Welcome to People I (Mostly) Admire. This is a special bonus episode — a live event recorded over Zoom and presented by W.N.Y.C. and the Greene Space in New York City. The event featured Steve Levitt in conversation with his Freakonomics co-author Stephen Dubner, covering topics such as the birth of this podcast, the power — and limits — of data science, and why Levitt’s efforts to make the world a better place usually anger everyone across the political spectrum.
Stephen J. DUBNER: Good evening and thanks for joining us tonight. I’m Stephen Dubner. I’m one of the coauthors of the Freakonomics book series, and I also host Freakonomics Radio, which is now celebrating its 10th anniversary. And the man on the other side of your screen is my Freakonomics friend and coauthor, Steve Levitt. Levitt, would you like to introduce yourself?
Steve LEVITT: Sure. Steve Levitt, and I have taught economics at the University of Chicago for the last 20 years. I write books with Dubner. And more recently, I’ve given up on academics and decided it would make sense to try and maybe have a little impact on the real world. So, I’ve started a center called RISC at UChicago that’s trying to do good. And, against my better judgment, I’ve started a podcast called People I (Mostly) Admire.
DUBNER: Now, you say “against your better judgment” as if somehow you were press-ganged into service. You were begging to have your own podcast, were you not?
LEVITT: No, it’s not that you forced me. It’s just that I pretty much try to avoid anything that has actual requirements and demands on me and/or deadlines. It breaks all my rules to actually do something where I’m obligated to someone else to show up at some time to do something.
DUBNER: So, Levitt, we should say that our partnership began quite a few years ago when I interviewed you for the New York Times Magazine article that ultimately led to Freakonomics. I thought I’d basically just interview you again tonight. Although you may end up turning things around since you yourself are now an accomplished interviewer with your People I (Mostly) Admire podcast. So that’s my first question: which side of the mic do you prefer?
LEVITT: Oh, God, I would take being the interviewee a thousand times over the interviewer.
DUBNER: So, okay. I return to the question: why on earth are you hosting a podcast if you so hate doing anything and especially interviewing people?
LEVITT: It’s not that I hate it. I really hadn’t — I’d never — literally hadn’t interviewed anyone in my entire life before the podcast. And I didn’t really understand the pressure, the intense pressure, that comes with interviewing, because when you’re the interviewee, “I’m just going to be who I am, and whatever happens, happens.”
When I’m an interviewer, I feel this infinity of possibilities. You could take the interview anywhere. And I feel enormous pressure about taking it in the right direction. At the same time while I’m trying to really engage with them and be empathic and whatnot— When I’m done with one of my interviews, I have to go home and sleep for four hours. I’m completely, totally exhausted.
So, I have found your guests interesting and your interviews really, really fascinating. But I’ll be honest with you, my favorite thing about People I (Mostly) Admire is how much of you is in it. So for instance, here’s you in conversation with, first, the actress and neuroscientist Mayim Bialik, and then the literary agent Suzanne Gluck.
LEVITT: So, whenever young people ask my advice about getting a Ph.D. in economics, I almost always try to talk them out of it. Getting a Ph.D. sounds fun and romantic. It seems like it will open all sorts of doors. But the truth is, really, it’s brutal, and it’s hard. It destroys many people’s self-confidence and sense of self-worth. Does that describe your Ph.D. experience at all?
Mayim BIALIK: Yes, it literally— I mean, it near broke my spirit.
LEVITT: I go even farther in my advice when people ask me about writing a book.
Suzanne GLUCK: What do you say?
LEVITT: I say, “Don’t write a book. Please don’t write a book. For sure don’t write a book if your reason for writing a book is you want people to read it.”
DUBNER: All right. So, Levitt, here you are on your podcast, telling everyone to not get a Ph.D. and to not write books, the obvious paradox being that you have a Ph.D. and you have written books. So, what’s the point you’re trying to make here?
