Joe House (8:45)

Thanks. It's great to be with you.

Holden Thorpe (8:47)

You have spent decades seeing the US Science system from several different angles. As chancellor of unc, as a professor, as editor in chief of science and its suite of journals. From your perspective, what makes the American science system so uniquely successful?

Joe House (9:08)

Well, I think a big part of it is how it was set up. It was set up by someone named Vannevar Bush who was involved with the Manhattan Project and around when World War II ended. And there was a lot of urgency around how is the United States going to continue to compete in the new world that was formed by the war, and especially in science, since a lot of things that happened between the US And Germany and Japan over science during the war and Vannevar Bush had a, what might be kind of a radical idea, but it turned out to be correct. And it's part and parcel of how we think about everything now, which is that the right people to do science in a completely unfettered way were professors at universities because they wouldn't be subject to the federal government setting the agenda in a way that might prohibit them from finding something that wouldn't have been found otherwise, and that they should have a lot of independence when they do that for the same reasons. And so Bush convinced the United States to have a federally funded research enterprise and to do it in universities with relatively independent university professors carrying out the research. And this was a complicated idea. It would have been easier to just set up a bunch of central facilities that the government funded because most of the universities at that time didn't have big labs that people are used to seeing when they drive around the university campus now. And all of that stuff had to be built and programs had to be put together to train students in the system. But what we ended up with, which is something that I have benefited from and administered my whole career from one perspective or another, is an apprentice system where professors, having done what their students would then do, learned how to be very independent scientists, how to follow curiosity where it led, and to bring up students and other trainees that we sometimes call postdoctoral fellows. Those are people who have finished graduate school but have a faculty position or a permanent job yet to train them to do research. And they train the next generation and on and on and on and throughout the system, it's all about finding things that other people haven't thought about before. And it's based partly on the idea that it's impossible to know what science you have now is going to turn into the applications of the future. But especially now with all the cuts and everything, you're seeing a lot of people talking about how almost every approved drug has its birth somewhere in an NIH funded study at a university. And this is, this is what's made the system great. And this is why people have traditionally come from all over the world to participate in it, which is another thing that we're now struggling with. But the fact that those folks are trained in the American system, even if they go back to their original country, it still gives the United States a lot of influence over the way the scientific enterprise plays out around the world. And a lot of us are very worried that that's going to go away.

Holden Thorpe (12:58)

There are some studies suggesting that every dollar spent on research, every dollar that flows from the NIH or the NSF to a university researcher, returns at least $5 to the economy. And if you take this statistic seriously, it really is like a kind of magic. There's almost nothing else I can think of that every time you put $1 into a machine, the machine spits out $5. You've had a really interesting career that allows you to see the origin of basic research, how that can be translated into technology, how technology can become the kernel of a new company. Can you explain a little bit about how this statistic can possibly be true? $1 of research returns, $5 of economic growth, right?

Joe House (13:53)

Well, some of it is the services and environment that goes to support the research. So there's a building there and there's people who work at the building and there's graduate students who come and bring their own money for things. And so some of that is just the additional economic activity that has to happen to support that $1. But then many of those studies are then perhaps followed up at other places which then spend more money on that. Or if industry takes interest in the research and they do something with it, that creates economic activity. And then because of something that we've had for a long time since 1980 called the Bayh Dole act, the universities are allowed to and in large part obligated to find a home for any technologies that come out of the research that they end up patenting with a licensee that is going to maximize the value of that patent. And so many, many university patents are licensed to drug companies and to companies in the IT space or aerospace companies or whatever IT is that that has a use for that patent. And then that creates more economic activity. But I think it's also important to remember that many of those dollars don't necessarily get multiplied. And that's because we don't always know where these things are going to lead.

Holden Thorpe (15:35)

Right. I like the way that you edited my original metaphor, because it's not as if there's a machine that turns every dollar into $5. It's more like the machine has this lottery function where some dollars create absolutely nothing and then other dollars are spent on something like Gila monster venom research in the 1990s that ends up becoming the basis for GLP1 drugs that becomes like a trillion dollar global industry. So the value of might not be just like a 5x payoff, it's something like a 10,000, 100,000x payoff. And you have this portfolio of bets that scientists are making as they're trying to discover truths about our bodies and the universe and the world. And some of them return something incredible and some of them return nothing. But we don't know which bets are going to return what beforehand. And that's why you want to essentially have the federal government taking a lot of bets. It raises a question. To a certain extent, the NIH and NSF can defend its spending with this sort of portfolio approach that I'm describing. But there's other people maybe say MAGA conservatives, MAGA tech folks who are saying, hey, you know, who's really good at making bets? Venture Capitalists, the private sector. Why doesn't the private sector just take on all of the science spending in America and just sort of, you know, put the NIH into the dustbin of history? It was a fine idea for the 1950s. It's not the best idea for the 2000s. Defend the idea that the federal government should be doing basic research funding.

Joe House (17:11)

Well, you're hearing a lot of that from the tech side, but you're not hearing a lot of that from the venture capitalists who invest in biotechnology or from the pharma business. They're very aware of the fact that the lead times are much, much longer, that there's far more risk. I mean, you discover some novel protein, it may in a test tube look like if you inhibit it, it's going to have a positive effect. But then you have to do it in a mouse, and then you have to do it in a dog, and then you have to do it in. Then you have to decide, are you going to invest a lot of money to do it in people? And for the venture capital industry to pay for all of that, all the things that would have to be done to find the one that works, that would be more than they can muster. And I should make a disclosure that I'm a venture partner at Hatteras Venture Partners in Durham. So I have a lot of experience watching these things and the ones that were developed in universities for a long time period before the venture capitalists invested in them. Now, on the tech side, there's a lot more of that kind of talk that the technologies because. And I think part of it is right now the two hottest things are quantum computing and AI. And it's true that the, that industry is making much more progress in those areas right now than academia is. And I know that because they're all competing to have papers in our journal and in our sister journal Nature, where most of these things tend to go when something big happens. And we are getting lots of papers from Google and Microsoft and Meta about this kind of thing. And so there is a lot of stuff going on there, but even that is rooted in ideas that were developed in academia. So I guess I would say, if I'm a venture capitalist, why would I start some skunkworks project where I'm going to spend money that I took from my limited partners with absolutely no idea where the product is going to be? I mean, most people who pitch to venture capitalists will tell you when I go in there, the first thing they say is, what's the product? Well, most of the time when this research is being done, nobody's close to the product. In fact, when I go to, when I talk to faculty members who want to start a company, the first thing I say is, well, what's the product? And some of them don't like hearing that. So that's not necessarily a natural thing. So the handoff between basic science and industry, and in particular the venture industry, really requires a shift, and it requires starting with things that don't come naturally to the venture capital industry at all.

Holden Thorpe (20:22)

Let's get to the news. The Trump administration has gone after science funding in a variety of ways. Thousands of people have been laid off from the NIH and nsf. Hundreds of millions of dollars in grants to scientists have been paused or cut. There are budget rumors of a 35 to 40% cuts to the total NIH budget. That would mean approximately 10, 15, as much as $20 billion cut from NIH funding, 50% cuts to the NSF. That would be another several billion dollars. We're essentially talking about cutting federal science funding into in half, in one year. What would a cut of that scale mean for American science?

Joe House (21:08)

It has many, many deleterious effects that could be very hard to reverse. I mean, first of all, there are lots and lots of students who are in the midst of their training who need that funding to finish. And one of the things that I think doesn't get said quite enough is if we're going to have less immigration, then making sure that we grow more scientists in the US Is going to be incredibly important. And so cutting off funding in the middle of someone's graduate school and losing that American trained scientist to the scientific enterprise is going to be way more costly than any money that you could save doing that. So I always start with that, because the young people and the young scientists that we're training in this country, many of them are very disillusioned right now. And I've talked to, I go to campuses all over the country, and I talk to students who have had their plans changed, and that is American talent that is being lost. And so I really have a hard time rationalizing the idea of restricting immigration and not training American students at the same time. That feeds on itself in a bad way. Then the second thing is a lot of things that were going on will be shut off in the middle. And those are things we already invested in. And so we're not going to recoup the money that was invested in that. And then the other thing that's going to happen is that not everybody will go, but there. But scientists love to do their science. That's why we're in this business. And if you can do it in another country, particularly if you have family over there or you don't have a reason to stay in the United States, then people are going to go off and do that. And then with what we've been talking about, there's the ideas that aren't going to be generated, the drug targets we're not going to have access to, the new ideas in physics that are going to give us the next advance in quantum computing, or the new ideas in computer science that are going to be used to make AI even better. Those ideas are not going to be generated, and we won't see that an industry right away. But in, in a few years or in longer, it's going to be very hard to build that back. And if it's built back, if it's, if it's rebuilt in other countries, well, some of those countries may be ones that we can collaborate with, and some of them may be ones that we don't collaborate with. And the thing that people need to realize is when the technology and the science is developed here, we have a lot of influence over how it is used. And so if we back off and China, for example, grows and it is growing there significantly. So we now get more submissions to our journal, which is one of the most selective in the world, from China, and we are coming close to, to publishing more papers from China than we do from the United States. And so as that continues to grow, if those ideas and those technologies are invented there, then you're essentially handing China, which many people see as an adversary, the ability to control how these technologies are developed, when that's something that the US has always had a long standing position in. And so giving up that lease on the future, not only is it a lost opportunity for the United States economically, but geopolitically, it's giving some country that we may see as our adversary the opportunity to control that part of the future. And that now I'm more on the side of thinking that all of science is one giant community. But for the people who care about the geopolitics of this, if they see China as an adversary, giving them the opportunity to develop all these technologies and receding from the competition doesn't strike me as a good way to do this.

