Shane Dunn:
Glad you joined us today. We have business school alumni, and we have current students whom we miss very, very much at Brandeis International Business School. And we also have staff from Brandeis and the International Business School joining us today. So we're grateful for all of your participation and attendance. And we hope you can stick around to the very end. But if you have to jump off sooner, that's completely understandable as well. We wanted to do a lunchtime faculty talk to capture as many people as we could. But we know during this environment, in this existing world we're in, there's no perfect time to capture everyone. But we're grateful for those willing to join today. So thank you.
Shane Dunn:
First off, my name is Shane Dunn. And I'm a senior director of development and alumni relations at Brandeis International Business School. Before we get started with our program, I want to learn a few ground rules to help us conduct a really excellent experience for all attendees, as well as our speaker today via this virtual environment on Zoom. We too of course at the Business School are figuring out how to best engage folks virtually. We have more programming coming that I'll mention to you later today. Brandeis is continuing to experiment with how we engage folks. And we know there's obviously a range of other programming that folks are engaging with online currently. So we're doing our best here. And I think you'll have a good time today.
Shane Dunn:
But it does start with some ground rules to make sure you have a stimulating experience. The first is, if you have not already or we have not done for you, please mute your microphones. And also turn off your video. It will just help us with bandwidth control. At some point if you want to speak, or you are given the chance to speak, if we get there you can definitely turn your video. But right now, please mute and video off. It will just help significantly. Thank you.
Shane Dunn:
The second piece is if you've not done so already, please make sure your name is updated. It's pretty easy if you click the three dots next to your screen in the upper right-hand corner. You're able to rename to whatever you want, but we of course hope you put your name in there. And you should feel free, because we have people in a lot of different areas to put where you're from. Are you in Boston or Massachusets, are you somewhere else around the world or on the country? It can't hurt just so we can kind of try to build some engagement and connectivity together as a group.
Shane Dunn:
And then the final thing is, you should feel free to use the chat as much as you'd like to connect with folks. If you see them, feel free to send a private chat. That also would be the way that we do questions and answers with our speaker throughout this session. So the first thing I would like to do before we jump into our formal presentation is take a poll, which I think will be popping up shortly. Yeah, here it is. So those of you who are able to see this, we have a quick poll. We're just trying to get a sense of who's with us. We have alumni, we have current students, staff, faculty, parents, or other. So we're going to take about 30 to 40 seconds here just to get everyone voting. We're going quickly. Thank you.
Shane Dunn:
Nice. So it looks about six or seven people still need to go. Five. 10 more seconds. Four more people. All right. I'm going to end the poll. Thank you for participation. Thank you all. So we actually have a lot of staff on the call, which is wonderful. So I know we have staff at the business school. We have staff in other parts of the university. So we're grateful to staff who are interested in hearing from Professor Scherbina, and just connecting with others. We also have several alumni. And as I said, we have faculty and parents, but it's also great to have a handful of students on the call with us today. I know you all are spread out around the world, and your whole semester is upended. But we're hoping you have a good experience right now. And we're glad we and our team, and that the business school could bring you a little bit of additional content to your experience this semester.
Shane Dunn:
So I'm going to now move us onto why you all are here, which is hear from our speaker. I am grateful that today we have Anna Scherbina, an associate professor of finance at the Brandeis International Business School joining us to talk about her research on the potentially, and now real, health and economic losses in the United States due to the COVID-19 pandemic. Anna will speak about her background in research for about 20 to 25 minutes. And then she will take clarifying questions. Or if you think there's something that you actually want to share throughout, you should us and engage with us via the chat window, and we'll try to get to your questions during our presentation. If not, we'll do a Q&A at the very end. Our goal here is to be done right around 1:20 P.M. Eastern Time, so that folks can get back to their days, including Professor Scherbina. So feel free to use the chat window throughout. We'll do our best, and I will do my best to moderate this. And I hope that it's a really good experience. So I'm now ready to turn it over to Professor Scherbina. Anna, please go ahead.
Anna Scherbina:
Okay. Hello, everybody. And welcome. And thank you, Shane and Gina for setting everything up. And for inviting me to give a talk. So I will right now discuss my ... Okay. Let me advance the slide. Okay. I hope you are seeing slides advancing. So this is the topic of my talk today, Evaluating the Optimal Health Policy Response to the COVID-19 Pandemic. And this based on the paper that I recently finished. So actually I just updated the paper with the new data. Because everything is changing so fast with this pandemic. The number of new cases is really growing exponentially. So the conclusions have changed relative to the paper that is published on the AEI website and on [inaudible 00:05:59]. And today I will present the results with the newest data.