LEVITT: So, look, I couldn’t believe more strongly in those two things: don’t get a Ph.D.; don’t write a book. And I think there’s this fallacy by people who get really lucky. I’ve gotten really lucky. Our book was some kind of a miracle, right? That we sold so many copies? People liked it. It opened so many doors. I mean, the fact that I got a Ph.D. and was able to be a successful economist was again super lucky.
But, to not take into account the fact that the typical experience of getting a Ph.D. sucks, and the typical experience of writing a book is you write the whole book and you sell seven copies— I mean, look, I say to people, “If it’s fun to write a book, write a book. If it’s like burning a hole in your belly to write a book, write a book. But just don’t expect anyone to read it.”
DUBNER: So, you mentioned that you’ve started this project at the University of Chicago called the Center for Radical Innovation for Social Change, or RISC. Tell us quickly what you’re trying to accomplish here.
LEVITT: So I’ve just got a group of 15 incredibly talented young people. And we are using the umbrella of academics, but not trying to write papers, not trying to get things published. But just looking at problems. And in particular, we’re kind of trying to find social problems that people haven’t solved.
And, look, the fact is, if it’s an easy social problem and it’s a popular social problem, people solved it a long time ago. The only ones that are left over are the ones that are either really hard or really unpopular. So, we’re trying to find mostly the ones that are not so hard but are super unpopular, like ones that a philanthropist wouldn’t touch because they’re worried that other people will be mad at them.
DUBNER: Give a for instance.
LEVITT: So, for instance I believe that the biggest single impact on the criminal justice system that we will have had in the last hundred years could come from using technology in a smart way. Essentially, just G.P.S. and other monitoring technologies on people who are under the jurisdiction of the criminal justice system. Okay. And I say those words, and everybody flips. I mean, everybody at that point gets incredibly irate: “Big Brother, blah, blah, blah, blah, blah.” And no matter what I say, like, “Don’t you think being locked up in prison is more intrusive than having someone knowing where you are on G.P.S.?”
But it’s really this miraculous thing in that if we can track people, potential criminals, people who have formerly been criminals and now have been released, and we can cross reference that with the list of the crimes and where they’re happening, we basically can catch people for sure when they commit crimes. If you can catch people for sure, they don’t commit the crime. If they don’t commit the crime, you don’t need to lock them up. So, you have this virtuous cycle where you can let enormous numbers of people out of prison, only lock them back up if they do wrong, but almost no one does wrong because of deterrence, save an enormous amount in personal hardship and spending, so, have lower prison populations, lower crime.
But the thing is, nobody’s doing it. And if you talk to a philanthropist— I’ve tried to raise general funds from foundations. And they just say to me, “Are you crazy? We would never fund that. That’s the worst project I’ve ever heard of.” And the right and the left hate it for different reasons. So, that’s when I know I’m onto the right track, is when the right hates it for one reason and the left hates it for another reason. I actually feel pretty good about that.
DUBNER: So, it sounds like you’re really setting yourself up to succeed there.
LEVITT: Look, but the thing is, we’re working with Cook County. And the sheriff there, Tom Dart, is an amazing visionary. And we’ve got a thousand of these bracelets on. And they’re working, and they’re changing people’s lives. People are being let out of jail that wouldn’t be otherwise. And they’re not committing crimes. And it’s— Look, I’m super passionate about it, because I really believe that a lot of the success that’s required when you want to change your world is: an idea can’t win the day.
People don’t— Other people can’t have the excitement and vision about your idea. But if you just show it to them in a way that’s so simple and so obvious, then, “Oh, yeah, why was I so upset about—? I don’t remember why I was so upset about it.” And we wrote about life insurance as a great example. One-hundred years ago, when people thought life insurance was the worst thing in the world — because how could you profit from the death of a loved one? — now, nobody thinks about that. They think about it in a more sensible way. So, we’ve got a dozen projects. But that’s the kind of thing we’re working on.
DUBNER: So, I have noticed, Levitt, that in your interviews as a host, that the center — the RISC center — is kind of creeping into a lot of the conversations. And it actually sounds like you’re on a recruiting mission, perhaps.