Holden Thorpe (25:51)

Science has reporters that have talked to the Trump administration about why they're doing what they're doing, why are they doing this?

Joe House (25:59)

Well, I think it's on several fronts. I Think some of it is just the cost cutting and saying the bureaucracy is too fat and so we got to get rid of spending. People like you and other people who know more about economics will figure out in the long term whether they're actually saving money or not. I think that's already up for debate. I think the second thing is it's just a way to go at the highly educated people in college towns who are seen as not part of the American story in the same way right now. I think it's a way to get at other parts of the universities that are seen as more of a problem by the administration. And then I think beyond the budget part, there's another part which is regretting or upset about the way things went in the pandemic or in the case of Robert F. Kennedy Jr. Long standing ideas that he has had that are counter to what mainstream science believes. So there's many, many elements to this. And one of the things that I like about what you're doing and what a lot of people are saying is that we can't just assign all these evil intentions to the people who are attacking us. We have to talk about what we've done to make it so that we're attackable and the ways in which that we have contributed to losing the confidence not so much of the politicians, but of the American people. And we need to be clear eyed and able to talk about how to reform within our world. And that's difficult in some of the echo chambers that we live in, as I'm sure you hear about some of the things that you write about. When I say things like, well, here's a place where the scientific community could have done better, I persist with that. But there are plenty of people in my world who think that's giving too much credit to the people who are attacking us and complying in advance or bending the knee or whatever it is the people in the echo chambers like to say. So I think talking about how and thinking about how we contributed to this and how we can change that in the future is really important.

Holden Thorpe (28:43)

Well, let's go right at that. Why did this happen? Why has science lost so much credibility and trust in the last few years among Americans? I mean, I know that this question is going to be seen by some as an act of victim blaming scientists who are having their livelihoods stripped away by the Trump administration. How dare I ask whether they deserve to have their careers destroyed. But there's an objective truth at the heart of this, which is that not only did the Trump administration get elected in a free and open election. But the preference for Republicans over Democrats in part reflects an enormous decline of trust in a range of institutions, one of which is universities and another being public health and American science. We're going to get to universities in a second, so please trust that. I'm going to set you up on that in a moment. But what do you think has happened to the decline of public trust in American science in the last few years?

Joe House (29:50)

Yeah, well, I think first of all, it's important to be really quantitative about what it means. The number of people who say they have confidence or a great deal of confidence in science is still well north of 70% in the latest Pew study. It was in the 80s before the pandemic, and we lost some during the pandemic. And it's actually started to rebound a little bit as some of the trauma of the pandemic has started to wear off. But I think some of the things that we did were specific to what happened in the pandemic. And part of that was being overly passionate and bombastic, even about the, the. Some of the ideas that were coming from the politicians or overstressing how important the scientific aspects of the pandemic were without weighing as many as much as maybe we should have the economic and psychological tolls that the pandemic had. And I was one of the people who was too bombastic during that time, and I testified in Congress that I should have done. I overplayed my hand. So I'm one of the people who's been pretty transparent about that, mainly because we tried the experiment and it didn't work.

Holden Thorpe (31:10)

Let's hear something specific. What did you get wrong? What did scientists get wrong that might have hurt public trust?

Joe House (31:19)

So I wrote several columns and so did many others and many scientists who spoke on social media about how we should always prioritize the health and the public health of everyone who might be affected by Covid, and that we shouldn't even wonder whether schools should be closed or open or not. And in retrospect, I don't think there's anyone who could look back now. It's always easier when you look back. Who would look back and, and not say, well, we could have opened sooner here, we could have opened sooner there. For universities, for example, in the fall of, of fall of 20 was pretty tough, but the. Certainly the spring of 21, it was time to, to get rolling there. And there were plenty of schools that stay. Stayed open and did well. And rather than writing columns about how that was too much Risk. I think we should have been more appropriately saying there are risks associated with this. People other than scientists need to decide how that balances against the educational and psychological and sociological risks. And that's why we have politicians. And so we're saying this, and if you have any questions, let us know. And we're glad that we're not the people who have to make these very hard decisions, rather than saying that the whole thing was obvious. And there are certainly several pieces where I did that, tweets that I sent out, but so were many other people who were writing in the pandemic during that time. And I think the lesson of that is that now we need to do a better job of helping people understand what the facts are and letting the politicians assume the responsibility of making the hard decisions about what to do with that.

Holden Thorpe (33:25)

I'm really glad you said that. I think this sort of self critique is really important. And I also think that, and I'm sure you agree the attack on science today is about something bigger than punishing scientists who experimented with activism and advocacy in 2020. I mean, the Trump administration isn't just cutting NIH, NSF funding at the agency level. They're going after very specific universities directly withholding billions of dollars of science funding from Harvard, Columbia, Northwestern, Cornell and more. And this seems distinct. There's a stated justification here that these cuts or threatened cuts are about specific, often prominent elite universities failing to address campus antisemitism or violating certain government policies related to diversity programs. Do you believe the Trump administration is being honest when they say this is why we're going after these universities, or do you think it's about something else?

Joe House (34:35)

Well, I think in a way that we were talking about with science, higher education has also lost a lot of connection to the American people. Of course, I agree that these attacks are not healthy for the future of America. But I don't think you can engage in fighting back in that without asking how higher education contributed to the loss of confidence that has happened. In the case of higher education compared to science, the loss of confidence is much larger than the loss of confidence in science is. I think that's because of a number of different things. One is the fascination that a lot of the media and just people in general have with the very elite, very selective universities, which do things that most people who go to college don't experience. The classes are small. There's a lot of resources there. And, you know, I wish people would think more about public universities like the University of North Carolina or even, you know, Land Grant universities that are in the south or in the middle of the country, they're training a lot of students. They're doing agricultural extension service. They're. They have big academic medicine efforts where they're training a lot of doctors. They're doing a lot of the things that, that Americans want from higher education. Not that. Not that Harvard and Yale aren't doing that too. They're just not doing it on the same scale. And so if you look at kind of the problems that you have at Harvard or Yale versus the number of students they're training and the number of people benefiting from it, it's easier to come to the conclusion that that's not a good calculus. And so I think what has happened with universities is that the things that they tend to emphasize success in research, which drives awards and Nobel Prizes and rankings and all those things over the things that Americans expect from higher education, which to me, the top three are undergraduate education, medical care, and athletics. These are the three things that most people who aren't affiliated with higher education know about. And universities, and I am one of the people who got the most involved in this, get all kinds of turmoil about whether athletics is being done properly. The public looks at that and say, look, I want to watch the games. Why are they. Why are they having a fight about this? But more specifically, in terms of the mission, undergraduate education tends to be the thing that people think about the most. But if you look at how universities behave, the faculty are often trying to get out of teaching more so that they can do more research, which is where the rewards are. And that is what the university tends to congratulate. Research is what the university tends to congratulate itself about. And in academic medicine, it's the same thing. A lot of the docs are trying to get out of working in the clinic so that they can do more research. But most people, if they drive by an academic medical center, they think what's going on in there is that they're mostly taking care of patients and training residents. So I think the universities need to recalibrate the functions that they focus on more to what the American people expect. Now, that's a pretty strange thing to hear from the editor of Science, who mostly benefits from their focus on research. But I'm not actually that worried about whether the universities can continue to produce great research if they have the funding to do it. I'm more worried about whether they can have the funding to do it. And that is predicated on meeting the expectations that the American people have, because the Academic freedom that we enjoy was all given basically in a bargain that we were going to do education and produce research in the public interest. And we're starting to lose academic freedom because people don't think that they're getting anything out of the other side of the contract.

