Re:Thinking with Adam Grant
November 9, 2021
Esther Duflo wants you to think like a plumber
Hey WorkLifers, it’s Adam Grant. Welcome back to Taken for Granted, my podcast with the TED Audio Collective. I’m an organizational psychologist, and this series is about rethinking assumptions we often take for granted about how we work, lead, and live.
Today’s guest is Esther Duflo, the MIT economist who won a Nobel Prize in 2019 along with two other economists-- MIchael Kremer and Abhijit Banerjee, her husband.
Esther was tenured in her twenties and has done groundbreaking experiments in the developing world to fight poverty. She’s the co-founder and co-director of the Abdul Latif Jameel Poverty Action Lab (J-PAL), a prolific writer and widely cited scholar, and has served on the Global Development Council at the White House.
Nice to meet you.
How did you end up becoming so passionate about studying the developing world?
So I think it started relatively early. I grew up in a very, uh, sheltered, intellectual middle-class life with my mother being a doctor and my father being a professor. But the one sort of a wrinkle to that, or a slight difference to that is my mother was a, and still is very active in, uh, organizations of doctors dealing with kids, victims of war.
And therefore she spent quite a bit of time in various countries, starting with Moroccan Sahara, and then in El Salvador, and in places that were dealing with various kinds of crisis—usually man made. And when she came back from the trips, she, you, she always organized a little slideshow for us.
We had these slideshows that looked like, you know, actual little squares, if you remember those probably not.
Oh, of course I do.
And we would watch them and she would explain to us what was happening and we would, um, and then she would also say, “This is your contribution to, to these kids that have a life that is so much less good than yours.”
And frankly, I saw this was pretty minimal as a contribution. And from then on, I kept sort of comparing my life to the life of the kids that I saw in these pictures, and thinking that it gave me a sort of a responsibility to do something about it. what is something wasn’t very informed and clear until much later.
So what happened that led you to get excited about economics as a tool for tackling poverty? Because it's not the first place that I would have looked.
No, it certainly isn't. So I had this idea that one day I would make a difference. And for some reason I had to holdout being a doctor, um, maybe related to seeing my mothers lifestyle. And then I, but I thought maybe I would do politics, but all of this was a little bit vague. So what really, what I was doing as being a good students and on the, for track to become an academic, I liked history.
So maybe I'll be a historian and, um, or the way a little bit forgetting or living to some later day, this ambition to perhaps do something about changing the world. And then I discovered economics by a chance by complete chance. I thought it was unreasonable to study just history and I needed to have some compliment to it simply because I had always been fascinated by history and I thought it was very dangerous to be guided by your eight year old self for your career choices.
So I figured, let me do something else too, in case this whole history thing doesn't work out. And I picked economics number one, because it, it seemed compatible with history and number two, because the professor was charismatic and interesting. But then when I started studying it, I hated it. I thought it was completely irrelevant, that it was a lame effort to put some equations on to describe people's behavior in a way that nobody believed, least of all the people who were trying to teach it, that the math was not interesting, that it was not describing anything.
So it was very clear to me after one year of economics, that that was it. But then I was left in an uncomfortable position because by that time I was living the gilded life of the elite French student paid to study in a very intellectually ratified environment. You know, I had everything to be happy and I felt very uncomfortable being so comfortable saying like I'm so far from what I wanted to do, like I'm not changing anybody's life and what I'm doing - if I become a historian, I will not change anybody's life that way. So I thought I had to make some sort of a break and I left for Russia for a year, my school had an exchange program, And there in Russia, I started doing gig work for economists, uh, translations, running around doing this, that, and the other.
And there, I realized that economists have so much power, I mean, academic economists, because they are listened to by policy makers. And it was like a mixture of admiration and horror because this was Russia at the, in the transition and the economists there were really defining the cost of this country, by the way, they did privatization, shock therapy and the like based on, I felt, very little knowledge. And so this really gave me my cost. I said, well, on the one hand, I'm going to try to acquire a little bit more knowledge on what I'm going to talk about, but on the other hand, and someone will probably be ready to listen to me. And that's really when I decided to come back. I also happened to meet Thomas Piketty in Russia, who, who was teaching at MIT at the time.
And he told me that at MIT, I would learn a much more practical economics than what I had learned in France. So all of this clicked nicely. I went to MIT, took my first development economics class and realized, ah, this is how I'm going to go back to, you know, what had been my mission from the get-go.
Wow. Okay. There's so many things I already want to follow up on there. The first one is you said it's dangerous to follow the career advice of your, or pursue the career interests of your eight year old self. So I had a similar experience with my, I guess my early exposure to economics. When I was a sophomore in college, I had declared a joint major between psychology and economics, and I knew I wanted to understand human behavior. And I thought the two worlds would be great to try to marry. And then I showed up in my first economics class and the professor wrote on the board that quality of life was a function of consumption and savings. I was like, that is the dumbest equation I've ever seen. My happiness has very little to do with how much I consume and how much I save, like do economists understand at all what people care about and what drives the quality of our lives?