Anna Scherbina:
So first, let me quickly talk about my background. I have a PhD in finance from Kellogg, Northwestern University. And prior to joining Brandeis, I was also a professor of finance at HBS and UC Davis. I am currently also a visiting scholar at the American Enterprise Institute. And for the last two years from 2017 to 2019, I worked at the Council of Economic Advisors at the White House as a senior economist. And you know how economists, they are responsible for different topics that they cover. So my main topic was obviously finance in fintech. I also looked at artificial intelligence policy, things related to technology. Obviously because there was so many things of trade, I worked on trade as well. I did a lot of work on cyber security, which kind of opened up my horizon to how important cyber security is in the current environment.
Anna Scherbina:
And I also worked with the National Security Council, with the people who work on bio threats. So that's our topic. So the National Security Council also was divided by topics, is the largest component in the White House. And one of the departments handled bio threats. So with them, I worked on the executive order to develop better vaccine technologies to fight influenza. And I was one of the authors of the CA paper on pandemic influenza, where we analyzed the possibility of pandemic influenza coming to the United States, analyzed what could happen, did a model of pandemic influenza. And the conclusion of that paper was that we need to develop better and faster, and more scalable vaccine technologies. Because that will help us not only in the case of a pandemic, but also in the case of seasonal flu, which nobody really talks about. But is really deadly, especially in some years when the strain is a little bit different from what we prepared for, could be really highly deadly. And there could be tens of thousands of deaths.
Anna Scherbina:
And little did I know that coronavirus also poses a threat. And so I actually used a lot of that knowledge that I developed to write a paper about COVID-19. So COVID-19 pandemic turned out really a serious threat. I looked up the numbers yesterday. And right now we have over half a million total cases in the U.S. And almost 22,000 deaths. And because we know that there's no widespread testing, a lot of people want to get a test, they can not. And some people are really even discouraged from going and getting tested. Like in New York City, the system is so overlong, people are not getting tested for example. And you could imagine people with the milder cases don't get tested.
Anna Scherbina:
There's speculation that maybe the number of cases is twice as many. And that's the assumption that I used in my analysis right now that we had about one million cases so for in the United States. So unfortunately right now, we don't really know too much about this COVID-19 virus. But we know some things. So we know that this virus is very, very contagious. It's actually more contagious than seasonal flu obviously. But also more contagious than pandemic flu that we analyzed in the paper with the different scenarios that we analyzed.
Anna Scherbina:
It's more contagious, and also more lethal. So if you look at the infection fatality rate, how many who are infected end up dying, we don't really know what it is for sure. Because we can't really observe all the infections. Because as I said, not everybody gets tested. But some speculate in their published paper that it's maybe 10 times deadlier than seasonal influenza. So with seasonal influenza, only .1% of the people who get it die. With this virus, some speculate maybe the infection fatality rate could be as high as 1%. As many as 1% of people who get it die. And some speculate it could be even higher than that.
Anna Scherbina:
And obviously it very much depends where you are looking. If you are looking at Italy, where the population is older and there are a lot of smokers, their fatality ratio is higher than maybe Germany, where maybe the population is a little bit younger. So what we also know about this virus, and again how it's different from influenza, is that while flu poses a higher risk for both children and older adults, with COVID-19, it seems like children are not really that much at risk. It seems that the probability of dying and probability of developing severe complications is much higher for older adults.
Anna Scherbina:
And what is true about COVID-19, and it's not necessarily true for pandemic influenza is that nobody has any immunity to it. With flu, sometimes older people you may find have immunity, because they've been exposed to a similar strain of flu some time back in the day. But with this COVID-19, it seems like nobody really has any immunity to it. Okay. So this is what we know so far. And before I get into my model, I just want to introduce some very basic terminology so that we're on the same page. And this is something that's the first thing that I learned when I was working on the pandemic flu paper. The R0. So what is the R0? Everybody talks about it all the time.
Anna Scherbina:
So for any kind of virus, R0 is very important, is the basic reproduction number. So it measures how infectious a virus is. So when a pandemic is just starting, and at the start nobody is immune, it measures how many people on average, and in fact in person infects over the course of their illness. So with COVID-19, we know that people are infectious for the next two weeks. So this is over the two weeks that a person is sick with COVID-19, they infect R0 people on average.
