An advocacy group claims NIH returns $2.50 on a $1 investment. But even a wasteful program pays salaries and buys supplies. My latest article breaks down the math and shows how that number changes when you question the assumptions. rebuildingscience.substack.c…

Adam is right that the dramatic skew of NIH funding toward older people is a serious problem. Younger scientists tend to generate fresher ideas and possess more current technical skills, while funding flows to established researchers through the old-boy networks that pervade academia. Too many scientists are still living off a high-profile paper from 25 years ago—whether or not that work ultimately held up, and in many cases it did not—using the prestige and relationships it conferred to absorb grant money long after their most productive years are behind them. But this misses two important points. First, many scientists continue to innovate at an advanced age. They are simply obscured by the larger and more visible group of older scientists who have been coasting on reputation and connections. Second, directing an NIH-funded laboratory requires more than scientific creativity. It resembles running a business: raising money, assembling a team, ensuring compliance, managing conflict, motivating people, promoting the work, protecting the lab from outside interests, and making sure the work actually gets done. These organizational and executive skills tend to develop with experience and are often more important than generating the original ideas. Still, the direction of Adam’s argument is correct: too much money in the NIH system is controlled by older, entrenched scientists at the expense of younger talent. One key is distinguishing senior professors who remain innovative from those whose funding is sustained primarily by past accomplishments and institutional power. Another problem is NIH’s current affirmative-action like preference for younger applicants, which disregards merit and only makes matters worse. The goal should be to fund the best scientists and the best work—not to substitute one form of age preference for another. I have many ideas about how to move in that direction, such as possibly abolishing the F- and K-series training grants.

This is a great point. Cancelled now describes everything from a disinvited lecture or temporarily reduced social-media reach to the complete destruction of someone’s career—rendering years of education and experience practically meaningless. Paradoxically, those who suffered the least serious consequences usually speak the loudest and receive the most attention, wearing cancellation as a badge of honor. Meanwhile, people who were truly devastated may be too traumatized to discuss it publicly, or they quietly develop ways to cope and move forward. The relatively minor but highly visible cases obscure the people whose livelihoods, reputations, and lives were genuinely shattered.

Somewhere a scientist is explaining a $47 purchase to someone whose job costs $180,000 a year.

Nickel-and-diming people over conflicts of interest is an age-old way to attack someone inside a highly bureaucratic system. Disguise personal hostility as an ethics concern, comb through every relationship, and look for a technicality to try to take them out. The insinuation that RFK Jr. went to DC for the money is absurd. He is in his seventies and has spent decades challenging powerful interests with enormous financial resources. If maximizing his personal wealth were the goals, cozying up to the people who control that money would have been a far easier route than fighting them. When you challenge the status quo, you give people invested in it a reason to attack you. Conflicts of interest are an easy place to go because almost anyone who has achieved any modicum of success has something that can be framed as compromising. In these hit jobs, whether an actual conflict affected a decision becomes irrelevant. The objective is to attack someone’s character and insinuate corruption without having to establish it. The insinuation becomes the punishment. Once the accusation circulates, the target has to spend their time defending their character while the people making it congratulate themselves for protecting institutional integrity. It is office politics at its ugliest, pretending to be public service.

