If you prefer to listen rather than read, this blog is available as a podcast here. Or if you want to listen to just this post:
I heard a story recently about 3M (h/t: frequent commenter Boonton). Supposedly, back during the SARS outbreak they decided they should build in some “surge capacity” around the construction of N95 masks. Enough additional capacity that they could double their production at a moment’s notice. It was unclear if they actually did that or if they were just thinking about it. And even if they had, it appears that the scope of the current crisis is great enough that it’s not as if this one decision would have dramatically altered the outcome. Still it’s hard to dispute that it would have helped.
The question which immediately suggests itself is how would the market have treated this development? In fact imagine that there were two companies, one who took some portion of their profits and plowed them back into various measures which would help in the event of a crisis and one that didn’t. How do you imagine that the stock market and investors would price these two companies? I’m reasonably certain that the latter, the one who took the profits and disbursed them as dividends, or found some other use for them, would end up with a higher valuation, all else being equal. In other words I would very much expect Wall Street to have punished 3M for this foresight, particularly over a sufficiently long time horizon where there were no additional epidemics, but even in the few years between doing it and needing it.
What this story illustrates is that attempts to maximize the efficiency of an economic system also have the effect of increasing and possibly maximizing the fragility as well. And while, in general, I don’t have much to say about the Coronavirus which hasn’t been said already and better by someone else. I do think that this may be one of the few areas where not enough has been said already and where I might, in fact, have something useful to add to the discussion.
To begin, I want to turn from examining the world we have to looking at the world we wished we had, at least with respect to the virus. And as long as we’re already on the subject of masks we might as well continue in this vein.
As the pandemic progresses one of the big things people are noticing is the difference in the number of infections between the various countries. In particular South Korea, Japan and Taiwan have done much better than places like Italy and Spain. There are of course a number of reasons for why this might be, but there’s increasing anecdotal evidence indicating that the availability of masks might be one part of the equation.
For example, Taiwan is very closely connected to China, and one might expect that they would have gotten the virus quite early. Probably before people really understood what was going on, but definitely well before the recent policies of social distancing really started to be implemented to say nothing of a full on quarantine, and yet somehow, they only have 235 infections, which as of this writing puts them below Utah!
There are of course numerous reasons for why this might be, but I’m more and more inclined to believe that one big factor is that Taiwan is a mask producing juggernaut. In fact as recently as a few days ago they pledged to send 100,000 masks a week to us. They can make this gesture (and I know 100k is actually just a drop in the bucket) because they’re currently producing 10 million masks a day! For a country that only has 24 million people. Meaning that while that won’t quite cover one mask per day per person it’s enough that if people avoid leaving the house unnecessarily and if some masks can be reused they have enough for everyone to be wearing one at all times when they’re out of doors.
South Korea is similar and the big challenge there was not that they weren’t producing enough masks, but to stop exporting the masks they were already making. Finally reports out of Japan indicate that about 95% of people are wearing masks. But more importantly reports were that even before the pandemic around 30-40% were wearing masks just as a matter of habit. Is it possible that this slowed things down enough to allow them to get on top of it once the true scale of the crisis was apparent?
As I was writing this post I did some research on the topic, but before the post was finished Scott Alexander of Slate Star Codex came along, as he frequently does, and released Face Masks: Much More Than You Wanted to Know which is a very thorough examination of the mask question. What he found mostly supports my point, and in particular this story was fascinating:
Some people with swine flu travelled on a plane from New York to China, and many fellow passengers got infected. Some researchers looked at whether passengers who wore masks throughout the flight stayed healthier. The answer was very much yes. They were able to track down 9 people who got sick on the flight and 32 who didn’t. 0% of the sick passengers wore masks, compared to 47% of the healthy passengers. Another way to look at that is that 0% of mask-wearers got sick, but 35% of non-wearers did. This was a significant difference, and of obvious applicability to the current question.
On the other hand, when we turn to the US and Europe, in contrast to Southeast Asia, there is definitely not a culture of mask-wearing and even if there had been, most of those countries apparently imported masks from places like Taiwan and China (a point I’ll return to) meaning that when those countries stopped exporting masks there were even fewer available here, enough that people started worrying about not having even enough for the healthcare providers. Once this problem became apparent various authorities started telling people that masks were ineffective. A policy which has since been called out for not merely being wrong, but contradictory, counterproductive and undermining trust in the authorities at a time when they needed it most.