Anna Scherbina:
So R0 is important, because if it's less than one, it means that the epidemic is dying out. Each person infects less than one person. So the number of new cases is decreasing. And eventually it will decrease to zero. If R0 is greater than one, then you will see a number of new cases initially increasing. Then reaching a peak, and then starting to decrease naturally. Because there are more people, more and more people develop immunity. So when you meet somebody randomly, there's a higher and higher chance that whoever you meet is already immune. So the pandemic just dies out naturally.
Anna Scherbina:
Just to put R0 into perspective. [inaudible 00:13:28] flu, it's on average across all the different studies. There was a paper analyzing R0 for different seasons of flu. It's on average 1.3. So it's neither high nor low. What 1.3 means is that the number of people is increasing initially, and then starts decreasing. So it's not really high, because we know that we get vaccinated for fly, and people have immunity from previous strains. So that's why it's not huge.
Anna Scherbina:
For Spanish Flu, right? Everybody talks about it right now, because that's the pandemic that is on everybody's mind. Analyzing the studies based on the data, R0 was higher than for seasonal flu. It was somewhere between 1.4 and 2.8. For Ebola, R0 is relatively high between 1.5 and 2.53. With Ebola though, so from my work I remember, people said why it was easier to contain Ebola is because people who are asymptomatic, they're not typically infectious. So you could tell who is infectious, because they have the symptoms. And with flu it's very tricky, because people could be asymptomatic and still infect others. And it's the same with COVID-19. It's very, very tricky. Can not just measure people's temperatures and say, "You are not infectious." So that makes it very difficult.
Anna Scherbina:
So measles, you know how everybody talks about outbreaks of measles. R0 is really, really high. It's between 12 and 18. With COVID-19, it's not as high. But it's really, really quite high. Like I said, it's higher than what we assumed in different scenarios for pandemic influenza. So different people have different estimates for what R0 is for COVID-19 based on different studies. Some studied the China, some studied the Princess Cruises, the cruise ship where a lot of people got sick. So they're different estimates between 1.5 and 3.5. I will use the same assumption for what the national R0 is. It's the same as [inaudible 00:15:38] college paper. I'll assume that the natural R0, the [inaudible 00:15:41] R0 for COVID-19 is 2.4.
Anna Scherbina:
So that means that each person who gets sick at the beginning of the pandemic when nobody else is immune, on average over the course of their illness will infect 2.4 other people. So the infection will spread really fast. Okay. All right. So what happens? So this is what my model shows, assuming that R0 is 2.4. And the person who is sick, is sick for two weeks. So that's the period over which they infect 2.4 other people. What will the number of new cases look like?
Anna Scherbina:
This is what the pandemic curve would look like. And of course I use my own model parameters. You could plug in different R0, you could plug in a different number of people who [inaudible 00:16:29] infected, you get something a little bit different. But this curve actually looks very similar to what the curve looked like for this imperial college paper that came out at the beginning of the pandemic. So I assume that R0 is 2.4, same as they. I assume that initially right now, a million people were already infected. And so this is what it looks like. The number of new cases increases really, really fast in the beginning. Then it reaches a peak some time in the middle of June. And then the number of new cases start decreasing, eventually decreasing to zero.
Anna Scherbina:
So why the number of new cases initially increase was because nobody is immune. You meet somebody randomly, you are very likely to infect them. And eventually, there's this herd immunity everybody talks about, when you meet another person, this person has a higher and higher chance to have been already sick and to have developed immunity. So that's why the number of cases drops. And you see because this pandemic is so infectious, it moves really fast. So the peak, if you don't intervene, the peak should have come really fast.
Anna Scherbina:
And so how many people will end up getting sick if we did not intervene? So how many people are sickened by a pandemic is called the pandemic attack rate. So I calculate 56% is the attack rate. So 56% would've gotten sick with the symptoms if we did not intervene. And using the infection fatality ratio I calculated from the CDC reports, and I assume that the Center of Disease Control can only observe 50% of cases, because the other 50% just don't get tested. So I calculated that infection fatality rate is .7%. So that COVID-19 is four times deadlier than the seasonal flu. So I calculated that 1.3 million people would've died. And so if you translate those numbers, people are getting sick, and people are dying into dollar terms. And I will explain how I do in just a second.
Anna Scherbina:
I would calculate the cost to the economy would over $9 trillion. Which is over 40% of the U.S. GDP. So how do we calculate the cost of the pandemic? So the cost has three components. The first is that people who are sick, they can not really work productively. They have to take the time off from work. And so their productivity drops. Second component is that some of those people are so severely ill that they have to go to the hospital, and medical care costs a lot of money.
