Intro. [Recording date: June 16, 2022.]
Russ Roberts: Today is June 16th, 2022 and my guest is the oncologist, Dr. Vinay Prasad. His first appearance on EconTalk was April of 2020 discussing cancer drugs and his book Malignant. Our topic for today is COVID-19: what we've learned, what we shouldn't have learned, and what we still have to learn. Vinay, welcome back to EconTalk.
Vinay Prasad: It's great to be back, Russ. And, of all the things I've ever done, I think I've gotten more positive emails by going on your show than anything else.
Russ Roberts: Well, when you cure cancer, I'll slip down to number two. But, I appreciate that. I appreciate that very much.
Russ Roberts: I want to start about vaccines. We recently talked to Greg Zuckerman about vaccines here. And, let's be honest: it is an extraordinary human achievement. I've waxed as eloquently as I can about the speed with which we developed the--'we'--with which a handful of people developed the vaccines, and the lives that have been saved.
How effective are they, though? Because, I just got COVID, two weeks ago. I was triple boosted, meaning I've had the original two doses. I took the third dose, I think about nine months ago.
And I was told that if I didn't get a fourth dose, I was a fool. And, by the way--this is one of my favorite cocktail party moments of all time--the person said, obviously I had allowed my political biases to overwhelm my rationality. And, I just said, 'You know, I don't think you know me so well. So, let's just leave that alone.'
But, I'm curious to hear your take on that, both the question of whether I should have gotten a fourth shot--if that's an answerable question. And, just generally, the fact that people who have the vaccine are still getting the disease.
Russ Roberts: What should we learn about this?
Vinay Prasad: Absolutely. So, I guess one more piece of background for your listeners is that, of course, in addition to being an oncologist, I'm actually an epidemiologist and biostatistician. I'm in--Department of Epi and Biostats is my primary appointment. And so, that's the lens through which I view a lot of COVID [Coronavirus Disease] policy--the epidemiology lens.
The answer to your question: I think I totally agree that vaccines were probably one of the greatest breakthroughs throughout the pandemic. The timeline was something that even the optimists weren't bullish enough about. Nobody saw it coming that quickly. Maybe the only person who saw it coming out quickly was Trump and people thought he was out of his mind to say that it was coming so quickly. It's truly a remarkable feat. And, I think there's no doubt about it.
The vaccines lower the risk of severe disease and hospitalization. And, that is very, very clear, particularly among older, frailer people. And that's the group you wanted to protect, because that's the group of people that was most hard hit by COVID-19.
Now like all medical interventions, the reason vaccines get into debates isn't that anyone debates that they have a value in an 85-year-old. It's that as you go down and down to lower and lower risk, and you get to--as we're going to talk about soon--very young children, or as you talk about additional doses, dose after dose, after dose, you get into the business of diminishing returns. And, at some point you start to wonder what returns are you getting at all.
And, the way in medicine we typically sort these out is larger and larger randomized trials. You're looking for smaller and smaller signals. You need larger and larger sample size. Though what we found with the regulatory system around vaccines and boosters--and second boosters--is as we've moved along, we've gotten less and less data.
So, for instance, you know, the regulatory authorization of that fourth dose, which in the United States is only granted for people over the age of 50, is actually a little bit more generous than what Pfizer wanted, which I think was over the age of 65. And, it was the FDA [Food and Drug Administration] gave them a little bit more. And, it was broadly on the basis that we can improve your antibody levels to what they had been a little bit before. And, not necessarily that we're proving that a 51-year-old man, for instance, who is in excellent health--no one in my mind has proven that person has a further reduction in severe disease or hospitalization by getting the fourth dose.
And, that's why I think this has become so debatable. Because, to be honest, there's a lot of uncertainty.
Now, I think most people would say that we believe the third dose definitely benefits older people.
And, the reason we believe that is observational studies that are done in many countries, but particularly the most notable have come from Israel, where people who get that third dose appear to have a further reduction in severe disease in hospitalization than people who don't get that third dose.
But of course, in Israel, there are some confounding factors. One is I think the Arab population has always been slower to get the extra vaccine. And, the people who are getting it are of higher socioeconomic status. And, they may be doing other things in their life that are--yeah.
And so, I think we have to put a grain of salt next to all these things.
And, the last thing I'll say before, [inaudible 00:05:17]--the last thing I'll say is: the places that are the most hotly debated are the places where you really wonder whether or not those extra doses matter.
In the United States many colleges, I think over 300, now mandate a 20-year-old college boy get the third dose. And, there's no exemption for whether or not he's just had COVID. By the way, he had two doses in COVID, he's got to get the third dose or he's going to be expelled from school.
That's an age group where the risks are much lower. The evidence is very weak that the third dose would benefit him. I don't think there's any credible evidence that somebody who had COVID and recovered and getting a booster in the aftermath has a subsequent reduction in severe disease or death from getting COVID again, ostensibly.
And so, I think that--and there's a downside, which is risk of myocarditis, which we can talk about. Which is the one safety signal that has emerged from mRNA [Messenger Ribonucleic Acid] vaccines. And so, that's the place where all the heat is because it's the most uncertain.
Russ Roberts: Let's talk about myocarditis which is--did I say that correctly?
Russ Roberts: That's a side-effect. It's funny: I would have felt uneasy mentioning a side-effect at some point in this journey. Because: Oh, people might misinterpret it and think you shouldn't get the vaccine. And, again, I want to make it clear: I gladly got the vaccine. I gladly got the two-dose vaccine. I got the booster.
I would also suspect that myocarditis is not the only side-effect. So, one, am I correct? Let me just add that when someone said to me, 'Oh, you have to get the fourth booster. What are you an idiot? You're 67 years old. You're at risk.' You know, my view was--and perhaps I'm crazy. I'm playing with my immune system.
And, I don't know if we know a lot about what happens when you keep asking your body to produce this, these antagonists, so these things we're introducing in our body.
And, it just struck me as: Well, the fourth dose would be relatively small in helping me. And, I wasn't comfortable or confident about the downside.
Now I should add that when I got COVID two weeks ago--I'm in pretty good health, I'm a little bit overweight. I don't think, I'm not obese, I think to the human eye. Although I may be to the medical establishment, let me confess. And, I'm 67.
It was really unpleasant. I coughed really badly, really hard for three days. I coughed well, but unpleasantly for three days. My stomach hurt just from the coughing. My nose was just running constantly. I was sneezing. I didn't run a fever. But, I haven't really felt like myself until about yesterday. So, it was two weeks of dragging around. I really wish I hadn't gotten it. And, I was really uneasy. I'd had a meeting with somebody older than myself the day before I exhibited the symptoms. And, I figured I'd given it to her. I don't think I did. Turns out, fortunately[?]. But it was really bad.
So, I would have liked to have avoided that.
So: Do we know anything about whether the fourth booster would've at least reduced that? And, what do you think the risks are?
Vinay Prasad: Okay.
So, I guess the last question first: I think the unfortunate thing is that all of these vaccines that we're talking about, it's the original Wuhan ancestral strain and the spike protein that strain made. And, now we have Omicron, which is to some degree, an immune escape variant. It has a different shape; and they're developing Omicron specific vaccines, which I'll come back to.
But, the challenge with that is, and many studies show, that as time goes out from your last booster dose, whenever that last booster dose was and whatever number it was, four or three, or even the second dose, as time goes out your vaccine-effectiveness--the probability that vaccine is going to protect you against getting COVID-19--goes closer and closer to zero. Some estimates say zero. Some say 12%. But, when you talk about vaccine effectiveness that low, then it's not a question of if I will get COVID-19. It's when will I get COVID-19.
Because it's only a matter of time if it keeps circulating and people have very, very poor vaccine effectiveness, we're all going to get breakthrough infection. And, I think even Anthony Fauci admitted that a few weeks ago.
In terms of the risks I put it in, I mean, two buckets. One thing to talk about is: Is there a risk from getting booster after booster of the original ancestral strain? And, I think largely people don't know. It's a bit theoretical. But, once upon a time, we do know one thing, which is there is something called imprinting or original antigenic sin. And, the basic idea is: many years ago, we knew that the cervical cancer was caused by the human papilloma virus [HPV] of which there are many, many, many types. The original Gardasil vaccine had four serotypes in it--I think 6-11, and then two in the 1930s. It had four different strains of the HPV in it. And we vaccinated a lot of girls with that.
Later, they developed a nonavalent vaccine--nine different strains of the HPV were in that vaccine. And, if you took a girl who had never gotten the four and gave her the nine, she'll have antibodies against all the nine different strains. But, if you took a girl who had gotten the four and gave her the nine, she has antibodies against all of them, but it's a little bit diminished to the new five and strongest to the original four because the immune system is being trained on the original four.
And so, this is the idea of imprinting original antigenic sin. How does that relate to this situation? If indeed, we get to the situation where in the fall we have an Omicron-specific vaccine, or next year we have a vaccine specific to the strain, it might be--I don't know it to be--but it might be theoretically possible that the fact that somebody has gotten more boosters of the original strain makes it a little bit--makes their response stimulate antibodies to the original spike protein epitopes, and not necessarily the new parts that we've changed.
So, that's a potential theoretical downside.
Of course, nobody knows. You can't prove it until it comes.
And then the last bucket, the non-theoretical downside. I mean, I think there are many side-effects of the vaccine, from fatigue, weakness, arm pain. I mean, these are the things. But, I think those are things when people mostly accept.
