— Raj Chetty
(Keynote address at the 9th National Housing Mobility Conference. September 20, 2023)
There’s no higher compliment, in my view, for an academic than seeing that your work has had a real impact on people’s lives, so this conference is incredibly meaningful to me. And of course, that impact is only possible because of the work you all do, so I’m very grateful for all of your efforts. As I was reflecting on what I’d talk about this afternoon, it occurred to me that my own life, the opportunities I’ve had, have really been because of housing mobility, not so much at the local level, but in an international sense. My parents grew up in low-income families in South India, and moved to the United States in search of the American dream, in search of better opportunities for themselves and especially for their kids, like countless other immigrants. And when they moved to the United States, which was back in the 1960s, the U.S. was truly a land of opportunity, much more so than it is today.
So in this first chart here, I want to set the stage for what I want to talk about this afternoon. What we’re looking at here is children’s chances of achieving the American dream, measured in a simple way: asking what fraction of children go on to earn more than their parents did by the year in which the child was born. And what you can see is that for kids born in the middle of the last century, back in the 1940s or 1950s, around the time my parents and others came to the United States, it was a virtual guarantee that you were going to achieve the American dream of moving up. We estimate that 92% of kids born in 1940 went on to earn more than their parents did. But if you look at what has happened over time, you can see that there’s been a dramatic fading of the American Dream such that for children born in the 1980s, who are turning 30 around now, when we’re measuring their incomes as adults, it’s become basically 50/50 as to whether you’re going to achieve the American dream of doing better than your parents: the American dream has become a coin flip. And so that broad trend is of great interest to economists like myself, because it reflects a fundamental change in the U.S. economy that we’d like to understand. I think it’s this very trend that underlies a lot of the frustration that people around the United States are expressing, that this is no longer a country where it’s easy to get ahead, even through hard work.
So why do I start with this chart? Broadly, in our research group at Harvard—Opportunity Insights—we’ve been interested in studying why the American Dream is fading, and what we can do to restore the American Dream going forward. And there are many different issues at play. And basically, the way I’ve structured my thinking and research agenda over the past decade or so is trying to understand the science of economic opportunity, what is happening here and what we can do to make progress going forward. That’s taken us in various different directions. Something I didn’t anticipate is that it would take us in the direction of housing policy – and housing mobility, in particular—as being a key aspect of what may matter for economic opportunity.
What I’m going to do today is spend a few minutes talking about how our research proceeded from the first chart, and what I see as the state of the field, and this is going to connect to themes you’ve heard about in sessions this morning and what you’ll hear about in the next session on the Community Choice Demonstration, which I think is very exciting. And then I’m going to talk about some of the key open questions and future directions in this field.
So let’s start with this trend of declining intergenerational income expectations. When you look at this trend, overall, it’s very hard to figure out exactly what may be driving it. Because there are numerous things that have changed in the United States over the past 60 years. And you probably can think of many different explanations: globalization, changes in inequality, the decline of unions, changes in levels of education, numerous things could have been driving this trend. What has really allowed us to make progress as a field, in the context of housing mobility, in the context of many other questions from a scientific perspective, is the ability to break down this national data in a much finer grained way, thanks to the power of big data. So, very analogous to how the microscope revolutionized biology, we’re at a point in the social sciences where we now have the tools and the data to look at these kinds of kinds of questions in a much more granular way, and thereby obtain much greater insight. And that’s basically the approach we’re trying to take at Opportunity Insights to this and many other questions.
So, what was the first step in that agenda? To understand economic mobility in a finer-grained way, we used data from anonymized tax returns, covering the entire US population, and we looked at a subset of kids who were born in the early 1980s in the United States, about 20 million children. We mapped those kids back to where they grew up, and we linked them to their parents. And we divided the US into 740 different metro and rural areas. And in each of those areas, we calculate a very simple measure of upward mobility. We asked, “What is the average income at age 35 for kids who grew up in low-income families in these areas?” And the first simple point that became evident to us about 9 or 10 years ago is that there’s an enormous amount of variation in your chances of rising up and achieving the American dream. Even in the present day, in the United States, there’s some parts of the country like the center of the U.S., a place like Dubuque, Iowa, for example, where kids growing up in families making $27,000 a year, on average, are making $45,000 or even $50,000, a year, a tremendous amount of upward mobility in a single generation.
Yet, there are other places like Charlotte, North Carolina, which I know is represented here today, and is trying to make lots of changes on these very issues. In the historical data, Charlotte, unfortunately, was a place where if you grew up in a family making that exact same income level $27,000 a year, one generation later, you’re actually making less than your parents were on average, which is kind of unbelievable, given the amount of economic progress that has happened over the past 30 years in America in general, and in Charlotte, in particular. So we started to recognize that place might matter in an important way. And furthermore, in subsequent work, it became evident that place matters not just at the level of different metro areas, but at a much finer geographic level.