And I'm wondering if I gave up on economics too soon. Do you think I should have rethought my distaste for those equations?
Yeah, I think you should have, but I understand why you didn't. And I think many, many years students have exactly the same experience. I had the same experience and I had given up. It's only because I needed to make a little bit of money in Russia, and because I thought it would be a way of being closed off to what was happening, that I got back in the proximity of economists.
Many people stop because that's what they are exposed to. Or maybe if they go to a second lesson and they start learning that a country's worst is related to the GDP or, or some such. And then we you know, people rightly concluded it's not for them. And I think that's, that's an error of… it's not even an error of the economics profession. It's an error of the way economics is taught, because if you look at what economists do, it's so varied and it's so diverse. And you know, there are still people who, who write down those equations, but there are many people who do other things like the behavioral economics you would perhaps have done if you had continued in the field.
Uh, but you have to have completed and survived two years of dispersant to get to it. And, most people don't. And that's why now this year, for example, at MIT, I teach two classes that, uh, are targeted to Freshmen. Right. I want to get them right out of high school before they’ve done any economics to teach them how you use economics to combat hunger or low education or to vaccinate people against COVID.
And I also want to teach them how we use economics to think about racism or climate change or that kind of thing. And we know, granted it's not going to be as rigorous and perfect as it could be, but at least it's perhaps going to give them a taste of the end of the journey, and give them a reason to undertake the journey in the first place.
They are clearly very lucky to have you doing that. And I'm sure they're excited to learn from a Nobel Laureate too. I think your arc is very different from the way that I normally envision a career. Right? So in, you know, in my case, uh, I, I was sort of, you know, I, I guess I wasn't quite disgusted with economics, but I was disappointed with what I learned and how oversimplified it seemed. And by contrast, I had fallen in love with psychology and I knew almost on first sight that I wanted to be an organizational psychologist. It seemed like you had a different trajectory that you had to fall out of hate with economics before you fell in love with it.
How did that happen? I mean, at some level obviously you were paying attention because economists had power, but then what shifted? What actually lit the spark or, or at least got you to say, maybe this isn't as terrible as I thought?
It's really getting to the, to this end of the journey when I was in Russia and I saw the, what the economist were doing, I could, I could picture myself doing something, not similar because I didn't like the type of macro-scale advice they were giving, but at least related in that I would put my science to the service of something useful for people, which is what these economists in Russia thought they were doing.
So after this one year of hating economics, when I came back to economics, I was already a graduate student and the easy math was replaced by reasonably hard math and the complete abstraction by some effort at looking at the world. and therefore, you know, armed on the one hand, uh, was this sense that this was all, there was a plan. There was a logic. I knew where I was going and I was going to do work that was ultimately interesting to me, but I had to learn, you know, to do my skills, uh. And on the other end with the fact that quickly enough, the skills, you know, turn into RPGs and things that were intellectually motivated, uh, gave me the, the, all the time and energy I needed to fall in love again to the subject. I had fallen out of hate relatively recently.
I think it's, it's so interesting then to see that not only in some ways did it take you a while to decide that economics were for you, but that when you finally did make that commitment, you took on a field that almost didn't exist yet. Right? As far as I know, when you started doing development economics, that was not a real option for doctoral students. Did people think you were insane?
It was an option, but it was a very, very, very, very marginal option. Now where I was very lucky is that by the time I got to MIT, I already knew I was going to do development economics. So if nobody had done development economists, I would have done it anyway because I was going back like returning to, uh, to, to my plan from the beginning.
But in fact, when I arrived, Michael Kremer and Abhijit Banerjee were teaching and they had done that leap of sort of going out of their well established field, uh, macro for Michael and, uh, theory for Abhijit and decided that they were going to teach development, which they were making up as they went along and they were making up with us.
And it was about, I don't know, five or six of us students in the room. And it's also exhilarating to see a field being created in front of your eyes and with your own inputs.
And yeah, it was a bit, you know, people thought it was insane in the sense that they were like, you're a good student. You know, you could just get a great job if only you did the normal field. And I thought, well, that's not, I didn't come here to get a great job. I came here to study development economics, and it was not even a question for me because I knew why I was there and I was going to do it regardless.
So if we fast forward, I don't know, couple decades, you wrote a book about radically rethinking poverty, and I think it's, it's one, it's a wonderful summary of so much of the research you've done, but it's also a perfect fit for the show because this show is about rethinking assumptions that we take for granted.
So I'd love just for starters, if you could talk a little bit about what common assumptions about poverty are overturned by your work.
So I think, depending on the flavor of the month, there are any number of wrong assumptions about poverty. And what's interesting is that they are very strongly held at the moment in time. And then they are kind of you know, next year, it's another one. So the one when I grew up and that I grew up with, and in fact, I grew up believing when I was, you know, uh, naively thinking, I was going to save the world one day, was that the poor were, uh, all desperate, and eh, they were all on the margin of starvation. And, you know, I grew up with it and with the French version of that, uh, and my mom's pictures from the desert and kids with distended bellies, and, uh, very, very, uh, howing lives and very little choices.