The one side effect that came sort of unanticipated and has really implications, I think, for safe vaccination. And, I want to stress that. That's my position--is not that we don't or do. It's that we do it as safely as possible.
And that is myocarditis. Myocarditis is inflammation of the muscle of the heart. Often precipitated after a viral illness or something. That's typically how we see myocarditis.
But, it was very clear--and the Israelis were the first to note--that after mRNA vaccination, particularly among boys between the ages of 12 and 26, and now we even see maybe 10- and 11-year-olds and up to 30- men, more than women, after the mRNA vaccine. More after Dose Two, and maybe a little bit more after Moderna than Pfizer, they have rates of myocarditis that are non-trivial. Israel's original estimate was one in 3,000 to one in 5,000 after Dose Two of the Pfizer vaccine.
Now, why does that matter? This is a serious adverse event. It often leads people to come into the hospital. They always say it's mostly mild. Okay.
But, it may not always be mild. There are definitely people who suffer from arrhythmias as a result. And, there are people who have a heart that's not squeezing as good.
And, it mostly recovers. But, again, there are still some people who are suffering long-lasting effects from this.
And, the reason I think it's so important is that it's very difficult for me to look a 20-year-old, healthy, college student in the eye. And, after they've gotten one dose, they come to your clinic and you're within the time period: Is it truly in their best interest to get that second dose right then and there? And, the answer is they're going to take a one in 3,000 risk of this, I think, non-trivial event. And, is the benefit--the further reduction in hospitalization--is that in the one in 3,000 ballpark, or is that in the one in 30,000 or one in 300,000, ballpark? Because, they're so young and healthy anyway.
And, I think that's why it's kind of a--that particular side-effect has a lot of implications.
A number of countries--Denmark has banned Moderna in people under 40. Kaiser Permanente in Northwest to here recently said, 'We prefer Pfizer over Moderna.' Other countries have spaced the doses apart.
And, all these things are done in efforts to lower the myocarditis risk while taking the benefit of vaccination.
But, then to come back to your situation, I think, had you gotten the fourth dose, I'm not sure you would have avoided COVID-19. You might've kicked the can down the road another month. I'm sorry that you had to go through it. But, I do worry that eventually we'll all have to go through it.
Russ Roberts:So, on the myocarditis, defenders of the vaccine for wider populations have argued that the risk of myocarditis from COVID is higher than the risk from the vaccines. So, the vaccine, even if it has a risk of myocarditis, is still better than getting it. Do you disagree with that?
Vinay Prasad: I guess I'd say--I've heard them say that. And, I would say that if you have a choice in this life, that it's better to get vaccinated and then get COVID, than get COVID before ever getting vaccinated. That's true as well. So, I'm a supporter of vaccination.
But, I think--here's what they have a problem with some of their analyses. Some of these studies are looking at the risk of myocarditis after vaccination, but they're including 85-year-old women with 20-year old-men. But the risk is hugely different between those two groups. So, if you stratify by age and sex, and dose, and product, you start to see different signals.
And, then the next thing I would say is that everybody, after you get every dose, there's a fork in the choice. I get one dose and I have a choice to get the second dose.
And, I'll admit personally, I rushed to get Dose One. I rushed to get it. I was a healthcare worker; I rushed. Dose Two, I was a little bit more ambivalent about, I wasn't sure that I needed it on that time schedule. I delayed it a little bit. And, Dose Three, I thought: I'm a 39-year-old person, no medical problems. I thought that--I wasn't convinced it was in my best interest to do. But, I did it because of course I'm victim of a mandate--which we can talk about.
But, back to your point. Is myocarditis--I think there is no doubt in my mind that for a 22-year-old man who has already gotten one dose of mRNA [messenger RNA, messenger ribonucleic acid]. And they're coming into the office for Dose Two--and let's say it's Moderna. So, it's the day-28 Dose Two.
The risk of myocarditis is higher if they get the Moderna; and then they will eventually get a breakthrough. Versus if they just go along and get their breakthrough. The risk of myocarditis is clearly higher. There's a number of papers that would prove that. Nature Medicine paper I've written some post about that.
So, I think it's not always true. And, I think that's important because that can allow you to tailor your vaccine policy.
Russ Roberts: What do you mean they get a breakthrough? A breakthrough of what?
Vinay Prasad: I guess I would say that I use the word 'breakthrough' to mean that anybody who has had any vaccine who nevertheless gets COVID-19, I call that a breakthrough infection. That's the term I would use.
Russ Roberts: I'm going to recommend a different phrase. A breakthrough sounds good.
Vinay Prasad: Yes. I see.
Russ Roberts: This is not good.
Vinay Prasad: It's not good.
Russ Roberts: Well, you can think about that after we're done.
Vinay Prasad: You're right.
Russ Roberts:And so, a lot of people argue, I think, from the data, that the immunity I now have as a sufferer from COVID-19 is better than if I'd gotten a vaccine. Is that correct?
Vinay Prasad: Well, I guess I would say that there's a bit of a survivor bias. Because, of course, if you imagine that a lot of people got COVID-19, some of them are not alive at the end of it. And so, among the survivors, their health is--you've weeded out the sickest people. So, I just want to point that out. That's often biased. But--
Russ Roberts: Very important.
Vinay Prasad: Yeah. But, there is a good study. It appears in MMWR [Morbidity and Mortality Weekly Report]. It looks at New York State and California State. And, what they do is they say: After vaccination, what's the risk of severe disease or hospitalization from a subsequent COVID-19 episode? And, after you've had and recovered from it. And, after you've had and recovered from it, and gotten vaccinated. And, I think it shows mostly that if you've had and recovered from it, going into the future, the risk was so abysmally low, the curve is, like, hugging the x-axis. And, after you've been vaccinated, it's still low--much lower than it would be if you're unvaccinated--but it's not quite as low as natural immunity.
Now, of course, there is that survivor bias there, too.
But, I guess I would say that I cannot tell you that you won't get COVID again. You might. In fact, I'm sure with enough time you will get COVID again. But, what I suspect is: you won't get it as bad. It will be milder and milder every time you get it. And, I suspect that your risk of getting severe disease or hospitalization from COVID now that you've had and recovered from it is very, very low. So, low that one wouldn't change one's life.
Russ Roberts:Let's talk about getting it again. In the early days, when it was much more severe, the Wuhan--what's the right language--
Vinay Prasad: Strain. Yeah. The Wuhan strain.
Russ Roberts: The Wuhan strain. It was a frightening time. I was afraid to go to the grocery. I was wiping down cans. It was--I was afraid to get my mail. People are still doing some of that by the way. It's fascinating. When I was recovered from COVID, but only, like, a week and a bit out, I gave someone a book and he didn't want to touch it. So, if you want to comment on that behavior, I'd be interested. I told him, I thought it was probably okay, given that it would've been over a week. And, I didn't think the surface thing mattered. But, you know, who am I? I don't even play a doctor on TV.
But, the interesting thing to me is that in the early days we all thought, 'Well if you get it, it can be horrible.'
A number of people got it, had a bad experience. But, never had to go to the hospital, never went on a ventilator. And, then we all said, 'Well, lucky you're now immune,' because that's the way we thought about things.
And, then there was this rumor--or, occasional possibility--that someone had gotten it again. And, at that time we all--most people responded to that-- thoughtful people were, like, 'Well, there are false positives. Maybe you didn't have it the first time. Maybe you didn't really have it the second time.'
Now we're suddenly in a world where, 'Oh yeah, you're going to get it again.' What changed there? Did we learn something that we didn't know before? What is going on?
Vinay Prasad: Well, this will also come under the section heading of, like, you know: things public health distorted, or sort of didn't say perfectly accurately, that have plagued us over time.
But, I would say that, you know, before SARS-CoV-2 entered the scene we had four circulating coronaviruses. And, people expected and knew and was-well documented that you could get it episodically every few years. And, that's why they would keep circulating. That: immunity against coronavirus might be good in terms of coronavirus, very rarely made someone very sick. And, often when it did, they were older and frailer and in the nursing home. I'm talking about regular, old coronavirus.
But, it was known that they would be inevitably immune-escape, and we'd get colds. The common cold is a mix of viruses--rhino virus, coronavirus, adenovirus--and it's well-accepted that we would just get it sort of cyclically.
And, I think most people who thought about it realized that SARS-CoV-2 would probably eventually fit the same pattern. And a number of Europeans wrote about this. And, I think they've always been more comfortable with that idea.
I think some of the early cases that we talk about, though, there is some ambiguity. Because, you always want to separate: 'Did you really get COVID--completely cleared it--and then you got it again? Or, did you get COVID and it just never really cleared?'
And of course, there's a fraction of people, particularly people who are immunosuppressed, that it never really clears. And, they may have had long-standing, just sort of smoldering SARS-CoV-2 for a long time. And, some believe that is the basis by which the variants have evolved--like Omicron: where did it come from? It came from somebody who was chronically immunosuppressed and just allowed it to replicate in them over and over again.
So, but why did we say that? I think two things are both true: That, One, that we will get COVID-19 again; But, we do know that, and we now have the evidence that if you get it a second time, the likelihood that you're going to get sicker than you got the first time is very low. I mean, I think it's far less than, I think if I recall correctly, I want to say three tenths of 1%. But, it's very, very low. And so, it's an unlikely event. And, in that case, I think that is a little bit reassuring.
