We at Science-Based Medicine have been sounding the alarm about unscientific and anti-scientific medicine and medical claims since 2008. In this endeavor, we have tried to remain as nonpartisan as possible, although it is a fool’s errand to try to remain apolitical, because medical science, like any science with incredibly important practical applications that result in government regulation and policies impacting it, cannot avoid intertwining with politics. After all, much of healthcare, especially for older Americans, is government-funded. So is a large amount of biomedical research, thanks to the National Institutes of Health. As those of you who’ve read SBM over the years know, I’m quite unabashedly pro-NIH, viewing it, for all its admitted shortcomings (some of which I’ve discussed over the years), as the single greatest engine of biomedical discovery ever constructed in history. I’ve also been unapologetic in arguing that much of the reason for that success is how, since World War II especially, the NIH has been designed to be as insulated from the rough-and-tumble of partisan politics as any government institution can be.
While it is true that each new Presidential administration, along with Congress, gets to choose the leadership of the NIH and decide its broad research priorities, in general, the actual day-to-day decisions regarding which science, as represented by the research grant applications received by the NIH every year, has been left to scientists. Applications go through rigorous peer review by scientists at NIH study sections and, as a result, receive a priority score. Advisory Councils staffed by scientists then rank the applications by score into percentiles and fund as many grants as the budget will allow, for the most part starting with the highest scored grants and working their way down. True, sometimes grants are funded out of order for various reasons, such as wanting to balance a research portfolio, but for the most part it is a combination of the scientific rigor of an application that decides which grants are funded. We can argue—and have argued—whether the NIH peer review system functions as well as it should to separate the wheat from the chaff. We can—and do—point out deficiencies in the NIH system for determining which grants are funded. However, I argue that the current system is nonetheless much better than anything proposed by the Trump administration, which, as I’ve decided, is trying to promote a new set of rules that would de-emphasize scientific peer review and place the power to decide which grants are funded much more firmly in the hands of political appointees.
There is, however, a strain of “free market” libertarianism/conservatism that wants to go beyond just, as I have described it, turning NIH funding into a system to support administration-approved science and punish scientists who dissent. They want to abolish the NIH completely and leave biomedical research to the “free market.” One such ideologue is Dr. Scott Atlas, a radiologist and senior fellow at the right wing think tank know as the Hoover Institution at Stanford University who rose to prominence during the COVID-19 pandemic as a COVID “contrarian,” minimizing the danger of COVID in his belief that public health measures that interfered with business (e.g., “lockdowns”) were anathema and must be avoided, which led him to be for a time a special advisor to President Trump on the COVID-19 pandemic in 2020. Last week, Dr. Atlas published an op-ed in The Washington Post entitled, Abolish the NIH, with the tagline, “A $48 billion government monopoly should give way to the market.”
Oh, goody.
You might laugh at such an argument, and you might well laugh after seeing how flimsy and deceptive Dr. Atlas’ arguments in favor of abolishing the NIH are. However, his argument that the NIH is hopelessly politicized to the point where it would be better to abolish it in favor of letting the “free market” handle funding scientific research is no laughing matter
The NIH is both a monopoly and isn’t important?
The first thing I noticed about Dr. Atlas’ arguments was the contradiction at the heart of it. He claims that the NIH has a “monopoly” on funding biomedical research but then tries to argue, in essence, that it would be no big deal if the NIH ceased to exist because private companies fund most biomedical research in the US to begin with. Seriously, I kid you not. Dr. Atlas starts by arguing that the NIH is a “monopoly,” naturally quoting Milton Friedman to do it:
In 1980, economist Milton Friedman said the National Institutes of Health should be abolished. Friedman said the same about another government research agency, the National Science Foundation. And when he was asked what the NSF should be replaced with, he replied: “Nothing.”
For those of you unfamiliar with Milton Friedman, he was one of the most famous and influential prototypical 20th-century “free market über alles” guys who argued for the privatization of nearly everything, including the elimination of government licensure of physicians. It’s not surprising that he would think that the government shouldn’t be funding scientific research, be it in biomedicine or other disciplines. He was in favor of abolishing everything in the Department of Health and Human Services unrelated to “some public health activities to prevent contagion,” while saying, “No, not the National Institute. No, those are mostly research agencies. No. No. That’s a question of whether the government should be involved in financing research, and the answer is no.” He was also in favor of abolishing the Food and Drug Administration, a very bad idea favored by some libertarians that I’ve discussed a number of times. Friedman’s position on government-funded research was ideological, not science-based, and so is Dr. Atlas’, as you will see.