Holden Thorpe (39:18)

I'm trying to reconcile two things that you said in the last few minutes, because one thing you said is that the model of science that was gifted to us by Vannevar Bush was all about the free inquiry of free intellects. It was the practice of science untouched by political threat or political interest. But right now, it very clearly seems to me that the Trump administration is using explicitly political goals to try to force these universities to change their policies and even to make billions and billions of dollars of research grants contingent on those political changes. So wouldn't you agree that the Trump administration, in its dealings with these universities, is making research funding exactly as contingent as Vannevar Bush hoped we never would?

Joe House (40:24)

Absolutely. And you could view that two different ways. You could say, yeah, he was right that this was going to be done eventually, or you could say, wow, it's amazing we made it 80 years before this happened, because it really was done basically in good faith. The universities willingly gave the leverage that is being used to do all this to the federal government. And they were either naive or just appreciated the money so much that they didn't want to think about it. When the grants continued to grow and the federal government's leverage continued to increase. They were just assuming that the, the, the walls that had been breached would, would continue to hold. And you know what? What Vannevar Bush worried about is being tested.

Holden Thorpe (41:27)

Now, I'm remembering what you said about the cost of decimating science when you said, and this was so smart. It's not just about the projects we are ending. It's about the projects we won't be able to do in one year, five years, 10 years. It's about the scientists that we won't be able to train in the next five and 10 years, the scientists whose careers are being destroyed right now by the depth of these cuts. Do you think what's happening right now at the NIH is going to change the face of American science forever?

Joe House (42:01)

I hope not. And I think the optimistic scenario is that especially support for biomedical research is politically popular. I mean, the Washington Post had a poll out showing that the American people favor not cutting biomedical research by 77 to 21. And there was a hearing that Susan Collins led where my boss and Barry Slickman, who runs the cancer center at UAB and several other people testified and senators on both sides in that hearing were very supportive of biomedical research.

Holden Thorpe (42:49)

You are mildly optimistic or at least holding out hope for the possibility that these proposed cuts to NIH and NSF will prove so horrifyingly unpopular in polls and be so unpopular in the Republican controlled Senate that we might be looking at cuts that don't actually materialize.

Joe House (43:11)

I'm hopeful that that's the case and it would fit with the history. I mean we've gotten bad presidents budgets for science for a long time and the Congress has usually rectified that and the NIH budget has grown over time. Now there are two things that are one is obviously the Congress is not quite as assertive now as it has been in the past and the other is it's unclear who the champion would be. So the really good NIH budgets have come when there was a Democratic president and a Republican Senate because the two Republican senators that have contributed the most to the NIH growth, In fact the two senators of either party who have contributed the most are Arlen Spector and Roy Blunt. Those two folks both took the NIH as an important issue and they both Arlen Specter led the doubling of the nih and the increases that we got when Obama was the president were led by Roy Block from Missouri. And it's unclear who is filling that role now, but someone always came into that role because biomedical research is politically popular. Cutting the NIH is, has always been politically unpopular. And I don't think the administration. Well, like I said, there's two elements to this. There's the cutting and the saving part. There's some of the bad karma about COVID and vaccines that were also dealing with. But from the cutting part, I don't think that the administration would be cutting the NIH this way unless it were part of their overall program to reduce the influence of universities in general. So if the Congress is able to work around that somehow, then maybe these cuts won't be as bad as they appear right now. But that is not to say that there haven't been extremely distressing and anxiety producing things happening. And every scientist who works at a university is perfectly justified in worrying right now and they deserve as much support and reassurance from their leadership as they could get.

Holden Thorpe (45:38)

Holden Thorpe, thank you so much.

Joe House (45:40)

Derek. Great being on with you. Thanks for all you're doing.

Holden Thorpe (45:44)

Many thanks to Holden Thorpe. Next we have Bhavan Sampat, a researcher and historian at Arizona State University, to tell us the deep story of the American science system and how it was forged in the chaos of World War II. Bhavan Sampat, welcome to the show.

Bhavan Sampat (46:03)

Thanks for having me. Longtime listener. It's great to be on this side of the mic.

Holden Thorpe (46:07)

You are here to help me tell a story, and the name of that story is the Birth of the American Innovation System. I want to understand how America went from a country with practically no federal science policy to a country whose science policy leads the world, guides the world, where the NIH is arguably the most important scientific institution in the world. And I think the major line that we need to draw here in our storytelling is that there's a science policy before World War II and there's science policy after World War II. So let's start with before World War II. If you reach back all the way to the 1800s, invention, it seems to me, is mostly a solo endeavor. When we hear about and read about the inventors and scientists from the 1800s and before, it's typically solo tinkerers. It's Alexander Graham Bell, the telephone, it's Elijah Otis, the working elevator. These guys worked alone. And even the idea of invention as teamwork doesn't really make it to the mainstream until Thomas Edison and his research lab shows off the incandescent light bulb and the phonograph and parts of the telephone and the working video camera with everything that they build in Menlo Park, New Jersey. But around the same time that the work of invention is evolving from individual tinkering to team based work, there's another critical vein of invention history that we have to tap into now that also clicks into the 1880s, and that is the birth of the hygienic laboratory. What is the Hygienic Laboratory? And tell us how it becomes the National Institutes of Health.

Bhavan Sampat (47:55)

The NIH traces its roots back to something called the Marine Hospital Service, which eventually becomes the Public Health Service. But as the name indicates, it's originally created to help provide relief to sick and disabled seamen. And this is, I think, the late 1790s. And it establishes in 1887 a hygienic laboratory to study infectious diseases again in the context of marine needs. This hygienic laboratory is moved in 1890 or 1891 to Washington, D.C. and starts to do work on epidemiology, including some work on vaccine development. The hygienic laboratory was very much intramural in the sense that the research was being done at the headquarters of what eventually became the National Institute of Health. And that happens in 1930 with something called the Ransdell act, which transforms the hygienic laboratory into the national institute singular of health, broadening its scope beyond just infectious diseases to start to think about things like cancer as well. Though at the time funding was quite limited for anything.

Holden Thorpe (49:07)

So let's pause the story here in the 1930s. Is the federal government a major funder of any scientific research in this decade? Or is science funding in the 1930s mostly coming from, say, private philanthropy?

Bhavan Sampat (49:21)

The US federal government was not a significant funder of scientific research before World War II, with the exception of agriculture. Agriculture was one field where there was considerable funding. In medicine, it's private foundations, primarily the Rockefeller foundation, also some others that are providing the money for funding scientific research. The other thing that is happening is you start to see in the late 1800s the rise of American research universities, beginning with Johns Hopkins University in 1876, University of Chicago in 1892. You start seeing new universities being created on the German model of actual science focused universities, right? And then the old colonial colleges and others start to transform in the image of the research universities, so the Harvard's, Columbia's, et cetera. And then you start to see interaction, university industry, interaction between the firms that are kind of increasingly starting to dip their toes into research and some of the universities that have started to do so as well.

Holden Thorpe (50:30)

World War II changes this. World War II invents everything that modern Americans think of when they think about American science and our innovation system. Let's slow down to tell this part of the story very deliberately. What happens in 1940?

Bhavan Sampat (50:48)

So it's June of 1940. German armies are marching across Western Europe and the US realizes it's going to get involved in World War II, and that the side with the better technology and the side that's best able to harness science to solve wartime problems is likely to have a significant upper hand. And this is both good news and bad news. It's good news because there are scattered scientific and technological capabilities that have built up, as we just talked about in the interwar period, in firms and universities across the country. The bad news is that there's little coordination of those pockets of excellence. There's little coordination of science and technology and no serious government apparatus to kind of harness the scientific and technological capabilities in the country. So there's this urgent need to mobilize science and technology to help confront the crisis. So 18 months before Pearl harbor, the US sets up what eventually would become called the Office of Scientific Research and Development, and it would be run by Vannevar Bush, who was formerly vice president and dean of engineering at mit, I believe at the time he was at the Carnegie Institute of Washington. He's this kind of patrician science policy. I don't even know if they called it that at the time. But a leader of the scientific establishment. And he, together with others, James Conant, the president of Harvard, Carl Compton from mit, the who's who of science at the time, kind of get together and convince FDR to set up osrd. What eventually becomes osrd?

Holden Thorpe (52:35)

Tell me how OSRD works. There are people around the country, as you said, little pockets of genius around the country that might have little ideas that, if activated, can build a product that will help us win World War II. How does Washington, how does OSRD find, evaluate and fund these science and tech.

