Search the Site

Episode Transcript

My guest today, Daron Acemoglu, is a professor of economics at MIT, and coauthor of a number of influential books written for a popular audience, including Why Nations Fail and Power and Progress. He is also, without a doubt, one of the greatest economists I have ever met.

ACEMOGLU: Historical processes really shape economic relations, and we cannot understand the economy today without understanding where we’re coming from in terms of history

Welcome to People I (Mostly) Admire, with Steve Levitt.

Daron tackles huge questions at the intersection of politics and economics. Why are some countries rich and others poor? How does democracy take hold and what factors allow it to survive? Who benefits and who loses from new technologies? I’ve left every conversation I’ve ever had with Daron amazed by his insight into how the world works. I hope that’s true again today.

*     *     *

LEVITT: I want to talk about your stature in the academic community, because I think even people who know and like your public facing work, they probably have no idea about what a giant you are within academics. So let’s start with citations of your work. That’s one of the most commonly used metrics of a scholar’s impact. I looked this up yesterday because I knew we’d be talking. You have 226,000 citations according to Google Scholar. And when I saw that number, I practically fell out of my chair. Wow! For purposes of comparison, I looked up my own citation numbers, and I’ve had a pretty good academic career. We’re the exact same age, and you have more than five times as many citations as I do. In fact, I think you are the most cited economist in the world over the last two decades. That must feel pretty good.

ACEMOGLU: Thank you for saying that, Steve. But, of course, citations are just one of the metrics, and I wouldn’t say that I have had more influence on academic thinking and knowledge than you.

LEVITT: The difference between you and me, I think, is that I write about fun stuff that doesn’t matter, and you write about important stuff that does matter, and people listen to you, and I think that’s valuable. But anyway, let me stop complimenting you. Let’s just talk about your books, okay? So back in 2012, you and James Robinson published a book called Why Nations Fail: The Origins of Power, Prosperity, and Poverty. And I remember at the time thinking it took a lot of audacity to tackle a topic that was so huge and so fundamental and so complex, that it wasn’t likely to have an easy answer. And yet you offered a really simple hypothesis. What’s the basic idea in that book?

ACEMOGLU: It’s about human choices. When people look at the big gulf that exists between rich and poor countries, they often think of factors such as geography or culture, which are essentially out of the control of most people. But how we organize society, which we broadly refer to as institutions — those are the root causes. So whether you have democratic governance or not, how you control corruption, how you structure the tax system, how you enforce or fail to enforce property rights, for whose interest you enforce property rights. But then we take one step away from perhaps the simplest economics approach, and we say: It’s all about politics and conflict when you think which type of institutions, which type of economic arrangements are going to prevail.

LEVITT: One thing I love about that book is the seemingly endless parade of case studies you present in support of that idea. But my single favorite one in the book comes from an academic paper that you and Simon Johnson and Jim Robinson published in the American Economic Review in 2001, I think. And this is just a brilliantly clever paper. Can you describe this paper you did about the residual impacts of colonialism?

ACEMOGLU: That’s a really important paper for my own intellectual development, because I have always been interested in history. And my thinking was: historical processes really shape economic relations, and we cannot understand the economy today without understanding where we’re coming from in terms of history. And one of the really important historical events, of course, is colonialism. When you look at the former European colonies today, you see a tremendous amount of inequality. Some places like the U.S., Australia, Canada, Singapore are very, very rich — among the richest nations — and some of them, many of the colonies or former colonies in Sub-Saharan Africa or the Caribbean or Peru, Bolivia are very, very poor. The right question perhaps isn’t what’s the impact of colonialism in the abstract, but think about what different types of historical legacies colonialism has created, and those legacies are really related to institutions. We started thinking about why is it that the U.S. got something very different from colonialism than, say, Peru. In Peru, they wanted to set up what we call an extractive system, an institutional structure so they will be able to extract valuables and human labor out of that area. And given the conditions that they encountered on the ground, they succeeded. In the U.S., they were more confused about what they wanted to do, and when they tried to set up similar extractive institutions, they failed. So at the end, self-governing institutions started emerging. But why the U.S.? Why not Chile or Brazil? This is where we started thinking, they probably encountered different conditions, and one very important one was whether they could actually go and settle in those places. So we wanted to understand why is it that people went to the United States and settled there, and they didn’t go to Nigeria and settle to the same extent. And one obvious answer is, of course, Europeans themselves did not have immunity against some of the biggest killers that were around the world, such as malaria, yellow fever, and other gastrointestinal diseases. As a result, the disease environment at the time had a first order effect on settlements, settlements had an impact on the early institutions, and then we documented the pathways via which early institutions lasted to today. And that was our strategy for teasing out the effects of institutions, but at the same time also bring the history of colonialism into economics.

