Paul Romer on the COVID-19 Pandemic
In this Bonus Episode of EconTalk, economist and Nobel Laureate Paul Romer discusses the coronavirus pandemic with EconTalk host Russ Roberts. Romer argues that the status quo of shutdown and fear of infection is unsustainable. Returning to normal requires an...
Nassim Nicholas Taleb on Antifragility
Nassim Taleb, author of Fooled By Randomness and The Black Swan, talks with EconTalk host Russ Roberts about antifragility, the concept behind Taleb's next book, a work in progress. Taleb talks about how we can cope with our ignorance and...
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.


Alan Goldhammer
Jul 27 2020 at 12:04pm

Russ – this is a great conversation and puts a lot of the risks of the pandemic into perspective. I hope this gets a broad readership.

Bob Smith
Jul 27 2020 at 1:38pm

I don’t know, the deeper the conversation went the shakier Mr. Taleb’s claims got.

I think Mr. Taleb provides a service to humanity by making people conscious of fat tail risks, but I am suspicious of his purported rigor.

I wonder how many people really understand his math, and whether the complicated equations give him a sheen of guru status without anyone really understanding them. They look impressive, so he must know what he’s talking about, right?

Mike M.
Jul 27 2020 at 3:56pm

A great discussion with a great “outside the box” thinker. The discussion provides lots to think about, especially the last five minutes. I really appreciate the discussion on the inter-generational discussion, makes you think really hard about the golden rule.

Jul 27 2020 at 4:53pm

Btw I hope to hear Dr Ioannidis soon on EconTalk to defend his position! In the meantime I recommend the series of interviews on the UnHerd YouTube channel.

Andrew Gisler
Jul 27 2020 at 5:28pm

Interesting discussion but I must admit I probably missed points and would benefit from listening again.
I didrdid Mr. Taleb’s article “Ethics of precaution’ and wondered two things.
1) what is his thought on the shutdown of elective surgery? It seems this is creating a situation where one risk is being avoided at the expense of individuals having to take on more risk (not getting treatment that they might need)

2) Must it be an all or nothing on the extent of precautions people must take. Example, essential places like grocery stores make sense to take extra precaution but why must it also be done for the non-essential places like bars and restaurants?


Michael Byrnes
Jul 29 2020 at 5:38pm

Must it be an all or nothing on the extent of precautions people must take. Example, essential places like grocery stores make sense to take extra precaution but why must it also be done for the non-essential places like bars and restaurants?

The basic idea is that taking precautions at non-essential places protects people everywhere, including the essential places like grocery stores.

A big factor in your individual risk for contracting Covid-19 is how many people in your community have it (and are infectious). If 1% of your community has it, you are relatively safe no matter what you do. If 30% of your community has it, your risk is much much higher because you are likely to be interacting with infectious people quite frequently.

The more universal the precautions are in your community, the less likely that there is significant spread and a high percentage of infected people in your comminuty and thus your risk is much lower.

Dave Neil
Aug 13 2020 at 12:23am

You’re talking about individual risk of contracting the disease, but if 1% of your community is actively infected then it most certainly does matter how each member of that community behaves.  It’s all about the transmission rate:  If the community behavior leads to an Rt of >1 then you get exponential growth and starting from 1% currently infected you will get very large numbers very fast.  Plus, remember superspreader impact: it just takes one person in a community of thousands to to infect dozens or even hundreds in just a matter of weeks.  And you might be the one who is already infected but don’t know it yet.  Meanwhile, 30% would be extremely high, only seen in the case of some cruise ships, care homes and prisons where your best course of action would be to get out, but of course none of those people are/were able to.

Mark Brady
Aug 20 2020 at 12:51pm

Sure, sounds very folksy. If I’ve learned anything from economics it’s that things that sound good often have bad unintended consequence and sometimes those unintended consequences are worse than the original, purported ‘good thing’.

Duke just reported that some masks that folks are wearing are actually worse than no mask at all.

So if everyone just did the ‘good thing’, the ‘reasonable measure’ but with the wrong mask it would cause more infections not fewer.

Dentists are describing a phenominon called ‘mask mouth’.
So if we’re all do the ‘good thing’ when it’s not really helpful, perhaps outside in the sun, but folks are increasing their chance of cavities and worse, maybe the we should be skipping the ‘good thing’.

Wisconsin State Gov is requiring employees to wear masks when they are alone and on conference calls.

When we take such a simplistic approach like, “hey this is always a good thing”, then we get dumb policies like this that will have real negative consequences.

Wasn’t it yesterday that everyone thought the world was going to be destroyed by single-use plastic straws and bags. We’re told that the climate crisis is about to kill billions. The massive increase in disposable masks are littering my neighborhood streets. Wouldn’t this be treated as hazardous waste in a hospital?

Can’t we discuss the viability of building heard immunity by allowing younger, healthy folks to go about their business (if they choose to) and contract this? In fact, this is the only way to test a vaccine. If the early vaccinated people do everything possible to hide from the virus, we’ll have no idea if this works.

Ajit Kirpekar
Jul 27 2020 at 6:07pm

Maybe I’m reading too much into his definition of science and dismissal of evidence, but I think they go hand-in-hand.

Empirical work without theory is just as empty as theories with no empirical evidence.  Otherwise you end up concluding that doctors make people sick therefore the cure should be to eliminate all doctors.

I also agree with the commenter above, the longer the conversation went on, more uneasy I felt about his stated claims. And fat tail seemingly lurk everywhere and I’m not convinced his estimation of the true costs is well known at all. These are dynamical systems were behavior continues to update an augment in so many different ways.

As an aside, I don’t know if Dr. Taleb will be reading these comments, but I wish he would learn to adapt his conversation style to different audiences. Pretending that everyone has an up-to-date training in probability theory and understands the lexicon is a mistake.

Jul 27 2020 at 6:27pm

Most of the statistical jargon went over my head, but here are some thoughts:

I have no problem with the better-safe-than-sorry reasoning behind masks, distancing & sanitation, as the potential upsides far outweigh the downsides. But let’s not pontificate about whether they actually work or not, because right now we just don’t know.

I have a major problem with the assume-the-worst reasoning behind shutdowns and lockdowns, as the downsides far outweigh the potential upsides, even in actual worst-case scenarios.

You cannot treat all pandemics the same and categorize them by default as fat tail events. The specific pathology and epidemiology of different viruses matter. Seasonality matters. Deaths (life years lost vs lives lost) vs cases vs infections vs periodic mortality matter. You have to put things into context and perspective.

Science is first and foremost about evidence, which is gathered and interpreted in order to infer/deduce and most importantly test properties and processes—not the other way round.

Science is not the same thing as logic, and mathematics is not science or truth but a language. What makes mathematical sense may not be true or even scientific.

Decision-making must be based on data and evidence as much as principle and protocol, especially in a fast-moving and evolving event like this pandemic.

I hope to hear Dr Ioannidis soon to defend his position! In the meantime I recommend the series of interviews on the UnHerd Lockdown TV YouTube channel.

Julien Lafleur
Jul 27 2020 at 6:57pm

I greatly appreciated Mr. Taleb’s point that science is about properties, not evidence.  The fact of the matter is that while evidence is necessary for science, it is not sufficient. At least some of the greatest science in our lifetimes comes from a blazing insight into properties, and evidence is only there to confirm or confound.

This misapprehension is ubiquitous today.  Listen to pretty much any podcast or news outlet, and they will proudly proclaim their sources, as though that fact that their data has been verified somehow makes their interpretation of those facts correct.

What other processes are possible given the evidence you’ve collected?  What piece(s) of evidence would force you to re-evaluate your hypothesis?

Granting that demonstrating causality is very hard, if you cannot give some explanation of process, then all the evidence in the world does not create science, it just fills spreadsheets.

David Gossett
Aug 10 2020 at 12:11pm

Wow. This is extremely well put… “At least some of the greatest science in our lifetimes comes from a blazing insight into properties, and evidence is only there to confirm or confound.”

You are right on the money, in my opinion. And this encapsulates the entire episode into a single sentence. Well done!

Tom Kelly
Jul 27 2020 at 9:07pm

In listening to your podcast for the first time in awhile I found myself disappointed that you didn’t bring up the arguments against a law requiring mask wearing.
First of all Talebs argument that two people wearing a mask increases probabilities by square. If one vulnerable person wears a mask and if that reduces the chances of getting a virus by lets say 98% the second mask reduces only the 2% probability. (I do not know the exact numbers) but his assumption is misleading. If some one is at risk….wear a mask..stay home.

Secondly, the Cold (56,000 died per CDC) Shingles, the Flu, kissing, sex and etc all can kill people.  When do we have the government tell adults what to do, avoid or wear? Taking it even further a sign or a t-shirt can cause aggravation in other people and cause stress or anger that can cause death. At what point do we give up our liberties? And on who’s information?

You also did not discuss, at least enough, the deaths that shutting the economy or borders will cause.

And pandemics are not vastly more dangerous as war. What the heck one nuclear bomb can kill 20 million instantaneously.  We better be prepared.

Taleb …and you have lost a lot of credibility for me. Your fear has gotten the best of you.


Robert W Tucker
Jul 28 2020 at 9:33am

Answers to some of your concerns lie in the differences between fat-tailed multiplicative risks and other risks you mention.

Martin Dertz
Jul 28 2020 at 11:35am

the Cold (56,000 died per CDC) Shingles, the Flu, kissing, sex and etc all can kill people.  When do we have the government tell adults what to do, avoid or wear?

My understanding of Taleb et al’s position from reading their paper is that those are all known risks so their harms are well known (no uncertainty). Government intervention comes when there is a novel pathogen with high rate of spread.

You also did not discuss, at least enough, the deaths that shutting the economy or borders will cause.

My guess is that’s because those things aren’t contagious so limited possible harm they can cause.

And pandemics are not vastly more dangerous as war. What the heck one nuclear bomb can kill 20 million instantaneously.  We better be prepared.

My understanding of Taleb’s point is that both pandemics and nuclear war are in the same class of risk. Listen to 9:21 where Taleb says

 If I tell you a billion people died, you know they did not die drowning in their bathtub in any given year. Whereas if I tell you, ‘A billion people died. Try to guess,’ odds are it’s going to be either a nuclear war or a pandemic. More likely to be a pandemic. It’s actually fattest tail.

Later he says

So, we spend a trillion dollars on weaponry, nuclear warheads. You can destroy Russia, or you can destroy any country you want a hundred times. But, we are still not able to have proper testing and/or proper quarantine.

So my understanding of his point is that it’s frustrating that for one of those risks people say “We better be prepared.” and for another they say “But what about the unintended consequences?” given they have the similar properties.


Jul 27 2020 at 10:55pm

But the point is by wearing a mask, lock down, etc aren’t u making the society fragile in long run (with assumption that the virus will persist & cannot be stamped out; long game is immune system adapting to viruses ) !  I understand in short term u get the some benefit of controlling hospitalizations.


Also, who is controlling the actual fragility in the system: uncontrolled population growth, rapid urbanization w/denser societies, faster movement of a people across borders, etc.

Mary Hirschfeld
Jul 28 2020 at 2:29am

Frustrating interview.  Looking at pandemics solely in terms of probability distributions ignores the fact that the sorts of measures that might hinder a small pandemic are unlikely to do anything to stop the big one that kills billions.  Meanwhile the precautionary measures exact a large toll:

The externalities involved in our choice to socially distance are not limited to the spread of this disease. They also include the impact on a deadly unraveling of the economic and social fabric that sustains life.

Jul 28 2020 at 6:27pm

Important point. Thanks for this reference.

Kingston Smith
Jul 28 2020 at 2:13pm

As someone who cares about human freedom, individual liberty and personal responsibility, I found myself disagreeing with almost everything Taleb said as usual. He is smart enough to understand the difference between an opinion and a statistical finding, so I can only assume he conflates the two intentionally. One cannot credibly claim in the same interview that statistical analysis supports both the idea that the role of government is to quarantine international travelers at the border, and that there is no risk in visiting his elderly mother in Lebanon when he wears his ridiculous double mask and goggles ensemble. Those are simply his personal opinions and life-choices based on his own fears, values, and risk tolerance, same as everything I’ve ever read or heard from him. Repeating fat-tail over and over again is not an argument. It’s a ruse for people who don’t understand statistics.

Aug 2 2020 at 9:36am

The thing about his repeating his favorite mantra of “fat tail” is that, unless I’m missing something, he often really means “long tail” (i.e., observations very far to one side from the mean)

Todd Mora
Jul 28 2020 at 4:47pm

I thought this was a very interesting discussion and I do not possess the statistical expertise to question Mr. Taleb.

I would have liked to hear more discussion based on the Russ Roberts axiom “there are no solutions, only trade-offs.”  With the management of a pandemic there are an infinite number of trade-offs being made in the name of safety.  The concern I have is who is make the decisions of what to trade off and what is the benefit?  There is very little rational discussion regarding the trade-offs being made, to the degree that the governmental agencies we are supposed to trust for making the decisions, is intentionally lying to the public.

I think there could have been more healthy skepticism in this discussion by both Russ and Nassim.

Jul 28 2020 at 5:30pm

Thank you for the interview. There is much that Nassim Taleb and you had to say that I learned from and/or agreed with. One important one being those about non-linearity and also that reducing transmission by a factor (50% in his example) can have a much greater affect on virus spread.  I do not think he is right, though, that epidemiologists don’t (or, as I think he said) are incapable of understanding this.