I think a number of you have heard about the Opportunity Atlas, a tool that we put out about five years ago, that allows you to look at the same set of statistics that I was just showing you, but now at the census tract level. Here, I’m just showing you a snapshot in the New York City area where, tract by tract, you can see what children’s chances of rising up are. The striking thing in this map is the spectrum of colors that you’re seeing within New York, from the darkest reds to the deepest blues. You can drive two miles down the road in New York City. And it’s like you’re going from Alabama to Iowa in terms of rates of upward mobility. So the roots of these differences in economic opportunity are hyperlocal. And this is the sense in which mobility across neighborhoods can potentially be quite important.
We’re by no means the first to think that neighborhood environment in place might matter. Folks in this room: sociologists, practitioners, for many decades have had that view, and a lot of their work has been motivated by that hypothesis. But it’s been surprisingly hard in the social sciences to actually document hard evidence that places have a causal effect on people’s outcomes. And so the key next step in this research agenda, which I think was really crucial for the field is understanding what places’ causal effects are on people’s outcomes. And again here big data was very useful. We were able to look at millions of families that moved between different neighborhoods, in New York and in cities across America. I’ll skip the details, but basically, the punchline of what we found is that places do matter. They matter in particular for kids, rather than for adults. And they matter in proportion to how long you are growing up there.
So now you can start to understand why the early studies of Moving To Opportunity necessarily focused on adults and older kids, because the young kids had not grown up yet when HUD was doing its final impacts evaluation, or when my colleague, Larry Katz, and many others were writing studies using survey data of MTO. They just weren’t able to look at the youngest kids. And so with this big data approach, we were first able to show at the national level that you get this very strong dosage effect: the more years you spend in a better environment, the better you do.
So then we were able to go back to the MTO data working with folks at HUD, linking that to the Internal Revenue Service tax records, and showed that you get the exact same pattern with MTO, a randomized control trial, as we saw in the Opportunity Atlas. Then things really started to make sense, in accord with the intuitions of people here who were working on these issues well before I was. We had kind of missed this proof with the research methodology we had before, and we learned we need to focus on the young kids who get that biggest dosage of moving into a better environment, who gain the most from making those moves.
With that set of research findings, the next thing that emerged that’s really influential in my own thinking about how to do research, both in this context and more broadly, is thinking about what this actually means for practice on the ground, and how we can make this work for more kids. And the pivotal moment here—which I think Sunia [Zaterman] mentioned in her introductory remarks this morning—was when Greg Russ [then executive director of the Cambridge Housing Authority] walked into my office one summer at Harvard. And I vividly remember he had a number of our papers marked up in red ink. He had read them in great detail. And he said, “You know, I think there’s a real opportunity here to work with practitioners in the housing mobility space to make a difference.”
The thesis was, working with Greg and others, like Stephen Norman and Andrew Lofton in Seattle, let’s try to think about how we might make a difference. What can we learn about how to make housing mobility programs more effective? What came out of that was a collaboration Creating Moves To Opportunity (CMTO) that I think a number of you are familiar with. It was basically a pilot study, set up as a randomized trial, to help families with housing vouchers move to higher opportunity neighborhoods by doing three things: providing customized counseling services, connecting people to landlords, and providing a little bit of financial support for things like security deposits, application fees, and so on in a tailored way. Essentially, the way I think about it abstractly is sort of a social support service to help you use the voucher more effectively without fundamentally changing the parameters of the voucher itself.
So we set up a randomized trial with the Seattle and King County Housing Authorities, we had almost 1,000 families come in to apply for housing voucher to in the standard way, and 500 of them—roughly half—received these additional services.. And we were then able to follow these families over time and ask what happened. And what you can see from our results, and as I think a number of you are aware, these additional services made a huge difference. In Seattle, the comprehensive mobility services cost about $2,500 per family, which is not a tiny sum, but relative to the monthly voucher payments that you’re making, it’s actually not that much of an incremental expense. This intervention dramatically changed where families chose to live. In the treatment group, 55% ended up moving to high opportunity areas, while in the control group only 14% did. Now, this was very exciting to us, but the next question was, of course, why is this happening?
One of the new things we did in the context of this study was to collaborate with the brilliant Stefanie DeLuca, who led a team of folks to talk with the families who ended up moving to these high opportunity places to try to understand what it was that was making a difference to them. I know that Stephanie touched on some of these themes this morning, so I’ll be brief, but we learned that the families valued the support that these mobility programs provide above and beyond the financial support of the voucher itself: emotional and psychological support, the flood of relief that comes from having a housing counselor helping you out, somebody who’s helping you broker with landlords and navigate an unfamiliar situation, someone who’s able to provide short-term financial assistance exactly when you need it, and tailored to your particular situation.