So that's one. Of course, there are people who live like that. That's for sure. Usually that's caused by war. And fortunately that's relatively a small fraction of the world. Uh, but many people live in poverty and live very different lives, where a lot of things is happening and they're actually active and they have a lot of choices and they make good choices and bad choices, and they have a fair amount of freedom.
At the other extreme. Another common assumption about the poor is that the poor are all, uh, you know, Bill Gates in waiting. They dream of being entrepreneurs and of, uh, taking their destiny in their hands and, and so on and so forth. And all we need to do is to provide the capital and step out of the way.
So these are two of them. Other ones that are quite frequent; this is about less in Western countries, towards people in poor countries, but a lot in Western countries to people in poverty in Western countries—the idea that the poor are fundamentally lazy, something got wrong with them. Sometimes it's, you know, made a little bit more politically correct by saying it's because of their circumstances. But we always come back to that. That fundamentally, they don't really want to get out of it.
So the poor are slothful, lazy, et cetera. That's the third assumption, which is not always expressed in this way, but fundamental engaging policies. So those are some of the three that we were trying to fight against with Abhijit Banerjee.
I want to talk about the last one in particular. Um, in psychology, when we try to explain why people consider the poor lazy, we often think about research on the belief in a just world, that people don't want to believe that they live in an unfair world. And so they rationalized people's condition as a function of what they deserve, or, you know, at least like you said, their circumstances, to say, “look, this isn't wrong and I'm not contributing to an ongoing injustice.”
Do you see it similarly? Do you think about it differently as an economist?
Yeah, I think this is definitely a part of the story. But for the next book, we did a small, uh, polls of people where we asked them, if there were, for example, universal basic income, would they stop working. And most people said that they won't stop working. And then a half the people, we asked them instead, if there were a universal income, would other people stop working and they say that the other people would stop working. And likewise for, you know, when we ask people whether, uh, um, health insurance without work requirement, does that make you lazy? That doesn't make me lazy. Would that make other people lazy, that would make other people lazy, et cetera. So this disbelief that, uh, others are much more sensitive to financial incentive than I am is pretty frequent.
And I think the belief in the “just world” is definitely a part of it. It's also, uh, you know, to be honest, a pretty convenient belief, because it justifies not doing very much, because if you try to intervene, then you would actually rob the poor of the incentives to better themselves. You know, you would make them slothful by making their lives too easy and rob them of the American dream.
There is also the outside world. And I think that belief, uh, of the poor being slothful is also continuously repeated by the media [and] by policymakers.
Our French president did the whole, like, uh, beyond the scene take to look cool. And then his idea of looking cool was to stay on cap, to rehearse on camera speech that said, “giving poor people money destroys their life by making them lazy.” So I think it's just very present. So I don't think it's just motivated belief. I also think it's, it's, what's in the air and it, it takes an enormous amount of convincing to move people out of this impression, and some people will not move out of this impression.
There was a very funny headline in the Wall Street Journal when several studies appeared about a proposal grant, which is [about] the help during COVID. So, there was a worry that the help was too generous and people would stop working. Then there were about a half a dozen academic studies showing that, in fact, not at all. And the Wall Street Journal says paying people not to work will lead them not to work. Everybody knows that except [inaudible] economists, it seems. And the title was “economics versus common sense.” So this, this, this ideology is so strong that like, if it's faced with data and facts, then the facts must be wrong. The ideology is right. So we also live in that world.
We do. So tell us a little bit more about the facts. What is, in your view, the most convincing evidence that whether it's, you know, a CARES Act or universal basic income, that that doesn't make people lazy?
So there is plenty of evidence. First, the CARES Act itself, which again, it was $600 weekly of unemployment insurance. It was studied by a half dozen different teams and they all came to the conclusion that it didn't have any impact on labor supply. That's the most recent piece of evidence and extremely striking in a way. Then, uh, from all over the world in particular, all over the developing world, there has been a series of cash transfers, ex experiment. They are usually conditional on health or education behavior, but not on work.
And Abhijit Banerjee took all of these studies together. And a lot of them were done in these very rigorous, randomized controlled trials. So you can compare the - exactly compare the people who got this very generous support and people who didn't. And if you look at their labor supply, they're [inaudible] of working is just the same. And the number of hours worked per week is just the same. And this is done in studies all over the world. So that's another example. Then examples, closer to whom in the U.S. is the, the, the Alaska, uh, um, and grant [inaudible] grant doesn't make people lazy and so on and so forth. So the evidence keeps coming.
And in fact, I haven't seen, so far, evidence that was lining up strongly in the other direction.
So, what are the missing motivations that aren't accounted for? When we assume that you know, that that giving people cash essentially is, is going to decrease their motivation to work, what do you see as the reasons why they keep working and sometimes even work harder?
Yes. So, um, everybody has in mind a utility function very much like the one that your professor put on the board in your first lecture of economics, which is, oh, what people want is money and rest.
But in fact, people want more than money in particular, one thing which kind of ties everything nicely together is dignity. And well, what does dignity mean? It is a sense that your life is worth living and that other people consider that your life is worth living.