There are a lot of things I strongly disagreed with — and, points of disagreement are more interesting/provocative so I will focus on them.

1) Saying that the World Health Organization lied about masks or, in Taleb’s words “worse than lied, they were incompetent”. I don’t think the WHO lied — the CDC is another question that I’ll get to below. And, (as I’ll also elaborate on below, I THINK IT IS HARMFUL. I think the WHO was genuinely were skeptical of the evidence for the effectiveness of masks for some time. I too find this puzzling but I also respect that they’ve studied the matter much more deeply than I and that it is more complicated than the analogies you two made to parachutes and seat-belts. The effectiveness of seat belts and parachutes is much clearer than cloth masks. It is not clear if (how) effective masks are if used properly and also what fraction of people use them properly.

Few people put seat belts on incorrectly but many (most?) people put masks on incorrectly. Personally, I find my glasses fog up when wearing a mask despite efforts to get masks that fit tightly. In addition to the annoyance, this suggests air is going around it, not through it. Does the sock mask Russ recommended really do any good at all (from my understanding, little if any good)?  Does wearing a mask for hours accumulate the virus (assuming I have it) and CREATE a potential hazard either by expelling a larger amount of it if I cough or if someone else comes in contact with it the mask. So, the idea that despite lack of clinical studies it should have been obvious is not as easy a judgment as you suggest.

1a) As to the statement that they intentionally mislead the public because they were afraid of the public buying masks in limited supply. First of all, I think that was the CDC, not the WHO. Second, I agree that was bad. But it was not silly as Taleb suggested that they didn’t understand “the market” or that people would make their own. Yes, people would (have) made their own. But, as you two seem to be unaware, they also, at that time, were actively hoarding N95 masks which they correctly understood to be more effective. And, this was at a time when N95 masks were is very short supply so that medical professionals treating COVID or high risk patients were having to reuse masks for many days. I live in NY. I saw the results of that (our mortality rates are higher than elsewhere because the hospitals were so overwhelmed care was inadequate in many cases).

I am not saying the ends justify the means. But the concern about public buying of masks wasn’t silly, as you suggest.

2) The reason I say that saying the WHO lied is dangerous is that the US has defunded programs to control Ebola, polio and malaria as a result (not, obviously, because of your comments, but you are contributing to the general perception. Certainly the WHO has made mistakes and arguably worse. They have generally been slow, as you’d expect of a large bureaucracy. But on balance they have been very helpful. I don’t think the world (or the US) benefits from either a less well-funded WHO or one more in the grip of China.

3)  Your casual assumption that people would not eat in restaurants without having to be told to is contradicted by the evidence. Yes, of course, many people have done so. But many have not. One example is Sweden. I was mystified this spring that people would be filling the bars and restaurants there, but I asked a friend who said “yes, the bars are full”.

4) “Don’t worry until 1000 people have died. This seems like very bad advice and Nassim didn’t take it himself. He began working on this long before it had killed 1000 people.

… to be continued  in another post.


Jul 28 2020 at 5:56pm

Continuation of previous post:

4) “Don’t worry until 1000 people have died. This seems like very bad advice and Nassim didn’t take it himself. He began working on this before it had killed 1000 people (which didn’t happen until February). And many countries did, too, which is good and proper.

Also, the “magic number” hardly defines a pandemic. Many diseases kill way more than 1,000 every year. There are reasons that new diseases need to be reacted to differently than things like tuberculosis (though, it would be good if more resources were put into tuberculosis, too). But the criteria for assessing risk should be much more nuanced and sophisticated than Taleb suggests and reaction needs to be much earlier — as he actually advocates.

1,000 may be a useful criterion for historical research but that also will introduce selection bias as it will ignore risks (e.g. SARS, I think) which had the potential to be pandemics but never reached that threshold.

The discussion suggests a focus on reaction to the pandemic after it starts rather than measures (that epidemiologists had recommended for some time) such as stockpiling N95 masks and respirators and creating protocols in advance.

5) Russ pointed out that people who ‘laugh at’ the notion that seat-belt-wearing leads people to offset that safety measure by taking more risk in other ways invoke it for masks. My response is a) yes, but the converse also seems to be true. For instance, Russ dismisses or ignores that when he presumes that mask-wearing is a good thing. And, it actually could have been more of an issue with masks than seat-belts.  The Peltzman effect is generally partial. That is, about 25% of the direct benefit is offset by behavioral changes. But, if people dramatically overestimate the benefit, then the net result could be adverse. This may seem silly since we now know that mask wearing (of cloth masks is quite beneficial) but hand-washing was a similar thing that seemed clearly beneficial but we now think its benefits were overestimated. I remember the days much was made of the fact that the virus persisted on surfaces for many hours but it now seems that risk was greatly exaggerated.

6) Finally, Russ suggested his father was sensible not wearing a bicycle helmet because he enjoyed the air rushing through his hair. Curiously, Nassim doesn’t comment on this but then goes on to talk about a) risk of variability v risk of ruin (suggesting that the second is much more of an issue than the first) and the importance of outliers. It would seem that Russ’s father was ignoring both of these.

Thanks for the provocative discussion and for reading this entire comment.


Jul 29 2020 at 10:38am

To play devil’s advocate with Taleb’s central thesis:

Is gerontocide minimization itself a greater harm if the cost turns out to be social and economic collapse?

Is this a risk worth taking, and- is it a risk society and government can afford to take?

Jul 30 2020 at 3:33am

When the stimulus passed it included help to hospitals affected by the covid-19 crisis. This created an incentive for hospitals that were struggling to label things as Covid-19 related that never really were, there only had to be “COVID like symptoms” not even an actual test. Labeling a death as COVID-19 related was 8,000 dollars in federal money alone. This has been acknowledged  by all the CDC/ health officials.  This leaves a situation where the death rate appears to be skyrocketing and not going down at all.
I’m disappointed in this podcast. I don’t agree with the guests claims about “fat tail” risks. The data coming from the government is contaminated with false COVID related deaths described above. The current numbers supporting any model are unreliable.



Michael Byrnes
Jul 30 2020 at 10:36am

This created an incentive for hospitals that were struggling to label things as Covid-19 related that never really were, there only had to be “COVID like symptoms” not even an actual test. Labeling a death as COVID-19 related was 8,000 dollars in federal money alone. This has been acknowledged by all the CDC/ health officials. This leaves a situation where the death rate appears to be skyrocketing and not going down at all.

What you have done here is made a sweeping and significant (if true) claim here which you have offered no evidence whatsoever to support. Not productive.

All-cause mortality rates (ie, the rate of death from any cause) is fairly consistent from year to year and has taken a big spike upwards in 2020, which is a strong indication that Covid-19 deaths are bring undercounted.

From March 1 thround May 30, there were roughly 122,000 excess deaths from any cause in the US, a number approximately 28% higher than reported Covid-19 deaths.

It doesn;t rain because people hold up their umbrellas.

Jul 30 2020 at 11:47am

It’s not a secret? This has been discussed by top health officials for a while now:

If I’m a physician at a hospital that’s struggling because of the restrictions (not preforming non essential treatments etc. ) I can check a box labeling a death as “COVID-19 related” and receive federal money or face laying off staff.

The numbers are inflated and reports are plagued by this bad incentive created from the CARES stimulus package.

It doesn;t rain because people hold up their umbrellas.

If someone who has COVID-19 get’s eaten by a shark, we find their arm and their arm has COVID… it’s going to get recorded as a COVID death.

Michael Byrnes
Jul 30 2020 at 10:21pm

The numbers are inflated and reports are plagued by this bad incentive created from the CARES stimulus package.

If a lot of people say something stupid, does that make it true?

I’ll spell the part of my argument that you opted not to engage with more clearly.

People die every year for all sorts of different underlying reasons. And the overall death rates, taking all deaths, not just Covid deaths, into account, are quite stable on a year-to-year basis.

Until 2020. Over a 3-month period (March-April-May) there was a massive increase in total deaths in the US, regardless of cause. Roughly 122,000, a higher number (by 28%) than the reported number of Covid-related deaths over the same period.

So, the same year we get Covid, we get a massive increase in the death rate.

If your claim was accurate, instead of seeing a bid increase in the overall death rate, we would see a relatively stable death rate, with the increase in Covid-19 related deaths offset by a reduction in other types of deaths.

But we don’t. We actually see more of a spike in all-cause death rates than deaths specifically attributed to Covid.

Another thing you do here is ignore the parts of the equation that cut against your belief.

Yes, it is possible that some deaths classified as Covid-related would have happened anyway. But the other side of that coin is that not everyone gets tested for Covid before dying, and the misattribution issue is going to work in both directions.


Armin Chosnama
Jul 31 2020 at 5:52pm

All-cause mortality went up around the same time as covid arrived, so it’s definitely logical to conclude that it’s caused by covid. But it’s not a very strong causative argument. Definitely not strong enough to call concerns about the logic or the data stupid.

Michael Byrnes
Aug 1 2020 at 7:04am

Armin Cnosnava wrote:

All-cause mortality went up around the same time as covid arrived, so it’s definitely logical to conclude that it’s caused by covid. But it’s not a very strong causative argument. Definitely not strong enough to call concerns about the logic or the data stupid.

Not a very strong causative argument? Viewed purely in isolation, apart from other knoweldge we may have, as simple “thing A” (Covid) and “thing B” (surge in all-cause mortality), it is fair to say that there isn’t a causative argument in a statistical sense.

But… we know that:

1. All-cause mortality is very stable year-to year

2. This year there has been a massive surge in all-cause mortality

3. This surge in mortality has coinicided with the arrival of a heretofore unknown resporatory virus that spreads more readily than the flu and has been demostrated to cause pneumonia, hospitalization, and death.

The obvious educated guess to make here is that Covid is causing the excess deaths.

Other scenarios may also be possible, but the sole alternative offered for consideration in this thread is not credible. (That alternative: people are dying of ordinary stuff peopkle often die from, but those deaths are being (mis)classified as Covid-19 deaths in people who are infected). Were that true, the data would show something different: relatively stable all-cause mortality even as deaths attributed to Covid-19 surge. That’s not what is observed.


Aug 1 2020 at 8:45pm

The increase in death rate is expected for something totally new to the world.
Saying “122,000” higher is a straw-man whenever you consider that number is like 1 percent of the total number of “confirmed cases”

I also said “COVID” related deaths. You’re comparing apples and oranges by comparing numbers year to year, the death rate isn’t something that’s linear.

Let’s cut down all the coconut trees from all the people that die from getting hit in the head with a coconut while we’re at it.



Mark Brady
Aug 13 2020 at 1:08pm

Excess deaths.

Emergency room doctors have reported a steep decline in the number of cardiac cases they are treating. Are people just having fewer heart attacks or are they just not going in for treatment since they are more scared of a virus than permanent heart damage.

Suicide rates are up

I can go on with deaths being caused not by COVID but the response to it. An economics podcast and its listeners should be able to discuss that we have a trolley problem with an 87 y.o. emphysematic on one track and a 45 y.o. small-business owner, father of three, small children on the other track. We should be able to discuss QAYL without be accused of geronticide.

Sylvain Ribes
Jul 31 2020 at 9:21am

I confess I’m no Taleb fan, but still this was underwhelming.

I agree with one of his main points “take cheap measures that have compounding effects”, but who wouldn’t?

Aside from that, anything that should’ve been layed out clearly and accurately in his discourse was not, and I felt like Russ was a bit more accomodating than usual.

I wish tradeoffs had been more discussed than in the handwavy “some industries are suffering, but not too many industries are suffering.” which is both wrong and euphemistic.

On quarantine rules, they can be handy in some very specific set of circumstances, but it’s demagogic pandering to insist on closing borders and quarantining travelers once the epidemic is endemic to a country. If you have thousands of endemic cases, a spare hundreds more will not change the outcome in any way. The fact of the matter is that in today’s world such rules probably couldn’t have been enforced quickly and stringently enough in most countries to quash the pandemic. Maybe we’ll do better next time if we manage to have some privacy-compliant contact tracing tools, but it’s not obvious to me.

The whole focus on the “1000 deaths+ or nothing” is patently absurd. Criticizing naive modeling while coming up with such an arbitrary conclusion (from an non-described blackbox model) felt very awkward for several reasons.

Indeed Nassim mentions mutations as one of the reason why we should be very watchful for those 1k+ deaths pandemics, but fails to address that mutations overwhelmingly occur to make viruses less lethal, not more so, but more importantly that it could happen to any strain of virus, not just the already deadly ones.


It’s also disingenuous to handwave away the concern that the governmental response to the pandemic might have been costly. Yes, ultimately a large part of the cost stems from people being afraid and not only from the coercive measures imposed from top to bottom. But people would have been much less afraid if the strong fear signalling from public policy had not come to be.


Finally, I wish I’d have heard spelled out specifically, from Nassim of all people, how important it was to apply the precautionary principle stringently when uncertainty was at it maximum, but how equally important it was to loosen it when the situation became clearer.

Because it’s also wrong to flat out assert that we know nothing about “morbidity”. We do know quite a bit, and for hospitalized patients, figures seem to be in line with patients hospitalized for influenza or for pneumonia.

Michael Byrnes
Jul 31 2020 at 11:35am

I agree with one of his main points “take cheap measures that have compounding effects”, but who wouldn’t?

In the United States today? Too many to count wouldn’t.

