Transcript of Extra: On COVID-19 Strategies with Robin Hanson

The following is a rough transcript which has not been revised by The Jim Rutt Show or by Robin Hanson. Please check with us before using any quotations from this transcript. Thank you.

Jim: Howdy. This is Jim Rutt and this is the Jim Rutt Show. Listeners have asked us to provide pointers to some of the resources we talk about on the show. We now have links to books and articles referenced the recent podcasts that are available on our website. We also offer full transcripts. Go to, that’s

Jim: This is another in our extra COVID-19 episodes. They’re shorter, the audio quality won’t be as good, but the intellectual content will be every bit as good. Today’s guest is Robin Hanson, professor of economics at George Mason University, and he has some very interesting and perhaps to some people startling ideas about how to manage the COVID-19 virus and what lessons we should learn from this to be ready for next time. Why don’t we start out, Robin, with the idea of dosage effect, something that I’ve seen relatively little in the media. The idea that the strength of the infection you get, and hence the risk of having the infection is in some fashion, linear or nonlinear, I don’t know. Probably related to the amount of exposure one gets. What can you say about that?

Robin: Well, I’ve seen three media articles over the last couple of weeks. One in the New York Times, one in the New Yorker and one in a British paper, so it’s not completely neglected. And the usual tone is, well people make the simplifying assumption that you’re either infected or not, but there’s actually the subtly, and it’s kind of interesting and there’s this literature on it. And in fact in the medical literature, there is this huge literature on what they call viral load because that’s relatively easy to measure. That is they look in your bloodstream and see how many viruses there are. And that correlates with a lot of things. People with higher viral load typically have more symptoms and die more, and all those sorts of things. Now the initial viral load is what we’re talking about as a dose. That is the moment you get infected, how big a chunk do you get?

Robin: Now people believe that that makes a lot of difference. And that’s one of the main rationales for wearing masks because not just to cut the chance of an infection, but to cut the dose that you get when you get it. But in terms of research, it’s thin because it’s hard to actually control and measure the initial dose. We don’t do a lot of infecting people on purpose in order to measure that. But there is a long history, not only do we have like two studies in the last few decades where we’ve seen an effect of dose on death rates for humans for a virus. One is measles where it’s about a factor of 14 another is SARS, where it is about a factor of 3. And these are different variations in the dose that naturally happen. But two centuries ago we used to do low dose infection of smallpox as a deliberate way to reduce death.

Robin: So for example, George Washington at Valley Forge, a big set of US troops in Canada got wiped out, which is why the US doesn’t control Canada at the moment. And then George Washington’s troops at Valley Forge are threatened. And so he adopted variolation, which was … reduce the death rate from a typical 20 to 30% from smallpox down to 1 to 2%. and that was a deliberate small dose. So we know there’s this big effect. So some people have talked about it in the media, but there’s two things they haven’t mentioned, and I’m not sure it’s because they don’t know about it or kind of scared to mention. One is that it makes lockdowns more problematic. So two extreme kinds of ways that you get to a dose, one is by say, kissing a family member or spending a lot of time close to a family member, and then you’re likely to get a big dose.

Robin: If they’re infected and you are kissing them first moment they got infected, then you’re going to get a lot. Another way you might get infected is say, touching a door handle somewhere out in the world where somebody else touched it in the last few hours. And in that case you’re likely get a small dose, or say walking past someone who’s were breathing out in the wild. And those low doses would have much lower mortality rates, again, according to the data we’ve seen let’s say a factor of 3 to 30. And so with lockdowns we are slowing the rate at which people get infected, but unfortunately raising possibly the death rate when they do get infected by a possibly large margin, that’s a concern. So we need this lockdown to be worth it on the other margins to be willing to pay that cost.

Jim: Has anybody looked at simulation including that effect?

Robin: I don’t know that they have. I haven’t seen anybody even mention that effect besides me.

Jim: I’ll pass that along to somebody, because at the Santa Fe Institute we have a number of epidemiology modelers in our community. I will pass that idea along to them and maybe connect you guys up and maybe somebody can run a model and say, “All right, how good does the social distancing in fact have to be to overcome the intensification effect?”