So it's a job that makes sense. It's an integration in a social network. It's a position in society. It's friends to talk to and so on and so forth. And in our societies and in most societies, it comes to, to a job and it comes to a job that, you know, you see a path with. Progress doesn't have to be that you want to become the CEO, but it means that you see what you're doing and why.
And that, that goes for a lot. Then the other thing is, you know, so people want to continue the life they have with the friends they have. That has a series of implications. But beyond the fact that people would be, would rather work than not work, another one is they would rather stay in place than move. Another thing we kind of assume is if people don't want to move for a job, they must be, they must be lazy because “Hey, it would be so simple.
You know, you've lost your job, making furniture in North Carolina. Why don't you go to New York? And you can be a security guard in a furniture shop. And there you go, you know, the great promise of trade is being accomplished. But the fact is by doing that, you're losing everything that makes your life worth living: your extended family, your support network, your house, and probably a lot of square foot of housing because you, you're not gonna find the same house in New York. Uh, maybe your career progression in your job, uh, and even the sense that you had a reason doing this particular job, as opposed to something else.
That reminds me of some research that Terry Mitchell and his colleagues did on what they call job embeddedness, which is the idea that if we want to predict people's turnover, over time, we need to pay attention to their fit with the organization. Is there alignment between their values and their interests and their skills, and what they do and where they belong?
We need to look at their links to people, and we need to look at what they would have to sacrifice to walk away. And one of the things they found in the research was that, um, you had to look at fit links and sacrificed in the community, not just in a job to determine whether people would really walk away, and it sounds like you've arrived at a very similar set of conclusions in a, from a very different perspective,
Partly because we are looking at the same data and presumably with as open a mind as we can.
Back on the point of dignity, the desire to be respected, also the meaning that you spoke about the, the motivation to make a difference. When I was doing my doctoral dissertation, I had done some experiments with university fundraising callers, and I found that just randomly assigning them to meet one scholarship recipient who had benefited from their work was enough to substantially increase the number of calls they made, the minutes they spent on the phone, ultimately the money that they raised.
And one of the first questions that I got, I think it must've been in my first job talk was, “But how does showing people the meaning of their work and seeing who they help—how does that compare against a financial incentive?” And I thought it was such a complicated question to answer, because I don't know what dose of meaning is equivalent to, you know, a hundred dollar reward and the closest I ever got to trying to compare was thinking through, could I design interventions that were equal cost?
So whatever amount of money financial incentives would require, could I substitute that for a chance to see the real impact of your work as, as a Nobel Laureate who spent your whole career doing the kinds of fields, experiments that I was grappling with in that moment, how would you think about comparing the effects of incentives and meaning and impact?
Yeah, that's a great question. One way to answer this question, is that, why do you want to know, uh, it's often in some sense, free to give meaning, or if you, uh, if you organize the job accordingly. Uh, and in, in some other sense, sometimes it's just completely impossible. So the price value you need of meaning is a bit hard to, it's a bit hard to equalize from context to context.
The reason why it's not an absurd question to ask is that, um, uh, magnitudes do matter. And not just science. And that's something that policy makers forget often, and they get like vague intuition that something works or it doesn't work, but without thinking, you know, how much, how much does it cost them to deliver that thing, that works. And ultimately when you have to make policy choices, it's pertinent and important.
I agree, but what I also found a little bit disconcerting about it was the implication that if, if meaning was as powerful or more powerful, we could just stop incentivizing people and compensating them for their work. And I didn't want to give any manager an excuse to substitute one for the other.
Ah, that's a great point. I don't think that would work. Eventually people need to eat too, unfortunately, whenever there is a possibility to make money then this, this, this wage is being exploited. So I, do think that, uh, that there is a way in which managers do attempt to provide, uh, substitutes for, uh, for money by giving meanings. But it's not always, I don't think you can fake it for very long.
I certainly hope not. We, actually had some managers who did try to fake it and they said, look, I'm just going to tell the story of the scholarship student. And I did not get the same effects in that version of the field experiment because in part employees were suspicious, they looked at that and said, Hmm, I think my boss might have an ulterior motive for trying to inspire me. And it doesn't have the same meaning to me as when I actually meet the person face to face.
So I tried this experiment or some version of this experiment in India in a different context. Uh, we had done some work on women as policymakers, so in India, they have a policy that force, uh, people to elect a woman as mayor every third election. And this is randomly assigned, so you can compare the places that had a woman mayor and places that didn't. And we compare them in the next series of elections.
What we show is that in the next election, women are more likely to run and women are more likely to be elected after the restriction is lifted if they had an experience of a woman as policymakers. So this is good since then, we've shown that this comes in part from less prejudices against women as ability to be leaders. And we were interested to see, and we were shocked at this process by just telling some people that women are actually just as good as men.
And so we did an experiment where we had, just before an election, we had actors doing a campaign on motivating electors in general. And in some version they were showing the performance of women that were elected in other places, and that had zero impact. So we came to the same conclusion that actually you have to do your experiences yourself. You, there is no substitute for someone trying to explain them to you, perhaps for the reason you're saying that I think people quickly glaze over as soon as you're trying to kind of preach to them.
