Skip to main content

Restoring the American Dream

Influential economist describes novel work tracking economic mobility and policies to boost it

Get all our news

Subscribe to newsletter

Our sense is that environment matters throughout childhood, not just at the very earliest years, and it's worth thinking about how you improve environment throughout childhood.”

Raj Chetty
Harvard economist and director of Opportunity Insights

Harvard economist Raj Chetty (right) takes questions from the audience about big data and upward mobility with IPR's Diane Schanzenbach, who moderated the discussion.

In his IPR Distinguished Public Policy Lecture on February 20, Harvard economist Raj Chetty discussed his research to systematically trace "the dramatic fading of the American Dream" by pairing big data with innovative models and methodology.

"With sophistication and nuance, Raj Chetty has demonstrated, in scholarly and public fora, the importance of place in terms of both racial justice and in terms of economic mobility," said Northwestern President Michael Schill, also a scholar of housing and poverty, in welcoming the 200-plus participants to the lecture. "His work has revolutionized the field and provided an example of what we do as scholars—and what IPR faculty, postdocs, and graduate students do—and how important that is to the future of our nation."

Chetty, the William A. Ackman Professor of Public Economics at Harvard University and director of Opportunity Insights, began with a dramatic contrast of the upward mobility of children born in 1940 to those born in the 1980s in the U.S.

"For kids born in the middle of the last century, it was a virtual guarantee that you were going to achieve the American dream of moving up," he said, with 92% of children going on to earn more than their parents. Yet four decades later, that figure dropped to just 50%.

This change reflects a "fundamental change" in the U.S. economy, he continued, with critical implications for U.S. society, politics, and elections.

Michael Schill

Northwestern President Michael Schill delivers the opening remarks.

Chetty began using big data when granted access to longitudinal anonymized tax records around 2010. The first study he conducted using that data was with economist and IPR Director Diane Whitmore Schanzenbach and others examining how the quality of a child's kindergarten teacher could affect the child's income at age 30.

"[It] certainly had a very big impact on the way I thought about things and the set of issues I started to focus on from there," Chetty said.

He described how his research group, Opportunity Insights, is "trying to understand how we can restore the American Dream" and the determinants of economic mobility, as well as providing evidence for scalable policy solutions.

In introducing him, Schanzenbach said, "Raj is one of the most influential social scientists around today, and as one of his former co-authors, I've seen his extraordinary research creativity firsthand as he examines and re-examines some of our most pressing problems."

Mapping the Geography of Upward Mobility | Explaining Variation in Upward Mobility | Understanding Differences in Upward Mobility | Examining Social Capital with Facebook Data Creating Evidence-Based Policies to Support the American Dream

Watch the video.

Mapping the Geography of Upward Mobility

Illustrating his talk with a series of maps, Chetty showed how anonymized tax data allowed his lab to trace the geography of upward mobility within the United States.

He pointed to the enormous variation in children's chances of rising up out of poverty across the map, highlighting areas of high upward mobility indicated in blue and green—in the center of the country and the coasts—with red and orange signaling lower upward mobility, like in the Southeast U.S. and the urban Midwest, including Chicago and Cleveland.

map.jpgAs an example, he pointed to blue-green Dubuque, Iowa, where children who grew up in families making $27,000 a year on average—which is the average household income of low-income children by age 30—are now making between $45,000 and $50,000.

"That's a tremendous amount of upward mobility in a single generation," he said.

On the other hand, children growing up in families also making $27,000 in red-colored Charlotte, North Carolina, make less money than their parents one generation later—despite an improving economy over the past 30 years that one would have expected to have lifted them out of poverty.

Comparing a place of higher upward mobility like Dubuque to one of lower mobility like Atlanta also permits the researchers to examine what happens when people move from one to the other.

"We can start to unpack what the drivers are of these differences in economic opportunity, with an eye towards potentially making changes going forward, that might give children better chances of rising up," Chetty said.

Explaining Variation in Upward Mobility

Chetty then dove into possible explanations for the wide differences in upward mobility around the nation.

He debunked the idea, touted by some economists, that differences in labor markets might be driving the gaps by showing how rates of upward mobility did not relate to job growth for the 30 largest metro areas from 1990 to 2010.

