Capitalisn’t: Coronavirus—a Cost-Benefit Analysis of the Economic Shutdown

Luigi: As you can imagine, this is the first podcast we’re going to tape in social distancing from home. So, I hope to apologize for some background noises, because we’re not working in the studio.

And we decided, given the situation, to go to a weekly podcast and focus our attention on the impact that COVID-19 has on our economy and what can we do from an economic point of view to fight back.

Kate: We’re going to cover this question of whether we should lift restrictions, whether we should try to go back to normal and potentially bear the sacrifices in terms of loss of life from doing that. As President Trump said, he wants to be able to return to normalcy by Easter. And so, we’re going to consider the consequences of that and whether it’s worth it from a cost-benefit perspective.

From Georgetown University, this is Kate Waldock.

Luigi: And from the University of Chicago, this is Luigi Zingales.

Kate: You’re listening to Capitalisn’t, and in particular, our new weekly series covering what’s going on with the coronavirus.

Luigi: So, Kate, how are you holding up in New York? New York has been hit particularly hard.

Kate: To be honest, it’s a little bit scary. I think for the past couple of weeks my fiancé and I have been trying to be exceedingly careful. We’ve been wearing gloves and face masks and using hand sanitizer for a long time. But yesterday it hit me particularly hard, and de Blasio had come out with a statement about his expectation of half of all people living in New York City actually getting coronavirus.

I’ve been able to see the line of coronavirus patients at the Brooklyn Hospital extending down the street, which is pretty wild and pretty scary. And I’m starting to wonder, what’s going to happen if, when I get it, are there going to be any hospital resources available to me? What if there’s no ventilators available? It’s a pretty scary reality. So, we’re trying to hunker down and only leave the apartment once every two weeks or so. But it’s been pretty difficult.

What about you? How are you, Luigi?

Luigi: Actually, Chicago is not being hit so hard yet. But the news coming from Italy is horrendous. And I have a lot of family and friends that are indirectly touched.

Kate: Are you OK?

Luigi: No, I’m OK. With all these stories about other countries doing so much better than the United States, they’re not. And a friend of mine who has a brother in Germany, her brother actually has been sick for a long time. He finally got tested four days ago, and he’s still waiting for the result of the test. And in the meantime, the hospital does not accept him, nobody accepts him. He’s at home with a wife and a kid. So, the situation is, I think, dramatic everywhere. And so, this is a difficult moment for everybody. So, we’re also used to dismissing President Trump’s comments.

President Trump: We cannot let the cure be worse than the problem itself. We’re not going to let the cure be worse than the problem.

Luigi: “Oh, this is another crazy point of President Trump.” However, there is some rationale in what he’s saying, in the following sense. Think about fatalities in car accidents. Every year, I think 25,000 people die in car accidents in the United States. Now, this is not leading anybody to say we should stop driving cars. People are trading off, in some way, human lives with some other benefits.

The important point is to try to figure out if there is a systematic way to analyze whether it’s worth it. Economists have elaborated one method to do that, which has a lot of limitations, but it’s a useful method, at least to check whether Trump’s argument is completely baseless or whether it has some reason. And this method is called cost-benefit analysis.

Kate: I’m not an ethicist or a philosopher, but my gut is telling me that there is a difference in a cost-benefit analysis when you’re thinking about a small-probability event and an event that’s pretty random, versus a near certainty of mass loss of life.

Luigi: I understand you are unconfident in this, Kate, but remember the cost-benefit analysis does not really place a value on individual lives. What they do is they try to estimate how much people are willing to pay for small reductions in their risk of dying. And this is done every day by the Environmental Protection Agency, the EPA, to decide what is the tolerable or optimal amount of pollution, because we don’t like pollution, but reducing pollution is expensive. And then you have to decide how much pollution you tolerate.

At some point, you want to lift restrictions, right? I don’t think you want to live like this forever.

Kate: Sure.

Luigi: So, how do you decide when that point is?

Kate: I don’t know, Luigi. I’m not president. I’ve never lived through a pandemic before.

Luigi: You mean, unlike me, who is?

Kate: Right. These are truly unprecedented circumstances.

Luigi: What I’m saying is, in making this decision it’s useful to consider, what are the economic costs that we’re imposing in terms of losses of lives? Having that tradeoff is useful guidance. It’s not the only guidance, but it’s useful guidance.

