The Good, the Bad, and the Ugly of the Covid Vaccine (with Vinay Prasad)
May 27 2024

covid-information.jpg The Covid vaccine saved many lives but so many mistakes were made in how public health officials discussed it, implemented it, and assessed its effectiveness. Epidemiologist Vinay Prasad of the University of California, San Francisco talks with EconTalk's Russ Roberts about what went wrong, the costs of the mistakes that were made, and what we can do better the next time.

Gregory Zuckerman on the Crazy Race to Create the COVID Vaccine
In the race for a COVID vaccine, how did a couple of companies who had never produced a successful vaccine make it to the finish line so quickly? Gregory Zuckerman talks about his book, A Shot to Save the World,...
Adam Cifu on Ending Medical Reversal
Why do so many medical practices that begin with such promise and confidence turn out to be either ineffective at best or harmful at worst? Adam Cifu of the University of Chicago's School of Medicine and co-author (Vinayak Prasad) of...
Explore audio transcript, further reading that will help you delve deeper into this week’s episode, and vigorous conversations in the form of our comments section below.


May 27 2024 at 7:17am

How can you still introduce the V to this episode as „saved many lives“?! They rather destroyed many lives -.- Incredible denial of reality from the establishment folks still…

Todd K
May 27 2024 at 11:22am

Vinay Prasad:

Certainly, randomized trials have shown the severe disease part, and some observational data has shown the death part. That’s a very good thing.

I thought this was interesting because I have heard Prasad strongly criticize observational studies may times over the past two years.

And [vaccines] did aid in ending the pandemic. By that I mean, COVID-19 will continue to spread. It didn’t stop the spread.

I don’t think the CDC data support the claim that vaccines aiding in ending the pandemic by much if at all. From April 2020 through March 2021 when there were still many unvaccinated elderly, 520,000 Americans died from/with Covid. From April 2021 through March 2022 when a very high percentage of elderly and vulnerable were vaccinated, 400,000 died from/with Covid. That is a small reduction considering that pandemics usually naturally get weaker in a second year.

Ken P
May 28 2024 at 11:14pm

I tend to agree with Vinay that there was a large benefit to high risk populations, and the RCTs back that up, but I agree that the observational data is pretty weak.

You make a good observation but it’s especially hard to compare across time periods with different strains, etc. I can think of harmful interventions in both of the periods you mention that would skew either period to a worse outcome than the other. In period one, there were many mistakes with vents and also pushing infected into nursing homes with the high risk.  In the second period, people were likely less healthy, after a year plus of getting less physical activity, gaining weight.

Todd K
May 28 2024 at 11:48pm

One thing I forgot to mention is that in late January 2021 when Biden was sworn  into office, the CDC  ordered test centers to lower the PCR testing cycles from  40 to 45 down to around 30. That automatically lowered the number of deaths classified as Covid deaths and this occurred three weeks after the Covid death peak of early January 2021.

The vaccines may have helped a little but not obvious.

Todd K
May 29 2024 at 6:16am

(That is, the vaccines may have helped a bit to end the pandemic, but not obvious.)

T Coddington
May 30 2024 at 12:19pm

I don’t personally see how claims of remarkable VE against death in the elderly can be squared with the data from summer 2021.

Ken P
May 30 2024 at 11:16pm

The PCR protocol change is something I wasn’t aware of.  I’m not very swayed by observation data but you did put some interesting analysis into your substack.  I wouldn’t put too much faith in any serological findings.  Antibodies decline rapidly so it means you haven’t been vaccinated or been infected in the last couple months, but it doesn’t tell you anything about last year.  I don’t think that impacts your analysis, just throwing that out there for information.

Memory cells (B & T) can rapidly respond and T-cell responses are likely key to protection.  T-cells are also apply less selective pressure than antibodies and therefore mutations are less escape prone for Tcells.

The number of ACE2 receptors is like 10,000 fold higher for certain health conditions and the elderly and their immune responsiveness, especially to new antigens is much lower and has less breadth in the response.  The T cell receptor repertoire is much more narrow for elderly.  For those reasons, it would not surprise me if a high percentage still die, but even 25% death reduction would be a great benefit with the low vaccination risks seen in this age group.  I’m tempted to revisit the Moderna RCT study, but I’m kind of burned out on the topic.

May 28 2024 at 11:55pm

Yes. Vinay usually says that RCTs is the only “real” evidence in epidemiology (and for good reasons).  If I recall correctly, there is no RCT evidence for these up-front claims that Vinay is making about severe disease, and he usually pays attention to that. But he also pays attention to balancing just enough on the “right” side of the Overton window. The RCTs showed (temporarily) fewer positive tests among symptomatic (in non-risk-groups). No mortality reduction. Not even less sick people compared to control group (other colds? flu?). One group of researchers later tried to pool available crumbles of data from the big pharma RCTs and concluded that the adenovector product may have been the better one for ACM. It is astonishing how these studies were designed (and halted), and we end up scratching our heads instead of sitting on solid conclusions.

Jun 1 2024 at 2:03am

FWIW, this is *the* reference statistically and clinically persuasive state-of-the-art RCT evidence for mRNA severe disease efficacy (hat-tip Vinay):

30 “severe cases” (as defined in the text) in placebo arm, versus none in treatment arm. Among those there seem to be 2 ICU visits. Not sure there is any reporting of the test-negative severe “non-case” background? In a study with 30k people over several months, are these numbers significant in any real practical sense? The magnitudes of the exclusions are much larger for example. Who knows.

Shalom Freedman
May 27 2024 at 11:24am

This richly informative conversation taught me many things I perhaps should have learned or known many years ago. The first is that public health officials are not necessarily motivated by considerations of truth and public good. Politics and self-interest intervene too often. The example given Prasad around the Trump- Biden election is an especially troubling one. Roberts and Prasad agree on the principle that the noble lie’ is not the right policy and the job of both the scientific researcher and the public health official is to find and tell the truth.

Prasad’s story about the difference in risk-taking between giving an experimental drug to a terminally ill cancer patient and a new cancer patient has not had the extensive treatments the far sicker person has had makes it seem that simple common sense is also often the basis of right decision.

It is possible listening to this conversation with Prasad to easily understand why he is so popular with the audience of ‘Conversations for the Curious’.

Prasad is not dogmatic, is open to evidence, makes the impression of being a doctor and person who can be trusted to search for the truth and tell it as he understands it.

Ed Hoopman
May 27 2024 at 12:24pm

I’m very skeptical of this interview, and I’m concerned that libertarian priors are enabling carefully framed inaccuracies to bleed through.

I’m not any kind of immunologist or medical professional, and I’m also concerned that MY priors and subscription to CW are in the way in this case.

Two examples I hear: one, that the probability of myocarditis is greater from vaccination than actually from getting covid, and two, that vaccinating those who’ve already had covid is redundant, as having the disease confers a similar benefit.

I didn’t go back and carefully review both segments, but I don’t believe the counterarguments are confronted.

In the first case, that myocarditis from vaccination is far more likely to be mild/easily treatable, while that induced by covid is much more likely to be severe. And it’s somewhat of a deceptive comparison if the likelihood of myocarditis due to covid is much less * in cases of covid among the vaccinated *.

And in the second case, it’s largely always been acknowledged that those who’ve had covid have similar benefits to those who are vaccinated, but those who are *both vaccinated AND have had covid* have a better response than those who’ve had one or the other.

Again, not a medical professional, coming from a prior position of strong pro- vaccination, I’m just noting that my critical thinking alarms are sounding loudly.

I hope Russ will have someone on who can rebut or otherwise present counterpoints to this point of view.


Ben Service
May 28 2024 at 2:09am

Good points but it does play into the point that we did some pretty dumb things during COVID that reflect badly on the whole profession and system and reduces trust in studies.  You might be right on all your points but how do we know.  It would be good if some people who were on the pro vaccine “side” of things and had power came out and said all the things we screwed up and why those things got screwed up.  I am pretty pro vaccination (I think I gladly got at least 3) but I still think I personally made some dumb decisions and also went along with some dumb societal ones.  Luckily a lot of people in positions of power have good records of what they actually said and did which enables an honest reflection.

Some humility would go a long way to repairing things but it could also backfire you can imagine how Trump “I’m never wrong” would use it.

Ken P
May 28 2024 at 10:44pm

In the first case, that myocarditis from vaccination is far more likely to be mild/easily treatable, while that induced by covid is much more likely to be severe.

Vaccination induced myocarditis primarily impacts the young and healthy, (16-19 appears to be peak risk) who had near zero risk of serious outcome from the disease.  Covid associate myocarditis predominantly occurs in unhealthy and elderly with severe Covid infection… particularly WHEN they have a serious infection. There’s a range of outcomes, but athletic teenagers who end up on heart drugs for years and have to give up sports etc, is not uncommon and is not exactly mild IMO.

The myocarditis signal was making it into the literature around 2021 and many other countries did the right thing by focusing vaccination efforts on those at high risk of infection.

And in the second case, it’s largely always been acknowledged that those who’ve had covid have similar benefits to those who are vaccinated, but those who are *both vaccinated AND have had covid* have a better response than those who’ve had one or the other.

That’s conjecture (even though it is often presented as a given), as is the idea that if two shots are good, three, four, or more is better.  The CDC, and drug companies had plenty of funding to run studies to measure the benefit of vaccinating the previously infected and efficacy of additional doses.  They chose not to run those studies.  That risk/reward ratio is an important consideration with any medical intervention.



Todd K
May 29 2024 at 1:41pm

Ed Hoopman wrote:

“And in the second case, it’s largely always been acknowledged that those who’ve had covid have similar benefits to those who are vaccinated, but those who are *both vaccinated AND have had covid* have a better response than those who’ve had one or the other.”

This is almost certainly not true based on the first study (Israeli) in 2021 that showed those with increased immunity from getting Covid had between 10 to 25 times the protection as those who without having Covid had the two doses of the mRNA shots.



Ben Service
May 27 2024 at 8:08pm

Good discussion and made me reflect on how I made decisions during covid, I mostly just toed the line as none of the impositions were particularly painful to me or my family personally.

It does make me sad now that I don’t know who to trust and how to make informed decisions about things I don’t have much chance of ever knowing enough about to make those decisions.  I’m an engineer that makes petrol, jet fuel and diesel and people can buy those things without giving them much thought, sure the marketing folk in my company will roll out some thing saying our stuff is better than the other companies but at the end of the day whatever is sold is fine 99.99% of the time.

The Trump argument was interesting, why is everything so partisan these days, where are the smart reliable bureaucrats that everyone trusts, did they ever exist?  If they did when did it all go wrong and why?

I now don’t know what to do about an annual flu vaccine, my guess is the risks are very low and getting the flu is not much fun so I’ll probably get one but how do I know this is actually the right choice.

May 28 2024 at 8:18am

Prasad makes the mistake of misstating the case for vaccine mandates as whether vaccination prevents getting and transmitting Covid entirely.

Even a 50% effectiveness against getting and transmitting Covid can significantly reduce the impact of the disease, particularly the stress on hospitals.

Jack B
May 31 2024 at 2:22pm

Agreed, was shocked that the “flatten the curve” point never came up in discussion. It was my understanding as well that reducing the impact on Hospitals during the pandemic was a large reason for the widespread vaccinations. I would have liked to hear that discussed.

May 28 2024 at 9:43am

Prasad’s discussion of the lableak issue is misleading. He ignores the varying degrees of skepticism by scientists and the fact that many lableak skeptics started out with differing views. Contrary to his claims, the authors of one of the key papers did not say a lableak was impossible; the argued it was implausible . In fact papers uncovered by the House Subcommittee indicated that some reviewers felt they gave too much credence to the lableak theory.


Furthermore, other critics of the lableak theory, including Ralph Baric, have also signed a letter asking for a better investigation.

May 28 2024 at 11:32am

As far as the political motivations around the election which were claimed here almost as fact I just need to quote the cliche’ “Extraordinary claims require extraordinary evidence.” I’m not saying that those claims are or aren’t true but for an interview that had so many sentences start like “according to this or that study” the conversation many times strayed into gut beliefs and so forth and kept expecting Russ to push back a little more here and there…

Gregg Tavares
May 28 2024 at 11:36am

I was a little disappointed to hear Russ belittle people in SF for still wear masks.

I lived in Japan for 15 years. People wear masks commonly. I’m here now and many people still wear them. Not just for covid but for everything. Colds, Flu, etc…

What’s wrong with them? Are they useless or do they help (for colds, flu, and covid). I thought their efficacy was well established by now. (note: I wasn’t sure about that and a friend challenged me to look it up a year ago. I did and at least at the time I found they’d been proven effective)

Here’s a statistic: The average person gets 2 colds a year. A cold lasts ~1 week. That means the average person is sick with a cold 2 weeks a year or 14 days. 365/14 is 26 meaning meaning the average person is sick with a cold, 1 of every 26 days. That means effectively, every day, one of every 26 people has a cold. Some of those stay home because they are sick. Some of them have not developed symptoms yet. Let’s just guess then that when you go to a public space (a store, a mall, a grocery store, a library, a restaurant), somewhere close to every one of 50 people you see has a cold. Add in Flu and Covid and the number goes up.