First, however, after having claimed that the NIH holds a “monopoly” on biomedical research, Dr. Atlas then pivots to argue that the NIH is no longer very important relative to what the “free market” and private institutions are already doing:
Abolishing the NIH would not create a funding vacuum. The private sector funds 78 percent of U.S. biomedical research and development. Venture capital has exploded. Investment focused on artificial intelligence in drug discovery alone surged nearly 36-fold between 2010 and 2024 — funding the overwhelming majority of innovative medical devices and building the AI revolution in biomedicine. The Howard Hughes Medical Institute, the Wellcome Trust, the Gates Foundation and the Chan Zuckerberg Initiative represent more than $150 billion in endowed capital. Nokia Bell Labs, funded by AT&T and then Nokia private revenue, generated 10 Nobel Prizes.
Wait, what? Make up your mind! Either the NIH is a “monopoly” on funding biomedical research, or it’s a relatively insignificant funder. Which is it? Nuance, of course, is not what Dr. Atlas is about. It’s also amusing that he even mentions the Gates Foundation here, given that he accuses it of having spent “two decades steering the NIH research agenda toward its own priorities — with agency officials as willing partners,” citing an “investigation” from a questionable (and arguably biased) source, Paul Thacker (yes, that Paul Thacker) as “evidence.” It’s also rather amusing that he cites the Gates Foundation, along with the Wellcome Trust, both of which are attacked in Thacker’s hit piece for supposedly having too much undue influence on NIH policies, as two of the potential private sector saviors who will fund research if there is no NIH. Let’s just say consistency and coherence are not major features of Dr. Atlas’ arguments.
At this point, I had to ask: How much per year is funded by that endowed capital touted by Dr. Atlas? Remember, an endowment exists to use the income stream from what it is invested in to fund whatever the endowment is set up to fund.. The principal itself is not supposed to be spent, except under special circumstances spelled out in the endowment agreement. After all, the idea is that the endowment will provide a continuous source of funding that will last indefinitely, something it can’t do if the principal is spent down. In any case, the combined returns on the funds in the combination of endowments cited by Dr. Atlas are not likely to be funding anywhere near ~$50 billion a year (the NIH budget right now). Later in the op-ed, Dr. Atlas touts investments in basic “high risk” scientific research to be funded by private foundations:
Howard Hughes Medical Institute funds investigators without concern about risk or yield. Wellcome has committed $21 billion through 2032, and Chan Zuckerberg Initiative has pledged at least $10 billion over the next decade, for basic or high-risk science.
The HHMI is a great institution, as are the others, and I do love to see private foundations funding higher-risk science. Did you notice something, though? If you do the math for Wellcome through 2032, assuming Dr. Atlas means the years 2027-2032, that’s $21 billion over six years—or just $3.5 billion per year through 2032. As for the Zuckerberg Initiative, $10 billion over a decade is only $1 billion a year. That’s only $4.5 billion a year, maybe a little more if Zuckerberg pledges more. That’s only one-tenth of the NIH budget. I note that Dr. Atlas doesn’t mention how much the Gates Foundation or other foundations have pledged, which makes me suspect that he knows it’s not very much compared to the NIH. Of course, the way Dr. Atlas phrased this not-insignificant commitment makes it sound like a lot more than it is, particularly compared to the NIH, which funds several times this much basic research, year in and year out, and has been doing so for eight decades.
It’s even worse than that, though. Dr. Atlas seems to be implying that Wellcome and other foundations can make up any shortfall in research funding that would result from abolishing the NIH. What he neglects to tell the reader directly is that the Wellcome Trust is a London-based UK trust that funds global research. (Indeed, its webpage cites committing £16 billion, which Dr. Atlas converted to American dollars without informing WaPo readers.) Its commitment is not to fund primarily US-based research. Wellcome funds research in the UK and around the world. Thus, only a fraction of the money touted above committed by Wellcome to high-risk research might reasonably be expected to fund American researchers. Seriously, how did this deceptive of an argument get past WaPo editors? All you have to do is click on the link that Dr. Atlas provided to realize how deceptive he’s being. As an aside, one of the first things I do after reading a guest post submission to our little blog is to click on every link used to back up a guest blogger’s arguments, to see if the articles cited actually say what the guest blogger says they say and to see if they actually support the blogger’s arguments. This is Editing 101.
