Corey Hajim: Hi, Chris, how are you? Chris Anderson: I’m very well, Corey, it’s absolutely lovely to see you. CH: It’s great to see you, too.
CA: Somehow, you’re always smiling, no matter how dangerous, weird, crazy things are. Thank you for that. CH: You don’t see me in the other room crying afterwards, but we’ll leave that for the other [unclear]. So Chris, this is week three of these conversations, how are you thinking about the people we should be speaking with? CA: I mean, there are so many aspects to this, right? There’s understanding the basic pandemic itself and all the science around that. There’s the psychology that we’re all going through, the mindset. And we’ve had speakers addressing both of these. And then I think increasingly, the conversation is going to be “what now?” “How do we dig ourselves out of this? What’s the way forward?” And there’s a couple of speakers this week focused on that. And I think it’s — These conversations are incredibly rich, because I think one of the things that people have got growing consensus on is that step one, we kind of get, right?
You shut things down, physically distance in whatever way you can, different countries have gone about it slightly differently, but basically that “flattens the curve,” ultimately, the number of cases, the number of infections slows down. And, but then what? Because you can’t go back to life as normal, when you’re living at home completely. You could do some things, but you can’t. And so that’s what we’re going to talk about today. CH: Right, it feels really hopeful to talk about some actions we can take besides just staying away from everybody else. So, well, I guess I’ll pass it over to you to introduce the speaker, and I will come back a little bit later to share some questions from our audience. CA: Thanks so much, Corey, see you again in a bit. CH: Thank you. CA: And yes, if you know anyone out there who has just got stuck on, “But how do people get back to work?” “Where do we go from here?” Those are the people who you should, maybe invite them into this conversation right now, because I think they’re going to be really interested.
Our speaker, our guest is a professor at Harvard, Danielle Allen. She runs, among other things, she runs an institute for ethics there, the Safra institute. And fundamentally, she’s thinking about the ethical questions about what’s happening here, but she has pulled together an extraordinary multidisciplinary team of economists, business leaders and others who have put together a plan, and I’ve been obsessed with this whole thing and how we find our way out. This plan is as compelling a plan as I’ve seen anywhere. So let’s dig into it without further ado, Danielle Allen, welcome here to TED Connects. Danielle Allen: Thank you, Chris, happy to be here. I’m really, really grateful to have the chance to have this conversation with you. CA: It’s — It’s so good, I just enjoyed our conversation over the last couple of days. This is such a complex problem. What I kind of want you to do is just go through it step-by-step, to see the logic of what it is that your team are putting forward.
First of all — Just the problem itself of how we get the economy going again, just talk a bit about what’s at stake there, because sometimes this is framed as “The economy? Who cares about the economy? People’s lives are at stake. So let’s just focus on that, don’t worry about the economy.” But it’s not as simple as that. I mean, as an ethicist, what’s at stake if we don’t restart the economy somehow? DA: Well we have to recognize that we’ve actually faced two existential threats simultaneously. The first was to the public health system. If the virus had been allowed to unfold unimpeded, our public health systems would have collapsed and that would have produced a whole legitimacy crisis for our public institutions. So of course we shut down, we had to do that, it was a necessary self-defense action that has, however, really devastated the economy. And that is also an existential threat, we can’t actually endure a closed economy over a duration of 12 to 18 months.
Nor can we really endure a situation where we don’t know whether we might have another two to three months of extensive social distancing. So we really need an integrated strategy, one that recognizes both of these existential threats and finds a way to control the disease at the same that we can keep the economy open. We call that combination of controlling the disease while keeping the economy open pandemic resilience. We think that’s what we should be aiming for. CA: So people who aren’t moved by the notion of the economy, capitalism, whatever, think instead about the millions and millions of jobs that were lost, the people who are desperate to make money. And I guess the lives that will be lost unless we solve this problem. DA: Absolutely, the economy is one of the foundational pillars for a healthy society with opportunity and with justice. You can’t have a just society either, if you haven’t secured a just and functioning economy that delivers well-being for people.