LEVITT: If there were 100 or 200 amazing teachers whose words were broadcast to every student, I think that gets at the point you made about access and how that could really level the playing field for people who are not as privileged as you and I have been.
BIALIK: First of all, I think we should talk about this off the air, because I think it’s an amazing idea.
LEVITT: I run a center at the University of Chicago, and we are always looking for smart, creative people who have ideas and who want to change the world. So, we will definitely keep a seat warm for you if you ever decide you want a change of career in that direction.
Ken JENNINGS: Oh, that’s very nice. Thank you. I will definitely keep that in mind.
LEVITT: Let me put my team on it.
Yul KWON: Yeah, seriously, Steve, I would love to work with you on this.
DUBNER: So, that was you apparently offering jobs to Mayim Bialik, who stars in The Big Bang Theory, and then Ken Jennings, the all-time Jeopardy! champion. And then, Yul Kwon, a multi-hyphenate who’s best known as a winner of the reality show Survivor. So, it sounds like your strategy overall is basically trying to skim the cream of nighttime television and get them to the University of Chicago to help you solve your problems.
LEVITT: You’ve seen through me. That is correct. And it’s working. It’s working great. So, Mayim and I are working on a couple of projects trying to do some social good. Yul actually made something he cared deeply about, which I also care deeply about, in terms of organ donation — he made a connection to me where RISC is now working with one of the big organ donation groups because of Yul. Ken Jennings, he was just being polite. But, yeah ,two out of three is not bad.
DUBNER: A few work friends of ours have pre-recorded some questions. So here’s a question that was sent in by Katy Milkman, who’s a Professor of Operations, Information, and Decisions at the University of Pennsylvania.
Katy MILKMAN: Hi, Stephen, Steven. What open question do you think economists should focus on in light of the Black Lives Matter movement?
LEVITT: I actually had a long conversation with Kerwin Charles on People I (Mostly) Admire — he grew up in Guyana, but he’s Black. I think we both agreed that it’s not the whole answer, but economics is a huge answer, right? If you could get something like economic equality, increased economic opportunities for African Americans, it would have a huge impact. I mean, that relates back to education. I mean, what I really quizzed Kerwin on is: everybody knows that, right? But how to do it is much harder. I don’t think we have a great answer to that.
But for sure, obviously, there are other issues, discrimination even conditional on economics. But Kerwin made a really good point. If you’re African American and you graduate from Princeton, you’re going to have a pretty good life on a lot of different dimensions. So, the challenge to economists is: how do we work on the education production function? How to make public schools better? How do we give the kind of skills that actually get jobs like data science knowledge or other tech things more equally distributed in society? So, that’s the obvious thing.
DUBNER: It’s pretty early days for having this podcast, which you mean to be an engine of discovery for new ideas or for collaboration, at least, with ideas that you’re having. I’m curious if you feel you’re making any progress, even tiny?
LEVITT: Really, I think I’m mostly just in the position of trying to publicize what other good people are doing. I mean, another of the interviews I’ve done is with a woman named Caverly Morgan, and she has succeeded somehow against all odds in getting a mindfulness class as a term-long class in all of the Portland schools. And I think that’s just really, really important. I think we neglect mental health in the school system, just because we didn’t think about it 50 or 100 years ago.
So, before People I (Mostly) Admire, I guest-hosted on the Freakonomics podcast where I did one on math education. And that turned out to be amazing. So, I basically just complained, absent much evidence, about how bad I thought math curriculum was. And that’s really turned into a movement. We’ve been working with 15 different organizations, Jo Boaler at Stanford, and others.
And we’ve got now actually a consortium; we put together a website called 21CMath.org where we have a lot of resources, where we’re trying to really make a difference. So, that’s one place. We started up a tutoring program that we piloted over the summer, remote learning and working with a bunch of school districts. It is a case where, much more than usual, I’m not just saying things, I’m actually doing some things.
* * *
DUBNER: Here’s another question from a colleague and close friend, John List, who is also an economist at the University of Chicago and who has appeared in many episodes of Freakonomics Radio.