Robin: So the big trade off is many people are hoping that we can just, with lockdowns, shut this virus down and most people never get it. If only 5% of people ever get this virus, then you’re saving the other 95% and that’s a pretty big gain. So you would be willing to pay a relatively large costs increasing that 5%’s risk in order to save the other 95%. but if we’re just talking about everybody getting it spreading that out over time, that’s a harder trade off in my mind because first of all you’re suffering that big economic cost over a much longer time of the big lockdown. And secondly, of course, the question is how much better is it to get sick when you’re not near the peak of the pandemic because you get more medical resources? Is that a factor of 50%, is that a factor of 5?

Robin: What is that gain and mortality of getting the medical treatment away from the peak? So, for example, people focus on ventilators and making sure everybody has them. But apparently people on ventilators, three quarters of them die anyway. Benefit bent on there can’t be much more than a quarter in terms of mortality rates. Which is nothing to sneeze at, I mean, if you can get that benefit, you want it. But again, at what cost? So as I said, there’s two main effects that people haven’t talked about. One is this dose effect increasing death rates in lockdown. And the other is the prospect of reducing death rates by just deliberately infecting people.

Robin: And this is what they used to do with smallpox centuries ago it was called variolation. And again, it was considered this great historical triumph to drastically reduce death rates from smallpox, which were really wiping Europeans out. Of course, this was something the Chinese and the Indians and actually a great many of the Africans who were brought to the US as slaves, they had undergone this treatment back in Africa. And so, say the United States, this information was available to them. But they just refused to listen or believe until prestigious enough people from the Royal Society, et cetera, talked about it.

Jim: Now you have been advocating straight up that we consider a program of structured variolation to attack COVID-19. Could you tell us about your ideas there?

Robin: Yeah. So as you know, a lot of people have a lot of ideas and a lot of ideas should be explored. There’s many different vaccines, people should test them, they should be tested as fast as possible. But this variolation concept I consider just very reliable. I mean the odds are very high that it works. Of course, you still should test it so I think we should try with a test of a 100 or a 1,000 people where you just try different ways to dose them and different ways it gets in the body, et cetera. Maybe antivirals as a compliment. Well, but after that I’m really quite sure, at least 80, 90% or more that you’ll get good results. And then say a factor of five in death rates, you can then drastically reduce how many people are affected. Now what you actually want to do is to attract volunteers for this.

Robin: You don’t necessarily have to force people, and then you have a place where they go to get infected, then stay isolated until they are recovered. If you just infect people but then let them go around, then you’re just accelerating the pace of the whole pandemic. Whereas if you isolate them as soon as you infect them, you actually can be slowing down the pandemic. It’s spreading out, flatten the curve. So the key concept is after an initial trial of roughly 1,000, figure out how to dose and properly infect people, isolate them in what I call hero hotels where you pay to go in and you can’t leave until you are recovered. And then try to get as many of these things open and as many as want to come in voluntarily and maybe even subsidize it, get health plans to cover the cost, et cetera. And that’s the key idea.

Jim: Yeah, certainly for the upfront costs, no reason not to incentive heavily. Pay $10,000, right? You’d get a shitload of volunteers for 10 grand, young and healthy people who [crosstalk 00:09:20].

Robin: It’s not just money, though. A lot of people are really hurting because they can’t work and they’d like to go back to work-

Jim: [crosstalk 00:09:29].

Robin: They’d like to socialize and as soon as you are recovered from this you are verified and certified to work and socialize.

Jim: Yeah, that’d be great. Part of the deal is you maybe get a small stipend to do it and then you get a certificate that says you’re immune. Now of course, that’s assuming, which seems very, very likely from the history of viral infections that you will get at least a year’s immunity from an exposure. I have yet to see that definitively from anybody though. Have you seen any reports on whether an infection actually does give immunity?