The whole focus on the “1000 deaths+ or nothing” is patently absurd. Criticizing naive modeling while coming up with such an arbitrary conclusion (from an non-described blackbox model) felt very awkward for several reasons.

He might not have made this point clearly, but I took him to be offering this as a rule of thumb – once you are at 1,000 deaths you are (somewhere) in the fat tail of the distribution and traditional statistical approaches will produce misleading or inaccurate results.

Because it’s also wrong to flat out assert that we know nothing about “morbidity”. We do know quite a bit, and for hospitalized patients, figures seem to be in line with patients hospitalized for influenza or for pneumonia.

Taleb was getting at the question of long-term morbidity following recovery. We may not know “nothing” about this, but we know very little in terms of what the risk is, how many affected people will have long-term issues, etc.

Todd Johnson
Aug 1 2020 at 3:32am

Hello Russ and Nassim,

Always good to hear your thoughts. Some contrary observations:

1. Masks, while I totally agree that mask can reduce transmission (I wear them at work every day when around close groups)…. what about the unintended consequences? Because we all wear masks, our teams (over 1900 people at our Mfg site) stand close together, sit close together in meetings, etc…. These are not N95 masks (which are effective), they are more like standard surgical masks or dust masks allowing plenty of air in through the sides, etc. If no one had masks, I believe it far more likely group sizes would be limited (fewer people in meetings, an empty chair between each person, like we required at the beginning of the pandemic, etc) and it is certainly not obvious which situation is better at reducing disease spread….. masked and congregate, unmasked and maintain physical distance.

2. The government does not make masks (about 20:15)…. I think that was a mis-statement.

3. Uncertainty… when you don’t know how bad the pandemic will be, the reaction is not, ‘let’s shut down society’… the reaction should be – ‘follow law and social norms’. Encourage responsible behavior by the population and encourage all elderly and sick individuals to take extreme care.. Breaking the constitution (local authorities wielding near unlimited control of ‘their’ subjects), forcing those who are not sick to stay home (destroying their ability to earn and likely increasing overdose deaths… approx 5,000 more in March, April, May 2020 than 2019), not allowing people to leave their house (vitamin D is a critical modulator of the immune system), etc. I could continue…. Don’t Panic is the appropriate response to uncertainty.

4. Are you aware that for the past 9 straight weeks. Every week, 2020 weekly deaths are lower than 2019 weekly deaths (CDC data… weeks 20-28 (May 2nd week to July 2nd week). And using the same publicly available total weekly death data, there are fewer total deaths W1-W28 2020 compared to 2019, 2018 and 2017 after adjusting for yearly increase in deaths (using W19-W40 from prior year as baseline to exclude flu deaths). I can’t explain this, perhaps 2 months of driving deaths in prior years makes for the difference?

Interesting discussion of black swan events. This disease isn’t a black swan, only the response can turn it into one. Stay safe.

Todd Johnson
Aug 2 2020 at 8:08pm

I have to mea culpa on item 4. The CDC data continuously updates even after the weekly data is listed at >100 complete. In fact after building the database 9 weeks ago and only updating the most recent week after it is listed as “>100” complete, I double checked the entire data set. 2018 was stable, 2019 fluctuated mostly in single digits or by dozens with a couple weeks changing by hundreds, the 2020 data, particularly the near current weeks changed by up to 1000s. With this level of fluctuation on a very clear attribute, it’s a wonder the public health officials have accurate information to make decisions.

Maria S.
Aug 1 2020 at 12:55pm

I guess when you’re a hammer, everything looks like a fat-tailed risk.

I suspect Roberts’ respect for Taleb and his other insightful work prevented him from really challenging Taleb’s claims.

Still an interesting and thought provoking episode

Aug 1 2020 at 3:44pm

Two issues

Masks: A little bit of inside baseball for the non-medical people. When you see a meta-analyis published in a prominent journal, it means the DATA IS BAD. Otherwise there is no need to publish a meta-analysis. A really good randomized double-blind placebo controlled trial referencing established biological process (N=1) with large effect sizes always beats a meta-analysis. The recent meta-analysis in the Lancet on social distancing and masks for COVID+ viruses using NON-RANDOMIZED studies concluded “low certainty” for efficacy of cloth masks in the community. e.g. the DATA IS BAD.  The data for social distancing is quite good for 3 feet, however; 6 feet, not so much.. So what happens, politicians and n0n-medical people (epidemiologists) make wild statements like “if everyone wears masks we can open up the economy.” ???. Guess what, people are not computer algorithms. They can only process so much prophylaxis at once. Our nurses in the COVID clinic have to instruct patients to take of their gloves because they inevitably wear them all day and get infected with self-contamination, rather than wash their hands regularly. Multiple countries and clinics suggest against masks because they create a false sense of security. Randomized studies (not anecdote), some not all, suggest cloth masks can actually increase infection rates in certain settings.
Current policy probably has increased death rates in terms of life years lost. COVID kills old people not young people.FULL STOP. Economic collapse and social deprivation kills young people not old people of Social Security (or government paychecks….skin in the game). FULL STOP.  While everyone was preaching about “bending the curve” in hospitals, no one was was considering about how to “bend the curve” in outpatient psychiatric medicine for the next 2 years to deal with suicide and drug abuse deaths directly related to poorly targeted lockdowns devised by credentialed but not educated public health specialists and epidemiologists. CDC director Redstone testified this month that they are already detecting the excess mortality signal in young people from “deaths of despair” in their data……
Old age veneration. ???? Biology says I’m dead weight now that my children are raised. I want young people to go out to bars, drink a lot, and have a lot of sex while getting infected. Today. Its looking like this strategy worked for Sweden. This is the prelude for them having children who will ultimately care for me in my old age and invent a lot of beautiful technology to address all sorts of existential threats like asteroids, solar flares, pandemics, super volcanoes, political instability etc. Lets close down schools so people in nursing homes can live longer??????? The lives lost from decreased fertility/fecundity may be the greatest tragedy of all…..


Michael Martinez
Aug 1 2020 at 6:18pm

How about let’s use some common sense. If you see people dropping dead around you, then yeah probably it’s a deadly virus. Otherwise no it isn’t.

Mark Brady
Aug 13 2020 at 1:23pm

Are you saying until you see someone “drop dead around you” then the common sense thing to do is assume there’s no dangerous virus? Or if I see someone with a severely compromised immune system drop dead of from an otherwise common virus, I should use common sense to determine my next step is to declare a pandemic and set my hair on fire? Maybe I don’t understand your comment.

Mark Raymond
Aug 2 2020 at 8:21am

Thanks for the interview. Absorbing conversation. I was intrigued by Russ’s comment that the role of government is to provide information so that people can make good decisions about risk. Does that mean that you think governments should allow, for example, people decide if they go to restaurants when there are high infections? In a pandemic, government should be deciding what people do and don’t do not the people themselves.

Aug 2 2020 at 10:24am

This conversation would have benefited from more clarity.  Both Taleb and Roberts almost surely mean something like “long tail” rather than “fat tail” when discussing pandemics and even that doesn’t really capture what they were trying to get at.  The point is more that events with a probability far from the mean (e.g., 4 billion dead) need to be addressed because the “effect size” (in this case, catastrophic civilization-wide collapse) is so much greater than events with probabilities closer to the mean.

Aidan T
Aug 2 2020 at 4:54pm

This was a thought-provoking discussion, and it made me change my mind on a few things, but I was disturbed by two things:

I am surprised that nobody asked the question “compared to what?” For example, when considering the risk to visit elderly parents, nobody compared it with the risk of not visiting them. For an 80-something mother is Covid-19 or loneliness a bigger threat? I suspect that it differs for each mother, but the question is a trade-off and not a binary. I am surprised to have to remind Econtalk of that.
Nobody applied Taleb’s great insight that you need to avoid the “absorbing barrier” to institutions. For example, every professional orchestra in the world is currently effectively bankrupt, and the solution can only be resuming performances to a paying audience (as I recently observed in Madrid, Spain). Is it worth it? Hell, yes. This involves some risk, but it is not “gerontocide” to advocate for it. Nobody wants anyone to die, but the virus gets a say too, and a world without public performance or social gathering is so dismal that we should take reasonable risks to preserve them.

Aug 2 2020 at 8:02pm

I have heard many discussions about climate change where arguments that it is a serious threat are dismissed-“the models aren’t good, it’s hubris to think you can understand the models, basically you can’t prove that it’s happening to some gold standard, so I can dismiss what you’re showing me”.

When Russ says:

They said, ‘It’s not scientific. There’s no evidence. There are no studies. There’s no clinical data on mask-wearing.’ And, I’m thinking, ‘Why is clinical data the only kind of evidence that is acceptable?’There’s no clinical data on whether it’s unsafe to jump from an airplane without a parachute at 35,000 feet. Right? So, when someone says, ‘Should you put on a parachute or not?’ you could say, ‘Well, there’s no evidence that it helps.’

I thought “yes, this is the frustration-a good descriptive idea is dismissed with “the evidence doesn’t meet my standards” And then this exchange:

You have suspicions. You don’t take chances. You refuse the ship.if there’s a 1% risk that the water has the bubonic plague in it, you don’t drink 1% of it, thinking, ‘Well, it’s probably okay.’ Or, ‘Less that 1%, it’s probably good.’Nassim Nicholas Taleb: There we go. There we go.Russ Roberts: You just don’t drink it. You go to zero. You go to the corner, we call it in economics.

Why don’t we “go to the corner” on carbon emissions? The black swan scenario that scientists are pointing to would merit that.

Captain Crunch
Aug 2 2020 at 9:46pm

I think that Taleb ignores his “skin in the game” principle. It’s easy for the governors and mayors to arrogantly proclaim “shut everything down.” They (and their government cronies) continue to receive their fat government (i.e. taxpayer provided) paychecks, benefits and perks while they put millions of people out of work and out of business. He also discusses the “precautionary principle;” as far as I understood, he means that drastic actions are merited because the downside of not taking action is millions of people dead. But on what is he basing his model & prediction? I suspect he used Neil Ferguson’s by now much discredited model. In years past, for various claimed “pandemics” his estimates were off by orders of magnitude (and that was without lockdowns); in some cases Ferguson predicted millions of people dead and it turned out less than 500 died.

Michael Byrnes
Aug 3 2020 at 6:37am

For example, when considering the risk to visit elderly parents, nobody compared it with the risk of not visiting them. For an 80-something mother is Covid-19 or loneliness a bigger threat? I suspect that it differs for each mother, but the question is a trade-off and not a binary.

One of the problematic issues here is that when you make a choice such as this, it isn’t just you and your elderly mother who must bear the consequences. By engaging in practices conducive to viral spread, one increases not just his own risk but everyone else’s risk.

Christine Thomson
Aug 4 2020 at 1:05pm

Russ, wonderfully informative interview, as usual. I am a better person, and wiser, because of your podcast.
A side comment you made to Nassim Taleb in the beginning of your podcast re your Dad refusing to wear a bike helmet. I’m sorry you didn’t point out, to your dad or to your podcast audience, the potential costs to others of a possible accident resulting in brain damage that a helmet would likely spare. Who pays those costs, beside your dad: hospital/doctor  bills? (possible life-long) home care? that would not have occurred if a helmet was worn. As I Once told my husband, a bicyclist who eschewed helmets; “you have a right to make your own choices regarding your own body. But what you do with my husband, I want to have a say in it.”



Doug Iliff, MD
Aug 5 2020 at 3:48pm

Adding -cide to any discussion tends to be conversationcidal, unless it is referring to inanimate objects; fungicide is OK, but infanticide is not.  Once Taleb tossed out geronticide, the exchange veered off into a cloud of undefined gassification.

Geronticide is murder.  Plains Indians practiced that, in a passive sense, to elders unable to follow the bison.  Is that what Taleb was suggesting that Western Europeans and Americans are doing if we think that the value of one Quality Assisted Life Year (QALY) is somewhere south of infinity?  If so, this would have been a discussion worth having.

People don’t like to think about putting a price on human life, and yet we do–behind the curtain.  When mammograms were pioneered, the thought was that it was worth $50,000 to achieve one QALY of survival.  Now, I understand, it’s something like $200,000 for a new procedure or test.  Without any debate, it turns out that environmental interventions cost around $9M per QALY.  Taleb suggested $1B, but Russ didn’t try to pin him down on what he was referring to, or if he were serious.

At some point in the future, we’re going to get a general idea of what we spent for one QALY from the governmental response to COVID-19.  It’s going to be very high, because the virus is mainly killing people very close to their expected lifespan.  In contrast, the response to the Spanish flu pandemic a century ago would have been cheap for two reasons– it was especially lethal for young people, so every life saved would yield a bundle of QALYs, and the social response did not cost a lot of money.

This is a discussion we need to have as a society before the next pandemic arrives, with a lot of input from economists, physicians, and ethicists.  Politicians will hide their heads in the sand, so the input needs to be explicit.  The financial cost of the response– measured not only in dollars, but in the indebtedness we bequeath to our descendants– must be considered.  In addition, it must be balanced against to the direct medical and social costs of the response itself, which will prove to be a challenging endeavor:



Steve Ravet
Aug 17 2020 at 1:35pm

Just wanted to say thanks for this post, it says exactly what I’ve thought about, but in a much more forthright and meaningful way than I would have been able to put it.

Shawn Buell
Aug 10 2020 at 9:57am

The trouble I frequently have with Taleb is encapsulated in his assertion that the purpose of Science is to describe “properties”; this is flatly wrong.