Anna Scherbina:
And then the most important component is the people who are dying. Each death is a loss to society. We would pay money not to die, right? We would pay money for medical interventions in order not to die. So you could actually quantify the cost of life. And I have ... So the cost of life is quantified through the value of statistical lives. I use an estimate from a paper from [inaudible 00:19:42], for people in different age bins. So on average, the cost of statistical life in my sample is over $6 million. Because so many people will end up dying if we do nothing, the total cost of a pandemic would translate in over $9 trillion.
Anna Scherbina:
So looking at this astronomical cost, we know that doing nothing is not an option. Then even though we don't have a vaccine, and we don't have any effective treatment, we have to intervene in some other ways to stop the spread of this pandemic. Okay. So what can we do? So what we're going to do is the non-pharmaceutical interventions. And this is something that has been done during the Spanish Flu in the 1980s. So non-pharmaceutical interventions are described in this famous Imperial College paper by Neil Ferguson and his co-authors that just came out. And it's also described in the Roadmap to Reopening, A Policy Paper by Scott Gottlieb and his co-authors. Scott Gottlieb was the former FDA commissioner who just stepped down recently.
Anna Scherbina:
So what do those non-pharmaceutical interventions entail? The first phase is suppression. So we're try to suppress the virus. We're trying to reduce the number of new cases to the point that we can manage the virus better. And this is what we're doing right now. We're in this lockdown phase, which is the suppression phase. And you have heard that it's working, because what we're doing is we're reducing the chance that each person meets other people, and passes on the infection to others. This is what the social distancing is all about, making sure people are not meeting as many other people. Because we don't really know who is infected, who is not. So just reducing the number of meetings helps us contain the virus and reduce the number of cases.
Anna Scherbina:
The goal of a suppression is to bring R0 to below one. To have a reduction in a number of new cases. And that's what is happening right now, it seems to be working. The number of new cases is dropping. Even though the number of deaths it seems like is not dropping yet, but it's because people who got sick maybe three weeks ago are dying right now. But the number of new cases is declining. And yeah, how do we do this? We just don't let people meet in large groups, don't let a lot of people go to church, close non-essential businesses, ask people to shelter-in-place, and so on.
Anna Scherbina:
So what happens next? So if you look at Scott Gottlieb's plan by the way, he has the four phases. He doesn't really call them that, but he calls this phase, phase one. Suppressing the virus. The next phase is going to be mitigation phase. We're not going to just return to normal. We're going to try to mitigate the virus. So we're going to reopen the economy to some extent, but still discourage people from meeting a lot of other people, because we still want to reduce the chance that the virus will be passed on to others.
Anna Scherbina:
The goal of mitigation is to bring R0 to a value lower than its natural reproductive rate. So a value lower than 2.4. So if you read the Ferguson paper, probably through the mitigation phase, because the restriction on people's lives are not as drastic as during the suppression phase, you will naturally meet more people. So they are arguing, you can not really reduce R0 to below one. But probably you could have it slightly above one. So it will still spread, the number of new cases will still grow. But it just wouldn't grow as fast if we did not intervene.
Anna Scherbina:
And people ask, "What could mitigation entail?" We're still trying to figure it out, and this is why there's all this discussion about, how do we reopen the economy and not let so many people get infected? So I think we're still going to very much encourage the elderly, the vulnerable populations to do social distancing. We're still going to ban large gatherings. [inaudible 00:24:07] we're going to ask people not to fly, to ask people to work from home as much as possible. And if we get the number of new cases down low enough, we could probably do a lot more in terms of contact tracing, and isolating people who we think are infected. Because it doesn't really make sense to isolate everybody if we could then identify who is infected and who came into contact with an infected person, and just isolate those people.
Anna Scherbina:
And with a lot of testing, if that becomes possible, we could identify people who are immune, and those people can return back to work and perform critical functions, for example in the healthcare or in law enforcement, or in the military. So that's the second phase, and we will see how well that's going to work. I guess we're all learning from China, and we're going to learn from Germany that is about to reopen, and see how to best do this.
Anna Scherbina:
The next phase according to Scott Gottlieb's plan is immunization. Once a vaccine becomes available we will start immunizing people. He's proposing we immunize people who are more susceptible first, like the elderly and people with health problems. And eventually everybody else. And then phase four is what we haven't really done after the last pandemic, we need to actually restock our medical stockpiles and prepare for the next pandemic threat. So the central question that I'm trying to answer right now, and that's what's on everybody's minds right now, how long should this lockdown phase, phase one, last? When should we start reopening the economy?
