Anup Malani and Milan discuss India’s second COVID wave—what we know, what we don’t know, and what we need to find out.
It has been a harrowing week for India. The country is reeling under the effects of a devastating second wave of the coronavirus, which is responsible for more than 300,000 new cases a day and more than 2,000 fatalities. And these official numbers are almost certainly a dramatic undercount.
To understand what is driving this new second wave of the virus and the global health implications of the surge, professor Anup Malani joins Milan on the show this week. Anup is the Lee and Brena Freeman professor at the University of Chicago Law School and a professor at the Pritzker School of Medicine.
Anup and Milan discuss India’s second COVID wave—what we know, what we don’t know, and what we need to find out. Plus, they discuss the findings of numerous serological studies Anup and his co-authors have conducted across India, the contested role of lockdowns, and the worrying prospect of vaccine nationalism.
Welcome to Grand Tamasha, a co-production of the Carnegie Endowment for International Peace and the Hindustan Times. I'm your host, Milan Vaishnav.
It has been an absolutely harrowing week in India. The country is reeling under the effects of a devastating second wave of the coronavirus, which is responsible for more than 300,000 new cases a day and more than 2000 fatalities. These official numbers, of course, are almost certainly a dramatic undercount of the true numbers. These worrying developments have come as a rude awakening to Indian citizens as the country's political leadership has consistently proclaimed that it successfully defeated the virus, that the economy was returning to full strength. To understand what's driving this new second wave of the virus and the global health implications of the surge, I'm joined on the show today by Professor Anup Malani. Anup is the Lee and Brena Freeman Professor at the University of Chicago Law School. He is also a professor at the Pritzker School of Medicine. And just yesterday, Anup was elected to the American Academy of Arts and Sciences, one of the nation's oldest and most prestigious honorary societies. I'm very pleased to welcome him to the show for the very first time. Anup, thanks for taking the time.
Thank you very much for having me.
So, Anup, before we get into the current crisis in India, I want to rewind the clock a little bit. If you go back to just a couple of months ago, politicians and policymakers across the country were publicly proclaiming that India had defeated COVID. The country was well on its way towards full recovery, and both domestically and internationally, there was a kind of widely held perception that India had dodged a major bullet. With hindsight, do we understand why it is that India was largely spared by the first COVID wave?
I think we have some ideas, but I think we want to question the premise first before we try to answer the question you raise.
First, we're not 100 percent sure what happened in the first wave. We do know that there was a massive undercounting of actual cases. There's talk of massive undercounting of deaths right now; if anything, that might have been worse the first time around. So, the first thing we need to admit is, it could have been that the first wave was meaningfully larger than we thought. So, that's the first thing.
But let's suppose that we did dodge it in the first wave and we're [only getting hit now in] the second wave. A second question I think we want to ask is, is the second wave actually worse than the first wave? So, when we tried to say, how did we manage the first wave, we need to figure out if, in fact, the second wave is maybe as good as the last one. I know it's hard to imagine that in the middle of a crisis, but it's important to remember that this wave – while the peak is higher and rose much more quickly than the first wave, the last wave was just a very long running wave that lasted about two months. So, we have yet to see what's going to happen here.
But then if we go back to the first wave and ask, why was it okay? I think what we're asking is not so much, why did a lot of people not get infected? I think a very high number of people got infected. I think if you look at seroprevalence surveys, we probably think that in the first wave, the official reports massively undercounted the number of cases. I think the thing that's remarkable about the first wave, if we thought we escaped the problem, is that it had a very low number of deaths.
Now, assume that those death counts are correct. There are a standard set of explanations that people have been exploring, but the jury is out. The first is some sort of cross-immunity because of existing coronaviruses that are endemic to India and maybe other low-income countries – India's experience wasn't necessarily isolated. Another possibility is the BCG vaccine. I think we're investigating both of those. There's an effort to study whether or not the BCG vaccine generates the type of antibodies that would be effective against at least the old variants of COVID. That's a study that is being undertaken. That's for the cross immunity. One of the challenges is figuring out whether, in fact, people that had antibodies to COVID-19 were individuals who would also generate antibodies to the old coronaviruses. That's just a hard question. It can be done, but it's [still] to be done.
There's also the discussion about genetic advantages. So, I think a lot of people may remember the Science article – I think a Science or Nature article – suggesting that there's a gene variant inherited from Neanderthals that that was quite common in India and in parts of Europe that might be protective against the most severe harms from COVID. I think that deserves some exploration.