Bhavan Sampat (52:56)

Projects over the course of the war? It enters into R and D contracts north of 2000. R&D contracts with universities and firms across the country involving most of the top universities and firms in the country at funding levels that are a pittance by today's standards, but really order of magnitudes greater than what had been funded previously. OSRD had a choice. It could have done things internally. It could have brought scientists say, come to Washington and do stuff here intramurally. But Vannevar Bush and others first thought that to be impractical, but second, didn't want to disrupt universities. And universities were very much reluctant to be disrupted, at least in that way. So they decide they invent. And maybe that's too strong of a word, but this is really the first serious exercise of the extramural R and D contract. They develop procedures for funding extramural research, including procedures around how to write grants, patent policy with respect to grants, indirect cost recovery policy with respect to grants, reporting requirements, et cetera. And so that's the vehicle. And a number of historians have said that the R and D contract is one of the most, perhaps with some exaggeration, one of the most important inventions of World War II.

Holden Thorpe (54:28)

I think the Office of Scientific Research and Development is the most important American institution that nobody has heard of. I mean, truly, outside of you and me and maybe six other people, I feel like practically nobody has heard of osrd. But if you've heard of the Manhattan Project, well, it was spun out of osrd. If you're familiar with radar or the microwave, those were OSRD funded inventions. In the early 1940s, a small group of European scientists came to OSRD with a mold that they called penicillin. They essentially said, we can't scale and test this new potential drug because Europe is submerged in war. Can you help us. And OSRD produces and scales and tests and delivers the medicine we know as penicillin. Maybe the most important medical breakthrough of the 20th century, the beginning of the antibacterial revolution, flows from OSRD. So after the war, just before Franklin D. Roosevelt dies, he asks Vannevar Bush to write a kind of thesis statement about the future of science. And Bush does not disappoint. He writes a very famous document, maybe one of the most famous documents in scientific policy history called the Endless Frontier, where he assigns government a starring role in the funding of basic research for the foreseeable future. Pick up the story here. How does the legacy of World War II help to create the innovation system that we still have today?

Bhavan Sampat (56:12)

Yeah, so it's important to note that even before the war was over, OSRD's effort was so successful that everybody knew that the era of small government and science funding was over. Essentially, we'd be funding science, we'd be funding. The federal government would be supporting research after the war. And the question was how, how we would do it. And I think it's important to recall that Bush wasn't the only player on the scene. Bush's kind of nemesis or frenemy during the war was a gentleman named Harley Kilgore, a senator from West Virginia who was an old New Deal Democrat and in some sense very different from Vannevar Bush. Vannevar Bush was this kind of patrician, Boston Brahmin, scientific elite kind of guy. Harley Kilgore, I have a picture of him in front of me right now is this kind of old slumpy West Virginia lawyer. Really didn't. I mean, he would admit that he doesn't really know all that much about science. He had a few staffers who did. He was really concerned about economic concentration. So he kind of made his name around anti monopoly kind of stuff. There was a lot to like about the OSRD effort from the perspective of Harley Kilgore. But there was other things he didn't like. He didn't like the fact that a lot of the money was going to friends of Vannevar Bush. Right. He didn't like the patent clause. The firms that got money got to keep the patents, and he thought that was a giveaway of public patents. Other parts were that it shouldn't be elite scientists in charge, but essentially it should be politicians, for lack of a better word, in charge of the process. Like he kind of said during wartime, okay, I get it, we need this elitism. But as we transition to peacetime, we need More political control. We need more democratic accountability of science.

Holden Thorpe (58:24)

And on the other side we have Vannevar Bush. What's his basic argument?

Bhavan Sampat (58:30)

Vannevar Bush was trying to make the argument to first taxpayers and politicians that funding pure basic research, fundamental science, with the scientific community calling the shots, can help. Will help is the thing that will help generate national security, health and economic welfare going forward. And that should be the cornerstone of post war science and technology policy. He was making the case not just to taxpayers and politicians, but also to the scientific community. And the case of the scientific community was it's not going to be like the war. There's going to be under the new National Research foundation, considerable scientific autonomy and self governance.

Holden Thorpe (59:28)

So while Bush and Kilgore are having this debate about what the future of American science should look like, something else is happening. Something that seems so unimportant at the time, but turns out to completely reshape the way science works in America. All of these medical research grants that were being administered by a group within OSRD called the Committee for Medical Research. This is like the wartime committee that's in charge of distributing all sorts of research research grants to fund medical stuff that can help our Navy and Army that's being wound down, the war's over and these grants that are outstanding in 1945, they have nowhere to go. What happens next?

Bhavan Sampat (60:18)

That's exactly right. Again, Bush wanted medical research at the National Research Foundation. Others had other thoughts. Nobody really thought it would go to the nih. But while Bush and Kilgore and their allies are fighting it out over the 1945-1950 period, something has to happen to all the research, including the medical research. And through some deft political maneuvering, Surgeon General Perran, NIH Director Dyer they kind of exaggerate the competencies of the NIH and the experience of the NIH in this particular arena. And they end up getting the remaining 40 plus the 40 plus CMR grants that have not yet expired. That triples the budget of the NIH and forms the basis for what eventually becomes its extramural research funding program. Nobody thought it would be a big deal. So much so that they hire. The person they hire to be in charge of it is the former Public Health Service venereal disease head Cassius Van Slyke, who's at home recovering from a heart attack. And they describe it to him as, and I'll quote here, an incidental lower left hand, side drawer type of activity. He positively wouldn't have to work more than two hours a day and probably not more than four or five hours a week. Wow, okay. Then a few things happen. First, a lot of the CMR contracts that were pending at the end of the war, they were for buying penicillin for a range of reasons. The price of penicillin tanks, that basically results in a budget windfall, increasing the budget further. Then Van Slyke and one of his colleagues, whose name I can't recall right now, they send out a letter which is sometimes, at least in science policy circles, referred to as the most naive letter in the history of Washington to medical school deans saying, we have limited funds available for research purposes. If you have investigators who need these funds, please let us know by return mail. And the response then is overwhelming. And that is the beginning of the NIH extramural research program.

Holden Thorpe (62:46)

Before the war, medical research in America is about $40 million, almost all of which comes from the private sector companies, philanthropies. Within a decade, the US government is spending $240 million on medical research. Today, federal funding for basic research is $50 billion. So this naive letter begins a trickle, which begins a flood. And now we have the modern nih, the modern crown jewel of global bioscience research. There's a couple habits and customs of the NIH whose history I want to trace. One of them is peer review. How does peer review come up in the modern nih?

Bhavan Sampat (63:36)

So, like a number of early NIH policies, they end up just following the model or adapting the model that was developed by the Committee on Medical Research during World War II. So I didn't go into it in too much detail, but cmr, they had to decide how to spend the money. And they had this two stage process. The first stage is they would have scientists from across the country. And during the war it was actually they were based at the National Research Council, evaluate proposals for their scientific and technical feasibility. And in the second stage of the process, they would have osrd, Vannevar Bush, together with representatives from the military, evaluate them for practical considerations in terms of how important they would be for helping win the war. That two stage process, the dual peer review process, is adapted, more or less adapted, by the NIH in its early days, where they have a system of what they today called study sections, where scientists from across the country are consulted on the scientific and technical merit of proposals. And the second stage of the process is institute directors and advisory councils make decisions about whether they fit the mission of that institute or the NIH overall.

Holden Thorpe (65:05)

Why is peer review controversial?

Bhavan Sampat (65:07)

Peer review is controversial for a number of reasons. And the reasons for which it's controversial have kind of varied over time. I think today one of the big reasons it's controversial, one of the kind of criticisms of NIH peer review today is that it's just really, really hard to evaluate, to reliably evaluate the scientific merit of large numbers of proposals at scale. That was a problem that Vannevar Bush anticipated, which is partly why I think he preferred institutional grants over bottom up peer review. Another is that it's an old boys club, that scientific peers give money to folks who look like them and that or think like them. So there are certainly, and this goes Back to the 1960s and 1970s, questions about whether the funds are being allocated to the best scientific ideas or to just if there's biases against certain racial ethnic groups, whether there's gender biases, whether there's biases against certain institutions that just don't have as much experience with grantsmanship and then biases against certain ideas. So in Alzheimer's, for example, the amyloid hypothesis, if you didn't do stuff along those lines, it was hard to be, I'm not involved in this, but at least anecdotally, it was hard to get funded on that kind of stuff. So those are some of the reasons why it's controversial. I think today, at least over the last five years or so, there has just been a lot of concern about just how burdensome the process is, how burdensome the process is to the agency, but also to the actual individual investigators and universities that have to write the grants, that have to go through all these requirements in order to get the money. So those are some of the margins on which it's controversial.