Here’s the funniest thing, something that made me chuckle when I noticed it. Dr. Atlas’s reference supporting the conclusion that the private sector funds 78% of biomedical research and development links to an article published by Victor Dzau and Keith Yamamoto entitled—ironically—Government-Funded Health and Biomedical Research Is Irreplaceable. As the title suggests, this commentary makes exactly the opposite arguments to the ones that Dr. Atlas is trying to make, something that Dr. Atlas surely knew when he included the link. Indeed, the authors argue that, even with private sector funding growing as a percentage of total funding of biomedical research in the US, NIH funding is nonetheless still indispensable to the US biomedical research enterprise. The authors even call government funding of research “unique in undergirding the science and technology enterprise in at least four distinctive ways” (more on those distinctive ways later). Truly, Dr. Atlas must think his readers are idiots. (Actually, he probably knows—as does any blogger or web designer who pays attention to the click-through rates on links included in posts—that only around 1-2% of readers will bother to click on any given link embedded in a paragraph like that. Indeed, a 2% click-through rate is considered quite high and a successful link placement. I’d be willing to bet that Dr. Atlas knew damned well that the vast majority of readers wouldn’t bother to click on the link.)
But, thanks, Dr. Atlas! Later in this post I will quote liberally from your source as I address the rest of your deceptive arguments. (I do so love it when the crank I’m dealing with provides me evidence and ammunition to argue against him.)
In the meantime, I also can’t help but note Dr. Atlas’ focus on artificial intelligence, which is currently the darling of tech bros everywhere. There is no doubt that AI is going to be very important going forward, but, unlike the large language model (LLM) chatbots and other forms of generative AI that are in the news, AI use in biomedicine has been a thing for quite a few years, in particular for predicting protein structures and for drug design. And that’s (mostly) a good thing. It is not, however, the be-all and end-all of biomedical research (or at least, the bulk of biomedical research), as Dr. Atlas seems to imply. For example, for these AI predictive models to become effective, they have to be trained on the findings of basic science, and where will that basic science come from? Moreover, the article cited by Dr. Atlas notes:
While AI-related biopharmaceutical companies represent a minority of overall deals, they have experienced substantial growth, especially in discovery tools, where funding surged nearly 36-fold from 2010 to 2024. The majority of this growth occurred during and after the COVID-19 pandemic, in which VC funding experienced significant volatility [8]. AI-related VC activity expanded to 19 states by 2024, though investments remained highly concentrated in California and Massachusetts. These trends suggest that AI is an emerging investment focus but may be influenced by regional innovation ecosystems and biopharmaceutical research resources.
See what I mean about AI-driven drug discovery having been a thing for a while? In any event, that biomedical research using AI has seen explosive growth in venture capital funding over the last several years is not an argument against the NIH or for the contention that the NIH is not necessary. First, these AI companies are a minority of deals, and, even if they weren’t, the NIH doesn’t work this way. It’s designed to fund basic and translational research that can then be turned into medicines, treatments, and other products, the latter task being more suited to private industry. Also, with a few small AI companies sucking up tens—or even hundreds—of billions of dollars in venture capital, the NIH can’t compete in that space with a yearly budget of under $50 million.
Finally, with respect to whether the NIH is important or insignificant, Dr. Atlas notes:
Look inside the grants at where taxpayer money actually goes. The NIH paid universities approximately $9.5 billion in fiscal 2026 to cover indirect costs such as facilities and administration expenses. These same universities accept grants from private foundations with little or no funds provided for overhead. The Gates Foundation caps its university overhead at 10 percent — a rate universities accept without protest. The disparity is easily explained: Private foundations say no, while the government readily spends other people’s money with no accountability. When the NIH attempted to impose a 15 percent cap in 2025, universities sued — and won.
I’ve discussed the issue of indirect costs covered by NIH grants and—for that matter—most government grants. In this, he is echoing a common right-wing attack on indirect costs as a wasteful boondoggle when they are anything but that. Even so, I will concede that the issue is complex, and there are arguments that the indirect cost rates negotiated by various universities might have risen too far, but Dr. Atlas is, once again, making a simplistic argument, adding to it more false equivalence.
