Can you explain to the committee why you chose to fund this dangerous research in China? On the advice of council, I respectfully declined to answer based upon my rights under the fifth amendment of the constitution.
I'm Neil Ferguson, a senior fellow at the Hoover Institution. And no, don't adjust your devices. Bill Whan's off this week, so I am sitting in his chair as moderator. However, my fellow good fellows are also here and they are the one and only HR McMaster and John Cochran who are also senior fellows here at Hoover. Hi HR and John. Our guest today used to be a visiting fellow at Hoover, but he got an even better job last year. Dr. Jay Bhattacharya is the director of the National Institutes of Health and I think the acting director of the Centers for Disease Control and Prevention. Welcome Jay or should I say welcome back.
Nice to be here Neil. Although it's an open question whether those jobs are better than being a professor at Stanford. We'll see over time.
Of course we'll get you to comment in more detail on that later on. I want to take you back in time to your last appearance which was November 9th, 2021. Back in those days, you were just a mild-mannered academic like the rest of us. And the conversation then centered around well an assortment of COVID-related matters. Remember COVID viewers, that pandemic that we've all kind of stuck in the memory hole. And one of the things we talked about was Anthony Fauci's performance as head of the National Institute of Allergy and Infectious Diseases. I think it's probably an understatement to say that at that time Jay was notorious, famous to me as a skeptic about much that Dr. Fauci advised us all to do during the pandemic and in some cases required us all to do. So Jay, can we start with an update? Is it fair to say that just about everything we've subsequently found out about COVID has vindicated you and other skeptics of lockdowns, school closures, and other measures? Or to put it really simply, was the Great Barrington Declaration right after all?
I mean, this is one of those things where it's hard — I mean yes, I think mostly it's right. I think the thing is it's kind of a cheat Neil because the Great Barrington Declaration argued for focused protection of vulnerable elder people based on the epidemiology of the virus which is that it really killed older people at much higher rates, orders of magnitude higher rates than young people. And then that there really were direct harms from the lockdowns, the school closures which are palpable. You still see echoes of that today. That was the standard wisdom in public health. The fact that it was controversial at the time is still a stunning thing to me. But yeah, I think we've seen now the wisdom of that approach which wasn't really novel to me. It was really, like I said, an age-old way we handle public health.
And I think — I mean you couch this in the context of Dr. Fauci and Tony Fauci who I guess would have worked for me if he had stayed on at the NIH. I think his is going to be a very complicated legacy. I think he had a really positive role to play in the HIV epidemic which continues to this day. But over time he made some decisions I think were very questionable about dangerous gain of function which are the research agendas that may have actually led to the pandemic eventually coming — not specifically directly but in terms of the research agenda allowed experimentation on viruses in ways that probably shouldn't have been allowed. He authorized that, oversaw it, advocated for it.
And then during the pandemic itself he played a very leading role in public health, which is actually an odd thing for a head of a research institution like NIAID to do. I sit as acting CDC director and NIH director — that's never happened in history. Those are two different missions, research and public health. He didn't have both those missions. He only had research. And I think he abused his position in many ways that's come out since the 2021 interview we had together. That's a source of a lot of consternation for me because I'm advocating for the continued relevance and importance of the NIH as an institution to advance American health. And you have that history of Tony Fauci's actions during the pandemic and before it — it makes it much harder.
You're probably too busy to read extracts from his diaries that Senator Rand Paul has made available. Any thoughts on what they're revealing?
I read — it's 1100 pages so I've not read through all of them but I have a friend sending me every time they find a nugget they send me a nugget. And I'm almost at a loss for words. There are places where he directly attacks me — like just leave that alone. But more broadly the way that he approached public health and the way that he approached science left no room for debate or discussion. He was the science. It actually wasn't scientific in one sense. It was an argument from authority on things where he really didn't have the institutional role or expertise really to make judgments.