But I would say my preferred explanation, the one that I'm spending the most time thinking about, is something called survivorship bias. And it has some interesting implications. So, the basic idea is, in India, individuals face a lot of disease burden, meaning there are a lot of diseases that could strike you down, and one implication of that is individuals with frail health, particularly frail immune systems, are lost to other diseases over time at a faster rate than, say, in developed countries like the United States or countries of Europe. What that means is the people that are left over have particularly robust and very strong immune systems. And that would lead you to have a robust response to COVID. So, if that's the case, then what's really going on is that people didn't die in the first wave of COVID in India in part because they'd already succumbed to other diseases earlier. So, yes, we had a lower IFR [infection fatality ratio] – it's because of the disease burden people generally face in India. But the flip side is that tells us that we lost those people much earlier. So, you don't want to celebrate the low IFR – you want to say, basically, we probably could have saved them so that at least they got to COVID.
One of the factors that you haven't explicitly mentioned is demographics. So, there are a lot of people saying that by virtue of the fact that the median age in India is around 27, 28 – we know that younger people, even though they might be carriers, are more likely be asymptomatic or not be hit as hard. Could that be playing a role here?
If you look at the total number of deaths, it is surely playing a role. The fact that we have an age structure that skews young relative to, say, developed countries – that certainly plays a big role. It doesn't necessarily explain why our IFR was different. And let me be clear about terms. So, IFR is the infection fatality rate. It is the probability that you die given that you are infected – and here it's infected with COVID. Now, we can calculate IFR just overall for the population – it's low for India in part because it skews young – but what we often do is we calculate the IFR given your age or conditional on age. So, we calculate what your infection fatality rate is if you're 20 to 30 versus the infection fatality rate if you're 60 to 70.
So, in other words, this is an age-adjusted fatality rate.
Exactly. And what's interesting about what we see in India is that, age-adjusted, India has a lower IFR based upon the official numbers, and meaningfully so – maybe an order of magnitude or larger. At least, that's what we found in, for example, our Karnataka study. So, the question we want to ask ourselves, maybe, is why is it that a 20-year-old has a – that's not exactly the right way to put it, but – a ten-times lower probability of dying given infection in India than in the United States? Why is that also true? And maybe even more so for the elderly – one of the interesting things that we find in our study of IFR based on a number of serological studies in India is that the relative advantage that the elderly have in India is even greater than the relative advantage that the youth have, which is a bit of a puzzle, actually. But in either case, our IFR overall and conditionally on age is lower. And that also means that your overall deaths are going to be lower.
So, you alluded to the fact that we have a number of serological studies; you yourself have been engaged in many influential seroprevalence surveys that were conducted during 2020. Could you outline for us what you set out to do when you design these surveys and what some of the key takeaways were? I know you did many studies – you did some in Karnataka, some in Mumbai, some with migrants in Bihar – but if you had to kind of identify what some of the key toplines are, what would they be?
The key one is that we're massively undercounting cases if we only look at confirmed cases. So, let's step back and think about how we count COVID cases. The first thing is that we go out and we do some testing – we meaning the government in this context. They do some testing. Now, once you do the testing, there are two ways that you can estimate how prevalent the disease is. One is just the total count, the total number of cases, which is the common approach. The problem with that approach is that if you don't do enough testing, you might not capture everybody that's infected. So, if there are more people infected than tests, you will definitely undercount.
A second approach is to calculate the rate of positivity on tests. Sometimes it's called the positivity test. That is, you calculate the following fraction: the number of positives divided by the number of tests done. Okay, that's the positivity rate. That's maybe a better estimate. But that has two separate problems. The first one is, if you are seeking out people that are infected to test, then you're going to get an overestimate of the fraction of the population that's actually infected. And the second is, one of the very interesting things about the way we do testing policy around the world – this is not specific to India, this is from the WHO down – is that when we see that the positivity rate is too low, we encourage countries to test more, thinking that the reason why the positivity rate is low is because we're not testing enough. And the problem with that is that that means that your positivity rate is a function of how much you're going to test. That causes confusion in trying to estimate what exactly the fraction of the population infected is.
So, those two solutions don't work. So, the question is, what's a better solution? And the better solution is, I want to randomly select from the population people to test, regardless of what they look like – whether they have symptoms or don't have symptoms, whatever their contact history is, I just want to test those people and estimate what fraction has the infection, and then I can get a good estimate of what the overall prevalence of COVID is in this population. That was the goal of these studies. And the result of them was, we were massively underestimating – maybe perhaps 40 to 100 times – the number of people infected [compared to] if we use the first method. And surprisingly, we were actually maybe underestimating if we use the second method. But either way, we're getting it wrong.
But doesn't that mean that the Indian anomaly is even more striking? Because many, many more people than official records suggest had contracted the virus. But the deaths were then even smaller as a fraction?