I think what I see as a vision for modern social science, is about bringing these different types of people working together: people doing quantitative work like myself and my team, and people on the ground, practitioners, like all of you who really understand how these programs work and make can make a difference, and people doing qualitative work like Stephanie. I think there’s a valuable interplay between these different methods.
Many economists would have the view that what really matters is giving people incentives to move to these higher opportunity places, maybe telling them where they are. If you take a totally rational economic view of the world, that’s the only thing that you might think would matter. In fact, that turns out to be totally wrong. If you just give people financial incentives and information, it doesn’t get us anywhere close to what the customized services do. And if you do kind of a “light touch” version of these services—so basically try to save some money and don’t help people on a one-on-one basis, do something a little bit more uniform— you get part of the way there but not nearly as far as really working with people kind of meeting them where they’re at. So the lesson from that is these kinds of mobility services can be really impactful, but they need to be done in this higher touch sort of way. And they can really dramatically shift the impacts of the voucher program.
Just to put a number on it, we estimate that the average kid who grows up in a high opportunity place as a result of being randomly assigned to the treatment group will go on to earn about $200,000 more over their lifetimes, relative to the children in the control group who, just by chance, didn’t get those additional services. So that shows you the value, the rate of return to society of that $2,500 upfront investment. And I think it’s that kind of quantification that that can help motivate policymakers and Congress to change things on scale. And so what came out of that sort of work and other related work that others have done, as you all know, is the bill that supports the demonstration that HUD is now doing, and a number of you are now involved in – plus the new housing mobility services funding that just went out. There are other policy efforts like the Family Stability and Opportunity Vouchers Act, which is currently being considered in Congress, a $5 billion expansion of the Housing Choice Voucher Program, with additional mobility supports, additional opportunity vouchers, and so on. All of this data helped shift the conversation, in terms of what we should do going forward in the housing mobility space.
What I want to do in the last couple of minutes here, is talk about what I see as some of the key open questions and future directions in the field. One question we’re very interested in is understanding the mechanisms through which neighborhoods matter. What is the underlying science of these differences in economic opportunity? Why do some neighborhoods give kids much better chances of rising up than others? Is it, in fact, just about the schools? Is it about integration?
One factor that I want to highlight that’s emerging in our research as being quite important is this idea of social capital, which I know Sarah [Oppenheimer] touched upon in her remarks this morning, and I just want to come back to it and show you why it seems to be so important. To explore this issue we created what we’re calling the Social Capital Atlas. This is constructed from a completely different data source than the Opportunity Atlas, using Facebook data, to measure the extent to which low and high income people are friends with each other in different counties in America. And it shows very clearly that in places with greater cross class interaction, like Dubuque, Iowa, for example, you have much higher rates of economic mobility, than in places like Charlotte, North Carolina, where you have much more disconnection across class lines, and correspondingly much lower levels of economic mobility. Who you’re interacting with and who you’re connected to, what your networks look like, what your social capital is, in that sense, is really a key driver of mobility and other related outcomes. And that’s part of the reason why housing mobility, which is basically one way to create more integration, can be so effective in increasing economic opportunity going forward.
A second key thing that we’re thinking about, and again, I know this came up in the opening remarks, is thinking not just about helping people move to opportunity, but bringing opportunity to people where they currently are. There’s no reason to debate which of these approaches is more effective, I think it makes sense to take a “both-and” approach here. So how can you create high opportunity neighborhoods? Right now, for example, we are looking at the long term impacts of public housing redevelopment to see what the second generation effects look like – stay tuned for that research next year.
Another preview of work we’ll be putting out in the next few months, is an analysis of how opportunities change across neighborhoods, like a new version of the Opportunity Atlas that evolves over time. This addresses a limitation of the Opportunity Atlas as it currently stands – as Phil [Tegeler] and others have correctly pointed out, the Atlas only shows you the outcomes for kids who are growing up in the 1980s and 1990s. Obviously, what we really want to know is what the outcomes look like for more recent generations of kids. And so now with enough data, we’re able to construct these maps where you’re able to look at how a particular county is changing over time: are things changing in Charlotte in more recent years? How are they changing for Black folks in Charlotte versus white folks in Charlotte? That is new data that we’ll be putting out. And we hope will shed light on these issues going forward.
Let me conclude by coming back to the slide that I started out with. You know, you can look at this picture as a disappointing one, showing that the American Dream is fading. But I think increasingly, we also have the tools to understand how we can change that trend going forward. And I think the work that all of you are doing is an incredibly important part of that picture.
Raj Chetty is the William A. Ackman Professor of Economics at Harvard University and the Director of Opportunity Insights (opportunityinsights.org)
Link to the full issue of Poverty & Race Journal, Volume 32 No.3 (August-December 2023)