Holden Thorpe (67:27)

One theme of my Research for Abundance is that many of the early NIH leaders like James Shannon, very explicitly warned that the NIH could not be allowed to turn into a bureaucracy that would waste scientists time. In 1946, Cassius Van Slyck wrote in the journal Science that quote, it is not desired that the preparation of these reports, meaning grants for applying for research, present any long tedious burden, end quote. Ten years after that, 1956, James Shannon, himself, one year into the job of NIH director, wrote an article in Science the journal, where he said, quote, the research project approach can be pernicious if it is administered in a way that it produces certain specific end products or provides short term periods of support without assuring continuity, or, or if it applies overt or indirect pressure on the investigator to shift his interests to narrowly defined work set by the source of money, or if it imposes financial and scientific accounting in unreasonable detail, end quote. This is just a stunning statement to look back on 70 years later because in so many cases, the very warnings of 1940s and 1950s leaders of NIH have become a reality for scientists today. By some surveys, it takes 40% of the typical scientist time to fill out paperwork. We see now in the Trump administration efforts of the government to essentially, what did Shannon say? Apply direct pressure on investigators to study certain things and not study other things. And surely that political pressure in different ways predated the Trump administration. Tell me this, where did this pile of administrative burden come from?

Bhavan Sampat (69:39)

Congress, as you mentioned, the early leaders, Van Slyke, Shannon, following Vannevar Bush, really did not want administrative burdens to kind of bog down science. However, the NIH budget, as you mentioned, grew rapidly over the 1950s and 1960s, and there was concern from kind of government oversight folks in Congress that we might be force feeding the nih, we might be spending the money too quickly and not having enough political accountability. That's almost a Kilgorean theme, by the way, though I'm not sure if Kilgore would have agreed with it, as. As such, the big change is something called the Fountain Committee investigation in the 1960s, where Fountain complains about financial mismanagement at the NIH, but also this idea that, wait, researchers can just change their mind about what they're doing and they don't have to report back. Right? We don't actually know how much time researchers are spending on federal grants versus other stuff. He complained that NIH has this tendency to romanticize research and try to convince you that it can't be subjected to any sorts of accountability. Originally, his complaints were sort of mild, but Jim Shannon, in one of his rare political missteps, ignores them and says, yeah, it's not that big of a deal. And Fountain came back hard with other reports, a series of reports that are increasingly critical of the nih, oftentimes spotlighting very specific cases, not the agency as a whole. But what comes out of all that is the regulations. So you have in the 1960s the first edition of the NIH Grants Policy Handbook, which some investigators have, some scientists have described as a series of thou shalt and thou shalt nots. And that has just increased in size, adding more and more stuff, some of it good, some of it around civil rights and human subjects research, but it just accumulates over time and no one really thinks about it after it's on the books. So we have this accumulation of administrative burdens that Originates with Fountain and pushes towards accountability that I think today puts a stranglehold on big parts of the process.

Holden Thorpe (72:14)

How do you balance the fact that the NIH has clearly done so much good with the fact that like many bureaucracies, it has evolved in a path dependent way and rules and regulations around it have evolved in their own path dependent way in a way that has arguably created, let's say, mandate drift. That maybe peer review is not the perfect system for evaluating thousands of grant proposals and maybe the administrative burdens and regulations around the paperwork requirements of being a scientist in America. Absolutely. Take scientists away from their trained job to do actual science. How do you balance in your own mind the good and the bad and the bad here?

Bhavan Sampat (73:03)

Yeah, I think an important thing to say first is just that I believe, and I've done some research showing that on average, the NIH delivers high value for money. I just think that that needs to be respected and in some sense preserved in any serious NIH reform efforts. That said, the budget, as I wrote in that Brookings piece, the budget of the NIH has grown a thousand fold in real terms since the end of World War II. Our knowledge about how to effectively spend that money has increased, let's just say less fold. And I think what's going on, or my sense is that there's just a certain rigidity in terms of how it does things and a kind of closeness in how it does things. And there hasn't really been an openness to experimentation, to thinking about itself as a learning organization, that we're trying to do this really hard thing, which is spend money on, invest money in science to generate the most returns for us and global health. And the way we've always done things might not be the right way to do things going forward. There's been this sort of resistance to learning, to experimentation. And I don't just mean formal experimentation in terms of like randomized trials that you may talk about with, with Pierre Azoulay, but I just mean rolling things out in a way that we can evaluate them later on, sharing data so other people can take a look. And I think part of that resistance is because it has protected its budget at all costs. It's kind of been a little bit reluctant to admit that this is a hard business and we might not be doing exactly right. And in some sense, I think going forward, institutionalizing the learning function at the nih. Let's figure out how to do this better function at the nih. The science of science is a really important thing to emphasize in reform efforts going forward.

Holden Thorpe (75:14)

Bhavan Sampat, thank you so much.

Bhavan Sampat (75:16)

Thanks for having me on.

Holden Thorpe (75:18)

Many thanks to Bhavan Sampat. Finally, we have Pierre Azulais, a researcher at MIT who spent a ton of time and energy studying what's wrong with the American science system, what's wrong with the nih, and how we could reform it rather than destroy it. Pierre Azoulet, welcome to the show.

Derek Thompson (75:39)

Hello and thank you for having me.

Holden Thorpe (75:42)

Pierre, you're someone who spent years critiquing the American science system and the NIH specifically, and you're here to make us smarter about how America does science and how we can do it better. But I think we have to begin with the news here. The Trump administration has fired hundreds of scientists and public servants from Health and Human Services, fda, cdc, nih. They've threatened to withhold or have plans to cut tens of billions of dollars of scientific funding and punish universities like Harvard and Columbia for not working with the administration. Before you administer your critique of nih, what do you make of the Trump administration's science agenda?

Derek Thompson (76:30)

I feel terrible. I feel it's a hatchet job. And my sense is that the goal is to make Harvard cry, not to improve the system. I think universities and pointy headed academics have been kind of designated as the enemies. I mean, J.D. vance was pretty upfront about that. And so thinking about ways to cut, you know, academics down a peg or 2 or 10 is viewed as a feature. Right. Regardless of the effects. Right. And my, my concern is it's kind of like a contractor would arrive in the house and would kind of, without any plan, start hacking at a wall without investigating whether that wall is load bearing or just an ornamental wall, just to see what happens. And here is why I think this metaphor is particularly apt. Like, you know, if you do that in a house, nothing terrible might happen for 10 years, you know, because maybe another wall can take the load for a while, but eventually the house might crumble, like when you least expect it, like far into the future. I think the challenge is that if those cuts have adverse effects on the rate of medical progress, we're not going to find this out immediately. We're going to find this out, you know, over time. And where I say over time, it's over the next 30 years. And so that will make it, it will make it hard to attribute back to those reforms potentially. But that doesn't mean that the effects will not be felt.

Holden Thorpe (78:21)

You and I both care about American science working better. We agree that the Trump approach is counterproductive at best. I want to use our time together to discuss the right reforms for American science. So let's start with a big picture thesis statement. What's wrong with American science? What's wrong with the way the NIH has evolved over the last few decades?

Derek Thompson (78:47)

You know, I think we've had two last decades in some sense where NIH could have an NSF to some extent as well, could have really taken the bull by the horn and think about, you know, rejuvenating itself for the 21st century. And they haven't done it right. And I think it's worth like talking about this point in a bit more detail because it goes to the heart of what NIH does. And this is really a unique, it's a uniquely American idea that the way we're going to fund science is by letting scientists propose ideas. So it's bottom up, okay? There is no grand czar deciding, okay, we should do more like deciding in advance where we need to invest, right? So scientists propose ideas, those ideas compete with one another and other scientists, the peers of the people who propose those ideas are going to adjudicate the relative scientific merits of those ideas. Because not every idea can get funded. That's always been the case. That certainly has been more the case in the past 10, 15, 20 years where the way to think about it is that roughly between, roughly 1 in 10 proposal actually gets funded. Maybe a little bit more than that, but not a lot more than that. It's very, very competitive. To me, that is what is most beautiful about the American system of supporting science is that it is bottom up, it is decentralized. And then the NIH has evolved some creative ways of within that system being able to be responsive to specific demands from policymakers who sometimes with very good reasons, feel like this area is being left unexplored. And so there are some kind of escape valves to allow the system to be responsive to those desires without kind of, without it devolving into some sort of like a series of top down dictates. And so whatever we're going to think about to try to reform nih, my claim is that that's something we should try to protect at all costs.

Holden Thorpe (81:11)

Pierre, from reading your work, I've gathered that you have several big ideas for fixing American science. And I think the biggest theme from your work is that the most important science funding institutions, nih, nsf, are biased against high risk, high reward research. In other words, that our system of paying for science rewards incremental and cautious science that's less likely to push the Frontier forward. Why does that bother you?