That's exactly right. Early on in the pandemic, people were focused on something called the case fatality rates. That is the following fraction: the number of confirmed deaths divided by the number of confirmed cases. Now, let's suppose that we count deaths correctly. Let's just assume that for a second. If you go from a case fatality rate to an infection fatality rate, the thing that you do is you replace the denominator – you replace the confirmed cases with the actual number of infections, maybe estimated from one of these population level surveys, and you're exactly right, it massively increases the denominator. And so the IFR falls relative to the CFR, and it's very low. So yes, the first-order effect of the results of the serological surveys was we suddenly thought that the IFR was much lower than before because we were kind of making inferences from a case fatality rate.
Those of us who are not in the scientific community have tried to become experts in the last year and a half. We've been trying to struggle. And so concepts get thrown around constantly, including this concept of herd immunity. I think a lot of people looked at the serological surveys and said, “Wait a second, if more than 50 percent of the surveyed populations in some of these areas had COVID antibodies, then India must be well on its way towards herd immunity.” Now, clearly what we're seeing now with the second wave was that was a mistaken view. So, in hindsight, what did we miss? What did we get wrong by kind of assuming that herd immunity was a threshold that we were either approaching or that we had crossed?
Okay, let's define concepts. So, herd immunity is important in epidemiology because we think that once we achieve some level of herd immunity, the spread of the epidemic slows – it doesn't disappear. It means the something called the reproductive rate falls below one. When the reproductive rate is below one, that means each person who is infected infects less than one additional person, and so slowly the epidemic is going to peter out. Now, that's why we care about it. We want to get our reproductive rate below one; we think herd immunity gets us there.
Now, the fundamental problem is that herd immunity is not an objective fact. It depends on how much people interact and thereby spread the infection. So, for example, if you imagine a society with a high level of activity, you're going to need a lot of people infected before you reach what's called herd immunity. Imagine a society with very few people interacting – say, rural India or even developed countries that are even more sparsely populated. You're going to need fewer people infected to reach that critical point.
What complicates it even more is that people change their behavior within a country over time. So, if you go from no lockdown and regular activity – say, in February 2020 – to lockdown and very little activity in April 2020, then all of a sudden, the amount of people that have to be infected to keep the infection growth rate down – to keep R-naught below one – that number is much lower. The problem is the following: when we relax from lockdown, and people interact more – and I'm not saying lockdown was the only thing that's responsible, a lot of people are self-protecting – but as we then relax our precautions and we engage in more activity, then all of a sudden, the threshold for herd immunity to control the epidemic rises. And so, all of a sudden, you're going to have more infections.
Now, what we really should be asking ourselves is something like the following, which is very hard to do because we haven't seen this for a long time, about a year and a half – we want to ask ourselves, if we resumed the level of activity that we saw in April or May 2019, how many people would have to be infected for the reproductive rate to fall below one? That would be the herd immunity threshold we want. But we haven't approached that level of activity yet. And I know we all say, “Oh my gosh, people are being too risk-taking, they're not wearing masks,” and things like that – I just don't think that we've reached the level of activity or nonchalance that we had back in 2019. So, we can't see what level of people need to be infected for us to reach herd immunity, if that makes sense.
That makes sense. But one of the things I think that is not well-understood – and it's there in your studies, but it may not get translated to the broader public – is that there's a lot of internal variation within India in terms of the prevalence of antibodies, isn't there? So, when you hear that some areas might have as high as 50 percent, there's a lot of internal heterogeneity.
Yeah. And I'll complicate a little bit more, because in fact the level of antibodies is declining, and that's not alarming, you shouldn't be alarmed by it, but it does make it hard to figure out how many people are actually infected.
So, let's break that down into parts first. There's a lot of heterogeneity. So, in July 2020, our study in Mumbai found that 55 percent of slum dwellers on average had antibodies, and only 15 percent – I'm simplifying the numbers, rounding them – but it's 55 percent versus 15 percent in non-slums. So, massive heterogeneity, and there are two possible explanations. One is, slum dwellers, even in lockdown, are in crowded housing conditions – that's my preferred explanation, I have a paper that kind of favors that explanation – using common toilets and things like that, whereas people in high-income areas, they each had an apartment with their own toilets. They didn't have to interact as much with others, even though it's a dense city. An alternate explanation is that people in slums were more mobile, and they went around. I don't think that's the right explanation. It seems like mobility patterns are roughly the same for slum dwellers and non-slum dwellers: it fell massively during the lockdown.
So, one thing is that slums had a lot more. Then, when we went to Karnataka, another thing that we found – this was around the same time, so June, July, August – we found that overall seroprevalence in Karnataka was very high, 46 percent across the state, but there was a big gradient between urban and rural. Rural areas had basically 10 percentage points lower seroprevalence, prevalence of antibodies, than urban areas – again, consistent with this basic idea that places that are more crowded are more likely to have been infected in that first wave. And, by the way, I just want to point out that these studies were done before – in hindsight, we know the actual peak of wave one occurs in September, October – these are the ones done before that.