And as a result, there were all kinds of instances where it seemed to me from reading the extracts that were sent that he was saying things publicly that he didn't believe. Like the very definition of hypocrisy, right? That the public role would say that it's a conspiracy theory that dangerous gain of function might have caused the pandemic. And the private notes would say, well, you know, nine of 12 scientists that I brought in told me this might be the case. And you see this over and over again — on school closures, on masks, on social distancing, a whole host of things.
There's this sense that he was playing a role, a public health role, conveying very confidently things that maybe he himself didn't even believe, if I'm believing the diary. And I just — it's just heartbreaking. It's heartbreaking to think about what might have been if someone with a little less sense of hubris had been in that position, had been a little more circumspect in how he communicated the really really conflicting data. I mean it was really hard to tell exactly what the — because it takes time in science to develop an idea about what we actually believe as scientists together. And so he was conveying certainty and as a result millions of people's lives were fundamentally altered often for the worst.
I mean I think especially school closures — the intervention that he did for school closures in 2020 led to now millions of school kids, not just in the United States because his words had resonance around the world — millions of kids being deprived of the basic right to an education that they had and they're still paying the cost of that. Anyways, it's a painful subject.
Jay, let me ask this. It's fun to make Fauci's kind of in the news and respectfully I declined to answer on the advice of counsel — you know, this is fun. But it wasn't just one guy. Deborah Birx was in it. The entire science community was in it. You and Scott Atlas were cancelled by your colleagues at Stanford, by all the large medical and public health institutions were all in on this and it clearly was some politics. It was kind of funny when before the election Republicans were for vaccines and Democrats were against it and then the election changed and all of a sudden whoops we changed sides on vaccines.
So it's not just the arrogance of one man. It's a whole set of institutions jumping in and doing it. It's not just the silliness of masking — screaming two-year-olds playing outside in a playground. Perfectly obvious to everybody this was a stupid idea. But it's also the censorship which you experienced, the censorship from the government, the Twitter files, the cancellations, the universities.
So in your job as NIH interim CDC director — I had to go look up exactly what you do — the NIH, the CDC, and the FDA are parallel. I wish you were in charge of it all, but you're not. But apparently, you have a little say at the CDC. What worries me most is the next pandemic. The catastrophic failure of trust not just in the science but certainly in the institutions of science which went all in on the last one and the institutions of the CDC. The next pandemic will come before we're ready. This one might be different. It might preferentially kill young people like past ones have. It might have a 10% fatality rate, not a 0.2% fatality rate. And the utter incompetence of these institutions will lead people to just ignore all advice. So how do you put that all back together again?
Yeah, so let me first take it from the most important thing from a personal point of view. Please don't wish the FDA on me. I'm not getting much sleep now and I'm looking forward to handing the CDC off to Keisha Scalff who's going to be an expert CDC director. So that's first. But I do appreciate your confidence in me, John. On the scientific community, I agree with you — wasn't just Tony Fauci. I mean, he's the — it's very easy to point to him because he pointed to himself, right? He went on TV and said things like — you know, if you question me you're not simply questioning a man, you're questioning science itself, right? So someone like that who takes on himself essentially this high ecclesiastical authority within the context of the scientific sociological structure — it's easy to point to him but it's the structure itself that's the problem, right?
And in fact this has been at the heart of how I've been thinking about my job at the NIH — what I'm really about. What I'm trying to do is reform the scientific community by making how we decide what's true really focus on replication. Because it's essentially democratization of an epistemological power, right? Rather than having high authority that gets to decide what's true in science — a top journal editor, a fancy professor, NIH director — instead it's what does the scientific data actually say? It may sound surprising, John. I mean you can see this even within economics sometimes, right? So what does Paul Krugman say determines what's true in economics? Well, we say no to that because we say we actually look at the data. We can have a scientific fight. The Chicago tradition is all about that scientific fight, right?
In science it's the same problem except that it's often not so public, right? And so but it became public during the pandemic — you had Tony Fauci getting to decide items of vital interest to the well-being of everybody on scientific matters and then all of a sudden all the scientific community falls in line underneath him. That points to a sort of sick scientific culture, a scientific culture that doesn't value replication, that doesn't value the junior scientist who finds a very inconvenient piece of data that overturns the whole thing. I've been working very hard at the NIH to reward that kind of structure because I think that's the long-term solution to the problem you identified.