John LIST: Hey, guys, this is John. First, congrats on 10 great years. Second, thank you so much for having me along on some of the ride. It’s been great to learn from you, and it’s been also wonderful to be part of some of your shows. So, here is my simple question. We’ve all been around policymakers a long time. And what I think is a fundamental disconnect is the distance between where we are scientifically and where current policies are in many cases.
So, what I want you to do is I want you to tell all of the academics — if you had to shake them and tell them one thing that they could do better to get more of their research to change the world through policies — what would you tell those academics to do? And secondly, for policymakers, what would you tell policymakers to look for and to explore to get more of the secrets that we have locked in our academic journals unlocked so some of those great ideas can change the world?
DUBNER: I like those questions from John. Levitt, you have any thoughts?
LEVITT: So, it’s absolutely true that the incentives that academics have fight 100 percent against implementation of policy. All the incentives for an academic are to have a great idea, to test it, preferably in some randomized study, to get it published in a top journal, and then to never think about it again. And go do that over and over and over and that’s how the accolades come. And virtually never does an academic paper in a top journal translate into policy. You have to go beyond ideas.
And actually, that’s exactly my motivation for RISC, the center I’m running. Reflecting on 20 years of nothingness, of no impact whatsoever, of constantly feeling like the ideas go nowhere, we’re trying to do exactly that next step. We’re trying to take really good ideas, simple ideas, and just prototype them in a way that a public policymaker needs to have zero imagination, needs to have zero innovation. We’re going to show you exactly the seven steps you need to do. We’re going to make it happen on a small scale, but in a real world setting with all of the kind of structure around it. And if it works for us and you want to do it, we’re going to be able to hand you that machinery so you can wholesale copy it, right?
So the secret to getting things adopted is to making the cost of adoption zero, to make it— Just copy it. So, what we’re doing with— In Cook County with the G.P.S. ankle bracelets, if that works well, we’re going to have software, we’re going to have protocols, so that someone who wants to do it doesn’t have to think at all. They just say, “Woah, that worked in Cook County. How do I do it?” And we can just mail them a big box with everything they need to do, and they can do it.
DUBNER: Right. So, you are sympathizing with the academics’ dilemma, which makes sense since you’re coming from academia, which is that they’re just responding to the incentives of how you get ahead in your profession. But you should also, I would think, be able to sympathize with the policymaker side. Especially— I mean, it’s tricky because when you say policymakers, sometimes that means the politicians, sometimes it’s not. Sometimes it’s a civil servant, but the politicians are certainly the ones out front.
And, at the risk of joining the cheering squad of an incredibly unpopular team at the moment, which are politicians, almost every politician I’ve ever gotten to know, even a little bit, it’s obvious that for the most part, most of them got into it for what we think of as the right reasons for public service. And then, they find that the incentives in this particular game that they’re playing are totally orthogonal to the ideas that they were getting into. And so, they need to do things for reputation. They need to do things for raising money and so on. And the thing about policy that comes out of academia is it’s almost always an interruption of the status quo.
There wouldn’t be a suggestion of policy if it weren’t new. And what politicians typically do is stay as close to all forms of status quo as they can. So, speaking to the policymakers out there, whether they’re elected or not, do you have any advice for how to use academic research that you truly believe in? And I guess it’s either be bold and courageous about it, in other words, not worry about all the actual political realities; or find some other channel to come forward with it, even if it goes somewhat against your personal incentives.
LEVITT: Yeah, so, I would go even further in defense of politicians, which is: most really nice academic results are somewhat misleading because they’re done at a very small scale with a set of extremely motivated and talented people as a hidden component. I’ll give you an example; not to make fun of Jeff Sachs, but a little bit to make fun of Jeff Sachs. Jeff Sachs wrote a book about how we could save a billion people on the planet or something. I can’t remember. But basically, he talked about, “Well, I took a bunch of Harvard masters and Kennedy School students and Ph.D. students. And we went to one per African village and put in mosquito nets. And these villages turned out really great.”