Then, in the middle of the epidemic, we did a study in Tamil Nadu, which found a similar sort of urban-rural gradient. But then we learned a second thing, which is very interesting – and it wasn't the first time we saw it. The second thing that's really important about that is not only is there heterogeneity across places because more crowded places are more likely to get infected – and thus actually have different thresholds for herd immunity, going back to your previous question – but the second thing we learned was that antibodies are a very short-term measure of infection.
So, let me give you two examples. The first one was Tamil Nadu. In Tamil Nadu, in October, we found that about 32 percent of the population had antibodies. But Tamil Nadu is right next to Karnataka. In Karnataka, in July and August, we're finding 46 percent. How is that possible that two adjacent states have such dramatically different results? One possibility is that, in fact, individuals, once they're infected, for a short period of time have antibodies – maybe three months, maybe six months – that are detectable, and then those decline. So, what could be happening is that Tamil Nadu hit its peak a little bit earlier and was on its way down when it measured 32 percent. By the way, we got some evidence that's consistent with this back in Mumbai, because in an unreported version of the study done in August rather than July, we sampled slums and non-slums, and we found that actually maybe antibodies declined in slums during that period by a few percentage points – not statistically significant, necessarily, but enough to make us wonder: even though official infections are rising, how are antibodies staying constant in slums? So, the way to think about it is, serological studies are very important, they tell us a lot, but you can only look back maybe three to six months and find out what happened – not all the way to the beginning of the epidemic.
So, if we come to the current crisis, you have written elsewhere that there are three explanations for the second wave – and I just want to put in a plug for your Substack, Research Notes. We'll put the link on the show notes so people can subscribe to the fascinating deep dives that you're conducting with real data. I want to ask you about these explanations. The first one that you point to is that the current spike we're seeing today in India started in a district, Maharashtra Amravati, the eastern part of the state, and that kicked off the spread. Now, as you point out, this is more an explanation about where the second wave started, not exactly why it happened. But how confident are we in this kind of origin story?
I would say only slightly confident. So, if we look at the timing of the infection, it does seem that Amravati had the most notable early infections in this wave, the most early outbreak in this wave. So, if you look at the data, you will see that Amravati in the second week of February had a spike. Okay. Now, the good news is that that spike went down after two or three weeks, but it is the first, it seems, to have a spike. And if you look out [further], you will see that adjacent districts seem to have a spike just a little bit later. So, that's all positive for suggesting Amravati's the cause.
But the problem is that there are other places like Pune that seem to have the spike almost as early as Amravati did, and there are places in between those two places that didn't have the spike until later, so it's somewhat inconsistent. So, I'm not 100 percent confident that's the case. Now, the real thing that we would want to use to identify whether Amravati was the cause is actually do gene sequencing of the strains that you find in Amravati versus Pune or someplace else. Now, that gene sequencing should be done – it's like doing fingerprints at the scene of a crime, and that's typically how epidemiology has been done – but we're still waiting on the full results of that. Hopefully, the government is undertaking that, but it hasn't been made public yet.
So, we don't know if Amravati was the cause. But you're exactly right: just because we thought that Amravati was the cause doesn't necessarily solve the problem because we have to figure out why and how Amravati was the cause. Was it because they had a population of susceptible people? Was it because there's new variants, and it just so happened that they originated in Amravati? We don't know the answer to those questions just yet. Those are the questions we ought to be asking.
So, coming a little bit deeper on the why, another explanation you point out has to do with behavioral response. So, if you buy this line of argument, we could be seeing this virulent second wave perhaps due to changes in people's behavior, which in turn has two parts. [First,] was there a change in behavior, and if so, why? [Then,] if we start with the [question of] was there a change – do we have any evidence to suggest that Indians changed their behavior in the run-up to the second wave?
So, first, if you just look at objective data that we have available – we don't have all the data available, but if we look at objective indicators of behavior, we don't see strong evidence that there was a massive increase just before the second wave. I do believe that there's a general relaxation over time in terms of people gradually engaging in more and more behavior, but in fact, we don't see a big spike in activity just before the second wave starts.
There's some people that have tried to find explanations. So, for example, for Amravati, some people say that in January, there were village-level elections around Maharashtra, and that might have triggered this. But we have to be very careful because the problem is this post hoc reasoning can lead you astray. Some people call this the post hoc propter hoc fallacy, which is, just because you see something and then you look for the thing that comes before it, that doesn't necessarily mean that the thing that came before caused this, for example, Amravati outbreak. So, I don't see strong objective evidence of change in behavior, just a gradual increase in behavior over time, so that's a reason to doubt it.