Robin: The key concept is that when you recover, that is your immune system pushing it out. So at the moment of recovery you are immune and that’s what it means, your immune system pushed it out. If your immune system wasn’t involved then of course not, but almost always it’s your immune system pushing it out. So the question isn’t are you immune, it’s how long are you immune? And that depends on basically the virus changing over the coming months and years. So there are some viruses that change really fast, such as HIV, in which case immunity hardly lasts at all. But that’s very rare. So we have good reason to expect just on the usual distribution of viruses that this won’t be such a thing. And we do see the mutation rate on this one being modest and typical, not enormous like HIV.

Jim: Yeah. In fact, I’ve read, though this was before, there was millions of hosts, that the mutation rate was quite a bit lower than influenza, which might indicate that you’d have a multiyear immunity.

Robin: Of course, the mutation rate is proportional to the number of people who have it, so you wouldn’t see a very high mutation rate- [crosstalk 00:11:06].

Jim: Yeah, that’d be a very important number for the CDC to be tracking and reporting out to the people who are thinking about these strategies, [crosstalk 00:11:15] is only three months it may not be worth the hero hotel.

Robin: Right. But we’re facing a really dire prospect here. The odds are really quite good that basically over half the world will be infected in the next year. Our attempts to suppress this are not going very well, there’s still hope and we should try, but we need a plan B. And a plan B that can cut deaths rates by a factor of five should be pretty damn tempting. It should at least inspire an initial test to figure out if it works.

Jim: Yeah, it’s really two parts that we win from this. One is cutting the death rate, but second is managing the socioeconomic implications on the backside of the curve, which is where our society really needs to start being focusing its attention. Because if we don’t handle this right we could crash the whole socioeconomic system in a very ugly way. And the earlier we get people back to work in a safe fashion, the better. So I think your proposal actually addresses two important factors.

Robin: Yes, I think the prospect of a lockdown for say five years until the vaccine is finally delivered is really daunting. I mean, we can handle a lockdown for a couple of weeks, but there’s not really much prospect of ending this in a couple of weeks. What many people are hoping is that they could reduce it within a couple of weeks enough so then massive test and trace would hold it down. But even that is really ambitious. It’s far more than we’ve ever done with previous pandemics. And at the moment, this lockdown isn’t actually cutting the growth rate to a negative growth rate. It might be slowing it to a lower positive growth rate, but it’s still positive at the moment.

Jim: Well, United States came late. If we look what’s happening and Italy and Spain and in California and Washington, the curves had been bent there. So I believe that in those places that have put reasonable lockdowns on, we will see a bending of the curve the next two weeks. But your point is still well taken, to manage the backside of the curve is going to require a very dynamic dance of a really strong surveillance, contact tracing which the United States has sucked at. And then the reimplementation of pretty draconian social distancing in the new hotspots on the backside. Because a herd with no immunity hotspots are going to reoccur and if you let them spread they’re going to take off again. We do have some hope though from places like Taiwan and Singapore and Hong Kong that with enough testing and with enough tracking and with strong enough quarantining, maybe we can manage it for the year and a half until we have the vaccine.

Robin: Again, it’s a week maybe. I mean we should pursue it but we really need a strong plan B because most likely we’re going to need to invoke plan B. So even places like Singapore and Japan who seem to have been doing okay as the places around them get more and more infected, they get more and more of a wave of things coming across the border to handle, then it gets harder and harder. And there’s a basic fact about exponential growth, is if you have a bunch of different exponential growths all added together, it’s the highest growth rate that dominates. So unfortunately, unless they can be completely isolated from one another, whatever’s producing the highest growth rate, then it leaks into the rest and then the total growth rate is dominated by the highest one. So it’s not enough to point to the three best places, you have to wonder about the three worst.

Jim: We don’t know what’s going to happen in the third world, whether the warmer, moister temperatures will cause it not to go exponential, whether there’s built in immunities for similar things, who knows? But that could be the actually untalked about major reserve if the global south becomes utterly overwhelmed by this thing.

Robin: Well, even in the United States where we haven’t traditionally had strong borders between the states, it just takes a few states like New York who don’t keep it under control, for whom a lot of people leave and go to other places, to overwhelm the good efforts of people in other places who would otherwise keep it locked down.