“Properties” are qualitative, but scientific data emerges from “quantitative” measurement. The qualitative is a sub-set of the quantitative.  The purpose of science is to create models of systems, matter and energy which have predictive power over those things’ future behavior.

This is why his claims about things like IQ don’t hold up to scrutiny; if we had a scientific theory which could predict the outcome of a complex system with 40% or greater accuracy (knowing nothing beyond a single data point; in this case IQ) we would be thrilled. It’s hard to find single traits of complex systems that have as strong an R^2 coefficient as that in any social science.

Dave Neil
Aug 13 2020 at 1:02am

Many commenters seem to be missing what I took to be Taleb’s main point.  When a novel disease meets the definition of an emerging pandemic and has killed 1000+ you are faced with devilish dilemma.  You can’t properly assess the risk of whether that disease will go on to kill 10 more, 1000 more, 1,000,000 more or 1,000,000,000 more.

August is a long way from Feb 11 when China reported 1000 dead and we knew next to nothing about COVID-19.  In fact, that seems to be the exact day it was named COVID-19.  The WHO did not declare it a pandemic for another month.  We didn’t know things like the age risk curve.  We didn’t know if a vaccine would be a short, medium or long term intervention, but with recent phase 1/2 results, vaccines look more likely rather than less.

Taleb is arguing that we need to do much better, much faster because by the time many countries reacted, it was waaaaaay too late.  From our vantage here in August, we seem to have dodged the mega-bullet this time around, but how will we react next time?  When the pathogen might be an order of magnitude more deadly.

One question I would have asked Taleb is what criteria you can use to ‘stand down’ your response?  If 1000 dead is a heuristic to trigger a crisis response, what determines when you have enough information to assume the more extreme outcomes will not eventuate.  You will obviously never have certainty, but we will need something to guide us.  Do we know enough about it today to conclude this is not the mega-bullet?  I feel like we do, but that’s just my 2 cents and I’m not sure Taleb would agree given his comments.

I also would have asked how practical ‘quarantines’ would be as a crisis response.  It’s worked well on a small scale for some repatriations and small island countries.  But is it doable for somewhere like NYC/London/Paris/Sao Paolo?  Do we stop 95%+ of flights to NYC in that week after Feb 11?  You can’t quarantine hundreds of thousands of people so you’d have to stop them travelling altogether, almost immediately.  Big call to make on Feb 11.  And will we be willing to do it over and over again, once every few years as these diseases continue to emerge?  Seems highly unlikely…

Mark Brady
Aug 13 2020 at 12:50pm


I’ve never been disappointed in an EconTalk up to now.

You always, or used to, make it a point to disagree with points you agreed with to ensure a balance. Even if a point was correct on balance, you would seek to expose the unintended consequences of an position. In this episode, you agreed with ridiculous points and never once stopped to point out the unintended, but very significant, consequences of our current times.

Let me state, I wear masks whenever I am supposed to, I veer into the street when walking to maintain a minimal distance. This isn’t political.

You never mentioned QALY once. You never mentioned the all of the Death Above Expected which are not COVID caused but rather caused by the response. You mentioned that we have no scientific evidence that jumping from a plane without a parachute… mic drop… wearing masks is a good thing. I would have appreciated illumination about whether I’m wearing the mask to protect me or thee. If there are certain masks that are good or bad (Duke just revealed some ‘masks’ actually make matters worse. Do they help if I’m 10 feet from you? 3? 20? If they are so great, should we wear them forever and ever? Is this good for the species in the long run?

What do nuclear weapons have to do with pandemic protocols? That seems more like a political point than an economics one. Trump rallies are super spreaders? Is there evidence that’s true? If there is then that’s fine but I have yet to see it. The 7-day moving average for OK shows a pretty clear “knee” around June 1 when the state “opened” and a straight upward trajectory right through the 20th when the rally was held.

I never expect post hoc ergo propter hoc reasoning from you without a disclaimer.

Then, the end was completely incomprehensible. The history of Geronticide? Does that has something to do with the current pandemic? Is someone suggesting that we should let the virus kill off the elderly? No context, no QALY… Was this innuendo?

You’re my solace in these overly political times. I need you to moderate a discussion. Two folks having a self-congratulatory conversation about how superior they are to the Cretans who question the officials who admittedly lied to everyone is not worth my time.





Jared Szymanski
Aug 14 2020 at 4:17pm

Taleb is ignoring the lessons of Antifragile. For almost all of healthy people under age 65 or so, COVID19 is not a “ruin” type of event. The risk is no greater and may be less than risk from many other pathogens that we face daily. On an individual level, as long as the risk of ruin is so low, getting sick occasionally actually improves one’s antifragility by training immune system cells, effectively tuning it up. This explains why some people who get COVID19 experience no symptoms, because they may have had a cold in the not too distant past caused by a similar coronavirus.

On a population level the same principle applies. A population that experiences a high rate of infection will be more antifragile as immunity is built up. Why don’t we focus more on antifragility than tail risk?

Steve Ravet
Aug 16 2020 at 7:54pm

I’m a little late to the posting here, I don’t know how  an NNT episode slipped past me. I think I heard that 1) pandemics are fat tail events, so we must respond to all of them, and one of the strategies is shutting down international travel, and 2) that any virus that kills more than 1000 people (in what population size?) is considered a possible pandemic. Do these two things really go together?

I think that in the episode there are assumptions that go unaddressed (ie that the only goal is preserving life, versus living life). I also didn’t hear any discussion of opportunity costs (for example living constantly under the threat of international travel being shut down). What is the cost to economic livelihood of living in the shadow of quarrantine? How is that balanced? Will it be abused? Is it appropriate to make that decision for 320 million people at once or should it be states? Or counties?

Finally there were some simple mistakes like saying that mass transit is a superspreader. Doesn’t most of Asia refute that claim? There is a wide variety of responses in terms of mask wearing (or not), lockdown (or not), social distancing (or not) and it seems that infection rates and deaths don’t clearly correlate with any of them.

NNT episodes are always interesting and thought provoking, and this one as well. I would echo the comment I’ve seen in other posts that the arguments seemed to get shakier as the conversation went on.  BTW I’m currently listening to the Ioannidis episode from 2018 (another one that slipped by) and enjoying it as well. Please have him on soon as I find his contrary ideas interesting.

Russ, as always, keep up the good work!

David Moeller
Aug 22 2020 at 11:12am

Taleb says that we should be careful during the early stages of a pandemic because it might be a really big pandemic. There’s uncertainty in the sense that we don’t know whether it is going to kill 100,000 people or 1 billion people. It’s true that we don’t know which kind of pandemic it is. But the unchallenged assumption is that we can significantly improve the outcome by human intervention. This might be true, and sitting here in August 2020, there’s some evidence that it’s true at least in the short run. Some countries that took dramatic actions do seem to have fared better so far in terms of virus case and death tolls. But it seems that it’s way too early to say conclusively. Maybe there are historical examples of successful interventions that I’m not aware of.

Comments are closed.


EconTalk Extra, conversation starters for this podcast episode:

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:

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

TimePodcast Episode Highlights

Intro. [Recording date: July 3rd, 2020.]

Russ Roberts: Today is July 3rd, 2020, and my guest is philosopher and author Nassim Nicholas Taleb. He is the author of the Incerto, his collection of books, Antifragile, The Black Swan, Fooled by Randomness, The Bed of Procrustes, </>Skin in the Game. This is his ninth appearance on EconTalk. He was here last in March of 2018, talking about Rationality, Risk, and Skin in the Game.

Our topic for today is the COVID-19 pandemic. I want to remind listeners, we're recording this during the pandemic. Audio may be below our usual standards. This episode will be released on YouTube with video, we hope. So, please go to YouTube, and search EconTalk, and subscribe. I want to thank Plantronics for providing today's guest with a Plantronics Blackwire 5220 headset.


Russ Roberts: Nassim, welcome back to Econtalk.

Nassim Nicholas Taleb: Yeah. Thanks again for inviting me. And, one little detail. I'm not technically a philosopher: I'm just someone interested in probability. So, whatever field is attached to that would be my field, but not quite philosophy. But, anyway.

Russ Roberts: Well, I like to also call you a flaneur, but that's even more pretentious.

Nassim Nicholas Taleb: Flaneur is my profession. Flaneur--you don't have a--but, actually, it's pretentious now, but it used to be a quality in Roman time, [foreign language 00:01:50].

Russ Roberts: Yeah?

Nassim Nicholas Taleb: Doing nothing. Or, not being guided by any constraint. That's pretty much what it is about. Yes.

Russ Roberts: And, in your usage of it, it often denotes wandering on foot, without a purpose.

Nassim Nicholas Taleb: Yeah, but if you try to make it productive, you're going to fail--

Russ Roberts: Correct.

Nassim Nicholas Taleb: Anyway, aimlessness is the idea, whether intellectual or physical.

Russ Roberts: I want to begin with a personal reflection. My father, Ted Roberts, used to say that even though there are snakes in the world, it's good to go barefoot so you can feel the grass between your toes. That was a statement about trade-offs that he embraced and shared, and I embraced eventually. My dad hated the popular wisdom of what he would consider obsessive safety. He never wore a bike helmet because he liked the feeling of the wind in his hair. My dad taught me that value of trade-offs at a very young age.

But, while I am his son, I am Nassim Nicholas Taleb's student, and I learned from you, Nassim, about the risk of ruin--that trade-offs become relevant when you're out of the game, either literally, as in wiped out in gambling, or figuratively, as in being dead. So, being 65, I have been exceptionally cautious over the last few months. Would you say I've done the right thing in responding to the pandemic?

Nassim Nicholas Taleb: Yeah. Okay. So, what I'm saying is that caution and prudence are perfectly compatible with your father's statement.

So, the central idea that risks of variation is extremely different from risk of ruin, particularly at a systemic level. Like, risk of car accidents doesn't impact--whether it goes higher or lower--doesn't make a difference in a risk of ruin for society. The example I gave in Antifragile at the individual level is: if I jump one meter, I'm probably going to get stronger. Actually, if I don't jump often, I will get weaker. But, if I jump 20 meters--you know, it won't happen.

So, there is a difference between medium-size variations at individual level and systemic at the system level. You must survive.

And, this is something we discussed last time, last chapter of Skin in the Game, which is something about the absorbing barrier, that conditional on being here, odds are there's some properties we didn't have. We are paranoid for large-scale risks and certain classes of risk. And, what I've tried to do with my collaborator is figure out what are the systemic risks we should be avoiding, which is liberating because it allows us to take a lot of risks elsewhere.

Russ Roberts: But, do you think that--just thinking about myself and my personal risk--not the societal risk--should I have been extremely cautious, as I, in fact, have been during this pandemic?

Nassim Nicholas Taleb: Yes. We actually wrote a thing on it, Joe Norman and myself, as a follow-up: that even my archenemies--people who consider themselves my archenemies, liked--like Sam Harris--

Russ Roberts: Well, we don't have to name them--

Nassim Nicholas Taleb: No, no, no, no. Sam is a nice guy because he went in and said, 'Listen. I hate--'. And, this guy hates me. He is my archenemy, but you know what? You got to read that piece.

And, in the piece, we follow the following very simple reasoning. It is very selfish for you to not reduce your risk, because a contagious disease has different properties. So, let me explain. Someone came up with a statistic that being hit by a car represented much larger risks for people. But, at the time it was true--

Russ Roberts: Than the pandemic --

Nassim Nicholas Taleb: Than the pandemic. Than COVID. But, then I said, 'Okay. But, if you don't behave conservatively, you will increase that risk dramatically because the collective risk doesn't scale linearly from individual risk, unlike other risks that do effectively have some scale transformation.' So, if I don't pay attention, the risk will be high for everybody else.

It's just like wearing a mask. I wear a mask not for myself, but because a person I'm going to infect will infect maybe 10 more, or, I mean, maybe up to 10 more--

Russ Roberts: Some number. Yeah --

Nassim Nicholas Taleb: Maybe up to 20 more. Some number more, maybe on average. So--

Russ Roberts: Depends if they sing in a choir or go to a urologist.

Nassim Nicholas Taleb: Yeah, exactly. Exactly. So, you wear a mask not just to protect yourself, but for the systemic effect.

And, very few understand that thing. And, last time we discussed rationality, we spoke about scaling. Like, some things may be perfectly rational for society, but totally irrational for any individual in it.

Russ Roberts: So, in the case of wearing a mask, if we all--and we'll get to masks in more detail because we've come to a moment in American history deeply depressing to me, where wearing a mask has become a partisan, ideological issue rather than a scientific issue.

But, the point is, is that wearing a mask has external effects that are beneficial. It's like being a nice person. You bear a very small cost. It is uncomfortable and hot. And, by doing so, you reduce the chances that you are asymptomatically spreading a deadly disease to, potentially, as you say, dozens of people.

Nassim Nicholas Taleb: Yeah. Yeah. So, it becomes a duty, sometimes, to act against your own interest--

Russ Roberts: A moral imperative--

Nassim Nicholas Taleb: It's a moral imperative. And, masks are the cheapest. I mean, think about it. You're inflicting huge economic cost by not wearing a mask. Because--I've computed a few things about convexity of masks and errors made by bureaucrats, initially. Because initially the World Health Organization [WHO], the CDC [Centers for Disease Control], all these people were against mask-wearing.