The point being, why would it upset you to see people in public wearing masks? Even if covid never existed it would still be a good idea (as it was before covid in Japan)

Jack B
May 31 2024 at 2:27pm

This was my reaction to this point as well. The characterization that people still wearing masks are doing so purely for political purposes was disheartening to hear. I like the economical perspective you bring to work hours lost, but it’s also just a matter of looking out for your neighbor when you may be unwell.

I think in 2024 we should be encouraging common sense masking. Examples include… when you feel sick, wear a mask in public; while on a crowded bus or train, wear a mask; in a crowded lecture hall or conference room where many are coughing; consider putting on a mask. Not that hard, and could make a world’s of difference on sickness.

Cody L. Custis
May 28 2024 at 12:19pm

Michael Olsterholm, of the Center for Infectious Disease Research and Policy, is willing to make the important statement: “we don’t know.”

To this day, policymakers and fools refuse to state that they acted on reasonable hypothesis that turned out to be wrong.  By claiming omniscient infallibility, they demonstrate total and complete fallacy.

Ajit Kirpekar
May 28 2024 at 12:59pm

I am midway through the episode and I have a sense of deep moral outrage at how politics  – in particular one’s aimed at Donald Trump – can distort incentives such that the life and death of the public do not outweigh them. Its disturbing.

I don’t like Donald Trump’s media personality either(I don’t know him personally so I can’t say what he’s actually like); but to let those feelings warp your natural incentives to “do the right thing” is infuriating to me.

Ben Service
May 28 2024 at 6:21pm

Both sides of politics do this though and it saddens me as well.  What is the cause of it, both sides have populist politicains who like to alienate and vilify people on the other side in order to create support for themselves, is this new or old as the hills, has it had a revival recently, is it a bad thing (my gut feel is yes but open to being wrong on this), if it is bad what is a path forward, what is the path forward going forward towards?  People talk about Trump derangement syndrome but I think Trump went a long way to cause this, he certainly isn’t the least divisive politician around, I am guessing there are politicians from the past who were also like this though so I don’t think it is a new phenomena.

I know I personally do get affected by who is telling the story and whether I like their type of world view or not, I do try to be aware when I am doing this though.  I’d challenge you to think of situations where you have done the same and recognise when you are doing it.

I am pretty leftie (probably would be called a libtard by folks on the right), but obviously open minded enough to listen to shows like Econtalk, Goodfellows, the libertarian and absorb the ideas.

Ajit Kirpekar
May 29 2024 at 5:30pm

I absolutely am likely to be affected by it. I think it takes a lot of effort to avoid being swayed by tribalism. That said, and this occurred before Trump – I feel this general hostility that comes with these things. Essentially, differences in opinion have allowed vulgarity and demagoguery to be the natural reactions.

Its utterly depressing but I have to believe that has been the case for a while. I imagine Milton Friedman dealt with death threats in his time. The booing of his Nobel Prize was just a stark example.

Ben Service
May 30 2024 at 3:37pm

Thanks for replying Ajit, sounds like you and I are on different political sides of most things but I think at least we could sit down and have a coffee and a pleasant conversation with each other.  The best thing about the EconTalk comments section is that is very civil from both sides of the debates, I suspect this stems from the way Russ runs his show and the listeners it attracts vs say marginal revolution.

Ron Spinner
May 29 2024 at 6:05am

Very interesting juxtaposition to hear that the head of Pfizer may have influenced the last presidential elections against Mr. Trump. At the same time that Mr. Trump is on trial for influencing the elections.

Vinay Prasad
May 29 2024 at 1:24pm

In response to Erik, “there is no RCT evidence for these up-front claims that Vinay is making about severe disease”
Incorrect. The Moderna randomized study shows an imbalance in severe disease, and this is highly statistically and clinically persuasive.

In response to Ed’s “myocarditis from vaccination is far more likely to be mild/easily treatable, while that induced by covid is much more likely to be severe”

This is incorrect. First, myocarditis after vaccination occurs in young healthy men who have nearly no COVID19 risk. While much is self resolving, MRI shows persistent gad abnormalities suggesting scarring, which may precipitate death from arrhythmia in future. Finally, a tiny percentage does lead to frank failure and death. The risk of harm is far greater than potential for benefit, particularly with doses after dose 2.

Myocarditis after vaccination is an idiopathic side effect that occurs in isolation from other malady. Myocarditis after COVID itself is most likely demand ischemia in people who are critically ill. In other words, one is an immunologically mediated phenomenon and the other is part of a constellation of bad outcomes.

Myocarditis after COVID19 (aka troponin rise) mostly affects old, vulnerable people who get really sick. Finally, the studies that document its frequency use the wrong denominator of EHR coded infections and not all people who got sick— who mostly didn’t test.

The final point is it is not risk of one vs the other. Vaccination does not stop COVID19, so it is the risk of COVID induced harms + or – vaccine harms.

In response to Jonathan, ”Prasad makes the mistake of misstating the case for vaccine mandates as whether vaccination prevents getting and transmitting Covid entirely.

Even a 50% effectiveness against getting and transmitting Covid can significantly reduce the impact of the disease, particularly the stress on hospitals.”

Here is why Jonathan is wrong. First, consider a world with and without vaccine mandates. A world without vaccine mandates still has a lot of vaccination, as some people want it. A world with mandates, might have slightly more, as some people get it only because of the mandate. Others may prefer to be fired (which has negative consequences). How many people got vaccinated in the US just because of the Biden mandates?

The answer is shockingly low. Btw 1 to 4% by my estimates. You can perform discontinuity to prove this to yourself.

Now does an increase in vaccination by 1 to 4% change transmission dynamics, when the vaccine has a 50% reduction in transmission that fades to 0% or perhaps even negative vaccine effectiveness at 4 months? The answer is absolutely not. As such it did nothing to shield hospitals from the overflow that they never had anyway in third quarter 2021.

Vaccine mandates slow transmission the same way watering your lawn stops forest fires.

Ajit Kirpekar
May 29 2024 at 5:35pm


I appreciate you taking the time to read and respond to the comments. It helps someone like me who is not in this field or in this data to reconcile the criticisms such that it can lend validity or credence to what I am learning.


May 30 2024 at 10:01am

I’d echo the comments above that this episode gave comparatively short shrift (if any shrift at all) to the counterarguments to Vinay’s points. Russ typically at least plays devil’s advocate a few times an episode, even when it is obvious he agrees with the guest; this time he largely just allowed Vinay to speak at length and nearly entirely unchallenged. This is troubling given Vinay’s views are far from settled science (as the controversy around him should make readily apparent), but is also especially disappointing considering Russ is very well-versed in statistics, and, as an epidemiologist, the vast majority of Vinay’s arguments are statistical in nature.


To Vinay specifically:

In the episode and your response above, you continue to “hide the ball” statistically and make simple mistakes of ensemble probabilities. Regarding myocarditis rates from COVID itself vs the vaccines, your argument appears to be:  Young adult men suffer myocarditis following vaccines at a rate of x (a very small number); given the risk of COVID in ~20 year olds broadly is very low, we therefore should not vaccinate young adult men. Of course  this is making the elementary mistake of applying the ensemble average (COVID risk in 20 year olds) to a very small subset with a known very rare reaction (those who end up with myocarditis following vaccination). Given your stated assumption that everyone will get COVID, the questions we really should be asking are (1) among specifically those young men who end up with myocarditis from vaccination, how many would have ended up with heart complications from COVID itself? and (2) among young men who, if unvaccinated, would end up with heart complications from COVID, how many cases of COVID-related heart complications would vaccination prevent? The rates you cite do not answer these questions.

But we don’t even need to get as technical as in the above. Instead, just address the elephant in the room directly: mortality. Mortality is the most objective and easiest endpoint to measure, and the simple truth is, even among young people, COVID does cause death, yet, to my knowledge there is no case of a person dying from a COVID vaccine, even after what is now several billion administered doses.

A point made by many (but perhaps most loudly by Nassim Taleb) that seems lost on you is that COVID is not a “disease of the elderly.” Based on the data, it has been clear for a while now that COVID increases the risk of death proportionally across nearly all age groups; the death rate stratification by age we see is a result of the fact that the baseline mortality rate already differs among these age groups. The average 100 year old is many times more likely to die in the next year than the average 20 year old; increasing the mortality rate for both by, say, 20% has a much more dramatic effect on the former than the latter: from 50% to 60% for the 100 year old, but from 0.14% to 0.17% for the 20 year old. We see many, many more 100 year olds dying, but that does not change the fact that both groups took on a 20% increase in mortality risk.

Lastly, while much of the masking debate effectively reduces to questions of morality not worth debating here, I was quite surprised by your comment: “I think most cultures value children ahead of adults and elderly,” which also went unchallenged by Russ. Are we to pretend that the two largest cultures in the world, China and India, don’t exist? A mistake made more bizarre still by the fact that I assume at least based on your name you have some degree of Indian heritage…

Jonathan Golden Harris
Jun 1 2024 at 2:18pm

Responding to:

Here is why Jonathan is wrong. First, consider a world with and without vaccine mandates. A world without vaccine mandates still has a lot of vaccination, as some people want it. A world with mandates, might have slightly more, as some people get it only because of the mandate. Others may prefer to be fired (which has negative consequences). How many people got vaccinated in the US just because of the Biden mandates?

The answer is shockingly low. Btw 1 to 4% by my estimates. You can perform discontinuity to prove this to yourself.

Now does an increase in vaccination by 1 to 4% change transmission dynamics, when the vaccine has a 50% reduction in transmission that fades to 0% or perhaps even negative vaccine effectiveness at 4 months? The answer is absolutely not. As such it did nothing to shield hospitals from the overflow that they never had anyway in third quarter 2021.

Vinay switched from wrongly conflating partial effectiveness with ineffectiveness to asserting minimal impact on uptake.

Other estimates of the impact of vaccine mandates are much higher than hiss. They show that they caused people in some countries to get vaccines earlier (an initial spike of 60% in uptake and a cumulative effect of 4-12 %, see ).

Certainly, these estimates are subject to the assumptions of the models and methods and will vary depending on the nature of the mandate.

If Vinay wants to restrict the analysis to the US mandate, one needs to consider that it was limited by the courts, so the relevant percentage isn’t the population level percentage, but the percentage of those subject to the mandate.

Furthermore, Vinay neglects to consider that fewer incremental vaccinations resulting from a mandate imply a lower social cost.

I would expect him to claim the difficult-to-measure “credibility” cost that can be used to refute any policy that someone disagrees with; however, if Dr. Prasad is concerned about public health credibility, perhaps he should be more careful to avoid exaggeration and accusations of bad faith in his statements.

[Note from Econlib Editor: This commenter, Jonathan Golden Harris, is the Jonathan already in conversation with Vinay Prasad in this thread. We do not usually allow changes in nicknames mid-thread on EconTalk, but we are allowing it this time in the interest of continued conversation.]

Michael Krogh
May 30 2024 at 5:50am

I’m troubled by the “masks don’t work” comments. 30 years of working in industries requiring masking and repeatedly measuring the effects of masking, improper masking, masking with illness, etc. suggests to me that masks do, in fact, work exceptionally well if worn correctly. I’m certainly understanding of the complications, from poor fit to just being stupid (how many people wouldn’t cover their noses?) There’s no question the ridiculousness of implementation as discussed is problematic (like preschool naps) and the tradeoffs with child socialization are legit concerns.
I have no problem discussing and weighing the effectiveness of mandates, the negative aspects of masking-especially for children, the stupidity of masking outdoors, the impracticality of expecting average citizens to properly use a mask, etc. But properly worn masks work. They reduce airborne bioburden by many orders of magnitude. The mask works, even if “masking” as a public health strategy may not.  We should make that distinction clear.

Todd K
May 30 2024 at 10:25am

30 years of working in industries requiring masking and repeatedly measuring the effects of masking, improper masking, masking with illness, etc. suggests to me that masks do, in fact, work exceptionally well if worn correctly.

Masks worn in industry are for particles that are far larger than viruses. A 2015 RCT of nurses and doctors showed that cloth masks let 90% of the influenza virus through and surgical masks let 45% of influenza through.

Many in Tokyo wear surgical masks in spring to block pollen from cedar trees that are in the surrounding mountains. The surgical masks likely help since pollen is 25 times larger than influenza or the new coronavirus.

To Vinay Prasad: Does it make sense to put so much weight on a Moderna RCT?

May 30 2024 at 11:39am

Agreed. The episode was very sloppy throughout on the policy vs physics question of masking. Despite paying lip service to this distinction at the outset (the question of the efficacy of a doctor wearing an N95 to enter a COVID patient’s room was mentioned then promptly ignored), the guest repeatedly presented evidence showing only that masking policies were ineffective (largely due to difficulties with compliance, as Vinay himself conjectured). But both host and guest then went on to ridicule those still masking (be it in San Francisco or choosing to mask children on airplanes), thereby implying that the evidence of policy ineffectiveness entails masking as an individual choice is irrational. Very sloppy indeed.