For most people, myself included, it’s just common sense that wearing a mask helps, the only question is how much? Based on evidence out of the countries just mentioned, and the SSC post I would venture to say that they help quite a bit. Also they’re cheap. Particularly when weighed against the eventual cost of this pandemic.
We’ve all learned many new things since the pandemic began, one of the things I didn’t realize was how bad the SARS epidemic was and how much the current precautions and behavior of the Southeast Asian countries is based on lessons learned during that epidemic. And while it’s understandable that I might have missed that (particularly since I didn’t start blogging until 2016) the CDC and the federal government on the other hand should have been paying very close attention. In fact, you would have expected that they might have taken some precautions in case something like that happened again or worse, started in the US. (Though to be fair, we don’t have wet markets, if that is where it started. As if we didn’t already have enough conspiracy theories.) Instead the US Government’s response has been borderline criminal. (Other people have done a much better job of talking about this than I could, but if you’re interested in a fairly short podcast just about the delay of testing that avoids sensational accusations, check out this Planet Money episode.)
To continue using the example of masks, I think it’s worth asking what it would have taken for the government to have a one month supply for every single person in the country. Stockpiled against a potential pandemic. According to this Wired article, before the pandemic 100 disposable masks were going for $3.75, let’s be conservative and round up and say that masks cost 4 cents a piece. From there the math is straightforward: 330 million people X 30 days X $0.04, is ~$400 million dollars, or 3% of the CDCs budget, or less than what the federal government spends in an hour. Still, I’ll agree, that’s a fair amount of money. But remember that’s the absolute maximum it would cost. I’d actually be surprised if once you factored in the huge economy of scale that we couldn’t do it for a 10th of that or even a 20th of that. And it would presumably have been cheaper still to just buy the necessary machinery for making masks and then mothball it, with a “break in case of emergency” sign on the door. Once you factor in all the potential cost savings, it’s hard to imagine that this would have cost more than $25 million (in fact if the government wants to offer me a $25 million contract to make it happen for next time I would be happy to take it.) And when you consider that it’s probably going to end up costing the US over a trillion dollars, plus the expected odds of something like this happening, you start to wonder why on earth they didn’t do this and countless other things that might have come in handy. (A strategic toilet paper reserve? I’m just saying.)
When you consider all of the budgetary cuts that were proposed for the CDC, which have emerged in the wake of the pandemic, and which generally involved only tens of millions of dollars, it seems unlikely that even $25 million would be allocated just for masks, but why is that? With a federal budget of $3.8 trillion why are we so concerned about $25 million? (It’s the equivalent of worrying about $25 when you make $3.8 million a year.) I understand people who are opposed to government spending, heck I’m one of them, but this also seems like one of those classic cases where people balk at spending anything to prevent a crisis, while somehow simultaneously being willing to bury the problem in a giant mountain of money once the crisis actually hits. It would be one thing if we refused to spend the money regardless of the circumstances, but if recent financial news is any indication we’re obviously willing to spend whatever it takes, just not in any precautionary way. (Somewhat related, my post from very early in the history of the blog about the Sack of Baghdad. Whatever the federal government and the CDC were doing in the months leading up to this, it was the wrong thing.)
One assumes that this desire to cut funds from even an agency like the CDC, where budgets are tiny to begin with, and where, additionally, the cost of failure is so large, must also come from the drive for efficiency we already mentioned or it’s modern bureaucratic equivalent. Which I understand, and to an extent agree with. We shouldn’t waste money, whether it’s taxpayer money or not. But given the massive potential cost of a pandemic, even if one never emerged, it seems clear that this spending wouldn’t have been a waste. But how do we get to there from here? How do we make sure this drive to save money and increase efficiency doesn’t create priorities which are so lean that they can’t spare any thought for the future. How do we avoid punishing companies who exercise foresight, like the example of 3M? Or how do we ensure that governmental agencies are making reasonable cost benefit calculations which take into account the enormous expense of future calamities, and then taking straightforward precautions to prevent or at least mitigate those calamities?
