Raj Chetty believes he may have found a way to reignite the American Dream. Using “Big Data,” Chetty says we can offer equality of opportunity to kids of disadvantaged backgrounds. The renowned Harvard Professor has received plaudits for his work in harnessing data to propose solutions to economic inequality. He sits down with our partners at Amanpour & Company to explain his work.

TRANSCRIPT

CHRISTIANE AMANPOUR: Now, our next guest thinks that he might have the key to reviving the American dream. Raj Chetty is a Harvard University professor. He’s using big data to understand how to give kids from disadvantage backgrounds better chances of succeeding from his own personal experience of moving to the United States at the age of 9. He’s made it his mission to thin about how to give other children the same opportunities that he had, and our Hari Sreenivasan sat down with him to understand just how that might work.

(BEGIN VIDEO TAPE)

HARI SREENIVASAN: So your work for several years has chronicled what I would say how the Horatio Alger story about, you know, pulling yourself up by the bootstraps, working hard, that that notion of the American dream is simply not true, at least by the data that you’ve looked at. Why is the American dream, as we’ve all been led to believe it, why is that fading?

RAJ CHETTY, DIRECTOR, OPPORTUNITY INSIGHTS: Yes, so to start with the facts, back in the 1940s and 1950s, virtually all kids in America would grow up to have a higher standard of living than their parents did. So for children born in 1940, for example, 90 percent of them went on to have a higher standard of living than their parents. And if you look at kids who were born in the 1980s who are turning 30 today when we’re measuring their incomes, that number is down to 50 percent. It’s a 50-50 shot as to whether you’re going to achieve the American dream of moving up. And so, that fading of the American dream, you know, I think is of tremendous concern from an economic perspective, socially, politically, and there are a variety of factors that I think play into what’s driving that trend, but at a macroeconomic level, a lot of it has to do with the fact that wage rates and incomes for people in the middle of the income distribution basically haven’t gone up over the past 30 years, so most of the economic growth that’s occurred in America has gone to the very, very top of the income distribution.

SREENIVASAN: And you’ve kind of pointed out three kind of big categories that where you’re born, what your education level is, and what race you are are huge determinants, and we don’t like to kind of think about that because, again, we’ve been led to believe everyone has an equal shot and everyone can get above their kind of status where they were born, but where they were born, I mean, down to the zip code level with the census track is enormous.

CHETTY: That’s exactly right. So it’s not just about broad, regional variation. It’s not just about the Midwest versus the east coast versus the southeast. It’s actually about block-to-block variation within cities. So in our most recent work, we’ve been developing a tool that we call The Opportunity Atlas which allows you to zoom in to specific blocks within cities and look at what kids’ chances or rising out of poverty are at that very granular level. And the remarkable fact that you see is that in virtually every city in America there are incredibly sharp differences in children’s chances of rising out of poverty that range from the types of levels we see in the countries that have the highest level of social mobility in the world. So these are often Scandinavian countries or countries like Canada where you see kids growing up in low-income families generally have pretty good odds of rising up, but then you also have a couple miles down the road some neighborhood where if you look at the same kids from low-income families, same backgrounds, their chances of rising up look lower than any country for which we currently have data. So to me, that’s fascinating because it’s both, you know, a challenge given the scale of this problem, but it’s also an opportunity, right, because it means the answer in terms of restoring the American dream is not about going back to a previous era or looking at a completely different set of countries. It’s often about looking two miles down the road and figuring out what’s going on there and how you can replicate that.

SREENIVASAN: Where are you getting this data? It’s not like you’re out there surveying people and asking them this or that. I mean, where is this information coming from and how big is it? What’s the scale?

CHETTY: So the modern era of social science I think is really fueled by the availability of big data, much as we all hear about big data being used in the private sector. Our vision is that such data can be used to tackle important social and economic policy questions. And so, in this particular case, a lot of the data we’re working with comes form anonymized census, tax, and social security data that allows us to follow very large samples of people over time. So many of the statistics that we’re constructing on children’s chance of rising out of poverty, they’re based on 20 million kids, all kids born in the early 1980s in the United States. And it’s that scope where we’re able to follow 20 million children over a 30-year period. It would have been impossible prior to the adment of the modern information age. That’s what’s allowing us to drill down.