And that is not to mention that to imply, as both Russ and Vinay do, that in 2024 those still masking in public are doing so to “virtue signal” (Russ’s words) or make some kind of political statement rather than protect themselves betrays a real political bias of their own (and an outdated one at that).

You can rest assured that someone with the presence of mind to procure and use something as substantive as a Flo Mask (which only became widely available after mask mandates were largely gone), is quite likely doing so out of concern for their own health (who knows what hidden comorbidities they might be living with as you silently judge them?), and not because they still suffer from Trump Derangement Syndrome.

May 30 2024 at 2:41pm


I know you are very eager to cite the fact you have been intellectually consistent on the topic of the vaccines, so I’m hoping you will address the following since I haven’t seen you do it anywhere else.

The flu vaccine. Many parallels with the COVID vaccines. For one, both are protecting against a disease that results in what many at this point consider mild illness in most of the population (many today even derisively describe COVID as “just the flu,”), with severe disease occurring in similar segments of the population (the old and unhealthy). The flu is somewhat unique in its danger to the very young, but this is more than outweighed by the fact that COVID has significantly higher rates of morbidity and death for the population in general.

Both vaccines also have a similarly checkered history of effectiveness, although the COVID vaccines are clearly the more effective of the two; even in its best years, the flu vaccine is at best 60-70% effective.

Both vaccines also similarly have serious, but quite rare, side effects. The most salient obviously being myocarditis linked with COVID vaccines and the flu vaccine’s link with Guillain-Barré syndrome (which is arguably a much more debilitating condition than myocarditis), among others. In both cases the frequency of side effects is similarly quite rare, on the order of 1 in 10,000 or fewer.

Lastly, the flu vaccine is, and has been for a while, mandated for most healthcare workers in order to keep their jobs.

I’ve never seen you expressing concerns over flu vaccine mandates for healthcare workers. Why not?

Todd K
May 31 2024 at 1:19am

I’m not Vinay but…

From 2022, coronavirus has been about as deadly as the flu and recently less deadly.
Influenza vaccines are closer to 30% to 40% effective and unfortunately, the band of protection is almost entirely between unhealthy 65 year olds and unhealthy 80 year olds. The mRNA vaccines did essentially nothing for those over 90 because their immune systems were very low to begin with and so many have comorbidities.
Flu vaccines have a far better safety profile than the Covid vaccines where 1 in 800 have had a serious side effect that required hospitalization with some causing death.
In May 2022, a Danish study showed that the mRNA vaccines were risky enough that the lead author said  healthy people under 50 should not take a booster based on original and Delta variants. That year, this became policy in the UK and the EU while the US was pushing for 6 month olds and up to get a booster shot.

Andy Nanneman
Jun 2 2024 at 1:56pm

Question regarding the technical definition of transmission in regards to vaccine research.  For years now critics of the vaccines have highlighted the fact that vaccinated people can acquire and spread covid-19, therefore they do not prevent transmission.  I recall significant outcries when a Pfizer spokesman stating that it was not studied.

I assumed that transmission must have a technical definition and that is lost when people sensationalize such a statement.  Does researching a vaccine’s impact on transmission mean specifically, given a person is infected with covid, they have a reduced probability of spreading the sickness if they are vaccinated?  I could see that not being studied but the spread of the disease would still be retarded if the vaccine reduces probability of being infected in the first place.  Or if it reduces duration of virus shedding.

Unfortunately people like Vinay focus on the fact that it can still be transmitted and that the policies were not nuanced.  It has been 3 plus years and I still do not think I have seen a nuanced discussion about regarding how that despite the vaccines do not prevent spread they do have some effect on “flattening the curve.”

Ralph Casale
Jun 11 2024 at 10:53am

I was disappointed in much of this interview.  I’ll focus on the two main reasons.

First is the failure to distinguish between the virus (Sara-Cov-2) and the disease (Covid-19).  When terms are used like “got Covid” it means a great deal whether we are talking about symptomatic disease or just testing positive for the presence of the virus.  Vaccines do not and can not prevent infection.  They can and do  limit virus proliferation once infected.  But a positive test for the presence of virus via PCR and via antigen have veery different sensitivities.  As the pandemic played out, we moved from only testing by PCR those with symptoms, to testing for antigen in those who suspected they were exposed.  So ‘got Covid’ meant different things at different times, and that needs to be detailed.  To be fair, the CDC also did a poor job of distinguishing between the two, further confusing rather than educating to populace.

Second is the way Risk was addressed.  Since Risk is first and foremost a perspective, the risk to different stakeholders needs to be considered more.  Yes, primary risk is to the individual, but broader risk to society is often a balance to be considered e.g. in vaccination of the young ( for example those in their 20s are less likely to ‘get Covid’ the disease after infection, but more likely to get infected and spread it to others due to social interactions).

Ralph Casale
Jun 11 2024 at 3:59pm

My point is that when you frame risk primarily from the perspective of the individual, and then criticize public officials for their consideration of that risk; you should consider and that the perspective of public officials is more focused on ‘risk to society’ than ‘risk to individual’.  How these can and often do differ is a more interesting discussion.

Gregory McIsaac
Jun 14 2024 at 9:25am

Around 42:00 Vinay Prasad said: The evidence pre-pandemic was pretty conclusive that randomized studies of community masking were generally negative and did not appear to work, and that’s why Fauci famously said on 60 Minutes in middle of March, ‘You don’t need to wear one. You might be touching your face and doing something else that has a countervailing benefit.’


But in the July 18, 2022 episode of Econtalk, (around 1:09:00) Vinay Prasad and Russel Roberts claimed that Fauci had lied about the lack of efficacy of community masking.  Vinay Prasad on the Pandemic – Econlib (


So did Prasad and Roberts lie to us in 2022?  Or were they just mistaken?


A lie is when we present something as true when we know it is false. In order to determine whether someone is lied or was just mistaken, we must know what is in their mind. This usually impossible but except perhaps when someone contradicts themselves.  And this might have appeared to have been the case with Fauci in March 2020 because he later changed his recommendation on masking to align with the CDC’s determination.  But, his earlier statement was not a recommendation for all time: he prefaced by saying “…at this time…”, indicating that recommendations could change as conditions and information changes, which they did.


To assert that someone lied without evidence of them knowingly telling a falsehood is ad hominem, arrogant and misleading.  It is also akin to some of  Saul Alinsky’s  “Rules for Radicals”. According to Alinsky, rallying the masses against an abstract policy, institution or idea is difficult.  So he recommended focusing on an individual (or scapegoat) for ridicule and demonization.  That seems to be going on with Fauci.


When someone makes a mistake, as we all do (and as Roberts and Prasad did in accusing Fauci of lying when he was presenting the scientific consensus on the effectiveness of community masking in early 2020), I think it more responsible to say people were mistaken rather than to claim they were intentionally lying.   If you claimed someone lied, and it turns out they were not lying, you may have some egg on your face to apologize for.  But there is always the option of ignoring the episode and pretending it never happened.


required, not displayed
required, not displayed

This site uses Akismet to reduce spam. Learn how your comment data is processed.


Watch this podcast episode on YouTube:

This week's guest:

This week's focus:

Additional ideas and people mentioned in this podcast episode:

A few more readings and background resources:

A few more EconTalk podcast episodes:

More related EconTalk podcast episodes, by Category:

* As an Amazon Associate, Econlib earns from qualifying purchases.

TimePodcast Episode Highlights

Intro. [Recording date: May 8, 2024.]

Russ Roberts: Today is May 8, 2024, and my guest is oncologist and Professor of Epidemiology, Vinay Prasad, of the University of California, San Francisco.

This is Vinay's fifth appearance on EconTalk. He was last here in August of 2023 talking about cancer screening. Both of his 2023 appearances made the top 10 of your favorite episodes of last year. Vinay, welcome back to EconTalk.

Vinay Prasad: Russ, thank you for having me, and it's such a treat to hear that, that the listeners liked those episodes.

Russ Roberts: Yeah, they were awesome.


Russ Roberts: Our topic for today is the COVID vaccine, and in particular recent paper of yours in the Monash Bioethical Review co-written with Alyson Haslam that we will link to. The title of the paper is "COVID-19 Vaccines: History of the Pandemic's Great Scientific Success and Flawed Policy Implementation."

Before we dive in, I want to observe that it remains surreal to me that the very issue of vaccines has somehow become a huge area of discussion, a lightning rod. Much of it seems to be, if not fact-free, cherry-picked by lots of average people--not experts--but some experts, too. Somehow it's become such a big issue. Does this seem strange to you? Could you have imagined this drama in advance of how it has played out? If someone said to you, 'Well, you know, it's going to be really controversial and people are going to lie and exaggerate and ignore key facts,' is that strange to you?

Vinay Prasad: It is strange to me, Russ. You know, there's something about vaccines that has captured people's attention and interest and perhaps even ire for--in the last 30, 40 years.

There's long been a movement, I think, in America and globally skeptical of vaccines, particularly childhood immunization, and maybe that has something to do with the fact that it's injected. If they were all tablets, that might have been different, you know. Maybe it has something to do with the fact of the age with which these vaccines are typically administered. For whatever reason, I think it has captured some small group of people's interest for a long time.

But, the COVID-19 pandemic, I think, naturally expanded that group. I mean, naturally, we were making policy decisions and scientific decisions--which we are going to talk about today--some good, some not so good; some self-inflicted wounds, I think, by policymakers. Some mistakes were made. And I think that naturally was just pouring kerosene on something that had been smoldering for a while, which is a skeptical anti-vaccine sentiment.

But, you're totally right. There's so many things we do in medicine that don't capture people's interest, like giving kids Tylenol or acetominophen, giving kids ibuprofen or something--that when they are sick--or all the things we do, pediatrics. But for some reason, vaccines has always gotten people's attention.

Russ Roberts: I think you're right about the injection part. And, I think it's important to point out, or at least clarify--tell me if I have this right--traditionally, vaccines are an example of the dose makes the poison. Right? You get a little bit of the disease, and it mobilizes your body's natural immune system to fight the recurrence.

And yet, this vaccine--there are different variations of it for COVID--but the most important ones were not of that nature. Correct? And therefore, you weren't giving yourself a dose of the disease. You were doing something else, it turned out, that may have been challenging. But it didn't have that emotional, I think, unintuitive, for maybe people, idea that, 'Hey, give yourself some disease for your own good.'

Vinay Prasad: Yeah. It's a really interesting observation. I think you're right that, you know, as far back as the days of Edward Jenner, vaccines have always been either a weakened version of the virus, a weakened cousin of the virus, a weakened viral strain or a lower dose--something to expose you to what was potentially harmful without the harm, so that you would build immunity so that when you came across it in everyday life you would have some resilience against it.

This vaccine, in a number of ways, is different, of course. I mean we'll talk about--there's a couple different platforms. Johnson & Johnson and AstraZeneca, which actually was withdrawn from global market yesterday, are both adenoviral vector vaccines using DNA [Deoxyribonucleic acid].

And then the novel--the most novel--formulations were the mRNA vaccines of Moderna and Pfizer, and those are probably the ones that have maybe created the most, I think, discussion, perhaps because they are new. But, also, there are some concerning safety signals across all these products. Oh, and I used that jargon again, so maybe you are going to want to unpack that.

Russ Roberts: Right. No: You unpack it.

Vinay Prasad: Safety signal. Russ and I were just talking before, and he said, 'In your guys' line of work, when you say safety signal, what you mean is there's something concerning there. It's a harm.' And actually, yes, that's what we mean.

I think typically, we refer to these things as safety signals when you discover some untoward adverse event, something you didn't expect, you didn't want linked to these vaccines.

And those really do play an important role in policy. And I think the big policy blunders around COVID-19 vaccination were not taking those signals serious enough, fast enough, and not acting upon them. So, we can get into that.


Russ Roberts: Let's start with the opening line of your paper, quote,

The COVID-19 vaccine has been a miraculous, life-saving advance, offering staggering efficacy in adults, and was developed with astonishing speed.

That is the good news, and it is certainly is an opening--close quote; before that, before I said, 'That's the good news.'

That certainly is an answer to anyone who might say, 'Well, Vinay Prasad is anti-vaccine.' You're not anti-vaccine. And so, talk about what was impressive and good about this rollout.

Vinay Prasad: Yeah. And I think that this is something that--I frame it intentionally this way because I think it's the truth. I think this is something that people don't fully admit. One, President Donald Trump launched Operation Warp Speed, which was a sort of economic [?incentive?] program to expedite these vaccines to market. And what the government did was offered to essentially incur the risk of all the pharmaceutical firms: 'We will aid you. We will provide guidance, and we will sustain any losses you take if these vaccines are unsuccessful.'