Okay. And they did that in like seven different African villages. I’m like, “Great, if you have a million Harvard Kennedy students, motivated people, who want to go to Africa and live in an African village and work tirelessly with incredible human capital to do it, well, maybe you could save the world. But most academic studies have hidden in them these secret inputs of talent and drive. And so, almost everything that looks good in an academic study fails when you do it at scale anyway.”
And actually, John List is really right at the forefront of that thinking, and how can we learn better. So, look, I don’t blame any politician for being skeptical about taking academic results and doing— And in some sense, in my little, tiny way, I’m trying to fix that. I’m trying to find a way to be an intermediary.
DUBNER: I should mention, if anyone wants to hear more about that, we made an episode of Freakonomics Radio called “Policymaking Is Not a Science (Yet)“. And Levitt, I remember when we were talking about you hosting People I (Mostly) Admire and what kind of people you’d want to interview and what kind of people you would not want to interview, on the “not” side was politicians. But I have noticed in several of your interviews that you seem determined to get nonpoliticians to run for office.
LEVITT: Do you ever think of going into politics?
CHARLES: I do not.
LEVITT: Would you ever consider running for office, getting into politics?
GLUCK: Oh God, no.
LEVITT: It strikes me that you would be a fantastic public servant.
JENNINGS: I admire public servants very much, but it just did not seem like a good fit for my temperament.
DUBNER: Okay. So, Levitt, that was, in reverse order, Ken Jennings turning you down again, Suzanne Gluck, who happens to be your literary agent and mine, and before that, Kerwin Charles, who is now the dean of the Yale School of Management that you mentioned earlier. But I have to ask, Levitt, since you’re pushing for everyone else to run for office and you say that your new RISC center is about finding evidence-based policy to make the world a better place; why are you not running for something?
LEVITT: So, I would not want to run for office. I mean, honestly, I wouldn’t mind being president. It’d be fun to be president if you could just be president and not have to do it. My dream is to have one of my friends get to be president. And so, without me having to do anything, they’re going to invite me. And they’re going to build a little kind of closet off of the Oval Office that I’m going to get to sit in. And I’m going to get to whisper about every possible policy issue. So, that’s kind of my dream. So, actually, it’s funny. By the time I’m done interviewing almost everyone, I’m dreaming about them being president.They’re not taking it up, though.
DUBNER: Some of my favorite conversations that you’ve had with people are extracting information or wisdom, and I’m thinking of one with Nathan Myhrvold, who’s the former C.T.O. of Microsoft; this was a long time ago. But he’s very, very, very polymathic. He trained as an astrophysicist and as a mathematical economist. And, he runs an invention firm and a lot of other things. But — and I don’t mean to pat you on the back too much or make you blush — I really love the wonder with which you receive some of this wisdom. So I want to play one clip.
Nathan MYHRVOLD: What happens with the candle is when you light the wick, that has just got a tiny bit of the solid wax in it, and you can light it. But then, the heat from that candle flame melts a little puddle of the paraffin wax, which then gets sucked up the wick, and it burns. So, the candle makes its own fuel because the candle flame can only really burn liquid or vaporized paraffin, not solid paraffin.
LEVITT: I had no idea how a candle works. And there’s so much about modern life that we just— we think we understand. But if someone said, “You know how the candle works,” I would have said, “Of course,” to think— I’m going to look at candles in a completely different way from now on.
DUBNER: So, did you really not know how a candle—?
LEVITT: Oh my God. That’s amazing. I mean, think about it. Your whole life, candles just burn. And then, all of a sudden, Nathan Myhrvold, who is an unbelievable genius and so much fun to talk to, he explains to you in a simple way how a candle works. I mean, don’t you love that?
DUBNER: I do love that. It also reminded me: have you heard of this idea called the illusion of explanatory depth that comes from psychology?
LEVITT: It’s that. We think we know stuff. And then, someone asks you anything and you literally cannot explain anything.