So all we have to do is remember back to 2008, and think about the impacts on things like suicide and depression and so forth, that flowed from that recession, so the economy is a public health concern in the same way that the virus is a public health concern. CA: OK, so talk about why this is such an intractable problem. People isolate, in many countries in the world now you’re starting to see the cases flatten and in many cases decrease. It looks like, whether it’s happened now in your country or not, that will happen sooner or later. So why isn’t that problem solved, we’ve beaten the virus, let’s get back to work? DA: That’s a great question and it really speaks to how new the experience for us is to encounter a novel virus. It just really hasn’t happened to our society in a very, very long time. So we are what’s called the susceptible population, meaning not any of us at the beginning of this had immunity. We were all susceptible to catching the disease.
For a society to be safe, it needs to have what’s called herd immunity. You could achieve that through vaccination or through people getting the disease. But it takes 50 to 67 percent of the population to get the disease in order to achieve that level of protection. We don’t expect a vaccine anytime in the next 12 months, possibly 18 months, so we have to recognize that that pathway is not open to us. And to get a sense of the magnitude of what it would mean to live through the disease to get to herd immunity, think about this: In Italy right now they estimate that about 15 percent of the population has probably been exposed to the disease. So you’d have to repeat what Italy has done three or more times, to get to a place where you can reasonably think that there’s herd immunity. And I think you can see that when you think of that picture, how destabilizing a process would be of just leaving things broadly open without disease controls.
So the real trick is whether or not there’s a substitute for social distancing as a method for controlling the disease. CA: Right. So Italy, even with that 15 percent has suffered at least 15,000 deaths, some people argue that it’s underreported by 50 percent there, it might be 30,000 deaths plus there, and as they come down the curve, there will be more to come. Multiply that by five or six, say, for the bigger population size of the US and the herd immunity idea per se doesn’t seem like a winning idea. I mean, it’s a horrible idea. DA: It’s a horrible idea, exactly. And we do have alternatives, that’s the important thing, we actually do have a way of controlling the disease, minimizing loss of life and reopening the economy, so that’s the thing we should all be focusing on. CA: And again, the initial problem is that if you just let people start coming back, as soon as they gather again in reasonable numbers, the risk is that this highly infectious bug just takes off again. DA: Exactly. CA: And so one scenario is that you have countries lurching from a little bit of activity here and then suddenly it explodes again and everyone has to retreat.
That does not seem attractive, that also just doesn’t work. DA: No, exactly. I mean, we described that as a freeze in place strategy for dealing with this. That is you freeze and you shut down all activity, and then that flattens the curve, you open up again, then you have another peak, you have to freeze again and so forth. So you have this repeated process of freezing, which just does tremendous damage to the economy over time. I mean the upfront damage is huge, but then there’s never space to recover from it, because of great deal of uncertainty and repeated applications of economically ruinous social distancing. So I think you’re really pointing to the features of the disease that make this situation a problem that it is. And there are really two that people should focus on. One is the degree of infectiousness. This is a highly infectious virus. So the comparison to the Spanish flu is a reasonable one from the point of view of degree of infectiousness.
Then the second really important point about the disease is that it’s possible to be an asymptomatic carrier. That is to be infectious, to carry the virus, and never show any symptoms yourself. Current estimates are still imprecise, but people think that about 20 percent of virus carriers are asymptomatic. And that is really the thing that makes it so hard to control. People don’t know they’re sick and then they become disease vectors, spreading it everywhere they go. CA: Yes, indeed. So talk a bit, Danielle, about your thinking about how we might outwit this thing. DA: So the alternative to social distancing as a strategy for controlling the disease is really massively ramped up, massively scaled up testing, combined with individual quarantine. So we are going to continue to need individual quarantine for those who are positive carriers of the virus, until such a point as we have gotten a vaccine. Now what does that mean exactly?
It means that the standard quarantine that aligns with the incubation period, 14 days is often what people talk about, in the conservative picture you might say twice the incubation period length, 28 days for individual quarantine. And we need that quarantine for people who are symptomatic and for asymptomatic carriers of the virus. Now the only way that you can actually run an individual quarantine as opposed to a collective quarantine regime, is if you do massive testing. We really need to make testing in a sense universally available, so that we can be testing broadly across the population. There are ways to target test, make it more efficient and so forth, but in principle, what one should imagine, is really wide-scale testing, tens of millions of tests a day, connected with quarantine for those who test positive.