That coordination, I think, was remarkably successful. Even the optimists weren't bullish enough at the speed with which the vaccine could be developed and deployed. We had a positive press release for the Pfizer and Moderna vaccines in November of 2021. And keep in mind: In January, we were barely starting to sequence the virus itself. So, all within one calendar year. I think that's a remarkable story about vaccine development.

And the next point I want to make is--and this is something that I think is really sort of clear from the evidence--is that if you were a 70-year-old person, an 80-year-old person, a 60-year-old person with medical problems, if you had not already had COVID-19 and you got that vaccine in January of 2021 where millions of people globally did, that vaccine remarkably lowered your risk of severe disease and death from COVID-19.

And we know that because the randomized controlled trials [RCTs], particularly the one by Moderna, showed a huge reduction in severe disease from COVID-19 from having received the vaccine.

So, I think there's some extreme skeptics out there who say, 'The vaccine did more harm than good in all age groups. The vaccine was a scientific bungle.'

It saved many lives. Some of the estimates, I think, are maybe misleading. Maybe there are estimates that are kind of speculation.

But, there's no doubt about it: If you were an older un-immune person and you got that vaccine in first quarter of 2021, you had a remarkable reduction in severe disease and death as a result. Certainly, randomized trials have shown the severe disease part, and some observational data has shown the death part. That's a very good thing.

And it did aid in ending the pandemic. By that I mean, COVID-19 will continue to spread. It didn't stop the spread. We can talk about that. There's some policy errors around the messaging around that. However, it remarkably reduced the death toll, and there are many people alive as a result of these vaccines who otherwise would not be had we not developed them.


Russ Roberts: We interviewed Gregory Zuckerman on his book about the creation of the vaccine, which we'll link to. I've talked about how--I think I was 65 when the vaccine came out, or maybe 66. I think 66. And of course, I was over 65. Which meant I had a higher risk of COVID! Which, it's one of those weird things where it's really not discrete. It's a continuous effect, and it also depended a lot on your general health. I was generally healthy. While, as listeners know, a little more heavy than I wish I otherwise were, I don't have the worst kind of comorbidities that COVID-19 wreaked havoc with.

Despite that, at that point, most of us were living in what felt like the Middle Ages and the bubonic plague. We felt danger all around us. I remember vividly going to the grocery store and putting my hands in Ziploc bags that I brought from home, touching on the handle of the cart, going through the groceries. I remember with horror when the cashier helped me bag the groceries with unwashed, unknown hands touching my cans, and probably washing them in the early days.

And when I got that first dose, as I've said here before, I felt like Superman. I felt like--other than kryptonite, which I didn't think was around--I had suddenly beaten the demon of COVID. I did occasionally remember I was still mortal. But I felt immortal there even with only one dose.

So, it was a remarkable and I think generally a wonderful thing. So, what's wrong? That's all great. What went wrong?

Vinay Prasad: Well, many things went wrong. I think that's all the good stuff, and I think it's important to acknowledge the good stuff as we get into the things that we did wrong. Because I think there are some people who are quick to dismiss criticism of the vaccine in any form as a blanket sort of anti-vaccine sentiment, as you articulated.

Now, what are the things we got wrong? I mean just a short list.

One, the rhetoric around whether or not the vaccine could halt transmission was incorrect and misleading.

Number Two, in the randomized control trials that led to vaccine approval, we could have explicitly tested whether or not the vaccine slowed or halted transmission. We did not do so.

Number Three, we started really well with older, un-immune people, but we quickly made the mistake of extrapolating our vaccine advice to people who had already had and recovered from COVID-19, an extrapolation that lacked evidence. An extrapolation that ran counter to a longstanding medical tradition of thinking that having had and recovered from a disease does confer some sort of durable immunity, particularly against the severe manifestations of that disease. So, we extrapolated that advice incorrectly.

Number Four, we pushed this in younger and younger populations, often using the brute force of the state and the schools to mandate this. Even into populations where the risk and benefit balance were more tenuous and uncertain.

What am I--Number Five, we started to discover safety signals. In other words, there were harms of this vaccine. We--the governments--were reluctant to admit those harms. They delayed interrogating those harms. And they failed to act upon those harms. And those harms did not fall equally across the age distribution. In many cases, those harms fell disproportionately among the people with the least to gain, the youngest people.

One of those harms is inflammation of the heart, myocarditis. That's a harm that particularly affects young men between the ages of 16 and 22. The risk of that was massive. The Israeli experience--which we can talk about--where the first estimates were 1 in 3,000, in the second dose of the mRNA vaccines, that risk of myocarditis far exceeded the potential benefit of those doses on severe disease.

So, in other words, that's a roundabout way of saying I'm fairly confident, for some populations, we actually started to inflict a net harm on these groups of people by recommending and mandating dose after dose after dose.

And then, the last thing is children. I think the evidence is quite weak, particularly children who have had COVID-19.

And then, finally, the benefits of an annual or biannual, perpetual vaccine policy, which is in the United States right now--we have essentially an at least once a year, possibly two times a year in certain age groups, vaccination for COVID-19 in perpetuity till the end of time. What's the evidence for that?

And so, this paper, which is a very lengthy policy paper, about 22 pages, talks about all of this stuff.

And then, one last point I want to make: It talks about all that stuff. But, the last point I want to make is we should talk about the initial vaccine approval. It came out after the election. Why did it come out after the election, not before the election, when Trump was saying it was going to come out before the election? Was that based on science or was that based on politics? And I think that's the first, most provocative part of this paper. And I think that's a very--your listeners might find that interesting.


Russ Roberts: Yeah. Let's go back to the first thing you mentioned, the transmission. It's a classic argument in economics. The technical term is externalities, the idea being that if you don't vaccinate, you're imposing costs on people who might have very high costs of getting vaccinated themselves relative to yourself; and therefore, even though it's your risk and your life, you might be exposing others to the dangers of the disease--others who cannot protect themselves or who the cost of protecting them is very high for health reasons.

And this argument was made over and over again, in the early days of the COVID-19 vaccine: That a mandate was, as you argue, the argument was it was ethical. It was justified. Because, you needed to help old people--and potentially children, although here in this case it wasn't true. But, the argument, say, in other vaccines would be: Well, you might think, well, it's just up to you. But, no, if you don't vaccinate, you're risking[?]--you're putting elderly and children at risk who--where it's very costly to vaccinate them, health-wise cost.

And, you argue that's not true. Was that true at the beginning, or just when the Omicrom variant came out?

Vinay Prasad: So, I think the Omicrom variant, which really sort of came out by the fourth quarter of 2021, clearly had vaccine escape--i.e., that no matter how many doses you'd gotten in the past, you could get Omicrom through that and you could spread Omicrom through that.

However, it was known as early as the summer of 2021 in the United States, from places like Provincetown, Massachusetts, where they had a sort of annual gathering of many men, 97% of whom had been vaccinated, but many of whom came down with COVID-19, that even some of the earlier strains could spread despite vaccination.

So in other words, I think we knew pretty early on that vaccinated people could get COVID-19 even after being vaccinated. That the vaccine would not be able to halt transmission.

But, I want to make one more point about the trial.

One of the points that I think we don't talk enough about is that all global economies are put on hold. I mean the pandemic has, as you say, it was like living in the plague. I mean, it has changed the lives of so many people. Governments are spending tens of trillions of dollars globally on pandemic economic relief and on the sequela of having diminished economic productivity. In other words, this is a huge global event.

We run the randomized control trial, and it would not have cost them much more and it would have been infinitely logical to do the following: In this Pfizer study, I think 40,000 people, we divide them in two groups. Half get the vaccine. Half don't. And we measure how many of those people feel like they have symptoms of COVID-19; we test them for COVID-19. That's the study design.

The study could have also been powered and designed to look at things--I think some of the critics are correct--to say, 'How come you didn't run a study in older people and actually look at all-cause mortality as an endpoint?' Could have done that. They could have run a different study in just over age of 80, and the endpoint could have been all-cause mortality.

And how come you didn't take your big study and basically say, 'Out of these 20,000 people in one arm, we're going to say we'll sample 5,000 people and test every one of their family members every week for COVID-19 to see does it actually slow the spread? Not only are they less likely to get COVID, are they less likely to spread it to their family members?' And you could have analyzed that in a randomized study. And it would have cost a little bit of money. But, in comparison to what we're spending on this, it's literally a drop in the ocean.

That, to me, is a missed maneuver.

And, I worry that, you know--that there's a reason why the company doesn't want to do that, obviously, because there's a very overwhelming chance that it doesn't do that and that would sort of undermine one of the claims of the vaccine. But, to me, it's very concerning that policymakers, like at USFDA [U.S. Food and Drug Administration], at CDC [Centers for Disease Control], and the White House, and EMA [Emergency Management Agency], that they didn't make the company generate those data. That's a huge missed opportunity. And, I think many of us said that at the time, and I really don't see much justification for not doing that kind of study.


Russ Roberts: And given that the government was subsidizing the creation of the vaccine by guaranteeing enormous purchases--which was probably a good idea--but they extracted nothing in return for that, which could have been incredibly inexpensive relative to the profitability of the product.

Two things: First, this issue about transmission. The people who took the vaccine and then still were able to spread the disease, the vaccine could still have been effective, you're saying. It didn't keep them from getting the disease. It kept them from getting a more dangerous version of the disease. Their symptoms were reduced even though they still had it. Therefore, they could pass it on to others very easily. Is that the key distinction there?

Vinay Prasad: Absolutely. That, in order to benefit you, all I need the vaccine to do is reduce the likelihood that I'm going to die of COVID-19, or have the breathing tube put down my throat from COVID-19. Which I think these vaccines do, particularly--well, I think they do that in people who have not already had COVID-19.

Okay. The next thing is: Does it work so well that it actually prevents me from ever even catching a milder version of this? The answer to that is: No, it doesn't.

That has different policy implications. This has implications for passports, for mandates, but not for the personal health question.

So, I think two things are true, that: The 80-year-old who rushed to get it was making a wise choice. But, compelling that the 20-year-old to get it to protect the 80-year-old was an unwise policy decision and has huge ramifications for trust in public health and for trust in vaccines.

And I'll just make one more nuanced point to the listener, which is that: one thing that was also unclear was that if the 80-year-old gets the shot and we compel the 20-year-old to get the shot on top of that, does that further reduce the 80-year-old's likelihood of death beyond the benefit they're getting from the personal receipt of the vaccine?

And, that was something that policymakers did not do a great job of studying.

Sort of this idea that--I think they had this fantastical idea that we could literally extinguish this virus and sort of make it get extinct by vaccinating everybody. That had sort of no basis, and also runs counter to what we know about coronavirus, which is that, prior to this, there were four circulating endemic strains of coronavirus, and we all get them over and over again in our lives. It's not the kind of thing you get once and you never get again. It mutates. It changes. And then, you get it again a few years later. Which is what's going to happen with COVID-19. We're going to keep getting it again and again every few years as long as humanity persists, I think.


Russ Roberts: Well, you mentioned the 20-year-old. I think what was even worse was smaller children--the closing of schools--because the idea would be--the argument was: we'll keep the kids out of school--that, the nine-year-old--because that way they won't be able to infect their grandparents when they go to visit. Do you think that was a mistake?

Vinay Prasad: Oh, absolutely. School closure was one of the greatest policy blunders--particularly in the United States--domestic policy blunders I think of the last quarter century.

Let's just put it in perspective. Only one nation, to my knowledge, was brave enough to keep schools open the entire time. That was Sweden, and they did that predominantly in kids under the age of 12. They never closed. Most of Western Europe, I think, was pretty good about it. They closed for six weeks to 12 weeks, something in that ballpark, places like Germany and Spain and Switzerland. But then, they reopened quickly by fall of 2020.

The United States, to my knowledge, is one place where--schools were closed in San Francisco from March of 2020 until the fall of 2021. We're talking about 18 months of school closure in urban enclaves, like Los Angeles and Washington, D.C., and San Francisco. This was a catastrophic policy decision. We see it through all the test score data: huge decreases in learning among these kids, often poor, minority kids who potentially have the most to lose from having school closure. And, the justification was that the school closure protected Grandma and Grandpa.

We've had a couple of really nice economic analyses on this topic, including one from the University of Southern California [USC] researchers in the policy school there; and including some from Germany, using a natural experiment of staggered school closures in the summer of 2020. The economists who have analyzed this rigorously find--some studies find no decrease in community transmission from closing schools. That's the German study. And some find a very, very marginal difference in transmission that's unlikely to have made a big impact in COVID-19 dynamics. That's the USC conclusion.

But, I think most people believe that it didn't accomplish the stated aims. It didn't slow transmission to Grandma, Grandpa, or it did so only very marginally. The kids had nothing to worry about anyway because most of them in the United States and elsewhere globally got COVID-19 without having gotten vaccination eventually, and most of them did very well. This is, thankfully, a disease that does not kill children the way influenza does. And, the loss to children was catastrophic in terms of learning losses and grade losses. And, economists will know that will have spillover effects on livelihood, employment, and life span.