LEVITT: If we had sat down to lunch as you started this paper, I would have told you there’s no way you’re ever going to, number one, come up with a simple explanation, and number two, convince anybody that it’s really true. And that was what was so startling, I think, to economists when we read this paper, is that by so methodically collecting data and integrating that with the history, you were able to trace this out. And I guess most economists, because we don’t think about institutions, it never would have occurred to me that England would set up different institutions, different sets of laws in places where settlers went versus where they were just trying to take advantage of indigenous people. But in fact, they set up really different rules in those places. And then, again, as an economist, I wouldn’t think that what they did hundreds of years ago, why would that affect their institutions today? So much has changed. But then you show so convincingly that actually these institutions just persisted. And then, of course, the last amazing piece of the puzzle is that then you can show that these places that kind of by chance got these different institutions hundreds of years ago, there’s incredible differences in the current level of economic development and incomes there. And for me, that’s exactly what economics should be about.

ACEMOGLU: Thank you for saying that, Steve. I think we were very lucky. And then we were also fortunate that I think the profession was really open minded and open to these different approaches. So we put a lot of effort, but luck was important in this as well.

LEVITT: Yeah, I think people underestimate the role of luck in academic research. Lots of times you have good ideas and you can’t find the data, or maybe you have what seem like good ideas and the data don’t cooperate, maybe because the data aren’t right or maybe because the idea’s not good. Let me just say, in my own experience, it’s rare that everything comes together. I would guess that of the papers I began thinking about writing, probably only one in 10 ever turned into anything any good. How about for you?

ACEMOGLU: It’s hard to know what “begin thinking about” means because I think I have lots of crazy ideas and then I have the good sense of not investing much time in some of them. But after I start a project, I think it’s one in two that gets abandoned — but that, of course, is only a selected subset of ideas and sometimes it’s very painful. You invest weeks, months, and then you realize this is not going to work. One thing I would say for abandoned projects, sometimes I learn a lot from that process. So it’s not all wasted time.

LEVITT: The concept of creative destruction — it’s the idea that new technologies or new ways of organizing society can provide huge benefits, but they often do harm to the existing elites who are extracting enormous amounts of wealth from the existing system, which means that the elites frequently resist innovations. You give fascinating examples in the book Why Nations Fail. Like there was this guy named William Lee who invented something called a stocking frame that dramatically sped up the rate at which people could knit. And one would expect this guy would get a hero’s welcome, but you tell the story in the book that it was quite the opposite.

ACEMOGLU: One of the things that the economic growth process brings is disruptive change. And this is a very complex issue. You can have new technologies be really negative on some populations. The Luddites were against mechanized weaving. But they weren’t completely wrong in thinking, that’s not going to be good for us. You know, weaver wages fell quite dramatically after the power loom was introduced. But there are other instances where new technologies are going to create some losers, but there are going to be many, many more and bigger winners out of it — consumers, workers, and so on. But there might be monopolies, especially politically connected monopolies, that would resist them. It’s not just whether I’m a monopolist but it’s also: I am a political elite. I have disproportional political power, and social and economic change is going to alter the balance of power in society. And the William Lee example is just a simple economic aspect of this, where people were really worried, including Queen Elizabeth, that once you have this much better knitting, this would create mass unemployment. And Queen Elizabeth wasn’t internalizing the plight of low-skilled workers, but she thought that this would actually destabilize her reign. And that’s why the pure economic issue of William Lee’s stocking frame became a political one.