One of the most obvious potential solutions, but the one that seems to generate the greatest amount of opposition, is the idea of increasing the price of items like masks during periods of increased demands. Or what most people call “price gouging”. Let’s return to the story of 3M and imagine that instead of price gouging being universally frowned on, that instead it was widely understood and accepted that if there was an emergency 3M was not only allowed, but expected to charge 10 times as much for masks. In that case they’re not just hoping to help people out when the calamity comes, they’re also hoping to make a profit. This is in line with the generally accepted function of business, and presumably stockholders might reward them for their foresight, rather than punish them for not being “efficient” enough. In any case maintaining a surge capacity for mask production would be a gamble they’d be more willing to take.
Notice I said 3M, which is different from people buying up thousands of masks and then reselling them on Amazon. As a generalizable principle, if we were going to do this, I would say that people should be able to raise prices for goods they control as soon as they think they see a spike in demand. So if someone had started stocking up on masks at the first of the year before anyone realized what was going to happen then they ought to be able to sell them later for whatever they think the market would bear. This early buying would have been a valuable signal of what was about to happen. But once the demand is obvious to everyone then 3M should raise their prices (and profit from the foresight of building a second production line) and Costco (or whoever) should raise their prices. I understand that this is not what happens, and that it’s not likely to happen, but if you want a market based approach to this particular problem, this is it.
A governmental solution mostly involves doing the things I already mentioned, like relying on the government to stockpile masks, or to proactively spend money to prevent large calamities. Though you may be wondering how subsidiarity, the principle that issues should be handled at the lowest possible level, factors into things. Clearly state or even local governments could also stockpile masks, or give tax incentives to people for maintaining spare capacity in the manufacture of certain emergency supplies. But as far as I can tell subsidiarity was long ago sacrificed to the very efficiency we’ve been talking about, and thus far I’ve seen no evidence of one state being more prepared than another. Though speaking of tax subsidies, it’s easy to imagine a hybrid solution that involves both the public and private sectors. 3M would have faced a different choice if the government had offered a tax credit for building and maintaining surge capacity in mask construction.
You could also imagine that greater exercise of anti-monopolistic powers might have helped. If you have ten companies in a given sector, rather than one or two you’re more likely to have one company that bets differently, and maybe that bet will be the one that pays off. Additionally globalization has also been a big topic of conversation, and was one of the first effects people noticed about the pandemic. Hardware companies were announcing delays for all of their products because they are all built in China. But we also saw this in our discussion of masks. Most of the mask production also appeared to be in Southeast Asia, and once they decided they needed the masks locally the rest of the world was caught flat-footed. Of course, economists hate the sort of tariffs which would be required to rectify this situation, or even improve it much, and they also mostly hate the idea of breaking up monopolies, but that’s because their primary metric is efficiency and as I’ve been saying from the beginning, efficiency is fragile.
Before moving on, two other things that don’t quite fit anywhere else. First, doesn’t it feel like there should be a lot of “surge capacity” or room to take precautions, or just slack in the modern world? Somehow we’ve contrived a system where there’s basically a car and television for every man, woman and child in America (276.1 and 285 million respectively vs. a population of 327.2 million) but somehow when a real crisis comes we don’t have enough spare capacity to do even as much as nations like Taiwan, Japan and South Korea? As I’ve already suggested, there are obvious difficulties, but doesn’t it feel like we should be wiser and better prepared than we have been in spite of all that?
Second, am I the only one who would have felt a lot better about the huge stimulus package if we weren’t already running a deficit of $984 billion dollars in 2019? This, despite supposedly having a great economy for the last few years? In any rational system you build up reserves during the good years that you can then draw on during the bad years. That does not appear to be what we’re doing at all. I hope the MMT advocates are right and the size of the US government debt doesn’t matter, because if it matters even a little bit then at some point we’re in a huge amount of trouble.