Russ Roberts: If that's true, what was all that conversation in the early days of the pandemic about herd immunity? Because what you're really saying is that there's no such thing. It's a phantom, a fantasy--that, if you can only get, literally only get it once, and then a bunch of people either have gotten it or have been vaccinated against it--which again was thought, I think by many of us early days, 'Well, I'm protected. Nothing to worry about now.' If it's true that you can obviously get it again, what's with this whole herd immunity thing? And what was--the countries that based their policies on a strategy like that? And, at various times I think that would include the United States. Certainly it was discussed as Sweden's policy. Or am I wrong? I don't know.
Vinay Prasad: No, no, you're right.
Russ Roberts: What do you think?
Vinay Prasad: I mean, I guess I would say that herd immunity really applies for a situation where you have a virus that after you've had and recovered from it or been vaccinated against it, you have a long-term, sterilizing immunity and you're unlikely to get it again. And, then there's a mathematical equation, one minus one over R-nought [R0], or the rate of spread that gives you the herd immunity fraction. And, of course, this is also something that Fauci famously stepped in because he was found to be kind of cooking the herd immunity number to suit his policy preferences and not what he actually thought.
Now why did people fall in love with the idea that once we get to this magical number, the virus will just go away? I think to some degree it might have been wishful thinking. And, maybe to some degree, I also succumbed to that wishful thinking: that maybe this vaccine is so good--and the original reports were that it was stunningly good.
And, we didn't have Omicron. We didn't see that immune-escaped variant right there. And so, I think many people thought that maybe we'll be in that situation.
The other thing that changed is the R-coefficient, or Rate of Spread. Omicron is explosive. I mean--and once R goes very high in that number, herd immunity goes over 100%. So, it's a moot point anyway: we're all going to get it.
But, I think it goes back to the models, the original Imperial College London model, which didn't quite come as promised. It was less than that. And, a number of the fearful models that led to sort of the global shutdown. Which is that all of these things are imperfect, rudimentary sort of surrogates for how the world should be. And, herd immunity and that concept, that equation, all the modeling, it's all very rudimentary. And so, I guess we shouldn't be surprised that it didn't actually hold up. But, I don't know if it was malice as much as wishful thinking.
Russ Roberts: We'll come back to Fauci in minute. I might try to defend him--won't be easy for me, but, I'm going to give it a shot.
But, I want to react to something you just said it was more body language and maybe some eyebrow work that may be not visible to those listening on audio only. You made what I would call a semi-disparaging allusion to lockdown and various responses that we had. And, I hope we'll talk a little bit about school closings and other things--we'll get to children in a minute, in particular with respect to the vaccine. I want to finish up the vaccine.
But, I want to ask the following question before we get there. In the early days of the pandemic, there was some experimentation that took place at the national level. And, some of that experimentation was due to various policy makers having more authority. Some of it was due to cultural differences across countries. Some of it was due to who knows what.
But, would you disagree with the statement that the countries that pursued the most aggressive lockdowns had the most success in preventing death?
So, my impression--I have mixed feelings about the lockdowns and that phrase, I don't even like that phrase actually, because it covers a wide range of both official policy and actual response. Because some places locked down and they really locked down. And, other places locked down and it was, like, kind of.
But, for the places that really locked down--I have a son who lives in Australia. Australia locked down. You couldn't get in the country. You couldn't get out of the country. When you got there, you had to spend two weeks in a hotel. And so on, and so on. But, they've had very low death rates. And, they may have paid a price for the lockdown in terms of economic liability, all kinds of human flourishing, factors that are relevant; but weren't they at least successful in keeping their deaths down?
Vinay Prasad: My honest view on this is that I think this will be one of the great economics questions for the next quarter-century, because there's a lot of things.
And, I'll just give you a few factors that I think about as to why it's such a hard question to unpack.
I think one thing to be said is that pre- this pandemic, the idea that lockdowns would even be in the toolbox was antithetical to public health. I think that's clear from writings of D.A. Henderson [Donald A. Henderson] and others. George W. Bush commissioned a commission to look into pandemic response, and nobody felt this was even in the toolbox.
It was only when people saw what China did in Wuhan that we even got the inkling that we could do this--because I think people worried.
And, then I think the other thing that allowed this particular policy response was the rise of technology companies. If it weren't for Netflix at home, and Amazon Prime, and Uber Eats delivering your food, and Zoom, so we could still--white collar workers like us--could still engage in our livelihood, I think you wouldn't have had this policy response. Like, people like us working at Twitter and Facebook, if we were being laid off in droves, I think we would find some way to come back to work and stagger our shifts and wear masks or whatever we want to do.
But, to your specific question, how does one know if a lockdown helped or hurt a country? You can look at these aggregate totals, but I think there's a few factors to consider.
One: Seed load. Seed load is something I don't hear people talk much about. But, seed load means that when the entire world panicked in the first two weeks of March, there were already so many cases on the ground. And, that number is probably very different in economic hubs like Western Europe and in the United States than a nation like Australia. I don't know what the seed load is. I don't know what the seed load per capita is. And, I don't know the interaction between seed load and population density. Yes.
Russ Roberts: And, by seed load, you mean what I'm shedding as a carrier of COVID? What do you mean by seed load?
Vinay Prasad: I mean, how many people out of a hundred thousand people in Australia and March 10th had COVID? And how many people out of a hundred thousand people in the United States had COVID? Because that's the original seed that led to the propagation, even if we put walls between all the countries.
The other thing about COVID-19 is it is a very chaotic, stochastic spread. It has a sort of a right-tailed distribution: a few events lead to a lot of people getting it rather than lots of events leading to few people getting it. And, that creates a lot of randomness in this situation.
And so, I guess coming back to the point: One can imagine that with a very low seed load, a lockdown can actually take a small seed load and extinguish it. But, once a seed load gets above a certain threshold, no amount of lockdown, even welding the doors shut in Wuhan, that kind of lockdown, Shanghai lockdown, even that cannot contain spread if it gets above a certain threshold.
I don't know the answer to this question. We don't really have broad-scale surveillance data from that time, but I think it'll be something for economic--economists and others to mull over. The natural geography of a place, the ability to control your borders, the ability to screen people on entry, very different from a country that abuts other countries, and a country that's an island nation. I think maybe even preexisting differences in immune susceptibility, perhaps. Maybe even differences in BMI [Body mass index], age, age structure of a population may have led to different COVID-19 outcomes, particularly age structures that are younger, where a lot of young kids will get it early. I'm just tossing out factors that might be implicated.
But, I guess everyone has a story to tell. We pick two countries and we say, 'Why did Denmark do this?' right? But, as an economist, the holy grail will be a model that you construct that explains data across many diverse settings and actually does it. And, I think it's a 20-year project. But, I look forward to reading those papers. I think they'll be terrific.
Russ Roberts: I'm not. I hope I'm alive to read them. And, the reason I'm not, the reason I'm not is that there'll be a variable called 'lockdown'. And, it gets back to my point earlier, which is--
Vinay Prasad: Yeah. Right. No, it should be--
Russ Roberts: how people actually responded in practice. And, I'll give you just as an example or at least a furthering of this question: After seven days of the virus, I still tested positive. And, I felt a lot better. I didn't feel a hundred percent. I felt a lot better. And, I think I was following somebody's guidelines--I felt it was okay to go out, go outside. It was okay to go inside if I wore a mask. And, I didn't test again. Not so much because I wanted to--'Well, what if it's still positive?' Just like, 'Eh, after this length of time, I know people can test positive for a long, long time.' I was probably shedding a very, very small amount of virus. I probably shouldn't get in somebody's face and talk to them, whispering for 10 minutes. But, other than that, I thought it was okay to go outside.
Many people made fun of me, by the way, here in Israel. Because they said, 'Oh, everybody has it. I mean, what are you talking about? Good[?] to go.' [?] after five days is you go outside. I was uneasy doing that. Maybe I was overly cautious, but I felt I should not spread it further. Was I being silly?
Vinay Prasad: No, but I think you were being a good person. I mean, that's the kind of person you are. Nobody wants to knowingly contribute to somebody else getting sick. I think we never wanted to do that pre-COVID. We don't want to do that now. But, I think in terms of the scientific question, which is by testing positive--and I suspect you used a home antigen test--
Russ Roberts: Yes, I did--
Vinay Prasad: By testing positive on a home antigen test, are you really infectious? And, how infectious are you? And, how infectious are you compared to the person who in that pre-syndromic phase don't even know that they have COVID? And, the answer is it's a very difficult question.
There are some studies that compare home antigen tests to like, can you culture the virus? And, they use that as a surrogate. But, these are all surrogates of the thing you care about, which is: Do you, Russ Roberts, actually infect someone else?
And, to my knowledge, I don't think--I can't think of any, and listeners can tell me if I'm wrong--I don't know of any study that has taken PCR [polymerase chain reaction] levels, antigen levels, or even any test you want and correlated against the gold standard of: Does someone infect someone else in the real world?
And then one last thing about lockdowns, just to come to your point. I absolutely agree. Lockdowns is not a monolith. It's many things. And so, I could imagine that if you're developing this sort of economics model, what you want is one variable should be 'voluntary behavioral change' using cellphone tracking data.
Because that means without any policy, the moment I see the cases going up on the news, I'm staying in my house. I'm going to go to the grocery store less often. Voluntary behavioral change, that's not a government policy.