LEVITT: He wanted a patent, right? And she refused to grant him a patent. 

ACEMOGLU: She said, “If I do this, what will happen to all my subjects?”

LEVITT: So I’m curious, I’ve thought about this a lot, and I’ve wondered, do you think that power is missing from formal economic models because it’s just really hard to integrate into the math or because economists just haven’t considered the role of power important enough to try to capture?

ACEMOGLU: I think both. So I am completely with you and that’s actually what triggered all of the work that I started doing with Jim Robinson and then with Simon Johnson. That’s what triggered the Why Nations Fail book, which you have so kindly explained in a masterful way. And that’s also what motivated my more recent book, Power and Progress. I don’t like repeating myself, but both books have power in their titles or subtitles. I think it’s so important and it’s understudied.

LEVITT: Another point that you make really forcefully in your book is the importance of property rights, and by that I mean having the confidence as an individual or as a company that the things that you own won’t be stolen by thieves or expropriated by the government. Living in the United States in the 21st century, it seems normal, almost inevitable, that there’s rule of law, there are property rights, that there’s democracy. But looking back over human history, and I hadn’t really fully understood this until I read Why Nations Fail, not only are all of these pretty rare, but the set of circumstances that led to rule of law and property rights and democracy taking hold, first in England and then spreading, it actually seems somewhat miraculous that at some point early on, an autocrat didn’t stop the whole process, derail it, and go back to this central control of all the assets. Is that your feeling as well? That we got lucky in some sense to get where we are?

ACEMOGLU: I think there is absolutely an element of luck or contingency. On the other hand, what we see, for example, emerge in England sometime in the 17th century with political power balance shifting away from the monarchy, that wasn’t the first time this was tried. The Roman Republic, ancient Greece, city-states in the Indian subcontinent, some civilizations in the New World, some in Africa, had also experimented with more participatory governments and some of them with different notions of property rights. Many of them did not survive. And what’s miraculous in some sense is that with all sorts of ups and downs, what started for the, perhaps, 20th time in the 17th century, gradually spread and became more successful. And at times it looked like it wasn’t going to. I mean, who in the middle of the Nazi and fascist takeover of Europe in the 1930s, would have thought, “Okay, the future of democracy is secure.” So we had a pretty rough patch.

We’ll be right back with more of my conversation with Daron Acemoglu after this short break.

*     *     *

LEVITT: I want to talk about your most recent book. It’s co authored with Simon Johnson and it’s called Power and Progress. Came out in May of 2023. So this is a book about technology and prosperity and it comes to a very different conclusion than most people would expect from economists as authors. You conclude that technological advances haven’t proved as wonderful as most economists would believe. You have a lot of examples in the book of technologies which create a lot of good, but almost all of that good is gobbled up by a set of elites, and it leaves the rest of society no better off. Could you run through a few of those concrete examples?

ACEMOGLU: So in the middle of the 18th century, the U.S. South got completely transformed by an equipment called the cotton gin, which enabled the cleaning of the type of cotton that the South was suitable for its growth. And once the cotton gin was in action, the U.S. South became the biggest exporter of cotton in the world. And it fueled the Industrial Revolution, the beginning of the Industrial Revolution was all about textiles and all about cotton. And many, many, many people made fabulous fortunes out of cotton plantations, cotton intermediation, exports, and all sorts of other things that are linked to cotton production. So you would think this is an innovation that just creates all sorts of benefits. But then look again, and you’ll see that the people who are actually doing the production of cotton, the enslaved Black workers of the U.S. South, they did not benefit. In fact, their conditions got much worse because they were moved to the deep south where the cotton plantations were. They started working longer hours under harsher conditions. So in this instance, who has power completely determined how unequally the gains from this new, very clever machinery were divided in society and they were divided very unequally.

