#
Too many of us think of the world in terms of binaries.
#
That makes it easy to make sense of the world.
#
The most common binary is that of good and evil.
#
That's why mainstream commercial cinema has heroes who are good and villains who are bad.
#
Keep it simple for the audiences.
#
Make it easy to root for one side over the other.
#
That's why there are so few nuanced gods in religion.
#
And even in our politics, in an age where so much has become so tribal, we think in
#
binaries, it's an easy way to make sense of a complex world.
#
In these pandemic times, one binary that is all around us is that of the lockdown.
#
So much of the discourse behaves as if there are two possible ways to act.
#
One have a lockdown, two have no lockdown.
#
But the truth is that this is not a binary choice but a continuum along which many other
#
The best option, in fact, might be one in which we don't have a complete lockdown but
#
nor do we let the pandemic get out of control.
#
What are these gradations?
#
How can they be calibrated?
#
What kind of metrics do we need to determine how much of a lockdown is needed?
#
And what kind of data do we need to make good decisions?
#
Most importantly, given the nature of our politics and given how incompetent and ill-equipped
#
our governments are, can good ideas even be implemented in practice?
#
Welcome to The Scene and the Unseen, our weekly podcast on economics, politics and behavioral
#
Please welcome your host, Amit Verma.
#
Welcome to The Scene and the Unseen.
#
My last episode on COVID-19 was recorded in the middle of April with Shruti Rajgopalan.
#
When I think of the time that has passed since then, it feels like it went by in the blink
#
When I think of all that has happened in that time, it feels like an eternity.
#
There has been so much pain and suffering across the world and especially across India
#
that I sometimes feel guilty to even be able to live a relatively normal life.
#
There is so much to process about all that happened.
#
Shruti and I spoke about how policymakers were faced with thankless choices at that
#
time, no matter what they did, countless lives would be lost and that cost would be viscerally
#
It would be seen while any counterfactual would be unseen.
#
And there is no way to calculate all the costs of any action, even in hindsight.
#
For example, we know that the lockdown has been a disaster.
#
There have been massive humanitarian costs, especially among the poor.
#
Not having a lockdown could also have been a disaster, as our healthcare systems could
#
have been overwhelmed and again the poor would have suffered the most.
#
And it's easy to say that hey, do a lockdown but implement it well so suffering is minimal.
#
But while that sounds good on paper, given how poor our state capacity is, that might
#
have been utopian to aim for.
#
In fact, as I argued in a column a few weeks ago, what India may have learned from this
#
is that our flailing state is a bigger disaster than COVID-19.
#
That earlier episode and the column will be linked from the show notes.
#
Today's episode, though, doesn't focus only on what has happened in the last few weeks,
#
but also on what we can do moving forward.
#
My guest is Anoop Malani, a polymath public intellectual based in Chicago who has a degree
#
in law, a PhD in economics and has worked closely in health economics over the last
#
He has written papers on past pandemics, the insights from which inform many of his
#
thoughts on the current one.
#
And he's not a believer in pontificating from a distance.
#
For years, he has been advising governments and working closely with them to bring about
#
actual change on the ground.
#
And in the case of COVID-19 as well, Anoop has put forward a plan for coming out of lockdown
#
that aims to minimize death and suffering.
#
It's a nuanced plan based around what he calls adaptive controls.
#
It takes a position that both removing the lockdown completely and keeping it going as
#
is our impractical solutions with huge costs.
#
Instead, we need to have an approach by which we lift controls to varying degrees depending
#
on local conditions and we keep adapting to new data.
#
He elaborates upon this in this conversation and he also shares his larger thoughts on
#
what the state and society can learn from this pandemic.
#
We spend the first 40 minutes of our conversation exploring his personal evolution as a thinker,
#
which I found fascinating and we start talking about COVID-19 after the break around the
#
Before we begin this conversation though, let's take a quick commercial break.
#
If you enjoy listening to The Scene and the Unseen, you can play a part in keeping the
#
The Scene and the Unseen has been a labor of love for me.
#
I've enjoyed putting together many stimulating conversations, expanding my brain and my universe
#
and hopefully yours as well.
#
But while the work has been its own reward, I don't actually make much money off the show.
#
Although The Scene and the Unseen has great numbers, advertisers haven't really woken
#
up to the insane engagement level of podcasts.
#
I do many many hours of deep research for each episode besides all the logistics of
#
producing the show myself, scheduling guests, booking studios, paying technicians, the travel
#
So well, I'm trying a new way of keeping this thing going and that involves you.
#
My proposition for you is this, for every episode of The Scene and the Unseen that you
#
enjoy, buy me a cup of coffee or even a lavish lunch, whatever you feel is worth.
#
You can do this by heading over to sceneunseen.in slash support and contributing an amount of
#
This is not a subscription.
#
The Scene and the Unseen will continue to be free on all podcast apps and at sceneunseen.in.
#
This is just a gesture of appreciation, help keep the thing going sceneunseen.in slash
#
Anoop, welcome to The Scene and the Unseen.
#
So Anoop, you know, before we get to the subject at hand, which is the pandemic and the lockdown
#
and the difficult few months that we've been having, tell me a little bit about your background
#
because I was sort of looking at your background and your CV and all of that.
#
And it seemed to me that you're like the forest gump of academics.
#
You clerked in the Supreme Court in the US.
#
And you know, you've also written a lot of papers on medical and health care issues.
#
So where did it all start?
#
Where did the journey begin?
#
What did you want to do when you were a kid?
#
Yeah, that's a great question.
#
I do feel a little bit like the forest gump of academics and also the real world doing
#
all those different things actually puts you in interesting places at different times.
#
So I got a chance to see, for example, 9-11 up close.
#
But let's see, I think it probably started in high school, you know, typical Indian immigrant
#
Parents wanted me to be a doctor.
#
And you know, I think I probably wanted to satisfy my parents, but my parents have very
#
So, you know, again, classic situation.
#
In seventh or eighth grade, I did a science fair project where I tried to see the impact
#
So this was back in the 1980s.
#
There's still a real concern about nuclear war.
#
And so I thought it'd be interesting to x-ray mice to see what the impact of radiation would
#
And so mice were easy to work with.
#
My dad was a doctor, so was my mom, but my dad was a doctor and he said he would figure
#
out a way for me to irradiate the mice at his hospital, the hospital that he had privileges
#
And so we went and we irradiated the mice.
#
And then I did, I tracked the mice for, I don't know, I guess weeks just to see what
#
their IgG levels were over time.
#
And I think my parents were super excited because they thought that I was interested
#
And what I realized back then was that I was actually interested in the research component
#
And so that science experiment went well, more for me than the mice.
#
And then I decided that I would be a little bit interested in research.
#
In high school, I decided that I would debate, and I don't know if you've ever debated,
#
but debate is basically about making, not just learning how to make arguments, but making
#
other people's arguments and leveraging those arguments in order to win a debate.
#
And what I always thought during this process was, why am I only repeating what other people's
#
Wouldn't it be more interesting if I could make those arguments myself?
#
That is to say, do the research that would help other people make arguments.
#
And so I think those are the two things I had going off to college.
#
The other thing I had was, in high school, my former rebellion wasn't to cut my hair
#
in strange ways or to wear crazy clothes.
#
My rebellion was to decide not to become a doctor, which in a family is constitutes rebellion,
#
which is not to say I wasn't going to go to college.
#
It was just, I was going to do something other than medicine.
#
So I went off, went to college.
#
I initially majored in political science, mainly because that seemed to flow naturally
#
from the debate activities I was doing in high school.
#
And then in college, I got to take some math courses.
#
And I realized I enjoyed math a lot more than political science.
#
And so while my major was political science, it's really just math classes I took for my
#
I mean, at least that's my main memory is the math classes.
#
I remember almost nothing about my political science classes.
#
And sometime around my fourth year, I must have been in my third year, so I was at Georgetown
#
University in Washington, DC, and there's a famous bookstore called Kramer's Books in
#
And I would go there with some frequency and just kind of hang out.
#
And they also had a coffee shop.
#
So back in the day, so in the 1990s, that was still relatively novel.
#
I would go and have a coffee, read books.
#
And I happened upon a book by Gary Becker, I think it's called The Economic Approach
#
And I, for some time, because of debate, just being curious, tried to understand how humans
#
I think everybody does this, try to figure out how it functions so that they can figure
#
out their role in it and how to interact with it.
#
And so I was doing the same thing, except I had a little bit of math in my pocket.
#
And so I was always trying to figure out how everything fit together, why I was attracted
#
to the concept of democracy, or what this notion of this thing called rights is and
#
how it made sense and what role should the government play, just kind of standard stuff
#
that a political scientist would think about, but from a mathematical perspective.
#
And I could never really get it all organized in my head.
#
I had intuitions, but not a structure.
#
And all of a sudden, I happened upon this book, and it's an amazing book.
#
And for the first time, I realized that people have been thinking in a similar way.
#
Remember, I wasn't an economics major, I was a political science major.
#
I wasn't reading Chicago economics at the time, certainly not Becker.
#
And so this is the first time I had an eye-opening experience where I said, look, this person
#
is thinking the way that I want to think using the tools that I'd like to be able to use.
#
And really, I want to study with this person or do similar sorts of stuff.
#
So then I decided to apply to grad school in economics.
#
Now I had a political science background, but I'd taken a lot of math classes, had some
#
So that was one decision I made, I was going to apply to economics grad school.
#
The other thing is that I was at Georgetown.
#
And if I remember correctly, my class, something like, I was in this program called the School
#
It wasn't the right best fit for me, but it was a program that I was in where you studied,
#
your focus was international more than, say, domestic.
#
And my recollection was something like 50 or 75% of students applied to law school.
#
I'm not saying all of them went to law school, but 75% I had heard had applied to law school.
#
So I'm a young kid, so I'm going to do what the crowd does too.
#
So I decided I would also apply to law schools.
#
So I applied to law schools and economics grad schools, and the only place that I got
#
into both was the University of Chicago.
#
So that's how I got to the University of Chicago.
#
And so you know the background there.
#
At the University of Chicago, the very first class you take in the economics department
#
is with Gary Becker, well, the first microeconomics classes.
#
So this is fantastic for me.
#
This is exactly what I'd hoped for.
#
I felt so lucky that I had this opportunity.
#
Around this time, I also met a person named Jim Heckman.
#
So Jim Heckman was teaching, I think he taught the second quarter econometrics course.
#
And I didn't have an interest in statistics before grad school.
#
And then I met Jim Heckman, and Jim kind of introduces you to the idea of causal inference,
#
which is to say, you know, if you say x causes y, in order to say that, you have to say if
#
there was not x, you would not necessarily, or you have less probability of getting outcome
#
And how do you think very formally about that kind of statement?
#
And so he introduced something called a potential outcomes framework, which I thought was fascinating.
#
Obviously, he'd been, you know, had seminal work in the late 1970s that introduced this
#
concept kind of in parallel with Don Rubin.
#
And so it was fantastic.
#
And he and I hit it off.
#
And so the second year I was there, I think I was his TA for that course.
#
And so those were the two kind of big influences.
#
We can talk about some of the other people I met along the way, but those are the two
#
big influences that I had.
#
And then Chicago is a very strange place.
#
Chicago is, you know, I think everybody associates with market economics.
#
And they take that approach very seriously in all aspects.
#
So Chicago is also known as a place that provides very little support.
#
So when you start out, you have to, even if you're a third year student or fourth year
#
student, you got to make it on your own.
#
That means you figure out what papers to write, you write them, and hopefully you're successful
#
Especially in the 1990s, there was this no sense that the graduate program was there
#
to support you in your efforts on the market.
#
Sure, you can get a recommendation, but it was really incumbent upon you to come up with
#
a paper, incumbent upon you to kind of get out there, you'll, you know, the thought was
#
if you're good, you don't really need a recommendation, your paper should do all the work for you.
#
And so I knew this during this time, and I was among a group of people that were quite
#
star, you know, that turned out to be stars.
#
And while I did well, I was always worried that these people were going to, there are
#
not that many academic jobs out there in top places.
#
And so I was always worried, you know, what was going to happen.
#
So some of the names actually will be names that now many people will kind of appreciate.
#
Yvonne Warnering, MIT, Ed Vitlisil, I think he's at Yale, Mathias Dapke, who's made quite
#
an impact in macro and family economics, among others.
#
And so I started worrying.
#
I was actually very curious about the Gary Becker connection, because, you know, when
#
I read that he was one of your advisors, you know, while you were doing your econ PhD,
#
it struck me that he was, you know, almost the last intellectual of a certain kind in
#
the sense that he straddled all these disciplines, you know, Milton Friedman called him, quote,
#
the greatest social scientist who has lived and worked the stock court and Justin Wolfers
#
called him the most important social scientist in the last 50 years.
#
And it struck me that in that sense, he was almost like a 19th century intellectual, that
#
he wasn't bound by a specialization, that he was, you know, delving into all these different
#
disciplines, not just economics, but also sociology, racial discrimination, crime.
#
Just a few days ago, I was reading his work on rational addiction.
#
And it's just fascinating work.
#
And it sort of comes from that classic impulse to try and understand human nature.
#
And it seems to me that when I look upon your arc and your interest in the variety of things
#
that you've done, was he a direct influence in that sense?
#
It would be hard to understate how important Gary was in the way we thought about the world.
#
So we had a set of tools, so some basic economic models and some techniques, some framework
#
And we were taught to think about everything that way.
#
So for me, the first approach is to think about how judges behaved, which I'll explain
#
in a second because I eventually turn to law soon after where I left off, but thought about
#
how judges behaved, how people answered surveys, how clinical trials functioned.
#
Almost everything I thought of, I thought through an economic lens.
#
And the reason was, was Gary said that you should and that it was fruitful to do so.
#
And he demonstrated it in terms of, you know, you've mentioned crime, we could talk family,
#
we could talk addiction, we could talk advertising, you name the topic and he's thought about
#
And he wasn't alone, by the way, there was a foundation there.