I think that it's totally fascinating. And I guess that was another question I wanted to ask you about is when I, when you look at your research on women in leadership, I think the role modeling effect is obviously powerful. You've seen that girls are mostly like more likely then to rise into leadership roles, that voter attitudes change, right, when you have women randomly assigned to these leadership roles.
I wonder how you reconcile that with moral licensing research, which often shows that once people have supported one woman, they can basically say, “see, I'm not sexist.” And then they become less likely to support a second one. And over time, the risk is then that the first woman who breaks the glass ceiling does not actually end up substantially moving people's attitudes and preferences.
Well, what's interesting in this case, is that the first woman that broke the glass ceiling, uh, they were not brought in there by the voter’s will, they were put in place there. So the glass ceiling was broken from above, with a big hammer, not from a thousand paper cuts from below. And so in fact, the voters hate the woman who is the first, the trailblazers. They're [have] very negative impressions against them. And those trailblazers have a miserable time. And they don't run again, in fact. Because I think people are quite resentful to be forced to have elected a woman. So it is interesting that it's almost against their best desires that they actually, they have to, they're forced to realize that actually it never occurred to them that the woman could run.
They didn't like the idea. Actually. They still don't like the idea. If you ask them, they are absolutely willing to tell you that no, they don't like the idea that women lead, but at the same time, there is another thing that is happening, which is there's so much dislike of the idea of there being a woman leader that they never experienced it. And because they never experienced it, they also thought genuinely that women were incompetent. And when they are forced to have a woman leader, they still don't like the idea, but they are forced to realize that actually women are just equally competent.
And, and so this is, I think the mechanism in this case, it doesn't mean that you don't have what you're talking about, which is all my best friend, my best friend is a Jew or something like that.
Wow. That's in some ways encouraging, right? That even in the presence of prejudice, you can still improve people's perceptions of women's competence. It's also really depressing that even after acknowledging that women are just as competent at leadership as men, they still say, “I prefer not to have women in leadership roles.”
Yes. It's... I find it more encouraging than depressing because I can accept the fact that, uh, you know, masochism is gonna take some time to go away. And in some sense, you know, I'm often pragmatic, as long as women can still sort of slowly make their way and show their competence and, uh, and, and make a difference. Then, you know, the fact that some grumpy men are not happy about it, or even many grumpy men are not happy about it, then it doesn't bother me that much.
And I think it goes back to another, [inaudible] of my research is this idea that the very notion of what preferences mean isn't is a bit an interesting, in a way, because people can tell you that they don't like having a woman, that doesn't mean they're not going to vote for them because they think they are competent. So they also try to express something and that maybe try to understand what the social norm is and to conform to it. Uh, but, but maybe these views are less, uh, uh, deep. Then people would have you believe and that the behavior is more fluid and in today's environment, where we insist more in the media and, you know, people like you and me insist more on the fact that, uh, people are increasingly polarized in term of their opinions and, uh, um, increasingly impossible to move.
Uh, what I'm finding instead is that their discourse might be stultified, but not necessarily what they actually do.
I love this point. One of the things that I have noticed for years in my research is that people assume you have to change attitudes in order to shift behaviors. And very often the reverse is easier. If you can get somebody to change their actions or to try out a new path, then you start to activate cognitive dissonance.
Okay. Well, why did I do that? I must, uh, I must have a slightly more complex attitude than I thought. And you also lead to a change in identity, right? Well, okay. I must be the kind of person who does that sort of thing, and between wanting to rationalize their prior actions and actually learning something about who they are from seeing what they do, sometimes behavior shifts.
So I guess a concrete example of that is if I go back to the fundraising car colors, I was trying to study the mechanisms through which meeting a beneficiary of their work, motivated them to work harder and more productively. And I could not get any change in their attitudes. In my surveys, I was trying to test whether they reported their work being more meaningful, they felt more valued by, you know, by the scholarship students, as opposed to devalued by, uh, the alumni donors who are screaming at them for interrupting their dinners and hanging up on them. And I just saw no movement. And some of that I think is it's just, it's harder to measure fine-grain changes in attitudes than very objective observable behaviors and, you know, in terms of time on the phone.
But I think part of it, it was also, they had to spend a few weeks, um, you know, kind of observing themselves, spending that much more time on the job, working harder, working longer, raising more money to, to internalize that, “oh yeah, this job is not harassing people over dinner. It's making a difference for students who can't afford school.” And it sounds like you've seen a similar dynamic in your research that that's sometimes moving people's behavior is the beginning of changing their attitudes and opinions.
Yeah, it's making the motivated beliefs that you were talking about earlier in the interview work for you in a sense. And it’s also recognizing that the beliefs and the attitude and the taste, they are the product of, of many things, they are the product of your environment. They are product of what your friend thinks, they are the product, as you say, about your own perception of what you are doing and why you are doing and your behavior.