LEVITT: That’s a great example. You also have other examples where you’ve got what people might call smoothly functioning markets, so not relying on coercion and slavery, where innovations, again, don’t have the expected benefits. Could you run through one or two of those?

ACEMOGLU: I think for that I would come to the more modern period because the first 100 years of the Industrial Revolution were still pretty harsh in terms of how workers were treated. But over the last four and a half decades, let’s say, the U.S. economy has been on an innovation binge. All sorts of amazing widgets, apps, and new equipment have been invented and put to use in the production process. Robots, advanced machinery, software systems, now artificial intelligence. You would expect that this is going to bring benefits that are broadly shared, but in the United States inequality has skyrocketed and in a very peculiar way where it’s not just that some people have benefited more. It’s that people with very specialized skills — management consultants, software programmers, people who have very high levels of education like medical doctors — their wages, their incomes in real terms are increasing very rapidly. But workers with high school education, even associate degrees, even some people with college degrees are seeing their real incomes decline quite sharply. So it’s a very lopsided way in which the economy has evolved over those last 45 years. And why? Part of it is because we’ve put a lot of that digital technology to use in automating work. And when you automate work, you take tasks away from certain groups of workers. Many of the high school educated workers used to work in offices or on factory floors as blue collar workers. Those have been the tasks that have been automated and they don’t have as good labor market opportunities. They don’t have access to good jobs, so they’ve lost out while other people have made a pretty good amount of money.

LEVITT: So you discuss not only that there are different kinds of innovation. So in particular, there’s automation that we just talked about. And then there are different innovations that complement workers, that make workers more productive. But what I found really interesting in your book is you make a point which is that it’s not deterministic whether you follow the path of automation or you follow the path of making workers more productive through innovation. Those are choices that arise out of the way society is structured.

ACEMOGLU: That’s one of my most firmly held views and it’s not like an ideological hang up — although by this point, it might have become so. But that’s what I started exploring 30 years ago when I first became an assistant professor at M.I.T. I started thinking about, well, you know, economists are making a lot of progress in thinking about what’s the rate of technological progress, what’s the rate of economic growth. But they’re not asking about what type of technologies we’re developing, who are we trying to help with those technologies — men, women, skilled workers, unskilled workers, workers in the U.S., workers in India, which sectors, which types of activities. And that process really culminated in thinking about choice — first as an economic matter, I’m going to look at the profits that I can make from different types of technologies; and then as a social process, there are going to be some pressures on me from society at large. I’m going to have certain ideas, what Simon and I call visions, which is the likely path that I can push this technology and what sort of damages on others are acceptable, and which ones are beyond the pale. So all of these issues come in this choice architecture. But a very important part of new innovations is creating new tasks, new activities for workers. Look at people around you. If you think of the tasks that people perform as medical doctors, academics, software programmers, upwriters, design workers — most of these tasks did not exist 50, 60 years ago. Even when you look at a well-defined profession such as a journalist, or a medical doctor, or an academic, most of the things that we do, such as podcasts, Zoom calls, doing complex visualizations — those were not tasks that an academic in the 1950s could do. The technology for creating those tasks was invented only over the last 30, 40 years. Those sets of tasks are really important for our ability to generate value and therefore for the incomes that we earn. So this is at the heart of how we create economic growth and how we distribute the gains from economic growth.