#
You know, before then you had, you know, Milton Friedman talking about the draft, you know,
#
obviously then more classic subjects.
#
But there's a wide range.
#
So there's this notion of, I mean, not in an aggressive way, but that economics should
#
be hegemonic in the sense that it should go to other fields and try to introduce mathematical
#
And now, you know, that's kind of, I think back then, when I say back then, I mean, prior
#
to the 1990s, I think that was extremely radical.
#
And in some sections it is still radical, but not really today.
#
I mean, today the idea that you would take applied math and apply it to every single
#
topic doesn't seem surprising.
#
And this epidemic slash pandemic that we're seeing now is a great example of this.
#
If you think about epidemiology, there's the non-technical shoe leather epidemiology of
#
going to find the people that are infected and then addressing that.
#
And then there's the applied math version of it, which is SIR and more complicated models.
#
And we are perfectly comfortable with that now.
#
And I think that part of that is, you know, the idea that you can use applied math tools
#
and a structure to address things.
#
The only thing that economics really adds to that is to say, look, we shouldn't just
#
use math for biology or physics or chemistry.
#
We should use math for human behavior.
#
And there's some simple ideas that can help you organize those mathematical theories to
#
whatever interaction humans have.
#
And I think the ultimate example of that actually is people applying it also to animals and
#
animal behavior, which is another situation where it's just like, look, every sort of
#
interaction you can see in the world, regardless of what's doing the interaction, there's a
#
value to using applied math and some basic underlying principles or axioms in order to
#
And it's super helpful in terms of prediction, right?
#
It's really hard for me to understand how people behave until I have a framework to
#
And these tools that Gary had helped me think about that.
#
Let's move on to law now.
#
What was your experience there?
#
Like, you know, you clerked for Sandra Day O'Connor in the Supreme Court.
#
And what was that journey like?
#
What fascinated you about law?
#
And were you sort of looking at the legal system through the tools of economics?
#
So that was an interesting journey.
#
So I think the initial interest in law was not merely, I kind of understated my interest
#
So I initially said that I was interested mainly because every else at Georgetown was
#
But that's entirely correct.
#
If you, you know, debate in high school or college, you're going to think a lot about
#
If you're a political science major, you're going to do it again.
#
And a critical component of political science is the law.
#
And so there was an underlying interest in the law.
#
You know, when I first went to University of Chicago, I did my economics PhD coursework
#
first and I started to lose interest in law because I found the economics so new and so
#
But while I was working with Jim Heckman, and not just as his TA, but as research assistant,
#
he introduced me to a gentleman named Richard Posner.
#
I'd never heard of Richard Posner before, which is probably more a sign of my lack of
#
education than anything else.
#
Dick Posner was at that time a judge, also a lecturer at the University of Chicago Law
#
School, one of the founders of the law and economics movement, along with folks like
#
Becker and Aaron Director, and Ronald Coase, who was also there at the time.
#
And he and I hit it off.
#
And we wanted to work together, either as RAs or as co-authors.
#
But at the time I was working for Jim Heckman, and Jim is notorious, both in terms of number
#
of hours you have to work, but also in not allowing you to do anything other than his
#
But he respected Dick Posner enough to let me switch jobs.
#
So I went from an RA for Heckman to an RA for Dick Posner.
#
And anybody that's had an interaction with Posner, and that's Eric's father, we'll find
#
He is extremely prolific, likes to think about everything and is very systematic in thinking
#
I think a little bit more than Becker, he's willing to deviate from standard economic
#
models, especially as his career evolved.
#
So it was very interesting to see that.
#
And I ended up being the kind of person that was defending the classic economic models,
#
because I was just a young student in economics at the time.
#
And he was willing to push them a little bit more.
#
So we had some interactions.
#
He convinced me to kind of resume my law education.
#
So I turned to law school.
#
And one thing I discovered was that law school is very hard, but it's not nearly as hard
#
as economics graduate school.
#
This is going to tie back into the Supreme Court in just a second.
#
I found, I didn't tell a lot of people at this time, but I found law to be a relief
#
from economics in terms of how much less work it required to succeed.
#
It was hard nonetheless, but it just was much easier.
#
And that showed in it, my grades in economics grad school were fine, but my grades in law
#
school were a little bit better as a result of the change in intensity, which I found
#
to be a little bit of a relief, as I said.
#
So that performance and my rapport that I had with some of the professors that I'd
#
met indirectly, so Dick Bosner and Richard Epstein and some of the others, allowed me
#
to do well enough in law school to apply for clerkships.
#
And I was able to secure a Supreme Court clerkship.
#
You get these clerkships ahead of time.
#
So I knew by my third year that two years after graduation, I would be clerking for
#
In the interim, I was going to clerk for a person, a good friend named Stephen Williams,
#
who on the DC Circuit in Washington, one of the main appellate courts.
#
So that's how I got the opportunity to clerk with Justice O'Connor.
#
And that was an eye-opening experience in the sense that it is the first time that I
#
went from just writing papers that maybe people would read or being an RA, in which case somebody
#
else is writing the papers and I was doing stuff in the background, to actually writing
#
It's actually the second year that I was writing things of consequence.
#
I would undersell it if I said I didn't have the same experience my first year clerking
#
at the DC Circuit for Stephen Williams.
#
That year, the year that I clerked in for Stephen Williams, we had the Microsoft case.
#
So this was the big antitrust case in the year 2000 that showed up before the DC Circuit.
#
And there I had my first big practical introduction to antitrust and got to learn a little bit
#
about the tech industry, which had just had a crash or was in the midst of a crash.
#
So that was my first opportunity to do something of direct importance.
#
But then, you know, there's always a layer above the appellate court that was the Supreme
#
Court that made the ultimate decisions aside from, say, Congress and the president that
#
would make these decisions.
#
And so it was my first chance to see that in action and also to see the process, not
#
just, you know, kind of help make decisions, but see the process.
#
And so I found that very insightful.
#
And were you thinking about a career in law at any point or in fact, to sort of rephrase
#
that question, were you as a person, were you driven by a sense of what you want to
#
be like, I want to be X or Y or Z, or were you more driven by problems that interested
#
you that I want to solve this problem?
#
For example, it was, you know, a problem that drove you to do your little test on mice when
#
So were you driven more by problems or was there a sense of I want this to be my career
#
and I want this to be my path?
#
How did you approach all of that?
#
Because you left the law soon enough and you kind of got back into, you know, academics
#
I don't think I had a grand ambition to be anything in particular.
#
I've noticed two things.
#
One is I like shiny new objects.
#
So if you present me with a new problem or a new thing to think about, I'm always attracted
#
and want to think about it and probably have too much of a tendency to drop what it is
#
that I'm working on to think about that new, look at that new shiny object and think about
#
that new interesting idea that you've got.
#
The second thing I guess I had a sense of is trying to organize my way of thinking about
#
the world, which is to say not only do I want to think about your new object, but I want
#
to figure out how that relates to everything else that I've been thinking about.
#
It's not like I have some notebook with a grand theory of the world, but I feel this
#
strong sense of trying to make things consistent and rethink some of the old stuff in light
#
of what I'm learning about new stuff.
#
That sense of wanting to understand the world, I think is what probably led me to academics.
#
I don't want to undersell the idea that going to a school where people judged you by whether
#
or not you became an academic, in particular, University of Chicago was really bad because
#
the standard was not only are you going to become an academic, but are you going to win
#
That kind of pushed you towards academics too.
#
I did think about law and I thought about law because along the way, when you go to
#
law school, it's very practical, you have to spend your summers and I spent one summer
#
only doing this, which is working at a law firm and I worked at two firms and both of
#
I realized I saw for the first time that outside of academia, there are some people doing some
#
very important things and that are incredibly smart.
#
I worked for a firm called Wachtell Lipton, Rosen and Katz in New York, which is one of
#
the premier M&A firms among other things in New York and just met spectacularly smart
#
people who were intellectually curious, but at the same time were very practical.
#
They were resolving matters, both in court and M&A deals, for example.
#
Then I went to a place called Kellogg Huber Hanson Todd, I think that's what it was called
#
back then, it might just be Kellogg Huber now, in Washington DC and again, same experience,
#
just really bright people.
#
I met a gentleman named Peter Huber and he was one of the founders of the firm and he
#
actually found it very interesting because he both continued to write books and had a
#
practical consulting firm slash law firm.
#
That's when I first entertained the idea of being a lawyer.
#
One other name I should mention is Ken Feinberg as a person that I met who had his own firm
#
in DC and he played a really interesting role.
#
He helped serve as kind of a private judge to resolve matters.
#
When you had the Agent Orange Victims Fund or I think he might have done the 9-11 Fund,
#
he was the one that would allocate money from the fund to different payments.
#
That was also very interesting because he was also very intellectual about it and thinking
#
about how do we not just figure out who is harmed, but practically how you allocate a
#
limit out of resources across these people and account for proof.
#
I think that that really got me thinking that maybe law is a real career.
#
I think I probably ended up in a law school first for two reasons.
#
One was because it was a nice compromise between the two.
#
It allowed me to see and respect these individuals who were actually making kind of intellectual
#
contributions in a practical way in cases or in transactional matters and at the same
#
time keep a foothold in academics.
#
One thing that people don't appreciate about law schools, especially modern law schools
#
and maybe this is more so about kind of top-tier law schools, unfortunately, is that they allowed
#
you to think academically about a broad range of subjects.
#
You're supposed to think about law, but they didn't penalize you if you thought about non-law
#
That was the other thing that I think led me back to law and in particular legal academia,
#
at least to start my career.
#
I think that answers your question.
#
Yeah, and it's interesting that you should describe the very smart people you met in
#
these law firms as intellectually curious and very practical because that also sort
#
of describes a lot of the, seems to describe to me at least a lot of the work that you
#
went on to do specifically in health economics, which kind of almost puts you a decade and
#
It's almost like you were on a trajectory to bring you where we are today.
#
So what were the sort of problems that the shiny new things, as it were, that drew you
#
to towards health economics?
#
So again, it's going to be individuals to start with.
#
So just around the time Jim Heckman had introduced me to Dick Posner, I started thinking about
#
I also met this person named Tom Phillipson, who's actually an important person in the
#
current context because I think he's the head of the CEA Council of Economic Advisors for
#
And he was a health economist and had some really interesting ideas.
#
One of these ideas was that you could use economics to understand how people behave
#
during epidemics and how you can use economics to understand how people make decisions about
#
So I began having conversations with him and he was, you know, he'd begun thinking about
#
So he'd already kind of had a head start in thinking about this, but I thought it was
#
And so we talked about that, but the one idea that I latched onto that led me to kind of
#
embrace working with Thomas a bit more was that he viewed, he had this concept called
#
And the basic idea is that when you do a survey, for example, for research, you want to do
#
a survey to see who people are going to vote for, you want to do a survey to see what income
#
people have, you want to do a survey to see, you know, how has COVID affected them.
#
You have to think of it as a labor market, or at least as a market where the respondents
#
are supplying information and the surveyors are demanding that information.
#
And there's some trade and you have to think about what the price of that trade is and
#
what the quality of the supply is here at information.
#
And so that really started influencing me, made me think a lot of differently or think
#
very differently about econometrics.
#
Econometrics, we kind of took the data as given.
#
We had some, you know, discussion of selection, obviously, with Jim Heckman, but it made me
#
think a little bit more about whether or not we could use the idea of a market to improve
#
statistical inference from surveys.
#
And so he and I wrote a paper together in 1999 about these and the role that incentives
#
play and how that changes the inferences that you'd make.
#
But anyway, the main thing that I got from that was, you know, Tom was very much in the
#
mold of Gary and allowed me to kind of think in ways that I hadn't about data.
#
So connecting a lot of the work that Gary was doing with the work that Jim was doing.
#
And I decided I wanted to work with this person more.
#
So we didn't initially work on epidemics.
#
I got to that a little bit later.
#
But we did work on, you know, another very interesting aspect, which is the idea of nonprofits.
#
And this also started moving me towards health because in the United States, about 60% of
#
hospitals are nonprofit, more, I think, if you do it by bed.
#
But so we started thinking about why are some firms nonprofit, some firms for profit?
#
And why do firms decide to, you know, sometimes choose each of these two different forms?
#
Why do markets tolerate it in equilibrium?
#
And why does it only happen in certain sectors and not in other sectors?
#
So we started thinking about that.
#
And again, that got me thinking more about health.
#
And ultimately, when I was writing my dissertation, I think I was able to put, you know, three
#
of these influences all together in a way that's very, I'd like to think very Becker-esque.
#
So I'd started, just like many grad students at Chicago, I went through many topics, and
#
each of them were actually fine.
#
So my first topic was on how judges behave.
#
The second topic was actually on whether or not, this is the law influence, whether or
#
not the felony murder rule, which we can dive into if you want to, but it's a very peculiar
#
rule that says that if somebody dies while you're committing a felony, you're liable
#
for murder even if you didn't intend it.
#
Whether or not that rule actually works and deterred, it actually has very complex effects.
#
And so that was the second paper.
#
But I rejected each of these papers.
#
And the reason is because the standard at Chicago was, are you going to write the thing
#
that is not only going to be published in a top three or top five journal, but that's
#
really going to, you know, kind of put you on the map and put you on a track to winning
#
a big prize at the end of your career?
#
And I didn't think either of those were that big.
#
But then as soon as I was done clerking and was turning back to like finishing my dissertation,
#
I decided to dump all that because I had this idea one day as I was driving to the airport,
#
which is to think about clinical trials.
#
I was thinking about clinical trials.
#
I don't know why I was thinking about clinical trials, but I had this thought.
#
I found it interesting that some clinical trials treated more people than other clinical
#
That is to say some clinical trials, you know, you look at one treatment and you put half
#
the people in control and half the people in that one treatment.
#
And then other trials, you'd have two treatments you're looking at and a control and you do
#
And I thought, well, that's really interesting because especially if it's a blind trial,
#
you know, you're going to have this differences, you're going to get the same treatment maybe
#
if there are common treatments across the two trials, you're going to have different
#
And then it occurred to me that there's this concept I'd read out about called the placebo
#
effect, where your expectations about a drug affect the outcomes that you have after consuming
#
And this is just because I was doing background research with Thomas on data markets, learning
#
how clinical trials work.