So Jay I want to talk a bit about your analysis of these wider problems. When you were being confirmed, you set out five crises in the domain of the NIH and it's kind of a good opportunity to check in now and see how it's going. Let me start with the first of these crises you identified which was a chronic disease crisis. And you said, quote, 'American health is going backwards. Life expectancy flatlined between 2012 and 2019, plummeted during the pandemic, still hasn't bounced back to pre-pandemic levels.' This is all during your confirmation hearing back March last year, and you said you were going to make NIH focus much more on addressing this chronic disease crisis. How's it going?
So substantively we have actually started to see increases in life expectancy and actually the most likely reason why has to do with something the NIH did a long time ago which is these GLP-1 drugs that have reversed the increase in body weight that goes back at least 80 years in the United States. For the last two years body weights have been declining and so that has some impact on chronic disease. As far as actually getting the research enterprise to focus on it — we've increased the amount of leeway that the institutes have on addressing chronic disease problems that include things like — you can think broadly about chronic disease — include things like autism.
So you know this autism data science initiative to allow people to focus on and ask research questions regarding the etiology of autism. To more fundamental things like heart disease, diabetes, Alzheimer's disease. There's a sort of warrant to all of the institute directors that are in charge of the research portfolios there to think broadly and think widely. I'm hoping that the answer to your question will come out over the course of the next few years as you start to see these new discoveries turn things around. I believe it's possible, but you just have to articulate this that we're doing it and we're going to do that. Ultimately I think my tenure as director will be judged 10 years from now based on whether the US is still in the same kind of chronic disease problems that it currently has or scientific advances have really helped turn those around the way it has for obesity.
So I mentioned there were five crises and the one that I find most interesting is what you call the reliability crisis. You said in your confirmation hearing NIH supported science should be replicable, reproducible and generalizable. This is a tremendously important issue that we've not really touched on much on Good Fellows — the crisis of replication in the sciences. Explain a bit for our viewers what that refers to and what you're doing about it.
Yeah. So I mean it's actually Neil it's absolutely shocking when I talk to people who don't know about this but it is a fact that a large chunk of the biomedical sciences and unfortunately not just the biomedical sciences published literature — when independent people look at the same question, they even just narrowly reanalyzing the data that's there or asking for new data or new approaches to the same question, they often find a different answer.
Like estimates like in cancer biology — basic cancer biology — 70% of papers are not replicable or the methods aren't described enough in a way so that replication could happen. And it has real consequences — not just cancer biology — in neurobiology, in psychology, in field after field. Just recently economics came under the same kind of focus. And the reason is common to all of these fields. The reason is that the way that we judge productivity in every single scientific field it seems is very narrow and not tied to actually finding truth. We measure productivity based on how many papers you've published and how influential you are as measured by citations.
That is not the same thing as I've discovered — now we have scientific heroes, we have Nobel Prize winners and all this other stuff. But the work-a-day work of science is rewarded based on those two measures. It's not rewarded based on have I found an idea that transforms a field that other people when they look at it find the same thing as me, right? That sort of replicated important idea being the key measure of productivity in science is not the actual reality in our scientific fields. Instead what we have is this sort of authoritative structure where a top journal editor gets to decide what's true in their respective fields. Tony Fauci gets to decide what's true in his respective field, right? And so many people focus their efforts in science in currying favor with the top authorities.
It's akin to what led to the situation before the first scientific revolution where high ecclesiastical authority got to decide things like do the moons of Jupiter move. And the first scientific revolution was a few very clever people with telescopes who got to decide whether the moons of Jupiter move. So I've been working on that. The NIH actually is a great place to do that. One is we are now funding replication work.
