Tyler Cowen on herd mentality and herd immunity

Hello and welcome. This is LockdownTV from UnHerd.com. Joining us today is none other than Tyler Cowen, one of America's foremost economists, author of a very successful blog called Marginal Revolution, a podcast called Conversations with Tyler, and he is joining us down the line from Northern Virginia, close but not within the DC metropolitan area. Hi, and how are you doing, Tyler?

I'm fine. Hello.

So there's a whole lot of stuff we would love to talk to you about. We are being inundated with data. We're trying to make sense of it. There's a lot going on in the world, and you have a wonderful combination of skills and a big brain that we're hoping to deploy to help get some understanding in the next half hour. The first question is, there's a segment in your podcast, which I'm a big fan of, where you say underrated or overrated. So let me ask you, COVID-19, underrated or overrated?

Well, you mean the risk of COVID-19? It depends where you are and exactly when you're talking about. So if you mean England right now, at this moment, very late in August, most people probably are overrating the risk. Tim Harford had a very good piece in the Financial Times. That said, there are many parts of the world, including my own country, the United States, where people are still underrating the risk. They don't use masks, they don't take basic precautions. They go to bars inside. To me, that's quite reckless.

There was a piece you wrote in Bloomberg where you were talking about this idea of phantom risk, and where at some point, the actual risk from the virus can be replaced with some sort of phantom risk, and it can have a tail after the actual risk has receded to levels that wouldn't normally concern everyone. Tell us about that idea.

Well, we see this also happening with terrorist attacks. Before 9/11, clearly people underrated the risk that terrorists would take box cutters and commandeer airplanes and crash them into buildings, right? It was a terrible mistake. But after the attack happened, for a long time, possibly even still now, people overrate that risk because it has become salient to them. So one of the biggest problems of recovering from COVID-19, when that point comes around, is that people will consider the mere existence of cases a reason to panic, and that will thwart the resumption of economic activity. Now, to some extent, again, depending on the region and the time, we still should be taking data and case numbers very seriously. But at some point when fatality rates are lower, possibly there's a good vaccine or an okay vaccine, it will be very hard for us to move away from our previous understanding of risk, and we will be overreacting.

So there's kind of an inflection point then where government's job moves from being one where you should be sort of emphasising the risk to make sure that your people are being sufficiently cautious, to a point where you then need to reassure them and try and reverse the kind of fear that you've helped in gender.

That's correct, and we won't know exactly what to occur. So as you know, your own government is now talking about people going back to work in offices. Well, are they at the point where that's the appropriate switch to make? Personally, I don't know, but I think the point is that no one knows, and what you will need is a frank, honest discussion of trade-offs.

You mentioned case numbers. We had Professor Sridhar here last week and she was talking about this idea of zero COVID, that in fact governments should try to drive down case numbers to zero, because that's the only way that people are going to be sufficiently reassured to reopen society fully. And I just wonder, is it a case almost of too much data with this question of cases? Because if we had case numbers for, for example, flu— and I'm not saying they're comparable diseases even, but it is a serious disease that can claim lives. If we had the amount of data we have about coronavirus about flu, and people saw charts and case numbers, and in every school they knew how many flu cases there were, and if you were flu positive, you would, you know, take yourself out of society, that could potentially have been a huge fear for many years now. Do you think it's a question of almost too much data?

Well, we're using the data in the wrong way. As a social scientist, I'm reluctant to say too much data, but I watched your interview with the COVID zero woman, and right away when you pushed her on, do you really mean COVID zero as in the number zero? She said no. And the project is called COVID zero. So titles of projects in pandemics, I mean, really should not be there for reasons of marketing. They should be telling people the truth and in some way instructing them. So to me, that represents a broader intellectual failure that is going on, which is that it is truly very hard to consistently think in terms of trade-offs in a pandemic, in general too.

Are there governments that you look at and think, 'They're getting this right'? I mean, which countries do you think right now we should be looking to as a model?