And, we said there were two things they did not figure out. The first nonlinearity, the first one, is they didn't think--again, they don't scale--they did not think that if I wear a mask and you wear a mask, we're not reducing the risk by P, the protection, by a mask, but by P-squared. See?

So, the idea--so, for example, if I have 50% probability of being infected wearing a mask, and you have 50%, the total would be .25. So--but, there's another thing people did not pick up. And, we saw it in a lot of papers. And, that is even more central: that, if I reduce the viral load by half, I don't necessarily decrease the probability of infection by half. I may decrease it by 99%--

Russ Roberts: Right. And, there may be a threshold level.

Nassim Nicholas Taleb: because of convexity--

Russ Roberts: A threshold level.

Nassim Nicholas Taleb: There is an S-curve. So, if you're in a left side of the S-curve, it must be that you're convex. Everything is convex initially, and then concave at the end. So, that's the nature of the S-curve.

So, these are the things that were missed by people who analyzed masks, initially, scientifically. They just looked at viral load reduction.


Russ Roberts: So, back in January of this year, January 26, you wrote a paper with Yaneer Bar-Yam and Joe Norman warning about the pandemic. And, it got some attention. But, if you were in your shoes then, looking back at that moment, which was--in America, nobody really was paying any attention to this--what were you hoping would happen when you wrote that paper? Besides raising an alarm, which you did.

Nassim Nicholas Taleb: Okay. No, no, no. Two things, and it was depressing that we had to do that. Our team was basically monitoring for pandemics because, in my opinion, they are the fattest tails, and fat tails means that these things could really go out of control. And, we haven't had many pandemics in history, but we've had very lethal ones. They represent an existential risk. And, we knew that, so you have to stop some things in the egg, when it's cheap to do so.

And, when Ebola started, people were not reacting to it properly. They were comparing a process that is multiplicative to risks that were in a Gaussian distribution. So, here we had fat-tailed risk--Ebola or any pandemic--versus people dying in their swimming pool, people falling from ladders, and stupidities like that. They do not compare--

Russ Roberts: Rare events--

Nassim Nicholas Taleb: Yeah, but you cannot compare these two, because one is multiplicative, and then the other one follows what I call 'Chernoff bound.' So--

Russ Roberts: Yeah: when I drown in my bathtub, you don't drown in your bathtub as a result.

Nassim Nicholas Taleb: Exactly. Exactly. So, they're not multiplicative.

And, then the way to look at it is they have extreme value properties that are very different. If I tell you a billion people died, you know they did not die drowning in their bathtub in any given year. Whereas if I tell you, 'A billion people died. Try to guess,' odds are it's going to be either a nuclear war or a pandemic. More likely to be a pandemic. It's actually fattest tail.

So, just to alert people to the nature of pandemics, that's the thing we have to worry about the most. And, we failed during Ebola because people kept using--especially psychologists--rationality arguments.

And, our argument was twofold. One, the multiplicative notion; and the second one is that rationality point that I just made--that rationality doesn't scale.

And, this is completely absent: the fat-tailedness and the scaling are absent from the Decision Science literature, which effectively makes people doing the right thing look like they're irrational.

Russ Roberts: Right. And, no one wants to look irrational.

Nassim Nicholas Taleb: Exactly. So, the literature coming out of that tradition of rationality and all of that is completely bogus. So, we knew that from the Ebola days, four or five years ago. And, we were waiting for the next pandemic. And, sure enough, we saw the sign. In January. And, then we--

Russ Roberts: In China.

Nassim Nicholas Taleb: In China. And, at the same time, I started studying a little bit history, in preparation for a paper that came out in Nature: Physics with Cerillo, showing, effectively, that pandemics are the fattest-tailed thing, and that you got to kill them in the egg.

So, in preparation, I read some history, and I discovered that, effectively, Eurasia--I mean, let's call it the Old World; let's call it Eurasia--dealt with pandemics in a very effective way. It was part of the economics of the time.

Russ Roberts: Explain.

Nassim Nicholas Taleb: Because--people, when they talk about plague, think that it's an instant phenomenon that killed a lot of people and disappeared. We had two plagues. The first one, Justinian's Plague, and then the second one is what is commonly known as the Major Plague in Europe, in the 13th century. 1300s, 14th century, 13th or 14th century. And, people have the illusion that, 'Hey, it came.' In fact, no. The previous plague stayed. And, it was accelerated by commerce along the Silk Road. And, as can be expected, commerce brings germs. And, people adapted to it, because the plague stayed there for 300 years.

And, to give you an example of what happened, in 1720 a ship left Smyrna with merchandise coming from the East. Smyrna, now in Turkey, at the time in Ottoman Empire, Eastern. It was considered the Levant at the time. Went through what is now Lebanon, Cyprus. Went to Genoa. The Genovese didn't let it in because they had something called lazaretto, which was a quarantine for all ships. When they suspected the Plague, it was enforced. Talking around 1720. The ship was owned by the Mayor of Marseilles. Okay? And, he bypassed the quarantine rules. Half the population of Marseilles perished.

Russ Roberts: You mean of Genoa.

Nassim Nicholas Taleb: No. Marseilles. Genoa refused it.

Russ Roberts: Oh.

Nassim Nicholas Taleb: Okay. Because they were smart, so you think about it--

Russ Roberts: Oh. It went to Marseilles. Okay.

Nassim Nicholas Taleb: Yeah, it went to Marseilles.

So, you think about it, that was the mechanism. Mechanism: The ship comes in. You have suspicions. You don't take chances. You refuse the ship.

Russ Roberts: Wow.

Nassim Nicholas Taleb: And, the cities--you look at it. You're in the Black Death, or that period, when they--some cities shut down. They did not accept any foreigner or any outsider, and even the residents coming back had to go to lazarettos, or the equivalent of quarantine.

So, quarantine was built in.

Russ Roberts: Was a standard operating procedure.

Nassim Nicholas Taleb: Procedure. Standard operating procedure, activated whenever people suspected something. And, the Ottoman Empire--you see, the Ottoman Empire--had a border with the Habsburg at the time. And, whenever you crossed from one to the other, you quarantined.

Russ Roberts: And, it's worth pointing out that 'quarantine'--its root is Italian, or Latin, which is 'forty'.

Nassim Nicholas Taleb: 40. It's 40 days. It was 40 days.

Russ Roberts: So, I think these people were--

Nassim Nicholas Taleb: Yeah.

Russ Roberts: Which was a pretty serious--we do two weeks here in modern times.

Nassim Nicholas Taleb: But still, two weeks works. I mean, two weeks is good enough. I mean, whatever reduction of connectivity is very good.

So, what happened is you have to realize--and this is something I saw when writing The Black Swan--that we have more connectivity today than we did before, so therefore things that are multiplicated will be vastly more multiplicated today. And, you have more of a winner-take-all, just like we have economic winner-take-all with Google thanks to technology.

I notice Pavarotti, for example, the singer--the opera singer. Because you had the television, okay, he really--

Russ Roberts: And, the radio.

Nassim Nicholas Taleb: And, the radio--

Russ Roberts: I mean, album. CDs [compact disks]. Yeah.

Nassim Nicholas Taleb: Yeah. So, you had--particularly for opera, where you need the visual--you had a winner-take-all effect, where a few opera singers were making hundreds of millions of dollars, and the rest were starving, or working for Starbucks, or singing at birthday parties.

So, you have the same concentration, the same thing will happen, with diseases, that you would have a winner-take-all disease taking over the planet. And, we have super-spreader events that cause even more fat tails.

To give you an example, urology conference in Las Vegas. And, in this case, what happens in Las Vegas doesn't stay in Las Vegas, gets redistributed to 200 countries. So, people come in, they get something, and then they distribute it. So, that, you did not have on the Silk Road. It was much slower, 30 kilometers a day maximum, you see.

And, because you had that barrier of religion between the Ottoman Empire and what's considered now as Western Europe, because you had that barrier, it was porous but not as porous as the world is today. So, you don't have natural limits to things.


Russ Roberts: So, back in January, when you were warning about this risk--of course, it hasn't taken over the planet exactly, and we'll talk about that in a little bit. It's been, obviously, devastating in different degrees in different places. But, what do you think we should have done in January? And, the right way to phrase that is, when the next pandemic hits, and we get COVID-23 instead of COVID-19, what's the right response at the national level? and then at the state level? the local level? and the personal level?

Nassim Nicholas Taleb: We wanted systematic quarantines, because the fact of having quarantine for people coming from China is foolish, not knowing that they can come from somewhere else, okay--

Russ Roberts: Yeah. Europe.

Nassim Nicholas Taleb: and the fact that it came from somewhere else.

Russ Roberts: Yeah.

Nassim Nicholas Taleb: And, the Chinese, actually. And, then, also you want to have the World Health Organization removed. It's a bureaucratic organization that has been devastating for mankind because of its incompetence. They prevented people from wearing masks. So--

Russ Roberts: Now, since then, the CDC did the same thing.

Nassim Nicholas Taleb: Yeah.

Russ Roberts: They have since confessed that they did that. At least that's what they're saying now, that they did that because they were worried about a shortage. So, they lied to people, if that's true; degraded the quality and credibility of their organization, and resulted in--

Nassim Nicholas Taleb: No, no. It's not just that they lied. They're incompetent. They didn't understand the compounding.

Russ Roberts: Well, maybe, yeah.

Nassim Nicholas Taleb: See. Now they're saying, also, they did not realize that--you know, they're bureaucrats. In their opinion, only the government can make masks. They did not realize that there is a market, that the minute you tell people, 'Hey, wear masks,' that people would find stuff in their closets that would work perfectly as a barrier, and we wouldn't be here.

What I think today--we thought of the mask situation late in the game. Initially we bought the argument that it was not transmitted by that mode. We thought it wasn't contagious. But, very quickly we figured out the mask situation, when Taiwan--when we got a statistical hint there--why is it that Taiwan, that has a lot of traffic from Wubai--

Russ Roberts: Or Wuhan.

Nassim Nicholas Taleb: Wuhan--was completely spared?

Russ Roberts: And it's claimed that Hong Kong has been completely spared as well despite its [inaudible 00:20:30].

Nassim Nicholas Taleb: And, look at the mask situation. So, you start getting statistical hints.

And, then, of course, we [?] convexity argument and realized immediately that they missed it. So, was very embarrassing. But, that's not the point. The idea--

Russ Roberts: You said 'missed it.' I thought you meant, 'misted.' Like, because of the virus. But you meant: 'missed it.' Sorry.

Nassim Nicholas Taleb: Missed it. And, then also there's something. They said, 'Well, there's no evidence masks work. Don't use them.' They cannot--.

Russ Roberts: I'm still hearing it. People are still saying it.

Nassim Nicholas Taleb: Yeah, 'There's no evidence mask works.' Hence, wear it, if you don't know. The whole idea of the Incerto, the central idea of the Incerto, is that when you have uncertainty in a system, it makes your decision-making much, much easier.

Russ Roberts: Rather than harder. It seems harder.

Nassim Nicholas Taleb: Yeah. It seems. And, let me give you [inaudible 00:21:18] example. This water: If I tell you, 'Well, I'm not certain about the quality of this water,' what would you do? Would you drink it?

Russ Roberts: No.

Nassim Nicholas Taleb: Okay. Very good. Okay. Very simple. I tell you--we have uncertainty about the pilot's skills. He could be excellent, he or she [?]--

Russ Roberts: Or incompetent.

Nassim Nicholas Taleb: Okay. We're not sure. We have uncertainty. What do you do? Right? You get off the plane.

Russ Roberts: No, you get on the plane for a while. The reason I like this example is that, if there's a 1% risk that the water has the bubonic plague in it, you don't drink 1% of it, thinking, 'Well, it's probably okay.' Or, 'Less that 1%, it's probably good.'

Nassim Nicholas Taleb: There we go. There we go.

Russ Roberts: You just don't drink it. You go to zero. You go to the corner, we call it in economics.

Nassim Nicholas Taleb: There we go. And, people behave that way. We're going to see how people's built-in instinct is vastly superior to what the CDC had, what a lot of scientists, what these decision-making professors, the Nudge group in the United Kingdom. All these people are completely incompetent when it comes to basic things that your grandmother gets.

Russ Roberts: Yeah. Your grandmother is right, often.

Nassim Nicholas Taleb: This instinct helped us get here.

Russ Roberts: Yeah. If she weren't right, we wouldn't be here.

Nassim Nicholas Taleb: Exactly.

Russ Roberts: Yeah.

Nassim Nicholas Taleb: There we go.


Russ Roberts: So, this--I'm going to talk about masks for a minute because what I found fascinating about it, in thinking about these kind of issues of risk generally: If you tell people that if you wear a seatbelt you might drive less carefully, they laugh at you. They say, 'Nobody pays attention. That's ridiculous.' But, those same people, for some reason, when we started to suggest that wearing masks was a good idea, 'Oh, no, no, no. There's all these other facts. You might touch your face more. You might think you're safer than you really are,' which is like the seat belt argument. They said, 'It's not scientific. There's no evidence. There are no studies. There's no clinical data on mask-wearing.' And, I'm thinking, 'Why is clinical data the only kind of evidence that is acceptable?'

Nassim Nicholas Taleb: Okay. Yeah. So--

Russ Roberts: It's absurd.