So, two weeks ago, when the problem was not so important for the United States but was very pressing in Italy, I wrote a post on the ProMarket blog trying to frame this cost-benefit analysis. I did a back-of-the-envelope calculation, and I calculated that letting the disease run its course could lead up to a $65 trillion cost in human lives lost.

Casey Mulligan, who’s a colleague here at the University of Chicago, did not spare his criticism on Twitter about my calculation. And he was kind enough to record a brief message suggesting why he does not believe me.

Casey Mulligan: I read a blog post by Professor Zingales where he reaches a conclusion that the United States should be willing to spend $65 trillion to avoid extra deaths from the coronavirus. This estimate is wildly inaccurate by a factor of a hundred or more. You could say that Professor Zingales had his thumb on the scale, but that’s true only if there were an elephant sitting on his thumb. In each step in the calculation, he exaggerates the data needed. He exaggerates the number of deaths. He exaggerated the fraction of deaths that would be preventable by shutting down the economy. He exaggerated the value of life applicable to those deaths, and he failed to consider the range of options for protecting our citizens from the harms of the virus.

Luigi: It’s not a coincidence that Casey uses the factor of a hundred. Short of that, my argument goes through, a shutdown of the economy is worth paying to avoid the loss of so many lives. Now, the listener can judge on their own. I reached my number as the product of the value of a statistical life and the number of additional lives lost by letting the disease run its course. As the value of statistical life for the mostly elderly people who will die, I use $9 million. People quibble about this, it can be seven, it can be four, it can be five. The difference is not that big.

The real important number is the additional deaths that might be caused by letting the disease run its course. Now, remember that all the official forecasts predict that roughly 75, 80 percent of the population eventually will get it if we don’t do anything about it. Imperial College came up with some estimates that are even higher. And everybody agrees that roughly 1 percent of the people will die even under the best medical conditions. So, we’re talking about 2 million people dead.

Now, the question is, how many extra people will die if there are no hospitals to receive them, no respirators to intubate them? And basically nobody knows this number. I did two things. One is, I looked at the difference between the percentage of deaths in Hubei province, where Wuhan is located in China, where this problem of scarcity of supply was real, versus the rest of China, where this probably did not exist. And the difference in percentage is 3.5 percentage points of difference, hence my sort of extra 7 million in deaths.

You can do another bottom-up calculation. We know that roughly 5 percent of the people end up in very critical condition. We also know that typically only one out of five of these dies, hence the 1 percent fatality rate. And the question is, if the five don’t get any hospital support, how many of them will die? If you get that four out of five will die, then you get back to my numbers.

Now, I would be delighted to learn that my numbers are overestimated, because that means that we’re going to lose fewer lives. But even if I assume that the value of life is only $4 million, which would be completely in contradiction with all the literature, way too low. And I assume that the number of deaths only doubled from two million to four, I still get an $8 trillion cost of letting the disease spread without any control, which is more than 30 percent of the US GDP. So, whether you want to believe the point estimates, I think the reasoning is what is important from an economic point of view. It’s not rational to let people die.

Kate: Look, I understand the point that cost-benefit analyses, and cost-benefit analyses involving the value of life, are used broadly in all sorts of policy applications in regular society, and that they can sometimes be useful in taking, like I said, small-probability events and consumer choices around those events and backing out societal implications for the value of life and how much we should protect a certain number of lives.

I think that we’re in a different situation right now, where the stakes are that we could lift all of the shelter-in-place orders that are currently in place, and we could try to return to normalcy and accept that that would almost certainly lead to the loss of life of many individuals. I don’t think that those are comparable circumstances, and especially when it comes to a situation in which you can point to a specific person and say, this policy will almost certainly kill you, I don’t think that the same sorts of rules apply. I don’t think that cost-benefit analyses involving the value of human lives should be used in those circumstances.

Ideally, we would be in a situation like in China, where every single day you’re having zero to one more cases in all of mainland China. And those are the circumstances under which you can lift the shelter-in-place orders. Are we going to get to that point in the United States? I have no idea. I hope so. And at that point we can revisit the issue. But if in six months we’re still experiencing tens of thousands of new cases every day, then we’re going to be in a totally different world. And I’m glad I’m not the person making the decisions in those worlds.

Luigi: One thing I would like to say is I don’t believe that economists have the right to dictate economic policy. I think we only are there to provide tools to help in the decision-making that must be made by elected representatives. And so, I don’t see the cost-benefit analysis as saying, you have to do this or you have to do that. But I see that as a guide to what should be done if you want it to be consistent.

