Economic Mobility Through Cross-Class Connections
5 questions with social capital researcher Johannes Stroebel
In most communities in the United States, kids who grow up in poverty are extremely likely to live in poverty as adults. But in some communities, kids who grow up in poverty have a good chance to move into the middle class later on.
Recent research has uncovered why upward economic mobility varies so greatly across communities, and it has to do with social networks. Communities with more economic connectedness — places where relationships across economic diversity are more common — have significantly more upward economic mobility.
We spoke with Professor Johannes Stroebel of New York University, one of the researchers behind these groundbreaking studies, to learn more about this research and to find out what communities can do to support economic connectedness and create places that support socioeconomic mixing, one of the four outcomes of Reimagining the Civic Commons.
This conversation has been edited for length and clarity.
Q: What are the top takeaways from your research on social capital and economic mobility?
Johannes Stroebel [JS]: We started thinking about this question because we have known for some time that upward income mobility varies across locations, but we have not understood the drivers. For example, a kid with poor parents growing up in Salt Lake City has a much better chance of rising up in income level over their lifetime than a similar kid growing up in the South. But what generates intergenerational upward income mobility in some places and not others?
People have thought about a range of explanatory factors, and one theme that we kept hearing about was the importance of social capital. In the research papers we published in Nature, we tried to empirically assess the strength of the relationship between social capital and upward income mobility and understand how differences in social capital might contribute.
Social capital wasn’t a particularly well-defined term, so we started by classifying concepts of social capital into three measurable categories.
The first is the cohesiveness of a community. This refers to the density of social networks, or how many of your friends are also friends with each other.
A second is connectedness, or the extent to which people with different characteristics within a community are connected to one another. Are Republicans connected to Democrats? Are old people connected to young people? In our research, we focus on economic connectedness, or how much poor people are connected to rich people within a community.
The third measure of social capital is civic engagement. This is the extent to which there’s trust in a community or people participate in their community’s public life.
To measure social capital empirically, it was important to measure the structure of social networks. Because the first two categories, cohesiveness and connectedness, are all about who is friends with whom. So we started a collaboration with Facebook and, working with anonymized versions of their data, came up with local measures of the type of social capital we wanted to study further.
One of our key findings is that variation in upward income mobility is highly correlated with only one type of social capital: economic connectedness. The extent to which low-income people are connected to high-income people appears to be a driver of upward income mobility, at least in a correlational sense. The evidence for a tight causal link is not quite there yet, but economic connectedness is highly predictive in terms of explaining variation in upward income mobility.
Q: Why do you think economic connectedness boosts economic mobility more than other factors? What is it about living in communities with a lot of connections across income levels that helps move people out of poverty?
JS: There are several possible mechanisms for this, and all of them are very plausible. The first is direct access to opportunities. The idea that, if you grow up with friends from higher-income families, things like accessing internship opportunities might be easier. You may have better access to these opportunities through relatively wealthier friends.
A second, related mechanism is access to information. Your higher-income friends might tell you about job opportunities that you otherwise would not know about. Similarly, information about how to navigate the college application process can be very valuable.
The third mechanism is around aspiration. A lot of people who reached out to us or commented on articles about our paper in The New York Times and elsewhere said this idea captured their experience of upward income mobility. They said some of the rich friends they had as kids shaped their aspirations and showed them there’s life outside of their immediate community.
It’s hard for us at this stage to say quantitatively which mechanism is more important. I suspect all of them contribute.
Q: In your research, you mention that our society has implemented some policies that expose people of different backgrounds to one another, but exposure does not necessarily result in economic mobility. Why is exposure alone not enough?
JS: When you think about why some places have more economic connectedness than others, you can point to two main mechanisms. One is exposure, or how many rich people there are around you, in your immediate neighborhood or in your school, church or social clubs.
The other mechanism is what we call friending bias. This happens when rich and poor people go to the same high school or live in the same neighborhood, but the friendship networks still remain fairly segregated. Sociologists call this homophily. It’s the idea that you gravitate to people who look like you, or who share your race, gender or social class.
We have seen a lot of public policies to promote exposure. This includes ideas like building public housing near wealthy areas or implementing school busing to create more diverse schools. And that’s super important, because exposure is a necessary condition for connectedness. But our research finds that the second component — friending bias — is equally important. It’s not enough to just have rich and poor kids show up at the same school. You need to make sure that they become friends with each other as well.
There’s a whole range of things you can do to reduce friending bias, and some of these things may be more immediately attainable compared with the large-scale changes often necessary to increase exposure. For example, we found that smaller schools have less friending bias. When schools are very large, it becomes easier for cliques of similar people to form. When schools, cohorts or classrooms are smaller, everybody becomes friends with everybody and schools become much more integrated.
Sports, from a team spirit perspective, seem to be very integrating. In high schools where people talk a lot about pep rallies and team spirit, we saw more integration across class lines. Participating in sports can also be integrating, but we need to be careful that sports don’t become too expensive, with, for example, lots of travel. Extracurricular activities that are expensive are generally correlated with increased friending bias.
There’s some evidence that schools with school uniforms might be schools with more integration. The physical design of spaces may also play a role. We found examples of schools with very high friending bias that are also running multiple cafeterias, and only one of the cafeterias serves free and reduced lunch. So the kids who are eligible for free and reduced lunch hang out in one cafeteria and the wealthier kids hang out in another, and that starts creating separate social groups.
It’s within high schools that we’ve done most of the analysis, but you could imagine the same type of analysis within communities, public spaces, public parks, museums — any place where people of different characteristics can interact. Those are places where I think friending bias is likely to be lower.
Q: Given the level of economic segregation in the U.S., what types of strategies do you think seem promising for increasing these types of cross-class connections, without changing a community’s residential makeup?
JS: I think it is about creating spaces where people will interact. One thing we found was that people exhibit much less friending bias in their recreational activities. For example, people go for after-work soccer, and this activity brings everyone together because it is something they have in common.
Friending bias is also extremely low at churches. That doesn’t necessarily mean churches are great for economic connectedness. They are often very segregated. Rich people tend to go to one church and poor people tend to go to another church. But if a poor person shows up at a given church, they are essentially equally likely to make friends with a rich and a poor person at that church. And again, I think that’s because they now have something in common that can overcome these other differences. Something that bridges the socioeconomic divide rather than making it more visible. So thinking of ways to create a common identity across class will help generate economic connectedness.
Q: How can parks, libraries and other public spaces bring people together across economic backgrounds? What should cities be thinking about in terms of the design, programming or management of these shared places?
JS: We don’t quite know yet, but one crucial thing to highlight is that we have produced a free, accessible data set and visualization tool called the Social Capital Atlas. It shows the exposure and friending bias for every zip code, high school and college in the U.S. Our hope is that people use it to learn what works and what doesn’t within their own community.
For example, a high school principal could check out their own school and then look at five or six neighboring schools. And perhaps one neighboring school has a really low friending bias. The principal could then try to uncover what’s going on in that school, and learn from it and communicate what they’ve learned to others.
It’s the same idea with public spaces. There will be some zip codes where friending bias is very low, and a mayor or a community group could say, let’s go into this community, understand what they’re doing, what’s happening there and what we can learn.
Our hope is that, with the data free and available, everybody can look at their own community, understand what is working well and then build best practices based on exactly these types of questions. How do you learn from the data to design public spaces in a way that wasn’t possible before? Our hope is that people are going to use the data to think about interventions — to both study what works and implement new ideas.
Reimagining the Civic Commons is a collaboration of The JPB Foundation, the John S. and James L. Knight Foundation, The Kresge Foundation, William Penn Foundation, and local partners.