#
And it occurred to me that this is a really interesting opportunity to study placebo effects.
#
So I wrote a paper very quickly on how you can compare outcomes across two trials that
#
are looking at at least one common treatment, but that have different allocations of patients
#
So different fractions getting treated to test for placebo effects.
#
And I immediately called up Thomas on as soon as I stopped, this was a, you know, like probably
#
like an hour drive at the end, I called Thomas and said, I had this idea.
#
He immediately saw that it was a great idea and he said, call Gary.
#
So I called Gary and he said, yeah, that's a fantastic idea.
#
You should write that paper.
#
And so within three months, I wrote a dissertation on placebo effects.
#
And so I knew also that that was actually a significant paper because here are two people
#
that thought not just it was an okay paper, but there was a really interesting paper.
#
And so that's how I kind of locked it in for me that I cared about health economics, which
#
is, you know, I had the data market stuff.
#
I had now this dissertation.
#
So health economics was going to be an area I thought was going to be very interesting.
#
And so that's how I ended up back at health.
#
And that's a good time to take a quick commercial break.
#
We'll come back after the break and, you know, get up to speed with bringing you to the current
#
And then we'll talk about that in some detail.
#
We'll be back in a minute.
#
If you're listening to The Scene and The Unseen, it means you like listening to audio and you're
#
That being the case, I'd urge you to check out Storytel, the sponsors of this episode.
#
Storytel is an audio book platform that has a massive range of audio books from around
#
Their international collection is stellar, but so is the local collection.
#
They have a fantastic range of Marathi and Hindi audio books.
#
What's more, I do a weekly podcast there called The Book Club with Amit Verma, in which I
#
talk about one book every week, giving context, giving you a taste of it, and so on.
#
Download that app and listen to my show.
#
And as long as Storytel sponsors this show within this commercial itself, I will recommend
#
an audio book that I liked on that platform every week.
#
My recommendation for this week is Mitra Marjani by Krishna Swapthi.
#
This examination of the unexpressed desires and sexuality of a married woman took conservative
#
India by storm when it was released in 1966.
#
And it still sounds as fresh and relevant today in 2020.
#
So do listen to it on Storytel.
#
Mitra Marjani by Krishna Swapthi.
#
Download the Storytel app or visit Storytel.com.
#
Remember the Storytel with a single L, Storytel.com.
#
Welcome back to The Scene in the Unseen.
#
I'm chatting with Anoop Malani and we've just reached that stage of the career where he's
#
become interested in health economics and we'll soon get to COVID.
#
But before that, Anoop, I want to take you back to something you said at the start of
#
the show, where you said that you had witnessed 9-11 up close.
#
And the tone of your voice gave away that it was something significant to you.
#
So can you elaborate on that?
#
So this was obviously September 2011.
#
I had just started my job as a clerk for Justice O'Connor at the Supreme Court.
#
And everything was fine.
#
You're just getting the hang of things.
#
And all of a sudden, one day we show up to work, we start pretty early and then somebody
#
runs into our office and says, planes crashed into a New York skyscraper, the World Trade
#
And everybody, just like everybody else, everybody has their stories about 9-11.
#
And at first we were really worried.
#
We thought that this was just a small plane and then we saw the second plane.
#
When you see this happening in real time, obviously your skin tingles, your hair stand
#
This is a significant event, at least for those that were in America, seeing what's
#
And then we had heard that there was a plane that had landed to the west of us.
#
So I don't know if you know about the geography of Washington, DC, but to the west of the
#
Supreme Court immediately is the Capitol and then to the west of that, down a ways a bit,
#
And to the west of that, if you go a little bit further across the river, is the Pentagon.
#
We didn't know that the plane had crashed in the Pentagon.
#
We saw smoke in the sky and thought the plane had landed in the White House, which really
#
kind of freaked us out.
#
The place went on lockdown, the justices were evacuated, we had to leave the building and
#
all these men with guns came out.
#
And so we ended up watching with crowds in hastily opened restaurants and bars along
#
At about this time, it just occurred to me, I was so worried about what was happening
#
I didn't realize my sister was actually in New York City, she was going to med school
#
And so I spent part of the day trying to find her and part of the day trying to kind of
#
get my bearings straight.
#
And so that event, I think, allowed me to see that these things that happen in the world
#
that you hear about, and this is not the first major event that happens, that they happen
#
to people and they're real events and they'll have real impacts in a way that's not just
#
a news story that you read about.
#
So that was the first thing that I realized that these are significant events.
#
And one of the things that I noticed, and this would happen to me repeatedly because
#
there were other strangely terrorist events that we've been close to, meaning not a participant
#
in to be clear, but just to like be adjacently affected.
#
It made me realize that there are kind of two ways to address this, you know, I'm simplifying
#
greatly here is that there's some of my colleagues who say like, oh, these events happen and
#
then they go back to their regular research.
#
And some people say these are events that one ought to think about more seriously and
#
understand the impact of.
#
And I noticed that I had, you know, kind of had the latter reaction, which is I wanted
#
to understand these events.
#
And one of the ways that we did it, that year that we had 9-11, we actually had three major
#
events at the Supreme Court related to D.C.
#
One was obviously 9-11, which dwarfs all the others.
#
But a second one was that we had a series of anthrax threats and attacks against government
#
offices, including the U.S. Supreme Court, just two months later.
#
And then towards the end of that year, there was a sniper going around D.C. shooting people
#
from a white van, everybody thought.
#
And so I was interested in that last episode, the D.C. sniper, people thought were targeting
#
people at gas stations.
#
And so people started avoiding gas stations.
#
And so my immediate reaction was, well, I wanted to kind of understand whether that
#
was a rational reaction or not.
#
And I wanted to see if we could determine that by looking at where people went to go
#
buy gas and based on that, kind of back out what their beliefs are about the threat.
#
That is to say, how much cost were they willing to undertake to go someplace else?
#
Now that research topic crashed and burned because I couldn't get access to the gas data.
#
But when these crises happened, I was interested in trying to understand those crises and not
#
so much try to, you know, kind of push them aside and go back to what I was doing before.
#
I guess it was the shiny new object, but had more salience because it had affected me or
#
the location that I was in.
#
And so that's how 9-11 kind of happened.
#
The one other thing I'd say about 9-11 that was a little bit controversial, not controversial,
#
but I think is a bit different is I feel that when events like this happen, people kind
#
of rush to conclusions about what that event means and what to do in response or what not
#
And I don't have that view.
#
My view is that at the beginning, there's just a ton of uncertainty and that one ought
#
to entertain a wide range of views and potential responses and discuss them before deciding
#
That's kind of controversial because, you know, the other thing I observed during this
#
is that not only is there a lot of uncertainty, but there's a ton of hurting, which is to
#
say people kind of heard around particular conclusions about an event and rushed to judgment
#
It's not irrational, by the way.
#
And the reason I know that is because of an important paper that Objet Banerjee wrote
#
back in I think 1991-92 about hurting, that it can be quite rational, but it could mean
#
that people rationally rushed to judgment and then the potential for mistakes could
#
And so at 9-11, I think the rush to judgment, which I think in hindsight we know was the
#
conclusion that the right response was a military response and that Iraq and Afghanistan ought
#
We learn in hindsight that maybe that might not be the right approach, might have been
#
much more costly than the event itself, but there's that heard to rush to judgment.
#
And so I think we see the same sort of thing with COVID, that same sort of dynamics play
#
out a little bit, but we should heed the uncertainty.
#
That's the other thing I learned, I think, from 9-11 and learning now from COVID.
#
And that also, just thinking aloud, it seems to me that some of the response to 9-11 might
#
have been dictated by the fact that the incentives on the politicians were to appear to do something
#
about the problem instead of taking their time and doing things which might make an
#
impact but are not that visible.
#
Do you think that played a part then and do you think that is also an issue now?
#
And in fact, that is a very important point.
#
If we were to write a book about this, I would say what you just said is a theorem, which
#
is that politicians feel the need to act.
#
And that causes people to act probably before the information warrants it.
#
Let me offer a corollary to that theorem, which is that people that are in the business
#
or have the jobs of gathering information to inform policy making are influenced by
#
what the politicians want to hear or what the policy consequences are.
#
They're influenced by that when they decide to gather the information, interpret the information.
#
That means that these stories that we hear about intelligence agencies feeling the pressure
#
to find evidence of WMDs weapons of mass destruction after 9-11 in Iraq felt like they had to provide
#
more evidence in favor than against, regardless of what the data actually said, because that's
#
what the people higher up wanted to hear.
#
And so you can imagine the same sort of incentives play out during COVID, which is depending
#
on whether or not your boss, which is your higher up in the bureaucracy, whatever government
#
you're in, do they want to hear that cases are high or cases are low, whether they want
#
to justify a lockdown or want to exit from a lockdown, that will influence what you report
#
And what that means is that we may not always get the best information.
#
And if you want to be more sophisticated about it, it means that you have to account for
#
the incentives that the producers of information have when you interpret that information.
#
So in some sense, it goes back to that data market concept.
#
But it's an important, I think, an important corollary to what you said.
#
And I was going to ask a question about incentives in the current case a little later on.
#
But since we are on the subject, I'll just ask it now and then we'll rewind and we'll
#
sort of come back to health economics and your other work on pandemics, which is that,
#
you know, when I think about what the incentives are for current governments, it strikes me
#
that, you know, whatever option the state chooses now, the costs are seen immediately
#
and the benefits are unseen.
#
So the incentives of politicians are to play it safe in specific ways.
#
For example, if I look at Prime Minister Modi earlier, his incentive was very much towards
#
calling the lockdown and, you know, playing it safe rather than have the dead spied up
#
because of the dead spied up from Covid, that would be, you know, something that would be
#
a visible problem, especially for governments which are particularly concerned about optics,
#
as is the case in India, where a lot of the sort of your incentive is towards controlling
#
And in a similar way, you have a situation today where, for example, I am, you know,
#
I'm in Bombay, Uddhav Thackeray is the Chief Minister of Maharashtra, and there is a lot
#
of pressure from the central government, you know, there's a narrative war that is happening.
#
There is a theory that at some point, the central government wants to impose President's
#
rule because Uddhav Thackeray is not being able to control the pandemic here.
#
And therefore, the incentive for Mr. Thackeray, whether or not he is responding to these incentives
#
But the incentive on Mr. Thackeray would be to keep the numbers low.
#
And one way of doing this is by not testing enough.
#
And in fact, on a daily basis, one hears horror stories locally of people who have had symptoms
#
for 10 days, 12 days, and they go to hospitals and the hospitals refuse to even test them.
#
So suddenly you have a situation where these optics matter and these other political dynamics
#
matter and pervert the incentives to an extent where the people in charge may not necessarily
#
be taking the best decision.
#
As someone who works closely with governments, have you found that to kind of be a worry?
#
I want to be very careful with my answer here because I want to continue working with governments.
#
So let me try a different approach, which is going to stay what you state, but in a
#
more general way, an oblique way, but then ultimately agree.
#
So let me turn instead to an idea from evolutionary biology, which is one of the things I like
#
This is an example I really wanted to show.
#
I was very excited the first time I was able to tell my kids this because I've been waiting
#
until they were ready to absorb it.
#
But in general, we think that more information is better.
#
We meaning humans, at least modern thinking humans that have university education.
#
We think more information is better.
#
The reality is that evolution doesn't see it that way.
#
That when you compare species, species don't have an unconditional incentive to obtain
#
information and truthful information.
#
They have only the incentive to gather the information that enables them to survive.
#
So for example, this can explain the difference between say humans who rely on sight, feel,
#
and sound versus say bats who rely on sound primarily.
#
And it's really you obtain the information gathering ability that's optimum for survival,
#
You don't gather all the information that you want.
#
So we talk about that in the context of evolutionary biology, and then I always ask my kids like,
#
so how does that affect other things and try to push them a little bit like a law teacher
#
might push their students?
#
And I get them to see that that's everywhere, that our incentives to gather information
#
are a function of what the utility or the value, the practical use of that information
#
is going to be, but that that practical use can also affect the way that we gather the
#
So you're not always getting quote unquote unbiased information about the world.
#
Now bringing it back to your topic, that same concept applies here, which is a government
#
official, a politician doesn't have an incentive to gather truth for the sake of truth, not
#
only in the mundane ways in the sense that information is costly and so you need to economize
#
on on expanding effort and resources on getting information, but also they might find that
#
certain conclusions have different payoffs or different conclusions have different payoffs.
#
So for example, in a democracy or in a place where public opinion matters, you know, what
#
the public thinks is important.
#
So in that sense, when you're gathering information, you're not really just trying to find an unbiased
#
estimate of what's going on in the world, whether it's the amount of COVID in the population,
#
things like that, what you're really doing is trying to solve a, you know, play a beauty
#
I don't know if you remember what the standard beauty contest, the goal is, and sometimes
#
stock markets are characterized this way, the goal is, is to figure out what the price
#
is that everybody else will want to price the stock at, not so much what the underlying
#
true value of the stock is.
#
And so that's the classic kind of beauty contest approach to or structure for information acquisition,
#
To some extent, I think that's what politicians are trying to do.
#
They're trying to figure out what is going to get the best reaction in the population.
#
And that's obviously going to be influenced by whether or not you are, you know, in a
#
lockdown now and think that what you want to do is keep that going for political reasons
#
or whether you're not in a lockdown, you want to get to a lockdown, I'm not going to pick
#
any leaders, I'm not going to pick out any leaders and point them out.
#
But it's also going to be the case that the dynamics between leaders really matters.
#
So if you take one position and somebody else takes another position, that's a political
#
threat, what you'll want to do is gather information that supports your views as opposed to that
#
other opponent in the political process.
#
And so again, it's going to skew the information that is produced.