Excuse me. CA: So weird. Anytime anyone coughs today, you go, “Oh, God, are you OK?”DA: Yeah, no, no, I’m fine, Frog in the throat, that’s all it is. CA:
So just to play out a thought experiment. If we had an infinite number of tests available, and after the curve has flattened and cases have gone down, everyone came back to work, but everyone was tested every day. Then what we think is that the tests will show up positive at the same time, or possibly even ahead of the time that people are infectious. But certainly, let’s say at the same time, regardless of whether they’re symptomatic. And so you could — Those people would immediately go back home and the rest of the population should be OK, we should be able to get work done, in that thought experiment, right? DA: Right, in that thought experiment, exactly, yeah. CA: But the trouble is, that that would mean doing, whatever, like, 200 million tests a day. DA: Right, exactly. CA: Which is many, many orders of magnitude more than we have and could even imagine ramping up to. So you have a proposal, and this is the ingenuity, the proposal, of how to potentially administer tests in a way that’s much more efficient.
Talk a bit about that. DA: Sure. So if you were going to use a purely random testing method to control the disease, you could probably actually get away with testing everybody every two or three days — I’m playing along with your thought experiment here — and bring the number down to 100 million tests a day. But even that is a magnitude that would take us multiple months to get to, let’s just say if we even wanted to try to do something like that. So the thing that you really need is smart testing. So rather than testing the population at random, what you do is you use testing to identify people who are positive, and then you add to that contact tracing or contact warning, we think about it in both ways. And what this means is that once you know who’s a positive test, you figure out who else has been exposed to that person over the previous two weeks, and they all get tested as well. So you start to identify a class of people who are a higher probability of being infectious and you test that group of people.
So you move away from random testing, you target it through contact tracing or contact warning. And then, depending on the level of effectiveness of your contact-tracing and contact-warning strategy, you can reduce the numbers. So on a moderately effective contact-tracing regimen, you could imagine doing 20 million tests a day. On a highly effective regime of contact tracing and warning, you could get yourself down to the order of five to 10 million tests a day. CA: And some countries in Asia seem to have pulled off a version of this strategy that has been effective. But it requires one of two things, if I understand you right, Danielle, it requires either just this massively scaled up, or potentially quite intrusive sort of manual contact tracing where you have big teams who swoop in to anyone who’s tested positive and try to unpack their complete recent social history. Or technology plays a role, and this is where it gets complicated, because you know, there are apps in some of the Asian countries, like, China has an app which most people are, I think, required to carry, certainly in Wuhan and elsewhere, where it’s very good at predicting whether someone may need quarantine.
And they will be required to do so. And so there are all these concerns about government control, government intrusion. You are in discussion about ways of doing some kind of technology that would be more acceptable in a democracy, and I’d love you to share what those are. DA: Sure, I’m happy to do that. So I think it’s an important thing to say upfront that the rates at which we would need to test per capita are higher, much higher than Asian countries used, because prevalence is much higher here. They caught it earlier, they had these tools built before the pandemic hit. As a consequence, they’re able to control it with a lower per capita rate of testing than will be the case for us. We just have to accept that fact at this point and recognize that massively scaling up is specific to our situation, because we weren’t ready before it hit. So then, yes, OK, if we’re trying to do the smart testing, trying to use tools, what can you do?
So we’re actually open to manual testing in the plan that we’ve developed, I want to just say that, and I think that society, we have a big choice to make, whether what we want is a big core of manual contact tracers who are tracing people’s histories and figuring out who they’ve been in contact with and who they’ve been exposed to. Or if we want to try to use a technological system. The important thing is there is a diversity of options within the technology space. So it’s really important to recognize that places like Singapore and China have used highly centralized data systems for supporting this. And so what happens is, sort of, you carry your phone around, and everybody is connected to a central data system, and then when somebody in the system has a positive test, that gets put into the app, and then their phone communicates to other phones that it’s been in proximity with over the previous two weeks, to alert people that they too need to get a test, OK? That’s the basic concept.