Anna Scherbina:
Some people say we should start reopening already at the end of this month. And you see governors arguing that that's too soon. De Blasio wants to keep schools in New York City closed throughout the summer. Some people argue it should go on until at least the end of August. And Bill Gates says it should go on for another 10 weeks. So how should we think about it? So this is what my presentation is all about.
Anna Scherbina:
Okay. So this is what I tried to analyze right now in this study. I say what happens if suppression is lifted now or after four weeks? Nobody's talking about right now, but after four weeks at the end of April, beginning of May. What happens if suppression is lifted and we go into this mitigation phase until the vaccine is available? Nobody's talking about fully reopening the economy until the vaccine is available, with no doubt we have to do something until some kind of treatment is available.
Anna Scherbina:
So this is my baseline scenario, if we reopen the economy at the beginning of May. And then I say, well compared to that, do we gain anything if we extend suppression a little bit longer? What if we extend it through the first week of May, the second week of May, and so on? And so we gain more, and what we'll lose by limiting the productivity of the people through the lockdown. And so what I compared is this trade off. Do we gain more by using suppression for more weeks, and preventing more people from getting sick and dying at that time, and also in the future?
Anna Scherbina:
Or do we lose more, because economy is obviously less productive. People working from home, working parents are less productive. So what is this optimal trade off? And this what I do this cost benefit analysis say, what is the optimal time when the benefit of doing the lockdown stops being higher than the cost of doing the lockdown. So that's the point that lockdown should be lifted.
Anna Scherbina:
Okay. So to analyze this, let's go back to understanding what happens when a person is infected. So we will first try to understand the cause of the pandemic, and how to quantify them. If a person is infected, they could be symptomatic or asymptomatic. And you hear a lot with COVID-19, some people just say they never have any symptoms. But of course the troubling thing about this is that people don't have symptom, but they still go on and infect others. So that's why we want to quarantine everybody, right?
Anna Scherbina:
So a person who is ... And by the way, that's the same thing with the flu. Some people just don't develop the symptoms, but they could still infect others. So if a person is infected and they are asymptomatic, they are fine, they're not going to die. They don't really take the time off from work. They don't become sick, don't use any medical supplies, et cetera. So they don't really impose any cost to the economy.
Anna Scherbina:
But the problem is with people who develop symptoms. The symptomatic people could have a very mild case. And maybe they just take a little bit of time from work, but they don't really require any medical interventions. Well, they may be again, a little bit more sick and maybe have an out-patient visit, and maybe use a little bit more medical services. They could have a worse type outcome where they would actually end up hospitalized and use medical resources a lot, but then they would survive. And the worst type of outcome is that a person becomes ill, they get hospitalized, stay in the hospital, and then eventually die. So those are the different outcomes that could happen.
Anna Scherbina:
And then we tried to quantify the cost of each of those outcomes. Obviously asymptomatic people is not costly. But the cost of each person is increasing the more severe their condition is. And so what are my assumptions? Of course I don't really know what's going to happen in the future. And at this time, we don't really have, or at least I don't have a lot of data to know how well the suppression is working. So far we know that it seems to be working, because the number of new cases is dropping. So I used two different assumptions for how well the suppression period, this lockdown period is working right now.
Anna Scherbina:
My more pessimistic assumption says that we're able to reduce the R0 to .7 It means that each infected person right now is infecting .7 people on average. Why are they still infecting people? Well, you still can infect people in your household, because you live with somebody. You still go to grocery stores, you can infect people there. And you can infect people in the hospital if you end up in a hospital. So then those people infect .7 people on average.
Anna Scherbina:
If we think suppression is working better than what I assume, I also use an optimistic assumption that would drop R0 to .5. So that means that each infected person only infects half a person on average throughout the duration of the illness. So what's going to happen next when we go to mitigation? As I said, that's a big unknown. We just don't really know how that's going to work. I wish we had a good and very clear plan what we're going to do. But we just don't really have it at this point. So it's a lot of speculation.
Anna Scherbina:
So I assume that if we dropped the number of new cases sufficiently enough to below 50,000 cases, we could do a lot more than if the number of new cases is higher. So if we drop the number of new cases to less than 50,000 cases, we could identify sick people and also do a lot of contact tracing. So say you got sick, "Who did you interact with in the last two weeks? Let's go and isolate those people as well. Because they may not have any symptoms, but they may be contagious. So let's isolate them." And through this, we could drop the R0 to 1.1. Just slightly above one. Meaning that the number of new cases will increase, but very, very slowly.