Derek Thompson (81:43)

So what's interesting is that, you know, there is, there has never been a funder of science in the history of the world, right. Who prided itself on, on, you know, encouraging low risk, low reward research. Like if you ask nih, they're going to tell you. But that's our all reason for being, of course, this is what we do. What's so great about risky research, right? And we can talk about what risk is in this context is that we have lots of reasons to believe that really important discoveries do not, you know, there are no easy pickings, there are no low hanging fruit. And so if you're not willing to risk failure, you're unlikely to discover something that's truly, that's going to have a truly transformative impact on science, on health. Now that doesn't mean that you would want everything to be high risk, right? It's kind of a portfolio. And so there's place for incremental advances as well. And we shouldn't dis them. But I think the concern is that in a harsh light of day, if you look at the NIH portfolio, there is really no mechanism that exists to systematically encourage scientists to take on, you know, bigger questions than they would otherwise. And that doesn't kind of invite, that is not kind of forgiving of failure along the way, at least as an intermediate kind of possible outcome.

Holden Thorpe (83:24)

Let's focus our analysis here on peer review, which we've explained a bit in a previous conversation. What's your beef with peer review?

Derek Thompson (83:34)

So we talked about how those proposals, right, emerge from the bottom up, right? So who's going to judge those proposals?

Joe House (83:43)

Right?

Derek Thompson (83:43)

They're going to be other scientists. There's going to be a committee, right. And that committee is going to evaluate the merits of the proposal and they're going to kind of vote on it. They're going to give it a grade, right. And there are lots of ways in which you can organize this process. But the concern is that, so there, there are, you know, conservatism could creep in at lots of levels, right? They're what the scientists who are proposing the ideas could perceive as being the system, whether or not it is true. Right. And it could kind of like choose more incremental projects because they feel, perhaps wrongly that they will be penalized if they don't do so. So that's one thing that could be happening, right. The second thing is that the people who are doing the evaluation themselves think that they, their job, part of their job is to ensure that you know, the taxpayer gets its money's worth, and so they have a duty to pick stuff that is likely to work out. And so very often that means the con, the combination of those two things means very often when scientists apply for a grant, they've done half of the grant already. They have that in their back pocket. They have preliminary results, and if they get the grant, what they're going to do with that money, they're going to complete the program. So they're going to do the other half and then wink, wink, not, not. They're going to use the second half of the money to prepare the next grant application so that when it's time to reapply, you know, they have new preliminary results that can see the next phase of their investigation. And you might say, well, that's just creative accounting. That's just like, it's just like, you know, when are you going to recognize there's a slight mismatch between what you promise and when it's going to arrive and when it's going to rewarded by the system. And you might say, no big deal. And I think the fear is that you're kind of on a treadmill. And if you step off this treadmill even for one nanosecond, it's going to be very, very hard to get back on the treadmill because no money, no ability to produce those preliminary results that would allow you to get into the system again, get into the queue again. And so because scientists are very aware of that. And remember, it's not just one individual typically, it's a team. Right. And particularly young people, careers are at stake. Like you want to, you want to make sure, you know, they stay on the team, that they're not fired prematurely, that they are not unemployed. Like you're like a small business person. Right. In some sense that might well lead you then to conceive of your research program as a series of incremental steps that are going to make sure that you and your team stay on the funding treadmill. And none of that research is necessarily bad, might be very valuable research. Right. But if everyone does that, then we're not going to take those kind of bold steps into the unknown.

Holden Thorpe (87:08)

Right. You're saying if the peer review process has a reputation of being biased toward plausible outcomes that already show some evidence that the experiment's likely to succeed, then science is going to move forward, one little teeny tiny likely success at a time. But maybe science should move forward instead. One big improbability at a time. That's how you discover Some of the biggest truths.

Derek Thompson (87:34)

Or at least it should also proceed like that.

Holden Thorpe (87:36)

Yeah, right. There should be a portfolio of incremental bets.

Derek Thompson (87:40)

Exactly. This word portfolio is really important. And the paradox here is, of course, scientists have a small portfolio of projects, right? So they're bearing a lot of risk. NIH as a whole has a very wide portfolio. They're very diversified. Like, they should embrace risk a lot more. Like they should want their. They should want their scientists to take on more risk. Like the conservative bias of individual scientists, right, who, like, they're worried about the continuity of their lab, they're worried about their career, they're worried about tenure. It's not rocket science. For why. There's a very legitimate reason. There are very legitimate reasons for scientists to be concerned about that. There is in some sense, I think, less legitimate reasons for NIH as an institution to be worried about scientists taking on too much risk.

Holden Thorpe (88:31)

And it's funny because you and I have spoken a bit in the last few months as I was doing research for the book, and I found that in 2012, this is just to give people a sense of how widespread this idea that NIH has become biased against risky and novel research. In 2012, the journal Genome Biology, which is one of the most prestigious journals in genetics, published published an article by Gregory Petzko, a biochemist and a member of the National Academy of Scientists. It was a satirical bit in which King Ferdinand and Queen Isabella of Spain in the late 1400s are mocking Christopher Columbus for not collecting preliminary data for his voyage across the Atlantic. And Columbus is essentially applying for an NIH style grant from the government of Spain to sail the Nina, the Pinta, and the Santa Maria. And quoting from the article, when King Ferdinand suggests that Columbus try a shorter trip, say, to Portugal rather than all the way to India or the Americas, Columbus exclaims, everybody knows that Portugal is immediately west of Spain. What will you learn from that? Not much of anything, Queen Isabella responds, but it can't fail now, can it? Besides, you've sailed to Portugal before, so the study section will know that you can do it. End quote. So this is satirical. And I'm not trying to suggest that people at NIH are being as absurd as suggesting that Columbus sailed from Spain to Portugal. But this is the general idea. There's a sense among scientists that the peer reviewers they're submitting their grants to are so conservative that you have to demonstrate your success to prove your experiment is a foregone conclusion, or something like it, in order to secure that next R01 grant for your research. Is that a fair sort of stylized summary before we move on?

Derek Thompson (90:23)

I think it's a great summary. I think it, you know, it hits close to home. And I think that, you know, if you, if you ask most scientists, including those who serve on those committees, you know, after a few beers, they would say, yeah, that's, you know, we're trapped in this system and no one likes it, but it's hard to see our way out of it.

Holden Thorpe (90:42)

So one major avenue of criticism that you've described is this institutional bias against novelty and risk and the sort of high risk, high reward science that's most important to push forward the frontier. I would say a second major criticism that you've made across your career is that the NIH grant application process has become so choked with paperwork that the typical American scientist now spends almost half of their time filling out documents rather than actually doing science. How did we get here that so much of the job of being a successful scientist today is understanding grantsmanship rather than doing science?

Derek Thompson (91:30)

I'm actually working on a big survey of management inside scientific lab, so we'll be able to get more granular data on that very soon. And I would say that grantsmanship is yet something different. Like they just spent a lot of time on the process of feeding the machine, supporting the lab. Right. But that grantsmanship is this really interesting word. It means a really important skill for a scientist to succeed in this system is having a knack at writing those documents in a way that will feel compelling to the referees, to the, to the committees. Right. And I think a very. And the big fear is that a lot of the stuff that contributes to make a grant seem compelling is not stuff that actually contributes to doing great science. Like there's a correlation there, but it's nowhere near one. It's much weaker than that. You know, when I, whenever I've sat, to kind of like think about writing a grant, which I tend not, you know, my research is cheap, so I don't have to, I don't have to do it as much. But there's, there is some benefit of kind of like, you know, thinking ahead about what you want to do, taking stock, you know, a little bit of perspective, taking a lot of sense making and putting that stuff on paper before you forge ahead, stream of consciousness style, that is not all bad. Right. But if 40% of your time is spent in just kind of like crafting the document in ways that you think is going to appeal. Right. That has an opportunity cost. Like you could have been at the lab bench you could have been teaching your postdocs or your grad students and something is being lost there. And so I think that's a meaningful thing to worry about. But paradoxically, I think the same thing that is responsible for not enough risk taking is also responsible for scientists spending too much time on the grant getting as opposed to the science, which is kind of an incentive system and embodied in this peer review, this process of peer review that encourages undesirable behaviors. And so there is a chance that a fix, and it's probably not a fix with a capital F, right? It's probably a number of fixes with lowercase Fs that reforming peer review in meaningful ways might actually stand a chance at both preserving the time that scientists spend actually doing research in their laboratories or in the field, and at the same time leading them to take on more risk where they think that there is a chance that those risks could be extraordinarily productive.