And they are much more fluid than we often think as economists, you know, inspired by Becca, we think they are what they are, and we take that as, as primitive. And I don't think that's true and also people just don't know what they believe most of the time. There are all these psychology experiments that I'm sure you know much better than me about people, for example, being influenced by their, the last number of their social security, uh, uh, number, uh, when valuing a bottle of wine.
Uh, if you make them think about the last digit of their social then they are going to value the wine closer to that. If, if they have a high number that you give a high price for the wine, if they have a low number, they'll give a low price for the wine. So this shows that on most things we just don't know, and a lot of things are complicated as you say.
And so our own notions of what we like or want or need is a little, uh, fuzzy and, uh, and, and therefore starting with behavior and starting with, you know, concrete advice as opposed to preaching is, uh, can be helpful in, in, in the context of COVID and polarization. I have a recent example of that.
So we did a very large experiment, where we sent about 50 million Facebook ads to people before Thanksgiving and Christmas. So 20 million before Thanksgiving and 30 million before Christmas. It was a group of doctors, um, who sent this very short video, asking people not to travel for Thanksgiving before Thanksgiving and not to travel before Christmas.
And we have randomized these ads at the geographical level, so we can follow the impact on mobility and we can follow the impact on COVID. And what we find is that people were quite a bit less likely to travel at Thanksgiving and at Christmas if they had received our ads. And two weeks later there is less COVID, which is a pretty rare example of a, a non-medical intervention that actually affects COVID.
But what's most interesting is, uh, that the impacts were just as large in heavily Trump country - counties as in democratic countries. Uh, and this is something that, uh, we didn't necessarily expect because there was this idea, especially after the election, that everything was politicized. And in fact, people were responding to our ads, the comments were extremely nasty usually. But at the end of it, people are just as likely to be responsive to these doctors from MGH telling them they should stay home for, uh, for Thanksgiving or Christmas.
Which sort of leads me to the belief that people actually will change their behavior if you give them information. The reason why they don't is more, maybe that they never really heard it because people are in echo chamber. So they never had a chance to get the message. Not that when given information that they wouldn't be able to, to, to react to it.
I'm struck by the specificity of that. You're not saying stay home all year, right? You're not threatening. People's freedom. You're just saying, you know, it'd be really great if you don't travel during this specific period.
Yes, I think this is absolutely a part of it, which is it's not about. Uh, and this is something which we found also in working on HIV and aids in Africa, comparing at some point we sort of run a horse-race between the government program, which is basically, uh, it's called ABCD, that's Abstain, Be faithful, use a Condom, or you Die. Which is basically pro- promoting with teenagers complete abstinence on the ground that, you know, that's the safest. Well, that's the safest if you do it.
So we tried that and contrasted it with a much more specific advice, which is avoid having sex with older men. Or we didn't even say that. We just give, uh, kids, information on the prevalence of HIV among younger men, older men and girls
And this was news to the kids because they always saw the g- the boy is the more dangerous people, not the serious, not the very serious men that paid for their school fees. And we found this to be, uh, extremely effective at reducing, uh, um, risky sexual behavior, pregnancies, herpes, HIV, and the general you know, don't go anywhere, et cetera strategy to be completely ineffective.
Wow. One of the surprising patterns in your data is that I'm thinking about when economists talk about revealed preference. I'm wondering if, if part of what you're showing is that it's often not until we see ourselves take action that we discover our own preferences, not just reveal them to other people.
Yeah, I think that's absolutely correct. And, uh, and, and people are confused about their own preferences. People are also confused about the preferences of other people.
So there is nice work on, uh, showing that uh, by Neil Bernstein, for example, showing that, that people, for example, Saudi men think that other Saudi men would not approve of wives working. But in fact themselves, they would approve of wives working. And if you make them, if you make them aware of what the social norm is, uh, then they actually, they say, “oh, really? It would be okay for people. Then it's also okay for me.” And then that leads them to be more welcoming of their own wife working.
That that I think is, is really nicely paralleled as some work that Dale Miller did on college campuses with, uh, with students who engaged in binge drinking, thinking that that was the norm. And all of a sudden discovered that not only do other people not want to do it that much, um, but they didn't even know that other people had those preferences. And once they did all of a sudden, they said, “oh, well, I don't - like if other people don't think this is cool. I don't think I want to do it either.”
Yes, exactly. This is, again, something that economists find it difficult to, to accept that, uh, you might want to do things because other people do it because what other people do is a signal of what you might like. Not just because you, you might be punished if you don't do the same. Uh, but it's definitely quite powerful.
And what this research shows is that people are often quite confused about even their reading of the signal, which is not particularly surprising when you consider that they don't understand what they want themselves.
When I think about the kind of work you do, you know, often going into developing countries and designing very large scale experiments to try to track the effects of policy changes. Um, it, it, you know, it sounds, it sounds like it's really fulfilling your, your original goal of having a career that makes the difference.
But you wrote at one point that economists are like plumbers. Can you explain this and help me unpack why plumber is a more desirable way of being in the world than we might assume?