By way of background for those not familiar with NIH grants, indirects are the costs over and above the dollar amount of direct costs that go to fund the research. Basically, the part of the budget for indirect costs goes to the university to help fund the infrastructure used by its researchers, including labs, costs involved with maintaining core facilities for researchers to use, as well as the bureaucracy necessary to administer the grants and to monitor compliance with federal regulations. I would never deny that there is a legitimate argument to be had over whether indirects can be excessive and add way too much to the cost of research funded by grants. In some cases, they can be 70% of the direct costs added to the overall grant. However, it must also be noted that indirect cost rates are periodically negotiated by each university with the NIH and that the NIH can therefore negotiate lower rates if it decides that it is not getting its money’s worth from any given university. Echoing past attacks on indirects, Dr. Atlas is trying to portray NIH funding of indirects as gross waste in a deceptive comparison with the indirects negotiated on grants with private foundations. I’ll give you a hint here. Universities can afford to accept the much lower (or even nonexistent) indirects on grants to their researchers from private foundations because the NIH funds indirects at a much higher level. One can argue whether the NIH in effect subsidizing the lower indirect rates paid by private foundations is good public policy or not, but that’s not the issue that Dr. Atlas is addressing. Instead, he’s deceptively arguing that universities should be able to accept much lower indirect cost rates associated with NIH grants because they accept them with grants from private foundations. Therein lies the false equivalence, which can sound convincing if you don’t know anything about how NIH grants are administered.
Dr. Atlas’s misleading arguments get worse.
Exaggerating the influence of politics on NIH
The other major claim made by Dr. Atlas is that the NIH is hopelessly political. As I like to say, no one—and I mean no one—argues that the NIH is above politics. However, the NIH is close to unique in that there has been for decades a strong bipartisan consensus that funding science based judged to be the best science by actual scientists is a good thing for the nation. And they’ve been right! The NIH is a large part of the reason why the US has risen to become such an international juggernaut in biomedical research. So, I actually agree with Dr. Atlas that politicization of the NIH is harmful, which is why I’ve written a couple of posts lamenting how the Trump administration is trying to pass rules that would devalue the role of scientific peer review in deciding which grants are funded while giving way more power to political appointees than they’ve ever had to decide which grants are funded. If these rules pass, then Trump policies mandating that only grants that advance administration priorities will be funded and that political appointees will be able to kill at any time funding of grants that the administration does not like will be codified permanently.
To his credit, Dr. Atlas does mention the same politicization of the NIH under President Donald Trump and HHS Secretary Robert F. Kennedy, Jr. that I’ve been warning about:
Then the NIH reorganized around a new political agenda under President Donald Trump. Last year, his administration terminated over 5,100 grants worth more than $4.4 billion on political grounds. Courts intervened. A federal judge ordered the grants to be reinstated. The Supreme Court, however, later allowed the Trump administration to cut nearly $800 million in NIH grants. Taxpayers fund all the ideological shifts, litigation and policy reversals.
Of note, however, this is the last example that Dr. Atlas mentions, and he engages in a lot of false equivalence, painting these instances as equivalent to actually making peer review almost irrelevant and giving the power of funding decisions to political appointees, because, you know, “DEI”:
This financial exploitation is compounded by political agendas, regardless of party. President Ronald Reagan’s administration delayed AIDS research funding; Reagan did not address the epidemic publicly until 1987, after more than 36,000 Americans had been diagnosed and nearly 21,000 had died. President George W. Bush restricted embryonic stem cell research on religious grounds. President Barack Obama reversed that ban by executive order. His administration created the NIH’s ideologically driven Sexual and Gender Minority Research Office in 2015.
President Joe Biden converted the agency into a racial and social-justice instrument: a diversity, equity, inclusion and accessibility system of almost 100 offices, committees and groups across 27 institutes and centers. His administration allocated $241 million for the Faculty Institutional Recruitment for Sustainable Transformation program to hire more diverse faculty. The money was granted to address underrepresented racial groups in science, claiming the representation gap is driven by “structural barriers” and “in large part by institutional cultures lacking necessary elements of inclusion and equity.”
It is interesting that Dr. Atlas leads with the Reagan administration, basically equating Reagan’s reluctance to address the AIDS epidemic in the 1980s with the NIH response. It is true that the Reagan administration had a horrible record on AIDS research, but that wasn’t the whole story. If you look at what really happened, part of the problem with the NIH response to the AIDS crisis in the 1980s was also largely structural. In the early1980s, as AIDS became a public health concern, the NIH system of grant funding had been set up to facilitate mainly investigator-initiated individual grants, a system that struggled to adapt to the crisis, resulting in delays in funding large research efforts. Moreover, the NIH was largely siloed from the CDC, and there was considerable difficulty in cross-agency communication and collaboration about AIDS, as well as even between the institutes of the NIH.
According to the NIH itself, which, admittedly, is not an unbiased source:
In June 1981, the Centers for Disease Control and Prevention published the first of many Morbidity and Mortality Weekly Reports (MMWR) related to the disease that would later be named Acquired Immune Deficiency Syndrome or AIDS. In 1982, the National Institutes of Health (NIH) identified associated HIV/AIDS funding. In May 1983, Congress passed the first bill with specific funding for AIDS research and treatment across the U.S. Department of Health and Human Services.