Holden Thorpe (94:48)

So let me try to do my best to represent NIH's defense here. And if there are people from the NIH listening, I can't promise that I'll muster the best possible defense. But this is just based on my own conversations with some folks at nih. I think one pushback from them would be, you just said, Pierre, that the NIH hasn't done anything to combat this impression that the agency spends too much on small bore, incremental science. But the NIH does have an hr, HR division, high risk, high reward. They have Pioneer grants that are designed to give younger scientists a bit of time and extra money to take these big bets. Why do you think the hr, HR, high risk, high reward division of NIH and Pioneer grants aren't going far enough. How would you essentially push back against the pushback?

Derek Thompson (95:43)

So first, I would totally acknowledge that they have done a ton to counter the impression that they are not doing enough. They're not encouraging enough high risk, high reward research that they have done. Right? But I think you need to think about how those efforts are designed, right? So I would say ultimately encouraging risk taking is about changing the incentives of scientists. You want to ground it in the incentive that scientists feel they labor under. And the biggest piece of the incentive system here is the time horizon that you have to plan your investigations, right? Bring something that might fail, at least in the short term, to fruition. You need a longer time horizon. And none of the so called high risk, Pioneer and blah, blah, blah type efforts at NIH kind of meaningfully change the time horizon that is provided to the scientists, right? So for example, in the case of the Pioneer grants, which are given to individuals, this is kind of a one off thing for them, right? So the way to think about it is that it's almost like a prize than a grant, right? We think you have a lot of promise here is a pot of money for five years. Go do some damage, which is good. But what we know with certainty is at the end of the five years, you're going back to gen pop, you're reentering the traditional system. So you still only have five years to try to kind of change the world. And that's not a lot of time. In my mind, what I see is the imprint of looking like we're doing something without really grounding the design of the program in a clear understanding of the incentives that scientific heads of laboratories actually face.

Holden Thorpe (97:59)

Pierre, We've done episodes on this show about science that fails to replicate. Like sometimes you get one gosh, well, result like, wow, we cured this cancer and then people try to replicate it and it fails and fails and fails and it turns out the finding isn't real. How do we balance the value of science that replicates with the value that we always want to be trying new things in science? This seems like a really tricky question, like how can it be that the NIH is, say, too biased toward funding science that's highly plausible, but also that clearly we need a certain number of studies that really are just trying to redo old studies to see if those findings actually hold up.

Derek Thompson (98:47)

You know, scientists left to their own device, in their own kind of intellectual communities can be quite good at developing indicators of progress that turn, turn, that will turn out in retrospect to be flawed. Right? So they are like, like they are, they are prone to fads, they're, they're prone to group think. The committees that allocate the funding can be captured by a particularly vocal, cliquish school of thought. And so one aspect of NIH reform is certainly to think about how do we make it easier in some sense for there to be kind of disruption into the established paradigms kind of on this, on a pretty regular clip, so that there are no sacred cows, or at least they don't stay, they don't stay sacred, you know, for long. Right? So like, very often, you know, scientists will become ensconced in the system. They've been very successful at it, you know, for very good reasons. They did a lot of things that were, you know, that really pushed the field forward. But then what, you know, my research shows is that maybe Also, they overstay their welcome. Like maybe at some point they, you know, they find themselves and their disciples and their former students in a situation where they can kind of really influence the direction of future investigations. And, you know, they become kind of obstacles to, you know, further progress for new hypotheses for new phenomena that are worth kind of investigating.

Holden Thorpe (100:40)

Turning now to solutions, we've talked about how the NIH funds science on a project by project basis. There's another way to go here, which that we could pay people rather than projects. We could give scientists money to just be themselves and explore their own natural curiosities rather than only fund a particular idea that they have for four or five. You've done some research on the Howard Hughes Medical Institute, which takes the latter approach, which gives a ton of money to individual scientists without asking them to apply for specific projects. So this is the people over projects model in a way. How does that work? What did you find?

Derek Thompson (101:22)

So, yes, so I studied scientists at risk of being funded by those two systems which embody different incentives. Right? So you're right. And that's the difference that people anchor on. Typically, NIH funds projects roughly. There are some experiments here and there, but that is the model. And HHMI funds people. Right. And then I was comparing those two systems and I was comparing the extent to which scientists under HHMI seem to, quote, unquote, succeed more or less than their NIH funded counterparts. And here my emphasis was because I was very worried about this issue of risk or that was the kind of the underlying thing that HHMI funding was, was supposed to enable. I was thinking about whether HHMI funded scientists succeeded in the sense of writing scientific articles that are kind of blockbusters, that kind of garner an extraordinarily high amount of citations and also at the same time more likely to sometimes face plant like basically produce articles that no one seemed to particularly care about. Right? Because should take on more risk. This is kind of what you should expect relative to their NIH funded counterparts. The number of publications, and certainly the number of publications per dollar was very similar in both cases. Right. Now the question is why? And here maybe I'm going to surprise you. I'm going to offer a spirited defense of project based funding as opposed to person based funding. So HHMI does a number of things. It's kind of a bundle. It's a bundle of attributes, right? So one is they fund people. And that's kind of the tagline fund people, not projects. That's the HHMI bumper sticker. And they're very Proud of it. But they are doing other things that I think might be more important. First, they are providing this longer time horizon. Like HHMI funded scientists in the era that I was studying effectively had 10 years to prove their mettle. They were basically given a pot of money at the start and they were being told, go change your field. Come back in 10 years. I'm caricaturing, but not that much. Whereas at NIH, it's more like four or five years max. And then the ax falls. What have you done for me in these past four years? And you better have a lot of lines on the cv otherwise you're not getting renewed. Right. There was that aspect of this, there was an additional aspect of insurance, which is that, you know, HHMI is a real. It's a grant. I mean, it's, it's not a prize. Like you need to stay hhmi. You want to stay hhmi. And it's very prestigious. You know, you're like Stanford and hhmi, MIT and hhmi. It's like that's how you. That shows up when you on the authorship list. So it gives you a lot of freedom, it gives you resources, financial resources to do your research. And the thing is, if you actually don't, if they think you haven't changed your field after 10 years, the axe falls and then you lose your HHMI status. But what HHMI does is that they don't kick you out suddenly. They ram down their support over the course of two years so that you have the time to kind of like adjust because you're going to probably need to start competing for NIH grants again. Right. And the fact that you know that insurance exists might well lead you to take on more risk at the start of the period.

Holden Thorpe (105:19)

Right.

Derek Thompson (105:20)

You know, it's worth it. You have a long time horizon to plan. So we have evidence, we have pretty strong evidence to suggest that what HHMI does is actually to encourage risk taking. But why it succeeds in doing so, I think is very controversial. And very often the people not project aspect is emphasized. I'm skeptical that is the real reason, but I could be wrong. And I wasn't able to kind of unbundle the different attributes to figure out which is the one that is really responsible for, for the results that we observed.

Holden Thorpe (105:59)

There's so many thoughts that leapt to mind as you gave that last answer. The first is that there is a sports metaphor here, which is that in baseball you have this concept of launch angle, that in the last few decades, the launch angle of swings in Major League Baseball has increased. And this higher launch angle increases both the odds that any particular swing will end in a home run and the odds that that swing will end in a strikeout. And there's a way in which what you're saying is whether or not launch angles are good for baseball. And we can sort of debate either way. There's something very important about thinking about launch angle in the context of building out an ideal scientific portfolio. You want to encourage some scientists to have a high launch angle that says, I have this idea, and I think if it succeeds, it can be massive for my field, but there's a decent chance I'm going to totally whiff. And if you have a classic NIH R01 grant that might be two and a half years into this process with nothing going for me, and that's going to look really, really bad for my career. And so we want to think about ways of designing for scientists to increase the launch angle of their swings so that, for example, you're thinking about it through the lens of time. Rather than fund you for four years, I'll fund you for 10 years. That's a really cool idea. And it brings to mind that in the 1950s, right as the NIH and the NSF were launching themselves, there were two other invention models that were coming of age in that period that I think are really interesting. One is Bell Labs and the other is darpa. So Bell Labs was the research and development lab of AT&T back when the government had essentially sanctioned AT&T as a monopolist. So AT&T was allowed to essentially operate as a government sanctioned monopolist. And they had this R and D lab called Bell Labs, which famously invented all sorts of stuff, the transistor, the solar cell. And scientists at Bell Labs had many advantages. But one of their advantages, speaking of launch angles, is that they were at a science lab owned by the largest and most profitable company in the world that was never going to go out of business. And so you could be this scientist at Bell Labs and, and take on a bet that might last 10, 20 years. Your company won't go bankrupt. You don't have to worry about it. And then the other model that I want to get your brain on is DARPA. So in 1957, Sputnik is launched into the sky. It looks like the Soviets have delivered a absolutely shocking technological surprise. And Dwight Eisenhower, among other things, creates the ARPA, Advanced Research Projects Agency, which is later known as the Defense Advanced Research Projects Agency, or darpa. And DARPA is also legendary in its own way DARPA scientists invent, among other things, the Internet, gps. DARPA has a slightly different model of scientific discovery which is oriented not around scientists begging the government for money, but rather oriented around this position called program managers. Can you talk a little bit about what program managers do do and how their position on the field, so to speak, is a little bit different than the grant funded scientists that we're familiar with with the NIH model.