Besides, how is it that Charlotte and Atlanta, two of the fastest growing cities in America by many measures, rank so low for economic mobility? The data show that these cities import their talent, he said. Many move to them for well-paying, high-skilled jobs, but they did not grow up there.

"[This finding] suggests in very simple terms that simply trying to get the Amazon headquarters to come to your city, for example, is not necessarily the solution to increase rates of upward mobility in your own city," Chetty said.

Viewing the map through a demographic lens, Chetty highlighted how places with larger African American populations, like in the Southeast and Midwest Rust Belt cities, show up in red and orange on the map. To parse this possible connection, Chetty and his fellow researchers paired the tax data with census records, which contain race and ethnicity information. The resulting map for Black and White men exposes the stark gap between the two.

"We're finding that if you take a Black boy and a White boy, who grew up in families at the exact same levels of income, same resources," he continued, "they have dramatically different prospects of rising up."

"So, race seems to matter. But there continues to be an important role for place," he said.

Understanding Differences in Upward Mobility

To understand what might be driving the stark differences in mobility, Chetty and his team used tax records to follow five million families who moved to new neighborhoods. They examined moves for families with children of different ages, including those who moved when their child was two and up to age 23.

They found that the younger the children were when the move took place, the more they earned at age 30. But the effects faded for older children, with those who were in their early 20s when the move happened seeing no gains at all.

Such data, he said, points to the incredible value of early childhood interventions, and how moves to better areas even at later ages like 10 and 15 still offer significant returns, providing valuable insights into possible policy interventions.

"Our sense is that environment matters throughout childhood, not just at the very earliest years," Chetty said, "and it's worth thinking about how you improve environment throughout childhood."

Examining Social Capital with Facebook Data

In turning to possible determinants of social mobility, Chetty then walked through the idea of "social capital," or the connections and social assets that can confer benefits, as one of the biggest drivers.

He described a recent Nature article in which he and his colleagues analyzed Facebook data to provide a more accurate picture of social capital. They pored over the data for 25- to 44-year-olds, or 70 million people—who represent 85% of the U.S. population—with 21 billion friendships.

The researchers set up a model to predict the users' income levels and then studied their social networks to track whom people were friends with and if these friendships crossed class lines as defined by income. They examined how many people below the median income had friends who were above it and vice versa. They uncovered results similar to those on economic mobility from the tax data: Blue-green areas show more interaction between friends from different income backgrounds and red/orange less. Plotting the rates of upward mobility from the tax data against this Facebook measure of economic connection indicates a tight link between the two variables.

"It's one of the strongest correlations that we, or others, have identified today," Chetty said.

He and his colleagues also plotted economic connectedness from the Facebook data against median household income by zip code, finding that those who live in richer places tend to have more high-income friends on average, for instance. They calculate that if children from low-income backgrounds lived in counties with the same level of economic connectedness as their higher-income peers their incomes in adulthood would increase by 20%.

In comparing median income data to social mobility data, Chetty underscored how the economic connectedness measure seems to matter more. While the mechanisms are still not clearly understood, he offered that social connections might lead to jobs or how knowing someone who went to college might set you on that same path. If we think such economic connectedness matters, he continued, then what does it mean for differences in class interactions, even in places with similar income levels—and how we might increase it?

Chetty drilled into another study using the Facebook data to explore possible determinants of the differences in economic connectedness. He offered two possible explanations, either income segregation, giving the example of rich kids going to one school and poor kids to another, or "friending bias," when both rich and poor kids go to the same school but remain socially segregated.


      IPR associate Ivuoma Onyeador asks a question during the Q&A.

The researchers used the Facebook data to not only measure friendships, but also to trace friendships back to where they were formed, allowing them to examine how much people do or do not interact across class lines.

Using data from the Social Capital Atlas, Chetty then showed how students at a seemingly diverse Evanston high school engage more in friending bias, while those at a diverse Chicago high school are more likely to interact across class lines.

"The key point I want to make is that what you need to do in a place like Evanston Township High School is very different from what you need to do in a place like West Charlotte High School where the problem is fundamentally about a lack of exposure," Chetty said.