Well, plumbers are really helpful. Uh, my father kept saying that, uh, the best for me to marry would be a doctor or a plumber, because you always need these people near you. Finally, I found neither. I don't know if it's the most, there's, it's a, it's a particularly desirable way of being in the world. It's just, uh, uh, good analogy for the work that, uh, at least some economists should do. Uh, not necessarily all of them, but, uh, you know, the economists that you and I like- that you and I didn't like when we first talking is economists that tries to be like physics. its tries to, to, to, to come up with laws of the world, laws of motions for the world. Well, you can do that for the physical world, but for the economic world you cannot, for the reason we have discussed for the past hour or so, behaviors are complicated. It's going all over the place. We have no control over most of it.
Some economists are more like engineers, which is they realize this is complex, but they say, “well let's isolate the situation and construct systems that are robust to, to a lot of, uh, of things that might be happening”. So for example, people who do school choice program, uh, school choice, algorithms tied to a program, you know, software system that are robust and, uh, uh, that can allocate kids to school in an efficient way.
But then once you realize, when you leave the lab and you go to the field, is that, um, the world is so messed up and so confused that you don't have the safety net of a bounded set of assumptions. There is always something that is not what you expected. Details are incredibly important. And you cannot, of course, when you, when someone comes up with a policy, a policy maker, or anyone comes up with a policy or program, they try to put their best foot forward.
They try to make, you know, to think of the details, the way they, the best way they can. But then it has to be that we're going to miss some stuff. We can't think about everything properly from the get-go. And therefore we have to be like plumbers. We have to be willing to try it out. And if it leaks to fix it, and if it leaks elsewhere to fix it and then do it and do it and try and by trial and error.
So the analogy was plumbing is both the fact that we just don't know, we have to try our best.t Experience helps, but, like a experienced plumber is better than a new plumber, but we will still make mistake and you have to be quite humble and how you, um, uh, in observing what you're doing and tinkering with it, and repairing it.
So that's the analogy was that it's desirable or not desirable. I don't know. It just, it just useful. And, and a lot of my work is plumbing. you know, it's not thinking about the huge question of, uh, when is the rich world going to give enough vaccine doses that we can actually immunize people in Africa?
Or do I, I hope someone is thinking about this question. I'm also trying to think about this question to some extent. But then there's going to be the next question is how to make sure that these doses go to as many people as quickly as possible to efficient campaigns that are logistically well-organized and that, uh, um, that people trust and won't have hesitancy and they will get vaccinated that also needs to be done. That's the plumbing part.
I love that. I think it's a great description of, of some of the probably underappreciated and extremely important work that doesn't always have the glamour attached to it that draws people to, to a new research question or project.
One day I was running and I was listening to a podcast and someone was lamenting the fact that, uh, after the developing the vaccine in record time, we were so slow at, putting the campaign in place in the winter of last year. And I was thinking to myself, I got the Nobel prize for that. Someone was saying, because they were saying you don't get a Nobel prize for organizing your logistical campaign. And I thought that in some sense, this is precisely what was rewarded. It's also useful. And I think it's recognized to be useful more than before. The world has become a bit more pragmatic.
I think you did a little bit more than what you just described to earn the Nobel prize. Right. But tell me, tell me how it's affected you, your career and your motivation.
I don't think it affected my motivation. I think my, my motivation comes from comes from that same, that eight year old self is still a driving me in that sense. motivation is really trying to kind of program by program thing by thing make it work. And it certainly has affected my everyday life, in the sense that I get more, opportunities to, to speak, and to, to convince, various people, policy makers, the public, et cetera, to, to think in this way and to, uh, to be pragmatic and efficient in their work to adopt certain policies, but more importantly, to adopt the kind of division of trying to be close to the ground and people's view in designing programs and implement in evaluating them.
Sometimes I feel a little bit concerned that, it's a huge asset and I should do the most of it. and sometime as I, as I live my day-to-day life in a sort of normal way, doing my teaching, solving some little bits of research or this and that, I'm wondering, oh, am I kind of not making the most of this opportunity to, to have the most influence then I feel, well, you know, I've always put one step in front of the other. I should continue.
I think that that might be one of the bigger, like maybe a, I don't know. I don't know if I would quite go this far, But it seems like one of the double-edged swords of American culture is on the one hand, you're constantly asking what have I done lately? On the other hand, you're like, oh my gosh, what have I done lately? Have I already peaked?
Yes. I think it's useful to keep in mind where we, what we were discussing a few moments ago, which is the little step that you take every day, you, you feel that they don't necessarily amount to much. And then it's when you step back and look how, you know, how they accumulated that you feel that, or maybe I have, uh, I have contributed.
And that's something which I also keep trying to tell, for example to my students, is that in particular, when you think about global poverty, it's pretty easy to get discouraged because there is a lot of it and, sometimes it gets better, but mostly the problems are land, and unless we are willing to, uh, to go, um, you know, problem by problem decompose the large problem of poverty in to a million smaller problems about how to solve that issue. And then that issue and that, and that issue, then we constantly make progress but then when you do that every day, you might have a sense that you've only dealt with very, very small things. And, uh, uh, once in a while, it's good to step back and think, “Well, that, you know, that amounted to something because you're not the only one doing that.”