Now, I want to issue a caveat, though – our measures of behavior are imperfect. So, what do we use for behavior? We use things like the Google Mobility Index, right? The Google Mobility Index only picks up people that are using location services on their phone and on their smartphones. Not everybody has smartphones, not everybody has the data plans to use the location services. So, maybe what we're seeing is, amongst the people that are maybe high socio-economic status that are using smartphones, there wasn't a big change, but that doesn't mean that we see observe or other types of mobility. And, in any case, even if we did have Google mobility data on everybody, it wouldn't tell us, for example, whether we were using our masks when we went out. So, there's no direct evidence that activity increased, but we don't have complete information about all the risk-taking that humans were engaging at this time.
But it would seem to me that the most obvious explanation, in some sense, is complacency, right? I mean, wherever we live, in battling this for many, many months, we all have fatigue, we're all tired, we're all fed up. In India, government leaders at the very top are sounding a triumphant note, issuing claims about having defeated the virus, [that] things are coming back to normal. So, is the simplest explanation just that people let their guard down? And they were, in that regard, following the cues that senior leadership was sending.
So, we want to be a little bit careful about this. I think we have a sense in India of saying if the government leaders do it, we follow. That's not entirely true. In some sense, what the government leaders do is a function of what the population believes. I don't think that in a democracy, the leadership deviates that far. With a few exceptions – demonetization, let's say – they don't deviate that far from what the population wants. If the population thinks lockdowns are too burdensome, leaders will begin to think the lockdowns are too burdensome. So, I want to be a little bit careful. It's not a herd following the leaders. It's a little bit give and take between the two.
The causal arrow goes in both ways.
Yes, exactly. That's right. But I do agree with you that people were relaxing their guard. I mean, again, with all the caveats, let's just use Google mobility. Google mobility was rising basically continuously since the perigee, the very bottom, of the economic crisis, which was in March and April of last year. Since then, it's been steadily increasing. There's been no notable uptick. But during a big stretch of that, after the September/October peak, you saw reproductive rates estimated from infections hovering around one in a lot of places. When people thought that India had gotten past the worst of it, that was a long period of three to four months, and all during that time, people were becoming more and more complacent. So, you have to ask yourself, yes, we did see an outbreak in, say, March, a massive one, the second wave, but why is it that we had this long period where people were relaxing and we didn't see the outbreak? So, it's not enough to just say complacency. As an explanation, we have to figure out why there was a big deviation in September, even though the complacency had been continuous all along.
So, this brings us to the kind of third explanation – and this is the one that scientists, of course, are spending every waking hour trying to understand – which is about the specific variant or mutation of the coronavirus that India is currently battling. What do we know about this variant? And if you think about sort of putting in layperson's terms, how does this variant relate to what we came to know as COVID-19?
So, the first important thing to remember is that viruses are constantly mutating. What varies across different viruses is the rate at which they mutate. And so one of the interesting things about this epidemic is, within a year, we've seen some notable variants even before we saw that 617 India variant – and there may be other Indian variants, too, right? We saw a UK variant, a South Africa variant, a Brazil variant. That tells us that the rate of mutation is meaningful.
Now, the open question is, what is the consequence of that variation? That new variant, is it more transmissible? Is it more harmful? And can you explain what we're seeing now? For reasons we talked about before, I think you probably want some combination of behavioral change plus something else to [explain such] a big spike in cases. So, I think the variants almost surely play a role. We have done some gene sequencing that suggests that variants – especially in India, it seems – are [present in] a high fraction of people that are infected. And so that's another piece of evidence suggesting it's more than just behavior – it's got to be the variants.
But, again, it brings us back to the question of, how bad is it? So, here's the way that I think about it. I have a framework in my head where I think of two types of people and two types of infection. The two types of people are susceptible people, people who have not been infected before, and then people with immunity – either immunity because they were infected with the old variants or because they were vaccinated with something that looked like the old variants. So, the two groups of people, susceptible and immune.
Now, if I take a look at the impact of the old variants on this population, you'll see that the old variant will have a relatively high IFR – at least, the same as we saw in the first wave – for people that are susceptible. But for people that are immune, it's likely to have a very low effect, a very small IFR, even relative to the first wave. So, the first thing you want to think about is, how much of this is just the old variants? And there, if we see an IFR that's very high, and we see it concentrated amongst the susceptible, then we think it could just be any variant or the old variants that are causing the second wave.
Now, who are the people that are susceptible? But we already have a hint, remember – back in July, there was the Mumbai study, which suggested that the non-slums had much less infection than the slums. So, we want to ask ourselves, are these infections showing up more in non-slums relative to the slums? And I think that there's some evidence that suggests that that might be the case. I had a very interesting conversation with someone who I won't name talking about who in their family and their extended household were infected – because they were reporting, they were sitting in Delhi, and everybody that they knew was infected. But they are higher socioeconomic status, and I said, “Well, what about the household workers? Are they infected?” And what's very interesting is that the household workers that stayed in the house were affected, but the household workers that went back to the slums were not infected.