LEVITT: So, there are technologies that automate tasks that people already do. And when those happen, the people who own the technology, they get all the benefits, and the people who are automated away — well, tough luck, right? They no longer have the job they used to have and a lot of their skills now are not useful anymore. They’re now competing in a market where there’s more people competing for fewer jobs, and they suffer.  And then the flip side is what you just described so well, which is that when technologies come in that allow an academic like you or me to be more productive — or even a good example might be in a typically lower skill job, the job of the auto mechanic has changed dramatically because technology now allows that auto mechanic to observe what’s going on in the car through all sorts of complex systems in a way that never could have been observed before. That job has been transformed, but not necessarily eliminated

ACEMOGLU: 100 percent and I love the fact that you gave an example from an auto mechanic because naive thinking might be, “Oh, we’re going to create new tasks, but they’re only just going to be for the very, very well educated.” And in fact, history doesn’t support that. The example that Simon and I give quite centrally in the book is Henry Ford’s factories, which were at the forefront of automation. They introduced decentralized energy sources, new electrical machinery, the interchangeable parts system, and then turning into the assembly line. So they automated a lot of tasks, but at the same time they created these new tasks for auto mechanics.

LEVITT: We’ve been talking mostly about the past. You also have strong opinions about the present. And you think that the big tech companies have made disastrous choices, focusing on machine learning and A.I. instead of working towards machines that are more useful to people. Is that a fair assessment?

ACEMOGLU: Yeah, I would say that’s a fair assessment. Let me put that in the context of the current A.I. debates. I am completely on board with people in Silicon Valley who say, A.I. is an amazing technology, it has a lot of potential. But the issues that we talked about, which is choice about the direction of technology, who is going to benefit from it, who is going to control information, whether we’re going to automate work or whether we’re going to create new tasks — all of those are up for grabs. And in the area of inequality and wages, I link this very strongly to different ways in which we’ve conceptualized what we want from computers. One idea, very influential, goes back to the brilliant British mathematician, Alan Turing. Absolutely brilliant person. But he also shaped the way that computer science and artificial intelligence evolved by emphasizing that what we want from these machines is autonomous machine intelligence. But what we want is what Simon and I call machine usefulness. We want them to be complements to humans. And you see that this is going to lead to a much more balanced vision where sometimes machines are going to perform tasks, but sometimes they’re going to enable us to perform new tasks as humans. And they’re going to put human agency at the center. In history, actually, in recent history, you see many periods in which this machine usefulness ideas are quite important in the industry. And when they are, they lead to phenomenally important innovations. The computer mouse, hyperlink, hypertext, menu driven computers, they all came out of this view that we have to use digital technologies to make humans more capable. But especially with the more recent A.I. wave, what we are doing is not putting machines in charge. The machines, even large language models such as GPT-4 and ChatGPT, are not that capable to be in charge. What we’re doing is we’re putting a very small elite who controls those machines, who controls those algorithms, in charge. So we’re back to power again.

You’re listening to People I (Mostly) Admire with Steve Levitt and his conversation with Daron Acemoglu. After this short break, they’ll return to talk about why the U.S. economy changed drastically in the 1970s.

*     *     *

In the time we have left. I want to get Daron’s take on the economic performance since World War II, and especially what’s been going right and wrong in recent decades. Although Daron is greatly admired by University of Chicago economists, I suspect he believes that Chicago-style thinking has done more harm than good.

LEVITT: So after World War II, the world at large experienced a kind of growth in prosperity that was shared that I don’t think there’s any precedent for it in the history of mankind. Now, one could say there are a lot of different reasons for that, but you would argue that it is really this idea of the right kind of innovations that are creating new tasks and jobs for workers, along with laws and strong unions that made sure these gains were shared broadly. You would say that’s the reason we had this amazing 40 year run. Is that a fair assessment?

