger awards—initial awards are only around $300,000, which is nowhere near enough to launch a company—and (3) make those awards much more quickly. NIH should consider hosting top venture capitalists as “executives-in-residence” to provide policy advice and oversee reforms.

Rather than look to reform these programs, Congress could consider scrapping the mandate that NIH engage with these biotech startups in the first place. NIH may not be the best vehicle for connecting basic research to startups that take treatments to market.

Reorienting Biomedical Academia Away from Soft Money

Across biomedical academia, universities expect many professors to raise a substantial portion (sometimes the full amount) of their own salaries through grants. Even long-time tenured professors can be at jeopardy of losing part or all of their salaries if they fail to bring in enough grants. This “soft money” system can discourage the best and most innovative science, because it incentivizes individual scientists to stick with the most risk-averse ideas that can get funded by NIH.

Soft money can create serious conflicts of interest as scientists adhere to their funders’ wishes in the interest of not losing their jobs, labs, or salaries. As a scientist at Stanford said, “Universities are really involved with tracking conflicts of interest like outside funders (pharma), etc. But nobody tracks the real conflict of interest: Part of your salary depends on getting grants! That means putting food on the table for your family. It’s a huge existential threat!”⁶⁶ Or as an NIH IC director said in an interview, “The salary issue is one that I feel very strongly about. The growth of soft money salaries has been very negative. It’s reasonable for people to get a certain amount of salary from grants. But to be paying 70-90 percent of someone’s salary isn’t healthy for the system or for those investigators. You’re going to be much more careful about taking risks or about publishing honest conclusions if you’re worried about your salary. The university’s investment in a person and their commitment becomes less if they’re not salaried. That changes the sociology and the culture if everyone is on soft money.”⁶⁷

Many universities and medical schools financially depend on the soft money system. Any solution would be difficult and would require careful thought and execution over many years. The HELP white paper suggested that Congress and NIH should “explore this problem in further detail and identify policy options that ensure NIH and institutions each provide appropriate degrees of support for federally funded academic researchers.”⁶⁸

Bruce Alberts, former National Academies president and former editor of Science, already proposed a partial solution: NIH could “require that at least half of the salary of each principal investigator be paid by his or her institution, phasing in this requirement gradually over the next decade,” or that “the maximum amount of money that the NIH contributes to the salary of research faculty (its salary cap) could be sharply reduced over time, and/or an overhead cost penalty could be introduced in proportion to an institution’s fraction of soft-money positions (replacing the overhead cost bonus that currently exists).”⁶⁹

NIH should form a committee with at least 51 percent non-NIH representation with the goal of creating a plan to reduce academia’s use of soft money requirements over the next five to 10 years.

Stop Requiring Preliminary Results⁷⁰

There is a well-known but unwritten convention around the way to secure funding from NIH⁷¹:

1. Do a series of experiments and select the successful ones.
2. Submit a grant proposal to NIH in which you present “preliminary data” from those experiments.
3. Get a grant.
4. Report back to NIH that your experiments were a success.
5. Meanwhile, use the new round of grant money to do further experiments in which you pick the successful ones to be the basis for your next NIH request.

This dynamic may appear as a performative charade, but it is driven by NIH’s requirements for preliminary data in grant applications. The initial NIH peer review process revolves around researchers demonstrating enough “preliminary data” to show that future research is likely to produce positive results. For example, one NIH webpage advises that “when you’re crafting your grant application, high-quality preliminary data can make all the difference.… You must also assess whether or not your preliminary data are sufficient to convince reviewers that your project has a high likelihood of success.”⁷²

Science worth pursuing is, by its nature, uncertain—true discovery begins where success is not guaranteed. At the frontier of research, the value lies not in predictability but in the process of uncovering the unknown. But in the NIH system, researchers are constrained to ask for funding only for ideas that they already know will work. As one top scientist said, “It’s an open secret that people write grants for stuff that they’ve already done—they know it’s feasible and has good preliminary data, and by the time it gets funded, they can take the money and do something else.”⁷³

This approach stacks the deck against any new and innovative research projects. After all, any truly new idea will not have years of “preliminary data” showing that it already works—by that point, it is an old idea, not a new idea.

Congress needs to emphasize to NIH that it should seek to fund new and innovative research, not just research that is already halfway complete, and that NIH should stop demanding “preliminary data” for almost all grants. It should reform its peer review processes to reduce the reliance on preliminary data and should do so in a rigorous experiment so as to test the results over time.