Creating Evidence-Based Policies to Support the American Dream

Chetty said his recent research indicates viable policy solutions to promote the American Dream of upward mobility by reducing segregation, investing in places, and recognizing the role that universities and colleges can play.

For policy ideas to address segregation, Chetty discussed results of a randomized control trial (RCT) of a low-cost program of coordinated support called Creating Moves to Opportunity (CMTO) in Seattle. It revealed that for those in the experimental group of 500 Housing Voucher Choice recipients, 53% of them moved to neighborhoods with higher mobility versus just 15% for the 500 in the control group. That meant kids in the experiment group could wind up earning about $200,000 on average more over their lifetimes, Chetty said.

More customization of such low-cost social support—such as housing navigation services with information about local schools and financial assistance—could further promote social mobility as well. These approaches are now being tested in HUD’s Community Choice Demonstration which launched interventions similar to CMTO in eight other U.S. cities, as well as finding their ways into congressional bills to expand housing voucher programs and corresponding mobility services.

"It's just an illustration of how I think academic research, even in this polarized climate, can have a really direct impact on policy and, ultimately, on people's lives," Chetty said.

The hyperlocal nature of upward mobility as revealed by his research also means deploying hyperlocal solutions to tackle investment in local opportunity. For instance, he discussed how the city of Charlotte, upon learning how low the Opportunity Atlas ranked it in terms of social mobility, worked to improve. Bank of America, based in Charlotte, hired and trained more than 1,000 low-income residents. The bank teamed up with a local nonprofit and community college to provide more training, and RCT results are showing dramatic increases in their income over time.

Last, Chetty explained how universities should do more to promote low-income students’ economic rise. Using data from Opportunity Insights, mobility report cards detail every college's contribution to economic mobility in America. He credits colleges and universities, like Harvard, Yale, and Northwestern, with their work to push the low-income kids who attend them to reach the upper-middle class or beyond, after college. The data, however, indicate that only 3% of students from such highly selective colleges and universities are coming from the bottom 20%. And for high-scoring SAT students, the higher-income students are far more likely to attend such institutions than those with the exact same scores from lower incomes, despite the increase in financial aid and outreach to lower-income students.

"That suggests that there's something that can be done within the higher education system itself to increase economic opportunity," Chetty said.

Watch the video.


Chetty is the William A. Ackman Professor of Public Economics at Harvard University and the Director of Opportunity Insights, a research and policy institute he co-founded in 2018 that uses big data to study the science of economic opportunity.

His research combines empirical evidence and economic theory to help design more effective government policies. His work on tax policy, unemployment insurance, and education has been widely cited in media outlets and congressional testimony. His current research, which he will discuss on February 20, focuses on equality of opportunity: How can we give children from disadvantaged backgrounds a better chance of succeeding?

He received his PhD from Harvard University in 2003 and is one of the youngest tenured professors in its history. He was previously a professor at the University of California at Berkeley and at Stanford University. He has received numerous awards for his research, including the prestigious 2012 MacArthur "genius grant" and the 2013 John Bates Clark medal, given to an economist under 40 each year whose work is judged to have made the most significant contribution to economic thought and knowledge.


IPR Distinguished Public Policy Lectures are given by prominent individuals who can speak to the use of research in policymaking and other issues. Past lecturers have included Atlanta Federal Reserve Bank President and CEO Raphael Bostic; Arthur Brooks, who was president of the American Enterprise Institute at the time; Princeton economist Cecilia Rouse, who currently chairs President Biden’s Council of Economic Advisers; Donna Shalala when she was U.S. Secretary of Health and Human Services for President Bill Clinton, and many others. IPR will welcome Yale University social psychologist Jennifer Richeson in spring 2023 as its next distinguished lecturer.

IPR is one of the country's oldest and most prominent interdisciplinary social science research institutes. The Institute's more than 160 award-winning faculty are among the top experts in their fields. Using rigorous methods, they conduct innovative, policy-relevant research, tackling some of the nation's most pressing social issues—from education and inequality to social safety nets and gun violence. IPR faculty experts train policy-minded scholars and doers, and they share their research widely with policymakers, foundations, nonprofits, and the media to support sound policy decisions.

Published: March 29, 2023.