That's the thing that I think the epidemiologists just did not understand. They were so singularly focused on COVID-19, they didn't understand that not going to school and losing two years of learning loss can shorten someone's life. That, to me, was I think one of the big blunders.

Russ Roberts: And some might be tempted to say, 'Well, but it's worth it to save some of the elderly or some of the grandmas and grandpas.' Being a grandfather now, which I was not in the time of COVID, I would happily accept the risk of a higher chance of COVID in return for my granddaughter having a better life. I think it's very hard for people to think in those terms, those kind of trade-offs. Maybe not every grandparent would have felt that way. I think many would have. Maybe most, if not all. And I wonder if the next time we have something like this whether that will be the response if we encounter a disease with this incredibly strong differential impact on populations based on age.


Russ Roberts: When you talk about the trial of Pfizer of the 40,000, you said we could have made them do a more richer trial analysis and find out what the cost-benefits were for different age groups. Was that what you said?

Vinay Prasad: That's one point, and we could have tested transmission directly. So, we would known both things.

Russ Roberts: Right. Test that directly. Now, it's true that you have come out and said that ex post. Ex ante, at the time, was there a loud outcry for that policy? And I ask because it's tempting to say that Pfizer's political influence insulated them from that request. It could have been just an error on the part of policymakers. It could have been a disagreement about what was important. There could be a lot of reasons. At the time, were people demanding it, in the industry--in medicine?

Vinay Prasad: Yeah. That's a great question.

Russ Roberts: In epidemiology?

Vinay Prasad: So, I think it's a fair point, and often people say that: 'A lot of your criticism is ex post, but what did you say at the time?' And for a while, on my Twitter page, I had a pinned post of every op-ed, video, and podcast I did in real time to show that many of these were my policy positions at the time, not just after the fact.

To this particular question, there was a New York Times op-ed written by Peter Doshi, and I believe it came out in the fall of 2020 prior to the vaccine announcement, and it specifically said, 'We ought to design and power these studies for severe disease, not merely having symptomatic COVID-19.'

And then, privately I will tell you that many of us were discussing this. So, I think that that was something that was said.

Now, what was said on the other side? Look, there are many things that we could have done differently in the original studies. One is: Did we really need a study from 18 to 100? Could we have done separate studies, like studies in--randomized trial in nursing homes powered for mortality; randomized trial from 60 to 80 powered for severe disease; a randomized trial in 40 to 60, looking at transmission and looking at symptomatic COVID-19; and then the last randomized trial from 18 to 40, looking at symptomatic transmission, severe disease, and death.

But, here's one distinction. When you start to break it out by age groups in multiple studies, studies can halt at different times. We have one study and the moment it halts, everybody on the control arm gets the vaccine. So, essentially, detecting safety signals is a wash. Detecting whether or not the vaccine starts to lose efficacy with time is lost because everyone's getting it.

But, if you have different cohorts from 18 to 40, you can run that randomized trial for six months or eight months or a year. That vaccine does not need to be made available as soon as 80-year-olds. In fact, there's a supply constraint anyway.

Even if you were to approve it, you don't have enough vaccine for 18-year-olds and 80-year-olds. You're going to get the 80-year-olds first. So, why not design these studies in a way we get more and more information, more and more safety information?

Now, it's easy to say that this is all in retrospect, but I think these are principles of medicine that are longstanding, and people knew that at the time.

The cynical answer is that the company, of course, has very little incentive to do these studies. The longer you run the study without crossing people over, the more likely it is that people start to find that more and more safety concerns are linked to your product.

And then there's also, I think, the good intention, which is that we're in the middle of a global pandemic. People want a one-size-fits-all. We're all concerned. We're panicking. We just want a one-size-fits-all solution. A single vaccine study, 18 and up--let's just get it done and just vaccinate everyone, and let the chips fall where they may. I think that's also part of the calculus.

I would say that one of the things that my essay did was talk about attitudes about vaccine. And I think that's very interesting, and I'll make a brief mention.

I went through many, many news stories, and I documented what were the news outlets saying about vaccination by calendar year. And, there's a table of quotes in 2020, and the quote said--and you have to remember the context. Trump is on television every day, and he's saying things like, 'The vaccine's going to be marvelous. You're going to get it. It's going to be wonderful. It's going to come out before the election. Everything will go away.' These kinds of very tall claims. He's running in an election year. He's saying these things.

Many people are writing op-eds saying, 'Be careful. Just because a vaccine is approved doesn't mean it's going to work. Be careful. Even if it increases antibodies, it doesn't mean it actually reduces the risk of dying. Be careful. Some vaccines in the past have actually increased the risk of getting the virus. Be careful. Vaccines can have safety signals that take years to be known. So, maybe we shouldn't rush into this.' Scientists are just publishing a steady stream of this.

The moment Biden wins the election, the same scientists, they switch 180. And then they say, 'The vaccines, they came out. Nothing to worry about. Safety signal is pristine. I've never seen data this good.' I've highlighted these quotes in my essay. I worry that some of it is politically motivated. And that troubles me.

And the final thing I'll say is: Pew [Pew Research Center] has public opinion polling on vaccination from May of 2020 till the election. The public opinions on vaccines cratered. I mean, this negative rhetoric in the media space in 2020, the election year, absolutely destroyed Americans' desire to take a COVID-19 vaccine. That's been documented by Pew.

And then, 2021 when Biden is in office, was an attempt to, I think, resurrect that and try to get people to take what they had already spent a whole year dissuading you of.

And then, the last thing I want to say, Russ, about the timing of the approval. Trump, many times, said that this vaccine will be available before the election. And, I don't think he said that because of his usual bombastic nature. I think he said that because he was told that that was likely to be the case.

And, let me explain why he felt that was likely to be the case and what actually happened.

When you run a randomized control trial--any randomized control trial that's blinded, where you don't know who got the vaccine and who got the placebo--you have to make decisions on: When do you look at the data the first time? Basically, what you do is you randomize people to vaccine or placebo. The primary endpoint of the study is symptomatic SARS-CoV-2, meaning people have to say, 'I have a runny nose. I have a sore throat.' And then we swab them. Then we mail that swab to Pfizer, who is running genetic testing on it. And if they find COVID-19, that's scored as an endpoint: One, Two, Three.

So, as the trial is running, people say, 'I feel sick. I feel sick.' They swab them, swab them, swab them. And Pfizer is running this, like, 'Oh, we have one positive, two positive, three positive.' But, they're blinded. They don't know which arm of the study they are.

But, once they have 32 positive results, they have to take--this is the original statistical plan called for--at 32 positive results, we're going to look at the data the first time. We'll look. And if the ratio of the positive results is 26 in one arm and six in the other arm, it's a winner. If it's 27-5, 28-and-4, it's a success.

If it's anything less than 26-6--if it's 25-7--trial keeps going. Right? So, this is called the stopping boundary. An extremely skewed imbalance in those swabs could stop the study at 32 events. That's the original statistical plan.

And, Pfizer knows how many swabs are coming in each week and how many are testing positive. And Trump knows that, too. So, I think Trump was very confident it'll come out in October because he knows--Pfizer tells him: 'With the rate with which these are testing positive, we'll hit 32 by mid-October.' I think that's what Pfizer is telling the Trump team privately. And that's why he's so bullish about it coming out in October.

Now, here's where it gets interesting. There's a huge movement among academic doctors, epidemiologists, and the press, and everybody who doesn't like Trump--I mean, that's everybody in--most of the elites, I think. They didn't want him to have it. They didn't want him to have it. So, they tried many ways to delay that first look. One was asking for a minimum of two months' safety data rather than a median[?] of two months' safety data. This was a proposal. That didn't make a lot of sense, scientifically.

Ultimately, under extreme pressure, Pfizer altered the statistical plan of the study and changed the 32-look to 64-look. They made a protocol amendment to change the first look and increase the number of events. The stated reason is that this would increase the statistical confidence that the result is valid.

But, that's incorrect, because at 64 events, less of a skew is required to trigger the stopping rule. One can imagine at 1,000 events, if it's 550 and 450, we stop. But, at 32 events, you need a huge skew, 26-6, you know what I mean? A bigger difference.

They changed the statistical plan in October. And, they did not do so, in my opinion, for any credible scientific reason. It was only to delay the time until they got the report and could stop the study.

And then, there is some reporting in STAT that Bourla himself suggested that they slow down on testing those swabs coming in to further kind of delay it. And, coincidentally, they took the first look at something like 90-some events which occurred right after the election, and then they put out their press release a few days after the election.

So, the argument that I made in the essay--which I think some will find controversial, but I think is true--is that there was no medical or scientific reason to not look at 32 events. To change the protocol. Changing protocol always, I think, is a problem in statistical plans. It's better to stick with what you started than to keep altering it. It was done, I think, almost entirely politically motivated. And, I think it actually killed people because that delay of even a few weeks in a peak of Delta wave will result in a delay of immunization at the most critical time.

And so I think it was--this is something that few people recognize, but I think historians will recognize that as an incredible, incredible policy plunder.


Russ Roberts: Very interesting. I think there are people in the world who think that public health officials are merely scientists looking for the truth. There are people who think that--no one would actually be so naive to think that the head of, say, a pharmaceutical company--you mentioned the head of Pfizer. His name is?

Vinay Prasad: Albert Bourla.

Russ Roberts: Bourla. So, that was the reference you made. I think most people understand that there's a conflict of interest there. The head of, the CEO [Chief Executive Officer] of a pharmaceutical company in the American system has a profit motive at stake, which is both what is fabulous about our system and not always fabulous, the American system.

And yet, somehow we have this romance about a doctor in a government role. Really, what you're suggesting is that they, too, respond to incentives.

I like to think that if there is another pandemic in our near future and these lessons are remembered, that a serious scientist in a leadership role in the public health system would remember the kind of lessons you're talking about and respond differently. But maybe not.

Vinay Prasad: Yeah. I think, while it's nice to think of public health officials and epidemiologists as impartial, I think many studies show that they're extremely Left-leaning. And then, while it's nice to think of science as separate from politics, I think science and politics have always been intertwined.

And then, I think the final point is that Trump is a uniquely triggering figure. I think we underestimate, but I mean, I work at University of California, San Francisco. Suffice it to say that the faculty here are not the biggest Trump supporters. Throughout the Trump Administration, I think we routinely saw faculty--he really knows how to get under people's skin. In part, that might be his appeal, to his own base: is that he angers the people that they don't like. But, I think that, to me, it's absolutely incredible that you could let this guy rattle you and get these policy decisions wrong.

I think there's two examples. One is the schools. There's very good public opinion polling data in the United States that at the end of June, many Americans--the American Association of Pediatrics, the teachers' unions--were all calling for school reopening.

Trump famously gives a speech at the end of June or early July of 2020 where he says, 'In the fall, all the schools have to reopen.' The same public opinion data shows a huge swing in sentiment about school reopening. The AAP [American Academy of Pediatrics? Or:[?] Affirmative Action Plan?] revises their position. They actually go against children.

I think it's naive to think that these organizations did not also suffer from, I think, what some call Trump Derangement Syndrome--where you literally say the sun rises in the West just because he said it rises in the East. You're so triggered by him, you can't focus on what matters.

I think school closure in the United States, many data show, it was linked to whether or not it had a strong teachers' union, whether or not the city was Democratically controlled.

But it was not linked to the spread of COVID-19, the cases of COVID-19, and hospitalizations of COVID-19. And Republican districts reopened much sooner, to their credit.

That, to me, is a problem. I say that as somebody who, as I've said before on this podcast, who considers himself a Progessive and progressive ideals at heart: which is that, as a Progressive, I believe that championing poor minority children and their education and their health is the Number One value of Progressives, even ahead of--and I think most cultures value children ahead of adults and elderly. And to give that up because Trump said it was literally cutting your nose to spite your face. I think that was terrible.


Russ Roberts: Well, good thing he's out of--oh, I forgot: he's running again. And we'll have the same kind of challenges, I think, in policy. You know, I've said this before on the program, maybe even to you: I have vivid memories of him coming out onstage in the early days of the pandemic, unmasked, standing cheek-to-jowl with eight other members of the Administration. And, by the way, they don't look happy. They're not keeping six feet apart from each other. They're grim. And you could say they're grim because they were in a pandemic. But, I think part of the reason they're grim is that they've got to sit there and listen to their boss, and he's doing it in a way--he's speaking unmasked and close to everybody.

And, I believe--maybe there's even evidence for this--but I certainly believe that mask attitudes were driven by whether you liked or did not like Mr. Trump.

In particular--help me out here. In California, there are still people wearing masks.

Vinay Prasad: Oh, yeah--

Russ Roberts: Who are now[not?] protesting and trying to maintain anonymity in an anti-Israel protest. It's just a social signal: 'I'm worried and I'm showing you that I care about you, so I'm wearing a mask.' Which means you cannot look at my face very well. You cannot read my mouth and connect to me as another human being. I found that cost of the pandemic to be non-trivial in our human relations with each other.