ACEMOGLU: Absolutely, yes, 100 percent. You summarize it really well. But what the U.S. economy achieved and what many of the other industrialized nations achieved during that period is really remarkable. On average, real wages grew about two and a half percent every year. So what that means is that if you look at a worker who starts at 20, By the time he or she is 40, their income would have almost doubled. That’s a breakneck pace of economic growth for workers. And then when you look at how different types of workers fared, was this just driven because there were some very well educated, very skilled workers who reaped all the benefits? No, it’s actually the opposite. Wages at the bottom grew slightly faster than wages at the top. So inequality remained stable or declined, not just in the United States but in many of these industrialized nations. And that’s because we used new technologies in a way that helped low-skilled workers who worked with their hands, workers who were employed in the service industry. Those people became more productive and employers then wanted to hire them and pay them more.

LEVITT: So that’s the story of post World War II until about the 1970s. And then, in the 1970s, everything goes haywire from this perspective, right? This is where inequality starts to explode and people with less education really, really suffer. So there might be many reasons why everything went haywire in the 70s, but the reason you put forth is that we made bad choices. And indeed, the one person who is most to blame for everything bad about America since the 1970s, you would say it’s Milton Friedman, the king of Chicago economists. Am I right?

ACEMOGLU: Oh, I wouldn’t say he’s the one to blame. I wouldn’t go that far.

LEVITT: Okay, he’s at the heart. I mean — ideas he had are really central to how economic activity changed.

ACEMOGLU: I would say the following. Look, I think there are different ways of running corporations. One is you essentially do it in a very cold, calculating way that says, let me run this corporation just for the profit motive. And that’s what we assume in economics, that most corporations are like that. But I think reality is a little bit different. If you’re a boss and suddenly you get a big windfall, what you thought was not a great product suddenly becomes a great product, most people would say, “Well then I would actually share some of those gains with the workers as well.” That’s not what we assume in our models, because we say, “Well, that’s more profit for the shareholders.” But many companies have this more complex social relationship with their stakeholders, their customers, their workers, perhaps even broader society. And that’s not always inefficient. And I think both kinds of corporations have always coexisted. But they have very different implications for who benefits from economic growth, and they are under different types of stresses. And what happened in the late 1970s and 80s is both ideological and economic shifts have created much more powerful forces towards the profit motive. And one set of factors for this was related to competition from Japan and competition from China and all sorts of other cost cutting pressures on corporations. But another one was ideological, and this is the famous New York Times Magazine article that Milton Friedman wrote and said, “The only social responsibility of business is shareholder value. That’s the world we should strive for.” And I think that really created a very different mindset among managers or a justification for managers and shareholders to sideline workers. And that took two forms. First of all, they didn’t share the gains as much. The fact that unions were getting weaker facilitated that. And second, it also fueled more use of automation, because if you think of your labor costs as a cost to be cut, then you welcome all sorts of automation devices that are going to enable you to cut those costs. So it was a broader social and economic change, and I think Milton Friedman captured the essence of it. And I think he was very influential because he was very eloquent, he was very respected, and he was a very good marketer of these ideas, but he’s not single handedly to blame.

LEVITT: So I want to end on a lighter note. I’m almost certain that you will not remember the first time that we spoke to one another, do you?

ACEMOGLU: I do. I do. I do. We even started, like, talking about some new research I was really excited about.

LEVITT: I don’t think — wait, that’s not the first. Yeah, I don’t think you remember the very first conversation.

ACEMOGLU: That was in the corridors of M.I.T., that’s right, yes. 

LEVITT: In the corridors. Okay. Even before that. Okay. It must have been 1994 or even 1993, so —

ACEMOGLU: ‘93, yeah, when I first arrived. I arrived at M.I.T. July 1993.

LEVITT: Okay, so in 1993 you had arrived and you were a faculty member. Okay. I was a graduate student and I remember thinking, “How does a guy from Turkey, who’s the exact same age as me, already have a job at M.I.T., the best economics department in the world?”

ACEMOGLU: Well, you know what I thought is that, you know, this guy is gonna go to much farther places than I am. Why is he still a Ph.D. student?