The National Cancer Institute (NCI) and the National Institute of Allergy and Infectious Diseases (NIAID) conducted most of NIH’s early HIV/AIDS research. By 1985, NIAID was the lead NIH Institute sponsoring HIV/AIDS research. However, HIV/AIDS posed many challenges which exceeded the mission of any one Institute, as it is a multi-system and multi-organ disease involving malignancies; opportunistic infections; and cardiovascular, neurological, gastrointestinal, and other complications. HIV/AIDS also affects people across the lifespan. Furthermore, behavioral and biomedical interventions are needed to prevent new infections. Thus, nearly every NIH Institute, Center, and Office (ICO) became involved in conducting or supporting HIV/AIDS research.
As the expanding pandemic required increased coordination, the Office of AIDS Research (OAR) was established in 1988. The Office was first established by the Assistant Secretary for Health and later codified in the Health Omnibus Extension of 1988 (Public Law 100-607).
Again, none of this is to say that the Reagan administration didn’t have an abysmal record on AIDS or that it didn’t impede funding for AIDS research. It most definitely did. Indeed, it’s difficult not to think, when I look at what the Trump administration is doing to scientific agencies, that I’ve seen this movie before. After Reagan took office, which coincided with the time period when AIDs was increasingly becoming a public health concern, he and his Office of Management and Budget Director David Stockman set about trying to slash the budget of nearly all federal programs other than the military, the CDC and NIH included, with Stockman calling for massive budget cuts. Indeed, that first May 1983 funding bill only allocated $2.6 million to AIDS research, and, according to the classic book on the AIDS crisis by Randy Shilts, And the Band Played On:
Congress would have to discern for itself how much money government doctors needed to fight AIDS. The administration would resist but not put itself in the position of an on-the-record funding veto. The epidemic’s research would survive from continuing resolution to continuing resolution, a game that would ultimately achieve some funding for the doctors while disabling any attempt to plan ahead for studies that might be needed as the scourge continued to grow.
So, for the moment, let’s grant Dr. Atlas the point that politicization of AIDS by the Reagan administration hampered the NIH response to the crisis, mainly by delaying funding of needed research and hampering the development of a coordinated research plan. Certainly the administration’s actions made overcoming the structural barriers at the NIH to creating integrated research programs to address the crisis far more difficult. Let’s even credit Dr. Atlas for also calling out the Bush administration’s ban on NIH and other government funding of embryonic stem cell research. Are those incidents a reason to abandon the NIH as hopelessly political and use as a pretext to shut it down? Of course not!
What Dr. Atlas is really about, obviously, is laying down a whole load of false equivalence. Notice how much verbiage Dr. Atlas expends on President Biden’s initiatives to promote greater representation of underrepresented minorities, a whole paragraph compared to the one sentence each devoted to Reagan’s hampering funding for AIDS research and Bush’s ban on government funding of embryonic stem cell research. That’s the “tell.” Dr. Atlas is trying to argue that what Biden and Obama did was equally bad to what Reagan and Bush did and to what Trump is doing now. Given Dr. Atlas’ ties to the “make America healthy again” (MAHA) movement and his prior support of the Trump administration, I will admit that it is telling just how bad the Trump administration is that even Dr. Atlas criticizes its proposed grantmaking rules for NIH (and all federal grants), but that doesn’t change the game he is playing, false equivalence based on right wing grievances against “DEI.” In any case, what Dr. Atlas is doing here is obvious. He is implying that all that dreaded “DEI” and all the efforts of Democratic administrations to level the playing field for underrepresented minorities in obtaining NIH grants, as well as to study health-related questions related to groups like transgender individuals represent a politicization of NIH science on par with what Reagan and Bush did, as well as—stretching credulity to the breaking point—what Trump is doing now.
Then, of course, consistent with his rise to prominence as a pandemic-minimizing contrarian, Dr. Atlas cries “censorship”:
Government funding of academia corrupts science in subtle ways, too: When institutions depend on the NIH — for faculty promotions, overhead and graduate-student support — dissent becomes institutionally intolerable. In the White House, I witnessed enforcement of the “harmful uniformity” that NIH Director Jay Bhattacharya, Martin Kulldorff and I described in National Review: Some physicians and scientists who questioned NIH leaders on pandemic lockdowns or masking were removed from social media platforms because of accusations that they violated misinformation policies. They were also threatened with loss of licensure by the Federation of State Medical Boards. Many stayed silent because their funding could have been at risk. The promise of NIH dollars distorts research, too. Now that Bhattacharya has declared reproducibility of research results a top agency goal, many scientists will predictably discover a dormant passion for repeating old experiments — and universities will be right behind them, operational overhead budgets in hand.