Nassim Nicholas Taleb: After 20 years of working with uncertainty--and, more than 20 years, of course, but, 20 years of publishing to it, I finally wrote a piece that has a sentence that summarizes everything I've struggled to explain. In my debate with John Ioannidis, I came up with a sentence: 'Science is not about evidence. Science is about properties.' It's not about [crosstalk 00:23:59] forecast.

Russ Roberts: Explain. Explain.

Nassim Nicholas Taleb: So, science is about understanding something. So, it's not about details.

For example, if you have absence of evidence, all right, it's still scientific to describe a process--

Russ Roberts: That this virus goes in your nose, and it comes out of your nose. So, if you cover your nose, and the guy across from you covers his nose, maybe it'll reduce--I mean, that just seems--

Nassim Nicholas Taleb: Exactly. It is scientific. It's scientific to say, 'Let's wear masks.' Because, if your decision--if you map your decision-making abilities or the decision-making--sorry--the rationale--in a scientific way, then you would get results that are very formal. But, this naive, evidence-based science is not there in sophistication yet, to grasp that. And, we're all interested in details about forecasting. But, science is not--

Russ Roberts: Hey, but wait a minute--

Nassim Nicholas Taleb: about forecasting single events. Science is about understanding the properties of stuff--

Russ Roberts: Let's talk about that. Yeah.

Nassim Nicholas Taleb: which may or may not include forecasting.

Russ Roberts: But, one second. I just want to add this famous, sort of onion-like example, which is: There's no clinical data on whether it's unsafe to jump from an airplane without a parachute at 35,000 feet. Right? So, when someone says, 'Should you put on a parachute or not?' you could say, 'Well, there's no evidence that it helps.' But, we understand gravity. We understand logic, and we would understand that you don't need a clinical study, and--

Nassim Nicholas Taleb: And, you understand asymmetry. You understand asymmetry.

And, this is what I'm saying, is that integrating in your decision-making rigorous science includes taking into account the asymmetry, the payoff assymetry: that, the harm you're going to get from wearing a parachute is vastly inferior, potentially, than the benefits you may get for it. That's it. Simple.

Russ Roberts: Right. It's the same with the mask, right? The mask is a little bit uncomfortable. It's a little bit hot. It costs $8. It's a bargain.

Nassim Nicholas Taleb: What do you mean it costs $8? You can make masks for 45 cents, using--

Russ Roberts: Yeah. 50 cents.

Nassim Nicholas Taleb: Listen--coffee filters work.

Russ Roberts: Yeah. You can also--I probably mentioned this before--get on YouTube and Google, or search, for 'sock, mask, scissors,' and you'll find a way to make a mask just with a pair of scissors and a single sock. It's phenomenal.

Nassim Nicholas Taleb: Yes.


Russ Roberts: So, let's talk about forecasting, because a lot of people have lost some unfortunate credibility because they picked numbers. They said--famous example in England, Neil Ferguson--not my colleague at Hoover [Niall Ferguson--Econlib Ed.]; the other Neil Ferguson--said that up to 2 million people could die from the pandemic. And, you could argue, 'That was okay because he said "up to."' Which is meaningless, of course. It doesn't mean anything. It just means it could be really bad--which it could have been.

And, you could argue it was important for him to scare people into being more cautious, and that was the payoff for this. It was that people became more careful when they heard that number. So, what's wrong with making a forecast like that?

Nassim Nicholas Taleb: Okay. I'm sharing a screen with you here. And, a paper that started with a debate with Ioannidis. And--

Russ Roberts: This is John Ioannidis, who is a past EconTalk guest, who I'm very fond of. And, we may get from him--

Nassim Nicholas Taleb: Yeah, I used to be fond of him.

Russ Roberts: Well, we may hear from him down the road. Go ahead. Let's see the image.

Nassim Nicholas Taleb: Talebetal_25062020_Fig1.jpg So, the title is "On Single-Point Forecasts for Fat-Tailed Variables." And, let me explain it very simply, and explain why science is not about forecasting, as people think. Sometimes it's about forecasting, but not necessarily so.

You see this graph here. I have a probability distribution that is very skewed. Actually, it's not an unusual probability distribution. It's common, log-normal, which is not the worst. I mean, there is worse--

Russ Roberts: And, for those listening to audio only, we're going to describe this in a minute, without the graph. But, carry on. Carry on, Nassim.

Nassim Nicholas Taleb: The average is slightly above 7. Okay?

Russ Roberts: Mm-hmm.

Nassim Nicholas Taleb: All right. And, the--

Russ Roberts: The variable.

Nassim Nicholas Taleb: The variable. And you only observe the average 15% of the time. Something above the average. Okay? 85% of the observation fall below the average. Okay? And, what's worse, that 50% of the observation fall below 1, which is like [13? 30?] percent of the average.

Russ Roberts: So the average is--for those not listening at home: the reason I love this image, and your discussion of it, is that it highlights something important that I've learned from you, which is: Most of our casual, non-real, academic explorations of risk are based around classroom exercises like flipping a coin. And, flipping a coin, or the distribution of height in the United States--these are what are approximated by normal distributions, where most of the events occur around the middle, around the mean. So, if you flip a coin a thousand times, you're going to get between 400 and 600 heads most of the time. You could get 700, but it'll be rare. You could get 300. That'll be rare.

The point of this kind of distribution--it doesn't look anything like a normal distribution. And, if you bring your intuition from the normal distribution into this picture, you're going to make some horrible mistakes. So, just to repeat: In this case, the average outcome is 7. But, as you said--

Nassim Nicholas Taleb: A little more than 7--

Russ Roberts: A little more than 7. But, 50% of the time, it's going to be less than 1. And that's because, and just to make it clear, that's because it has a long tail to the right, and you're going to get outliers relatively often. Unlike the case of flipping a coin, where you can't get more than a thousand heads in a thousand tosses. Or height--you can't get anybody over, say, 7.5 feet tall. And, wealth, or other things that are distributed--risk of a pandemic, death rate--

Nassim Nicholas Taleb: Okay. So--

Russ Roberts: You can have big numbers.

Nassim Nicholas Taleb: Talebetal_25062020_Fig2.jpg Yes. But, now I have a second graph to show you, which is going to depress you. Or, maybe--you know, I think that understanding uncertainty makes me happy because I know there's not that many risks out there. I just worry about those. And, that makes it--

Russ Roberts: It also makes you feel smart, Nassim.

Nassim Nicholas Taleb: Exactly so. But, that makes it vastly more fat-tailed.

Russ Roberts: Yeah.

Nassim Nicholas Taleb: Okay? So, in this case, the average for this distribution--if it exists, assuming it exists--will come from one observation. So, you can have a million observation. One will determine the average.

Russ Roberts: Meaning? Explain that. I don't understand.

Nassim Nicholas Taleb: Meaning--okay. So, fat-tailed means, like, if you take the wealth in America, you realize that 1% of Americans own 50% of the wealth.

Russ Roberts: Okay.

Nassim Nicholas Taleb: Okay. And, it gives you contribution by outliers that can be monstrous, like Jeff Bezos is wealthier than--I don't know--the bottom hundred billion American, 200 millions, whatever it is.

Russ Roberts: Whatever it is. Yeah.

Nassim Nicholas Taleb: Okay. So, the concentration. It is concentrated. These pandemics have effected, are very, very concentrated, in the sense that: One observation will determine--unlike the normal that we saw earlier, which gave us the intuition, 85% of the observation are above the average--in this case, only one single observation is above the average.

Russ Roberts: One.

Nassim Nicholas Taleb: One.

Russ Roberts: And, why is that important?

Nassim Nicholas Taleb: It's important to realize that you can't work with averages, and you cannot work with forecasting. There are techniques from extreme value theory, where you can understand perfectly what's going on, have a representation of reality--

Russ Roberts: So, let me try to--

Nassim Nicholas Taleb: That mean you have representation of reality, but it's not describable by an average, nor can it be captured by testing people's forecasting.

Russ Roberts: All right. So, let's take out the screen sharing. Let's go back to you and me [Zoom/YouTube screensharing removed] you and me.

Nassim Nicholas Taleb: Okay.

Russ Roberts: I want to try to [inaudible 00:32:21] this.

Nassim Nicholas Taleb: Okay.


Russ Roberts: So, let's talk about why this is important. If I'm an expert, and you're doing a coin-flipping exhibition, and you're going to toss a coin a thousand times, and I'm called in to ask, 'What's my forecast?' So, I say, 'Well, I think he's going to get about 500.' And, I'm going to be in the ballpark.

Nassim Nicholas Taleb: Exactly. Because--

Russ Roberts: I'm going to be in the ballpark.

Nassim Nicholas Taleb: Exactly. What you just described, parentheses, is the law of large numbers.

Russ Roberts: Yeah. Right.

Nassim Nicholas Taleb: It does work for thin-tailed distributions. So, if you have large enough a sample, or large enough a repetition, you can figure out what's going on.

Russ Roberts: So, the--another way to say it--and, again, to put it into folksy terms that I also got from you--is that you care about--you don't care about the average depth of the river if you're going to walk across it and you can't swim. You care about where it is deepest. And, if the average is three feet high, but there's a large stretch where it's 60 feet deep and you're going to die, that's what you care about, not the average.

And, here you're saying that if the average death rate from this pandemic forecast, which is based on a zillion assumptions also, is a million people, that that prediction is essentially meaningless. Because there's no sense in which--there's so many factors are going to determine where we are on the distribution that aren't going to be subject to the Law of Large Numbers.

Nassim Nicholas Taleb: Exactly. So, in another words, if the real average that can come from pandemics is something like 2 billion, more; or 3 billion--right? 99% of the time you will observe something under a million.

Russ Roberts: Right.

Nassim Nicholas Taleb: Okay? So, which is--

Russ Roberts: But, that doesn't mean--

Nassim Nicholas Taleb: Which is why we should not forecast numbers.

That's the first reason why a single-point forecast is highly unscientific, naive. What I call naive, [thinkatory ?] statistical approaches. Okay.

Russ Roberts: Say that again. Let's go back to that example, because I think it really helps people.

So, the mean, the expected value, the average--you run a simulation, the equivalent of a simulation. And you run a hundred plagues, and you find out that the mean number of deaths is a billion.

Nassim Nicholas Taleb: Yes.

Russ Roberts: And, that's because there's a few times when it's 5 billion and 7 billion, and there's a lot of times when it's quite low.

Nassim Nicholas Taleb: Exactly.

Russ Roberts: And, then it becomes--the actual event occurs. And, because people are cautious or for whatever reason, we get a good draw from the urn, and the death rate is maybe only a million.

Nassim Nicholas Taleb: Yeah, and it's not necessarily because people are cautious. It's mostly, largely, because diseases, even COVID, which we think know something about it, we know nothing. Because it can mutate. Sometimes it can start mildly. We classify as pandemic that have fat-tailed events that have killed more than a thousand people.

Russ Roberts: But, the point is that--

Nassim Nicholas Taleb: And, let me tell you why this is important.

Russ Roberts: Go ahead.

Nassim Nicholas Taleb: It means that you don't have to worry about anything that hasn't killed yet a thousand people. See, we separate the classes. Something that has not killed--every time you have a virus in town, it doesn't[?] matter.

Russ Roberts: It does not matter.

Nassim Nicholas Taleb: It doesn't matter. You don't have to worry. The minute it kills--if it's in the category that it has killed more than a thousand people, then you have to worry about that.

Russ Roberts: Because it could be a fat-tailed monster from--

Nassim Nicholas Taleb: It is a fat-tailed monster.

Russ Roberts: But, it could be a monster that could kill a billion.

Nassim Nicholas Taleb: Exactly. You have to kill it in the egg--

Russ Roberts: But, I want to--

Nassim Nicholas Taleb: as we're approaching a thousand. Yes.


Russ Roberts: I want to get at this point about the naivete and unscientific nature of these forecasts. Somebody who says, 'I have simulated this. The mean number of deaths is a billion.' Then the actual thing happens, and only a hundred thousand people die. You don't want to say--he might be an idiot--but you don't want to say he's an idiot because it was an inaccurate forecast. He's a--

Nassim Nicholas Taleb: That's exactly the point. So--

Russ Roberts: You're saying he's an idiot because you cannot--I would say it differently: 'Expected' doesn't mean an 'average'; don't mean 'typical,' here.

Nassim Nicholas Taleb: Exactly--

Russ Roberts: We run those words together.

Nassim Nicholas Taleb: Exactly. We said, 'You're an idiot to forecast,' and 'He's an idiot to critique your forecast.'

Russ Roberts: Right.

Nassim Nicholas Taleb: Okay. So, this is what we said in the paper. And, we said one thing. We said, 'Science is not about accounting. It's not bean counting. Science is about understanding a phenomenon.'

Russ Roberts: A process.

Nassim Nicholas Taleb: Now, the second thing I would say is that, effectively, some people have abandoned completely that naive forecasting, and study properties in risk management very effectively. The Dutch. Because you know it's a country, it's a low country: so they are worried about the level of the water. They are not interested in the average. They're interested in the maximum.

Russ Roberts: Yeah.

Nassim Nicholas Taleb: Okay. And, then there's a whole science of extrema coming from statistics, applied in insurance, applied in risk management, applied in finance, applied in many fields extremely successfully, that look at what robust claims you can make about maxima.

And, that's the one I used in my paper with Cerillo. We used exactly the same techniques used for the expected maximum or the average maximum--which is different from the average and different from the maximum--water level. And, of course, we modified it. We had to find robust tricks. It was complicated. But, over the years, we developed a technique to deal with extreme value theory applied to some classes of phenomena--namely, wars and the fattest of fat tails of all, pandemics.