So, maybe what's going on is we have a bunch of susceptible people that are in the high SES [socio-economic status] that are being infected now, whereas they were largely spared or relatively spared in the first wave. And so the first wave is – I'm simplifying greatly – is a low income wave, where the poor got infected, and the second wave is a high-income wave, where finally the people that managed to protect themselves because they could from the first wave are now coming out and they're getting infected. It doesn't explain why it all happened at once, but that would be consistent with the idea that there's a bunch of susceptibles in the population.
Now, that's for the old variants. But now let's suppose what happens with the new variants. Now, with the new variants, I would expect that the susceptibles are going to have a similar IFR: they're just as harmed by the old variants as the new variants. What's really interesting is, for the people that are reinfected – either because they were naturally immune because [they were infected during] the first wave or because they got vaccinated, which is a relatively small percentage, it's really about people that were infected first wave – what's happening to those guys? And those are the ones that really worry me. So, if they're infected with the old variants, low IFR, but if they're infected with the new variants, we really need to understand whether or not their infection fatality rate is high or low. If it's just that they're infected again, then we just have waves of reinfection, but the costs are going to be lower. But if their infection fatality rate is high, we should be very worried. Because what that means is not only is there reinfection, the reinfection is with new variants, and that the new variants keep a high IFR. It doesn't take a lot of foresight to think, if we could experience a second wave in that group, we could experience a third wave, and so on.
So, that's really the issue. What I pay attention to is, who's being infected, susceptible versus not, and then second is, of the people that were previously infected, can we identify the ones that were infected with the new variants? And what is their mortality rate once they are infected? We don't have a lot of information on that. All we have is the overall infection fatality rate – actually, just the case fatality rate. But that's the big puzzle that we need to answer if we need to know what the future looks like.
And one of the most puzzling things, Anup, that I've experienced in my own social network is that certainly many of the people I'm in touch with – who would fit in that higher socioeconomic status bracket – were not infected the first time. They have become infected this time; their children are also being infected this time. They may have been infected last time and just didn't present with any symptoms, they may have been asymptomatic, but this time they're actually presenting with symptoms. So, could that be a result of the variant itself?
The short answer is I don't know. And it's important that we acknowledge that we don't know. We don't have a lot of data. I do want to point out a few things before we over-interpret this. The first is, if this is a higher income wave, it's going to get a lot more notice because, as it turns out, we just do a better job of giving press attention to people that are higher income in India than lower income. That's been the case for a long time. So, that's the first thing.
The second thing is, we've got to be careful about extrapolating from anecdotes. So, yes, I see some symptoms, but I don't know what the prevalence of those symptoms is, so I just want to be a little bit cautious. Because sometimes if we make the wrong inference, we could go too far in one direction or the other.
By the way, I just want to be very clear what I'm worried about. One of the hidden dangers of this entire epidemic is that we close down schools. Schools have been closed with few exceptions for a long time. India is a country that needs to increase its human capital – human capital meaning education levels. It needs to increase that. We've kind of just paused that for a year. If we suddenly overestimate what the impact is on kids now, you can imagine another year where we pause schools, and India can't afford that. We will see massive consequences from that ten years from now, twenty years from now, and we will regret it. So, that's why I want to be very careful about what we say is happening to children. I don't want to underestimate the impact, but I also want to overestimate it.
But let's turn back to that impact. The thing that I want to make clear to people is that immunity, whether it's acquired through previous infection or through vaccination, does not protect you against infection. It protects you against disease. Your immune system is not on the outside of your body acting as a shield that stops you from getting infected – it's on the inside of the body. It prevents the infection from causing as big a harm as if you are not immune. It reduces your probability of death, reduces the probability of serious hospitalization, and so on, but it doesn't always protect against every kind of symptom. So, if you have low-grade symptoms, that doesn't mean that you don't have immunity. Now, what may be possible – and it's consistent with some of the anecdotal data – something that we can extrapolate from it is that the original variants [were] largely asymptomatic [while] the new variants are a little bit more symptomatic. But you're still protected. Even the young are protected against the serious consequences of hospitalization, and especially serious hospitalization and death. So, it is possible, consistent with symptoms, that that's true.
Now, the short answer is still that we don't know. We need to do better surveys to figure out what the prevalence of this infection is, what the prevalence of symptoms is, and that's sometimes hard to measure because people want to sometimes hide their symptoms because they're afraid of quarantine. So, then you want to look at things like, what is the death rate? What I really want to know is, what is the mortality rate by age for the current round of infections? And then, secondarily, let's sequence these folks and try to figure out if we're seeing new variants or old variants, and the big prize is if we can relate the mortality rate to the old and new variants by age. Then we can answer your question and then we can see how serious the second wave is and what the likelihood of a future wave is and set policy appropriately.