SREENIVASAN: So when it comes to neighborhoods and long-term outcomes, you’ve been studying, for example, voucher programs that have existed for a long time. Your previous research found that there wasn’t a net difference if you just gave people money to move, and then you did kind of a different experiment in Seattle. Explain what you tried.

CHETTY: So the whole idea of these affordable housing programs is that they’re supposed to give families access to higher opportunity areas where they and their kids might thrive. But puzzlingly, we see that most families, something like 80 percent of families that receive these housing vouchers still live in relatively high-poverty, low-opportunity neighborhoods. So, what we didn’t see at all was a pilot study where we asked, so why is it that families aren’t taking those vouchers and going and finding housing in these affordable areas, where we think their kids would have much better chances of escaping poverty? Is it because of preferences, so maybe you want to stay in the neighborhoods where you currently live, because it’s close to your family, close to your job, there might be many good reason you might want to stay in those areas? Or is the degree of segregation that we’re seeing in many of these cities, driven by some sort of set of barriers where families might not have assistance in the search process, they might not have the information they need. So, we did a pilot where we, for a randomly selected set of families, there were about 1,000 families involved in this study in Seattle, they had applied through the regular process for a housing voucher, half of them, randomly selected, we gave these additional services to help them help ease the search process. Basically remove some of those barriers. Identify landlords who might be willing to rent to you in high-opportunity neighborhoods, give you information about where those neighborhoods were, provide a little bit of assistance in saying, here’s a unit you could go check out and so forth. And what we found is that little bit of assistance up front, which actually only increases the total program up front cost by something like two percent, that dramatically shifts where families choose to live. So, in the control group, about 15 percent of families live in these high- opportunity neighborhoods within Seattle. In the treatment group, that jumps to 55 or 60 percent, so the majority of families are now choosing to live in neighborhoods where we estimate that their kids will go onto to earn an additional a $200,000 over their lifetimes, as a result of that simple move, about 5, 10 miles away from where they were living before.

SREENIVASAN: You’re also making a distinction. You’re calling these high- opportunity zones, not necessarily low-poverty zones.

CHETTY: Yes.

SREENIVASAN: What constitutes a high-opportunity zone?

CHETTY: Yes, exact — and that’s very important. So, what we’re defining as a high-opportunity zone is a place where we see in the data, kids who grow up there end up having high rates of upward mobility. So, it’s an outcome based approach. It’s just asking in a very direct way if — let’s say you’re a parent deciding where to live, a simple way to think about it is, where have I seen kids in the past who grew up in these neighborhoods? Where have kids gone on to do well, in terms of maybe having a high rate of attending college, high level of earnings and so forth and so on. So, it’s just kind of a, where do we see good outcomes. Now, that could be related to factors like having lower poverty rates, factors like having better schools and so forth, but what’s so powerful about these data is, we don’t need to rely on proxies for what might or might not predict rates of upward mobility. WE can just kind of directly measure the thing that I think matters to many of us, where are you going to achieve the American dream.

SREENIVASAN: One of the cities that you looked at, Chicago, for example, let’s say there’s a white kid and a black kid growing up friends and neighbors in a more distressed neighborhood, but their economic outcomes are going to be very different because of race. I mean, a lot of times in science studies we see correlation. You’re just saying, straight up, causation, this is the factor.

CHETTY: Yes.

SREENIVASAN: Holding for all other things.