Anna Scherbina:
And if we still have a lot of new cases, over 50,000, I assume that mitigation, all these restrictions on large gatherings, encouraging people to work from home, encouraging people not to go to concerts or churches could achieve the R0 of 1.3. So that's something kind of on the order of magnitude of seasonal flu. So that I think is reasonable. And of course if we don't really manage, then mitigation could have an R0 much higher than that. So then my analysis would not be right then, we should keep suppression even longer. I'll talk about it in a little bit.
Anna Scherbina:
And so yeah, I think I'm kind of running out of time a little. Okay. So I'm going to move on. So I have some additional assumptions that I talked about, I assume that 60% of people are symptomatic, 40% are asymptomatic. That's based on the data that is coming out. Only 50% are diagnosed, so right now we're starting with one million people being immuno-infected. So what I really by my research shows is that if you keep the suppression in place, to let's start with R0 .7. If you keep the suppression, let's look at the first graph, which is on the upper right.
Anna Scherbina:
If you keep the suppression in place for only four weeks, and then you go to mitigation, you see how the pandemic curve rises and falls before the vaccine becomes available 18 months from now. That's my assumption. So really it does not really achieve as much as it could've. We could've just gone to mitigation right away, it would be the same exact outcome because a pandemic curve will play out before the vaccine is available.
Anna Scherbina:
What if we start the lockdown after eight weeks? Look at the next graph on the right, the upper right. You see that now the curve moves forward in time? And the right tail is a little bit cut off, to some extent cut off because the vaccine becomes available. So let's go to the lower left corner, and look at what happens when we keep suppression for 12 weeks. Now it looks actually better, because the entire pandemic curve doesn't really have a chance to build up, because it's cut off by the time that the vaccine is available.
Anna Scherbina:
And your goal is to ... Let's look at the graph on the lower right. If you keep suppression in place for 17 weeks, then the pandemic curve just doesn't really have a chance to build up at all. You're running into the time that the vaccine becomes available. But the question is, does it make a big difference if you do 12 weeks versus 18 weeks? Do you really achieve that much relative to the cost of [inaudible 00:34:29] the lockdown? And I say no, you have to stop before 17 weeks, because the gains at some point, they stop. They start being higher than the losses from the cost of the lockdown.
Anna Scherbina:
And if you switch to .5, you see that because we achieve much more with suppression right now if R0 suppression is .5, if it's working so well. You can actually stop it faster, because it reduces the number new cases so much more in a shorter period of time that you could stop it sooner. And when you go to mitigation, you won't have a lot of time to build up the pandemic curve before the vaccine is available. So that's kind of the bottom line.
Anna Scherbina:
And so if you look at the incremental saving for extending suppression beyond four weeks, you see on the top you have the graph with R0 equals to .7. That is assuming that suppression is not working as well as the bottom graph. So you see the returns of extending suppression by each additional week beyond May 1st are really high for a while. So it really makes sense to keep extending suppression by more and more weeks beyond May 1st. But at some point, you see that the curve flattens out. So it means that even though there are additional benefits, the increase in benefits of extending suppression beyond a particular point don't really necessarily exceed the cost of a lockdown. And that's the whole point.
Anna Scherbina:
So then I tried to estimate what the cost of the lockdown may be incremental to mitigation, because we know we're not going to just reopen the economy to the way it was before, we'll still have to do mitigation. We'll still have to stop people from gathering in large numbers. So I estimate that those incremental costs of the lockdown are $36 billion. So let's look at those graphs. So when do the incremental benefits of the lockdown remain higher than the incremental costs of the lockdown?
Anna Scherbina:
So if you look at the top graph when R0 is equal to .7, the incremental benefits are higher than the cost for the next 17 weeks. And then that should drop below. When R0 is equal to .5, the incremental benefits are higher than the cost for the next nine weeks, and then they drop below. So if we believe that suppression is working really super well right now, that R0 is equal to .5, then we should keep suppression in place for the next nine weeks, until some time in the beginning of June.
Anna Scherbina:
And if you do that, you are able to save the economy $3.7 trillion. Which is 17% of the U.S. GDP. How? Because you prevent people from getting sick, from missing work, from using medical resources, and from dying, which is something that we need to quantify. And so then if you think that the costs of the lockdown are higher than expected, then we should optimally shorten the time of suppression. If you think suppression is not working as well as what I assume, then you should optimally extend it. Because then you could reduce the number of new cases more, and you could go into mitigation with a lower number of new cases. And if you think that mitigation is going to work worse than I expected, then you should optimally extend the lockdown for a longer period of time than I calculated there.