Which brings us back to our topic. If, as I’m claiming, all of the modern methods are unworkable, we might ask what have people done traditionally, and the answer for most of human history would involve families, and to a lesser extent tribes, along with religious groups. And I suspect that there are quite a few people who are gaining a greater appreciation for family at this very moment. In my own case, I have deep stocks of many things, but toilet paper was not one of them (an obvious oversight on my part). As it turns out my mother-in-law has a ton (not from hoarding) and so rather than show up at Costco at 7 am, or buy it on the black market I can just get it from her. And if she hadn’t been able to help me I’m sure that my religious community would have. (Just to be clear I still haven’t burned through the TP I had on stock at the beginning of things, I’m just laying the groundwork to make sure I don’t get caught flat footed.) This whole story is an account of surge capacity. Though it may not look like it at first glance. But think of it this way, when you need help, having a single child that lives on the other side of the country doesn’t do you much good, but when you have five kids, three of whom live close by, you have four times as many resources to draw on in a crisis, and potentially six times as many depending on what you need.
Going even deeper, friend of the blog Mark wrote a post over on his blog which I keep thinking about, particularly in relation to the current topic. He talks about redundancy, fragility and efficiency as it relates to biological processes. In other words, how does life solve this problem? He gives the example of building a bridge and compares how an engineer would do it versus how a biological process would. While the engineer definitely wants to make sure that his bridge can bear a significantly greater load than whatever they judge to be the maximum, beyond that his primary goal is the same as everyone else in the modern world: efficiency. The biological process, on the other hand, would probably build a bridge made up of dozens of overlapping bridges, and it might cover the entire river rather than just one stretch of it. In other words from an engineering perspective it would be massively overbuilt. Why is that? Because life has been around for an awfully long time, and over the long run efficiency is the opposite of what you’re striving for. Efficiency equals fragility which, as we’re finding to our great sorrow, equals death.
I suspect that some of you are either already suffering financial difficulties as a result of the pandemic or that you will be soon, so rather than ask for donations, let me rather make an offer of communication. If anyone needs someone to chat with feel free to email me. It’s “we are not saved at gmail”. I promise I’ll respond.
A truly difficult modern puzzle. My favoritte line (maybe, there were lots of great ones ) “…We’re obviously willing to spend whatever it takes, just not in any precautionary way.” I had never considered our relative willingness to gorge after the fact but not spend much less ahead of time.
One of the big problems with stock holders rewarding big companies is that we’re dependent on passive stock holders, and ill-incentivized mangement. So not only are passive stockholders looking for quick wins, management is looking for quick bonuses THAT DO NOT GET CLAWED BACK. That’s an even worse incentive than the general shareholders I think. And FINALLY, we then bail these behemoths out when they screw up. I’m not so sure there is a fundamental problem here so much as bad rules that actually encourage this behavior. But maybe there is – the fundamental problem may in fact be that you just can’t prevent money printing governments from bailing these business out – that’s a cultural and legal muscle that as far as I can tell nobody has and is weakening. It may be a reason empires always eventually fail. Perhaps even the length of human lives has something to do with this inability. A career is only 40 years and maybe only 20 of that is spent with much decision making authority – maybe not long enough for management to consider the time scale optimal for the longer-term company.
I was listening to Will Durant this morning and he commented that religion gives people really good reasons to do things that are really important for reasons they are blind to. Ultimately we need some kind of cultural reasons to be people that believe in precautionary action.
Great stuff!
Thanks!
The bailing out problem is huge. And the very definition of moral hazard, and which as you say people have largely stopped paying attention to.
Which Durrant book? Sounds like Lessons of History, which I just finished.
The concept of moral hazard is one will take more risks because “someone’s got my back” and this additional risk will actually cause the unwanted event to happen more. Gov’t subsidizes flood insurance, people build more waterfront houses, the number of houses that flood goes up.
I’m not sure what the moral hazard is here. “Hey you should have set up your pizza joint to operate by being closed 2-3 months out of the year?” Moral hazard is less applicable to unprecedented events.
Some of the price gouging…. few years ago we had Sandy here in NJ and oddly, gas stations were taken offline. At first people thought it was just a matter of getting power restored to the stations. Not true, I saw stations who had full tanks of gas bring in small generators to run their pumps and empty their tanks. The problem was stations weren’t getting gas deliveries. Gas stations turn over their inventory once every other day or so…..