Jim: Very good. Well, I think this is a very important idea and I hope that people think about it. I will pass it along to the modelers that I know at Texas and Northeastern, and where else? A few other places, the people in our Santa Fe Institute community who are doing a state of the art, and in fact some of the two of them are part of the team that Trump and his people are relying upon.

Robin: Well, [crosstalk 00:15:46] to talk to them.

Jim: Okay. I will make the offer.

Robin: Well, I did some modeling myself before I realized that this dose effect was an important deal. Just that the idea of deliberate infection, even without a dose effect. And even there I found gains and I analyzed the difference between deliberately infecting the young and healthy versus the old and sick, and who else you might put in quarantines. And I found compared to sort of a baseline model where you randomly put people in quarantine, a model where you deliberately put the oldest people in quarantine and you deliberately infect the youngest people gives a 40% reduction in death rates. So I was pretty excited about that before I realized there were these factors of 30.

Jim: Cool. And you know, these are folks who already have big comprehensive models take the consideration [inaudible 00:00:16:29]. So perhaps you could merge into-

Robin: I would very much prefer somebody’s professional experience model to the modified to include these facts rather than to rely on my amateur model.

Jim: Yeah, I’ll make those introductions as soon as we’re done or at least by the end of the day. Onto our second topic, you recently posted that long ago nations didn’t have standing armies. In fact, the United States didn’t even have much of a standing army before World War II. I think our army was smaller than Portugal’s. But today goddamn huge amounts are spent on standing armies all around the world. Shouldn’t we be getting serious? This pandemic is not going to be the last one. Hell we’ve had four or five of them in the last 20 years. And you know, I’ve come around to the view, people talk about wisdom.

Jim: That’s a topic I’ve always been skeptical about seems a little woo for my taste. But I’ve concluded that perhaps operational wisdom in the modern world consists of two things. One, understanding the concept of fat-tailed distributions, and the second the concept of exponentials. And pandemics look like they hit on both, I mean, it’s a fat-tailed distribution and we’re going to continue to have pandemics. And this is by far not the worst possible pandemic. We could have something that was more contagious and more lethal though there’s a balance there, as we know from epidemiology. And so this idea of building much greater standing capacity seems like a very good one. Tell us more about that.

Robin: Well, so that was just a tweet I made that some other people resonated with and it does make sense, but I’d like to focus less on capacity and more on flexibility. So what I’m struck by is this very basic phenomena in the world and human history, and in the history of biology, where you tend to have a split between specialized solutions to specialized problems and environments, and more general flexible solutions that are better able to adapt to changes. So for example, as you know, humans in particular, we’re not the fastest runners, we’re not the sharpest teeth, et cetera. We specialized in being general and flexible and we found a niche where in environments that changed rapidly enough or where we ranged over wide enough environments we could win compared to more specialized predators that we were competing against. And someone in industry in society, you may have a firm that’s in a very stable industry and a very stable product.

Robin: And then along comes a new product or a new variation and they just fall apart because they haven’t acquired flexibility and generality. They’ve just been tuned to that very particular environment. Whereas say like famously Intel or companies like that, their motto was everything’s going to be changing all the time, don’t get stuck with any one thing, constantly be ready for change and flexibility. And militaries, similarly as you may know, many times in the past a military has lost a war because they were so ready for the previous war that they couldn’t flexibly enough change and adapt to the new war. And part of the reason the allies won in World War II is that we happen to be exceptionally flexible compared to competitors. And I’m concerned that not only our militaries over the last few decades haven’t actually been facing difficult wars where they needed flexibility to win, but many of our other social institutions have atrophied.

Robin: Even our legal and regulatory institutions have assumed an environment is going to be pretty stable as they’ve seen it. And then when you get a big disruption, like the one we’ve just seen, they just can’t adapt. They’re just not ready for it. And honestly, that is some of the biggest problems we’ve seen with the response to COVID-19, is that our regulatory institutions, our medical institutions, just couldn’t adapt very fast. They weren’t ready for that. There was no place. So if you want to put a standing army in response to medical problems, I’m happy to devote the resources. But if you just give more resources to the same organizations run the same way, I fear they will atrophy in the same way, they will get locked down into rules and regulations and procedures that work well over stable decades and then fail terribly when all of a sudden, things change fast and they need to adapt fast.