Then, on top of that, business closure, school closure, stay-at-home order, shelter-in-place order. And, then you could even add, for some other countries, the more aggressive metrics, like different types of restrictions on that--masking orders.
And, then the last thing I would say is one should, I mean, each of these things should be queried and to see if it has an impact.
Russ Roberts: Yeah, we'll get to masks, I hope. So, but that's a perfect example--mask mandates. For the last few months--I often take a bus to work. And, for quite a while you were required to wear a mask on a bus in Israel. And, almost everyone wore them, but a lot of them were wearing them under their chin. And, that's not in the data set, usually.
I just want to say, by the way, that when I ask people who have, or who've gotten COVID, 'Do you know where you got it?' almost always say, no. I got it from my wife, and heroically took the brunt of it. She had a much milder case. Now of course, it's not really true that I got it from my wife. That's not a fact. I did get it the day after she got it.
Vinay Prasad: But, you both may have gotten it the same place. Yeah.
Russ Roberts: Exactly. So, it's just kind of a--yeah, kind of a--anyway.
Russ Roberts: Let's talk about children. Young children: It's recently been approved, I think recently been approved that the children under the age of five are now eligible--eligible? Is that the right word for the vaccine? Encouraged to get the vaccine? What's the right language?
Vinay Prasad: I guess five and up have emergency use authorization. And, under five, the FDA [Food and Drug Administration] voted yesterday to give it. But, they haven't yet given the EUA [Emergency Use Authorization]. But, it's imminent. It's coming. So, they will be allowed, permitted to receive vaccine under emergency use authorization and not vaccine approval.
Russ Roberts: Okay. Yeah. Which is--
Vinay Prasad: Maybe that's not even more clarifying.
Russ Roberts: Well, I hate that Emergency Use Authorization. It's not approved--
Vinay Prasad: It's not approved. But, it's authorized--
Russ Roberts: Yeah. It's authorized. Well, it's not really funny, unfortunately.
So, I've noticed on Twitter that there's something about that this Emergency Use Authorization is based on 10 cases. Is this true? Ten? What does that mean? What happened there? What's going on there?
Vinay Prasad: Well, that number comes from the number of cases that occurred after the third dose of the Pfizer BioNTech vaccine in kids under the age of five. And, it is a 10-case difference. And, then they're saying, that's why we have, quote, "80% efficacy." I think that's a very problematic statement. Yeah.
Russ Roberts: Whoa, whoa, whoa. Let's slow down. Let's let's unpack that. Because, first of all, I thought it was a trial based on 10 kids. That's the way it comes across on Twitter.
Vinay Prasad: Oh, no, no, no. I see. 10 events. Yeah.
Russ Roberts: So, let's slow down. Describe it again.
Vinay Prasad: Okay. Maybe I'll back up a little bit and give some broader, and then I'll tell you why they get that number from.
The broader context is basically: After the adult vaccine trials of 18-and-up, drug regulators had a choice in terms of what kind of evidence they need to move down to younger ages. And, of course, either below 16 or below 18, depending on the particular vaccine. Moderna was below 18 and Pfizer was below 16.
And, the choice they went with was Immuno-bridging. And, what does Immuno-bridging mean? Immuno-bridging means is that we're no longer going to ask the manufacturer to prove that you reduce symptomatic SARS-CoV-2.
So, the initial randomized trial of the Pfizer vaccine and the Moderna vaccine, the primary endpoint was symptomatic, laboratory,confirmed COVID-19 infection. You had to have symptoms. You didn't have to have severe symptoms. You could have just had a sore throat and a cough. But, you had to have a positive test. And, that was the primary endpoint.
Russ Roberts: You had to have a positive test, meaning to be counted as a person who had COVID.
Russ Roberts: So, there was a group of people who were the control, a group of people that were given the vaccine. The people in the control, some of them got COVID-19; and you're saying that definition of 'getting it' was symptoms.
Russ Roberts: Plus a test. And, similarly, the people who were given the vaccine, some of them of course got sick anyway, but most of them did not. And, that was wonderful. And everybody said, 'Great, hooray.' The number that was banded about was 92 or 95%.
Vinay Prasad: Right. Absolutely.
Russ Roberts: Which is very confusing because it's not 95% of the people didn't get it. It's a complicated calculation. We don't have to do that right now. But, that was the original way that the vaccine was evaluated.
Vinay Prasad: Absolutely. It was a 90 to 95% relative risk reduction. They were 95% less likely to get it. But, the raw number of people who got it was I think about 1%-and-change. A very low number of people actually got it. And, the other arm was 95% reduction from that number. And, that was in older people. That was, I think--I found it very credible.
There also was a signal in that trial--the trial was very large: 40,000-person randomized trial--that not only are you preventing the infection, you're preventing severe disease and hospitalization. There was a difference in severe disease and hospitalization. And, I believe it was statistically significant, although it was not the primary endpoint of the study and the study wasn't designed for it; but they happened to find it. And, that's what made people like me so reassured that this was going to be a great thing. And that's why we popped the champagne.
When they moved to younger and younger ages, I think they had a choice. The choice was: Are we going to make the manufacturer prove this over and over again that they reduce symptomatic disease and/or even hospitalization? And, there are some purists in evidence-based medicine--and I put myself in this camp--that felt like they really needed to make the manufacturer prove that. And, the reason being as you drift down in ages, the risk to the kids is even lower and lower. And, the potential upside is--
Russ Roberts: Both. The risk of--
Russ Roberts: Having a severe reaction to COVID is-- or getting it--is lower and lower?
Vinay Prasad: In fact, it's so low that a 4-year-old has a lower risk than a triple-vaccinated, 40-year-old, for instance, you know--in terms of a severe outcome from COVID-19. An unvaccinated 4-year-old is lower than a triple-vaccinated 40-year-old. Just to put it in perspective.
So, it's very, very low. And, to do that study, to prove a reduction in severe disease or even symptomatic disease, you need large randomized control trials, maybe hundreds of thousands of people.
Now, many people believe that those are financially not feasible and practically not possible. But, I want to remind people that Jonas Saul himself did a 400,000-person randomized trial of the polio vaccine many decades ago. I also want to remind people that Pfizer has been well-compensated for their vaccine. Their revenues this year will be $100 billion dollars plus. And, maybe $50 billion is on the backs of Paxlovid and the vaccine. So, we're giving these manufacturers a lot of money.
And then, the last thing I want to say is the appetites to enroll your child on this study, I think is very high. Even though vaccine rates in children in the United States are about one in three, that one in three is very motivated. And so, I think that we see vaccine rates in children plateau at about 33% or something like that in 5-to-11. Go ahead.
Russ Roberts: You're saying vaccine rates of the COVID--the COVID vaccine?
Vinay Prasad: Yes. 66% of parents are not doing it, so far.
Russ Roberts: Okay. When you say enthusiasm--appetite--there's still a large group that thinks it's a good idea and they're doing it. And so, we could use that data, you're saying?
Vinay Prasad: Yes. If you did a randomized trial and made it easy to enroll, you'll get 1 million people in one week. Because there's 1 million people who really want their kid to have a shot at it.
And so, I think all of the usual barriers to randomization are not present. You could have done the big studies.
Russ Roberts: Well, except that it's not going to be a random sample of the adults of America with kids under the age of five--
Vinay Prasad: That's absolutely true. It'll be associated--
Russ Roberts: because [inaudible 00:40:00] by income.
Russ Roberts: Your earlier point.
Vinay Prasad: Correct. Between placebo arm and vaccine arm, it'll be randomized. But, you're right, it's a select subgroup. But, to be fair, I think that also applies to the original COVID trials. Yeah.
Russ Roberts: Sure. Absolutely.
Vinay Prasad: But, the FDA decided to go a different way. They went with what they call Immuno-bridging. And, Immuno-bridging basically means: Prove to me you have the same antibody levels in these younger ages as older ages had. And, that's the primary endpoint of the study. If you happen to also have a detected difference in symptomatic disease, that's great. But, we're not asking you to prove that.
And, that comes to this youngest trial that you're talking about. In the Pfizer study, when you randomized kids, they initially planned two doses; and they had two doses to show they had the sufficient antibody level. Two doses of the Pfizer vaccine in kids under the age of five failed to generate that antibody level, so they added a third dose in a protocol modification.
And then in terms of the actual rates of symptomatic SARS-CoV-2--and you could put a link to it, maybe I'll send you the figure--there's a nice, cumulative-incidence figure that just shows almost overlapping curves. There really didn't appear to be much of a difference. For the other product, maybe we're talking about a 20 to 30% reduction in symptomatic disease.
And, the study has no power for severe disease. I mean, there's so few cases, you can count on one or two hands how many cases there were. There's just no ability to see any signal for severe disease.
So, putting this all together, this is a vaccine to the original Wuhan strain. It generates antibodies to the original Wuhan strain. I'm not very confident there's much of a difference in symptomatic SARS-CoV-2 if you give it to a kid. I have no confidence that you lower severe disease. I mean, there's just no signal.
And then the last bit of nuance is: Since these vaccine trials have been going on, kids have been getting infected with COVID; and in the United States, the CDC [Centers for Disease Control] says at least 75% of kids in this age group have already gotten COVID. And, it's very difficult to take a kid who's already at low risk, who already had COVID, and already did well, and give them a vaccine that will even lower that risk even further. I think that's very difficult. And so, that's why I think it's a very controversial space.