For about a week and a half we had bad gas shortages. Word would come that one station was getting gas, there’d be a two hour line there. Pop up operators appeared in parking lots with hand painted signs selling generators and a ‘full tank of gas with each purchase’. After a few weeks of this nonsense the gov’t put in a simple rationing system. License plate ends in an even number, you can buy gas on even days and vice versa. Broke up the panic and the long lines stopped. Things got better and a week or two later things were running again and the system was stopped and things went back to normal.
The price signals of scarcity are helpful but they can also become counter productive. When a hospital that needs masks gets outbid by 100 mom & pop outlets that are trying to price speculate, for example, is that helpful? Price run ups take on a life of their own and depart from the strictly ‘rational’. Consider Bitcoin. Do we just give the hospital a blank check when there’s a pandemic? No, we say you can buy one box per house per month but the hospital gets the first one.
Imagine you are in the grocery store going to buy some toilet paper and the first guy in fills up two carts, taking nearly a quarter of their stock in one shot. Shrug, market forces? No, the supermarket, knowing riots are bad for business, says two items per customer until the crises is resolved.
One can imagine a mixed approach where as you say hospitals get first dibs, but that whatever is left over is priced such that people don’t fill up their entire shopping cart. Saying one per customer does something similar, but it’s not as robust and is easy to abuse (there are many stories of people just sending all the members of their family in once. Increasing prices solves that problem and it solves the bare shelves problem, if the shelves are going bear your prices are too low.
Agreed. It’s a ‘soft barrier’ to the hoarders. Yes you could make repeat trips to get around the restriction but that takes time and energy. If the point is to break the panic, I think it will do.
I’ve seen toilet paper come and go several times now. If I had no toilet paper I would still have been able to replenish. Interesting story, back in 73 Johnny Carson had a joke about toilet paper shortages and it actually caused a surge in buying and a stockout. Toilet paper is easy to go on shortage but it’s unlikely to ever be a real problem. You can hoard TP but you’re not going to use it any faster so the shortage wipes itself out…literally. Hoarders are removed from the market for months.
In terms of whether the masks make the difference, I’m not so sure. There appear to be individual ‘superspreader’ events early on. The SARS on the plane case may just be one example. It takes one spark to get kindling going into a fire, but you may make many sparks before you get that one.
China has a culture of mask wearing yet that didn’t stop the virus. The gov’t imposed a massive shutdown and now it’s looking like they pulled off something amazing. Italy and China are about even in cases but Italy is 23 times smaller than China. If China did not contain it and we had 1B getting it with a 1% or more fatality rate you’re talking 10M dead. S. Korea wears masks but they imposed a massive testing and contract tracing regime. We are turning away people from testing unless they have serious symptoms. Japan might simply have been lucky. Maybe not as many sparks landed there, plus a mask and social distancing culture made their ‘kindling’ a bit damp for the few that did. Or maybe this is going to be a wave with different countries peaking at different times and Japan may be rising as we go down.
There’s so much speculation based on so little evidence I think the best thing is to not endorse any one theory. How many times in the past few months has the slow trickle of information completely upended theory after theory?
As to why some counties got this under control (or it never got out of control … yet?) again I can’t sign onto any individual reasoning. Even the story in S. Korea makes no sense. Tracking works for a maximum of maybe a hundred people – 200-300 if we’re generous.
Not thousands. At that point we need to start shopping for other explanations. And as you point out, we can’t point to a ‘flattened curve’, because it didn’t go through the population (as far as we know), it just went away leaving millions of presumably susceptible people alone. Why? I hear lots of ideas, but they don’t match the evidence. I expect we’ll know more in the postmortem, which is an unfortunate way to find out. I just don’t expect solid answers right now.
I have to say I find reading economists on this issue annoying for that very reason. They run with wild speculation too easily. They assume they understand how healthcare works, they are too eager to toss out improvised ‘solutions’ without appreciating that medicine has been working on these problems for centuries. I had the same problem when I dealt with cancer in the family. Why not just cut the tumor out? Why are we stopping chemo, it seems to work and it has no bad side effects let’s keep going to make sure every bad cell is gone. Wait here’s a pdf of some study that says turmeric may help…. there’s reasons the professionals don’t get excited by your googling.