Well, clearly, there are some governments that got it right, but they're no longer models. They acted so early, so well, that they have it under control. Taiwan, New Zealand, you know the whole list. But we can't just now copy them and wake up and say, "We're going to keep this from entering our country at all." So I think right now we're entering a period of high uncertainty as to what is exactly the best thing to do. But it seems to me the one front that really has enormous potential for success is biomedical research. Those institutions work fairly well, and I would like to see my government and also yours do everything possible to accelerate that work and in particular be ready on the side of distribution and have prepaid contracts up front with biomedical institutions doing the work that will save lives and make it safe for business to reopen. And it seems to me that is the thing we're doing that is working.

So that's a kind of practical step. Which I think most people would sign up to. Do you feel like on the kind of harder to pin down cultural questions, you know, that we talked about this idea that fear can kind of actually really damage a society long term, do you think there's anything the government can do about that or should be trying to do? Or do you think it's now kind of out of government's hands and it will be up to consumers and voters to decide when they are comfortable to resume life?

Well, there's a science of risk communications, as you know, and my government in particular has disregarded it. So I think top officials from the federal government should consistently wear masks, even if everyone around them is being tested all the time, just as a symbolic gesture. Uh, the way you get people to actually obey mandates is not to scold them. So the left wing in my country, I feel in some ways has been counterproductive. The way you get people to obey mandates is first to be clear and to be non-threatening and non-panicky, and then to express your own vulnerability and present the mandate in the form of a request that you do not wish to be endangered in a gentle way rather than scolding or lecturing. So the United States in particular has been a disaster at risk communications. President, governors, a lot of people running universities and schools We've just gotten it completely wrong.

And do you feel like some of those schools are not— basically not communicating those trade-offs? They should be reassuring as well as emphasising the risks? I mean, a lot of those schools have closed again, having reopened briefly, or they're saying that they're only going to partially open. Are you in favour of a more— a greater emphasis on the benefits of reopening those institutions, even if there is a residual risk?

Here's what is happening. Individual institutions do not want COVID risk on their books, so to speak. They don't want to be associated with a cluster. So they will say, well, if we reopen, we'll have so many cases. But of course, you might have had those cases from the young people mingling anyway, say, out there in bars. What we need is better data to understand the net risk. And I mean that word net very literally, the net marginal risk of reopening a college or university? I don't think we know that yet. I don't see many people working hard enough to figure it out. People are looking at the gross risk. Well, if we reopen, you know, we'll have this test positivity rate of, say, 7% instead of 2%. That's relevant information, but it's not dispositive. It's information wanting to pass along the COVID reputational risk somewhere else.

So, what about putting numbers and putting— studies on the risk of not opening? I mean, would you like to see that? Would that help to communicate the trade-off and say, 'Okay, this is the risk of opening, this is what we perceive to be the risks of not opening'?

Absolutely. I have on net nudged people in the direction of being more favourable toward opening. But again, I wish to stress the context-dependent nature of the decision, especially in my country. It really depends where you're at. And what are your institutional capabilities, what kind of testing you can do, many other questions. But it seems to me that opening is somewhat underrated in the United States right now.

Part of the reason for that must be politics. I mean, we are now— the conventions are over, we are now into the general election campaign in the US. The next 2 months is going to be nothing else. How much do you feel that the kind of political heat has negatively impacted decision-making around this pandemic? And do you think that after November 3rd, no matter who wins, or maybe depending on who wins, there is a chance of cooler heads and more of a kind of rational approach to this?

I'm not sure we'll see cooler heads. We may see a flipping of positions. But I would say as a general rule, if you see a discussion of COVID-19 or a tweet about COVID-19, and it's somehow tying the issue to President Trump, the odds are the expected value of that information is negative. I would say this as someone who is not a supporter of the president. I think quite possibly he's the worst president in my lifetime. Uh, but people confuse opposition with Trump to stances on COVID, and the overwhelming desire now is simply to attack Trump. So we've had many failures in my country. Say, from the CDC or the FDA early on in the pandemic season. And we're now trying to rewrite history as if all of those are failures of President Trump, who has been politicizing the agencies. Well, that's a valid complaint. But the early failures, say, on the testing front were not due to President Trump. They were intra-agency failures. And again, you're seeing this revisionist view. Everything must be done to make Trump look bad. And one thing we failed to discuss early enough was the possibility of semi-herd immunity. And because Trump said, "Well, at some point this will go away," that became something unacceptable for the intelligentsia to talk about. Now, I'm not convinced on herd immunity, but I certainly understand there's a reasonable chance it could be the case right now in some regions.