Russ Roberts: Well, we don't have--well, let me ask a first question. There's a bit of a debate breaking out in Twitter recently about the risk to young people. Let's not say young people--children. Young children. When I mean young children, I mean five and under.
So, in the early days of COVID, the argument was flu is more dangerous than COVID for young people. And, for old people it's--they're both unpleasant and dangerous, flu and COVID--but COVID's much worse. And, there's a big debate about whether we should have focused all the efforts--not a big debate, unfortunately: there's some debate--whether we should have focused all our efforts on the 80-and-older, or the frailest and weakest and immuno-challenged folks. But, this question about young people I thought was sort of settled, the young children--there's no--it just has such a low risk. And, there've been a couple studies published recently that said, 'No, no, no. Actually it's really, it's dramatically higher than flu and they should be protected also.' Is that correct? Do you think it's correct? Do we know?
Vinay Prasad: I mean, I think I know the blog post that got all that coverage and I believe it's an incorrect blog post in a number of ways.
So, I think listeners of your podcast can go pull up a chart from the Financial Times. And, this Financial Times chart shows the infection fatality rate by age. And, it also shows--in two times: March 2020 and then March 2022. And, in both time periods, and particularly in the new time period with the circulating--and this is from the United Kingdom--from the circulating strains: The risk of a child who gets COVID-19 from getting it at those ages is much lower than it is had they gotten flu.
So, nobody wants to say this. But, I mean, if your 4-year-old had to get either flu or COVID, it's better for them to get COVID than the flu.
Okay. So, I mean, I think that's one fact.
The reason this statistic exists, that somebody said, 'COVID is actually worse than flu,' is that in the Omicron winter wave in this country, many, many children got COVID-19. And, some of those children ended up being hospitalized. And, if you look at the total number of kids who became ill with COVID-19, it's a lot. I mean, in one short period of time, a lot of kids got it.
And so, if you look at the aggregate number of, say, hospitalizations or deaths, it might look bad.
But, there's two problems. One is: We still do a very bad job of differentiating somebody who comes in the hospital with COVID or from COVID. Hospitals have universal COVID-swabbing on entry. Everybody getting into a hospital is being swabbed.
So, if you come in and you've been in a car accident and you get swabbed, I mean, that may inadvertently be coded as a COVID death or even a COVID hospitalization. If you break your arm in[?] a test positive for COVID. And, there is not to my knowledge, really a well done study that really pulls these two apart. Even today, we still struggle with that.
We never looked for flu the way we look for COVID. We didn't screen everyone for flu on entry who are asymptomatic. That was never a practice.
And so, we're looking much harder and we're finding it much more often.
And then the other thing I think that's missing is that parents don't necessarily care about what the risk was to my kid in December. Parents are caring about what the risk is to my kid this December.
And, I think that if that's the case, then one should acknowledge that the fact that maybe even 75 or 80% of kids have had and recovered from COVID means the risk to your kid this December from COVID is even lower. And, flu is actually probably much more salient a risk, because flu can actually reinfect people much more severely than COVID.
Russ Roberts: And, let's be clear. This is not the Olympics. We don't care about which is really riskier.
Russ Roberts: And, the other thing I want to emphasize--and I think this is really hard for people to remember--it could be that the risk of COVID is twice as high as the risk of flu for death. But, it's still very low.
And, sometimes that doesn't matter, but sometimes it does. Sometimes if it's a very unlikely event, you don't want to do something that has a high cost.
I mean, the thing that--just to take, skip ahead a little bit and what I'm hoping we'll talk about--the relentless masking of small children in social settings and creating fear in children of becoming close to, physically close to their peers, is, I think, a bad thing.
I know it can't be quantified. I don't think it--itmight not be measurable in any way. I don't think we're going to know whether kids can recover from that emotionally or whether it's just an unimportant game. It's no different than going out on Halloween. We'll never know.
But, my intuition is that human beings were designed to talk to each other, through a--and to see each other's faces and our eyebrows and smiles. And, they're not unimportant.
So, we could debate that. But, depriving a child of that level of human interaction--to reduce a risk that's tiny, tiny, tiny--is something an economist would bring up. And, many people would respond by saying, 'But every single death, even one, is a tragedy.' And, my view is that it certainly is a tragedy, but millions of children being emotionally stunted is also a tragedy.
And, I don't have an easy way to compare those two. I would never say, 'I know how many emotionally stunted kids equals how many deaths.'
But, I think we should be aware of both of those effects. And, I understand that the death is measurable--sort of, it kind of isn't for the reasons you just talked about, which is a whole 'nother, horrible part of this public health issue here. Which is that the death counts are slightly disturbing. Having said that we have some measures of, we do have pretty good measures, I think of unusual death rates overall--
Vinay Prasad: Excess death--
Russ Roberts: Whether it's all due to COVID--excess death. Some of it could be due to other consequences or policy responses to COVID. But, it's not irrelevant. It does bring one--it is thought-provoking.
But, my point is, is that I just don't think we should ignore these other costs. And, I'm not suggesting have an easy way to take them into account, but I think they're relevant.
Vinay Prasad: Russ, I absolutely agree with you.
I mean, just one quick point about Excess Death. I think obviously Excess Death is observed, minus Expected Death. That's the calculation. And, I think there's a lot more nuance to Expected Deaths than meets the eye.
One is that as populations keep going year after year, there's constant aging, and so expected death should always be age-adjusted for the population, ticking up a little bit older. And, that's not always done in some of these studies. Anyway, but I'll put that aside. But, back to the kids.
Russ Roberts: That's important.
Vinay Prasad: It's important. Yeah.
To the kids: I guess it's not just my--I mean, my intuition also, everything I've read or written about this topic--I mean, I think that almost, the majority of our response in shielding children from this virus has been irrational and not proportionate to the risk they face.
You know, when we were kids, our parents would happily put us in the back seat of a car--and we lived in Cleveland, Ohio--and drive to Florida to visit relatives. Now that's a long drive. There's a risk to the child to being taken on that long automobile ride. It's a low risk. But, it was a risk they felt was worth it so we could visit loved ones and relatives.
In the New York Times, they have an article in the last week where it talks about some of the most extreme parents and what they've done in response. Because they are legitimately worried about their own children. And, I don't blame them.
But, what they've done is, they talk about three-year-olds who have never been on a play date. Two-year-olds who say, 'Is he real mommy?' when they see another child. Kids who have never seen the lower part of the face of another child. Kids who don't go on, have never been to a Thanksgiving, never had a birthday party.
This is, I think--history will view this as very disproportionate and cruel.
School closure, I think, is the most abysmal policy choice because as you and your listeners will well know is that school is a great engine for upward mobility and for wealth, which is a proxy for health, and living longer.
And, I think we will see declines in life expectancy as a result of the impact of school closure.
And, then I finally, I share your sentiment that it is--the American Academy of Pediatrics, I think to their discredit, they had a tweet that said, 'No evidence shows that a baby not seeing the lower part of a face is detrimental.'
And, my response to that was, I think that's technically true. No evidence shows that. But, no society has ever done that to children for so long.
And so, one would not expect to have a nice historical group to go to, because people would have thought it absurd. We can talk about our masking policy on kids. I think we deviate from World Health Organization here in the States, and it's taken sort of a religious fervor beyond what I think the evidence supports.
Russ Roberts: Let's turn to masks next. But, how old are you?
Russ Roberts: Yes. I thought you said, we said, I think you said something about 40.
When we took that car trip, my sister and I, and my brother--and I was born in 1954. So, we're talking 1962. My brother hadn't been born yet. But a typical car trip was, we made a big car trip from Moses[?], like Washington to Memphis, Tennessee, to Lexington, Massachusetts. A big V. And, my sister and I fought--I was eight and she was five or so--and or something like that. And, we would fight constantly over who would get to lay down in the shelf of the back window during the ride, because not only did we not have seat belts, but we didn't even sit in the seat.
So, when you said, 'Oh, you put the kid in the back seat,' you're talking about the fact that car travel is, on average, more dangerous than airline travel. And it's not safe, a hundred percent safe. It's dangerous, as most things worth doing in life are.
Russ Roberts: Let's talk about masks. I see numerous smart people wearing cloth masks, or--I don't know if they're cloth. They're more like paper. They're not the--so, let's talk about three different kind of masks.
The first thing, by the way, is that in the early days there was that study--I think that came out at Duke University--that showed this crude, non-scientific, really bad study. And, the gaiter where--
Vinay Prasad: That was worse--
Russ Roberts: The person who wore a thing that went down to their neck did worse.
And so, my synagogue banned those gaiters.
And, I explained to the--and then someone would, would show up with one, not knowing it was banned. And, they were treated as sort of a renegade poisoner. And, I would reassure them that the evidence that their gaiter was not effective was really quite small. Bad evidence. There was no evidence for it. So, it was a stupid. It was a really bad study.
But, let's talk about three types of masks. There are cloth masks: there are like that gaiter, the thing that people might wear over their face, nose, or mouth--their nose or mouth. There are the N95s, which are kind of a hard shell plasticky kind of thing. They're less pleasant to wear.
And, there are also the N95 knockoffs, by the way. Which I wonder about.
And, then there's this thing that comes in a box. The mask is square, is rectangular. It has a bunch of layers. It looks like something you would see in a medical ward. So, I see a lot of people wearing them as if they are in armor. And, I do--I don't make fun of them. But, I often observe that I worry that it's a form of theater. And, I'm curious what you think we know about these different kind of masks.