They also run with ideas that are frankly unworkable and even if they were could not be implemented in a matter of weeks. For example, having all the old people segregated from the rest of society as it goes about its business. Ohhh wait, a lot of old people live with younger family and nursing homes are run and staffed by mostly younger people. These are the same people who would go ballistic about gov’t planning when, say, Elizabeth Warren thinks we should undo Facebook’s merger with Instagram.
@Mark you’re missing the asymmetry here. There is zero evidence that masks cause any harm. There is some evidence that masks help. Given how cheap masks are, there would have been almost no downside to stockpiling masks and possibly significant upside.
Beyond that it was not my intent to “push a theory” merely to assemble all of the evidence for masks (which Scott Alexander did a better job of). And to point out that using that as an example we could have been way more prepared. Particularly if we have, apparently an unlimited budget to do so.
I do agree, masks don’t harm and they probably do help. On the other hand when this thing is over every house is likely to have a box or two of N95 masks and there will be multiple stockpiles. It is unlikely we’ll ever have this particular shortage again.
Also hospitals will probably adopt a protocol of using masks all the time rather than just with infectious patients and during epidemics. The production of masks, gloves etc. will be higher going forward.
Taleb is both a blessing and a curse. He speaks in dense mathematical language, but mathematically he’s a blowhard and show off, so mathematicians aren’t impressed by him. To the regular guy, he’s hard to understand. Personality wise, he attracts a dedicated fan following but turns everyone else off….which is the best way to ensure you won’t be listened too.
All you have to say is remember the mayor from Jaws? Close the beaches? For a shark? Absurd lose all that money for something that’s never happened! Then the shark eats a bunch of people.
The Asian nations didn’t just know how to prepare for this, they had multiple scares.
Right, but there’s a difference between ex ante reasoning and ex post. Ex post, we can say, “this one intervention has a possible marginal impact and no downsides; we should implement it since we know today we’re having a situation where we need it.” Ex ante, we have to say, “these five thousand possible mitigation strategies all cost a little, but cumulatively they cost a fortune for countless possible small risks. Let’s hold off implementing any one strategy unless we have good evidence it works.”
Not because the risk is small, but because the benefit is unknown and there are countless possible tail risks to consider, so we want to make sure our mitigation strategies are at least effective. Is it effective to reduce a 0.1%/yr. risk by 5% when there are at least hundreds of risks near that level to consider? Not unless resources are infinite.
That’s not irrational. I think it’s easy to take ex post reasoning and ask why it wasn’t considered ex ante. It’s two different types of calculations.
I’m hearing this argument all the time now from people who resist admitting, essentially, that the Talebians were correct. It would have more weight if we weren’t spending a fortune right now to deal with the after effects of this particular disaster. For the money we’re spending we could not only have dealt with the pandemic tail risk, but a bunch of other’s as well.
I understand that we’re dealing with predictions which are difficult to quantify, but I doubt you could come up with a sensible list of precautions that the government should have taken which exceeds the cost of what they’re spending right now. It’s not as if there are 10,000 potential risks out there, and even if they are pandemic would have to have been in the top 10 easily. If not top 3.
Yes I agree this is ex post reasoning, but give the probability of this disaster, and it’s impact. I am totally unconvinced that this shouldn’t have been thought of before hand particularly given that all the South East Asians did precisely that!
I’m not asking for some crazy precaution, I’m asking for the US to be able to look at SARS and copy the precautions taken by the countries who actually went through it.
I understand that, but I think we just don’t assign the same degree of efficacy to the efforts of the Asian response. For masks, it’s easier when you’re already making lots of masks in-house to have a lot of masks around. That’s less, “we were prepared” than it is “we happened to be the people who had this specific resource”. My suggestion above that we use masks for more industries outside of a crisis would solve this as well as having a stockpile.