Let me address that last bit first. Dr. Atlas isn’t a fool. He knows damned well that the NIH had been looking at ways to improve reliability and reproducibility of NIH-funded research for years prior to Trump’s re-assumption of the Presidency a year and a half ago. I’ll just cite this 2014 article by Francis Collins and Lawrence Tabak on an initiative by the NIH to enhance reproducibility to show you what nonsense it is for Robert F. Kennedy Jr. and “Podcast Jay” Bhattacharya to act as though the current administration is the first one ever to address reproducibility in science and to imply that the NIH never really cared about this issue until Trump became President again. The NIH didn’t tell social media platforms to deplatform doctors and scientists for spreading “misinformation,” and the NIH has nothing to do with state medical licensure or medical specialty board certification. As for the rest of the grievances about suppressing “dissent” cited above, these are standard right-wing talking points. One notes that nowhere in that dump of a paragraph does Dr. Atlas show an example of a researcher being denied NIH funding because he was a COVID “dissident,” or even staying “silent” for fear of being denied NIH funding. Surely, now that Trump is in power, if so many scientists were afraid to speak up for fear of being defunded by the NIH, Dr. Atlas could have cited at least one example. He doesn’t. He just expects readers to take his word for it.
Exaggerating the contribution of the private sector to research relative to the NIH
If you really want to see what Dr. Atlas is about, as well as how flagrantly he’s willing to misdirect and deceive, all you have to do is to look at this passage in his op-ed:
Private capital, not NIH direction, produced many of the most celebrated medical advances of the past two decades. Checkpoint inhibitors — drugs that allow the immune system to recognize and attack tumors it ignored — have converted metastatic melanoma, lung and kidney cancer from near-death sentences into more manageable conditions. These drugs reached patients through industry investment. Semaglutide, reshaping obesity treatment for tens of millions through drugs such as Ozempic and Wegovy, was developed by Novo Nordisk’s research chemists without NIH direction.
This is utter nonsense, an easily demonstrable false reading of medical history. For example, let’s look at immune checkpoint inhibitors, something I’ve discussed before in relation both to their having improved the treatment of certain cancers, to how COVID-19 vaccination appears to potentiate their effectiveness against certain cancers, and to false claims about COVID-19 vaccines causing “turbo cancers.” They are indeed a breakthrough and have in the space of a relatively few years greatly improved the treatment of certain cancers. While it is true that industry shepherded drugs that inhibit this immune process through to becoming FDA-approved, where did the basic research come from that discovered the immune checkpoint system to begin with? It turns out that James P. Allison and Japanese researcher Tasuku Honjo were awarded the 2018 Nobel Prize in Physiology or Medicine for their co-discovery of groundbreaking discovery of a cancer therapy achieved by inhibiting negative immune regulation, which directly led to the creation of immune checkpoint inhibitor drugs. It is true that Dr. Allison is also a Howard Hughes investigator, but NIH funding was at least as important to his discovery of CTLA-4, a key regulatory molecule in the pathway targeted by checkpoint inhibitors. In his 2002 review article on this discovery, Allison acknowledges support from “the Howard Hughes Medical Institute and grants from the NIH (NCI CA40041 and CA57986) and CaPCURE (J. P. A.).”
I’ll give Dr. Atlas half a cherry-picked point. HHMI and NIH both contributed to funding the research that led to the discovery of this pathway, at least in the US. But what about his claim about drugs targeting GLP-1, such as Ozempic and Wegovy? Once again, Dr. Atlas tells only half the story. The article that he cites comes from Novo Nordisk and describes how three of its scientists sought to “create a version of GLP-1, a natural hormone that helps the body produce insulin and regulate appetite.” That’s all well and good, and no one is taking away from the achievements of these scientists, but who discovered GLP-1 in the first place and worked out its biological importance and mechanisms?