#
Now, you know, in a best case scenario, what would happen is that the data would be all
#
out there unvarnished, and then you would just have to make arguments.
#
I think that's the kind of the best case view of the world, you know, kind of the way that
#
legal trials operate, which is all the data is presented and the judge decides as between
#
arguments that are made by the plaintiffs and defendants.
#
But the reality is, you have to think about the incentives to gather information and the
#
way to mask that information.
#
And I think, you know, what I'm looking for, and I imagine the most sophisticated politicians
#
do this, is that they realize that when they want to make decisions, they want to consider
#
not only the political aspects, but they'd like to see all the information.
#
And I imagine the most sophisticated, or I'd like to imagine the most sophisticated politicians
#
understand that their underlings have to be given incentives to provide all that information
#
and not the information that the underlings think the politician wants.
#
And so that way, the politician can make the best judgment, whatever the reason, whether
#
it's to compete in a political contest or to do well by the population or both.
#
So I guess the thing that interests me in this context is not so much, you know, what
#
information are we getting now, but as we look forward, what kind of structure should
#
we put into place so that despite these incentives, that unvarnished information comes forward?
#
Yeah, it makes absolute sense.
#
In fact, you know, when I mentioned Modi and Uddhav Thackeray in my question, I did not
#
intend to condemn them for the choices they may or may not be making.
#
They are rational people responding to incentives, and I think what one needs to do is not blame
#
individuals, but look at the system and the incentives the system provides.
#
So I think the answers are structural.
#
And when you mentioned the beauty contest, I immediately, you know, thought of the story
#
of the two guys who are in a forest when a tiger spots them from a distance and starts
#
Have you heard that story?
#
Yeah, you just have to be faster than the other guy.
#
So they're both running and one of them is the other guy that, you know, we are never
#
going to outrun the tiger.
#
And the other guy says, I don't have to outrun the tiger.
#
I have to outrun you, which is the way those specific incentives are structured.
#
And it also then strikes me that, you know, let's again sort of zoom back and come to
#
maybe a decade before covid strikes.
#
And you know, everyone knows that there is a possibility of a pandemic like this happening.
#
Everyone also knows that it's a very low probability event.
#
Now the thing is, if there is a low probability event, which has a relatively high and visible
#
cost of prevention, you know, if it's really low probability, no matter how high the cost
#
of the event might be, the temptation of the politician is always to avoid the cost of
#
preventing it, because he can always be questioned on that.
#
And that will, of course, carry an opportunity cost, whereas he could just spend his time
#
and effort somewhere else and not worry about the pandemic.
#
In that sense, for countries which hadn't, you know, been exposed to this kind of a pandemic
#
And of course, you know, Southeast Asia was better prepared because they had been through,
#
you know, other similar, much smaller pandemics.
#
But for the rest of the world, is it then inevitable that they were going to be underprepared
#
when covid came around?
#
The short answer is yes, and it was rational.
#
So in the 2000s, in the first decade of this century, I wrote a paper with a colleague
#
of mine, Albert Choi, where the thing that provoked it was this exactly this dynamic,
#
which is, you know, why weren't we prepared for 9-11?
#
Why weren't we prepared for global warming?
#
Why weren't we prepared for whatever it happened to be?
#
And, you know, some crisis happened.
#
I forget what the triggering event was.
#
It might have been just 9-11 itself.
#
But everybody was blaming the government.
#
And I thought that's kind of silly because, you know, it's rational for the government
#
not to act before a crisis in the following sense.
#
If you survey the literature right now, there are tons of warnings out there.
#
OK, there's warnings about global warming.
#
There's warnings about financial crises.
#
There are warnings now about epidemics.
#
The reality is there's even more than that.
#
There's warnings about the risk of asteroids and so on and so forth.
#
And governments can't address every single problem, OK?
#
There's limited resources.
#
So they have to choose.
#
And the problem is that you have experts in each of these areas saying mine is the most
#
important crisis or mine is the most significant crisis.
#
And the government doesn't know which expert is correct.
#
So the challenge for the government is to figure out how do you decide.
#
Now obviously, they're going to act on some things or another.
#
So for example, you know, the personal relationship between an expert and a politician might cause
#
that politician to weigh one expert's view a little bit more than the others.
#
But in general, or an expectation across these politicians, there's no particular thing that's
#
going to stand out that has a credible signal except for one thing, which is a crisis.
#
So when there is a crisis, you know, terrorists crash in two towers in New York and then the
#
Pentagon or when, you know, actually the stock market crashes because of, you know, subprime
#
lending in 2008 or when there's a actual disease that spreads, then politicians know they have
#
a credible signal that there's something very important and that they have to take that
#
And that's a really kind of rational way to act, especially when it's not just a one-time
#
event that will never occur again, but it's an indicator of how serious a class of threats
#
So I think that, you know, 9-11 taught governments that terrorism was something to take seriously.
#
Whether or not it overreacted or underreacted, that's a separate issue.
#
That's where the theory of herds comes in.
#
But that's one thing that I think was the signal of credibility.
#
The regulations that come after the 2008 crisis, I think is another example in the United States.
#
And I think that after COVID, we'll be much better prepared for COVID 2.0.
#
And it's this sort of theory, I think, that can explain why it is that East Asian countries
#
like Taiwan and Korea, et cetera, were better prepared for COVID.
#
And that's because they saw in 2003, SARS and that event, that prior crises, which didn't
#
hit all countries, but hit a subset of countries, kind of got them to see that this is a credible
#
And countries that SARS did not hit, again, said there was no crisis, so maybe this is
#
And so they didn't take it seriously epidemic control.
#
By the way, it also explains why it is, perhaps, why Kerala has done relatively better than
#
other states in COVID responses, because they saw the Nipah virus crisis.
#
And so they saw this is a real risk.
#
And so we have to prepare for it.
#
And so to some extent, yes, Kerala did fantastic, but part of the success that Kerala had was
#
that it had an early experience with the crisis, made it realize that this was something worth
#
investing in, worth preparing for.
#
So your theory is correct.
#
And would it also be fair to say, and again, this is a brief digression, but since you
#
brought Kerala up, that one of the factors would also have been that they have stronger
#
systems of local governance.
#
And a crisis like this really becomes difficult to control from the top down the further away
#
you are, because your information is that much diffused and the incentives get messed
#
So the more local the actual governance is, the better the flow of information, the better
#
the incentives, and the more chances that you're going to handle something like this.
#
Yeah, I would say there's two components about Kerala.
#
So I want to think a little bit more about the local, but let me go to the other one
#
first and then come back to the local.
#
So one of the things I always found very interesting about Kerala, there are various people that
#
have made the argument.
#
I think Amartya Sen has made this argument, and among many others, that Kerala performs
#
very well relative to other states in India.
#
And the argument had been, for some time people thought that it was because the communist
#
government was doing a good job.
#
I'm not entirely sure that's the case.
#
I think Kerala was doing well even before communists took over, if you look at historical
#
data on literacy and things like that.
#
But I do want to give credit to the communist legacy a little bit.
#
And I want to tell you that there's costs and benefits, and then get back to the local.
#
So the other state that comes to mind as I think about this is Vietnam.
#
Vietnam has done remarkably well if the data are taken at face value through this COVID
#
And it's not just in COVID, by the way, I had a separate line of research the last few
#
years thinking about slums.
#
And Vietnam is very interesting too, because from what I can tell, not very many slums,
#
especially given its per capita income.
#
And so one sense that I get is that one of the things that communist governments do is
#
And that is to say, build governments that are very capable of acting in whatever domain
#
And in this context, there was a high return to having a lot of state capacity.
#
And so that makes me think that legacy of having a communist government in charge really
#
built up the state in a way that was helpful for COVID, both here in Vietnam, that would
#
Now, there are costs associated with that too, I don't recommend communism as a peacetime
#
government very much, actually not at all.
#
And the reason is because I think that having the government play a big role in peacetime
#
means that you crowd out the private sector, which in peacetime is, I think, a more effective
#
method of providing goods and services to the citizenry in peacetime.
#
But certainly in crises, communist governments might have an advantage worth investigating.
#
So I think that's part of it.
#
Now, in terms of the centralization versus decentralization, which is, I think, the point
#
that you were making initially, I'm not sure, I do think local governments have a strong
#
role to play in having local state capacity and have a strong role to play.
#
But it's also important to remember that there's a trade-off, which is the information on COVID
#
That is to say, you want to be informed about what's going on elsewhere, not just what's
#
going on locally in order to be able to make informed judgments.
#
Then you want to be able to act locally.
#
So you want both local capacity but kind of a global awareness of the risk to be effective.
#
So I think the thing that makes this kind of interesting from Carol is they had a little
#
bit of both at a state level and maybe at a global level.
#
They seem to have a, again, maybe I'm speaking prematurely, but they seem to have a sense
#
of the magnitude of the event and they had the ability to think about it and act on it
#
at a local level and to do so in a manner that wasn't, at least from an outside perspective,
#
wasn't as bogged down by political competition.
#
So I think that's the other aspect of it.
#
And just an aside, how a pandemic can invert our sense of values where normally we know
#
the damage that communism has done throughout the world in the 20th century, but suddenly
#
in a pandemic, this aspect of communism comes to the rescue.
#
Similarly, I think all sane people would agree that urbanization and population density is
#
That's why, you know, the history of humanity is one of migration from rural areas to cities.
#
But now that same population density is, you know, turning out to be a temporary curse
#
while this pandemic rages.
#
Let me kind of take you back, let's say three months, you know, when people are just beginning
#
to realize that COVID is something serious.
#
And at this point in time, and as you've spoken about at length and we'll discuss, there isn't
#
a binary between lockdown and no lockdown at all.
#
There is a gradation of actions and conditions in between that we can start, you know, thinking
#
about and measuring and working towards.
#
But at the start before, you know, let's say in early March when, especially in India,
#
when people began to be aware of this, how do you look at the policy options before the
#
policymakers, before the governments?
#
And it's easy to sort of in hindsight talk about all the mistakes that have been made
#
and many have been made.
#
But what was your thinking at that point in time about how India should tackle it?
#
So the first thing that I thought was that we need more information.
#
Now we didn't have a lot of information about what was going on.
#
The information infrastructure was just being built, but that was the first thing I thought.
#
Then I thought, you know, until we get the information ready, I can see the value of
#
having a lockdown in the sense that it would allow you to avoid an irretrievable loss.
#
What do I mean by that?
#
I mean that, you know, just let's use the simple framework, which I actually don't think
#
is wrong, that there's really a choice between economic activity and deaths from COVID.
#
And you don't know what the magnitude of the deaths from COVID are.
#
You kind of have a sense of what the value of economic activity is, but there's a lot
#
of variation in what the potential health harms could be.
#
And it's irretrievable in the sense that once somebody dies, you can't bring them back.
#
I know as an economist, I can just put a dollar value on that, but I want to understand that
#
or appreciate that there is a kind of an irretrievable loss to this, or at least in the way that
#
people think about this.
#
In that context, you might want to be particularly conservative at the beginning and try to do
#
what is possible, maybe a temporary lockdown in order to stop the spread of a potentially
#
bad disease, focus really hard on getting information about the nature of the disease,
#
focusing not just what's going on in India, but other places in the world.
#
And on the basis of that information, make a subsequent decision about whether or not
#
to extend the lockdown or relax the lockdown or do something else.
#
That I thought seemed like a quick rational response that you could have.
#
The only thing that I would say there is that I thought that the calculus was going to be
#
different for the United States than for India for many reasons.
#
But the most important one is that the cost of a lockdown varies dramatically across those
#
two countries for two reasons.
#
The first is you have a much lower income population profile in India.
#
A lockdown was going to affect the poor more than it was going to affect the rich.
#
People that are on salaries can sit and take a few weeks off work without really losing
#
their income, or at least they have a savings buffer, whereas people that are poor don't
#
And so lockdown is going to have a disproportionate effect on the poor, and India had a poor
#
So that was the first consideration.
#
But the second thing, the reason I thought that lockdown was maybe not as obviously an
#
answer for India is that there's a difference in state capacity.
#
So the United States has a pretty good capacity to enforce the lockdown, notwithstanding the
#
riots we're seeing as a result of the police brutality against George Floyd in Minneapolis.
#
In general, the United States has a lot of state capacity to enforce this lockdown.
#
I was worried a little bit about India's capacity to do so.
#
I have to say that I was actually kind of surprised and forced me to come up with the
#
theory for how India was able to enforce one of the strictest lockdowns.
#
I think I have one, but it did tell me that it couldn't last for a long time.
#
It would just be harder as people chafed.
#
And so that would be the reason why I would think that while lockdown might be the right
#
answer for India, it would be a shorter duration lockdown than you'd see in other parts of
#
And I think that's been a source of tension.
#
I mean, it's TBD to be determined whether or not India's lockdown will actually be longer
#
than other countries or not.
#
I think that we're seeing that after June 8th, there will be a relaxation under the
#
newest MHA orders, all those states are free to extend, and you're starting to see some
#
relaxation in the United States, especially in the South.
#
But those are the two reasons why I think a lockdown is a good idea, but B, I think
#
that lockdown probably would have, I would have expected it to be shorter in India than
#
Yeah, we are recording this on May 31st, by the way, and the episode airs on June 7th.
#
So if many things have changed by then, don't blame Anup for me.
#
So you know, you mentioned that you had a theory about how the lockdown in India would
#
Okay, so I think that if everything went well, meaning the government responded appropriately
#
in an idealized way, what would happen is you do a lockdown, you'd start gathering
#
a lot of information, you'd start preparing hospitals and things like that.
#
So what does that mean?
#
That means you would use this opportunity to ramp up testing, to figure out what was
#
going on, to make sure that the lockdown was working, etc.
#
And the second thing you do is build out beds.
#
That's really what this whole, you know, when people say flatten the curve, you know, having
#
I'd worked with SIR models, to some extent, before this in a few research papers.
#
And so my sense was, you know, when you do a lockdown, you suppress your transmissibility,
#
that doesn't mean that unless you have your lockdown goes forever, it means if you let
#
up from the lockdown, the epidemic returns because your transmissibility rises again,
#
you're just delaying the kind of the jump in infections.