And yet, it became an incredibly intense fight, again, akin to vaccines, over what should have been a mere factual matter.

Vinay Prasad: Yeah, absolutely. I think that the more Trump didn't wear it, the more Liberals said, 'We got to wear it twice as hard.' And, organizations like the American Academy of Pediatrics were deranged, in my opinion, recommending it to two-year-olds and using the power of the state to compel it. That was Biden's Head Start program--compelling it in a three-year-old--where literally, as a really famous researcher told me once, he said, 'This is one space where you don't even need to know anything about the literature. You just need your eyes to know that making a two-year-old wear a cloth mask is the dumbest thing on Planet Earth.'

So, that's One.

And, Two: you're absolutely right. In liberal strongholds in the United States, there are, of course, you see it in the protesters. But, even outside of that, there are many people who still mask to this day in San Francisco because they are maybe personally worried. And I think that's also--and that's not the case if you go to Texas, for instance--rural Texas, you don't see any of that. So, it is balling[?] by political lines.


Russ Roberts: But, since you and I both care about the truth and have no political biases whatsoever, either of us: What I was trying to hint at there--not hint at, but imply--is that I was very pro-mask in the very opening days of the pandemic. I thought it was a low-cost, prudent thing to do is to wear a mask, which is why I was upset that Trump did not wear one. But, as the pandemic wore on and as it was clear that there was this human cost of interaction--especially for children, I found it incredibly painful that children would be looking both at each other, at other children and at their parents and grandparents without fully seeing their faces.

Was that an evidence-based feeling that that was wrong? I mean what have we learned about masks that's true? I really don't know. Because, many would argue, and I'm going to defend the Left-leaning folks we're both talking about--who I'm accusing them of being virtue signalers--isn't it better safe than sorry?

Vinay Prasad: Yeah. I guess I have a lot to say there. I mean what does the evidence show? Well, we could start with the pre-pandemic evidence was that generally community--the one thing to draw a distinction is that masks, it's many different interventions. Mask-wearing is like the ICU [intensive care unit] doctor going into the COVID-19 patient's room: should he wear an N95? That's one question.

The question I think most of us care about is the community question, which is: Should people in the community be advised or mandated to wear masks? Should those masks be cloth masks, surgical masks, or N95 masks? And how low do you go in terms of age? Do you take that down to, say, eight-year-olds or six-year-olds or two-year-olds, as in the United States? I think those, to me, are the more interesting policy questions.

The evidence pre-pandemic was pretty conclusive that randomized studies of community masking were generally negative and did not appear to work, and that's why Fauci famously said on 60 Minutes in middle of March, 'You don't need to wear one. You might be touching your face and doing something else that has a countervailing benefit.'

Then, of course, I think there was an activist campaign in the United States to recommend it. And I actually didn't disagree with them too much because I think you could say, 'Listen, out of an abundance of caution, out of prudence, we don't know. Just let's do this. Let's try.' What I do think is also another evidence/generation failure that governments like the United States and all of Europe--with one exception, DanMask--we didn't run any studies to actually figure it out. We didn't run cluster randomized trials. We didn't figure out if we actually got the policy right.

And then, at some point very quickly, by the summer of 2020, it was so politically polarized that I think nobody wanted to generate the evidence. If you generated that evidence in a well-done study, one side is going to be right and one side's going to be wrong. And that's going to have tremendous political implications. And I think it was actually thwarted. I know personally--like, from, anecdotally--that there were some discussion of trying to do it in places like United Kingdom, but there were some staunch people who thought they already knew the answer who didn't want those studies to be conducted.

I certainly think that--right now, I think my personal opinion is that when I see people in the store wearing it, when I see somebody on an airplane making an 18-month-old baby wear a flow mask--which I saw recently on an airplane--I think that's crazy. And I think that--like, forcing even your own child to wear a mask in 2024--I think those parents need someone to talk to them, at a minimum, and tell them that that doesn't make a lot of sense. And in many ways, I think it's cruel.

And you would have said that to somebody in 2019 if you see them masking their kid like that in an airplane. You would have said, 'What are you doing to your child? This is a kid, man.'

But now, somehow, we don't go there. So, I think we've talked about masks before on a prior episode. But, yeah. Now we have a new Cochrane Review. And, again, the new Cochrane Review is pretty negative. There were only three randomized trials run globally during the pandemic: Guinea-Bissau, Bangladesh, and Denmark. They all have limitations. And, I don't think they really move the needle forward on the evidence-based that much. I mean I think we're left in the same place. And that's what the meta-analysis suggests, too.

Russ Roberts: When I see people out in public now wearing the masks--which, by the way, I almost never see here in Israel and see in some numbers in the United States when I'm traveling there--I try to give those folks the benefit of the doubt. I assume they have perhaps some serious condition that this gives them comfort.

But, the other thing I want to say is that when the disease first started, we were first aware of it, we weren't sure much about how it was transmitted. We weren't sure much about it.

And I remember--you alluded to--I remember people saying, 'Just don't touch your face, and you'll really reduce your risk of the disease.' And, Fauci had said--and we've talked about this before on the program--that I think he was trying to reserve masks for healthcare officials, so he was discouraging public use at the time. I wish he had, instead, encouraged prices to solve that problem.

But, put that to the side. People said: 'Just don't touch your face.' And, I started looking at how often people touch their face. I'm touching my face now. It's almost impossible for a human being to not touch their face. It's just an interesting thing that I learned from the pandemic.

But, the point is that we learned fairly quickly that it was airborne.

So, why wouldn't masks help a little bit? What was the reason that we now are so skeptical? Just on common sense grounds? Randomized controlled trials are difficult. People do other things. They don't wear them all the time. They don't wear them correctly. Et cetera. Given that it's airborne, why wouldn't caution be a bit more effective there?

Vinay Prasad: I mean, I think that the mechanistic reasons are, You know, like: Do I think it's possible that it has a small benefit? Sure. But, maybe it gets washed away with everyday life.

For instance, yanno: If we go in a room for 30 seconds each, and you have COVID and I don't have COVID--we could have done these studies, too, these challenge studies, which people didn't do that many of. But, we could have put two people in a room for 30 seconds and neither of us are masked. I think there's a certain probability I get COVID-19, and if I wear a cloth mask and you wear a cloth mask, I don't know, maybe it goes down a little bit.

But, now let's change it from 30 seconds to one minute to two minutes to eight hours. You know: That's when I think you start to get into the real world. So, if two people are working in an office side-by-side for eight hours in the same stagnant pool of air, does the cloth mask still have a benefit?

And at some point, you've got to think that, 'Okay, all the particles are circulating for the last two hours here.' And eventually, some of it is slipping through your mask. Even if you think masks have benefits. I mean, that's one mechanistic reason.

The other mechanistic reason is if we're working in the office for eight hours together, eventually I'm going to take a sip of coffee. You're going to have a drink. It's going to slip down. Or the mask becomes soggy or it's not filtering as much as you think it filters, and all these sorts of mechanistic reasons.

But, ultimately, I think the beauty of, sort of, the bigger studies is that it accounts for all those things: how people use it, behavioral compliance, time I n the room.

You know, one of the things about the pediatric requirement in the United States was the kids in daycare had to wear the cloth mask--these kids who are learning facial cues and talking, except for the two hours where they all lay on mats next to each other napping in the same room. Then they have to take it off because it's a suffocation risk which, to me, is who is setting that policy? You have to literally be the dumbest person I've ever. And the CDC set that policy. Sometimes they have the gall to say, 'Who are these people questioning us?' And, 'Why don't people trust us?' You really set a policy that--

You can imagine a high school dropout from Mississippi would think that, 'Why are you making my daughter mask? That's pretty stupid. I mean, like, they're napping.' The CDC [Center for Disease Control] says we have to do it.

How will people judge you if you set such a clearly--I call these policies trust shredders. Like, if you set that policy that's so counter to just basic common sense, you will shred all the trust you have. And, you know, I'm not going to blame people for not trusting you again.

Russ Roberts: And I assume you have nothing against Mississippi. You only used that example because their education and test scores are not the highest? Why did you pick Mississippi?

Vinay Prasad: I have nothing against Mississippi. But, yeah, I mean just a place where I think they have a lot of college dropouts. I'm from Indiana. Listen, Indiana's got a lot of high school dropouts, too.


Russ Roberts: Let's talk about a new study that came out on side-effects of the vaccine, which is sort of the flip side of some of this. There's an efficacy question, and there's a question about the dangers from the vaccine itself.

So, this study that we're going to talk about had a pretty large sample.

Vinay Prasad: 99 million.

Russ Roberts: 99 million. Like, that's a dream come true. If there's anything bad about these vaccines, surely we're going to find it.

And, what was the bottom line of that study and why are you skeptical of it?

Vinay Prasad: Well, I should say that I want to commend the researchers for doing that. I think it's an important step. I did learn a lot of things by reading that study. This study comes out in the journal, Vaccine. It's a very large analysis. Its basic method is simple. They look at the rates of certain conditions, I think something like 30 or 40 things they think might be linked to vaccine in the time period before vaccination. And then, they look at it after people get vaccinated to see if there's an increase in that after vaccination, comparing it to sort of a baseline rate of these events.

And what they found was, I think, lots and lots of potentially concerning safety signals. Of course, the things we know about: the heart inflammation or myocarditis. They found that signal. They found blood clots, which are linked to Johnson & Johnson and AstraZeneca--the adenoviral vector vaccines. They found things that people were a little bit concerned might be linked, but we didn't have really a lot of proof that they were linked--like, low platelets or immune thrombocytopenia purpura, low blood counts of one of the clotting factors.

And then, they found a whole bunch of other things that were linked, including, like, a demyelinating disease that affects kids, febrile seizures, a whole bunch of other concerning safety signals.

Now, some of the points I want to make are: Number One, the authors think their method is not biased by reporting. I think it is biased by reporting. It's an under-estimate.

And let me explain why. Sample size is brilliant. It's a huge sample. But--and I think they are pretty good about knowing whether or not somebody got vaccinated and when they got vaccinated. I think that's a pretty rock solid data point.

But, one of the data points that I think is weaker is the event itself. You know, if you come in with low platelets or you come in with abdominal pain and that's due to a blood clot in the abdomen, it takes a lot before that's coded in the chart as mesenteric artery thrombus or a clot in that artery. What does it take? It takes the doctor being, like, 'Oh, boy, maybe something's going on here.' The doctor has to order the CT [computed tomography] scan. CT scan has to show the result. The result has to be coded in the chart and on an ICD-10 [International Classification of Diseases, Tenth Revision] coder, an electronic medical record code.

There's lots of reasons to worry that that wasn't really done so well. You know, a young person comes in with belly pain. It's easy to say, 'We looked for some basic things, but maybe it's just minor. It'll feel better in a day.' You might miss a blood clot that was there, for instance; and that doesn't get coded in the chart.

One can start to think about all the other things. If it affects populations you're normally not looking for, those kind of safety signals in, you might not be diagnosing it. You might not be coding it correctly. I think people who do electronic medical record research know that these datasets are often full of missing data, and they're really not robust.

So, I say all that to say, that: I worry that, in some cases, that the safety signal--in other words, that the risk of these products--is actually much bigger than what these researchers found. I think that's the direction I would worry more in.

And then, the final point I'd say, the reason this whole paper gets me interested and concerned is: I think public health has to admit that there were some populations that were harmed as a result of vaccination. For instance, like, pushing AstraZeneca vaccines in 18-year-old women in 2021, which was done across Europe, resulted in some women getting cerebral vein thrombus in the brain and having brain herniation and dying and some women getting blood clots.

To me, it's hard to believe. Like, even a little bit of safety signal in an 18-year old, healthy 18-year-old, may offset any potential benefit of that AstraZeneca vaccine in that 18-year-old because the risk of COVID--and bad outcome from COVID was so low in that 18-year-old already.

And, we have proven mathematically that for a man who is, like, 20 years old, dose two is harmful. I mean he's going to get COVID either way. He's going to have some risk of COVID outcomes either way. But the risk of myocarditis is so much worse than the risk of bad COVID outcomes that you should not be giving him more doses.

And, I think people who had had COVID and recovered from COVID, many of them were harmed by getting vaccines. What I mean here is mathematically, in aggregate, there was a net harm to that group. Not that there are anecdotal stories of harm--which could be acceptable if there was a net benefit to the whole group.

Now, why is this important? I think this is one thing that public health refuses to admit. They refuse to admit that any of our vaccine policies and mandates actually harmed a sub-population of people. They came up with elaborate studies that--they keep saying that, 'Well, COVID-19 is worse than the vaccine.' Those studies are so catastrophically flawed, and they should know it.

And I just want--and maybe I'll talk about that for one second, why those studies are flawed.

Like, if you want to know how often does a vaccine cause myocarditis, we can look at the record and say: everyone who got a vaccine and how many people got myocarditis. You could do that kind of math.