It turns out that the discovery of GLP-1 and how it can stimulate insulin release in the pancreas began in the late 1970s and early 1980s in two labs, one at Harvard University, where Joel Habener had been using then-new recombinant DNA technologies to “elucidate proglucagon amino acid sequences from cDNAs and genes isolated from anglerfish in the early 1980s…and the rat proglucagon cDNA and gene sequences followed shortly thereafter.” Daniel Drucker joined the Habener lab in 1984, and his transfection experiments in fibroblasts, pituitary cells, and islet cells established that GLP-1 and GLP-2 could be liberated from proglucagon. His 1987 PNAS paper established that GLP-1 directly augmented glucose-dependent insulin biosynthesis and secretion from β cells, the key discovery showing bioactivity. In the same lab, Svetlana Mojsov showed that laboratory-synthesized GLP-1 could trigger insulin production. (As an aside, according to Wikipedia, GLP-1 derivatives that Mojsov had synthesized while working in Habener’s lab were patented without her knowledge as peptides able to prompt the release of insulin, but with Habener as the sole inventor, and, even though some patents were amended, Mojsov still fights for a full share of credit for the discovery of GLP-1.) Elsewhere, in Copenhagen, Jens Holst and colleagues did pioneering work in characterizing GLP-1 further and showing its ability to slow gastric motility, an effect responsible for the utility of GLP-1 drugs to promote weight loss.
As was the case with the discovery of immune checkpoint inhibitors, funding sources were mixed, with the NIH funding Habener’s research (Public Health Service grants AM30834 and AM20349), with Holst being funded by Danish sources (Novo Nordisk Foundation, Danish research councils). Also, Drucker’s later work was funded by the Canadian government after he moved to the University of Toronto to take a faculty position.
That’s not all, though. Research that led to the identification of GLP-1 mimicking compounds was also funded by NIH and other government agencies, mainly the Veterans Administration:
The story of GLP-1 (glucagon-like peptide-1) therapeutics did not start in a pharma boardroom — it began at the lab bench. Back in the early 1980s, scientists supported by the U.S. government were already unraveling clues that would lead to today’s drugs. An NIH-funded researcher, Dr. Jean-Pierre Raufman, discovered unusual molecules in Gila monster venom that affected the pancreas. This odd line of inquiry was driven purely by scientific curiosity (why can a Gila monster go months without eating?) — the kind of high-risk basic research that public funding makes possible.
Fast-forward to 1992: Dr. John Eng, working at a Veterans Affairs (VA) hospital in New York, isolated a novel peptide from that same venom. He named it exendin-4. When Eng tested exendin-4 in diabetic lab mice, the results were remarkable. Blood sugar dropped dramatically, and unlike the body’s natural GLP-1 hormone, the effect lasted for hours. Eng had found a way to mimic GLP-1’s benefits without its biggest limitation (our native GLP-1 gets cleared within minutes). This was the eureka moment researchers had been hoping for — a breakthrough born in a government lab.
It’s worth underscoring just how much public science paved the way here. Academic endocrinologists had first identified GLP-1’s potent effects on insulin release in the late 1980s, and early clinical experiments showed it could normalize blood sugar — but only briefly, and with side effects. Eng’s exendin-4 discovery provided the missing piece. And tellingly, his employer — the U.S. taxpayer-funded VA — declined to patent the discovery. For three years, Eng struggled to get any pharmaceutical company interested in developing it into a drug. The notion of a diabetes treatment derived from lizard venom seemed far-fetched at the time; many industry experts were skeptical that it would ever work.
You get the idea. Yes, Novo Nordisk developed drugs designed to mimic GLP-1, but there would have been no GLP-1 known for them to try to mimic had it not been for public funding of basic research by the NIH and the Canadian government, as well as by the Danish government and, to be fair, the Novo Nordisk Foundation.
This brings us to what Dr. Atlas is either missing or intentionally not telling you. When he argues that, if the NIH were to be abolished, private foundations and industry would pick up the slack funding basic research and high-risk projects, he’s intentionally conflating drug development with the findings from basic science research that make drug development possible in the first place. If basic science doesn’t discover proteins and other biological molecules that can be mimicked or inhibited chemically, there are no targets for drug development, and that’s where NIH excels, Circling back to the article that Dr. Atlas cited to imply that NIH funding is a relatively small part of overall US funding for biomedical research, let’s look at what the authors call the “unique benefits of government-funded research,” and I thank Dr. Atlas for saving me some work by misleadingly having cited the article:
First, the federal government supports the vast majority of the basic research that establishes our understanding of biological processes and determines their molecular mechanisms—in other words, this research describes in detail the “operating systems” of cells and organisms. Knowledge of how these processes and mechanisms can go wrong and cause disease reveals “targets” for therapeutics or technologies that can treat, cure, or prevent disease. In other words, government-funded knowledge discovery enables industry-funded innovation and development of diagnostics and treatments that improve health and save lives.