Right. And, I mean, we saw that in the UK as well, because herd immunity became associated with Donald Trump Dominic Cummings and Boris Johnson in the early part, and it became this kind of word that no public official would ever utter because it was political kind of dynamite. But it is actually a biological phenomenon that has been talked about for decades. You know, I just spent time in Sweden, and there Anders Tegnell, who is their epidemiologist, he believes that the numbers coming down so low in Sweden are at least partially the result of immunities building up. Um, do you buy that?

I think we need to be very careful. So possibly a few regions have some kind of herd immunity. That would include southern England, many parts of Sweden, possibly Madrid, possibly Lombardy, possibly New York City. I call it semi-herd immunity or imperfect herd immunity because we don't know how permanent it will be. So one micro foundation for herd immunity is that you have these super spreaders, and once they are taken care of, they have an immune response or they're somehow removed from the system. Some of them even may die. One possibility is then you're simply pretty safe ongoing. But another possibility is that the potential super spreaders rotate over time. And as winter comes, as your social configurations change, as you as offices reopen. In essence, new individuals step into those super-spreader roles and your temporary herd immunity configuration is gone. Now, I think we don't know. Again, I want more data. I want there to be a kind of social and scientific obsession with answering that question. I know it's not easy. Uh, so I'm not sure about herd immunity, but I do feel in the United States in particular, it has been underdiscussed for a long time for political reasons. And say the New York Times only in August starts to seriously consider different notions of herd immunity. But as you said, it's a very well-known scientific concept. We did not have many people, serious people, telling us, say, in March, well, we're not at all sure about herd immunity, but, you know, there's a 20% or 30% chance this could happen and this should enter our calculations. Public health experts are too risk-averse to speak that way. So what you get from them is more of a status-conscious kind of calculation, and not the actual expected value truth.

Actually, you talked about labels for movements and phenomena. Herd immunity is an unfortunate label, because it sounds like you're treating people like cattle or sheep. In Sweden, they've started calling it collective immunity, which sounds kind of more virtuous and like a kind of group— collective action.

It may not be collective, right? So it's simply that some pockets have limited transmission to a fairly low level, and many things could change that, especially countries where many tourists come in. You could lose your semi-herd immunity through the arrival of tourism. I wouldn't say that's the biggest risk for Sweden, but certainly it's an issue. So I also worry about the people who have pushed herd immunity ideas Some of them are responsible, but many of them do not talk probabilistically enough. And there's a kind of once-and-for-allism, like when Michael Levitt will just say on Twitter, "It's over," like a two-word sentence. Nassim Taleb's critique of Levitt that he won't think probabilistically is essentially correct. And it's irresponsible when people push the herd immunity idea without also putting forward its possible limitations.

So what does this mean, do you think, policy-wise? I mean, if we give you the keys to the castle, and we say, right, you are going to decide policy, what would your metrics be for changing restrictions and guidelines? If case numbers is a kind of false friend and can get us panicked about things that don't matter, would you just look at hospitalisations and numbers of people getting seriously ill? Or how would you kind of run this risk level assessment?

Well, which castle is it again that I'm ruling on?

For starters, let's give you the United States since you're just some miles from DC there.