You could add in, by the way about the canned washing[?], and the Washington cans and mail and other surfaces.
Vinay Prasad: You know, it's a good question. And listeners who are interested in this section can find myself and colleagues, we published on the Cato Institute website, a, like, 80-page review of mask--community masking.
I guess I want to draw just a couple of distinctions here, which is, one, the N95 mask. When we talk about N95 mask, what it means is literally it's filtering particles of a certain size and 95% of those particles. And, as a healthcare worker, I've worn an N95 mask often--when I go and visit a patient with tuberculosis and now a patient with COVID-19. And, as a healthcare worker, you have to be fitted. Which means: you put the mask on and go into a booth and they put a sort of an odorous chemical. And, you have to say if you can smell it or not. And of course, you can smell it, it means it's leaking around the mask, and that's not good. And so, it is--
Russ Roberts: You don't just like pinch that little metal part at the top? Really?
Vinay Prasad: Yeah. More than pinching the metal.
Or even worse, I see people wearing, as you point out, the knockoff KF94, or N95s, with huge gaps by the nose. Or, you know, the air is just traveling by the path of least resistance.
So, I guess, one thing I want to say is that I'm very confident that if I go in the patient's room with an N95, the patient could have active tuberculosis or COVID and filling the whole poorly ventilated room. And, I'm very confident that I won't get ill. One, because I'm fit-tested. Two, because I know going in that he's got COVID and I know to be vigilant. And, it's a time-limited thing. I'm in there for 10, 15 minutes, and then I'm out.
That's very different. To say that the, quote-unquote, "We know masks work." I agree: We know under that circumstance masks work. Better to do that than to go in without anything on.
But, that's very different than the policy of: Should we advise average people to wear a mask in their day-to-day life, that sort of community policy?
And, that's always been highly disputed. There have been studies, pre-pandemic, of healthcare workers and influenza season being randomized to different masks, re: yielding sort of diffuse conclusions.
And, the-pre pandemic consensus was largely because there was no evidence. And, this is why when Fauci got on TV the first time he said, 'We don't advise you to wear the mask.' The pre-pandemic consensus was that: we just don't know that this works. And so, we cannot advise it.
In a very short period of time in the first six weeks of the pandemic, in the throes of, I think, a lot of anxiety, people did a full 180 on that and started to endorse cloth mask, community cloth masking.
I didn't disagree. I mean, I think you can do things out of the precautionary principle all the time. But, what I really heavily disagreed with were that there were ongoing efforts to run cluster randomized trials in developing Western--sorry--in developed Western nations that were, I think, actively thwarted by sort of a, maybe even well-intentioned propaganda or lobbying effort to get people to do it. And so, the United Kingdom didn't run any cluster randomized trials of cloth masking. And, the United States ran zero. No nation in the whole world. And, nobody actually ran any in children. And, nobody ran any in the place of greatest dispute: kids from two to five. So, we are going to leave this pandemic with very little sort of really well done studies on this topic. And, it was an imminently answerable question.
The next thing I'd say, is that--
Russ Roberts: Wait a minute. What should we have done? What should we have done? What do you mean by a clustered randomized trial? You would've done what?
Vinay Prasad: Yeah. So, I'll tell you about the one that was done and that'll give you--you know, I think we should have done in the United States. So, Abaluck, Jason Abaluck from Yale University, who is an economist--he ran, to my knowledge, one of two cluster randomized trials. The other was in the Guinea-Bissau. It's not yet reported. And, he ran his in Bangladesh. He chose rural Bangladesh because he knew at the time of his study--and I believe this was in the fall of 2020, early 2021--that they had very little prior COVID-19 there. The baseline zero prevalence was very low. And, he randomized hundreds of villages to three different strategies: No masking, cloth masking, or a surgical mask. And, these are a little bit different than the masks you see. I think he actually specifically had these manufactured.
And, he--as all good economists--not only did he do a randomized trial, he also had different incentive structures to get people to wear it. He had blind people observe villages to see what the level of compliance was.
He found many notable things. One thing he found was that control villages maybe had something like a 10% compliance rate with mask wearing, even though they weren't given a mask and weren't told to wear masks. The intervention villages, it went up to, like, 40%. So, he did have some delta on that masking. And, the rate of COVID--and the primary endpoint of the study was you had to have symptoms: you phoned a number and they wrote down how many people had symptoms. But, then they used blood-test-confirmed COVID-19 to validate that it was COVID-19 and not just that you were worried you had a tickle in your throat.
And, by that primary endpoint, the control villages had a rate of COVID-19 of, I believe, 0.76. And, the intervention arm villages, it was 0.69. So, that's the statistically significant 11% risk reduction. But, we're talking about three quarters and two thirds of one percentage point. A very tiny effect. And, actually it was only true for the group that had the surgical masks and not the group that got the cloth mask. Cloth masks did not work in that study. So, I think that's why a lot of us felt like cloth masking was an ineffective strategy.
One of the problems of the Abaluck study, and he may--and I know he'll disagree--but, one of the problems is there's an imbalance in randomization. In the villages that he gave a mask out, about 10% more people signed up to participate than the villages where he didn't give a mask out. And, of course, this is all randomly generated villages. So, how could that be? That's a statistically significant 10% imbalance.
And, I think the answer is that when they went to randomize people, the villagers knew that they were going to get a freebie in the villages where they had the mask. They pulled up in a, maybe a big truck and they had boxes. And, then the other village, they didn't think they were going to get a freebie. And, this is a bias called concealment bias. Randomization wasn't concealed from the participants. And, what that led to was a 10% imbalance in signup.
And, the reason I'm saying this is because I think the results of this study hinge on one thing: That 11th person who signed up in the villages that got a mask versus the 11th person who didn't sign up, is that person equally likely to phone that number and report their symptoms if they are ill? And, I would argue that they're not equally likely. They are the kind of person who only is signing up because they think they're getting a freebie. And, if they weren't getting a freebie, they're not going to sign up. And so, if they have symptoms, they probably don't really care that much about your study. And, they may not call you.
And so, I think that the entire result should we have a little asterisk next to it because it might just be due to imbalances in randomization. So, that's my thought on it.
Russ Roberts: But, isn't it such a tiny, tiny incidence in that setting that it's not a very reliable--
Russ Roberts: And, it's not probably a setting where a lot of people go to choir practice or basketball game in a crowded arena or sit in a big restaurant with 60 other people. I don't know--it just strikes me as maybe not that informative in either direction. But, that's what we have, you're telling me.
Vinay Prasad: That's all we have.
But, I will also add to you: and that's why I say that what we should have done. We should have done that exact same thing in the United States. I mean, we should have matched municipalities and cluster randomized them.
And, I mean, if you think about how--I mean, you'll know better than me--what will be the global losses from this pandemic? I mean, $20 trillion or even in the United States, I mean, tens of trillions of dollars. And, I think a nation that is willing to, I mean, print $5 trillion in money and lose $20 trillion in GDP [Gross Domestic Product], but is not willing to spend, you know, $10 million, a hundred million, on running sort of just basic public health randomized trials? I think that, to me, is a problem.
And, then the last thing I'd say is: The CDC complains that trust in public health is reaching an all-time low. And, I think that there's evidence of that to be the case. And, this issue like masking is supposed to be a science issue. It's become a political issue: I know that you're voting for Biden when I see you wear the mask; and I know you're not, when you're voting for Trump when you don't. And, it wouldn't have been a political issue if we had taken it over as a scientific issue. If we had run the study, I think we would have diffused a lot of it. But, politics thrives in a evidence vacuum. I mean, there is no evidence that cloth masking helps or hurts. And, there's no evidence it helps or hurts kids. And, naturally the Right will say it's child abuse. And, the Left would say that if you don't do it, you're a child abuser. It becomes a political issue; and that's very dangerous for society, I think.
Russ Roberts: So, I've made the following observation, which--I think it's true. It might not be, but I think it's true: In the early days of the pandemic, I was a big supporter of mask wearing. I felt in the early days--and I'm still not against it, although I am for kids. Or would be for my own kids. What you do with your kids is up to you. But, I felt at the time, a relatively small cost for adults and potentially high benefits. So, wear them.
I remember when President Trump came out for a press conference early in the pandemic surrounded by 10 other people, all really standing close together, none of them wearing a mask. And, I thought, 'That's a terrible--,' I mean, it'd be so easy for him to send a signal.
So, the thing I think is true, which--unverifiable; won't be a study 20 years from now about this; nobody's going to be arguing about it. But, I always wondered if he had come out religiously wearing a mask, and said, 'Everybody should wear a mask,' I wonder if the mask religion would be reversed--
Russ Roberts: And, that the Republicans would be saying, 'Masks are the only, or the best, thing we've got.' And, the Democrats would be saying, 'It's an infringement on our civil liberties. How dare them--how dare they tell us to do something that we have no evidence for?' But, that experiment will never be run. So, anyway.
Vinay Prasad: Democrats would say exactly, as you say; and they would also say that we know facial recognition is very important for young children and we should not put it on young children. And, they would also say: we know people of color are disproportionately harassed if they're wearing a mask or are feared, or something like that. And, it's unsafe for people--I think Democrats would've had that opposite view.