So, one other point is, when are we going to decide to restart the economy? And at some point, we need to make this decision. The lives that might be lost if we restart are not zero. But we don’t want to continue until forever. So, I think that having an instrument like this is useful to guide us. But ultimately, the decision is a political decision.

If I put saving a large number of people as a priority, I think that we can pay an economic cost for that. I always said that if you want to vote for Brexit, you know you’re going to lose some money out of that. But if the sense of pride of being independent from Europe is worth so much to you, that’s a legitimate decision. I don’t think that we economists can dictate people’s preferences. We can only tell them, look, this decision is in contradiction with what you’re doing when you drive or what you’re doing when you set the limit for the amount of pollutants that are in your drinking water. Sometimes political decisions are made on the basis of emotions rather than rational calculation. So, I think it’s useful to have guidance that exposes this potential calculation. And then, if people want to go ahead anyway, I think that’s fine. I don’t think it’s my role to decide what to do.

Kate: This is a discussion that I didn’t particularly want to take part in. Partially because I’m concerned about the opposite conclusion, right? What if you did a cost-benefit analysis, assuming that, say, a million elderly people would die, and you came to the conclusion that it was worth it for them to die? Let’s say it was an undisputed conclusion that it was worth it to restart the economy and let those people pass away. What then? So, I’m going to abstain from this conversation and let Russ Roberts fill in for me.

Luigi: We’re very glad today to have as a guest Russ Roberts, who is a research fellow at the Hoover Institute at Stanford and, most importantly, the very famous host of EconTalk, a great podcast about economics, where he interviews people on values, economic issues. So, Russ, thank you for joining us.

Russ Roberts: Oh, great to be with you, Luigi.

Luigi: And I gather from Twitter that you are not very convinced, not only of my cost-benefit analysis, but, if I understand correctly, of using cost-benefit analysis in this case. Can you elaborate on both?

Russ Roberts: Yeah, I think there are two issues here. The first is, are there any tradeoffs here to talk about, right? Are there any tradeoffs of any kind to talk about? And so, there’s a lot of really depressingly un-nuanced punditry on this question right now. There are people who are saying, at the two extremes, we should do whatever it takes to save a single life. I think that’s incorrect. As an economist, I don’t think that’s correct. And, at the other extreme, we should make a monetary calculation about the value of life and act accordingly. And that, I would say, is effectively what you did, although it was a little bit of what you might call a reductio ad absurdum.

I guess my discomfort with the exercise of trying to make an explicit calculation is that inevitably, some of the most important things don’t get quantified. And so, we put those to the side as economists, and let’s say we can’t measure the full cost, or we can’t measure the full benefits. We say, well, here’s the cost and benefits of what we can measure, and of course you should keep in mind that, fill in the blank. That’s problem number one I have.

Problem number two I have is that, although I understand as an economist that we act often as individuals and sometimes via public policy as if our lives have a monetary value, putting a monetary value on them is really a difficult analytical exercise. We’re able to do it effortlessly by picking a number. But when we start to think about a little more of the complexity of it, is an 85-year-old’s life the same value as a 20-year-old’s life? Is willingness to pay an accurate measure of what we should use when thinking about saving lives through public policy? It’s a really complicated thing. And then, if it extends over any sizable nonmarginal change, the idea that we’ve included all the costs and benefits . . .

I think what provoked me most about your estimate was that I think you said we should sacrifice up to three years of GDP to save X numbers of lives. And I’m thinking, if we turned off the economy for three years, the impact of that is more than just lost GDP. There’s a whole set of really important things, and thinking that we could just say, multiply three times current levels of GDP to get that cost, I think it’s just a mistake. And it leads to a false sense of confidence, because we’ve attached a number to it that has decimal points at the end.

Luigi: I think that, actually, I agree with most of what you say. The cost-benefit analysis has a lot of limitations. I wish that you and other people would remember those limitations when they impose it for every form of regulation. In the sense, we know that, for example, the access to proxy rules was struck down by the court because no cost-benefit analysis was done. There is a very strong pressure to exercise cost-benefit analysis everywhere. So, I’m actually glad that you think this has limited value, because it misses a lot of things. And I agree with you, and I mostly agree with the fact you said, I used the analysis as reductio ad absurdum, which if you don’t know Latin, means as an extreme example, not because I literally believe those extremes, but to show that the counterargument that people say, like President Trump, you cannot shut down the economy, because, after all, there are few people dead, is wrong even if you don’t care about all the other things.