LEVITT: Yeah, okay, you didn’t know anything about me yet, because you had never talked to me, okay? So, I thought to myself, I better start doing whatever Daron is doing. To begin in that direction, I read a paper that you’d published, and I don’t think you had hardly any papers published back in 1993 or ‘94. And I remember I picked up a copy of your paper that was published in a British journal called The Economic Journal. And I was too shy to actually go to your office and talk to you, but one day I was in the bathroom and I was standing at the urinal. And you just happened to walk in and stand at a nearby urinal. And I’m not sure why I had the courage to talk to you at the urinal, but not in your office. But I said, “Hey, I enjoyed your article in The Economic Journal.” And I remember you being flabbergasted that I had read it and then, still standing in the bathroom, I asked you a bunch of questions about the paper and you answered them. And that was the very inauspicious way in which you and I first met and I suspect that many people, if they had been a fly on the wall watching that first awkward interaction, they would not have imagined that things would turn out so great for both of us.

ACEMOGLU: Yeah, we’ve been both lucky, but I think it was very clear from the beginning that you were going to high places, Steve, I must say.

LEVITT: Ah, thank you for that. You were one of the few who thought that. Certainly my classmates didn’t think that. What’s next for you? You’re so good at research, and it would be pity for you to move on. But I could also see you playing an important role in public policy or politics, either in the U.S. or an international organization, or — and you being a person of Armenian descent, born and raised in Turkey — in one of those two countries.

ACEMOGLU: Thank you. I have actually thought about what would suit me best and I don’t think politics would. I think there’s just too much backstabbing in politics and too much grandstanding. I would love to have an influence on policy because I think there are important issues, but I hope my influence would be from afar. But I want to keep on doing what I’ve done over the last few decades. Academically, I think I want to continue on the analysis of A.I. I think this is going to be a formative technology with huge consequences for democracy, for mental health, for inequality. So I think we need more economic analysis of that, how to regulate it. I also want to think about something we touched briefly. Economic relations are not separate from community interactions. So how do we think of local communities and how they shape economic incentives and how what’s going on with the economy overlays with the changes in community structure, segregation, less civic participation in this country and around the world. But I also want to bring some of the other things that we talked about a little bit more formally, which is about ideas. Ideas play less of a role in both political science and economics than they deserve. New ideas shape how new political coalitions are formed, how we view new economic arrangements. So how do we actually make sense of new ideas coming up? Which ideas are going to become accepted? How do we think of the conflict between ideas or the marketplace for ideas? I think those are really interesting questions that I want to explore more.

When I started studying economics, I wanted to answer exactly the type of questions Daron has been tackling over the course of his career. But by the end of my second year in the PhD program, it was clear to me I did not have the talent to make progress on those big questions. It just wasn’t the way my brain worked. Thank God I had the good sense and the self-awareness to build a great career around answering fun, little questions. Every once in a while, I wonder what might have been. What if I had gone after the big, tough questions? Maybe I could have done something amazing. And I always come to the same conclusion. There’s only one way I could have been successful in that path: if Daron let me be his coauthor. I don’t think I would’ve added anything, but I could have stolen some of the credit. If you want more of Daron Acemoglu’s thinking my two favorite books of his are the ones we talked about today, Why Nations Fail, and his most recent book, Power and Progress.

LEVITT: Now it’s the part of the show where we answer a listener question and Morgan, our producer, as always, joins me to help out.

LEVEY: Hi, Steve. So a listener named Megan had a question about why suicide negates life insurance policies, particularly standard ones that are included in a typical corporate benefits package. What do you think?

LEVITT: So this is a great question and I think the right way to think about it is by working through the incentives of the various parties involved. So let’s start with the insurance companies. It’s really easy to understand why they would exclude suicides from payouts. What could be easier than for someone who’s suicidal — often a young person with a very long life expectancy — so the life insurance policy’s really cheap, could get a huge policy and it could be financially very costly to the insurance companies. So it’s really easy to understand why this exclusion would be in place.