There is evidence to what you're saying when it comes around school closure. There was a week in June 2020, where Trump came out publicly in favor of school reopening. And, if you look at public opinion polling--and I think Vladimir Kogan from Ohio State has done this--you'll see an inversion, by party, in whether or not schools should be opened the moment Trump put his finger on that issue. And, that's why we see today Left-leaning progressive school districts in this country were the longest closures. And it was Texas and Florida that reopened quickly.
And, I guess the reason it really bothers me is that, you know, just because he's a--I mean, a broken watch is right twice a day. He was right about the schools. And, I think many people just irrationally opposed him.
On the masks I'm actually very close to you, Russ. I never minded wearing it. It's no big deal. I wear pants. I wear a lot of things that--just out of social convenience. I never minded it.
But, I do mind that--what I do mind that we didn't generate the data. That really bothers me. I would have loved to see the CDC, on many issues, run that study.
Russ Roberts: Yeah. Well, there were a number of--yeah. Just a small technical question: So, and by the way, I just want to say for listeners, Vinay is only visible from the waist up. So, his claim that he wears pants, there's no evidence for it. Just want to say that.
Vinay Prasad: It's not falsifiable.
Russ Roberts: Exactly.
Russ Roberts: So, I'm sitting here breathing, and I'm putting out particles all the time. And, I assume there'd be a way to measure when I'm wearing one kind of mask versus another kind of mask: How much of my breath gets out a certain distance? Can't we get some evidence like that? I mean, don't we know something about the size of COVID particulates that are troublesome? And, we know what a cloth mask does versus paper mask versus--can't we get some, I don't know what you would call that--micro evidence?
Vinay Prasad: And, I think there are--I mean--I think when you see statistics, like the N95 is x- percent, and this is y-percent, it often is a type of study like that. They are typically performed--a mannequin in a chamber measuring exhaled particles.
But, just to draw a line, let's just take two- to four-year-olds in California. Throughout most of the pandemic, two- to four-year-olds by law in daycare had to wear a cloth mask. That was the public health policy.
Except--except--when they nap. Because that's a safety risk when they're napping.
And, they nap, actually, Russ, all next to each other in the same room. They just kind of unroll some yoga mats and they all nap. And, that room of course often has stagnant air.
So, I think the question I always asked is that: How much good could this be doing, if it's filtering some fraction of particles, when y'all take it off and spend two-to-four hours taking a nice siesta? I mean, how good could it possibly be doing?
And so, that's why I always think that, you know, the kind of study you're you're talking about, in my mind, is a bio-plausibility study. It's like early drug development. But, the proof is in the pudding, which is sort of real world, empirical, public-health kind of studies.
Russ Roberts: Yeah. Yeah. Oh, for sure. And, I--I haven't talked about this on the program for a while. But, my understanding of the six-foot mandate was that it was based on a single airplane observation about one person who had a different coronavirus--MERS [Middle East Respiratory Syndrome] or whatever it was. And: How many rows away people got sick three days later. And, therefore we just decided: it's six feet. I mean, it's not science. Which is okay. I mean, it's--
Vinay Prasad: I always say that if somebody's trying to persuade you of something and the first thing they do is pull out a seating chart from an airplane, it's not science. I don't know what it is. It's not science, they've got to pull out that seating chart.
Russ Roberts: How about a restaurant? Because that was the other one. There was this restaurant where they had their air conditioning on--
Vinay Prasad: Hair dressing. Yeah.
Russ Roberts: I haven't heard that one. But, in the line of the restaurant, everybody in the tables with the ventilating, they all got it. Interesting; could be true. It's a small sample.
Russ Roberts: Let's close and talk about the public health establishment. In a recent episode of this program--I can't remember which one--I said that Fauci had misled the American people. And, I was--a listener wrote me and said, 'No, it's not true. That's a lie. He just didn't have complete information at the time.' My memory is that he conceded that he had lied and that he had done it in the name of preserving supply of mask for healthcare providers.
Now, I've said many times in this program that just figuring out if that's even what I just said is a true statement is so overwhelmingly difficult now because of the polarization and politicization of information on the Internet that I'm not totally sure that's true. But, that's my memory. Can you speak to that? And speak more generally of whether the public health experts have been honest and what that's done?
Vinay Prasad: So, it's a good point. So, I guess I want to start by saying that actually I have a lot of respect for Tony Fauci; and I've actually met the man face-to-face. I did my training at the NIH [National Institutes of Health] and I have a tremendous respect for him. And, I think he's a important public servant in the history of this country.
But, I tend to agree with you that not only did he mislead or lie, he has admitted as such, at least on two occasions.
One occasion is the occasion you describe where between early March and maybe about six or eight weeks later, he gave conflicting reports about the efficacy of masking. He first said that on 60 Minutes, very clearly that nobody should community mask: It's silly. He made a joke: 'You need to touch your face. It's not going to work.' A few weeks later he was endorsing it. He later said that: 'Well, I lied to you the first time. And, the reason I did was a noble reason, which was I wanted to preserve mask supply for healthcare workers.' But, the problem I always had with that argument is: When you came back the second time and recommended it, you didn't recommend people wear the N95 or the surgical mask. You recommended they take their T-shirt and cut holes in it and take their socks--
Russ Roberts: Socks--
Vinay Prasad: Yeah. And so, if you were always going to recommend a homemade cloth mask, that would have never jeopardized the supply of the masks to the healthcare providers.
And so, there's a bit of a tension there.
And then, to be honest, what I really think happened is I think it was the second time that he lied. The first time, I think--and actually, I think historians will someday say that he was actually very much in line with WHO/CDC [World Health Organization/Centers for Disease Control] guidance and the preponderance of public health thinking.
There was a very quick reversal in public sentiment. I think Jacob Hale Russell has a very brilliant article about this in Tablet Magazine, "The Mask Debacle," where he talked about that change in public sentiment.
And, I think when he came out the second time, he was maybe overly enthusiastic and upsold the mask, maybe in part because he wanted to get people to try it. And, maybe he wanted to offer something. I mean, maybe he had a different noble intention.
The second instance that I know he has admitted to lying was to the reporter, Don McNeil. Don McNeil, of course now the fired New York Times reporter--and, that's a whole 'nother story--but, Don McNeil had an article where he asked Fauci, 'What is the herd immunity threshold?' And, Fauci gave a number that varied--maybe 70, 75, 80, 85%--he kept changing the number. And Don McNeil asked him, 'Why are you changing the number?' To which he remarked: 'I look at the public opinion polls on vaccines, and I see what percent of Americans are going to get it. And, then I think to myself, I can nudge this up a little bit.'
Meaning that--I think, what he's saying is that: If I push it up a little bit, I can motivate people to make a saluatory choice.
And then, I think you can go back--there's a lot of comments of when Ron DeSantis reopened schools in the Spring of 2020, I think Fauci said a lot of things that were wrong, that kind of oversold the risk to kids and actually may have further poisoned those efforts and kind of shifted the dynamics around Teachers' Unions.
So I think, like all public figures, he's fallible. He did make mistakes. I think it's true of anyone in that situation.
My own view is that at some point--as is often the case for spokespeople--you get caught in these lies a little bit too often that maybe it's best that you're no longer the spokesperson.
And, I think some time in the Biden Administration, they felt that, because he was no longer on TV all the time, and he kind of stepped back a little bit.
Russ Roberts: My view on this is that, like you say: Everybody's human. We make mistakes. I think it's important to admit them. I have no idea whether he had, the fact that he admitted that he misled people doesn't mean he actually did. He could just be covering his embarrassment, that he didn't know what was going on. Who knows? That's not so important.
But I think admitting that you deliberately--whether it was true or not--misled the American people, shows a lack of respect for the average person.
Now, you could argue the average person deserves a lack of respect; and maybe Tony Fauci feels that way. Or maybe he doesn't. Doesn't matter. But, I do think to come back to [inaudible 01:13:50], I think very deep insight that we have responsibilities and it's hard to live up to them in 2022. People tend not to.
I think if you're an expert in a position of authority, you need to tell the truth. And, if you make a mistake, you need to admit it.
And, I don't think that's anybody's strong suit. I think people like to be--they like the platforms that they perform on and they're not so keen on our responsibilities.
So, that's my--I don't want to say [inaudible 01:14:22] Fauci. I don't think he's alone. I think he's got an enormous amount of company.
And, I think all of it has diminished the ability of people to trust people who are experts.
And, there's some good things about that. It's not all bad. But, I worry about the next pandemic. I worry about the next crisis of any kind. And of course, we're in many of them right now. And, people are saying things I think are often not true. But, they justify it by saying, 'Well, I have to say it that way, otherwise people do x, y, z.'
And, anyway, why don't we close--
Vinay Prasad: Wait, I just want to say this one comment on this.
Russ Roberts: Yeah. Go ahead. Go ahead.
Vinay Prasad: I mean, I think what you said is really brilliant and is really the way I feel, which I think is that scientists in particular, when put in public health positions--even if they're put by politicians--I think they have a duty first and foremost to say the truth.
And, one of the things about the truth is often the truth is uncertain. Uncertainty isn't great when you want to motivate armies to go to war. Uncertainty isn't great when you want compliance. But, uncertainty is honesty. And, if you fain certainty when you don't have it, you will easily lose credibility. I mean, there's that video clip of Walensky [Rochelle Walensky, Director of the Centers for Disease Control--Econlib Ed.] saying that 'We know vaccinated people don't carry the virus.' And of course, she, at the time, already had some hints that was incorrect.