Anup, everything you've just said also is an input toward understanding the role of lockdowns, right? Because we are seeing several states, localities – whether it's the state of Maharashtra, the National Capital Territory of Delhi – instituting pretty stringent lockdowns trying to flatten the curve. Economists, as a result, are already downgrading India's economic growth forecasts because they're seeing a drop in economic activity, they're seeing a drop in mobility. We know that the previous nationwide lockdown had pretty dramatic consequences on a whole variety of fronts: education, livelihoods, income, so on and so forth. How do you think about the role of lockdowns? Because there is a very polarized debate, not just in India but in this country as well, about whether we should resort to them or not as a as a primary tool in the toolkit to fight this pandemic.
So, two things: first is how important lockdowns are, and second is what the impact of lockdowns are and how we should think about those. So, one of the interesting things I think we learned in the United States about lockdowns – where we had a lot of data on what people were doing and when lockdowns happened – [was] we learned that a lot of people reduce their activity even without lockdown. So, counties without lockdowns had massive reductions in activity even as compared to counties with lockdowns. So, the important thing is not so much [whether or not lockdowns were] responsible for reductions in behavior, but really what's surprising is that humans protected themselves even without a lockdown.
So, in this epidemic, my guess is that you would see in India in the second wave a voluntary reduction in activity even without lockdowns. The question is, should you do lockdowns on top of that, then? There are some places where clearly lockdowns have a bigger impact than others. For example, it can control whether or not as many stores are open and how long, it can control whether there are large gatherings or not, and importantly, it can control schools. So, I think that there's going to be a lot of control, a lot of reduction in behavior without lockdowns, but there are certain things that lockdowns are going to enhance in terms of controlling activity, reducing the spread of infection.
But then we have to think about the cost of lockdown, and I think that's critical in India. So, I think we learned a lot [from] what happened last year when we saw a nationwide lockdown. The first important thing is, as you pointed out, income fell, but the really important thing, I think, is that income fell the most for the poor. It's a massive, massive reduction. Now, fortunately, while we see a massive inequality in the amount of income reduction – everybody saw a massive reduction in income, but the poor much more so, maybe 90 percent of their income for a few months – we didn't see as big an inequality in consumption, meaning somehow the poor were able to borrow money, or had small stocks of savings they could rely on – probably more borrowing – to be able to get through. I don't know if that's going to happen if we have a second lockdown. It might be that their ability to borrow and make it through is a little bit less. There are some things that we did right in the first lockdown. I do think our analysis of what happened with food prices suggests that, in fact, a lot of essentials were supplied through the first lockdown. Perhaps we can do that with the second lockdown to kind of mitigate some of its effects. But we do have to acknowledge massive inequality.
So, the first thing I think we have to think about is the cost of a massive reduction in income, particularly on the poor. The poor still haven't recovered from that first lockdown. The second thing we have to think about is what I said before: if we do lockdowns, I think that there are immediate harms and there are long-term harms. The long-term harms are when we focus on schools. This is why the question you asked about whether or not children are symptomatic is critical. What I want to know is their death rate, and does that warrant keeping schools closed for an extended period? I know there are holidays coming up, but what happens after the holidays? That consequence we won't see for a long time, but it'll be very, very large. So, that's the second thing that I would focus on in terms of lockdowns.
The third thing I'll focus on is just I think there are interesting political – and I'm using the word “interesting” as a euphemism – there are interesting political implications of lockdowns. I think in the United States, I worry that the lockdowns are going to cause a generation of young kids, when they get older, to be strongly opposed to lockdowns. So, our use of lockdowns now is going to generate a backlash that later on might make it difficult in the next pandemic to use them as a tool. We need to be judicious in that. we haven't talked about that in the context of India. Is the use of a lockdown changing the culture such that there's a lot less support for that government control in the future? That takes a tool out of our toolkit for pandemic 2.0.
So, those are the things that I would think about in the context of a lockdown. The real answer to all of this is, we should vaccinate, because vaccination gives you a lot of the protection of a lockdown without the cost of a lockdown. And so, if nothing else, what we should be walking away with is we should vaccinate very, very quickly, and we shouldn't let our guard down. Even if we're in a period where it seems like the second wave is over and we're all okay, there could be a third wave. We need to keep up vaccination.
So, maybe this is a good place for us to end the conversation: on the subject of vaccines. You read what some global health experts are saying, right, and they say, you know, vaccine production is not really the issue. The issue is more about purchasing, it's about distribution, it's about allocation – or reallocation. That's number one.
However, at the same time, we're in the middle of a brewing firestorm between the United States and India – not yet officially, but certainly on social media – where a lot of Indians are pointing to the U.S. administration, saying, “You guys have invoked something called the Defense Production Act, which ensures that U.S. contracts for vaccine materials are prioritized before other contracts can be fulfilled.” So, it's not technically an export ban, but it has a lot of the outward appearances of an export ban. And many people in India are saying, “Look, we're dying, we need to ramp up vaccine production distribution, and you have become an impediment.” So, in this broader context, what should the U.S. posture be vis-à-vis global vaccination supply chains? And do you agree with the premise that we've kind of got the production part figured out and it's really about the distribution and allocation?