Two — which sends a signal to the scientific community this is important because right now it's not seen as important. Two, we are developing measures and metrics for scientific productivity that include replication and good practices around replication — like do you share your data, do you share your code. We just had this big S-index challenge where this kind of data sharing — it's a really important thing to be able to point to for science — is measured and therefore will be rewarded.
Three, we're going to have a journal, a place where you can publish your scientific replication work. Right now you can't — we'll have an NIH journal. We're working on that. We'll see that hopefully later this year. We'll establish a new journal where people can publish their replication work and share that information with the community.
And four — this is probably the most important of them. We're going to tie the search functions for scientific papers — right now you search for them on PubMed which is where you search for scientific papers at the NIH or everywhere and you see the journal and you see the authors and you may see the abstract. We're going to add a replication button. You click the replication button and then you'll get a knowledge graph of all the related papers and then you'll do an AI summary of each idea in the paper — what the related literature says — and you can have a comment button where we can disagree with the AI and have a conversation. And it'll put the emphasis on what do other scientists believe about the ideas that you publish rather than was it published in the top journal like the New England Journal or something. So in a way this illustrates how you can use the sort of power that you have which has the potential to be ecclesiastical — to empower science as it really should be practiced.
Okay. Here's the good news. There's one of your crises that I know you've solved and that's the crisis of scientific dissent. And you must have solved it because there are lots of people attacking you. Including the authors of something called the Bethesda Declaration — I think 37 pages denouncing your management of the NIH. Is this a sign of success or something else?
It's absolutely a sign of success. Right? So one of the first things I did is I issued an academic freedom policy at the NIH. So if some scientist comes up with a result — it actually used to happen — you'd have to often get for politically consequential or controversial results, scientists would have to get clearance from my office sometimes in order to be allowed to send it out for publication at all. Under my new academic freedom policy, if a scientist has a result, the only clearance they have to get is based on handicap accessibility and all that other stuff, nothing substantive. They're allowed to send it out even if I disagree with it, right? So I don't want to police scientific results.
And then the Bethesda Declaration people — what they want essentially in effect is a return to the status quo ante at the NIH. Not just on scientific freedom issues but also on like support for DEI initiatives and a whole other range of things I don't think are a good idea. So I just — I've turned the page on that. Is it completely — I mean we — these are cultural changes, they take time, but from a policy point of view I think I've made my position clear.
So you're probably wondering what happened to HR because I must be guilty of suppressing his free speech because he hasn't said a thing so far. And I think it's because we're talking about stuff that's quite far from his comfort zone. But we're about to get to the geopolitics of all of this because you can't talk about gain of function research, which was one of your crises, without talking about China. And I'm curious to know what you currently think about the Chinese part not just in COVID but more generally in the field of biomedical research because it feels to me as if one of the big changes since 2020 is that the Chinese have vaulted forward in terms of the scale and sophistication of their research. So this is a chance for HR to get into the conversation because if China's caught up or is catching up in this area, this is a big deal.
No, to add to it — the lab leak, the WHO, and China's ability to do stuff much much cheaper than we do it because there's a whole lot less paperwork.
Yeah. And just the context Jay of artificial intelligence combined with bioengineering — how you're thinking about that threat, what are the kinds of capabilities that we need in terms of rapid diagnosis, vaccine prototyping — I mean just that whole area of threat that is really centered on the PRC and the work that we know is now open source. I mean for many years there we had classified knowledge of them designing bioweapons. So yeah, I'd love to hear your thoughts on what you're doing in that area.
I'd also love to get your reaction HR for what I'm doing because I think this is one of those areas where the future of the country and in fact the future of the world really depends on getting it right. So just to set the stage because the problem I was addressing or thinking about in that confirmation hearing a year and a half ago is that as I said earlier Tony Fauci and the NIH had embraced a philosophy of risk tolerance regarding some very dangerous work.
The argument that they made was that you could take viruses, manipulate them, make them more transmissible, potentially more dangerous to humans, and then triage them to decide which ones were most likely to cause pandemics. The possibility that that research program might have led to the pandemic is a I think a very real one. I mean I know the scientists disagree. But in any case we as the NIH need to make sure that doesn't happen again, right? That research program.