Again on masks, the evidence suggests they might slow the spread, which is different from stopping it in its tracks when it hasn’t gone through 1% of the population yet. That’s not something we can attribute to masks. And as I mentioned above, contact tracing can’t work at the level of spread in China, South Korea, or the US. There has to be some other explanation for how they got to where they are, and whether that’s something we can implement here. An honest assessment of the proposed explanations for the stoppages in East Asia shouldn’t conclude that it was accomplished through the use of masks and contact tracing. It seems people are latching onto those explanations because they’re available, not because they make any kind of sense.
As to fighting against Taleb: I’ve agreed with Taleb since before the current crisis about the disproportionate focus on historical risk versus future tail risks. I’m not resisting admitting that he was correct, I thought he was before now. Honestly, if I had a few billion dollars back in November for tail risk mitigation I’d spend it on … asteroid tracking. That’s still more of an existential threat than a pandemic even today. If you had a choice between mask stockpiles and asteroid tracking I feel like you should absolutely go with asteroid tracking. – and that won’t even solve the problem, just tell us how much time we have before we all die!
This is the real problem with tail risks, and one that hasn’t really been discussed by Taleb (maybe it has – I haven’t read all his stuff, but his followers don’t talk about it much). When you start worrying about tail risks you’re on fairly solid ground. You know near-tail risks fairly well – a magnitude 7.0 earthquake, for example. The farther out on the tail you get, though, the more uncertainty there is in the system.
It’s easy to look at the current crisis and say, “why didn’t we do this cheap simple thing?” but even for a pandemic we can get lost in a garden of forking paths. Maybe for the next pandemic we should be ready to isolate old people, but then we’re just reacting to the last pandemic. How about instead we prepare for all the possible ways a pandemic can hit disproportionately: by age (children, healthy 20-somethings, old people); by gender; by race; by physical ailment – including but not limited to heart disease, overweight, RA/psoriasis, etc? Sometimes pandemics are respiratory, sometimes cardiac, neurological, integumentary, renal, etc. Why aren’t we more prepared for an ebola epidemic? What about the many invasive species we know are currently spreading to the US from their natural environments around the world – and bringing countless diseases not just to humans but also to key crop species? What about the fact that nearly all rubber trees come from a vanishingly small gene pool and are susceptible to the kind of viral eradication that wiped out bananas back in the 1950’s? Should we be taking some precautions about genetic diversity of rubber plants, since the next world war could easily be fought over rubber when the next – inevitable – pathogen sweeps through?
The answer to all of the above is, “Of course we should be preparing for each of these possible – and foreseeable – disasters!” The problem with the Talebian observation of tail risks, though, is this issue of proliferating possibilities.
In contrast, the problem Taleb very frequently discusses is that people tend to choose FUTURE mitigation strategies by looking to PAST problems. If, as a result of SARS-CoV-2 we assume the next pandemic will come from SARS-CoV-3 we’re not learning lessons from Taleb, but in spite of him.
Sure, there’s likely to be another pandemic, but maybe it won’t be a coronavirus. Maybe China will clean up its wet markets. Maybe their population will decrease as a result of population controls and modernization so fewer diseases come from East Asia. Maybe instead Africa will become the next epicenter, and we should be looking there for what specific PPE to stock up on. Maybe surgical gloves instead? Nobody knows.
After this, people will stock up on masks and take precautionary measures to keep ventilators around. Then the next pandemic will be some offshoot of necrotizing fasciitis, and everyone will ask why we didn’t think of stocking up on gloves – it’s so cheap! We’ll keep chasing this rabbit, but never get ahead of it.
To be fair, I’m not arguing against stockpiling masks in the future. I’m saying it’s unfair to complain against past planners for not preparing for the next crisis in the Right Way. Likely we won’t do better ourselves, even if we try harder, because we don’t know what we should have done until it’s too late. Not without true prophetic counsel.
Perhaps, but I firmly believe that even without revelation that there are precautions we can take which have a positive ROI, which will help with future crises, and which we mostly don’t take because we have too much focus on short term efficiency.
That’s a valid critique, and one you made in the article above. Maybe not just that we don’t take them because we need to see short term efficiencies, but also because we expect the world, and solutions to our problems, to be efficient. We balk at anything that looks highly inefficient, assuming “there must be a better way”.