Fund Graduate Students and Postdocs More Fairly

NIH funds R01 grants and similar grants for many graduate students and postdoctoral researchers working for principal investigators (PIs) on those grants. This structure leads to problematic results as universities admit too many graduate students compared to what can be supported in the academic marketplace, and PIs hire too many postdocs as a form of cheap labor.

The numbers tell a depressing story:

• PhDs awarded in biological and biomedical science in 2021: 8,149.⁷⁴
• Postdocs supported by NIH grants in 2022: 28,953.⁷⁵
• First-time R01s that NIH awarded to new researchers in 2021: 1,513.⁷⁶

In sum, NIH grants collectively fund three times more doctoral candidates than could get an academic job with an NIH grant, and there is a 19-year backlog of postdocs competing for new academic jobs.

While postdocs play a vital role in advancing science, the current system often leaves them spending many years in temporary roles with little pay and uncertain career prospects. NIH should consider limiting NIH-funded postdoc appointments to two years. This would reduce the likelihood of early-career researchers becoming stuck in extended postdoc roles that delay career advancement and offer little long-term stability.

These changes may be met with resistance, particularly from institutions that have grown reliant on large pools of low-paid postdoc labor. But they would reflect the significant expertise of postdocs and discourage overreliance on a labor model that often delivers diminishing returns for both researchers and science itself.

Just as important, NIH should require greater transparency around career outcomes. Currently, too few universities provide clear data to trainees about the job prospects of their past postdocs. Without such information, early-career scientists may enter long academic training pipelines without understanding how slim the odds of a tenure-track role may be. NIH should mandate that all institutions with five or more NIH-supported graduate students or postdocs publish, at minimum:

• average time to PhD;
• average length of postdoc positions; and
• career outcomes at 1, 5, and 10 years (e.g., academic, industry, or other roles).

Calls for this kind of transparency date back decades—including recommendations from the National Academies in 1998⁷⁷ and an NIH working group in 2012⁷⁸—but progress has been minimal. As Princeton professor and president Shirley Tilghman noted, voluntary guidelines have failed: “Unless you have a stick, this won’t happen.”⁷⁹ Now is the time to provide that stick.

Funding More Diverse Organizations

NIH’s extramural funding has flowed almost entirely to traditional universities such as Harvard and MIT, academic medical centers such as MD Anderson Cancer Center, and long-established entities such as Cold Spring Harbor Laboratory.

Yet there are many new types of organizations that deserve support. To take just one example, “focused research organizations” (FROs) have emerged in the past few years. An FRO is tightly focused on a single mission, has a limited time frame, and seeks funding as an organization rather than as an individual scientist. These organizations have shown potential, as demonstrated by recent efforts to significantly reduce the cost of brain mapping.⁸⁰

NIH should establish a new program to incubate and fund new types of organizations, including FROs. NIH could issue an annual call for proposals, selecting five to 10 projects for initial funding. It would then provide multi-year, milestone-driven grants, with flexibility for the occasional pivot. And it would seek to partner with private foundations and philanthropies to co-fund and sustain successful organizations.

Not all biomedical science needs to be performed within the traditional universities and similar entities. Better results could be achieved by spreading the money to other organizations with different models and different ideas.⁸¹

Nearly all of this paper’s proposed reforms, even the structural ones, can be accomplished directly by the HHS Secretary, the NIH director, or IC directors. Congress has already been quite specific in delineating who has the power to enact certain types of NIH reforms in Title 42, Section 281(d), of the U.S. Code:

Nearly all of the other reforms—such as adding more experimentation with peer review, launching replication projects, and trying to prevent academic fraud—can be accomplished by NIH directly.

That said, there is a systemic problem with NIH-wide initiatives: The NIH director actually lacks the authority to require all the ICs to work together on, say, a joint data project for all of the more than 100 repositories that currently exist. Moreover, a number of ideas are likely to be unappealing to NIH, because they might bring embarrassing information to light. (For example, a search for academic fraud might find that more NIH-funded scientists had been engaged in fraud.) Thus, many reforms may be much more likely to happen with congressional legislation (or appropriations language) or a White House executive order.