So, when we started warning about pandemics in January, it was not the cry-wolf variety. We told, 'Don't worry about anything else. Just worry about these things, things that are multiplicated, particularly pandemics, and here is the evidence.' It's not like, 'Oh,' people were supplying us with arguments[?], 'Oh, if we listened to you, you'd never make money. You'd stay home all the time.'

Russ Roberts: Stay home all day.

Nassim Nicholas Taleb: No. No, no, no. The entire point of the precautionary principle is to replicate what that old trader told me.

Russ Roberts: Which is what?

Nassim Nicholas Taleb: 'Take all the risks you want, but make sure you're in tomorrow.'

Russ Roberts: Yeah.

Nassim Nicholas Taleb: Which is how Goldman Sachs, taking a lot of risks, have survived, so far, 161 years.

Russ Roberts: Don't eat the nest egg. Don't [inaudible 00:38:53]--

Nassim Nicholas Taleb: Make sure you survive. Hopefully you'll invite Ole Peters, people who worked on ergodicity. Standard economics has flaws in it because they don't taken into account this absorbing barrier.

Russ Roberts: Yeah. The other thing I think you've emphasized over the years, which I find very helpful, is that a lot of modeling takes place because it's convenient, not because it's right. And, dealing with expected value, which is the average, is easy. And, so we do a lot of it.

Nassim Nicholas Taleb: Yeah. We're not against modeling. They are parameters for fat-tailed distribution that are very robust. We're not against modeling. We're against naive modeling. It doesn't mean that because you have a model, that it's right.

But, on the other hand, it doesn't meant that mathematics cannot be used effectively and rigorously to capture empirical phenomena. You just have to be very careful. And, the interesting thing--I just wrote this book, and actually it's out today. Okay. The Statistical Consequences of Fat Tails. And--

Russ Roberts: Fantastic bedside reading, I suspect, for people who have trouble falling asleep. It's a technical book, right?

Nassim Nicholas Taleb: Yeah. Yes. Very technical. And, it's sort of the technical Incerto. I have the Incerto for what I call the philosophical literary books, and this is sort of like a backup that I've been writing over the years. And, the first volume is out. Today.

And, the entire message of the Incerto is that a lot of things that people claim come out of statistical models, don't come out of statistical models. For example, linear regression. You cannot do it with fat tails. So, it's not science. When you violate fundamental mathematical rules, it's not science anymore.

Russ Roberts: Well, to me that's like the insight of Ed Leamer, in his essay we've talked about here before, "Let's Take the Con Out of Econometrics"--

Nassim Nicholas Taleb: [crosstalk 00:41:02]. Exactly. Exactly.

Russ Roberts: where he's basically saying that if you're constantly trying different variations of the models, of what variables you include, you can't use the standard statistical significance tests because you've violated the underlying assumptions under which they were generated.

Nassim Nicholas Taleb: There you go. That's one of them. So, I went through, like, 18 violations by saying that, for example, you cannot do linear regression if you don't satisfy the Gauss-Markov theorem. I didn't come up with that theorem. They are using that theorem as justification.

Psychologists use correlation. They cannot use correlation when you have non-linear effects, for example, because it's going to be a linear coefficient of linear association. And, stuff like that.

And, then you go deeper with the principle-components analysis, sector analysis. So, the idea is that they are violating the rules of statistical inference.


Russ Roberts: I want to talk a little bit about fat tails and pandemics to make it a little clearer. Let me see if I understand you correctly. I hear you're saying the following--because you corrected me at one point. You said, 'Well, it's not necessarily how people behave.' A pandemic of pandemics is one that is more infectious, more fatal, etc. So, we often don't know those characteristics of the virus when it's spreading. So, to get that tragic catastrophe--

Nassim Nicholas Taleb: You can tell pretty quickly. You can tell very quickly.

Russ Roberts: But the catastrophic, far-tail outcome of one billion dead--when do we need to worry about that? Is that because it'll be a particular [?]--

Nassim Nicholas Taleb: The minute you cross a thousand. Okay. We've tested. There's this magic number, whatever it is. It could be 200, 300 people. But, we said,: Okay, as a heuristic, let's study the class of those that have killed more than a thousand people in today's population equivalent in history.

Russ Roberts: Right. But, COVID-19 doesn't kill young people much, hardly at all.

Nassim Nicholas Taleb: It killed more than a thousand people.

Russ Roberts: I understand.

Nassim Nicholas Taleb: It's not that true[?]. First of all, we know nothing about COVID-19. We're, what, five months into the thing?

Russ Roberts: Yeah. Maybe.

Nassim Nicholas Taleb: And, most infected people were infected recently because it's still growing. And, you have about, what, 500,000 people died, today?

Russ Roberts: So, far.

Nassim Nicholas Taleb: So, far. There's one thing we can make. We know nothing about morbidity. Nothing about--

Russ Roberts: Explain. Explain.

Nassim Nicholas Taleb: Okay. When you get sick, okay, you come out of a disease, sometimes you come out of it stronger, but in some cases you have side--

Russ Roberts: Oh, I see what you're saying.

Nassim Nicholas Taleb: some effects, either of the treatment or the disease itself.

Russ Roberts: You would survive, but your lungs are damaged, and you're going to live much shorter--

Nassim Nicholas Taleb: A lot of young people have damaged lungs. We don't know.

Russ Roberts: We don't know.

Nassim Nicholas Taleb: [crosstalk 00:43:56]--

Russ Roberts: We don't know.

Nassim Nicholas Taleb: But that's the whole idea, is we know nothing. We know very little about COVID today. Until now--

Russ Roberts: So, the lesson, then--

Nassim Nicholas Taleb: Till now that they started treating it differently from respiratory thing. Now they understand that there could be some autoimmune effects and reactions. So, you cannot rule out the effect on young people.

Russ Roberts: And, what I'm thinking is that we've had some scary diseases--Ebola, SARS [Severe Acute Respiratory Syndrome], etc.--and not that many people died. And, I think it lulled people like myself into a false sense of security. 'Oh, these things are overblown. It's not so bad.' You understood, and your coauthors understood, that's the wrong way to think about it. What was I missing? Did this one kill a lot more people because we didn't react correctly? Because it's a different kind of virus?

Nassim Nicholas Taleb: Okay. No. Okay. The idea that you should treat, as we warned in January, absolutely all pandemics the same way.

Russ Roberts: Okay.

Nassim Nicholas Taleb: The minute they kill--and, the reason we use the number killed, because you have a lot of pandemics that are not that consequential, like the common cold. They're not consequential. So, you should treat them all as something that can potentially mutate, once you enter that class.

One thing that people don't realize is that--in hindsight, of course, you know what happened. You know that Ebola was not a big deal.

Russ Roberts: Right.

Nassim Nicholas Taleb: You know that SARS was not a big deal, in hindsight.

In foresight, when you're standing facing uncertainty, you should behave according to a protocol. What is my protocol here? The protocol is we have something to take seriously. And, let's find ways to pull out of it.

One thing I question is that people believe that fighting the pandemic is costing us money. It's not fighting the pandemic that's costing money. It's the pandemic. And, let me explain. If you're rational, would you go to a restaurant in New York City?

Russ Roberts: No.

Nassim Nicholas Taleb: Would you go to a bar?

Russ Roberts: No.

Nassim Nicholas Taleb: You would not go to a bar.

Russ Roberts: No.

Nassim Nicholas Taleb: Okay. All right. So--

Russ Roberts: Well, I would if I wore a mask, the other person wore a mask, and I didn't stand talking to him for 20 minutes.

Nassim Nicholas Taleb: [crosstalk 00:46:18]. Okay. All right. So, as a bar, then, [inaudible 00:46:18]--

Russ Roberts: It's a weird bar.

Nassim Nicholas Taleb: you have to wear a mask inside. Okay. So, what killed--

Russ Roberts: I say that because I was at a brewery yesterday, outside, in the countryside. Yeah.

Nassim Nicholas Taleb: Outside, that's fine.

Russ Roberts: Wearing a mask. But, yeah.

Nassim Nicholas Taleb: Yeah. Okay. When we don't know what's going on, in the beginning--we don't know what's going on. So people have risk aversion. They don't want to get on a plane that has 1% probability of crashing or an unknown probability of a crash.

Russ Roberts: Right.

Nassim Nicholas Taleb: People have that aversion. And, all it takes in New York City for some restaurants is a drop of 20% of revenues. That's all it takes for them to go bust.

Russ Roberts: Sure. They're already on a thin margin, anyway.

Nassim Nicholas Taleb: Exactly.

Russ Roberts: Almost all of them.

Nassim Nicholas Taleb: They're already working on such a thin margin.

And, I read a statement by the chairperson of Delta, saying, 'Hey, take care of the clients. They only make money--the plane is full, we only make money on the last four passengers.' So, there you go. Would I get on a plane? No. Now, I would, but of course--so, you realize that a lot of the economic contraction that we have is a result of uncertainty.

Russ Roberts: It's fear. No, it's mostly fear.

Nassim Nicholas Taleb: It's fear and uncertainty, and we didn't know much about it. And, that's a major cause of economic contraction. The positive trade-off between government shutdown and the business--

Russ Roberts: Not so important. We shut it down. Most people shut down. We shut ourselves down.

Nassim Nicholas Taleb: Of course. Of course. [crosstalk 00:48:01].

Russ Roberts: Because we're afraid. And, rationally so, is your point.

But, I want to come back to this understanding about risk because: why is any--this might've been different. It might not have had much of an effect. What was different about this one, relative to Ebola and--

Nassim Nicholas Taleb: It was not. It was not different.

Russ Roberts: Why has it spread so much more? Why is it so much more deadly? Is it the virus itself?

Nassim Nicholas Taleb: Ebola had some different characteristics, but you couldn't tell about this one at the time, and we still can't tell about this one, because this and Ebola can change. Ebola kills people faster and more effectively. It kills a high number of carriers. Okay? This one doesn't. But, things have changed.

Russ Roberts: Right. But, your point is that--

Nassim Nicholas Taleb: And, it could change here. And, it was Ebola--you can't really say, 'Okay, let's not worry about Ebola, because it's going to kill enough people before they--'

Russ Roberts: Spread.

Nassim Nicholas Taleb: '--before they spread it to others, so therefore it's a good disease.' You don't know. You should treat them all the same way. That was my idea, that all of them--and, we don't have to worry too much about pandemics.

Russ Roberts: Because?

Nassim Nicholas Taleb: Just to have a good mechanism.


Russ Roberts: So, we're going to come to that. But, I want to just put a period at the end of this piece of the conversation because I think I understand it better, which is, to come back to your point about statistical characteristics, you would be an idiot or a fool to look at past pandemics and say, 'Well, they really didn't kill so many people, so I'm not worried.' Your point is that, that past experience is not predictive. The average of the past 12 pandemics is a poor predictor of what to do, to expect of the next one.

Nassim Nicholas Taleb: Exactly. That's what we told the epidemiologists. And, we had an impactful paper, Cerillo and I, because it was presenting things that physicists understood, mathematicians could grasp, but epidemiologists not so much, you see.

And, then it's a mathematician who helped spread our message, not the epidemiologists, about the fat-tailedness of the process and why you should take it seriously. You should take all fat-tailed processes seriously. [inaudible 00:50:13]. We spend a trillion dollars in America, okay--give or take a hundred billion or so, no?--on defense, directly and indirectly.

Russ Roberts: Okay.

Nassim Nicholas Taleb: Okay? So, it means that we are not willing to gamble with certain things without security. You agree?

Russ Roberts: Right. Yup.

Nassim Nicholas Taleb: Okay. So, you may argue, as I do often, that this causes more instability in the world and stuff like that, and--

Russ Roberts: It has consequences--

Nassim Nicholas Taleb: Okay. This has consequences; and that it's not probably in our best interest to have--you know. Anyway. But, the point is, we spend that money. We don't have to spend as much for pandemics. And, pandemics are vastly more dangerous than wars.

Russ Roberts: So, what's the protocol? You said we should just follow the same protocol with every one of them.

Nassim Nicholas Taleb: Sure: mechanism for quarantine, reduction of--and identify super-spreaders. You don't have to lock down things. In this case, it's very easy. Super spreaders are subways, elevators, and big gatherings like the Trump convention.

Russ Roberts: Weddings. Yeah.

Nassim Nicholas Taleb: The Trump convention, for example.

Russ Roberts: Yeah. Trump rally--political rallies.

Nassim Nicholas Taleb: Political rallies. Okay. And, then if you have masks, then you can mitigate. Okay. So, okay, you can attend a game, but everybody should wear masks. Because now we understand that open air--but, when you don't know, you have quarantines and elimination of super spreaders, till we figure out what's going on.

Russ Roberts: My complaint--I have a few, but one of my complaints is that we don't know much about what's going on, even right now. And, one of the key roles for the government, to me, should be gathering information so that we could make good decisions about what's risky and what isn't risky. We don't have good testing mechanisms. The government did an atrocious job in the United States.

Nassim Nicholas Taleb: There we go. So, we spend a trillion dollars on weaponry, nuclear warheads. You can destroy Russia, or you can destroy any country you want a hundred times. But, we are still not able to have proper testing and/or proper quarantine.

Russ Roberts: But, quarantining is tricky, right? So, when China--in China, as far as I understand it, they used basically the military and said, 'You get the disease--'. They stopped people at checkpoints. They took temperature. They did testing.