Jim: Of course, it’s not just pandemics that this is about, it’s financial panics. It’s potential cyber attacks on our key infrastructure. It’s solar flares. In fact, one of things I’m contemplating is that we really ought to have a department of wicked risks looking at complex systems risks, that involves surveillance, sense making, decision making and then pass off the action doing to other people. Because the thing that shocked me the most about this damn thing, a lot of things have shocked me about it, but one of them is how unbelievably inept the whole west has been in taking a pretty damn clear signal by say the 20th of January that there was a nontrivial risk and some hedging ought to be done at the hundreds of millions of dollars of level and partial mobilization begun just as an intelligent risk. The same way a trader with a portfolio would of hedged the risk when they saw a trend going adversely.

Jim: And yet as far as I could tell, the chief executives of the major countries were caught blindsided, weren’t even aware of this damn thing until the 1st of March. So there was five or six weeks, which is about 10 doublings of the damn thing, before the sensors who had detected that there was something going on that needed to be hedged and mobilization needed to be started. Nothing happened. Or very, very little. And so the whole combination of monitoring and sense making and decision digestion, just as you said, it seems to have atrophied and failed in our society in a way that is surprising and bad.

Robin: So you know that the Romans had this policy to always be at war somewhere so that their military didn’t lose the edge and flexibility and experience to be able to fight a war. The question now is, you can assign a budget and an agency and tell them to be flexible, but how do you make them actually be flexible? So we might ask, how did these Asian nations succeed in being more flexible? A part of it is they are more autocratic and you’re looking at the high tail of the most functional autocratic ones and a lot of the low tail at the dysfunctional ones. Part of it is they faced SARS, a scary epidemic just a decade ago.

Robin: And so they created adaptations for that and were more ready for that. So we didn’t face that. We weren’t scared, so we didn’t set ourselves up for it. I’m not sure what else to explain other than just sort of look over the last few decades, a lot more has been changing in Asia. Their governance has had to deal with rapid rates of change in a lot of aspects of society, whereas our government and our society has been pretty stable for 70 years and a lot of people think no need to change the way we do things. Things have been going pretty well.

Jim: Yeah, I think that those are all real issues and of course it also goes to our governance. I mean, again, look across the west, the leadership that we have, where’s the Winston Churchill’s, where’s the FDR? Where’s even the Joseph Stalin’s, right? It’s basically lightweights and clowns that are the chief executives across most of the west.

Robin: And that can’t be the formal system, it’s still roughly the same formal system that has to be culture and our informal professions and other habits.

Jim: Yeah. Well, unfortunately we’ve made the profession of politics toxic, right? No FDR would put himself through the grinding humiliation of the current political process to become president of the United States when he could sit back fat, dumb and happy, and his with his millions of dollars on his estate. Very, very few people, I posted on this several times, you know, I know 20 people personally who would be better chief executives in the United States than the current one. And that’s just people I know. But not a single one of them would run for dog catcher because of the negative consequences of putting your name out into the public sphere these days.

Robin: I’m afraid in a democracy, the voters have to take a bit of the blame.

Jim: Yeah, at the end of the day, they do. The voters and the institutions around the voters. The media. Indeed.

Robin: The media responds to the voters, largely. It’s the voters who drive the media and the voters … Now maybe you would think of like the institutions of law or civil service professionals as also responsible for the resulting current state of those institutions and areas.

Jim: Yeah, we do need a institutional cleanup. Maybe this will be a call, probably not, but at least there’ll be some people with ears to hear that will start the process going. Well, Robin, I want to thank you for as always an incisive, intellectually powerful conversation and I think our audience will have learned quite a bit now.

Robin: Nice to talk to you again.

Jim: Great talking to you. It’s always great and I will send those connections out to the epidemiologists later today.

Robin: Okay. Take care.

Production services and audio editing by Jared Janes consulting. Music by Tom Muller at