The nation stands at a pivotal juncture for reforming NIH by refocusing its mission and making it more efficient. NIH and HHS leadership should embrace the structural reforms identified above. These reforms would not only bolster scientific integrity and accelerate breakthroughs but also help restore public trust in America’s leading health institutions.

NIH collectively represents the world’s largest funder of biomedical research, with a 2024 budget of nearly $49 billion—nearly 1 percent of the total U.S. federal government budget,⁸² with 27 ICs and over 100,000 funded projects or grants in place at any given time.⁸³

Yet it came from humble beginnings: NIH webpages regularly claim that NIH actually “traces its roots to 1887, when a one-room laboratory was created within the Marine Hospital Service.”⁸⁴ A one-room laboratory is a far cry from what NIH looks like today. How did it get from there to here?

In 1891, the “Hygienic Laboratory,” with its one staff researcher, moved to Washington, DC, but it was not until 1901 that Congress awarded the laboratory $35,000 .⁸⁵ Skipping over a few details, the Marine Hospital Service was ultimately renamed the Public Health Service in 1912, and a congressional bill authorized the lab to do research on diseases and pollution.⁸⁶

The Public Health Service and the lab occupied themselves with issues such as sanitation at military bases and treating influenza patients in 1918.⁸⁷ In 1930 the Ransdell Act renamed the laboratory as the National Institute of Health,⁸⁸ and in 1937 Congress created the National Cancer Institute (NCI, which later officially became part of NIH).⁸⁹

Even then, NIH was only a fraction of the expansive institution it is today. The annual appropriations for NCI were a mere $400,000 in 1938 and rose only to $549,000 by 1946. (At the same time, a division that ultimately turned into the National Center for Research Resources had a $2,866,000 budget in 1946). In the post-war years, however, appropriations increased dramatically: $14 million for NCI in 1949 and $14.5 million for research resources.⁹⁰

By 1950, NIH included NCI and the forebears of the National Heart, Lung, and Blood Institute; the National Institute of Mental Health; and the National Institute for Dental and Craniofacial Research with a total budget of $52.7 million. By 1959, several other institutes had been added, and the budget was $291.8 million.⁹¹ And by 1969, under the leadership of famed NIH Director James Shannon, more ICs had been added, and the total budget exceeded $1.1 billion.⁹²

As time went on, Congress added more ICs and reorganized the existing ones. By 1998, there were 23 ICs, and the overall budget was $13.7 billion.⁹³

The ICs proliferated and evolved. Consider one example of many: In 1948, Congress established an “Experimental Biology and Medicine Institute,” but by 1950, that had transformed into the National Institute of Arthritis and Metabolic Diseases.⁹⁴ In 1972, it became the National Institute of Arthritis, Metabolism, and Digestive Diseases, and in 1986 it became the National Institute of Arthritis and Musculoskeletal and Skin Diseases, which still exists.⁹⁵

Around the turn of the century, NIH doubled in size. With wide bipartisan support, Congress aimed to double NIH’s budget within five years. In 1999, Newt Gingrich, who had just stepped down as the Speaker of the House of Representatives, wrote in The Washington Post, “The highest investment priority in Washington should be to double the federal budget for scientific research. No other federal expenditure would create more jobs and wealth or do more to strengthen our world leadership, protect the environment, and promote better health and education for all Americans. For the security of our future, we must make this investment now.”⁹⁶

Over five years, the total NIH budget grew from $13.7 billion in FY1998 to $27.2 billion in FY2003, with annual funding increases averaging 15 percent.⁹⁷

The spending surge could not last for much longer, particularly given the worsening economy and other priorities such as post-2001 spending on terrorism and military engagements. Thus, the NIH budget flattened out and even declined for over a decade in inflation-adjusted dollars. Appendix Figure 1 shows the NIH budget from 1998 to 2024 in both nominal dollars and 2024 inflation-adjusted dollars.

NIH spending sharply escalated from 1998 to 2003, then declined (in real terms) from the 2003 peak for more than a decade. NIH spending finally started to increase in real terms in the final year of the Obama administration (2016) and under President Trump. In real terms (which we cite largely because of the high inflation of the Biden years), NIH spending has flattened out since 2020.

During this period, NIH has scaled up to 27 ICs (this number was capped in statute in the NIH Reform Act of 2006), plus the Office of the Director, which effectively leads a “catch-all” of NIH-wide initiatives and grant programs that span NIH or are not a good fit within any of the ICs.

Footnotes