Anna Scherbina:
And then I have some thoughts about how to best reopen the economy. So maybe right now I don't really have a lot of time to talk about it. But that's along the lines that I talked about earlier. Maybe it doesn't really make sense to lockdown and prevent everybody from working to the same extent. Because some people are healthier, so they have lower risks. And especially if they don't interact with other people, if they don't really have anybody they live with, then they have a relatively lower risk of passing on the infection to others. Maybe have those people return to work. Additionally what we have right now, we have people who are very valuable to the smooth functioning of the economy, scientists who develop vaccines, you have law enforcement, you have firefighters. So maybe have some kind of score based on how valuable it is to have a person back.
Anna Scherbina:
And what people have already discussed in the past is that in some fields, like let's say farming or construction, you could probably return to work safely and not be at a close distance to other people. So maybe have those people return to work. And what obviously we'd know about COVID-19, I use a lot of assumptions for pandemic flu. But what we now are starting to understand about COVID-19 is that people who are survivors of coronavirus, they may have long-term health consequences, which I have not quantified. So then we should take those into account, and add them to the cost. And if they're in fact a very expensive long-term consequences to people's health, then we should actually extend the suppression to longer period of time to prevent even more people from getting sick.
Anna Scherbina:
Just to summarize, if we do the lockdown right, it's a really important tool that we got to achieve our objective to reduce the number of people getting sick now and in the future, reduce the number of death, we could actually benefit the economy tremendously to the order of about 17% of the U.S. GDP. Just a last slide. Additional references. Some additional reading. So if you're interested, I have the paper which is posted on SSRN. If you're interested in looking the famous Imperial College paper, I have a link here, The National Roadmap to Reopening by Scott Gottlieb and colleagues is also posted here. I'd like to look at Paul Romer's webpage. He's doing some simulations. And then finally if you want to contact Shane or me, I have our email addresses. And thank you again for listening. So if you have any questions, we'll just go back to Shane right now.
Shane Dunn:
Yeah. Thank you, Anna, for going through that. I've actually heard from a lot of folks who are really grateful, and just acknowledging this great work and your presentation. So thank you first. I am going to give you three questions just to try to get to as many as we can with short time we have. And you can answer however you're able in the few minutes we have. The first, there is a question going back, you don't have to go to an earlier slide around the phases you mentioned. Obviously one is to prepare for the next pandemic the best we can. There has been reporting or acknowledgement that maybe the current administration might have not followed some previous pandemic plans from a pervious time. So how does government if this happened again, ensure we are actually prepared and can follow through on recommendations? That's number one from your perspective, and how the economic models dictate some of that.
Shane Dunn:
Number two is, we have a PhD class of 2014 alum, Avina, who works at the IMF, who has a question on lockdown exit strategy. "If R0 is hard to calculate in real time, but we have the ratio of new cases to cumulative cases, can you translate your assumptions on R0 to this ratio that governments can use to have a timeline to exit the lockdown?" Let Avina or me know if that is clear or not. And then the final question I have here, Anna, if you don't mind is, how robust are the results compared to using other basic epidemiological models? And
Anna Scherbina:
Okay.
Shane Dunn:
... there and see if we get any more in the next couple of minutes.
Anna Scherbina:
Okay. Yeah. Thank you very much for all these questions. So the first question, how to prepare for the next pandemic. I think our supplies got some more depleted with the swine flu, and somehow never restocked. So I don't really know where the failure came from. But when I was working on bio threats, specifically on pandemic flu, I was actually quite happy that somebody is working on it. And there were a lot of agencies involved. There was money being spent on trying to find new vaccine technologies. And I was actually quite happy that people were really thinking carefully about pandemic flu and understanding that that's a real threat.
Anna Scherbina:
So I don't really know where the failure came from in terms of coronavirus, why people were not preparing for the novel coronavirus. But I think the easiest way to do it is just to, after this pandemic is over, hopefully soon, just restock everything. So restock the ventilators, restock the face masks, restock everything that we have depleted. And then continue just expanding the horizons beyond just pandemic flu. Just think about coronavirus or maybe any other infections, and do tabletop exercises on those. And do the scenarios, "What could happen in case of those other pandemics?"