As I noted above, if you don’t understand the basic biology behind human physiology and diseases, you can’t use that knowledge to develop drugs or treatments. Next:
Second, the federal government funds scientific inquiries that are not seen as profitable targets for industry, such as research on rare diseases. Because these conditions affect small patient populations, private industry lacks the financial incentive to investigate them or seek therapies. In aggregate, however, rare diseases affect 25 to 30 million people in the United States alone (NIH, 2025b). Federal agencies like the NIH provide targeted support to researchers who investigate rare genetic mutations and disease mechanisms and infer therapeutic approaches. Such research has led to the identification of previously unknown disorders, improved diagnostic accuracy, and revealed treatments that may never have emerged through market-driven research. Moreover, it is not infrequent that the investigation of a rare disease has uncovered new information or employed a novel technology useful for deeper understanding or approaches to treating more common ailments.
Or, as Geoff Cook puts it in the specific case of GLP-1 drugs:
This public-to-private handoff is a textbook example of how healthcare innovation often works. The high-risk discovery phase was underwritten by taxpayers, and the product development phase by private firms. There’s no question that companies took on tremendous expense and risk to turn peptides into safe, mass-produced medicines. They ran large-scale clinical trials, navigated regulatory hurdles, and scaled up manufacturing. The partnership between public science and private development in this case has yielded phenomenal results: GLP-1 drugs are arguably the most significant advance in metabolic health in decades, offering hope to millions struggling with obesity and diabetes.
Indeed, a recent survey found that NIH funding was involved in the development of 99% of new drugs approved by the FDA between 2010 and 2019 and that the NIH spent at least as much as the pharmaceutical industry developing drugs.
There’s another thing that these libertarian bros (and they’re almost always bros) forget. If research is only funded by private industry and private foundations, then private industry will only fund research that (it hopes) will result in products that will swell its profits, and private foundations will be driven by the personal biases of their founders and boards of directors. The political process might be messy, but when you have an institution like the NIH that is (mostly) insulated from the worst partisan politics and dedicated to choosing which science to fund based on merit (regardless of whether you agree or disagree with how it determines which projects are scientifically meritorious), I would argue that’s way better than not having the NIH at all.
Next:
Third, the federal government plays a unique role in enunciating national strategies and priorities for health and biomedical research and bringing academia, industry, and patient groups together around common goals. Through presidential initiatives like the Cancer Moonshot or the All of Us Research Program, the government defines key areas of focus, coordinates large-scale research efforts, and allocates funding to drive progress in pressing health challenges.
Of course, central coordination of anything is anathema to people like Dr. Atlas, no matter how much it can contribute to the public good. Indeed, one of his criticisms of the NIH was that Reagan interfered with its funding and thus its ability to coordinate a national strategy for AIDS research.
And:
Finally, and very importantly, as part of its cost-sharing arrangement with academic institutions, the federal government pays a substantial portion of the cost of educating and training the next generation of science, technology, engineering, mathematics, and medicine (STEMM) professionals. This is the workforce the nation relies on to sustain and expand US R&D leadership, including in the health and biomedical sphere.
Again, to people like Dr. Atlas, this sort of training is anathema, as they believe from an ideological standpoint that it is better for the private sector to mold young minds into the next generation of scientists, physicians, engineers, and STEMM professionals.
Whenever I encounter a libertarian fever dream fantasy like the one laid out in Dr. Atlas’ WaPo op-ed, I’m always half-tempted to ask if I could have some of what the writer is smoking, even though I don’t do weed—or any form of cannabis, for that matter. His delusions regarding who will fund basic and high risk research if the NIH were ever to be abolished are on par with the delusions of libertarian “intellectuals” like Nick Gillespie and Ronald Bailey, both of whom have falsely argued that the FDA kills more people than it saves through its supposedly horrific bureaucratic hurdles to drug approval and advocate for eliminating the FDA in favor of private testing labs like United Laboratories (UL). In reality, contrary to his stated claim that the NIH is politicized beyond hope, I doubt that Dr. Atlas really cares about the politicization of science given how enthusiastically he participated in it during COVID-19. Rather, he wants to eliminate the major force countering the ability of those with politics that he likes to control biomedical research in the US.
I do wonder, however. The libertarian “free market will fund scientific research” guys are bound to clash with the Trump-aligned guys who are happy to keep funding the NIH at a reduced level, as long as Trump’s political appointees control who receives research grants. Who will win? I don’t know, but I do know one thing: Whoever wins, science in the US will lose.