CHETTY: You can’t overstate the importance of race and economic mobility. So, some people like to think that class is a dominant factor and maybe conditional on your income class, race is of secondary importance. That is absolutely not true in the United States. So, we have seen, with data covering essentially the entire U.S. population, take a black kid and a white kid, not just in Chicago, anywhere in the U.S., growing up in the same neighborhood, families at the same income level, same wealth level, both growing up in two-parent households, think as many things as you can to make those two families look identical, except for the race dimension, and you see vastly different chances of upward mobility, specifically for boys. So, there are big differences in rates of mobility for black boys relative to white boys, much lower chances of climbing out of poverty. But for black women, you’re odds of rising up, controlling for you parent’s income are about the same as for white women growing up in low-income families. So, there’s a gender by race innersectionality here, where it’s really about men — black men, who are facing challenges in rising up.

SREENIVASAN: So, what’s the biggest possible influence to give a young black boy a shot?

CHETTY: Yes, so you might naturally think of things like the criminal justice system, right? So we see incredibly high rates of incarceration for black men growing up in very low income families. I think that’s one of the tragic features of the current state of affairs in the United States. And, just so you know, that could be one aspect, thinking about how you reduce interactions with a criminal justice system. That could involve things directly, you know, in the space of mass incarceration, sentencing laws, and so forth. But it could also involve things earlier on in the childhood development process; changing access to schools. Another very strong predictor we find is the number of black fathers in a neighborhood is very strongly predictive of black men’s outcomes. If you grow up in a neighborhood where there are more black fathers who are present, we see that black boys have significant better outcomes. Interestingly there’s no correlation with the outcomes of black women or white women and white boys. So it’s something very specific. A plausible explanation is that this is about mentors. For example, if you see people following .

SREENIVASAN: Somebody to model.

CHETTY: Exactly, a role model. Somebody who’s career pathway you can follow of your gender or someone who, you know, looks like you; you can kind of follow in their footsteps. That makes a big difference.

SREENIVASAN: Well, and that sort of pivots back to the other part about race. Right. If you’re in a neighborhood and there is a criminal justice system that has incarcerated a large population of older black men that you might be able to look up to, there you are having a lower outcome and then the cycle sort of.

CHETTY: That’s exactly right. That’s exactly right. I think these things — there are lots of feedback groups here where, you know, you intervene in one part of the system, it might in the next generation have — can create sort of a virtuous cycle or a vicious cycle.

SREENIVASAN: So, you know, education has always been one of those ways that people say OK, this is going to be the opportunity. If I can get myself to college I’m — I’m going to make it out of this particular station in life. And I’m going to have this. And — and you have looked at access to schools, outcomes from schools, the mobility rates within schools, what do the numbers show? Is college the way out of where you are?

CHETTY: Yes, so I think there’s truth in that as an aspiration. Unfortunately, as a reality colleges are as segregated in America as neighborhoods are. So in other words if I’m a kid from a high income family, my odds of meeting a kid from a lower income family in my childhood neighborhood are just about the same as my odds of meeting a kid from a lower income family in the college to which I attend — the college that I attend. So the colleges that kids from high income families attend are very different from the colleges that kids from poorer families attend. And importantly the colleges that kids from higher income families attend, as you might expect intuitively, they tend to be the more selective ones. You know the elite private colleges, often the flagship state public institutions like the University of Michigan, Ann Arbor or U.C. Berkeley. And those are institutions where we see terrific outcomes for children. Kids who graduate from there go on, perhaps not surprisingly, to have high levels of earnings and so forth. But there are very, very low income kids at places like Harvard, Yale, Princeton.

SREENIVASAN: There’s a recent story headline that I just remember reading the other day. It said 43 percent of students that Harvard admits are legacies, jocks, or the kids of donors and faculty.

CHETTY: And look, I mean I’m a Harvard professor. Right. So I mean I recognize that there are trade-offs that institutions like Harvard make in order to support research, in order to support many different missions. But I think at some point we are, as an institution at Harvard and more broadly in the United States in higher education, going to have to confront in a — in a very direct way what our ultimate mission is. And if it’s to contribute to social mobility, which I think many of us view as central goal, I think one needs to at least think seriously about the trade-offs and the types of policies that you just described.