Nassim Nicholas Taleb: You don't have to go that far. The--

Russ Roberts: What quarantining would you support?

Nassim Nicholas Taleb: No. Quarantine [inaudible 00:52:53]. Listen, the Federal government's job is not domestic quarantine. It's international travel. At least check people coming in. Right?

Russ Roberts: Yeah.

Nassim Nicholas Taleb: They were not checking people--

Russ Roberts: Still not doing it, I assume.

Nassim Nicholas Taleb: Still not doing it. Okay. Check people coming in. That's what they do. And, then you quarantine people who are suspicious cases.

And, states should probably do the same. You're allowed to block New York City and say, 'Okay. We can check you,' like some Italian towns have done.

This is the victory of localism because it showed that, just like the Plague. The Plague was a victory of localism. Some towns fared a lot better than others. And, here some states will fare better than others.

But, the idea that it's cost you money is bogus because you can do it at very cheap costs.

Russ Roberts: Well, it does cost you money, but there's different ways--there's different amounts, depending on what you choose.

Nassim Nicholas Taleb: Yeah, exactly. No. It does cost you money, but I don't think it should. And, now we're going to adapt to an economic life that is adapted to these things. I think. And, the way--

Russ Roberts: How will it change? What do you think will change? 53:58

Nassim Nicholas Taleb: We're antifragile. I mean, we learned from this. It's like when a plane crashes, the probability another plane crashes drops.

I think the odds of us suffering from the big one--because this is not the big one, by the way. The big one can be some antibiotic-resistant bacteria coming out of hospitals, something crazy. So, now we know how to handle it. And, we know how to--

Russ Roberts: Are you confident about--how confident are you about that? So--

Nassim Nicholas Taleb: I am very confident. And, the fact that you and I are doing Zoom now, something you never wanted before--

Russ Roberts: No.

Nassim Nicholas Taleb: is a testimony to the change of times, and how we're adapting, and some people are better off. Like, we have this conference, the Real World Risk Institute conference, and we're going to be able to accommodate many people who could not afford to come spend a week in New York.

Russ Roberts: Sure.

Nassim Nicholas Taleb: And, we never thought of it before, of doing something online. The university system is going to adapt to it, and it'll be lot cheaper to get an education, because you can stay in India and get a class from any person in the world.

So, there are things that the adaptation is leading. As you know, the economy likes stressors, leading to a beautiful--somewhere is adaptation. Somewhere. And, then, of course, some industries are suffering, but not too many industries are suffering. The industries that are suffering are--

Russ Roberts: Travel.

Nassim Nicholas Taleb: Travel and--

Russ Roberts: Recreation.

Nassim Nicholas Taleb: tourism, and--

Russ Roberts: Yeah.

Nassim Nicholas Taleb: And, even that can adapt. I've been traveling. I will travel. Two masks. And, I feel safe traveling under these conditions. But--

Russ Roberts: Let me ask--I'm going to push you on that. Because, I want to go visit my 87-year-old mom, who lives by herself in Alabama. And, I feel very uneasy about getting on an airplane, and potentially picking up the virus in that airplane, and killing her. Although she's healthy. She's a healthy 87, thank God. But, you know, I don't feel comfortable enough wearing two masks and goggles. And, we have a selfie of you--you've posted it in a fantastic essay on mask.

Nassim Nicholas Taleb: Yes. Yes.

Russ Roberts: But, I don't know whether that's going to make it safe for her, for me to come visit. Will it?

Nassim Nicholas Taleb: I think--I mean, I've been looking at data. And, again, I don't want to make a statement now--

Russ Roberts: That would go against what you just spent an hour talking about, yeah.

Nassim Nicholas Taleb: No, no. No. Data could give you a hint--

Russ Roberts: Yeah.

Nassim Nicholas Taleb: that data. Particularly, we have a few cases of people infected who, in a high-volume business--say, like stylists in a salon--two infected people and then, an exposure of 15 minutes or however it takes, that led to zero contagion. So, we have enough. And, now we're going to accumulate studies for meta-things, that tells us that if you have two masks--and I'm not sure it's necessary to wash your hands, but I would do that anyway, and use these things--you should be okay traveling. And, soon we will have more data, when we examine those employees who spend a lot of time on airplanes. And--

Russ Roberts: Well, I find it incredible.

Nassim Nicholas Taleb: Yeah, yeah. Yes.

Russ Roberts: I find it incredible that we don't have those data regularly collected--

Nassim Nicholas Taleb: No, no.

Russ Roberts: and discussed. It's so important. It's so valuable.

Nassim Nicholas Taleb: The Chinese are good at collecting stuff.

Russ Roberts: Yeah.

Nassim Nicholas Taleb: Hong Kong. A lot of people have data. It's just it's not yet put together in a way that I'm comfortable discussing. But, I'm giving you ahead enough statistical signals that, if I had to go--and, I have to go see my slightly younger mother in Lebanon, which entails a much longer exposure--I'm going to hop on a plane. I'm just waiting for my travel agent to call back. See? So, I'm going to get on a plane and take that risk. Two masks plus other measures. Actually, I [crosstalk 00:58:01]--95.

Russ Roberts: Do you drink water?

Nassim Nicholas Taleb: I don't have to worry about water, but I--

Russ Roberts: No, but do you take the mask off, while you're on the plane, to drink the water?

Nassim Nicholas Taleb: If you're seated far from other passengers, the exposure is not linear. You see? For a minute or two, it's not going to hurt you.

Russ Roberts: Okay, so don't--listeners out there: Do not accept this as medical advice.

Nassim Nicholas Taleb: Exactly.

Russ Roberts: Nassim is not a doctor. He only plays one on EconTalk. Let me--

Nassim Nicholas Taleb: No, no. The whole thing, the other thing--let's talk about doctors, where people think that understanding medicine, which is understanding a human, single human, allows you to understand the collective. That's another skill[?] transformation.

Russ Roberts: Yeah. It's a good point. I learned that from you also, Nassim: If you want to figure out how to gamble, don't talk to the carpenter who built the roulette wheel.

Nassim Nicholas Taleb: Exactly. Exactly. Exactly.

Russ Roberts: Even though he should be an expert. 'He built the wheel. Who could know more than him?'

Nassim Nicholas Taleb: Exactly.


Russ Roberts: Let's finish by summarizing. Let me try to summarize what I hear you are saying for the next pandemic, as well as the rest of this one. Take inexpensive ways to mitigate risk, such as wearing masks. At the national level, close your borders until you figure out what you're dealing with. At the local level--are there things at the local level ought to be done? Or are those all going to be at the individual--

Nassim Nicholas Taleb: New York is quarantining people coming from hot spots like Florida and--

Russ Roberts: Texas.

Nassim Nicholas Taleb: And, Texas. And, New York is quarantining them. And, also as sort of reaction, because Florida was quarantining New Yorkers.

Small, little measures like these--you don't need to bring cases to zero. You just want to make sure they don't overwhelm your system and that they don't collapse your infrastructure. You see?

Russ Roberts: Yeah. Well, it's more than that, though--

Nassim Nicholas Taleb: And also--that's the first thing. And, then the other one is, if--pandemic will be imploding just like the Plague or the Spanish Flu, naturally, when it--I don't believe in herd immunity as much as in reduced connectivity. So, but, I have two more points I'd like to make here.

Russ Roberts: Sure. Could I stop you? No, I can't. Go ahead. Make them.

Nassim Nicholas Taleb: Yeah. One thing we did not discuss is geronticide.

Russ Roberts: Yeah. Talk about that. Although I think I disagree with you on this. So, go ahead.

Nassim Nicholas Taleb: Okay, so--

Russ Roberts: Geronticide is the murder of the elderly.

Nassim Nicholas Taleb: Murder of the elderly. Okay. The ancient world not only did not favor geronticide, but, to the contrary, they worshipped the elders.

Russ Roberts: They honored them. Yeah.

Nassim Nicholas Taleb: They honored them. I mean, the Senate is a council of elders. Senatus is an old person. Same with the Greek name, and then the same in Arabic. The title 'sheikh,' it means 'old.' And, here we have it worse than there[?], I think. Okay? So, now, you realize that some societies understand the Golden Rule: 'Treat me the way you want to be treated when you're older, and respect that.' And, we have had, and Mediterranean societies--most of us are issued a visa in Mediterranean societies--

Russ Roberts: Well, most of the people on this podcast are, anyway.

Nassim Nicholas Taleb: Okay. Coming from--

Russ Roberts: This episode.

Nassim Nicholas Taleb: the ancient world--people coming from the ancient world--Babylonian, Egyptian, Mediterranean, Greco-Roman--they had values that, basically, for them, geronticide is never an option. Never.

So, would you, would you, would you--if I give you a thousand dollars, would you kill your grandfather? Of course not. Now how about if I give you a billion? Okay. So, now let's negotiate. So, this is Northwestern European society.

Russ Roberts: Which?

Nassim Nicholas Taleb: Making claims, utilitarian claims--something, someone, infected Northern Europe with these, this calculus of, 'Yeah, no, if we let old people die, it'll save us x in economic terms.' That's not a question that you're allowed to ask in a classical society. It's offensive.

Russ Roberts: There's no trade-off.

Nassim Nicholas Taleb: I find it offensive.

Russ Roberts: There's no trade-off.

Nassim Nicholas Taleb: There are trade-offs you don't make.

And, this is also central, and our paper explained that, for the tail risk, you have no trade-off, because, you know, eventually you're going to die.

The ramification, this one, even if you look at it from a utilitarian standpoint, it's not a trade-off, because you don't know where it's going to stop. So the idea--and, America is the only Northwestern European--sorry; it's not quite Northwestern European, but in population, and habits, and language--that has full age equality, no age discrimination. So--

Russ Roberts: And, more than that--and more than that--we devote an enormous amount of resources and money to the health of the elderly, and they're less--

Nassim Nicholas Taleb: Exactly. And, the rationale is that--it's not the elderly, not a different class. The young people--you got to think ergodically--will be the elderly.

Russ Roberts: Yeah. That's hard for people to notice.

Nassim Nicholas Taleb: Okay. So, the other thing is, in other words: We're going to take care of you. So, we're telling the young, by taking care of the elderly, 'We're going to take care of you.'

Russ Roberts: We hope. Yeah.

Nassim Nicholas Taleb: Plus, there are other things involved in this discipline of taking care of the elderly, because you cannot have both--I mean, capitalism is not about ingratitude.

Russ Roberts: It shouldn't be.

Nassim Nicholas Taleb: You see. Yeah. I mean, the Greeks let the mules that were used in building the Parthenon--they gave them privileged grounds instead of letting them die because you don't need them anymore. To the contrary, they gave them privileged ground.

Russ Roberts: They honored them.

Nassim Nicholas Taleb: They honored them. They honored them by giving them great ground. Great food and stuff like that.

So, it was, it was--there's something, there's a transmission of intergenerational gratitude that people don't get.


Russ Roberts: Well, I'm going to take this in a direction maybe you're headed, which is, to me, this is related to your insight of soul in the game, which is that: you could argue it's irrational to take care of a donkey or a mule. It's just an animal. Especially in ancient times. You could say, 'Well, yes, they helped build the Parthenon. But, they're animals. We could let them die.'

But, the idea that you honor everything that is worth honoring, even when it's not, quote, "rational," has a sweet, sweet aspect to it. And right now, we're looking at you, Nassim, you're wearing a lovely sport coat and shirt, and I am confident you are as well-dressed below the waist as you are above it, because I know that's your motto.

Nassim Nicholas Taleb: Yeah. You can check.

Russ Roberts: Yeah. See? Oh, it's jeans. Jeans. Well, it's a nice look.

Nassim Nicholas Taleb: No, no: I always wear jeans. And, jeans--

Russ Roberts: But, you don't wear pajamas when you're on Zoom.

Nassim Nicholas Taleb: No. No, I don't. Even when you don't see me. Even when you don't see me. I described it in The Black Swan. There's an idea that Visconti, the filmmaker--when he handed the box that contained jewels, made sure there were jewels inside of it.

Russ Roberts: Yeah. In the movie. Even though it was just a prop.

Nassim Nicholas Taleb: Exactly. You're not supposed to see them, but he made sure it was authentic. There was this idea of being authentic.

When you write about something, you don't want to be an imposter. You want to talk--I mean, people may never know that I can't write in Chinese, may never know that I don't know Chinese.

Russ Roberts: Right.

Nassim Nicholas Taleb: But, you can't do it. You should never be an imposter. Likewise, I don't like people who write about probability not knowing it, and then use fortune cookies formulas and verbalistic formula.

Particularly this is present among psychologists. They are fake. It shows. There's something about a fake person. You notice because they will be fake elsewhere, you see.

Russ Roberts: Well, I like the idea that practicing authenticity, even in places where it's not observable, builds a habit that will protect you in cases when you might forget and go out in your pajamas to the grocery. I think it's a beautiful idea.

Nassim Nicholas Taleb: Yes. Yes. And, also in fields. For example, if you only write about subjects that you didn't discover on Wikipedia, but things that are natural, it looks clumsy a little bit, compared to the researcher who knows how to fake and write reports. But people detect it. People detect authenticity.

Russ Roberts: My guest today has been Nassim Nicholas Taleb. Nassim, thank you for being on EconTalk.

Nassim Nicholas Taleb: Thank you very much. Thanks a lot. Thanks again.