If you want to know how often does COVID cause myocarditis, the question is: Who had COVID in the dataset? If you look at electronic health records in the United States, some people had COVID, but some people didn't have COVID. But, among the people who don't have COVID in the record, many of them actually had COVID because how does it capture COVID in that dataset?

In order to be in the dataset as COVID, you had to be so sick from COVID you came to the doctor. They swabbed you and put it in their computer.

But most of us either had it mild. We didn't test ourselves. Or we tested at home and threw away the test. And that's not in their dataset. So, the denominator is not all COVID-19 infections, but the worst of the worst COVID-19 infections--those that warrant medical care.

And then, the numerator is how many people get myocarditis or whatever from that.

And they compare that against the denominator of everyone who got the vaccine and how many get myocarditis. And to me, that's such a flawed comparison.

You obviously are missing the majority of COVID-19 that didn't do harmful things. Because those people didn't come to the hospital. Right?

And, so, to me, I think a lot of that literature is flawed, and it is clear to me that it was a net harm in young men, particularly.


Russ Roberts: So, this is a true story. It happened to me. I moved to Israel. Before I moved here, I got two Moderna shots. When I came here, everyone said, 'You should get a booster. You should get a third one.' And so, I went to my health clinic, and they didn't have Moderna. They had Pfizer. And that alarmed me greatly. So, while they were preparing me to be vaccinated, I'm desperately on my phone trying to find out if it's okay to get a booster of Pfizer on top of Moderna. I did it. I didn't die. I didn't die in the office, and I didn't die in the aftermath of the office. So, I've had, at that point, three COVID-19 vaccines.

So, I'm at a party, and someone asks me if I've been boosted. I said, 'Yeah.' And they said, 'How many times?' I said, 'Well, I've had three shots.' And then they said, 'Are you going to get another one? Because a lot of people said you should get one every once in a while?' And I said, 'No, I'm not.' And he said, 'Why not?' And I said, 'It just seems like we don't know a lot about what this vaccine does in the long run to one's autoimmune system, and we're playing around with my body. And my body is a complicated, highly--it's a well-oiled machine, of course, and a tremendously impressive thing.' But, I said, 'At this point, COVID risk seems very small. The risk of dying from it seems very small.'

And he looked at me and he said, 'I think you're letting your political views get in the way of your common sense.'

And, I'm proud to say that--had it been 10 years before that, I think I would have gone berserk. I just looked at him and I just said, 'You know, you don't know me very well. Let's just leave it at that.'

That he thought that my disinterest in a booster was because I was pro-Trump, for example, as opposed to pro--anti-risk, anti-downside risk--was almost amusing. It took some effort to not get upset.

But, I'm curious what you think we know, if anything, about both long COVID, which very early on got scare stories in the paper--that people with COVID, their lungs were never the same and so on. And long impacts of the vaccine, especially people who are getting boosted over and over again. Do we know anything about either of those that's reliable?

Vinay Prasad: Yeah. That's a great question. I mean, I guess I'd say--for a second, I just want to say my medical philosophy, which is pre-COVID. Because I think it goes to your point, Russ. Which was that I think for those of us who practice medicine, we imagine there's a sort of continuum of people you see. We see some healthy people--and I'm a cancer doctor: I see people who are nearing the end of life with cancer that has relapsed many, many times. And, we have different appetites to try things, I think, based on the person in front of us.

If somebody comes into my office and they have multiple myeloma and they have progressed through four different prior therapies and their cancer is getting worse, and they say, 'Doctor, I'm getting tired every day. My blood counts are taking off. I mean this cancer is going to kill me. Can we try--' and then, you name the drug. We try some new drug that has an uncontrolled study of 20 people, and some people felt a little bit better. 'Can we try this new drug?' I think I would, and most cancer doctors would say, 'That's reasonable. We don't know everything about the drug. We don't know for sure it's going to benefit, but we know that this disease has taken off and you've tried everything that works. We either try this or we're get watch you die.'

So, I think we have an appetite for risk and uncertainty in a very sick person.

Now, as you go down that continuum, newly diagnosed myeloma, the FDA [Food and Drug Administration] has just had a new statement saying they're going to tolerate more risk. But I say: this is a condition where the median survival is 15 years. Do you really want to risk getting a drug that could cause nerve damage or Parkinsonism? The risk-benefit becomes more uncertain. I think you're more likely to be precautious and say, 'Less medicines is better. Let's go with what's tried and true.'

You can reserve those sort of unproven things for later if you need it. So, we have different attitudes. I think there's some disagreement there.

But, with healthy people, people who come in your office, they feel totally fine. They say, 'I'm totally fine. Is there a way you can make me better off?' I think my attitude in medicine has always been extreme skepticism. There's a really extreme conservatism in the sense that you really need a lot of evidence before you start mucking with them, before you say, 'Well, you ought to take this medicine or that medicine.' Why? Because they're healthy. The odds are in their favor, you know? And what's the evidence you have that you're really making them better off? That's One.

And, Two: throughout most of history, most of the things we did to healthy people that we thought made them better off made them a lot worse off. A healthy kid with acne, we radiate the face. Okay. Well, now they got a lot of cancers. I mean we did so many horrible, barbaric things, not understanding safety, not understanding what we were really doing. We did radiate acne, in this country. A lot of people had radiation to the face for acne; but it worked a little bit in the short term.

So, you start to get that--so, that's my background coming into it. And I view myself as a healthy young person. I would say that I personally am extremely reluctant to take medicines and do medical procedures unless it's proven to benefit me.

Now, back to your question. I think that's exactly how I feel about dose after dose of the COVID-19 shot, which is that if we wanted to know the answer, we could ask Pfizer and Moderna, who have earned $100 billion from these products, 'Hey, run a hundred-thousand-person randomized study each year. Look for severe disease and death. Look at all these outcomes. And let's see if dose nine versus the prior eight doses and the ninth dose has an additional benefit. Let's see.' 'Does it vary if you're in a nursing home and you're 80 versus if you're 60 and you're walking around? Does it vary if you're 20?' We can make them do these studies.

The USFDA [U.S. Food and Drug Administration] has repeatedly, I think, abdicated that responsibility and omitted those studies. And if anything, the level of evidence for the ninth dose is much less than the level of evidence for the first dose. They actually did do a randomized study back then. Now they do it based on antibody data and sort of a lot of speculation.

So, my view is--and this has nothing to do with politics. This was my view in medicine long before the pandemic, which is that the bar for dosing healthy people has to be high. Certainly, I think year-after-year boosters do not meet that bar. And I have written some essays saying that I'm not going to get another one until you prove it to me that it benefits me. And, I think that has nothing to do with politics.

Now, the next part of your question--go ahead. You want to talk about that? Yeah.

Russ Roberts: Go ahead.

Vinay Prasad: The next part of your question, long COVID and long side effects.

I mean, it's going to be very difficult on all these topics. Long COVID, there's a lot to unpack there. I think it means many different things.

If you get so sick with COVID, you come in the hospital and we have to put a breathing tube down your throat and ventilate you, I have no doubt in my mind that the path to recovery is not going to be easy. It's not going to be linear. It's going to take a long time. And we have long seen people really sick with any respiratory infection--like, influenza or even some of the older, sort of, respiratory viruses--have a long path to feeling recovered. Some of them may never feel the same after being on the vent for a couple weeks.

To me, the unique thing about when people say long COVID is, some people believe you can have asymptomatic or very mild COVID-19. It's very mild. It's like a cold. But then, four weeks later, you have a brain fog or you feel fatigued all the time, and that's a sequela of the COVID-19.

I think that that's a very bold claim. And I think that's a claim that is disputed. I think that some people believe that--you certainly feel that way: You feel very fatigued. You feel brain fog. I'm not sure it's attributable to the COVID-19.

Some longitudinal studies say that the people more likely to experience it--it's linked more closely to personality traits like neuroticism or underlying anxiety or depression at baseline. Just like among people who herniate a disk in the back: Some people have persistent back pain. Some people don't. It's not always linked to the anatomic characteristic, but sometimes social support. And, because pain is a combination of mind and body.

So, that, to me, is always the question with COVID-19.

So, insofar as the vaccines in first quarter of 2021 prevented severe disease, I believe they did. Then I believe that they did reduce the number of people who have the disability after being on a ventilator, because they produced people on the ventilator.

But, insofar as people think the vaccines lower your risk of having brain fog when you get a cold in 2026, I think that's entirely unproven, and I personally am skeptical of that claim. And so, I've never seen any data to support that claim. And I wouldn't be getting a shot because, like, I'm worried about brain fog in 2027 from getting a cold in 2026, for instance.

In terms of long vaccine side effects, I think these mRNA [messenger ribonucleic acid] vaccines, every single time you dose a man, there's a risk of myocarditis. It's about one in 10,000 for the third dose, one in 3,000 for the second dose. That never goes to zero.

So, I do worry that with added dose after dose after dose, at some point the benefit-harm balance is going to be--I think it's already tipped for 20-year-old men; but for 40-year-old men, it's going to tip, too. In 50-year-old men, it's going to tip as well.

Then the final point: Why is the global vaccination policy--at least the U.S. vaccination policy, I would say the United States--why does the U.S. vaccination policy have the exact same recommendation for somebody who has had COVID twice versus somebody who has never had COVID-19?

And, Paul Offit, the vaccinologist, was interviewed about this, and he said that when the Biden administration took office, there was a private vote of five scientists and the vote was 3-2. The vote was 3 in favor of saying they should have the same policy and 2 against, saying they should have a different policy accounting for some immunity from having had the virus.

We know very clearly, the people who have had and recovered from COVID-19 will definitely get COVID again, but they're not going to get severe disease and death nearly at the same frequency as they would have prior to being un-immune.

And that's the question here. Does the vaccine provide a further reduction in severe disease or death beyond the protection you get from having survived it once?

And I think that that 3-2 vote was wrong. I think they got it wrong. The two people in dissent were right. But, to me, what's most concerning is that a private vote among five people in the Administration has set a policy that has such sweeping implications. Literally, this policy meant there are nurses in the United States who were fired from their jobs. There are people laid off from work. People had to change the sector of employment.

We forced college students to get this. We threw kids out of college if they didn't do this. I mean, we have a professor at University of California system who was fired because he didn't want to--Aaron Caretti was fired as a professor of bioethics, of all things, because he didn't want to do it because he had had COVID-19.

So, you are setting a policy that's really ruining the lives of people. It's not conferring the externality of benefit to third parties because it doesn't stop transmission, and it's based on a 3-2 vote among people who are really kind of guessing. That, to me, is a problem.

That, to me, is part of what I talk about in the essay about how could something that's scientifically so impressive be bungled?

And the last thing I'd say is--you talk about mistrust. I mean, every piece of data we have right now shows Americans are not vaccinating their kids for measles, mumps, and rubella like they did before. I mean, everything is plummeting. Vaccines are plummeting. Trust in expertise is plummeting. Trust in science is plummeting. You mentioned a couple times, what if there's a next pandemic? If there's a next pandemic in the next few years, I think America will tear itself in half.

There is so much distrust of the expert community, so much deep distrust, and I think most of it is well-earned distrust. I tried to tell them not to mask the two-year-olds. That's going to backfire. I tried to tell them. But, it's such deep distrust that if we have H5N1 spread--avian flu spread, for instance--I think literally half this country is going to do one thing and the other half is going to do the exact opposite. And I think there's going to be people moving states and massive upheaval. I think if people start to use the police state to enforce things like mask compliance--try to arrest somebody who is not wearing a mask or something like that--there will be violence.

I think it's a pressure-cooker situation. As easy as it is for the government to point the finger and say, 'Well, it's Joe Rogan. It's all the misinformation spreaders who created this distrust,' I think they have to look inward. They are largely responsible, from bad policy decisions, firing nurses who got COVID-19, forcing college kids to boost, making babies wear masks. Those decisions undermine their entire credibility and moral standing, and I think they did it to themselves, in my opinion.


Russ Roberts: There's a book by Gary Greenberg, which I love. I've interviewed Gary a couple times. I don't think I've interviewed him explicitly on the book, although it's just when the book came up. It's called The Noble Lie. And, it's a different perspective on the human proclivity to take a strong position--sometimes in the face of evidence to the contrary--in the name of the higher good.

So, for example, I think--we talk disparagingly of people who may have responded to financial incentives, but I think for some folks who were, quote, "pro-vaccine" long into the pandemic and long after--maybe it was a wise idea for many populations--they would say to you, I think, in a quiet conversation, maybe not for public record, that: 'Well, there's this anti-vaccine movement out there, so I'm always going to exaggerate the benefits of vaccines because I need to fight that dishonest view on the other side.'

And I just want to make the argument--it's actually, I think, in The Theory of Moral Sentiments by Adam Smith--that's such a dangerous, slippery slope. It's really much better to advocate for the truth. And, although the noble lie is always tempting, one of the worst parts of it, I think, is convincing yourself that your motivation is good, especially in times when it may have more complicated drivers.