Also remember Taiwan has kind of frosty relations with China, might not have been much travel to begin with so when the lockdown started it was mostly spared. Hong Kong too had those anti-China protests going on with nearly half the population out in the streets shutting down most workplaces.
I think there’s good cause to do some serious mask research after this is over. Slate Star Codex’s piece makes it clear a lot of the studies are more lab conditions. If I’m working on live ebola virus I want 100% protection. But if I’m just using a mask on a regular basis if this cuts my risk of infection 40%, say, (includ. colds and flu which are a pain too) then that’s very helpful. It’s also helpful to know if this due to motivating behavior that’s not directly mask related. When I’m out now I more or less never touch my face until I’ve washed my hands. If the benefit of the mask is 99% of keeping people from face touching, you may discover you have a replication problem in these studies. The masks that reduced infection before will do so only because they train people to avoid face touching. Next time there’s a shortage you may see no increase from infection. Ditto for hand washing which no doubt will happen more after this event than it did before.
What we really need to know is how much infection potential is out there from just walking around, assuming you are keeping 6 ft from people and they aren’t coughing or sneezing without covering? How much do masks both high quality and the ‘improvised’ work there? Or are masks more a psychological tool to remind? If that’s the case we can experience a prolonged mask shortage as long as we recognize the need to compensate.
Health care workers, on the other hand, have different concerns The masks for them are much more functional because they are working in environments with high virus concentrations. There is some evidence that being subject to a high dose of live virus is more likely to cause the serious illness.
That brings up yet another unknown. I’m thinking here infection is not like pregnancy….which needs only one sperm. If you’re only encountering, say, 100 viruses walking down the street and your poorly worn mask blocks 60 of them, does that matter? Maybe you need to be hit with a few hundred thousand viruses to catch infection? I don’t think they know and there’s probably no easy to way to study it.
This visualization would seem to indicate that there was significant travel of infected people from China to Taiwan:
https://www.nytimes.com/interactive/2020/03/22/world/coronavirus-spread.html
As to your last point on dosing, Robin Hanson has been on about that for the last few days you should check out his stuff.
Seems like Taiwan struck early and used multiple tactics. They started checking passengers arriving from China on Dec 31st. Stopped the export of masks and then nationalized the mask production industry by Jan 31. Plus intense monitoring and tracing seem to make it not that different than S. Korea.
I think what you’d really want to test your theory is a mask wearing nation not that far from China that nonetheless didn’t do much more early on. I don’t think you’re going to find that. The type of country that wears masks is probably the type of company whose gov’t is also really on the lookout for epidemics.
https://foreignpolicy.com/2020/03/16/taiwan-china-fear-coronavirus-success/
Hanson is one of those who annoy me the most. The ideas he has are in fact being explored but I feel like he pounds on them as if he was the only one smart enough to have them or the medical establishment is stupid.
For example, I see him talking about variolation (when they used to take a smallpox scab and have someone eat it to give them a ‘poor man’s vaccine’ which actually worked better than you’d think). I’ve heard several times now doctors on podcasts discuss convalescent serum….basically taking the blood serum of those who recovered and either giving it to someone very sick or giving it to 4-5 health care professionals. The idea is introduce the antibodies from the blood that defeated the virus to provide protection or treatment.
On the other hand I may just be cranky.
Is there another way to increase capacity than just having excess unused production capacity lying around? For example, have more professions wear masks on the daily (like food prep, etc.) but switch to essential personnel during a crisis. Another method would be to make your production line capable of switching over production in a pinch so you can produce a needed product during a crisis.
A simple incentive for this is to have mandatory shutdowns of non-essential manufacturing during something like this, then provide a bunch of education to companies about how they can make themselves essential during a crisis. Or preferential purchasing agreements during a crisis, where Uncle Sam agrees to but a certain amount of the emergency production (and thereby replenish emergency stores at the end of the crisis by pulling from the overproduction).
I think the mask problem will resolve itself and never be seen again. What I think we need is a stronger concept of public health. Hospitals are part of an integrated system that comes alive during an epidemic but our health care model is highly individualistic and set up to be profit centers for individual cases. Chronic diseases are well tuned to the profit centered model but not so much infectious diseases.