In China and Singapore the data structure for doing this is highly centralized. There are, however, a lot of innovative apps under development right now that depend instead on a very privacy-protective structure where the data lives on the individual user’s phone and through a combination of encryption and tokens users of phones can communicate with other users of phones, but the data is not centralized. So in that regard, it becomes more of a peer-to-peer sharing, sort of concept of friends warn friends that they should probably go get tested. Then you would have a central repository of test data, but the truth is, we already have that, because all influenza tests for example, already roll up into CDC and Health and Human Services databases, so that they can track influenza patterns every year. CA: So tell me if I understand this right. You would carry on your phone an app that would, when you got, say, within six feet of another human carrying that app, the phones would exchange a Bluetooth — using Bluetooth technology they exchange a kind of token that says, “Hi, we connected.” But it’s encrypted.
And that is not communicated to a central server, that is on the phone. But if either of you in the next week or two tests positive, your phone will be able to communicate to all the people which it exchanged token with, to say, “Uh oh, someone who you were close to in the last two weeks has tested positive. You’ve got to isolate.” That’s basically how it works, it’s done that way. DA: Exactly. CA: And then after, what, three or four weeks, the tokens can actually autodelete? They go, they’re not there anymore. DA: They expire, that’s right. Because you only need the most recent two weeks’ information or data about where you’ve been and what other phones your phone has interacted with. So that’s the really key thing. CA: Alright, we’ll come back to that in a minute, but let’s see what our friends are asking online. DA: OK.CH: Hi, Danielle, hi, Chris. Yeah, we’ve got a lot of great questions, people are super interested in how this is all going to work.
There’s a couple of questions I’m trying to cobble together here. I think people are really interested in your thoughts on the United States health care system. We have so many underinsured and uninsured people and the changes that you might make to that system, I mean, does that situation make things worse, and what changes would you make to the system so that we’re not as vulnerable in the future? DA: So that’s a great set of questions, and so just from the point of view of the testing program, it is absolutely critical that the testing be free. And so there is absolutely, a sort of necessary feature of this, which is about, kind of, universal access element to the health system. And so I’m sure there will be tweaking that’s necessary in the existing health system to achieve that. We’ve also without any question seen vulnerabilities that relate to and stem from our fragmented health system.
So I think there’s a much bigger, longer-term question to be had, or conversation to have, about how we overcome that fragmentation. So yes, I do hope this moment will be a spur for that longer-term conversation about improving our health system and really achieving that universal coverage that we so badly need. CH: OK, thank you, I’ll see you both again in a little bit. CA: Thanks, Corey. So let’s stay with this tech issue for a bit. And the sort of civil rights or privacy questions that it might still raise in some people. So one concern is that surely, if your phone is able to contact these other phones, someone somewhere is ultimately going to reverse that and we’ll have some kind of record of your, you know, everyone who you’ve connected with, and that might be concerning to some. Is that a legitimate concern? DA: I think it is, I mean, I think we’ve been working hard on this question and really trying to think it through and when you talk to legal experts and civil liberties experts and so forth, everybody starts with the same premise: assume failure.
Assume that you’ll have a data breach. Think for that and what kind of protection you want in that regard. And so when you think that way, you of course are trying to minimize any likelihood of that happening, so hence the privacy-protective structure of phones communicating with phones, data living on the hardware of the phone, not in the server, etc. And then also you would want a kind of democratic accountability feature, so for example having the Department of Health and Human Services have an auditing function to audit whoever is manning the server or controls the server through which the tokens are exchanged you would want to audit their functionality and how they’re using the data. But then again, you presume failure, that somebody’s reverse engineering, the audit system fails in some fashion. What’s your protection then? The answer to that would appear to be upfront legislation that prohibits the commercialization of this COVID testing data.
So that anybody who in any way tried to commercialize it in any kind of way, would be subject to legal penalty. So I think that’s how you build the fence up upfront in the expectation that somebody would find the way to crack it. CA: And then there’s a set of questions around how you get this app out there at scale, because it’s only effective if, say, two thirds of the people who are working are carrying it, right, something like that. DA: Right. CA: And so short of authoritarian “everyone must have this app,” I guess there are ways that are interesting to say to people, one, this is a really useful app, it will alert you quickly if you’re at any risk. But two, to get to the kind of scale we’re talking about, you might have to say to people, “Look, we’re slowly going to come back to work, industry by industry, company by company, and the deal for you to come back and break isolation, the societal deal, is that you have to be willing to carry this app.” And you could, for people who didn’t want to do that, I guess you could have some protection, you can’t lose your job for that.