DUBNER: Right. Levitt, how does a zipper work? How does a ballpoint pen work? How does a toilet flush? So, I’m curious, now that you’ve learned how a candle works, did that lead to any deeper insight?
LEVITT: Look, the people who talk about real joy in life, it’s the people who are able to look at a candle and see all of the wonder in the world in the candle. I’m not as childlike as I’d like to be. But in those moments where I can get back to it, I think there’s just such joy that goes with that kind of insight. I love it. I mean, I got to go to dinner at one of the fanciest restaurants in the world with Nathan. My God. It was the most amazing experience. I mean, I don’t like that kind of food. I don’t like fancy food. I like fast food. But to actually hear him describe the making of it was the most incredible gift.
DUBNER: The psychologist Angela Duckworth from the University of Pennsylvania is my co-host on the No Stupid Questions podcast. She’s also the author of the wonderful book Grit. And here’s a question from Angie.
Angela DUCKWORTH: Do you think understanding the hidden side of everything actually changes behavior for the better? And if so, do you have any examples from your own lives?
LEVITT: So, trying to look at the hidden side of everything is essentially all I do all day long. And I think it’s fun if nothing else. And the entire world becomes a sandbox to try to figure out, apply things. And I think a good example of where it worked is on Covid. So, I think a lot of people’s reaction to Covid was a sense of being overwhelmed and helpless. And I think my reaction was, “Okay, there’s a problem. I’ve spent my entire life studying data and studying human behavior. What better problem than Covid to try and figure things out?” And I think— I’m not saying anybody listened to me. But I do think the things that I was saying made sense.
DUBNER: Give me an example. What have you contributed?
LEVITT: So, for instance, it is completely and totally clear from the beginning that if it’s true that it can be transmitted by people who don’t have symptoms, then we will get absolutely nowhere if we don’t test people who don’t have symptoms. Okay. And so, in a world in which no one gets tested unless they’re sick already, we aren’t going to do very well. I think the other really simple thing is, from day one, I loved it when the authorities, in the very beginning, said, “Look, if you’re a regular person, do not buy masks. Masks do not work, and we need every single one of them for the people who are in the hospitals.”
Okay, both of those can’t be true. It can’t both be true that masks don’t work, and that we need them for people in the hospitals. It’s also true that our only possible way out of this is if masks really work because we’re not testing very much. We’re not contact-tracing very well. We’ve just got to put on our masks and pray.
DUBNER: Well, you also liked the idea of a big incentive for testing, including being entered into a lottery that might pay you a few million dollars, for instance. Right?
LEVITT: Yeah. So, that’s true. In a world in which you had ample tests and people really were going to test and the government had a sensible program about testing, then I think a question that you run into is: well, why would anyone test? It’s not that much fun to have that thing stuck up your nose. And Paul Romer, who I’ve interviewed for People I (Mostly) Admire, his solution involved testing every single American once a week.
And you’d have to go to Walgreens or something and tip your nose back and have somebody, 52 times, stick this thing in your nose when you have no reason to think you’re sick. And Paul just thought, “Well, people will do it because it’s the right thing to do.” And my thought was, “You do that to me once, the chance that I’m ever going to show up in Walgreens again is essentially zero”, and so, my view—
DUBNER: Although we should say there are more and more saliva tests. Also, you didn’t mention vaccine development, which is— I mean, it’s amazing when you read the history of every vaccine development in the world and how it took what’s going to be 10x longer than this. It’s pretty remarkable.
LEVITT: Yeah. I’ll say the other thing that’s really interesting in this setting about vaccine development. In general, there is a real avoidance of wasting resources, of running, say— Look, we’re only going to use the best vaccine more or less in the long run, but we’ve got 10, 20, 30, 50 vaccines running in parallel, incredibly expensive. And it’s actually a great example of when there’s something valuable, you should have this enormous amount of redundancy. And people fight that usually. Of all the really awesome things we’ve done in Covid, that stand out heads and shoulders above.