#
The value of that is that you prepare.
#
So you do testing so you know where to respond, but also importantly, you build up beds so
#
that the peak is not, while you might have the same number of infections, the death from
#
those infections would fall.
#
And so that was the idealized response.
#
So that was part of the answer to your question.
#
The other part that surprised me that India was able to sustain this lockdown for so long,
#
even though it is known to have a very low rate of police per capita, that is to say
#
low administrative state capacity, was the following, which is it occurs to me that when
#
you have very little police force, it's actually easier to enforce a lockdown.
#
And the reason why is because once, if you can get the population to just kind of agree
#
to a lockdown, like credibly they stay home, then it's really easy to find violators.
#
Because if there's very few people on the streets, the few people that you find, you
#
can go after with a limited resource, a limited number of police officers.
#
That is to say, it's easy to detect violators with even a low police force in a lockdown.
#
It's when you get to that intermediate stage where you begin to relax your lockdown, that
#
I think India will have more challenges than, say, the United States.
#
The reason is because it's harder to find the violators of whatever rules that you have
#
as there are more people that are actually out in the street and interacting.
#
And that's going to quickly overwhelm the police force in India.
#
Whereas in the United States, you might actually be able to enforce that.
#
You not only have a better population monitoring infrastructure, but you also have more police
#
And so that's how I, at least that's tentatively where I stand for trying to understand how
#
India was able to have a very harsh lockdown with low police force per capita.
#
It also makes me a little bit pessimistic about the ability to some extent of doing
#
However, to me, I see that as a challenge.
#
The challenge that I get from that is how do you devise enforceable gradual social distancing
#
as opposed to complete social distancing under a lockdown?
#
And as Jude mentioned earlier, my view is that gradual social distancing or moderate
#
amounts, especially targeted, is going to be better for achieving the proper balance
#
between economic activity on the one hand and controlling infections and deaths on the
#
other hand, both of which are, I think, very important values.
#
But that's a profound insight.
#
And I have three related thoughts before I ask you further questions about the different
#
gradations of social distancing.
#
One of the papers I'm going to link from the show notes is the paper you wrote in the Journal
#
of Development Economics with a couple of co-authors called Learning During a Crisis,
#
a SARS Epidemic in Taiwan, where you speak about exactly this aspect of how the people
#
themselves, even though there were very few cases of SARS, the people themselves started
#
avoiding public places and restaurants and malls and all of that.
#
And gradually, the pandemic everyone was scared of just petered out.
#
And you refer to how an informational cascade can play a part in determining mass behavior
#
where they can overreact in one direction or the other.
#
And it seems that those kind of informational cascades probably did lead to more people
#
staying at home during the initial part of the lockdown.
#
So you're right that even though we have less police per capita, they would have had less
#
policing to do because there'd be fewer breakers of the law.
#
That's quite fascinating to me.
#
My other sort of thought was, and I discussed this in an episode I linked from the show
#
notes, which I did with Shruti Raj Gopalan on COVID, where she sort of spoke about how
#
the cost of not having a lockdown would have been disproportionately on the poor.
#
And the cost of the lockdown are also on the poor.
#
Now the thing is, the thought that this brings me to is that unfortunately in India, the
#
And therefore, when I go back to the incentives of the powers that be, I see two things.
#
Number one is that the costs of the lockdown are less visible than they would be if the
#
pandemic was allowed to rage.
#
And therefore, it is sort of easy for them to build narratives that completely ignore
#
that and say that, no, we'll keep the lockdown going and we'll keep the numbers down because
#
the other consequence is that if they remove the lockdown and the numbers go up drastically,
#
they'll be blamed for it.
#
It is safer to continue the lockdown as it is.
#
And the other aspect of this is that therefore, even though we want more information, even
#
though we want far more testing, their incentives are actually to do less testing because the
#
less testing you do, the lower the numbers will be and the better the optics for you
#
to make the case that, hey, we did the lockdown and it worked.
#
How do you sort of feel about this might be repeating my earlier question about the incentives
#
of politicians, but this seems almost an insurmountable problem to me when I think about it.
#
So you raised three very important questions.
#
So the first is learning during a crisis.
#
The second is about the impact on the poor.
#
And the third is about testing rates.
#
So let me try to address those in turn.
#
So first, let's start with the Journal of Development Economics piece.
#
So that was with Dan Bennett, among others, a colleague of mine at the Harris School.
#
And the interesting feature that we found in the SARS data coming out of Taiwan was
#
that even though this was a medical crisis, visits to hospitals fell dramatically.
#
You can see that in the claims data.
#
And so then we tried to figure out what was going on.
#
And what we kind of deciphered through looking at the data was that what had really happened
#
was that people were worried that they were going to get infected by going to the hospital.
#
So they avoided hospitals.
#
And that, paradoxically, that actually helped control the epidemic because you had one less
#
area where you were congregating.
#
And one could make the argument that hospitals actually were a vector for transmission of
#
And we speculated that this might actually have slowed down the progression of SARS.
#
But I want to stress two things about this paper that I think are interesting here.
#
First is just a little bit of academic history.
#
So back in the 1990s, epidemiologists were already talking about SIR models.
#
This had been going on for decades, obviously, of disease, even diseases that afflicted humans.
#
But they never really explicitly modeled how humans reacted.
#
And there were two scholars in parallel that had introduced the concept of something called
#
One was Thomas Philipson, who was on my dissertation committee with Gary Becker.
#
And the other one was Michael Kramer, who just won the Nobel Prize in economics.
#
And they posited this idea that in the midst of a crisis, that is to say, in the midst
#
of an epidemic or any disease, and they were thinking about things like HIV, that people
#
would take precautionary behavior on their own in response to the level of risk.
#
And when the level of risk was high, they'd do a lot of it.
#
And when the level of risk fell, they'd take more risk.
#
People would do less social distancing or other protective efforts.
#
Now, what they hadn't thought about was how people gauge that risk.
#
Their assumption was people could immediately see what the prevalence was, like the truth
#
on prevalence, and then adjust their behavior.
#
And so what the JDE, the Journal of Development Economics paper, tried to do is think about,
#
well, how do people form those beliefs in the first place?
#
And what we realized is they were not looking at raw data.
#
They were looking at, in addition to whatever data that might be made available in the press
#
or whatever, they were thinking about what their colleagues were doing, what their neighbors
#
And so you could get these big hurting effects, where people really kind of strongly distanced
#
themselves in response, so this ties back into some of the work that Abhijit Banerjee
#
So that was the important thing.
#
It's important to remember how people learn about disease risk to understand how they're
#
going to voluntarily respond.
#
So that was the first thing, I think, that came out of that Journal of Development Economics
#
paper that was interesting in this context.
#
People are learning about the crisis, and the amount of social distancing we'll see
#
after the lockdown ends will be a function of what they learn.
#
Testing policy is going to be important, but also what other people in the street are doing,
#
what their neighbors are doing.
#
And that's a little bit harder to predict.
#
The second thing that was really important that came out of this was how to do cost-benefit
#
analysis in this context.
#
That is to say, the question you asked was, is a lockdown a good idea or not?
#
And I think the initial simple way that people thought about it was, well, there's this trade-off
#
between health and economic activity.
#
And if you lock down, you get health and you get less economic activity and vice versa.
#
And then there was this wave of belief that actually, no, no, no, that if you released,
#
you would get less economic activity, that the disease caused less economic activity
#
because people would voluntarily social distance, even if there wasn't a lockdown.
#
And I think that there's some truth to that as well, except I think you have to combine
#
both views to really understand what's going on.
#
And then the Development Economics paper, the SARS paper, I think, helps me think about
#
So I think the way to do cost-benefit analysis in this context, that is to say, ask yourself,
#
what is the impact of COVID versus what's the impact of lockdown, is to understand the
#
And I apologize if this is a little bit a longer discourse, but I've been thinking a
#
So think about the chain of causation that you've got going on here.
#
You go from COVID or the disease, the disease causes some policy changes, and then those
#
policy changes cause some outcome along with the disease.
#
And so if you're trying to figure this out, you can break that up into two parts, which
#
is to say, first, think about what affect different policies, what the outcome would
#
be in terms of number of disease, amount of disease, amount of economic activity with
#
and without a lockdown, let's say, just to be simplistic about it.
#
And then think about what policies are caused by the disease.
#
That is to say, given the disease, what policy is the government going to pursue?
#
And that's how you can figure out what the impact of COVID is, separately from what the
#
impact of the lockdown is.
#
Does that make sense before I proceed?
#
Yeah, that makes absolute sense.
#
In fact, that was going to be my next sort of question to you.
#
So this is what, and I'm teaching a development economics class, actually, just finished teaching
#
development economics class.
#
We did a little segment on this, and this is kind of the way that I describe it to students
#
So let's take the second part of that step, which is, what is the effect of policy on
#
And imagine that there's two policies out there, lockdown or no lockdown, just for simplicity.
#
And then think about two types of outcomes, people dying from the disease and economic
#
The important thing to remember is that when you have a lockdown, you have a mandatory
#
social distancing that will, let's just for simplicity say that non-pharmaceutical interventions
#
like a lockdown actually reduce disease.
#
This should reduce the disease, but it will also reduce your economic activity.
#
If you don't do a lockdown, you will still have some distancing, but it will be voluntary.
#
People will decide what distancing do on their own.
#
We call this voluntary social distancing.
#
So you'll have some voluntary social distancing, which will also have these effects of lowering
#
the number of cases and also decreasing economic activity.
#
And the real question is whether or not you will have more social distancing with voluntary
#
social distancing or with mandatory social distancing.
#
And if you think that it's going to be more with mandatory social distancing, then you're
#
in the standard framework that people initially had, which is the lockdown will cause more
#
distancing, which will reduce the disease, but also reduce economic activity more.
#
Alternatively, if you think it's the other way around, if you think voluntary social
#
distancing will cause more, then you'll have the opposite view.
#
And ordinarily you would think that mandatory social distancing is more significant than
#
voluntary social distancing.
#
But that's not obviously the case.
#
And the evidence for this is actually Sweden, which has pursued a voluntary social distancing
#
policy and has seen remarkable reduction, not just in cases.
#
Now, I don't want to get into a debate about Sweden, but they've seen a remarkable reduction
#
in cases, but importantly, relative to say the alternative modeling prediction, but importantly,
#
they've seen a big reduction in economic activity, suggesting that it's theoretically possible
#
that voluntary social distancing does more than mandatory.
#
I don't want to extrapolate from Sweden because not all societies are like Sweden.
#
My kind of prior view is the mandatory social distancing will actually achieve more social
#
distancing than voluntary.
#
And so I think that the classic framing, which is that you get more disease reduction, but
#
also more reduction in economic activity with mandatory social distancing than voluntary
#
That is, I think, the right framework, although the degree of benefit might be smaller because
#
in the counterfactual you would get some voluntary social distancing.
#
So that's the first thing, the conclusion I kind of came to.
#
The second one, though, I came to is it took me back to Hayek, actually.
#
So one important difference between mandatory social distancing and voluntary social distancing
#
is that mandatory social distancing is coordinated, in this case by the government.
#
And the question is, is coordinated social distancing better than uncoordinated, kind
#
of free for all voluntary social distancing?
#
Now, ordinarily, you'd want to say the answer is yes, but I want to think a little bit more
#
And so what I suggest is thinking about cabining the different costs and benefits of social
#
distancing at the individual level.
#
So when I'm deciding whether or not to socially distance or I'm thinking about the costs and
#
benefits of social distancing to myself or in my vicinity, I think about the following.
#
Obviously, there's some private benefits.
#
I've reduced my risk of infection.
#
There are some private costs.
#
When I distance, I can't go shopping or go to work.
#
My lower income might lower my consumption.
#
But there's also a non-private component of it in my immediate vicinity, which is the
#
So when I go out and don't socially distance, I might risk spreading the infection.
#
And it's also the case, though, that when I don't go out, I won't provide demand for
#
So I'm going to affect other people's incomes.
#
So there are personal costs and benefits to myself.
#
And then there is these external costs and benefits of engaging with the world or not.
#
Now, the way that I think about this, and this is very much in a kind of a Hayekian
#
view of the world, is that I probably have better information about the costs and benefits
#
to myself, particularly for the economic benefits, and especially when there's a lot of heterogeneity.
#
Some people benefit a lot more.
#
So if you're a salaried worker, the benefits are smaller than if you're a daily or hourly
#
But society might have better information on the externalities.
#
And more importantly, they might do a better job of or have better incentives.
#
When I say society, I mean government, they might have better incentive to account for
#
And so when you're trying to decide whether or not I want uncoordinated social distancing,
#
voluntary social distancing, or mandatory, you're really making a decision about whether
#
or not you think that the heterogeneity and private benefits and costs are bigger than
#
the social externalities.
#
If you think externalities are more important and that the government has more information
#
about that, you want the mandatory.
#
If you think the private heterogeneity is more important, you want the voluntary social
#
And I think that's the second important component of thinking about whether or not it's better
#
to do lockdown versus a rely on voluntary social distancing.
#
So now you've got the two important things to think about when deciding whether or not
#
you think lockdowns are going to...
#
What effect the lockdowns are going to have on both cases and economic activity and which
#
is better from a welfare perspective.
#
Once you have that, that tells you normatively whether you should support a lockdown or not.
#
And again, I'm simplifying lockdown versus no lockdown, obviously you can do graduated
#
But then in order to figure out what the impact of COVID is, I think people are a bit confused.
#
What they need to think about is COVID not only had an impact on deaths, but they also
#
had an impact on policy, which then affects those things.
#
So the next step that you need to figure out is what is the likely impact of COVID on policy
#
This really gets back to some of the original questions that you had posed about what their
#
That's where that comes in to play.
#
So if you want to ask yourself the question, what is the impact of COVID?
#
You can't do that without thinking about what the impact is on policy.
#
Policy is kind of a folded into that question.