If there's — when I wrote an op-ed just recently in the Wall Street Journal where I made an analogy to what Enrico Fermi did — the great Italian physicist — during World War II he did this experiment to launch the nuclear age, the first nuclear chain reaction. And he was very worried that what if that nuclear chain reaction doesn't stop — it blows up the world. He did a calculation — brilliant mathematician — realized that the risk was zero and then he did the experiment.
And so the president signed an executive order last year and now we have a structure in place to enforce it — to ask and basically require scientists to do that calculation before they do any dangerous gain of function research. And that executive order includes an order for the United States essentially to stop collaborating with our adversary nations on scientific topics.
What we did is — and now that's really important from a geopolitical point of view because if you go back 25 years the US invested a tremendous amount in the biomedical research enterprise of China. In fact almost every single top biological scientist in China — their entire infrastructure, post-docs — involved US government investments. And now they are at a level where they have a fantastic biomedical research enterprise.
And it's legitimate and you know they have — they do faster randomized trials than we do. They violate — they have very different ethical standards regarding human subjects protection, which I think poses problems. But it's a legitimate adversary to us on this setting. And our investments with building up their biomedical research enterprise I think are better directed to the United States and to American scientists.
You bring up AI and I think it's a really really important question related both to the Chinese state actors involved and potentially other non-state actors. So the AI tools that we have now have the potential for people to develop formulas that would allow the development of biological agents that could pose a lot of harm.
And at the same time those same AI tools have the potential of unlocking scientific ideas that can solve the chronic disease crisis, make people healthier at scale, inexpensively. And so it's a really interesting balancing act HR to figure out how to do this right.
And the key thing to me is we have to allow the AI tools that exist to have their effect on the good. To protect people from the threat that these kind of AI tools in the hands of actors — I think the key thing there is state actors might be different than non-state actors — the key thing there is rather than regulating the AI, which seems like it's not possible right now to do, it's to regulate the access to the precursors that would allow you to develop those tools in the first place.
It's similar to the way we deal with nuclear threats. It's very difficult to control the knowledge about how to build a nuclear weapon. There's I think a Princeton honors thesis student in 1980 or something actually wrote his honors thesis on how to build a nuclear weapon, right? So that's available to bright people. What's not available is how do you get plutonium, how do you get centrifuges — those things are not easily available and we use the access to those precursors as a way to control that risk. I think it's going to have to be some structure like that on AI.
I think the DGRF — the dangerous gain of function risky research policy — I think is a big step forward. We're not going to have lab-created pandemics based on NIH funding or US government funding again as a result of these structures.
Jay I know we don't have limitless time with you and given that you're doing two jobs still I just have to ask you — and you'll know that I'm going to ask this — about vaccines. Because you said recently you see no proof that any kind of vaccine causes autism and that seemed like breaking with your boss the Health and Human Services Secretary Robert F. Kennedy who's repeatedly cast aspersions on vaccines. So is your relationship under strain with him or are you finding a modus vivendi?
No, actually quite the opposite. I love Bobby actually. I think he's an amazing leader and I've really admired and been very happy to work with him. Let me just characterize his position as I understand it as carefully as I can. His position is that in many cases in many vaccines we just haven't looked right. The Institute of Medicine in 2016 issued a report that suggested that the evidence based on some vaccines regarding this connection to autism for instance has been looked at and not found anything. But that for other vaccines there hasn't really been a look. And he's not wrong on some of those vaccines.
It's been very difficult to get scientists to ask the question because it's seen as a death to your career even to ask the question. And so I think it is accurate to say that for instance for the MMR vaccine — when I look at the data, when I look at the literature — this is far from my direct expertise so if those who are listening disagree with me, send me more data and I'll try to learn more — but I don't see a link and the best data sets I've seen, the best information I've seen around this suggest that the study would have found a link but it didn't. So it seems unlikely. That's for MMR.