DUBNER: Levitt, here’s a question from a listener, Rachel Grimmelmann watching on Facebook. Her question is, “How can we use behavioral economics to motivate people to be more cautious when it comes to Covid-19, i.e. wearing masks and so on, especially young adults in college?” I like this question. And my view from trying to do an episode on this very fact is that what we talk about behavioral economics, or basically nudging or using behavioral science to influence human behavior, has been mostly a massive failure in light of Covid and public health behavior.
We’ve found no evidence that really any of those smart, small nudges could cut through the noise of mass concerns. And I’d love you to tell me I’m wrong, Do you have any countervailing evidence?
LEVITT: Look, they don’t really work very well. In general, nudges get you a couple percent. They get them really cheap and for free, whatever, so they’re useful. But I can barely walk down the street without somebody from some company or organization saying, “We want you to use behavioral economics to transform everything about what we do.” And it turns out that really, honestly, it doesn’t, like you said, work very well.
And whenever you say the sentence, “I want behavioral economics to transform whatever I’m doing,” you should just cut the word “behavioral,” and say, “I want to use economics to transform.” Because it is, behavioral economics, good for a couple percent. But regular economics, the kind of economics that we do in Freakonomics, which is about incentives—
So, you can use financial incentives. Financial incentives would work on Covid. I guarantee you there’s enough money. The value of somebody who’s sick not infecting other people is so huge to society that we could pay people enormous amounts of money not to infect other people if we chose to. But social pressure — I mean, my God, if I walk down my street in Chicago outside the University of Chicago and I don’t have a mask on, people scream at me. They tell me I’m an awful person and I fear that deeply. So anyway, social pressure is an enormous amount. But that’s regular economics, about incentives.
DUBNER: So, here’s a related question. This is coming from Cori Blackburn, who’s watching on Facebook. Cori wants to know any words of advice for fledgling data scientists to catapult a career into something meaningful. And I just want to add to the question or maybe a preface to the question, which is, I have a sense that many, many people think that data science and big data are the latest magic bullets that will solve absolutely everything. And again, I see a lot of evidence of good intention and poor outcomes. And I see a lot of—
It’s almost like moral hazard, where people think, “Oh, now, the data scientists will figure out X, Y, and Z.” And it strikes me as a good tool to have if you have people who know what they’re doing and they know how to ask really good questions, but that’s not what I see. So, tell me that first. And then, get out of that depressing mode and give Cori something to be less depressed about.
LEVITT: Yeah. It’s a super insightful point that you just made, Dubner. I couldn’t agree more that a lot of times people think of big data as a substitute for good thinking as opposed to a compliment to good thinking. When you are really thinking smartly and cleverly about a problem and you have really great data, you can often make a lot of headway. And without both, you sometimes can’t. But really, big data has proven immensely powerful at — and data science — immensely powerful at a certain class of problems, which is prediction problems.
If you’re trying to predict the future, if you’re at Netflix and you’re trying to figure out what kind of movie I would like, these tools have been amazing. If you’re trying to figure out, by looking at a picture, whether in that picture it’s a cat or a dog or of a person holding a gun — amazing progress. On most of the questions we care about, really on any question where the question is “Why did something happen?”, these tools aren’t that great. When I work with firms that have data scientists there, what I find almost uniformly is that they operate in an incredibly complex space. They’re very concerned with technicalities, with techniques, with things being hard. And I think the answers are often very simple, right?
And so, I try to always do simple things and try to relate them in very basic ways. My favorite kind of graphs are what I call big bar-little bar graphs. They’re graphs that have one really little bar and one really big bar. And those are the kind of graphs that I show to C.E.O.s if I’m trying to convince them of something. And the C.E.O.s say things like, “Wow, that makes sense to me. I don’t understand how you take the same data that my data science team has, and I never understand anything they’re saying.”
So, I might say, “Oh, well, the answer is just to do things really simply.” But I think it’s more complicated when we think about incentives. Because much of the power that comes to data scientists in firms and organizations is because they’re completely and totally inscrutable. And the other people have no idea what they’re doing. And by having a set of skills that no one else has, you can wield power because no one can understand what you’re doing. You have a very special talent.