Well, we are extremely grateful that you are willing to do that, that you haven't, you haven't let the, the daily feeling of a drop in the bucket keep you from pursuing, what do we do to improve the whole ocean, so to speak. I think it's, I mean, it's, it's remarkable to me that as I look across the different topics you've studied, um, you call yourself a plumber in some ways.
I think that you're the kind of scientist that every social scientist should aspire to be because you've done work that really enriches our fundamental understanding, but also that has tremendous practical, practical impact on the world.
And I'm reminded of a Donald Stokes book where he talked about pastor’s quadrant, uh, where he said, look, it's, you know, you don't have to choose between basic and applied research. It's not a single spectrum. Uh, there are separate ones, right? You can, you can do pioneering work on germ theory and then also figure out how to pasteurize milk. And that's a hell of a career for a scientist. And I, I just think you're an incredible role model for that. And I wonder if there's anything you've learned from having one foot in the world of, of science and the other in the world of policy and practice about how we can all do more of both.
Yeah, That's a great question. Because often that is something that paralyzes people a little bit, that if they do work that's too useful, they might not get the academic recognition that they want. But then if they do work that's too pointed at academic, they might not be as useful in the world. And I think I'm fortunate that at least in the field of developmental economics, the two are aligned because if we go back to the plumbing work, you could say, well, you know, it's plumbing.
It's like you really not making science progress too much by doing plumbing, but in fact, that's not true. Because, um, often by, by addressing a plumbing question, you sort of uncover a whole scientific, like fundamental scientific point that had been kind of overlooked. So for example, to go back to school choice, uh, you could say, well, you know, we have understood the whole thing of how to, uh, to have algorithm to allocate, kids to school. Uh, this is now all done, the science actually complicated, but, you know, maybe we thought it was largely done. And then my colleague, [inaudible] started to go to talk to a lot of parents and school systems to get those systems implemented, and they thought they had the best system and a feature of the best system is that, um, it cannot be manipulated.
But whenever you went to school and tried to explain to parents that there is nothing they can do to, you know, be strategic about it, their best bet is just to say what they really want. People didn't believe him. People thought that they still had to be careful. They could game the system, not put their favorite school first in case they don't get a sport, et cetera.
So you might think that's a plumbing problem. You have to just try to convince those pesky parents that really, they have to give their real choice. But in fact, taking the problems seriously you now have to think about, you know, what's the mechanism that is consistent with this type of preferences. So you now have a mechanism that not only is not manipulatable, but that you have to be able to convince someone that it's not manipulatable and what are the properties of this mechanism, et cetera. So it opens a new sort of area of thinking that nobody had even thought about until you kind of went into the real world.
And that's just an example, but a lot of the practices like that, which is the, the real question that people have in the real world, they often force you to think more deeply about your assumptions about behavior. And when you find the results, that's the opposite of what you found, what, what you thought then likewise, it makes you think. For example, one of the first randomized evaluations that Michael Kremer did, which sort of started a whole movement of randomized evaluation development, it was to put textbooks in school.
And he picked that example because he wanted something convincing. He wanted to show clearly that you can show that textbook's work and it's so powerful. And then they fund the textbook that they were not working. And there was no impact of giving textbooks to kids and he didn't believe it. He redid the experiments, still, he found no impact. He redid the experiment, changed the test course, et cetera. Still he found no impact.
So in the process, two things happen. Number one is that you really find how one has to do this experiment. So we did a lot of learning on the statistics and the practice of experiment, which is fundamental science in a way. And also because he kept finding no impact except on the best best students that sort of was the first thing in understanding much better what happens to school in developing countries and why they are not more effective and how to makes them more effective. And then we can have some thinking on the theory of that, which has history and political science and economics, which leads to, oh, how would we design intervention in this course, in this world, which leads to practical power problem programs that can be implemented and tested and practice and back and forth in this way.
So exactly like your, your, your pastor example, which is that the, the work keep building on each other's. Sometimes you on an applied end of it. Sometimes you move to the more conceptual end of it. It's often in writing a book or writing a review article that you make some of your thinking kind of glued together because each of the individual projects is like a dot and then it sort of gives you a pointillist painting at the end of it.
Um, and that way you, you live in the best of both worlds, but the best kind of advice is not to get stressed out of where you place yourself for any given project, because it's a marathon, it's not a sprint, there'll be many, many such projects.
Well, Esther, I want to thank you. This has been such a delightful and fascinating conversation and your curiosity and your knowledge are both just infectious. And I think the world is a better place because we have you trying to figure out how to make things better for people who have it the very hardest on earth.
So I'm really grateful for everything that you do.
Thank you so much. It was very fun. Bye.
Taken for Granted is part of the TED Audio Collective. The show is hosted by me, Adam Grant.
Our team includes Colin Helms, Eliza Smith, Aja Simpson, Michelle Quint, Banban Cheng, and Anna Phelan. This episode was produced by Cosmic Standard and mixed by Jacob Winik. Original music by Hahnsdale Hsu and Allison Layton Brown.
When people complain that economists are arrogant, your reaction is ...
Male economists are arrogant.