When we introduce regulation, we say the initial standards for dioxin, whatever, we are implicitly putting a value on life and making a tradeoff calculation. I don’t think you are against that idea.

Russ Roberts: No, I think it’s important to try to get a measure of what costs are. I don’t think that’s a foolish idea and that it’s required in certain government settings is, I think, probably a net benefit. What’s interesting about this for me is that I don’t want to defend what President Trump said, and a lot of people have either parodied that or taken their own reductio ad absurdum, which is that, oh, what he’s saying is that people should die to preserve the value of the stock market, or people should die for “the economy.” And I wrote on Twitter that no good economist should ever say we should do something for “the economy.”

And what I added, by the way, is that what we care about as human beings, we care about human flourishing. We often struggle as economists to try to put a monetary value on different aspects of that. So, when President Trump foolishly or stupidly says, “We need to let these people die because of the economy, we don’t want to sacrifice the economy,” well, that’s a horrible thing to say. I abhor that. But I do think there’s something to be said about what we will absorb, what we will accept, what we will live with in terms of quality of life in the aftermath and in the short and long run in response to this crisis.

Just to take an example, I don’t want the airline industry to be nationalized, literally, to be literally nationalized. There’s an argument to be made for it. I also don’t want them to be bailed out. At the same time, I don’t want hundreds of millions of people to live off the government for years and years, if—this is the important if—if there is an alternative. And my view is that there are substitutes for shutting down the entire economy other than, say, the food supply chain and health care, which are the two most important parts right now that we must have. Short of nationalizing them, short of putting everybody else in deep freeze and then giving us welfare payments, that has an unknown cost. I don’t know, can we really afford $2 trillion every two months until this goes to zero? Are we going to have hyperinflation? I don’t know where we’re going.

So, I think that’s the challenge when we think about trying to be analytical about this and saying, oh, well $2 trillion, that’s a bargain given how many lives we could save. It could be. If it’s $2 trillion every two months, it’s not going to be so much of a bargain, because by the way, that $2 trillion is not the real cost. That’s a transfer. It’s a common mistake. But there is a limit to how much we can transfer without risking hyperinflation, a debt crisis, things that might fundamentally alter the United States in ways that I think even you would be unhappy with, Luigi.

Luigi: No, no. Russ, I agree with everything you said. But I don’t understand why you don’t want to have, among your reasoning, some measure of the cost of life you can save. And the piece I wrote for ProMarket two weeks ago was under the information that was available back then, and it was a ballpark estimate to say, can we sacrifice a lot to save these lives? And the answer is, yes. If there are more efficient ways to do that, I’m all in favor.

I agree, and I wrote another piece on ProMarket later that I agree, the lockdown by itself is not going to resolve the problem. It’s going to possibly slow down the diffusion, which means saving some lives because there is a capacity constraint. But it also means extending the pain for a longer period of time. And the system that, in my view, does work is a system where you have at least a temporary lockdown but a lot of tracing, a lot of isolation, like was done in Korea, like was done at a much lower level in the region I come from, Veneto.

And you can see the difference between Veneto and Lombardia in the number of deaths, it’s unbelievable. However, there is also another important issue, because the way Korea implemented the tracing system is extremely intrusive of personal privacy. So, I am in general in favor of protecting privacy. I think here there is a bigger reason to sort of sacrifice your privacy. But I want a system in place that this does not become the norm. It’s like in an emergency, in a war, you do things that you don’t want to do in a normal time. This is a war and we need to do it, but we also need to make sure that it is not replicable.

Russ Roberts: This is a real love fest, Luigi. I agree with almost all of that, except I will put two caveats that I think are important. As you point out, I think things that are working elsewhere in the world might not work here for a whole bunch of reasons. I think one of the tragedies of Italy is that, say, compared to Asia, the population, the citizenry responds differently to a government order than elsewhere. And I understand that. It would be true here.

Luigi: Actually, I’m sorry to disagree here, because I think that there is a lot of stereotyping here that is mostly wrong. First of all, Italians—

Russ Roberts: Those Italian mayors screaming at people to stay at home who are jogging and walking their dogs and doing things that are unnecessary, those aren’t representative of what’s going on, on the Italian street?

Luigi: Have you walked around in Chicago or New York, and seen how many people go without a mask?