But, certainly, this loss of trust--which is happening in all kinds of aspects of, I think, modern life--it's just a devastating loss of ease by which we can make decisions. We find ourselves at sea relying on sources that are often not true. So, comment on that.

Vinay Prasad: Well, that's so well-put. I mean, I think that you are absolutely spot on that many people are exaggerating, either saying things like, 'Oh, well, we know it to be completely safe. COVID's going to be worse for you anyway.' They say those things and one of their justifications is that, 'Well, the other side is saying things that are not true, so I have to lean in: I have to exaggerate a little bit to just sort of balance that out.'

And to me, that's absolutely unacceptable as a scientist and as somebody who cares about reason and truth, where I think, again, no matter what anyone else is doing, you have to keep a laser focus on what the truth is, what you believe the truth is, how to articulate it best.

Sometimes what that means is: it's easy to be a doctor and in January of 2020, tweet a picture of you rolling up your sleeve and say, 'Hey, listen, I'm 27, too, and I'm getting the shot. You get one, too.' But the truth is: really, there was a lot of ambiguity around that decision, particularly for people who had had COVID. It would be much better to focus on the things we know with some confidence or certainty, that older people should get it first, and then to acknowledge that if you're 27, maybe it's reasonable to do it, but don't feel too much pressure. We can hold off a little bit.

To me, it's always better to pursue the truth doggedly. I do hope it eventually wins out. I worry with COVID-19 that--I do worry it may not win out, ultimately.

I just think--one point. Many of the people who got the policy decisions wrong are winning all the national awards, and that concerns me. I mean it concerns me to see the FDA Commissioner, who I think made a lot of mistakes, winning an award for the American College of Cardiology for evidence-based public practice, which I think he did a weak job of.

I think--we didn't even talk about it, but there's been a lot of concerning revelations about what the scientists believed about the origin of the virus. Did it come from the lab or not? The House Committee is investigating. They find emails and SLACK [Searchable Log of All Communication and Knowledge] messages from these scientists privately acknowledging massive uncertainty while publicly they're saying, 'Absolutely not. 100% natural origin.' But privately, they're saying, 'We don't know, but we can't tell that to them.'

That is absolutely devastating and absolutely unacceptable.

So, to me, I feel like I'm not sure what will happen. Places like the National Academies of Medicine--I think they would probably disagree with some of my stands, and they're certainly rewarding people who were pro-school closure, pro-masking kids, pro-mandatory boosters of 20-year-old people who had had COVID. Those people are winning the awards. And so that, to me, suggests that we're not ready for an honest reflection--at least in the academy. Which, the academy and the public in America, I think, are almost entirely separate now. There's two groups of people who are not talking to each other.

And I think a lot of my professor colleagues do not have their finger on the pulse of what the average American thinks or feels, and I think, to their detriment, they're going to make some big policy errors in the future.


Russ Roberts: I should have asked you before. I don't know if I've ever asked you. Are you vaccinated?

Vinay Prasad: Oh, yeah. Well, of course. My personal decision was: The first dose--I think at the time I was maybe 36 or something and I had never had COVID-19--I was very interested in getting Dose One of the vaccine. And I ran out and got Dose One of the vaccine. I was reading the observational data showing, like--Dose One already gives you, like, 94% of the reduction in hospitalization, at least in a couple of nice New England Journal papers.

I also had tons of side effects from Dose One. I felt wrecked--pins and needles in the arm and it lasted for so long. So, I was extremely reluctant to get Dose Two.

I also knew that the timing of the two doses, which were 21 days apart for Pfizer and 28 days for Moderna, I knew that that was arbitrary. That was created merely to get the trial to complete in a tight timetable. Immunologically, spreading the two doses further apart would be much more reasonable; and that's what they did in the United Kingdom, of course, to their credit. So, I did not get Dose Two on schedule. I dragged my feet. And I only got Dose Two many months later when I received an email saying that I would be fired as a professor if I didn't get Dose Two. Then I got Dose Two.

And then, a few months after that, I wasn't going to get Dose Three, but then I got an email saying I would be fired on Monday if I didn't get Dose Three. So, I got Dose Three.

And, some people criticize me, saying, 'Well, if you really believed in your principles, you'd let them fire you.' I said, 'Are you out of your mind? That's ridiculous. I take Advil one day and--.' Look, I'm concerned about safety signals, but let's keep it in perspective. To be a doctor in America, you certainly have to do more than one day of misery in your life. It's a big sacrifice. I'm not going to give that up over one little shot. I'm going to suck it up, take some Advil, take the shot, and get back to work.

And then beat them in the long game, which is changing, I think, the attitudes of people around the error in their policy.

So, then after three doses then, now, of course, they've given up. They threw in the towel. I mean, they couldn't mandate it anymore. I think there's enough faculty member who would revolt. So, we're in a détente. The university has thrown in the towel.

Unfortunately, people like Aaron Caretti, he was fired. He's not been rehired. I think that's an embarrassment for the University of California system to fire--by all accounts, he was a professor that students loved. He's a nice guy. He had had COVID, recovered from COVID.

In my mind--and also, that's another point we want to make. You can have a mandate, but what should the penalty be?

I mean if you go back to some of the original mandates around vaccination, the penalty was a fine--$100, $200. There's some penalty. Should the penalty be loss of employment and a prohibition on employment in that entire sector for perpetuity? Should you get the death penalty for not getting this vaccine?

And I think many people did get a death penalty because they still can't get rehired by hospitals that still have the mandate.

Now, a lot of hospitals in particularly Republican areas are removing the mandate so they can hire these people back. But, three years of loss of employment for not doing this--I'm not sure we're helping people.

I mean that's, in my opinion, the death penalty. So, I think penalty should be proportionate to the evidence as well, and that's a longstanding principle of public health ethics. It was entire overkill. So, maybe you should have fined them. Fine them $500. Fine. Something that we can agree on is a deterrent. But what they did, I think, was unjustified.


Russ Roberts: So, let's close. You wrote a book a while back with Adam Cifu, another EconTalk guest, called Medical Reversal. I happened to interview Adam about the book, not you. And, in that conversation, we talked about how vertebroplasty--the injection of cement into the spine after a fracture, into the vertebrae of the back--is actually no better than the placebo of opening the can of cement and letting the person smell it.

And, in a recent episode, I revealed how, faced with this decision for my mom who had fallen. She's 91, had a compression fracture. She tried wearing a brace, unbelievably uncomfortable, realized that was not going to be a feasible solution. I didn't think I could talk the doctor into trying the placebo effect and opening the can of cement or the tube of cement as a noninvasive treatment.

He did the treatment. She's 100% pain-free.

But, as an economist helping my mom face that decision with my siblings, I went against the so-called literature of this technique that seems to bring relief, but when tested in a clinical trial did not, at least against this placebo.

But it highlighted--forget whether that's the right decision or not. I think what's interesting is how hard it is to make decisions in these kind of areas.

You and I are talking about vaccination and how many to get and whether to get it and what policy should be. But, the fact is we have a lot of evidence and we still have a lot of uncertainty.

So, what's your reaction to that story about my mom and the general kind of issues about so-called, you know, the 'gold standard' of the randomized controlled trials: are[?] finally establish whether vertebroplasty works? Versus the uncertainty we face in these kinds of decisions, that probably can't be totally removed, ever?

Vinay Prasad: Well, that's a great story. First of all, I'm glad to hear your mom is feeling better and doing well. That's the most important thing, and I think that's how your mom feels, how you feel.

And I guess--I mean, I think one thing to say is that when we wrote about that in the book, which was written in 2015, we were talking about, I think, two paired studies in The New England Journal in 2009. They're exactly as you described, Russ. Prior studies had shown vertebroplasty was better than, say, doing nothing.

But, what happens if you randomize people to vertebroplasty--where you open this polyacrylamide cement and inject it into the fracture? Versus, you take them to the room, the procedure room; you open the cement? You let them breathe that acetone smell, that nail polish smell. Then you throw it away and just inject some saline or don't do at all. And the answer was both groups felt better. So, it was really sort of a placebo effect. That's what those papers showed.

Now, one thing I should say is that: One possibility is that since those papers were done, maybe these interventional radiologists have invented some new cement or new needle technique that is actually better.

And so, I think that one thing to always admit is that perhaps a larger placebo-control trial done next week might actually be positive. So, I don't know. So, I'm open to that possibility. I do think that if they're going to do it and make money from it, that they should have some obligation to, like, generate that evidence or not have it paid for.

The next thing I would say is that this is so true for everybody. Which is: the beauty of that thing was that people on the control arm actually also felt better. People who didn't get it done--there were many people who said their pain went to zero or they felt better.

There's sort of a very famous story in medicine of the internal mammary artery ligation, which is an artery in the chest that doesn't go to the heart. And if you take people with anginal chest pain--which is sort of chest pain caused by blockages in the arteries of the heart--in the 1950s, they used to tie off this unrelated artery in the chest. Many people got off the table and said, 'Doc, I feel great now.'

Then, of course, they did a sham--two sham-controlled studies, randomized them to do it or say you did it and didn't do it. And both groups felt better. So, placebo effect is powerful. It's real.

I'm actually--the real policy question is: Well, should doctors actually do placebos? Probably through most of human healing, a lot of what counts as traditional medicines and ancient healing arts is a placebo--an elaborate, you know, put your hands on the patient, put them in a room, have some incense burning, creating this whole environment where you can coax them into a more positive mental state.

Should doctors do that?

And I'm actually open to that as long as sort of three things are met. I think the use of the placebo is reasonable if you can do it in a way that's not deceiving. There are some randomized studies where you randomize people to placebo. You tell them it's placebo. The control arm is doing nothing, but they still feel better on placebo. This is an IBS [?irritable bowel symptom?] study. It should be as noninvasive as possible. If possible, noninvasive and minimize harms, and it shouldn't get in the way of some proven, more effective alternative. So, I think there's a lot to be explored there.

But, your mother's story, I think, is so telling. And I think what it reinforces is: Yeah, there will always be more than randomized data to make these decisions. Ultimately, your mom doesn't care. If it was placebo or not, she feels better. Right? That's, I think, one of the most interesting things about medicine.

Russ Roberts: The thing I'll reveal about myself--I didn't reveal it in the last conversation that I talked about this on the program: When we were trying to make this tough decision, made tougher by the fact that my mom is 91. Thank God, she's mentally totally competent. But, she didn't take as many statistics in nursing school as her son did, who took econometrics. So, in theory, I'm an expert--who loves her. And I should make a really good, informed decision on her behalf. But, what I want to confess is when I was trying to make that decision, I thought: I wonder if that test, that control trial, I wonder if the two groups were really equally in extremis? Were they equally in pain?

And surely there's some bias--there's some publication bias--of finding the placebo effect.

And so, I thought: That makes me want to err on the side of intervening and getting the cement.

And by the way, I have a doctor friend who does this procedure. He does it with local anesthetic. Her doctor--myy mom's doctor--actually does it with general. So, it's just yet another reason not to do it because it's a much riskier intervention.

But, I managed to convince myself that the randomized control trial that was in your book probably wasn't as foolproof and ironclad as it might seem to be.

And, yet, I didn't look that study up.

I did not give it a careful evaluation. I used my confirmation bias, I think, that I wanted to intervene. I wanted to be helpful to her. I didn't want her to wear that brace. And if at worst it would be the placebo effect, it was worth risking in the face of the general anesthesia risk.

So, I just want to confess that. Even though I care deeply and often evaluate studies very aggressively for their reliability, in this particular case, I didn't.

So, it tells you something about me and maybe about, in general, how there's a bias toward intervening versus not intervening.

Vinay Prasad: You know, I think I agree with you, Russ. I agree with your reasoning. I don't disagree.

I guess the only thing I wonder is: Should the manufacturer of that cement be able to charge $3,000 a vial? If you can get--that's the question. Because we could do the same procedure and use--if it is, in fact, the placebo effect, how much should they get to charge for that cement?

And then, the next question is that, I mean, at the end of the day, in some ways you did the right thing, because the worst is, placebo effect, she'll feel a little bit better. At best, there's some real effect. And, as long as the bill is not prohibitive to her or something like that.

That's one of the interesting things about placebo. And thankfully, she didn't suffer any complications.

So, I think that's a way it could have gone wrong. The cement could have gone somewhere it's not supposed to and something bad could have happened.

And I think that's one of the reasons why--I also say that sometimes I wear my doctor hat or my policy hat. When I wear my policy hat, I sometimes say, 'This medicine, the data's so bad. We can't cover this medicine, this hospital.' Then the patient's in your office. You wear your doctor hat, and you say, 'Well, you know what? It's not my job right now to decide what we're paying for or not. It's already covered.' And, they want to try it and they know the uncertainty. We're going to try it. So, I think sometimes we wear our son hat and wear our econometrics hat.

Russ Roberts: My guest today has been Vinay Prasad. Vinay, thanks for being part of EconTalk.

Vinay Prasad: Thanks for having me, Russ.