#
So that's, I think, where people are not thinking as clearly as I would suggest about separating
#
the idea of impact of COVID versus what is the right policy.
#
So that's how it split that up.
#
And the JD paper kind of ties back into this because it sheds light on the private voluntary
#
social distancing aspect of it.
#
Let me pause there so that you can kind of provide feedback before I turn to the other
#
aspects of your question, which is the discussion you had with Shruti on poverty, which I think
#
is super important, and then the implications of low testing rates.
#
I mean, this is incredibly fascinating and I'll take time to process it.
#
But a few thoughts that come to mind.
#
Number one, I'd say that, and I think you'd probably agree with, you know, that original
#
sort of framing of the choice as lives versus livelihoods is a little simplistic because
#
no matter what choice you make, whether you have a lockdown or no lockdown, if you accept
#
a binary of just those two options, you are going to affect both lives and livelihoods.
#
My second sort of thought from there is that we will forever be in an epistemic fog.
#
There is no, you know, as regards the costs and benefits, there's no way to really say
#
how the counterfactuals would have turned out in the very long run.
#
And this is something that therefore will be sort of determined by whatever ideological
#
proclivities one might already have, which, you know, and that brings about the curious
#
situation where, you know, because Prime Minister Modi was for the lockdown and President Trump
#
has sort of been speaking against it, that you have the left and the right on opposite
#
sides of the debate in India and the US, where in India, the right wing is sending WhatsApp
#
messages about how important lockdowns are.
#
And it's the other way around over there.
#
And it strikes me that this debate may never really be settled because there is that epistemic
#
And obviously, normatively, therefore, you and I would say that, look, people have the
#
best idea of the trade-offs available to them and let them make their own decisions.
#
And you know, as in the case of Taiwan, you might find voluntary social distancing actually
#
solving the problem if people are just allowed to make their own decisions.
#
I mean, the counter to that, of course, would be what about the externalities?
#
But as you pointed out, the externalities cut both ways, that if people don't socially
#
distance, they are putting others at risk.
#
But if they do socially distance, they are affecting the economy, the demand in the economy,
#
and that will also take life somewhere down the line in unseen ways.
#
So you know, I don't know what's the resolution of this, because do you think that that epistemic
#
fog can ever be lifted?
#
Or would you agree that this is something that will be debated for decades about what
#
was the right thing to do?
#
And it will really depend on which tribe you belong to.
#
So I would say I'm somewhat optimistic.
#
I'm cautiously optimistic.
#
So the first thing I'd say is, I believe that in a few years, over the course of the next
#
few years, you will see more and more studies that more and more rigorously, although not
#
perfectly, assess the causal impact of lockdowns.
#
And obviously, you can't solve everything, because you can't observe the alternative
#
It doesn't exist, which is where we didn't do lockdown in India, or we released on different
#
days in India, things like that.
#
But we will see a combination of epidemiologists and social scientists get together and try
#
to come up with counterfactuals.
#
The only kind of pessimistic view I have is that, you know, I believe that the people
#
that are going to write papers that suggest that lockdown worked are people that had strong
#
priors that lockdowns were good ideas, and vice versa.
#
But my hope is that by seeing papers on both sides of that, we'd be able to get a better
#
conclusion about what is likely to be the case, although I still think I agree with
#
you that there's going to be a lot of uncertainty about what's going on, but hopefully we'll
#
be a little bit more informed.
#
I think the thing that is going to be a little bit more complicated, which you also raise,
#
which is this heterogeneity, I think that there are different impacts on different people.
#
That's an important component of this, and that's true both on the disease spreading
#
So we're finding out now that, you know, maybe different people have different proclivity
#
to spread the disease, either because they have more or less contacts than other people
#
But we're also finding those heterogeneous impacts in terms of what the economic consequences
#
So, you know, if you've got a good savings buffer, you're going to be less affected than
#
if you don't have as good savings buffer.
#
If you're a salary versus wage employee.
#
And when we think about the impacts of these things, we have to understand, we have to
#
appreciate that heterogeneity.
#
That's becoming an important part, kind of a mainstream idea in econometrics these days.
#
But we have to make sure that we do that in policymaking as well.
#
The reason why that's important is because it can lead to a situation where we have COVID
#
I'm not making a prediction that we're going to have it.
#
But if we have COVID 2.0, we have to appreciate that even after learning, it's possible that
#
different governments are going to do different things.
#
Which brings us back to your idea that the left and right are on opposite sides in India
#
and the United States, which, by the way, incidentally, nicely ties in with your idea
#
about political incentives at the very start.
#
Which is, you know, very early on, there's some hurting that occurs.
#
And the hurting doesn't necessarily always take you in favor of pro-lockdown or against
#
But there's some sort of hurting occurs for each of the different groups.
#
So the right in the US might have heard it to no lockdown and the right in India might
#
have heard it to a lockdown.
#
But then subsequent incentives are about sticking to your position.
#
And that kind of affects your incentives to acquire information.
#
In the future, I don't expect that to be any different.
#
I just expect it to be a little bit more informed.
#
I think the way that if I had my brothers or if I could, and I've been thinking increasingly
#
about this, is what do we do when COVID 2.0 occurs?
#
We don't want the same sort of hurting to make a decision.
#
Ideally, we'd like is the best possible information to inform the decision making.
#
There's going to be a political role in the sense that the politicians have to decide
#
what the balance is between economic activity and deaths.
#
That's a philosophical slash political issue.
#
I don't think that economists or social scientists necessarily have the answer to that.
#
But they should be able to say, given the way that you're making a trade-off, given
#
the way that a democratic society or government makes that trade-offs, what is the best way
#
My hope is that there will be less hurting the next time around.
#
I think that's what you see in East Asia now with COVID 1.0, given the SARS experience.
#
My hope is that you'll see that again in the future for all the countries of the world,
#
Let's sort of move on to the cost of the lockdown and how do we think moving forward?
#
You've done some very thought-provoking work on what you call adaptive controls, which
#
sort of goes against the notion that there is either a full lockdown or there is no lockdown.
#
You have a notion of adaptive controls where, number one, you argue that things should be
#
done with gradations, depending on how bad the situation is in different parts of the
#
country and different parts of cities, and two, you argue that this kind of approach
#
would actually be better than either approach, keeping the full lockdown or removing the
#
Can you elaborate a little bit on this?
#
Yes, we've been talking about this idea of a more gradual response to the crisis than
#
As I said before, I think that there is a strong argument for a lockdown at the beginning
#
of a crisis because you don't know what risk the disease poses and you need a little bit
#
of time to protect against the worst possible consequences, especially high deaths and high
#
death rates and gather information.
#
But once that's done, once you do that, you want to probably not use lockdowns as your
#
You want to do something that's a little bit more moderate.
#
The reason for that is just because there's a high cost to lockdowns.
#
You're reducing economic activity, and especially for the poor, that can have massive consequences.
#
You want to avoid those high costs of an extended lockdown.
#
What is a better approach?
#
I think a good analogy for what a better approach would be is the way that you drive a car.
#
When you drive a car, you have two goals.
#
I know you have many goals, but two primary goals.
#
One is you want to make sure that you're driving within the speed limit at a reasonable speed.
#
And the second thing is that you want to pay attention to obstacles, whether it's a pothole
#
in the road or a pedestrian running out, you want to be able to brake to avoid that.
#
But when you do that response, you drive moderately.
#
When you're just trying to maintain the speed, you don't fully press on the accelerator for
#
an extended period and you don't fully press on the brakes for an extended period.
#
You try to maintain a gentle adjustment of gas and brake to keep your speed target.
#
And then when you see an obstacle, that's when you brake, but you brake temporarily
#
and then you resume your course.
#
Now, the reason why it's a good analogy for the epidemic is that you kind of want to do
#
As you're going along with the epidemic, your goal is not to just immediately stomp out
#
What you want to do is you want to maintain the reproductive ratio, reproductive rate,
#
I should say, for the epidemic that's below one, so that it's gradually being driven from
#
And so I think of that as kind of your speed goal.
#
And then every so often, you're going to have these outbreaks and in response to outbreaks,
#
you need to contain them.
#
That is to say, you need to slam on the brakes.
#
And in this analogy, the way to think about it is that social distancing is like your
#
brakes, allowing economic activity as your accelerator.
#
And you want to avoid extremes, except in rare cases, you want to modulate back and
#
And so that's kind of the intuition behind adaptive control.
#
We just add a few other things for adaptive control to make sense.
#
The first one is that you have to choose your targets, so like the reproductive rate or
#
the trajectory of death or what your bed capacity is relative to hospitalizations.
#
You figure out what your trigger is, and then you gradually respond to that trigger.
#
That is to say, there's some mapping from what the data suggests COVID is doing to your
#
policies, but the policy response is gradual to avoid these extreme cases.
#
And the other thing that's important about this is keeping policy local.
#
Goes back to what you were saying about Kerala to some extent.
#
The disease is infectious, meaning they're going to be local clusters, unless you allow
#
international travel or something like that.
#
But even then, it's going to be kind of local clusters.
#
And so you'll want to cordon off areas based upon those clusters and not stop all activity
#
So you want to treat different areas differently.
#
And so the way that we've been saying it is, to the extent that you can draw cordons around
#
areas, say districts or blocks or even wards, if you have the capacity to do that, by that
#
I mean you can stop travel between the two areas, any two areas, then you should treat
#
those two areas differently and just monitor COVID in those areas.
#
And when COVID becomes severe, you tap on the brakes a little bit, have more social
#
And when COVID has subsided, you can allow more activity.
#
And you just go back and forth that way with the goal of keeping the reproductive rate,
#
for example, below one.
#
And over the course of a number of months, maybe six months, maybe nine months, you'll
#
be able to basically get back to the way things were before, in some places much more quickly,
#
But this is a nice way to balance economic activity with, and it's a nice principled
#
way to balance economic activity with infection control.
#
It's also really good when you don't have a lot of testing capacity, by the way, or
#
the ability to test, even if your labs could do all the tests, just that logistics, you
#
can't master them for logistical reasons or political reasons.
#
And the reason is because you will still observe outbreaks, people showing up at hospitals
#
And so even without testing, you will see those events.
#
And as long as you're doing things locally and can cordon off areas, you can keep the
#
outbreak from spreading.
#
That's another big benefit of adaptive control is that it accounts to some extent for inadequate
#
or suboptimal levels of testing.
#
So I have a follow-up question.
#
But before I get to that, let me kind of sum it up so you can tell me if I sort of got
#
Essentially, what you're pointing out is that, you know, at whatever point a lockdown
#
is completely removed, you're going to have a massive spike in deaths or a spike of a
#
And there's going to be an overall number that's going to be quite high, whether you
#
do it today or whether you do it a month off.
#
All you're doing with the lockdown is you're delaying that spike, but that spike is inevitable.
#
However, what you're saying is that what you can do is if you gradually release it,
#
you know, in a graded manner across different areas, depending on local conditions, then
#
you can control the overall number of deaths in the long run.
#
And this is the best way to do it.
#
Is that correct before I go on to my question?
#
And then the other part of our answer is to build the tools that allow each government
#
Now, what I'm sort of thinking of is one, you spoke about the triggers.
#
Now, it strikes me that whatever trigger you choose, let's say you choose a trigger of
#
the reproductive rate, which is how many people each infected person is passing it on to which
#
you want to be less than one or whatever, you know, depending on what the metric is,
#
you decide on what level of control and what level of social distancing you're going to
#
Now, you know, as we know, any metric that you come up with can easily be gamed.
#
And especially when you consider the incentives of politicians and you know, my listeners
#
keep joking about how I only talk about incentives, but I've got to get back to it in this context
#
because the incentives of politicians will be to keep that number down.
#
And there are many ways to do this, such as by not testing enough.
#
And in any case, we don't have adequate testing capacity to the best of my understanding,
#
even after, you know, a couple of months of lockdown.
#
And you know, the other part of this also is that at whatever, you know, different graded
#
levels that you do, it does one really have the state capacity to sort of enforce that.
#
And what that also requires is that people higher up the chain of power have to then
#
empower, you know, local authorities below them more than they would normally be comfortable
#
So it seems that there is firstly a political economy problem here.
#
And secondly, on the conceptual issue of, look, whatever metric you choose as a trigger,
#
But are there triggers that can't be gamed?
#
Are you looking at a combination of triggers like the amount of people who show up at hospitals
#
or just the overall death rate absent of cause?
#
You know, and so I'm just thinking aloud that when it comes to all these little practical
#
problems of who's going to enforce this stuff on the ground and who's going to take the
#
decisions, how do you think about that?
#
And it's a two part question.
#
One is, how do you think about that?
#
And two is, you actually have a lot of experience of working with governments and advising governments.
#
What's that experience been like?
#
So before I begin answering those two questions, let me just say that in my opinion, you can
#
never talk enough about incentives.
#
So you should just keep up with your approach to that.
#
I think incentives are critical to good policy making.
#
You have to make policy in light of the incentives, the private incentives that people have to
#
So let's now turn to these questions that you have.
#
So the first thing I'll say is when you set targets for adaptive control, it is super
#
important to think about setting targets that are actually meaningful in terms of the information
#
So for example, if you just set a reproductive rate target, as you said, reproductive rate
#
is crudely estimated by looking at the number of new cases minus the number of recovered
#
cases minus the number of deaths.
#
And you look at the trajectory of that from day to day.
#
That's typically how we calculate the reproductive rate without going into a lot of the details.
#
Obviously the confirmed cases that's manipulated by or affected by the testing rate, not only
#
your testing rate per capita, but also who you're testing.
#
You can test the people that are high risk or low risk to manipulate that number.
#
So you want to say, okay, well, I've got to be a little bit careful about this.
#
One approach is to change the trigger to something that's a little bit less easy to game.
#
So look at just deaths.
#
So deaths can be hidden, but it's harder to hide than confirmed COVID cases.
#
Obviously if there's COVID confirmed deaths, then obviously you still have the same testing
#
issue, but deaths might be less difficult to hide.