LEVEY: Okay, but Megan raises the point about corporate benefits packages, meaning someone gets a job and then a life insurance policy is part of the package that goes along with that job.

LEVITT: So I think Megan is exactly right, because who really thinks that someone who’s suicidal is going to research which companies have the most generous life insurance policies and would go to the trouble of applying to and trying to get a job at one of those companies? So I think Megan has really nailed it. I just think in practice this is just arcane enough and complicated enough that I wouldn’t expect insurance companies to try to change that because it doesn’t really work in their benefit in the first place.

LEVEY: Okay. But you were really excited about this question. Do you have anything else that made you excited?

LEVITT: I do. I actually had a student 20 years ago named Hsin-yu Tseng who wrote a paper on this topic that looked even more carefully and more deeply at the incentives involved and found really remarkable results. So one important institutional detail we haven’t talked about yet is that in the U.S., it’s only in the first two years of your insurance policy that if you commit suicide, it isn’t paid out. So after that it’s treated like any other cause of death. So Hsin-yu thought more deeply about incentives. He put himself in the shoes of someone who’s suicidal and he thought, “If I wanted to help my family, I could buy a life insurance policy. But rather than committing suicide in an obvious way, I could do it through the form of an accident.” And indeed, when he looked at the data, accidents were highly elevated in the first two years of having a new life insurance policy. As soon as that two-year window passed, the extra accidents, they disappeared and instead the suicide rate quadrupled. Really, I thought it was such a fascinating paper because it captured what happens so often in economic systems, which is the life insurance companies put in a policy, they think they’ve solved the problem, but then clever people with strong incentives turn out to be far more inventive than the insurance companies, and they find a way to get paid out in their policies that the insurance company hadn’t really imagined.

LEVEY: So Megan raises a great point that societal norms around suicide and mental health are changing and suicide is now thought to be really a byproduct of severe mental health issues, rather than something someone is doing to themselves. Do you see change in societal norms influencing insurance companies?

LEVITT: When I first thought about Megan’s question, I have to admit I was a little bit dismissive. I thought there’s no way that the insurance companies would ever change this rule, even for corporate programs. But now, as I put myself in the shoes of a life insurance C.E.O. and I think through the changing norms and I think about how society views mental health, I think I would be inclined to make that change on corporate programs. So not for individual life insurance, but for corporate life insurance programs because I don’t think it would cost life insurance companies very much financially, but it would be a reflection of how society feels and it would feel like the right thing to do. It’s not that often that we get a listener question that makes me think about something and change my mind. But Megan, you have succeeded today in changing my view on this issue.

LEVEY: Thanks so much for writing, Megan. If you have a question for us, we can be reached at That’s We read every email that’s sent. We look forward to reading yours.

In two weeks my guest will be Cat Bohannon, who has written an absolutely stunning book. It’s called Eve: How the Female Body Drove 200 Million Years of Human Evolution. 

BOHANNON:  The most important thing we ever did was get our hands on the levers of reproduction to overcome our most basic problem, which is that we suck at making babies. 

As always, thanks for listening and we’ll see you back soon.

*     *     *

People I (Mostly) Admire is part of the Freakonomics Radio Network, which also includes Freakonomics Radio, No Stupid Questions, and The Economics of Everyday Things. All our shows are produced by Stitcher and Renbud Radio. This episode was produced by Julie Kanfer with help from Lyric Bowditch, and mixed by Jasmin Klinger.  We had research assistance from Daniel Moritz-Rabson. Our theme music was composed by Luis Guerra. We can be reached at, that’s Thanks for listening.

LEVITT: I wish everyone thought it was an honor to be on my podcast.

ACEMOGLU: Of course they do, of course they do, Steve. They may not be as upfront about it as I am, but I’m sure they all do.

LEVITT: No, they mostly turn me down.

Read full Transcript


  • Daron Acemoglu, professor of economics at the Massachusetts Institute of Technology.



Episode Video