I would vastly improve risk communications at all levels of government. I would do everything possible to encourage mask wearing, recognising that pure coercion is not always the most effective way to do that. I would invest radically enormous sums of money on a super short time frame to collect better data on how COVID-19 has spread, better data on immunity, look into theories of herd immunity, safety of ventilation systems, how offices work, how super spreading might work. And on a super short hyper-accelerated timeline, learn much more about those issues. I would continue with Operation Warp Speed on the vaccine front, which I think so far has done remarkably well. We now have new procedures in place where approval of new tests is actually coming quite quickly. I'm very happy about that, but that's only very recent. And more long-term, I would look into how to make the CDC a better functioning operation in terms of actually doing things. I don't feel I know from my current distance how to do that, but that's a medium-term priority, though it's not something you can accomplish in the super short run. So those would be some of the things I would do. But also pre-purchase for, say, monoclonal antibodies, the notion of a pre-purchase programme from the government, akin to what we're doing with vaccines. Very likely I would do that as well.

Okay, so since we've got this fantasy of President Cowen for a moment, you are an economist.

No, not president, you had me ruling with the keys to the castle.

Okay, fine. Yeah, the supreme overlord. What about the economic front? Because, you know, you are an economist, and clearly we are entering uncharted waters economically. I mean, we haven't had reductions in GDP levels like this ever. What would you do that neither Steve Mnuchin in the US or Rishi Sunak in the UK have so far done?

I think in the United States, the overall macro policy response has been actually pretty okay, all things considered, given the time to formulate a response. In general, as an economist, I'm pretty skeptical of fiscal policy. Anyone who reads me will tell you that. But this is one of those rare cases where a huge dose of fiscal policy is appropriate. We did it. Of course, it was highly imperfect. Our biggest mistake was probably not getting enough into very small businesses. But that's intrinsically harder to do, right? Easier to help out a large business. We need another package of aid. It's being negotiated right now. Probably it won't be enough for state governments, but we don't know. We're on an okay track in the United States, and I know less about the UK fiscal response, but it seems to me from what I do know, it's been broadly okay, all things considered.

So what happens in the coming months then when companies start going bust? I mean, let's take, you know, the airline industry, for example. How long can it survive at these radically reduced levels of travel? And who do you save and who do you let go under?

Well, if airline companies go bust, which they do in my country really pretty frequently anyway, the assets are not destroyed. They regroup. In essence, the debt is written down and people are back to flying whenever that is safe enough. So that's far from my biggest worry. It's really the smaller businesses I'm worried about. But the sad truth is some considerable number of them probably ought to go bust because consumer demand simply will not be there and we need some resource reallocation. But I would readily admit that's the one area where the US and I think also UK policy response has failed the most. But again, it is an intrinsically harder thing to do.

So we've talked about the kind of health data and policy around that. We've talked a little bit about the economy. Do you worry for the culture more generally? I mean, you know, do you think people who worry about the decline of Western civilization or worry that introducing fear and a threat like this at a point where we are self-doubting and unsure of what we stand for could be really dangerous? Do you think that's it's over the top and it's overly fearful, or are you worried more broadly, big picture, for the kind of state of our culture after this shock?

Well, one should always be concerned, right? It's partly how you keep things on track. But I'm not as especially concerned about the current moment as most people are. Just a simple example, go back to the Spanish flu, you know, 1919 and the like. I'm not sure what are the causal relationships, but that was followed by the Roaring '20s, a very different mood. Now, I'm not predicting we'll see the same thing this time. You know, probably not. But I think that shows we don't have a very good understanding of how cultures evolve. And the notion that now today we have Facebook and thus somehow all of democracy will crumble seems to me a gross exaggeration and likely a complete untruth. So the United States, United Kingdom have long and glorious histories, they've always had major problems. Very, very likely our cultures will get through this. But of course, those cultures need a lot of repair work.

So what role do you think the election might have in that? I mean, let's fast forward to the 4th of November.

Let's—

I mean, I've got to get you on record. Do you have a view as to what's going to happen? Can we ask you that?

Sure, I follow the betting markets, and I think the last I looked, Biden was a 60-40 favorite. Since I'm not betting in those markets against the market estimate, I guess I must think my estimate is roughly the same. So Biden does not have by any means a lock on it, but he's definitely favored because unemployment aside, the COVID response has been poor. Trump has alienated a lot of people and he's not seen as reliable. And seniors in particular feel threatened, and they were Trump's earlier base in those Midwestern states he had won are up for grabs. Possibly Texas is up for grabs. So I would think Biden is more likely to win.