And, we've seen some things that have shaken my faith in public health.
One of which is the resignations of Marion Gruber and Phil Krause. Marion Gruber and Phil Krause were the Deputy and Director of the U.S. Food and Drug Administration Vaccine Product Division--positions they've held for decades. They withstood the political pressure of Donald J. Trump: they didn't resign under Trump.
But, in the course of the pandemic, when we got to universal boosting, they resigned. And they've written many op-eds in Washington Post. They wrote a paper in the Lancet. They were always critical of the idea that we needed a booster for all ages. They were much more willing to consider booster for elderly, but not for younger people.
They also, at one point, asked that the sample size of the kids' vaccine studies be increased, because I think they were also very concerned that maybe we're not measuring what really matters to kids--which is severe disease. And, I think that their resignation--in the New York Times was well-covered--that they resigned because they felt White House pressure for them to approve boosters. And, they didn't feel that was fair.
That, to me, speaks to loss of trust.
And, I guess I would say that I'm very worried, Russ, that what might happen going forward is that this science--which I truly believe in is this objective reality that whether you're a Democrat or Republican, there is some science. And, there's some things we can say with certainty. But, I worry that science will increasingly become a branch of political parties: and, they'll have their scientists and we'll have our scientists. And, we will literally be saying different things.
And, if we get to that state, I think we will truly have lost something, one of the things that have always made America great, which is: some shared sense of science.
And, I think we will be extremely vulnerable to, as you say, any crisis--it could be climate, it could be astronomical debris. It could be any crisis where you need smart scientists, policy makers, and public preference to, like, 'That Venn diagram needs to interact.' It's going to be, whatever the issue there's going to be two polar views. There'll be no way to settle the issue. And, I think the net result will be paralysis and just total failure. I'm very worried.
Russ Roberts: Yeah. It would be really bad if real science started to look like economics, where they have their scientists and we have ours, whoever they and we are.
Russ Roberts: But, let's go to our closing conversation, our closing issue.
I thought--naively, as I think you're hinting at here--that in the aftermath of a two-year-plus, very tragic experience where millions of people have died, that we would learn some lessons that we'd all agree on. We haven't. This is clear.
And, I think, if I had to pick one thing we ˆhave learned, I think even though vaccination is not universal, I think most people understood correctly--and, I think correctly--that vaccination is a good thing. For whom, how many times, that's a difficult question to answer.
But, that's kind of like all we've got
And worse than that, I think a lot of the things we've learned about vaccination aren't quite true. Like, 'Once you've been vaccinated, you're not going to spread the disease!' Those kind of things are just horrible that they got put into the intellectual space.
So, talk about what you think is the most important thing we should learn--have learned--going forward.
And, then--this is a little more pessimistic: 'Are we really--there will be more--and are we unprepared for those in terrible ways because of the way we've handled this one?
Vinay Prasad: More pandemics?
Russ Roberts: Yeah.
Vinay Prasad: I mean, I guess what have been the strengths of what we've done, I guess, and then what, yeah.
So, I guess I'd say that there were a few groups that actually did lead the way. There's the United Kingdom's Recovery Trial. And, they were able to, in a short period of time, randomize tens of thousands of people to different drugs. And, they're the ones that told us dexamethasone works. And, they told us that convalescent plasma mostly doesn't work. And, they gave us a lot of useful insights.
And, I think the development of things like that was a great boone.
Vaccination--I think the science was a boone. I mean, the mRNA platform was incredible. It didn't exist before and very likely it will be leveraged for many other things in the years to come. I think the initial vaccine and for the elderly saved many, many lives. And, I think it is probably the single greatest accomplishment.
I always put our COVID response in a couple buckets. There's things you do that lowers your risk of bad outcomes when you meet the virus; and there's things you do that delays the time to meet your virus. And, there's things you do that you think delay the time, but doesn't even do that.
And so, wearing cloth masks outside, like I see people; or, boarding up playgrounds. Those were things we did that we thought delayed the time to meet the virus, but probably don't even do that.
Russ Roberts: There are things you do that really do delay the time--you know: being a hermit, staying in your house, preventing your kids from ever playing. They do. But, in my mind, those only make sense when you're waiting for something from the first category, a risk reduction thing.
And, once you've had all the risk reduction, I don't think those make as much sense to me. Because, the virus will circulate forever.
The thing I think we did the weakest is, is my bias, which is: I think we succumb in medicine so often to this fallacy that something worked well in a high-risk situation so we should not question it as we go lower risks. I mean when it comes to cholesterol medicines, statins work brilliantly if you've had a heart attack and high cholesterol. And, now people talk about, well, should every 18-year-old take a statin. I think it's a very different question. I don't know: I'm not persuaded. But, there are cardiologists who say that maybe it should be every 18-year-old. Some even say it should be in the water supply.
We know chemotherapy works very well if you have certain advanced cancer. Should we look earlier for pre-cancer and treat you with chemo? I'm still unsure about that.
And, I think the same pattern is true with COVID-19 where the vaccine works well in older, frailer people. Does every four-year-old, even one that just had Omicron, does that four-year-old really need it? And, then all this: Should it be mandated? Should it be required? I think those are all problematic spaces.
I think the final thing that I would leave your listeners with is: I think there were perverse financial conflicts of interest. They weren't always the ones people thought, but there were some that I think are really problematic.
In my mind, the testing industrial complex is one. These whole pre-departure testing to come back to the United States was finally dropped. There are many prominent physicians who are consulting for those testing companies.
A lot of kids on college campuses are in this sort of quagmire of testing every week. And, if there's a lot of cases, they close; they make them stay in their dorm rooms. This is largely based on evidence generated by the testing industry, which has a very strong motive to keep us testing forever.
But, a number of European nations have gone the other way. They say: Don't test kids unless they're really sick and need to go to the hospital. Don't test yourself unless you are eligible for some therapy. This is something that we might have to, to some degree, live with. And, the more you test, the more you can sort of paralyze systems.
The same for, I think, conflicts of interest around Paxlovid and things like that. Because I think Paxlovid works wonderful for an unvaccinated person at high risk. But, I really don't know if it works for the 30-something-year-old tech worker who's gotten three shots in San Francisco who tells me he's taking it. And, I think that they were those kinds of questions.
Russ Roberts: Normally I was going to end here, but while you're talking, I had a thought and we didn't talk about it. It's strikes me as kind of important.
One of the extraordinary parts of the vaccine platform you alluded to earlier in this last part was that we could adjust it to the flavor that we have. Why are we still giving out Wuhan booster? That makes no sense. I don't get that. Like, as soon as Omicron hit, and I know many fewer people have died Omicron than Wuhan or Delta. But--it's unpleasant, right? And, people again told me to get the fourth booster. I didn't do it. But, it probably would've been a better idea--it would've been made more sense to take a fourth booster--if it was Omicron-compliant, whatever the right immunological word is. Why didn't that happen?
Vinay Prasad: I think that--
Russ Roberts: Do you know?
Vinay Prasad: No, no, but I totally agree--
Russ Roberts: Was it like a big stockpile of the old one, you want to use it up first?
Vinay Prasad: That was one thought I had, actually. I did have that thought--that they have set their commercial apparatus to make so much, and they do want us to exhaust that supply before they retool. That was one thought.
But, your point is 100% correct. The boone of the mRNA [messenger RNA] platform was that it was so quick to change direction. And, yet we have not changed direction at all.
There's one risk to changing direction, which I just want to leave your listeners with. There's talk right now that we're going to go to an annual COVID-19 shot. What should be the level of evidence for that? In my world, I'd love to see you prove to me that shot works in the Southern hemisphere when they're having their winter surge and then bring it to the Northern hemisphere for our forthcoming winter surge. I worry that the FDA is going to have a regulatory apparatus that says: 'Here are three strains. If you can prove to me, you can hit these three strains, you can get marketing authorization. You don't really need to have large human trials proving it works.'
And, the reason I'm a little worried about that is--in addition to the fact that they might miss; it might be for a strain that actually didn't turn out to be a problem--is that we never fully understood why we have myocarditis. And, there may be something about that spike protein that has some part of it looks, something like something on the myocyte, on the cardiac cell. And, it creates an autoimmune reaction against your heart. If you change that spike protein, it's possible that you have less myocarditis; but I think it's also possible you have more myocarditis. There might be some strain that has a lot of myocarditis.
And so, I feel like one of the reasons they didn't retool is they felt like they were in an acceptable place in terms of safety and some amount of efficacy that was diminishing, and that they felt that if they retool quickly, there's a risk that they'll be more efficacious, but there's also a risk of unanticipated safety signal. And, I think that might have been a barrier to them. And, they said from a--I mean, I hate to say this, but if I was in their financial meetings, I would say that the strategy from their financial standpoint for the company would be: You want to get people to be indebted to your product in perpetuity. That's the only way you're going to make a lot of money for years to come. So, you need sort of an annual shot. We're going to get the annual shot. You need to make sure it's safe. But while we've already retooled our apparatus to make a lot of the original strain, we got to try to get that out as much as possible.
Russ Roberts: And--there's no music on EconTalk, but in general--except for the theme music at the beginning and the end. But, this would be the place where we play some ominous violins or maybe an organ.
My guest today is Vinay Prasad. Vinay, it's been great talking to you. Thanks for being part of EconTalk.
Vinay Prasad: Thanks for having me, Russ.