And so, I have the luxury of being an outsider, of being a bigshot economist so I can do simple things and people don’t second-guess me. But if you’re a data scientist, I would say experiment, right? So, sometimes do things simply, and sometimes do things super-complicated. See what kind of results you get. Look, I do think just by virtue of being in data science, I think you’re in the right spot. I think the future belongs to data science. It’s just a great job. So, I encourage everybody who’s interested in STEM stuff: take a shot at doing a little data science. And if it appeals to you, go with it. It’s a fantastic investment.
DUBNER: So, Levitt, to wrap things up, I wanted to ask you about how you wrap things up on your show, People I (Mostly) Admire. Here’s a clip that we put together from the last few minutes of multiple interviews.
LEVITT: All right. Last question. Last question. All right. Last question. It seems to me you’ve lived a really good life. Do you have advice about living a good life? So, what advice would you give on leading a life worth living? So, how about on living a great life?
DUBNER: So, Levitt, that sounds a little bit like a midlife crisis in radio playing out before our ears. What I really want to know is why. A lot of people who interview others, you do kind of develop a signature question, often to end with, or a signature saying. And I’m curious why this is the question you chose. And really what I want to know is if you’ve gotten a good answer yet.
LEVITT: So, honestly, let me say, I wouldn’t have known I did that. It’s totally unconscious. It is a reflection of my own thinking. And the people I’m talking to have often had great success, often in different dimensions. And I think in the broader sense, it seems like that’s the right question.
DUBNER: Well, let me ask you this. Name a person — known, unknown, living, dead — whose life you look at and say, “Oh, that’s the way to go.”
LEVITT: Oh God. E. O. Wilson has been one of my idols. He taught me in college. He’s the guy who studies ants. And his class fundamentally changed my life. I mean, it’s the only person I can say, any teacher, who I would say, “Wow. what I learned from that person, I live my life totally differently.”
DUBNER: Give me an example.
LEVITT: Among other things, I had an intense fear of death. As a child, I would wake up in the night screaming, afraid of death. And E. O. Wilson taught me not to be afraid of death. And he did it in a weird way. He basically made me understand how unimportant I am because I think this fear of death in large part stems with a very overblown sense of self and your importance in the world. And he convinced me in a very nice way that I’d mattered not at all.
But then, he brings you back. You’re like— He’s maybe like bootcamp or something where they rip you, they cut off your hair and they rip you down to nothing. And he said, “Look, you don’t matter on a cosmic scale. But among the people that you interact with, you’re everything. On a local level, you’re incredibly important.” He said, “So, the right way to think about leading a life is on a local level.”
And I don’t know. For whatever, like— Maybe if I took it now, I would think it was stupid. But as an 18-year-old, it blew my mind, and it’s weird. My grandfather, he was someone who enjoyed life in a way that I’ve rarely seen anybody enjoy life. And then, he committed suicide when my grandmother was dying of cancer because he was ready to move on. And to me, that was an incredibly influential thing. He basically committed suicide shortly after I took E. O. Wilson’s class. Honesty, that, along with the class, it freed me up to really live a better life.
DUBNER: Well, Levitt, it was great to see you. I missed you. Let me just ask you, do you feel now that you were the interviewee and not the interviewer that you don’t need to go sleep for five hours?
LEVITT: Oh, I’m ready to go. No problem. This was nothing for me.
* * *
People I (Mostly) Admire is part of the Freakonomics Radio Network, and is produced by Freakonomics Radio and Stitcher. This episode was produced by Matt Hickey with help from Mary Diduch. Our staff also includes Alison Craiglow, Greg Rippin, and Corinne Wallace; our intern is Emma Tyrrell. All of the music you heard on this show was composed by Luis Guerra. Thanks to WNYC and the Greene Space for their assistance with this episode. To listen ad-free, subscribe to Stitcher Premium. We can be reached at firstname.lastname@example.org. Thanks for listening.