But, I mean, can you picture society making the choice that it is reasonable to require people who want to come back to work to carry that alert technology with them? DA: So this is the hardest question. We know we don’t want an authoritarian model, such as the one used in China and Singapore, so we have to figure out instead how to activate that thing, which is sort of the most important democratic resource or asset, namely solidarity. So what is it that, from a solidarity perspective, it’s reasonable for us to ask of each other? That has to be the frame for deciding how we approach this. And so one aspect of this is really, truly the building a culture of opting in to this. And there are examples of this. So for example, New York has tackled HIV testing through a program that goes by the label “New York Knows,” and it started out with labels of “Manhattan Knows” or “Brooklyn Knows,” and so forth, of the different burrows.
And what this program is is one that is owned by community organizations, community partners, that do the job of spreading the word and recruiting people into testing programs. And New York has the goal of having every single New Yorker be tested for HIV, so in other words, it’s established as an expectation, that universal participation, and it’s activated a network of community partners and organizations, to cultivate that commitment to solidarity. And so I think, in all honesty, that that would be a really huge part of what you would need to do in order to tap into solidarity, to have this work. I’m sure that we would see some amount of requiring in different context, I think that’s a very hard one, because you don’t want to generate labor discrimination problems. And so the model there, to think about and to sort of figure out what are our parameters, what we think is fair, connects to things that schools currently do, for example, when they require that students show vaccination proof before they can start the school year and things like that.
So there are multiple states that do that in schools for vaccines. Would schools do the same thing, what’s the sort of labor, the workforce question like, I think that very much remains to be worked through, but it’s a hugely important question. CA: I’d be curious what the watching audience thinks about this, maybe you could enter a comment on it. But I mean, is it reasonable, in the world that we’re in right now, for a company, let’s say, to say, “Look, we want to get back to work, but we want to do so and respect the safety of all our workers. That means that for you to come back to work, you need a test showing that you’re negative. And you need to carry this app so that we alert people quickly if there’s a problem.” Is that — “We won’t fire you if you don’t come in, but if you want to come back to work, that’s what you’ll have to agree to.” Is that a reasonable chance? I’m curious what people think.
Is there any other way to get — Sorry. DA: I mean, again, there is precedent for this in the sense that drug testing works this way in many employment contexts, right. There are many roles where people have to do routine drug tests as a part of preserving their job. That was a hotly debated issue in the 1980s, people sort of think back when that sort of first came in, and there was a lot of concern about it. We have managed to develop a regime for that, that has achieved an equilibrium of a kind. So I imagine that something is possible in this space, but we would have to draw on the prior experience with things like drug testing in the workplace, I think. CA: I mean, one problem that we face when you think about these big systems introduced is that in the past, there’s history where something got introduced, you think of the PATRIOT Act that came in after 9/11 and a lot of people have a lot of problems with that Act, and it gets renewed relentlessly, relentlessly, and here we are, nearly 20 years later, and it’s still with us.
So that creates quite a high bar for any standard that we push out here. How do we persuade people that this is custom-made for the current situation that we’re in, and it’s not going to be picked up and subsequently abused by companies or by government? DA: That’s an absolutely critical question, and I think we have a lot to learn from places like Germany, which are really, really strong and rigorous on privacy protections. Perhaps having some of the highest privacy-protection standards in the world. And Germany, over the course of the last few weeks, has articulated an approach that definitely picks up several of these elements. So there are ways of building in privacy structures that are meeting the standards of the German privacy framework, and so I think for us, that’s a really important place to look to, and learn from them how they’re structuring it, to achieve those privacy protections. CA: Danielle, you’re an ethicist, among other things, as well as a political theorist, and is it, as you think about how to apply ethical questions to this, is it inherent in a situation like this that there are going to be trade-offs, that there is no “perfect solution” that we just, you know — These things are fundamentally — You’ve got two goods that are fundamentally in conflict with each other or if you like, avoidance of two evils that are going to clash.