So let's say that happens, and you're right on that. Do you think the whole atmosphere changes immediately? You know, how much repair work do you think a President Biden would have to do? Or do you think the fact of his election and him being kind of a reassuring figure who's been around for decades would just instantly change the whole political atmosphere?

No, I don't think things will change as quickly as people believe. And in some regards, Biden would become more like Trump. I don't mean in terms of style, but Trump's willingness to let this rip. So Biden has said, 'Well, if necessary, we'll just close down the whole country again.' I know he said that. I understand why he said that. I'm quite sure he doesn't mean it. He doesn't even have the authority to do it. And if you look at the responses of Democratic governors, many to the left of Joe Biden, or for that matter university presidents, you don't find them acting the way Biden has said he would act. In some regards, they've been more like Trump. Again, not stylistically, but Americans do not want their businesses closed. You may or may not think that is prudent, but the response we have, we've had for long enough and in enough places. You can't just say it's one very bad president. It is to some extent what Americans have chosen. Maybe there's a cultural sickness to that. That deserves honest discussion. But I think one also has to be honest not to just blame it all on a small number of individuals.

So overall, do you feel pessimistic? Or would you say you feel optimistic about, I guess, we have to say, compared to everybody else? Do you think optimism is underrated at this moment?

Well, relative to Twitter, I'm certainly optimistic. But relative to actual living, breathing, walking American human beings, I'm probably right at about the consensus. So that's where I would put myself. Keep in mind media, of course, push negative narratives on that, right? So you need to make your own calculations to adjust for that.

Final question for you, Tyler, before we let you get on with your day. So you're known for this philosophy of state capacity libertarianism.

Yes.

It's intriguing. And I just want you to explain what it means for us.

In this context, we need a government that in terms of public policy mostly is devoted to supporting capitalism, not destroying capitalism, not redistributing all the wealth, not bringing in socialism. But that said, one of the best ways to support capitalism is that your government can actually provide the important public goods. Starts with the national defense, law and order, also protection against pandemics. Which we had neglected because we hadn't had major pandemics other than HIV/AIDS for a long time. And that one we performed poorly on. So now is an instance where you really need a government that can get some particular things done and produce the right public goods, or you will end up with much less liberty, much less capitalism, high rate of business failures, everything we're seeing right now. So that's state capacity libertarianism. It's like a capitalism. Yes., but you do need your government to have its act together.

Which of the two candidates do you think is closer to state capacity libertarianism?

Well, certainly not Donald Trump, but Biden, I don't know. He's a cipher to me. I mean, I remember seeing Biden on TV, the Watergate hearings. I think I was maybe 11 years old, and now he might be president. It's just very hard for me to process. Yeah, maybe he's just some embodiment of the professional politician. In a way that has to be modeled. And maybe that modeling is more important than whatever you're trying to project onto Joseph Biden himself. I don't know. I would say the problems in America are very often systemic and rooted in the citizenry and maybe less based in the personalities of particular leaders than you might think. And when I wrote about state capacity libertarianism, I made the term up. I wouldn't quite say as a joke, but I thought, well, what's a term I could come up with that would be so apparently self-contradictory and so cumbersome that no one will ever use, because I don't really believe in the terms. I would say be data-driven. Try to be as epistemologically honest as you can. Always be asking questions. Don't get too set in your beliefs. Take in as much quality information as you can. And at the end of the day, yes, kind of figure out where you've ended up, but don't obsess over the labels. It's going to make you a worse decision maker.

Great advice. Tyler Cowen, that's what we're trying to do here, and thank you for helping us do that today.

My pleasure, thank you.

That was Professor Tyler Cowen, Professor of Economics at George Mason University in Virginia, telling us to set aside our labels and political biases and try to follow the data and the numbers and come up with rational policies. So that sounds quite sensible to me. Thanks to him, and thanks for watching. This was Lockdown TV.