SREENIVASAN: You know speaking of policies right now and at least on the democratic campaign trail, we hear a lot about different variations of wiping out student debt or trying to figure out how to solve for it like Andrew Yang has a universal basic income plan. Will those sorts of policies — let’s say best case scenario somebody, whoever is president, gets their hands on the economic levers and is able to pull some of these switches, would they have an impact?

CHETTY: So let’s take each of those in turn. So something — I find some of these policy instruments are a bit blunt. So when people talk about free college, for example. So that sounds good and I think potentially moves in the right direction, but as in many other settings we find that there’s tremendous variation across colleges in terms of their impacts on kid outcomes. And so I think a solution like we just need to make all colleges free is a bit too gruff. We need to provide access to the institutions that are really propelling kids upward and we need to figure out how you improve outcomes in the institutions that aren’t doing as well. And I think that message goes not just for institutions of higher education, but all of the various pillars of our society that try to create upper mobility from the K-12 Elementary School to various other programs that we have from housing vouchers like we talked about to welfare programs. Designing them in ways such that we’re actually supporting the specific programs that create upper mobility rather than a broad brush thing.

SREENIVASAN: I know you haven’t taken, you haven’t worked for a campaign, you haven’t work for an administration. So obviously an active choice. I’m sure people come to you and say hey listen can you be my economic policy advisor, can you do x or y. What are you hoping, in the next five years, ten years — this is just — the data said that you’ve had access to, and these are the things that you’ve started to tease out a bit.

CHETTY: Yes.

SREENIVASAN: Are there — what other lines of inquiry are you intrigued by now?

CHETTY: Well, what I want to use these data for is to define in a scientific way answers to the question of how we can increase the quality of opportunity in America. So what I like about studying that issue is that I think it’s, even in this very politically partisan time, one that people from both sides of the aisle can embrace. I think everyone in America believes in equality of opportunity no matter your background, no matter where you came from. This as an abstract ideal to aim for, right. So, part of the reason I myself don’t get directly involved in partisan politics is I want to take an apolitical scientific stance to answering these questions. And I think we can identify things like better ways to design our affordable housing programs. Better ways to design our school systems. Ways to finance higher education that will lead to higher levels of upper mobility. I think we can systematically go down the list and understand the recipe for higher rates of upper mobility in certain parts of the country, certain neighborhoods and replicate that throughout the United States.

SREENIVASAN: In a way you’re talking about desegregating the ways that we have broken up by geography, by education, by race. It’s (LAUGHTER) it’s a big – big thing to tackle.

CHETTY: It’s absolutely a big thing to tackle. But I think a lot of these things, this comes back to the feedback loops that we talked about, where they can build on each other. So in one generation if you figure out how to help some kinds rise up that might reduce the amount of segregation in the next generation. Which itself then leads to better opportunities for kids from lower income families. The one place where I would hesitate on that conclusion is coming back to race. So one aspect that we have not talked about yet but I think it is very important, the context of race. Is that even if you look at kids from the most affluent families, my expectation when we were doing this research was that at some point race would become unimportant. If you were from the richest families in the America, maybe race would start to play a secondary role. If you went to the best schools, grew up in the best neighborhoods and so forth. And a really disappointing finding in the data for me was that that’s totally false. Even if you grow up in a family at the top 1 percent of the income distribution you go to the best schools in the city of New York. You go to the – you’re living in the best neighborhoods, you’re growing up in a two parent family with quite a bit of wealth. You still see that black boys who grow up in those families have much higher chances of falling down the income ladder than white boys do? So white boys who grow up in affluent families tend to remain at the top of the income distribution in the next generation? But black men, unfortunately, end up due to various structural forces falling back down to the middle class or even the bottom of the distribution. That, I think is quite distressing because often we focus on how can we help kids from disadvantaged neighborhoods rise up. But if we’ve got this constant treadmill downward pressure, we’re never going to actually narrow black white disparities in the United States. Unless we address the problems of the black middle class and upper middle class. And figure out how we can help them stay there. This is one where I think we need to think hard about how we fix that treadmill.

SREENIVASAN: Raj Chetty, thanks so much for joining us.

CHETTY: Thank you, Hari.