Russ Roberts: Yeah. We don’t like it either. No, I’m making a different kind of stereotype. I’m saying that in certain countries, of which I would say the United States is one and Italy, I think, is another, we’re a little less eager to comply with government restrictions, and we’re certainly less eager to accept draconian monitoring and punishments for violations. That may cost us lives. It may be a terrible mistake. But certainly, the example you gave of, say, testing and tracing, following people on their cell phones with the idea that that would disappear after the crisis was over, a lot of Americans are going to be very uncomfortable with that. Maybe they shouldn’t be. Maybe they should recognize this as a unique situation. But that’s my first caveat. So, I’m agreeing with you, except to say that I think culturally, and I’ll just pick on Americans here, I don’t want to pick on Italy, I think a lot of Americans are not going to comply with some of the things that we might think would work better than others.

And so, while I agree with you that certain things might be more effective, I think some of them are hard to implement in a large country, in a country with the culture of America, et cetera. But I want to make what I think is a more important point, which is that in a time of war, in a time of fear, in a time of insecurity, people are prone to accept policies that they wouldn’t accept at other times. And they’re right to, and that’s part of what your message is, and I agree with that message.

Here’s the problem. People like to exploit that. They like to use that as a way to gain power. One of the things that worries me that’s not getting a lot of attention, maybe it shouldn’t, but I think it should, is whether some of this is going to be exploited for personal gain for all kinds of people. Certainly for businesses, but also for politicians. And I’m extremely uneasy with the current president of the United States, or any president of the United States, having the kind of authority that we’re eager to often give them in the face of, say, 1.5 million deaths.

If you tell me there’s going to be 1.5 million deaths in America and not just old people, but a lot of young people as well, if we don’t do fill in the blank, X, Y, Z, a lot of people are going to say, “Oh my gosh, we have to do it, you’re right, oh my gosh.” But if it’s really not 1.5 million and if there’s other ways to get there that don’t encourage the acquisition of power in Washington, I think that’s something we ought to be thinking about and talking about. And I worry that the cost-benefit analysis approach that you used as a reductio ad absurdum can feed into that kind of what I think is dangerous.

Luigi: Russ, I agree. But, however, you have to be careful because this information, in my view, is already in the hands of the president, in the sense that Apple, Google, Facebook are able to trace you better than the Korean government can trace you. If I am President Trump and I want to get information today, I can easily get it. I just say, I will pass a law to break up Google or break up Facebook if you don’t give me that information. He’s going to get that information on the table the next morning. We know that. So, that’s the reason why I think that actually having a system to gather the information that is reviewed by the judiciary, like wiretapping. And of course, there are some exceptions. Apparently, some stuff was granted that shouldn’t be, nothing is perfect.

Russ Roberts: It’s inevitable.

Luigi: Yeah, it’s inevitable. Overall, I don’t think it’s such a bad system. So, I am willing to do it because I think that I already lost my privacy to Google and Facebook a long time ago.

So, there’s one thing, I don’t know, this is a concluding thing. Maybe he has something else. But let me say before, I want to actually commend you for inviting me to this debate, Russ. And I think this should be an example, because we started to trade messages on Twitter. And Twitter is designed to escalate. And I think that Russ, who is wiser than me, understood that and actually sent me an email and said, why don’t we talk? And the result is when we talk, we agree more than we disagree. And I think we have slight disagreement, but I think, overall, I don’t think I’m overselling it, we agree more than we disagree. Had we continued on Twitter, we would have hated each other for the rest of our lives, probably.

Russ Roberts: That’s right. And I think the other point to make, which I think is glorious about the current technological landscape, for all of its flaws, is that Twitter, I think of it as sort of a window or a door. It’s the beginning. It’s something you open that leads you to something more rich and deeper and interesting. That’s in its ideal sense.

The nonideal sense is I open my door into your house and I scream invective and venom, and you scream it back, and gestures are made and shots are fired, and people get entertained by that, by the way. But that’s not what I want to do on Twitter.

The other point is that, not only did we stay friends, but I learned something from our conversation today that I wouldn’t have otherwise learned. And the people, I hope, who are listening learned something about the subtlety and nuance of our agreement and disagreement that they would never have gotten from a Twitter war. So, I think this combination of technologies is really a glorious thing. And I hope, if not Tim Cook and others, I hope President Trump’s listening.

Luigi: Unless he’s busy tweeting himself. And thank you very much, Russ. I learned a lot, and not only in economics but a lesson about life. Thank you.