Derek Thompson (109:24)

Earlier we, we talked about investigator initiation as kind of the cardinal principle of NIH funding. If there is one. Well, ARPA is kind of the reverse. ARPA is like screw investigator initiation. We have a direction, we have a true north. In this particular class of technologies, there's a trajectory that is kind of like embryonic, but it already exists. And that's the thing we want to push on. And we're going to assemble a dream team of people and we're going to really assemble it, we're going to engineer it and we might actually kind of like we might force a national lab to work with an academic, to work with an entrepreneur. And that might be a team, right? And we're going to give them a pot of money and we're going to watch them like a hawk. There's a trajectory, there's a way of quantifying progress. And every six months to every year, we want to see that, you know, we want to see that you're actually pushing in the right direction and there is tangible evidence of progress. And if not, if you're not making progress, we're going to withdraw the funding. That's okay because you know, we're not holding it out again, you know, against you, where we were asking you to do something that was pretty hard anyway, right? This money is going to be given to some other team who's also pursuing an equally kind of seemingly out of reach kind of goal. But it's really important to understand that the ARPA model is very top down and at the top sits the program manager, which is this very rare person with exquisite tastes, with a foot in the business world and a foot in the science world. And so the rate limiting step for the ARPA system is to find those very rare individuals, right? So I think it's wonderful we should have a diversity of models. And so, you know, the idea of an ARPA H ARPA for health is not a crazy idea. There are differences between health and defense. Most importantly, with defense, there's only one customer, right? Whereas you know, anywhere else where it's, whether it's energy or health, transitioning technologies to the market takes, you know, is a challenge unto itself. But, you know, the. The idea of an ARPA H is another way in which we could think about the kind of strengthening the ecosystem of funding for biomedical research. But I don't think that's a way in which we should be strengthening. Nih, in my mind, if it exists, should be outside of nih because NIH is culturally not aligned with everything that makes ARPA different and potentially successful. So the fact that the current ARPA H is actually part of nih, that the director of ARPA H reports to the director of NIH makes me very bearish about the prospects of ARPA H. But I could be wrong. That's my intuition.

Holden Thorpe (113:03)

Yeah. The way I described it in the book is that arpa's a little bit like the Hollywood producer model. You got a Hollywood producer who's like, okay, I have this great idea for a movie. It's called Titanic. It's Romeo and Juliet, but it's on a fated ship. I gotta find a director. I gotta find a screenwriter. You got Leonardo DiCaprio. You gotta pull Kate Winslet into it. And they go do that. They pull together all the talent necessary to make that movie. The same thing has happened over and over again in DARPA history. So, for example, DARPA invented the first Internet. It was called Arpanet, the first prototype of the Internet in 1962. And you had these luminaries like Licklider and Bob Taylor. And they said, oh, you know, I know this guy at Carnegie who can work with us. I got someone at Harvard. I know someone at mit. I know someone at rand, I know someone at ucla. And they basically bring together this team to work on a proto Internet. And through the construction of this offline network of minds, they build the first online network of ideas. It's this beautiful illustration of the principle that program managers can succeed in taking, I love the way you put it, ideas that are already past that pure embryonic level and saying, oh, all these ideas, they work, but they're not speaking to each other. Let's have them speak to each other. Why can't this model work if we scale it out? Why are you pessimistic about situating this program manager species in the nih?

Derek Thompson (114:43)

A sense that. A potentially erroneous sense that what gets you in a position of administrative control at NIH is creating a system where good science and good research proposals emerge from the bottom up, as opposed to kind of crafting the problem areas yourself. Like, I don't see anything in the NIH DNA that would make me predict that they would be particularly good at doing this. Right. It's interesting that, you know, one feature of ARPA that people have often thought is important is that the director of DARPA reports directly to the Secretary of Defense, not to some, not to kind of some head of procurement within the DoD. It's really treated as it's kind of separate thing. And then they are managing very carefully the communication between darpa, which is supposed to go crazy, right, and the rest of defense procurement, which is supposed to not go crazy in some sense.

Holden Thorpe (116:02)

So I'm very interested in developing what I'll call an invention agenda for America in thinking about what are the set of policies and the set of institutions that would lead to the best inventions in medicine to improve people's lives. And I think one interesting challenge of building such an agenda is that invention itself is so mysterious. There is no formula for organizing science in a perfect way to optimize our chance of discovering the most important truth about aging or pancreatic cancer or Alzheimer's. Why is that? Why don't we know more about the perfect formula for producing great science?

Derek Thompson (116:47)

Because by definition, you're venturing into the unknown. Like, by definition, it's unscripted. I mean, I don't know, it sounds corny, I think, but I. I think it's true. It has to be the least scripted, prescribed in advance activity that human beings could engage in. So that doesn't mean though, that we can't gain knowledge about it, but it means that we have to really maintain a stance of great humility as we're doing so. And so we should resist hatchet jobs. Right. I think the metaphor that I would like to use here is that what we need is kind of a determined surgeon, maybe a reconstructive surgeon. You're not a supposed. I have some ideas. I have some tentative priorities that I think could unlock things potentially if pursued with some kind of continuity over 5, 10, 15 years. It's not flash of the pan. And we won't know if things are actually working for a while because it takes a very long time for the discoveries of today to percolate through the entire vertical chain from research to discovery to actual treatment that meaningfully impact people's life. Like that process itself can take 15, 20, 30 years. This is why you don't want hatcher jobs. But also you don't want to do anything that is going to be stopped at the next change of administration, because that by definition will not be the time horizon we need to meaningfully impact things.

Holden Thorpe (118:41)

Where would a Pierre Azulay reform agenda for NIH begin? Like, let's take a policy like peer review. How would you change peer review?

Derek Thompson (118:53)

Everyone paraphrases Churchill. So I'm going to do it right? It's kind of the worst system, except for all the others. But I think one mistake we make is that we treat it as one uniform, unchanging object. And if you think about the way peer review actually functions, it has a lot of attributes. And you could think about changing the makeup of the system as a way of aligning the goal of getting a grant and the goal of doing great science to a greater extent. So, for example, there's a lot of skepticism that the committees add a ton of value, that they are actually good at judging which projects are actually meritorious. And so one proposal that has gained a lot of support, I think, is one of a modified lottery, where what would happen is that, okay, scientists would, you know, propose ideas, there would be kind of a first cut, like where maybe we would eliminate, maybe even using AI, you know, obviously bad proposals. You know, let's say we triage 50% and we're saying, okay, you're not getting money. But then for the proposal that survived that first cut, we just randomize. I'm picking on that idea because it's not my idea, and I'm kind of dubious. I actually think committees add value. I just would like them to add even more value. But the point is, the really important point that I want to make is that it's not an implausible idea, but we should test it. We should actually craft a randomized control trial where we fund proposals either randomly or using the traditional system and see, at the end of the day, what seemed to have worked better in stimulating important ideas. Important programs train, you know, young scientists and build research that other scientists find useful. And to me, that is the meta reform that enables everything else, basically turning the scientific method on ourselves and treating it as something that is susceptible to imperfect perfection over time. And so that is, you know, you would think that scientists would love that, right? What's better than the scientific method?

Holden Thorpe (121:39)

Scientists love to conduct the scientific method, not have the scientific method conducted on them personally.

Derek Thompson (121:45)

That's right. That's right. That's been my experience.

Holden Thorpe (121:49)

Pierre Azoulet, thank you very much.

Derek Thompson (121:52)

Great. Thank you so much. It was fun.

Holden Thorpe (121:55)

Many, many, many thanks to Holden Thorpe, Bhavan Sampat, and Pierre Azoulet. Thanks to you for listening through this entire long show, maybe the longest show that we've ever produced. On plain English. I'll leave you with a quick takeaway. And it's the first takeaway. American science is too important to not be obsessed with making it better. And American science is too important to destroy. My fundamental fear that animated putting this episode together is that I think we're in the process of destroying what deserves to be reformed. Thanks for listening. We'll talk to you next week.

Derek Thompson (122:36)

Sa.