#
So you might say, I want a different trigger, which is deaths or trajectory of deaths.
#
Alternatively, you can decide that you're going to modify your trigger and say, it's
#
going to be a combination of your reproductive rate and your testing rate.
#
So if your testing rate is very low, then I'm going to impose more social distancing
#
independent of what the reproductive rate is.
#
Or if you do a high testing rate, then I'll focus on your reproductive rate.
#
And that way you give people the incentives to test.
#
Now obviously when you devise those rules, you need to think about two things.
#
A, it's not just the testing rate, but who you're deciding to test.
#
So testing policy is important.
#
This is something that people have been harping about since the beginning of the crisis.
#
It's important not just to test the people that show up at the hospital, which are people
#
that are already sick or test symptomatic people, which avoids the asymptomatic people,
#
but also to test the community.
#
So you're getting both symptomatic and asymptomatic and not people that are selected based upon
#
the severity of their illness.
#
So that's one approach that you can take, which is focus on the quality of the testing.
#
The second approach you can take is just be more sophisticated in the way that you do
#
your analysis of the data.
#
So this is another thing that we're thinking about.
#
We don't have a lot of community tests in India, virtually none, frankly, from what
#
But we have a ton of data on tests already done and features about those tests and characteristic
#
They gather actually a lot of data on whatever it is that they do.
#
If you can get access to those data, that is to say, either academics or researchers
#
getting access to that data, or if you're a statistician sitting in the government,
#
the question is, what can you do with even that selected data?
#
If you know what the testing strategy is, we think that there might be ways to interpret
#
And so that's another approach that you can take.
#
You can either modify your trigger to have better testing that's higher quality testing,
#
or you could do more fancier statistical footwork to try to make the best of whatever data is
#
So even if people try to keep the testing rate low or select, maybe you can correct
#
So that's the second way that you want to do.
#
But you want to adjust your trigger to be robust to the data that you get and kind of
#
issues with that data, including incentives to gather that data.
#
So that's the way it works with triggers.
#
Now, the problem with that approach obviously is that's great in an ideal sense, setting
#
good triggers, but the same people that are going to make a decision about lockdown are
#
going to make a decision about adaptive control.
#
And so if they have an incentive not to get accurate information, they're going to have
#
an incentive to make sure that their adaptive control triggers are also not getting accurate
#
So that, I think, is the challenge with implementing adaptive control.
#
And we're thinking about how to actually do that.
#
And that's a nice segue into your first question, as it turns out, sorry for such a long prelude,
#
but it's about decision making within the government.
#
And you're right, I've spent some years working with different governments, either in a kind
#
of adjacent to the research that I do, which is often on trying to understand the impact
#
of different government programs, like Rastya Swasta, Bhima Yojana, or other programs like
#
Mission Kakthiya, trying to work with the government to evaluate those, you learn about
#
what the government's decision making processes are, and you then can kind of get a sense
#
of they often ask you for help, and you can get a sense of how they're making decisions
#
and what they care about.
#
And the other approach is through the, which you'd mentioned earlier, I have this program
#
called the International Innovation Core that sends people, graduates of Indian universities
#
or people that are working as consultants or programmers or engineers in India to go
#
work not just in the private sector, but to go work for the government on some government
#
program for one to three years to help implement it.
#
We think that's a valuable tool for them to see how the government functions and to
#
improve, you know, provide support to the government.
#
As part of that, we also get some purview into decision making, not on the research
#
side, but the implementation side.
#
And in that process, I've learned a few things.
#
The first is there are people that have, you know, people certainly respond to incentives,
#
political incentives, but at the same time, there are a lot of people in the government
#
who really are just trying to make Indian society better.
#
For those people, I think it's better to think not in terms of politics as being the aim,
#
but politics being a constraint.
#
And your goal should be, how do I help them make policy subject to political constraints?
#
Okay, not seeking the first best, but looking for the second best.
#
And that's both in the short run and in the long run, setting up structural features of
#
decision making that enable them to have more latitude, more relaxed political constraints
#
And a lot of that is a function of, you know, building relationships over time.
#
You help governments for a little bit.
#
You build individual relationships with particular bureaucrats, IS officers and state officers.
#
And then through that relationship, you build the trust that gets those people to listen
#
to you a little bit more.
#
So that's one approach that you can take.
#
And the other thing is to broadly help build an ecosystem where there are more officers
#
And that's making, for example, investment in the training of bureaucrats.
#
It's fostering both a dialogue in academia and in the media where it's not immediately
#
critical of the government, but it's more understanding of what government constraints
#
So there can be a level of trust between bureaucrats more generally and people like us that are
#
trying to discuss, critique, inform policy from the outside.
#
And so I think that's another important lesson that I learned.
#
This is a long process.
#
I think we all have a role to play, but it's important to remember that government bureaucrats
#
have constraints and we need to account for those constraints.
#
So I think that's my answer to your decision making.
#
You're going to have to remind me of your second question, though.
#
This was my second question.
#
You answered the first question during what you called your prelude to answering this
#
But just a brief follow up on this, you are also talking with the governments now and
#
So how receptive have they been to your ideas of adaptive controls?
#
Do they come to you with preconceived notions of what they already want and they want backup
#
Are they also desperately trying to figure out what is the right thing to do and find
#
Usually when I have these conversations with governments, I don't come in knowing exactly
#
Well, I always have a notion of what to do, but I'm confident that I don't know the full
#
And the reason is because very often, in almost every conversation, I get new pieces of data
#
from the government, whether it's facts on the ground shifted or whether it's the political
#
constraints that are in operation.
#
By the way, political constraints are real reasons.
#
I think it's naive to just ignore those issues.
#
They are real constraints and they sometimes have rational basis as well.
#
I mean, we're a democratic society and so this is a side effect of a democratic society.
#
So I find that what typically happens is I have some notion, they give me some data.
#
I try in real time to see how my answer has to adapt to those realities that they're telling
#
And if I can answer it right away, that's great.
#
Otherwise, what I do is I say, I need to think about this, give me a day to write you a one-pager
#
in response, which is very often my answer and the key is just providing that back very
#
quickly so that we can set up for the next conversation.
#
But yeah, that's typically the dynamic that we have.
#
And I have to be honest, there's heterogeneity, but I will tell you there's also selection.
#
What do I mean by that?
#
I mean that some bureaucrats are more open to outside advice.
#
There are obviously differences in how much they are responding to private incentives
#
versus social incentives to make Indian society better.
#
I will not deny that, but I also have to acknowledge that the folks that are willing to talk to
#
me are probably going to be more on the side of being responsive, being open to these ideas
#
And so I understand that I'm seeing a selective population, but based upon that selective
#
population, I'm actually, I think more optimistic than a lot of my close friends are in this
#
context even though I know that it's a selective population.
#
Yeah, yeah, it makes a lot of sense and in fact, you seem very hopeful and optimistic,
#
which is, you know, maybe it's, you know, a result of my having just lived in India
#
and lived in Bombay all this time that I sort of have this very dark negative view of government
#
and the way incentives function, but I'm hoping to learn that I'm wrong in some way or the
#
Let's turn to the economy now.
#
I mean, there's no question that there has been a devastation, you know, whether it is
#
because of the watching of the lockdown or whether it is ultimately because of COVID,
#
which caused the lockdown to happen.
#
Those are sort of separate debates and, you know, they are intertwined.
#
But one very interesting aspect of this devastation is what has happened to migrant workers where,
#
you know, tens of thousands of migrant workers have left the cities and gone back, almost
#
in a process of de-urbanization.
#
And what is especially tragic about it, what is heartbreaking about it is that many of
#
these people from the reports that we hear have completely lost faith.
#
You know, there will be people, for example, when the Bangalore lockdown was being lifted
#
and there were migrant workers on the road who wanted to walk back to UP.
#
And they were asked, you know, why do you want to walk back?
#
The lockdowns lifted, your job will be back.
#
And they said, no, no, we want to die at home.
#
We've had enough of this.
#
And there has been that deep loss of trust.
#
And the dehumanization that has taken place has possibly driven many of them away from
#
the cities permanently, in the sense that you hear people saying that no matter what
#
happens, it's not about a livelihood anymore.
#
We will not go back there.
#
Now, this is, of course, just so disturbing at so many levels, at the humanitarian level,
#
And the deeper repercussions are that urbanization has always been an engine of growth.
#
You know, and there was a sense that India is urbanizing rapidly and it's a good thing.
#
And people are themselves making the choice to go to cities.
#
And that's been the overwhelming sort of tide of movement from rural areas to cities where
#
they are part of these large economic networks where, you know, they can contribute so much
#
more and get so much more back.
#
And now, perhaps, I mean, I don't even know if it's the first time in history that it's
#
happening with, at least in India for this short term in the last couple of months, a
#
movement has been the other way.
#
And it seems like it might not be a temporary thing.
#
What are your feelings on this?
#
Like, is it going to have long term consequences?
#
And how should we deal with that?
#
How should we approach this?
#
So, I think that's a critical topic.
#
But before I turn to that, I want to go back to your last question just a little bit, actually,
#
your prelude to this question where you said that I'm more optimistic than you are.
#
And that's because you're sitting in India.
#
I think that's a really important point, actually.
#
And if you will indulge me, I want to go back and give you a little bit of history for myself
#
to understand why I'm at where I am, and I think why you are where you are, and what's
#
So, I'm more optimistic.
#
But I wasn't always more optimistic.
#
When I was in college, one of the things that I wanted to do, you know, when I was growing
#
up, I was always, we'd only went on vacation to one place, which is India.
#
Every two years, we'd go on a month long vacation to India, visit our relatives, and
#
we wouldn't have a vacation in between.
#
Then, you know, when I went off to college, I wasn't part of that process as much as
#
I couldn't take a month.
#
But I decided to spend a summer working in India for Mother Teresa at Ashadan in Mumbai.
#
And during that time, I got to see poverty up close and personal in a way that I didn't
#
even get to see when we were visiting family in India.
#
We're not all prosperous in India, but not to the extent that I saw when I was in Mumbai
#
And I thought that it was very important to work in India and that you could do a lot
#
more good there than hanging out in the United States.
#
But at the same time, I had an experience with just regular bureaucracy in India, whether
#
it's taking money out of the bank, whether it's navigating traffic, or whether it's
#
navigating the medical system for some of the people that we saw there.
#
This is also in the middle of the AIDS epidemic in India too.
#
So I saw a little bit of insight into how we were treating HIV and AIDS patients.
#
And I became hugely pessimistic, hugely pessimistic, and then spent almost the next 15 years avoiding
#
India, focusing on the United States, maybe 20 years, 15.
#
And it was only that distance that allowed me to forget, become less cynical, and re-embrace
#
Now, I still had that history, so I knew that it existed.
#
And so I approached in a different way, which is I just lowered my expectations, and I forced
#
myself to reassess what I should expect of India.
#
It's also the case that I also stared at GDP growth rates in India, and per capita GDP,
#
and saw that India had made remarkable progress from, say, 85 to even 2005 or 2010.
#
And that there was something significant going on in India, so there was a lot of potential
#
I think those two things brought me back and made me much more optimistic.
#
Really it's idiosyncratic.
#
I had interaction with a number of positive, with bureaucrats that really inspired me in
#
Karnataka and elsewhere that kept me on this track.
#
Obviously, I get disappointed as well, because I think my expectations are a little bit different.
#
And being able to go back and forth from India to the United States helps renew that optimism.
#
And one of the things I like to tell folks that are, you know, my colleagues in India
#
that are a little bit more pessimistic is to understand that, you know, if you got deeply
#
involved in American politics, you too would become very cynical.
#
And so India is not always all that bad, and it's good to sometimes step away.
#
So anyway, that's my kind of brief repost to the differences between us.
#
So I would say, Amit, next time, come on out to Chicago, I'll introduce you to some Chicago
#
politicians and then, you know, take a few weeks off.
#
We'll be a little more closely aligned.
#
So, but anyway, back to migrants.
#
So yeah, I share your concerns.
#
I think I believe that urbanization is key driver growth in India, and that, you know,
#
we see very few societies that are high income societies or high fraction of their population
#
is still engaging in agricultural labor.
#
It is unlikely to be the case that if India wants to be high income, that will continue.
#
And that implies that people not only move out of agriculture, but they're very likely
#
move into cities, which are responsible for disproportionate amount of India's growth.
#
And I was very optimistic about that prior to COVID.
#
And you know, my big concern was why don't more people move in?
#
How do we make cities more accommodating?
#
How do we build infrastructure in cities and things like that?
#
COVID throws a wrench in all that.
#
I think the lockdown in cities was hugely painful to not just migrant income, but broadly
#
the poor in cities, especially the interruption of essential services.
#
I also think that that in combination with the higher case load per capita and higher
#
death toll per capita in cities, make cities less hospitable.
#
And the way that I think about it is, you know, I think of migrants as kind of the marginal
#
That is to say, they're the last people in and they're the ones that get the least benefit,
#
but they are getting a benefit.
#
So if you reduce the value of cities, those are the first people to leave because they're
#
the ones that are made, you know, they go from being, you know, seeing cities as a small
#
positive benefit to seeing cities as a negative as compared to people at higher up on the
#
So it's not surprising to me that you're seeing these numbers of 7.5 million or 11 million
#
people leaving cities to go back to their home state or to rural areas in the same state.
#
Now the real question is, is this a good thing or a bad thing?
#
And what's likely to happen in the future and what can we do about it?
#
I think that in the short run, you know, I can see why people leave and why they're going
#
The real challenge though is just, you know, how well will the states like UP, Bihar, etc.
#
absorb these people both in terms of COVID, are they able to effectively engage in quarantine
#
of the high-risk individuals so that the, so that COVID doesn't spread in rural areas
#
I think that's one of the big risks that we have right now with the return, very short-term
#
risk from the migration.
#
It's very much like the studies that it reminds me of is the early study done in the United
#
States showing that before the lockdown, when COVID, we now know in hindsight, was really
#
affecting New York City, before the lockdown happened there, people fled New York City
#
to other parts of the US.