For a whole bunch of other vaccines, I think Bobby's right. There just hasn't been people that looked. And just more broadly, the idea that we know everything there is to know about every single drug is just not true. I think we should — and when you have an issue like this where it matters to the public a lot — our job as scientists is not to judge the public for asking the question, but to answer it as best we can with the best data we can in the most convincing way we can without prejudging the answer.
And so Bobby and I get along just fine. He is actually a very bright man. He reads scientific papers. We argue about them sometimes. He likes those kind of conversations. He's not dogmatic. I think the picture that's painted of him in the press is just deeply unfair. He hired me because he wants my take on it. I don't come in saying I know all the answers just because I was a Stanford professor. I just don't want to recapitulate the sort of structures that I've been trying to reform.
But I want to bring this back to your main quest as I understand it which I heartily applaud. You said it briefly — people who ask those questions know their careers are over. There are so many questions in science that don't get asked because people know their careers are over. And I'll put in my plug. My wife has been bugging you to get people to ask the question of whether putting babies on their backs is a good idea or not. That's classic. And there's retrospective research to be done, there's meta-analysis to be done, none of which gets done. It comes to the basic thing which I think you highlighted but I want to emphasize — we never know for sure and there's this image that a study is done, it is published, that's the answer once and for all. The FDA falls prey to this — we're going to do a study, this drug is safe and effective once and for all and then you never look again. Whereas science is all about looking again over and over and over, plumbing into the details, finding all sorts of things slowly but surely that are not right.
That's not really rewarded in science. So plus the politics, plus the young people aren't doing it anymore. So as we're sort of closing — your plan for how to fix this big bureaucratized monolith of politicized science funding and the culture that expects one study answers it forever. I mean I think your answer to RFK was not there's no
Evidence. There's no evidence. Now, let's go look and find out, which will take decades. We'll never have an absolute answer on a lot of these questions.
Well, we may we may have an answer. It's just you know, sometimes we reach a conclusion. But I think the key thing, John, the key thing right now, the problem right now in the sociology of the sciences is the unit of productivity is a published paper. The unit but really the unit
Citation. Well, but that and then then all that flows from that, right? So once you have the published paper, then you can get the citations, then you get the promotions, you get the authority, then you have the you get to be editors and you get to decide truth in your field. The unit of productivity in science ought to be an important replicated idea and people that
Or the replication of the idea.
Yes. What but what? Right. Right. Well, I mean that important replicated idea will have many many many people to credit for it. The person who had the idea, the person who tested the idea, the people who replicated the tests, the people who developed the discussion, the structure to discuss it, all of that is like part of a unit of productivity that matters. So we have the wrong unit of productivity in science and as a result, you have essentially a choke point where you can stop scientific questions from being answered. They're inconvenient, right? You had you had I think it was like the Lancet or something that that that for the first time in its history or something. They they they they tapped somebody to be they they they endorsed the endorse word Joe Biden to be president. You know, you have essentially this merging of this polit like raw politics with science. And then you have a structure that allows people at the very top of that hierarchy to set standards that determine who advances in fields, what questions can be asked. We just need a democratization of that power. And a lot of my thinking around how I led the NIH is around that, like how can we reform the scientific structures to make the asking of questions, to reduce the policing of asking of questions, to allow scientists to advance when they have important ideas that are replicated that make a difference both in their fields and more broadly. It's just it's frustrating John because I know you and I have lived this in economics forever. To see these structures play out in other fields as well in science. It's it's although science is an incredibly powerful tool, unless we are able to do that, we won't really unleash it for the good of all.
Well, Jay, we need to let you get back to your two jobs. I think you've illustrated in this conversation why our old friend Scott Atlas is probably wrong to argue that the NIH should be abolished. So long as you're in charge of it, I think it must be a force for good. So thanks so much for joining us. Keep up the amazing work and come back when you've achieved all you can achieve to join us at Hoover.
Neil, thank you so much. HR, John, it's so good to see all three of you. And grateful for your questions, grateful for your thoughtfulness.