NIH’s $2.50 Return Doesn’t Add Up An advocacy group's talking point turns spending into a claim of economic gain without establishing how much value was added or how much would have happened anyway. Every $1 spent on NIH supposedly generates $2.50 for the economy. That claim confuses money changing hands with wealth being created. The number comes from the advocacy group United for Medical Research. Its FY2024 analysis reports $94.58 billion in economic activity associated with $36.94 billion in NIH research awards. Divide one by the other and you get $2.56 per dollar, advertised as a return on investment. A wasteful program also pays salaries, makes purchases, and sends money through businesses. Those transactions do not establish that the spending was productive or that society came out ahead. An economic-model calculation can count activity even when the underlying work accomplishes little. That is the central weakness of the $2.50 talking point. It offers a number about spending as evidence that the spending is worthwhile. To examine it, we can construct a simple model that separates the reported activity from the value added and asks how much would have happened anyway. Start with the denominator. The headline ratio uses $36.94 billion in research awards, rather than the entire NIH budget. Using the earlier model’s approximate $47 billion budget denominator gives: $95 billion ÷ $47 billion = $2.02 in reported activity per budget dollar. This comparison has a specific limit. It puts the covered research-award activity over the broader agency budget. A complete agency-wide calculation would also need the activity associated with NIH’s other expenditures. The $47 billion figure remains a working budget assumption pending reconciliation with the precise FY2024 budget definition. The distinction matters because a claim about selected research awards should not quietly become a claim about every dollar NIH spends. Next, separate gross output from value added. Gross output counts transactions across stages of production. GDP measures value added, subtracting intermediate purchases. Adding up transactions along a supply chain does not tell us how much value that chain created. Our model treats the reported activity as gross output and applies an illustrative value-added share of 58%. That factor approximates an economywide GDP-to-gross-output ratio. It is an assumption, not an NIH-specific estimate, and its application requires confirmation that the underlying activity figure measures gross output. With that assumption: $95 billion × 0.58 = $55.1 billion in estimated value added. Relative to the assumed $47 billion budget: $55.1 billion ÷ $47 billion = $1.17 per dollar. Under this construction, the ratio is already down to approximately $1.20 before accounting for alternative uses of the resources. Now address what would have happened anyway. The money does not originate inside NIH. Neither do the people, equipment, buildings, or supplies. Those resources have potential uses elsewhere. Counting activity around NIH-funded institutions does not establish the increase in activity across the whole economy. A scientist might work elsewhere. A supplier might serve another customer. Public funds might finance another program or remain available for private spending. The relevant comparison is between the economy with NIH spending and the economy under an explicit alternative. Treating everything associated with NIH funding as an addition to national output skips that comparison. We capture this issue with an “additional share”—the fraction of estimated value added that would not have occurred otherwise. If the additional share is 100%, all of the estimated activity is assumed to depend on NIH spending. If it is 75%, one-quarter would have happened elsewhere. If it is 50%, half would have happened elsewhere. These are scenarios. The model does not establish which share is correct. It shows how much the result depends on an assumption that the headline leaves unresolved. The calculation is straightforward: R = (O ÷ B) × v × a where: O is reported output, rounded to $95 billion. B is the assumed NIH budget, $47 billion. v is the assumed value-added share, 0.58. a is the assumed additional share. R is additional value added from the covered activity per budget dollar. Substituting the inputs: R = (95 ÷ 47) × 0.58 × a = 1.172 × a. The scenarios give: 100% additional activity → $1.17 per dollar. 75% additional activity → $0.88 per dollar. 50% additional activity → $0.59 per dollar. 25% additional activity → $0.29 per dollar. Under these assumptions, the ratio falls below $1 when more than approximately 14.7% of the activity would have happened elsewhere. That is a consequence of the model’s inputs, not evidence that displacement actually equals that amount. These figures should not become a replacement slogan. They depend on the budget denominator, the conversion factor, and the assumed additional share. They also describe value added rather than government revenue, so $1 is not a fiscal break-even point. But the exercise exposes the weakness of presenting $2.50 as an established net return. The headline does not settle how much value was added, how much activity was displaced, or whether the resources could have produced more elsewhere. Scientific discoveries and health improvements require their own evidence. Their value cannot be established by counting payroll and purchases. This model does not estimate those longer-term benefits or provide a complete comparison with alternative spending. NIH should be judged by what its funding achieves. A program can spend billions, generate transactions throughout the economy, and still waste resources. Following the money as it changes hands does not prove that spending it was a good investment. Disclosure: This analysis and draft were developed with AI assistance. I reviewed and edited the final version.

The doomsayers claim AI will make professions like radiology obsolete. But there has never been a better time to study radiology. We still have so much to learn about how radiation interacts with the human body and how those interactions vary across disease states. I studied some of these questions in relation to metabolism and found them fascinating. Fully understanding these interactions is a scientific challenge on the scale of understanding aging and human disease themselves. Automating image interpretation does not solve those questions. The opportunity is to automate mundane tasks and free people to pursue the difficult questions that make the field worth studying in the first place. That work could eventually allow us to answer radiation-related questions about health, disease, and treatment that we cannot say anything about today.

When politicians and advocacy groups cannot defend the status quo on its merits, they place the most emotionally useful victim they can find in front of the budget. A sick child, dying cancer patient, or struggling student becomes a human shield for every failed program, bloated bureaucracy, and protected interest behind them. NIH defenders routinely do this to validate a nearly $50 billion agency and demand that its budget be doubled, tripled, or, why not, increased to trillions. An anecdotal case, however, establishes neither durable success nor that NIH was uniquely necessary. It answers nothing about a grant system that rewards incremental conformity-driven projects, forces scientists to spend enormous amounts of time on grants, funnels overhead into administrative empires, concentrates funding within the same legacy institutions and old boys’ clubs, and pours money into large bureaucratic center grants that are largely rent-seeking. One patient cannot substitute for an audit of an agency. One anecdotal and possibly even unrelated story cannot account for the other $50 billion.

This article is spot on. You cannot repair a decaying university by adding a classics course. Many conservative activists act as though adding civics and great books courses while removing a few critical-theory courses will change the institution. Universities are sprawling bureaucracies built to absorb money, expand administrative power, protect internal networks, and avoid accountability. They can capture almost any new center, rename contested programs, and wait for stakeholders to lose interest. Small adjustments to the course schedule leave the underlying institution intact. The politics and other hot-button controversies are downstream consequences of that structure.