#
And now we realize that other parts of the US that COVID that you see there can be, genomic
#
analysis suggests that you can trace its ancestry back to the New York strains, meaning that
#
the flight from New York is really what spread COVID around the country.
#
And so I'm a little bit worried that that's going to happen now, the people in cities
#
who are infected are going to go to rural areas that have largely avoided infection
#
And now you're going to see small outbreaks in rural areas, complicating control.
#
So that's a very short-term concern I have, and the question is how states are able to
#
handle that and what kind of, can they do humane quarantine to mitigate those effects?
#
So that's the immediate consequence.
#
And then of course, the next consequence is, you know, what is the impact?
#
Not only are you going to see growth in cities, a big driver of growth in India, but you're
#
also going to see overwhelmed systems, perhaps in these states that weren't expecting this
#
You know, just to give you a sense, we're working with Bihar, Bihar is a population
#
of 100 million roughly, and they're expecting to get two to three million back.
#
So two to three percent of their population is coming back, which is a large number.
#
It seems small, but from a social organization perspective, that's quite a large influx.
#
So that's going to be a big thing.
#
Now the question is, what is the consequence going forward?
#
Like, are people going to, you know, are they, is this just a temporary return to get some
#
Remember, it's still the height of the summer.
#
There's still agriculture going on, so there's some rural demand.
#
Maybe once cities, you know, control the infection, lockdown, it gets lifted in the big cities,
#
then maybe these people will immediately return, in which case we're back to the same path
#
Obviously, there's going to be a short-term economic crunch, especially on the poor, but
#
hopefully we'll be able to return in a year.
#
The alternative view is that people have just revised their views of the value of cities.
#
If their belief is now that epidemics are possible and are going to come again, then
#
it's going to take a little bit more for them to move to cities, which means you're going
#
to see a slowing, a slowdown in the rate of urbanization in India.
#
First an immediate drop, and then a slowing down of the return, which is hugely problematic
#
from a perspective, from the perspective of India's economic development.
#
Now the question is, what do we do about that?
#
And I can think of two things, or at least I'm thinking along two lines.
#
The first is, for a lot of the states that are getting migrants back, if they can look
#
beyond the short-term crunch, they might see an opportunity.
#
And what I mean by opportunity is, you know, if you're a state like Bihar or UP and you're
#
getting a lot of migrants back, historically, you're a state that has economically lagged
#
And one of the reasons we think is because there is not as much urbanization in your
#
You see a lot of small towns, but you don't see big cities.
#
Big cities are much more common in the more developed states, especially in the south.
#
And so the question is, can you figure out a way to get your migrants, instead of in
#
sometime in the future returning to Mumbai, Delhi, Chennai, et cetera, can you get them
#
to go to cities in your state?
#
So this is a great opportunity for you to urbanize quickly, provide a path for your
#
migrants to go to local cities and grow those cities.
#
Take your tier three cities and make them tier two cities or better.
#
So that's, in some sense, an opportunity.
#
That's what I encourage states in the northern belt to do with their returning migrants,
#
see that as an opportunity.
#
The second thing is that, you know, when we think more broadly about cities, this is the
#
chance for us to kind of reassess how to make cities better, especially for the poor.
#
And the fact that a lot of migrants have actually left gives you an opportunity to really restructure
#
You know, the first and foremost is you need better infrastructure.
#
Your cities are growing faster than your underlying street and sanitation infrastructure, and
#
you really need to kind of invest in that.
#
That's the first thing.
#
To do that, obviously, you'll need more local control and local financing that's a little
#
That's going to have to be part of the answer too, but really build that infrastructure.
#
The second approach is to become a little bit more hospitable to migrants, to labor
#
And we're going to see that it's very hard to operate a city when a large fraction of
#
your drivers, security guards, household help, et cetera, construction workers disappear.
#
It's going to be hard to restart the city.
#
You're going to appreciate that that labor is essential.
#
Once you have that appreciation, it's going to be important to do things to make life
#
easier for those people.
#
That means, for example, you know, one of the big issues with slums is that there's
#
the reason why we have slums is because we have restrictions on the growth of housing
#
supply in cities, housing regulations, FSI regulations, things like that.
#
And if we could relax some of those, that would be great.
#
Another restriction is just so much land that's owned by the state.
#
If the state can start handing over land rights to slum dwellers, that'll entice people to
#
come back and feel a little bit more secure living in cities.
#
I think that's the second thing.
#
A third issue, maybe this should have been the first issue, is to start really investing
#
in public health, meaning allow the private sector in the healthcare system to deal with
#
regular transactions for health, but have the public infrastructure.
#
The public hospitals really focus on public health.
#
So things like infectious disease and trying to help poor individuals make sure they have
#
Having that kind of renewed focus, narrow focus would be good.
#
And cities really need to be the start of that if they want to get these migrants back.
#
So those are the three things that I would think would be really important.
#
I just want to stress again, one really, really important component of that second part is
#
that not only do you provide housing to these individuals, to low income people, but you
#
get rid of a lot of the required occupational licensing and other regulations that make it
#
really, really hard for people to work in cities, whether it's as a shoeshine or starting
#
your own small pawn shop or something like that.
#
You've got to make it easier for people to economically thrive in cities without all
#
COVID is an opportunity to revisit that.
#
We wouldn't have been able to do that in peacetime because of a lot of vested interests.
#
But now the vested interests are kind of back on their heels.
#
And so while that's happening, this is a real opportunity to make life a little bit easier
#
for poor individuals, poor people in cities.
#
Those are great and insightful points.
#
I'll quickly take the opportunity to plug episodes that I have done in the past.
#
Some of these subjects, they'll be linked from the show notes.
#
I've done episodes on FSI and rent control with Alex Tabarrok.
#
I've done an episode on slums with Pawan Srinath where he spoke about how slums actually play
#
an essential role because they're the entry point of migrants from outside the city to
#
And the solution of that, of course, as you pointed out, is increase the housing supply,
#
changing our regulations around FSI is a great way of doing that.
#
I have also recorded an episode with Prashant Narang on this kind of licensing within cities,
#
which is such a problem.
#
But that'll release after this one, perhaps in a couple of weeks.
#
I've taken more than a couple of hours of your time, so I'm sort of now going to ask
#
you to sort of look into the economist crystal ball and sort of look ahead and tell me what
#
are, you know, if you look five years into the future, what is sort of a best case scenario
#
and what is a worst case scenario?
#
In other words, what gives you hope and what gives you despair?
#
Like I've heard it said, like someone framed it in a way on Twitter, the optimistic side
#
of things, which really struck me where I forget who this was, but this person basically
#
said that COVID is like a vaccine, which will give us the antibodies to fight future pandemics,
#
which might otherwise be worse.
#
And I guess that's a very optimistic view.
#
But what gives you reason for hope and despair?
#
Okay, so I'll answer both parts.
#
The first is I want to kind of give you worst case, best case.
#
So the worst case scenario is I think we already touched upon, partly.
#
I think that there's a risk that you're seeing de-urbanization that's going to last for longer
#
And that's going to slow growth structurally in India.
#
Another thing that I think is very important is I fear that we're in for a rude macro awakening.
#
So let me explain what I mean by that.
#
So when we think about what COVID does, you should think about whether or not you think
#
that there's a lockdown or not.
#
There's an aggregate supply effect and aggregate demand effect.
#
The aggregate supply effect comes first, either voluntary social distancing or mandatory lockdowns
#
restrict your ability to work.
#
That reduces the amount of economic activity and in general will tend to raise prices.
#
But then there's a second effect that goes on, offsets that a little bit, worsens part
#
of it, which is that because there's less supply, there's less income that people have.
#
They can't work, so their incomes fall.
#
So that reduces demand.
#
So you have both the aggregate supply effect followed by an aggregate demand effect.
#
The net result is a massive reduction in the amount of economic activity, although an ambiguous
#
Now governments have responded outside of the public health context by typically trying
#
to address what they can, and that is the aggregate demand effect.
#
The way they do that is they increase supports to people.
#
So in the United States, we gave out a bunch of checks to people to help supplement their
#
And in India, you do some of that as well, although to a much lesser extent.
#
And the idea here is basically do what we can.
#
We can't address the aggregate supply issue because that's necessary for the or uncontrollable
#
because of the COVID crisis.
#
But the demand side, we can give people income.
#
Now the question is where governments got that money from in the first place.
#
Now ordinarily, if we were to think about just one country having, suffering a shock
#
like this, they would get a loan from the World Bank or from other countries going on
#
to the sovereign debt market to get money.
#
But here we have a situation where the entire world is wanting to spend more.
#
Now there are some countries that had the surplus to do so, China, for example.
#
But most countries didn't have that surplus.
#
They were already operating with massive amounts of debt.
#
And so in a world where there's more debt and everybody wants more money, there really
#
is only one way that you give stimulus checks.
#
It's by printing money.
#
Now we may mask it in many different ways, but what we're doing is we're printing money.
#
Now anybody that knows how inflation works, like you use the quantity theory of money,
#
you have a standard equation, Fisher's equation that tells you what the price levels are.
#
And what it says is that the total supply of money is equal to the quantity of money
#
that's available times the number of transactions that you have with money per day.
#
And that gives you the daily supply.
#
So not only is how many rupees there are, but how many times a given rupee circulates
#
in a day tells you what the total supply of rupees is in that day.
#
And in COVID, you have the following situation.
#
Because you have this aggregate supply effect, people are not able to buy and sell goods.
#
Velocity of money has fallen.
#
Now you can print more money and you're increasing the money supply, but you're not going to
#
have a big effect on prices because again, velocity is down.
#
So if total supply is the amount printed times the velocity and velocity has gone down, you
#
can print more and kind of keep prices stable.
#
So now imagine what happens if over the next few months you relax the constraint on supply
#
so that velocity rises.
#
So now you've printed a whole bunch of money and now you're going to relax constraints
#
Velocity is going to rise.
#
You're going to see a massive increase in effective money supply later on.
#
Not right now, later on as velocity rises.
#
And that's going to trigger a binge of inflation.
#
And you're going to have all the issues that come up with inflation that we've seen in
#
the past and negative consequences, except this is going to be a weird inflation crisis.
#
Because the inflation crisis where you've got this stock of money that's finally being
#
released to circulate in the society and we're going to have to figure out how to solve that
#
I'm not going to tell you that I know the answer to that, but one of the kind of the
#
worst case scenarios that I worry about is a global hyperinflation problem or a global
#
If we were 50 years ahead and everything was cryptocurrency or electronic money, the way
#
you'd solve it is you'd proportionally reduce everybody's money stock just the same way
#
you would inflate, you would deflate, but we're not in that position.
#
There's still money that's there and you have to figure out a way to kind of pull money
#
out of that system in order to control that inflation.
#
So I worry a little bit about this.
#
My prediction, if I were to make one, is that that's one of the causes of the worst case
#
scenario and that's something that macroeconomists will be thinking about that they don't have
#
an answer to right now.
#
Yeah, and it could hurt the poor the most because inflation first and foremost has attacks
#
So after what they have already been through, the vicious cycle just continues.
#
So let's not dwell on only the negative.
#
Let's try the positive because I think on net, I'm positive.
#
To me, that stuff just tells me that this effect, it's not just the health component
#
that's going to take a few years to resolve.
#
It's going to be the economic component that's going to take a few years to resolve.
#
But what I'm hoping is that the positive outweigh the benefits.
#
This is consistent with my general optimistic view about the world that I've learned about
#
So the first is the issue that you pointed out, which is maybe we'll learn to handle
#
future epidemics better.
#
I'm optimistic that that'll be the case.
#
All of us have a role here to play, which is to the whole government's accountable to
#
doing that preparation going forward.
#
So we should keep the pressure on for that and provide the assistance that we can.
#
The second thing is I'm hoping that we are able to take this kind of de-urbanization
#
as an opportunity to renew and strengthen our cities, and I'm optimistic that we'll
#
You know, what's amazing to me is that almost everybody I talk to, and I understand that
#
that's a selected group, but almost everybody I talk to with very few exceptions generally
#
has this view that cities are a driver of growth in India, that we need to do better
#
by migrant workers and slums.
#
And we often agree on what the policies are.
#
And the real issue is how do we get the government to engage in those policies?
#
To me, that's really a problem of trying to understand what vested interests stand in
#
the way and how do you come up with good second best policies to buy off those vested interests
#
so you get better policies.
#
That to me is the challenge going forward, but I'm optimistic that we'll be able to
#
challenge it to address it with urbanization in India as well.
#
The more complicated thing I think is really, you know, what do you do if there are structural
#
features of the political system in India that make it so that it cannot achieve, for
#
example, the trajectory that, say, China was on pre-COVID?
#
So this has to do with issues of political competition leading to corruption in ways
#
that China is able to control better than we are able to control.
#
This is the sort of things that people like James Robinson or that Chang Tai Shea, both
#
colleagues of mine at the University of Chicago, debate and that I think are an important debate
#
for us to have in India.
#
So in that sense, I am optimistic, but I think that we have to keep our eye on the ball going
#
I hope to see that, you know, in 10 years, India is just on a better path, especially
#
if we're able to address a lot of the issues that you've raised in prior podcasts and use
#
this opportunity to make those changes.
#
Anup, thanks so much for being so generous with your time and insights.
#
You've given me a lot and I'm sure you've given our listeners a lot to process in the
#
course of this conversation.
#
So thanks a lot for coming on the scene and the unseen.
#
Thank you for the conversation.
#
If you enjoyed listening to the show, do check out the show notes, which has many relevant
#
You can follow Anup on Twitter at Anup underscore Malani.
#
You can follow me at Amit Verma, A-M-I-T-V-A-R-M-A.
#
You can browse past episodes of the scene and the unseen at scene unseen dot I-N.
#
Thank you for listening.
#
Did you enjoy this episode of the scene and the unseen?
#
If so, would you like to support the production of the show?
#
You can go over to scene unseen dot I-N slash support and contribute any amount you like
#
to keep this podcast alive and kicking.
