Yes. Very, very roughly the NIH is 8% of all medical research funding in the US.
According to DiMasi et al¹ in 2013 a fully approved drug cost $2.870 billion USD, if you divide up the cost of all the failures to be paid for by the successes. They also say that this cost has been growing at 8.5% per year. If we take them at their word then that growth adds up to $7.639 billion USD per drug in 2013 dollars. There’s been 38% inflation since 2013, so that is actually $10.542 billion USD in 2025 dollars.
According to the FDA website² there have been an average of 48 approved drugs every year between 2021 and 2024. Drug development takes ages, probably over a decade, but I’m going to ignore that. 48 * $10.542 billion USD is $506 billion USD.
The NIH budget³ is $48 billion USD per year, of which they spend 92% on research. $44 billion USD is 8.0% of $550 billion USD. But actually that is apparently the 2023 budget amount, so maybe we should instead say that in 2023 it was 9.3% of the 2023 total. Either way it’s a small fraction of the research done in this country.
I know that there are other funding sources than just the NIH, but even so I really don’t think that the adjustment would need to be all that difficult.
¹ Joseph A. DiMasi, Henry G. Grabowski, Ronald W. Hansen,
Innovation in the pharmaceutical industry: New estimates of R&D costs,
Journal of Health Economics,
Volume 47,
2016,
Pages 20-33,
ISSN 0167-6296,
https://doi.org/10.1016/j.jhealeco.2016.01.012.
(https://www.sciencedirect.com/science/article/pii/S016762961...)
Abstract: The research and development costs of 106 randomly selected new drugs were obtained from a survey of 10 pharmaceutical firms. These data were used to estimate the average pre-tax cost of new drug and biologics development. The costs of compounds abandoned during testing were linked to the costs of compounds that obtained marketing approval. The estimated average out-of-pocket cost per approved new compound is $1395 million (2013 dollars). Capitalizing out-of-pocket costs to the point of marketing approval at a real discount rate of 10.5% yields a total pre-approval cost estimate of $2558 million (2013 dollars). When compared to the results of the previous study in this series, total capitalized costs were shown to have increased at an annual rate of 8.5% above general price inflation. Adding an estimate of post-approval R&D costs increases the cost estimate to $2870 million (2013 dollars).
Keywords: Innovation; R&D cost; Pharmaceutical industry; Discount rate; Technical success rates
I mean this in the best faith, but I feel this is still sidestepping the core issue: foundational science and applied development are just not fungible. The kind of research NIH funds: long-horizon, high-risk, mechanism-level discovery, is not something corporations will just pick up the tab on if public funding disappears and is very different from the also decades long drug pipeline. Not because they’re incapable or unwilling, but because it’s economically irrational within the incentive structure today. Of course big pharma companies have money, they will figure things out and maybe even spend more funding on basic science. But very few shareholder-driven companies are going to fund a 10-year exploration into, say, the epigenetic triggers of cancer with no guarantee of a product at the end. They might take safer steps and invest in foundational research tangential to existing targets, and yes, this does launch a whole other debate on efficiency of this vs NIH moonshot research.
Yes, they do some early-stage R&D work, but it’s almost almost always downstream of publicly funded breakthroughs. They build on what’s already been de-risked. NIH isn’t just 8% of the money, it’s the first 8%, the part no one else wants to touch because it’s too uncertain, too slow, and too unprofitable in the intermediate term. Trials are expensive, that's where 80% of R&D today goes.
> The kind of research NIH funds: long-horizon, high-risk, mechanism-level discovery, is not something corporations will just pick up the tab on…
Except that there are plenty of counterexamples to that idea. Bell Labs employed thousands of researchers and they made many foundational discoveries alongside the thousands of inventions that were merely development of the existing phone system. The list of foundational discoveries made there is so long that is hard to imagine. Just off the top of my head it includes the transistor, the integrated circuit, the operating system, lasers, fiber optics, CCDs, solar cells, etc, etc. Nyquist and Shannon worked there and discovered the mathematical foundations of information theory, including entropy, bits, communication channel capacity, etc, etc.
They were ridiculously productive and fantastically profitable, but these days with taxpayers funding research most corporations don’t bother any more.
The articles you link to have much to say about how research is currently done, but do not even purport to say that this is the only way it could ever be done.
Agree, and I have no idea what's the optimal way for research to be done. My claim is that the way research is done today works for fairly small impact on taxpayer burden, and without natural experiments with other countries with spending of a similar scale to the US, it will be difficult to see which mode is best.
Bell Labs was amazing, but it also was funded in large by profits from an emboldened AT&T with monopoly power. Is monopolistic profit funding more efficient than taxpayer funding? No idea.
Regardless of inflation, AT&T made way less money in the 20s than any pharmaceutical company does today. If they could afford the research budget back then, then the pharmaceutical companies of today can too. It had nothing to do with them being a monopoly.