Yeah, so that's a great question. So, the first thing I'll say is, I do believe that there's a real chance that we've applied the Defense Production Act too aggressively – we meaning the United States – and applied it too aggressively by covering inputs that are not essential for us completing our vaccination campaign but that impede, for example, India producing and other places producing vaccines that are necessary for the global campaign. I think that's bad from a public health perspective, I think it's bad from a foreign policy perspective. I think in the next twenty to forty years, the United States and India relationship is going to get stronger, and this is not the best way to start out on that path.
But I want to be a little bit cautious here. Vaccine nationalism, of which this is a strain, also afflicts India. Remember when India was having difficulty with the new cases and wanted to increase its vaccination? One of the things that the population complained about was that they were exporting vaccines. And so the exports were curtailed dramatically to serve the domestic population. That's exactly what the U.S. is trying to do. The only different thing between the two is that the U.S. doesn't need to do it as much as India maybe needs to do it. From a vaccine nationalist perspective, the United States is well on its way to vaccinating the entire population. It really faces a problem of vaccine hesitancy. Whereas in India, I think its first order [problem is] a problem of vaccine supply. So, that's the first thing that I would say: they have similar strains of vaccine nationalism, but India maybe has a slightly better argument in this context.
Now, in India, it's very interesting that we should be viewing the pandemic as a war, right? If we were invaded by a foreign country, we would devote all our efforts toward building our weapons capacity. We would repurpose industry. We would not let up just because there was a lull between battles, we wouldn't stop producing guns – we would just continue this process in preparation. Yet, for some reason, when we are attacked by something other than humans, but just as deadly as any war that India has faced for a long time, maybe ever, we've kind of let our guard down, as you said, in between the first and second wave. Yes, we did do vaccination, but we should be vaccinating like crazy. (That's a technical term I just came up with.) But we should be vaccinating at a very fast rate. And that means not worrying about pricing or things like that – it should just do everything possible to build up production capacity.
In fact, we should do that worldwide. India is not the only place where we ought to do that. And so that means investing in global supply chains. The U.S. should ramp up production of inputs into vaccine production, whether it's in the U.S. or abroad. It should allow the export of that once its own needs are satisfied. We shouldn't expect it to suspend all vaccine nationalism if we're not, either.
But then India should itself be doing a lot of things to help out. First, approve a lot more vaccines. It took a long time before [inaudible]. There's foreign approval, domestic approval. [But] it should be as fast as possible. Second, we need to think about – to the extent that we have production capacity limitations – more flexible dosing. Same issue that's coming up in Europe, which is fractional dosing, delaying second doses, things like that. Anything that we can get would be tremendously helpful. Within India, we should be focusing on production capacity. The debates over what price the Serum Institute and others should get along the way – let's fight those battles later. The first order issue is, let's get the vaccine production up, imports up, and vaccinate as many people as possible.
That's probably our best protection against not just a second wave, if we had done it before, but also potentially a third wave. If we were surprised by the second wave, we should certainly not be caught twice and be surprised by a future third wave. I'm not predicting that it's going to occur, but it would be imprudent not to plan for it. And so that means investing in production.
Then the last thing is really logistics. There's this sense – and I don't want to criticize too much – but there's a sense in India that the government has to do stuff. The U.S. had this sense too, initially. If you looked at its first rollout, it was very slow because the government was doing stuff, and it was only when it finally resorted to using private channels – the Walmarts, the Mariano's, the CVSes, the Walgreens – that it massively ramped up to, like, three million doses a day. India should think about the same thing. If it wants to speed up, it needs to start using the private sector for distribution, and quite aggressively. Yes, it should prioritize, but even within priorities, it needs to use those private distribution channels that are well developed. It's not just the government that can do this. One thing that I would note is that everybody thinks about the government providing health care in India, but as it turns out, the government only provides about a quarter of India's health care. Three-quarters is done through the private sector, and we should be doing a lot more of that with vaccination.
My guest on the show this week is Anup Malani. Anup is the Lee and Brena Freeman Professor at the University of Chicago Law School and a professor at the Pritzker School of Medicine. Anup is a rare beast because he understands public health, he understands economics, and he understands the law. I think you've seen all of those traits on display. It's a Saturday morning in Chicago; you have many, many other things to do. Thank you so much for taking the time. I think this was a massively enlightening conversation both for people in India who are listening as well as people abroad who are trying to get a handle on the humanitarian crisis that's going on. So, thank you so much.
Thanks, Milan. Thank you for having me.