Anna Scherbina:
The second question I think was about R0, how to calculate R0. Yes, so I think one could look at the data. So if one has really good data, and one can do the number of new cases by week, and has a lot of good data in terms of testing that you really are testing a lot of people, then it should very easy to calculate, right? So if you know that one person is infectious for two weeks, and then you see what the initial number of cases were, and you see what the number of cases is in the next two weeks, you could actually back this out from data. But we don't have good data, and that's our problem right now. Just because not too many people get tested. And it's really hard to know. Maybe somebody knows, but I just don't have the good data right now. And the last question was about government ... Shane, what was the last question?
Shane Dunn:
How robust are the results compared to using other basic...
Anna Scherbina:
Ah, right. Yes. So I have a very simple model. In my simple model, this is really what I need to just show what I want to show. I assume that everybody mixed with everybody else. A person in Boston can meet a person in California. And that's probably not realistic. So the people who do the epidemiological models, they probably have geography specific models. Because we know that probably we in Boston can meet other people in Boston, but we don't really travel right now much outside of Boston.
Anna Scherbina:
So I think you could have a real improvement on my simple model by incorporating the geographical element. And that's why people in terms of mitigation, they say, "Maybe we could do this mitigation by geographical region." Maybe in some rural areas where the population density is not that high, we could likely find there are not a lot of new infections. And we know people don't travel in and out of this area that much. Maybe somewhere up in Maine, maybe we could let those people go back to normal. But maybe not in New York City, because the population density is so high. And the chance of just running into somebody on the street is so high, and so many people are infected. But people from Maine and from New York that probably don't mix that well, especially with discouraged travel. So I would say that's a really clear way in which this model that I have can be improved.
Shane Dunn:
Anna, this might be our last question. One more from Jennifer, who's an MBA alumni. She's curious how effective you think certain financial innovations such as pandemic bonds may be in terms of fighting current and future pandemics.
Anna Scherbina:
Yeah. I don't know actually. I should read about this innovation. I haven't read about this. I think financial stimulus is very, very important. How we set it up, how we structure it is going to be extremely important in terms of not getting into a deep recession. Because we know some people can not work right now, they're running out of savings, they are losing jobs. Maybe they're just kind of losing houses, apartments. So we need to find a way to support those people right now through the hardship. Because we know it's for the common good, and we'll come out on the other side. And hopefully the personal balance sheets are not going to be too much weakened, and same with corporate balance sheets. That's why we're doing all these cheap lending to small businesses.
Anna Scherbina:
Because we know that their number of customers has decreased. So we don't want all those businesses to go bankrupt. We don't really want people to go bankrupt, because we know bankruptcy is just a dead weight loss to society. It's like a loss of intellectual property, a loss of assets. So we want to structure the stimulus the best that we can.
Shane Dunn:
Thank you. I'm going to give us 10 more seconds. If I don't see any more questions, we'll sign off, Anna. But giving us one more chance. Okay. Well, Anna, do you have any final words before I close us up?
Anna Scherbina:
Yeah. Well, thank you so much for participating. And it's actually something that is so important for us. I really hope we could get it right. And I really hope so much that we could prevent people from dying in the future. This is so important right now to really try to get it right. And thank you so much for all thinking about how to do it.
Shane Dunn:
Yeah. And I of course echo that. Anna, thank you for doing what you've done to bring light to this situation, and in your previous work and obviously in your current work. We're grateful to have you here at the Brandeis International Business School. And I want to point folks to this final slide that's currently on the screen. We will follow up with an email with links to Anna's paper. And then also we are recording this session, so we're happy to share this. And you can let us know if you want us to share it with some other folks. We'll put it on our website. You can feel free to email me, sdunn@brandeis.edu or Anna if you have questions or are looking for resources. I heard from folks during this session that they're eager to follow up. So Anna, hopefully you'll get some emails after this.
Shane Dunn:
I want to let everyone know just finally, especially our alumni who are still on the call. Again, in this new environment, new world, Brandeis University including the Business School, we're planning a range of events in the next several weeks and months, and beyond to engage alumni with a range of topics and social opportunities. So be on the lookout in your email. Reach out to us if you have any ideas. One event the Business School is hosting next week is about leading organizations through crises, which is what we're going through right now with a couple of our more distinguished alums, as well dean Katy Graddy. So be on the lookout for that, and we hope to bring more programming to you in the future. In the meantime, stay safe, stay healthy. And again, thank you for joining us today, and have a great day.
Anna Scherbina:
Thank you.