And that we’re not going to get away sort of untainted to some extent, we just have to try and make the least bad choice? DA: It’s a great question, and I think, I tend to formulate things as being about hard choices and judgments, rather than being about trade-offs. I think trade-offs often suggest that you can precisely quantify this degree of monetizable harm against that degree of monetizable harm, and I think that’s actually not as helpful to us in this current moment, to be honest. So in effect, I think the most important thing is that we clarify our core values. And so the way we’ve tried to articulate that is to say we have a fundamental value in securing public health. We have a fundamental value in securing a functioning, healthy economy. We have a fundamental value in securing civil liberties and justice and constitutional democracy. And so then the question is, given that set of fundamental values, what are the policy options that actually do secure all of those things?
So in that regard, at the end of the day, you know, there’s a bunch of libertarians in the group that we work on, and a lot of us come out very strongly, sort of, privacy protecting, liberty protecting point of view. And so we’re not here to sacrifice those things. We’re rather here to find a solution that aligns with the values that we bring in to this problem. So that’s how we think about the decision making. CA: Talk a bit more, actually, about the group that you’ve pooled together over this. I know that there’s a TED speaker Paul Romer, an economist at Stanford, who’s, I think, a key member. Who else is in the group? DA: Well, Paul was a key member. I’m afraid we parted ways to some extent, because he’s advocating random testing, so the sort of 100 million tests a day direction, and he’s not a fan of the contact-tracing approach, so he does have, you know, he’s sort of at one end of a kind of libertarian spectrum on that and my view, however, is that testing 100 million a day is far more intrusive than smart testing supported by privacy protective contact tracing.
I also think it’s really important to throw into the mix the fact that collective social distancing is a huge infringement on our civil liberties. We keep forgetting that. The alternative is not contact tracing versus nothing, it’s contact tracing versus social distancing. We can’t go out, we can’t form associations where we get to be together in person, churchgoers can’t go to church right now. You know, political parties are having their conventions postponed. If that’s not infringement on our civil liberties, I don’t know what is. So from my point of view, the civil liberties conversation is one about the contrast between the kind of infringement that is produced by social distancing versus the kind of infringement or reshaping that would be imposed by contact-tracing regime. I didn’t answer your question about our group. CA: Go ahead, it’s just amazing this thing is moving so fast in real time.
Talk about some of the other people who are in your group. DA: Sure, so Glen Weyl is an economist at Microsoft, a political economist, he’s a really key figure and he is really an innovative mechanism design thinker, who is really good at kind of, figuring out how to craft incentive structures and so forth that help people make choices in socially productive ways, in ways that are also freedom-respecting, and so forth. So he’s really been helping us think about the design of the policy pathway, Rajiv Sethi is another economist, Lucas Stanczyk is a philosopher at Harvard who has been scrutinizing the civil liberties and justice questions. I mean, that is his line of work, those are the things he’s most committed to, and that’s what he’s doing. We’ve reached out to a number of public health groups for regular consultations, so they’re not as directly part of our group in the sense of advancing a policy, but in terms of informing our epidemiological understanding, we’ve relied a lot on folks at the Chan School of Public Health at Harvard.
So lawyers as well, Glenn Cohen, who directs the Petrie-Flom Center for law and bioethics has been a critical member, Andrew Crespo also at Harvard Law School, Rosa Brooks at Georgetown Law school, I could go on, I’m missing key people, critical scientists. Actually, there’s a great paper on solidarity by Melani Cammett and Evan Lieberman that people should check out too. CA: It’s exciting that one of the impacts of this, and I’ve seen it in other areas as well, this crisis is really breaking a lot of cross-disciplinary lines and bringing people together in unexpected combinations, which is good. DA: Yes. CA: So how, if this plan got general acceptance, how — I mean, obviously, the clock is ticking, this is urgent, what would it look like to move this forward? Give a sense of what you think it would cost, give a sense of who might own it, like, what would it take to actually activate this giant idea? DA: Alright, so it’s a big price tag, so I hope you’re sitting down — I’m glad you