Turning a sick child into a human shield for a spending demand is political hackery at its most cynical. The published evidence involved one patient, short follow-up, and very early results. It also represents less than 0.1% of what NIH does. It does not prove that NIH was uniquely necessary or that the work could not have been funded and performed elsewhere. NIH stifles innovation by forcing scientists to spend enormous amounts of time writing and rewriting grants, rewards safe and predictable projects, and funnels money into bureaucratic programs and the same legacy institutions and boy's clubs. Too much funding supports paperwork, overhead, large bureacratic center grants, and other administrative structures instead of giving the best scientists the freedom to pursue their best ideas. A rough estimate is that less than 25% of what NIH funds is worthwhile. This is how NIH avoids accountability. Present one emotionally powerful case as a definitive NIH triumph, then use it to demand another $50 billion while brushing aside the vast majority of what the agency actually funds.

Universities lose their monopoly the moment talent finds another path to opportunity.

It is encouraging to see alternative models of higher education begin to emerge. Higher education is a trillion-dollar industry waiting to be disrupted. This is one small beginning. Hopefully hundreds more alternatives will spring up, compete, and show that universities are not the only way to prepare talented young people for the future.

There is a persistent misconception that every major medical breakthrough comes from NIH. Followership is often reframed as leadership. The graphs below show academia arriving well after industry proved the therapies worked. Consider GLP-1 drugs. I regularly attended obesity conferences throughout the 2010s. GLP-1 was nowhere near the center of the academic conversation. It was a small blip in the NIH obesity portfolio. By 2020, less than 5% of the NIH obesity grants had GLP-1 in them. Instead, the obesity/CVD fields were dominated by blowhards peddling stories about branched-chain amino acids, ER stress, and whatever drug target fit their latest grant. Most or all of these theories produced nothing remotely helpful for people. Meanwhile, industry was doing the work academia now likes to describe as bench-to-bedside translation. Novo Nordisk scientists developed liraglutide and semaglutide. The company designed, funded, and oversaw the pivotal clinical programs that established what these drugs actually did in people. The clinical timeline is illuminating. The first GLP-1 drug reached patients in 2005. Liraglutide received an obesity indication in 2014. Semaglutide produced 14.9% mean weight loss in STEP 1 in 2021. SELECT later showed a 20% relative reduction in major cardiovascular events in overweight people. NIH’s role in producing these breakthroughs was small or inconsequential. Once industry made the clinical success impossible to ignore, academia flooded into the field. NIH-funded project records mentioning GLP-1 increased to 326 in provisional 2026 data.

I enjoyed this talk from Bill Gurley. He identifies the central failure of the pandemic. Institutions and invidividuals with ego and legacy in mind reached conclusions without evidence, then attacked people who continued asking questions. COVID became so politicized and institutionalized that universities, NIH, scientific journals, and professional societies now have positions and reputations to defend. That makes them poor judges of their conduct. The Proximal Origin paper is a good example. It was rapidly published and publicly promoted by Fauci, even though the evidence was unclear and the authors had privately expressed far more uncertainty than appeared in the paper. We need independent commissions composed of smart people with no stake in the answer. People with Feynman’s independence during the Challenger investigation or David Baltimore’s willingness to raise the furin cleavage site when doing so was deeply unfashionable. One commission should examine the origin of COVID. Others should examine school closures, lockdowns, public-health messaging, and their long-term consequences. HHS/NIH could establish them, but the institutions being examined should not control their membership or access to evidence. Give serious people six months, access to the records, and freedom to follow the evidence. The purpose should be understanding and prevention rather than protecting institutions or settling political scores.

Anya appears so certain Carnegie Mellon will protect her that she advertises the arrangement. She even apparently feels so entitled to complain that the university damaged her reputation, as though its failure was not protecting her cleanly enough. The university issued a lawyerly statement that invited people to assume her leave was disciplinary, although it was apparently a preplanned sabbatical. She of course says the university never contacted her. This is how a university protects favored people during a scandal. It expresses enough concern to reduce public pressure while avoiding any commitments. Then it waits for the attention to move elsewhere. The calculation is that DeAngelis and the other critics will eventually move on. Anya can return to campus, continue promoting the same views, and nothing materially changes. Meanwhile, she trolls for more attention and taunts her detractors by flaunting how little consequence she expects.

Good point made by Steak ’n Shake. Universities have a similar governance problem. Their boards seem to exercise little influence and face little accountability, while government funding holds up institutions that ignore the people paying for them. Shareholders need a meaningful voice when corporate structures decay. Taxpayers need the same in universities. Public confidence in higher education has fallen sharply. That loss of trust should matter when public funding and institutional accountability are considered.

Children ask why until they understand. Adults form a committee after the second question.

Nothing says open inquiry like getting angry when someone asks the wrong question.

These university press releases are designed to project institutional value. They are not proof of what the institution has delivered or why it deserves continued public funding. Even if this work proves valuable, and I remain skeptical, nothing here explains why the funding needed to be concentrated at Harvard. That concentration is the product of a legacy NIH system that repeatedly sends public money to the same institutions. NIH should change its priorities and redirect future funding as science and public interests change. Harvard has no permanent entitlement to federal research money simply because it benefited from the previous system.