Who’s In Your Inner Circle?

The transcript below may be for an earlier version of this episode. Our transcripts are provided by various partners and may contain errors or deviate slightly from the audio.

Shankar Vedantam:

This is Hidden Brain, I'm Shankar Vedantam. Look up at the sky in most places in the world and you will see flocks of birds. Look carefully and you'll see these flocks are usually made up of the same kind of birds. Pigeons flap around with other pigeons, crows with crows. Starlings congregate with, well, lots and lots of other starlings.

Somewhere along the way, someone coined a memorable phrase: birds of a feather flock together. Sociologists have noticed that this phenomenon is not restricted to birds. Humans tend to do the same thing, except because we are complex creatures, the things that draw us together are not just physical similarities, but cultural constructions such as social class, shared interests, and common cuisines.

We think nothing of the fact that music lovers hang out with other music lovers, or that book readers congregate in book clubs, and that gardeners form their own societies. The internet has made it possible for people with all manner of niche interests to discover others who are exactly like them.

When we do think about this at all, we usually say it's a wonderful thing. Well, it is, and it isn't. This week on Hidden Brain, the unintended consequences of enjoying the company of others who are just like us.

The next time you attend a wedding, take a close look at the couples around you. The couple getting married, sure, but all the other couples, too. How many couples do you think have both members vote for the same political party, or belong to the same race, or speak the same language?

The answer is obvious. Most people married to each other have similar backgrounds. They are usually of the same race or ethnicity. They might even have attended the same schools, or work in similar professions. This hasn't always been the case. The old trope on television was the business executive or law firm partner, invariably a man, who ran off with the secretary.

The new reality is that law firm partners tend to marry other highly-paid lawyers. Doctors get hitched to other medical professionals. Especially in well-to-do communities, it's vanishingly rare for people to marry others from starkly different backgrounds. Matches that span significant educational divides, where one person has gone to, say, graduate school while the other hasn't finished high school are extremely rare.

At Stanford University, Luigi Pistaferri studies the consequences of such matching. He uses the usual tools of the economist: models, math, and data. But besides being an economist, Luigi is also a Jane Austen fan.

Luigi Pistaferri:

The most important novel of Jane Austen, Pride and Prejudice, is centered around the fact that the Bennet family wants to have all the daughters be married to people from hopefully higher social classes.

Shankar Vedantam:

You might remember the story. It centers around one of the Bennet daughters, Elizabeth, and her unlikely relationship with Mr. Darcy, a very wealthy bachelor.

Alison Steadman, as Mrs. Bennet:

His name is Darcy, and he has a mighty fortune and a great estate in Derbyshire.

Shankar Vedantam:

Their romance is unlikely because Elizabeth is supposed to be well beneath Mr. Darcy's social class. When the seemingly mismatched lovers finally overcome many hurdles and get together, we are left with the treacly conclusion that love conquers all.

Colin Firth, as Mr. Darcy:

I have been a selfish being all my life. As a child, I was given good principles but was left to follow them in pride and conceit. As such I might still have been but for you, dearest loveliest Elizabeth.

Shankar Vedantam:

As a Jane Austen fan, Luigi loved the story, but as an economist, he doesn't buy the plot.

Luigi Pistaferri:

It's the comical aspect of it that is interesting, together with the fact that it seems from watching the movies that the Bennets are actually a lower social class, when in fact they were probably themselves in the top 10% at the time. The three, four girls, I don't remember how many sisters there were, they were all literate, and they could have servants in their house, already tells you that the Bennet family actually is not poor. But given the enormous amount of wealth inequality there was at the time, it looks as if they're in lower ranks of the society when in fact they were pretty high up.

Shankar Vedantam:

Luigi says that if you think about Pride and Prejudice through the lens of an economist, in fact, if you think of love and marriage through the lens of an economist, you start to see things differently. You start to see lots and lots of what sociologists called homophily: birds of a feather flocking together.

Luigi Pistaferri:

We economists tend to think in terms of markets, so here I'm referring to what we call the marriage market. The marriage market homophily means someone marrying someone similar to themself. That could be cast in terms of wealth, it can be cast in terms of levels of education, it can be cast in terms of income. People from relatively high social classes try to marry someone from the very top of the distribution social classes. That would be Mr. Darcy, for example, in the novel. The research in economics is a form of homophily. We call it assortative mating.

Shankar Vedantam:

Assortative mating is what you see when doctors pair up with other doctors, or lawyers marry other lawyers. It goes back to that idea we discussed that most people tend to choose partners who are similar to themselves in background, in education, in wealth.

Several powerful forces drive assortative mating. Think about how and where people tend to meet romantic partners. On dating apps, you can filter potential matches based on education, professional background, age, and physical appearance. And the places where people are most likely to meet potential partners in real life? Those tend to be school and work. Situations where you are overwhelmingly likely to find others with similar interests and educational backgrounds.

Luigi Pistaferri:

There are frictions, what we call frictions, costs of actually exploring the universe of potential mates that you can have. And so, given that you don't have a lot of time to do this, you tend to marry people you encounter, or you meet with in places where you spend most of your time.

It could be the workplace, it could be the place where you went to college, and indeed, we observe a lot of the matches actually occurring, happening through these two channels. There's a lot of people who end up meeting the person they marry at work or in college. This is a very common occurrence.

Shankar Vedantam:

On a recent visit to Chicago, I heard that 50% of University of Chicago undergrads end up marrying someone else from the same school. Now, that number seems improbably high, but the fact remains: homophily is a real thing in the marriage market. What's striking, Luigi says, is how movies and popular culture celebrate a world that is very different than reality. He cites a famous movie from 1990.

Luigi Pistaferri:

The modern Cinderella story is Pretty Woman, the movie with Richard Gere and Julia Roberts.

Julia Roberts, as Vivian Ward:

Hey sugar, you looking for a date?

Richard Gere, as Edward Lewis:

No, I want to find Beverly Hills. Can you give me directions?

Julia Roberts, as Vivian Ward:

Sure, for five bucks.

Richard Gere, as Edward Lewis:

Ridiculous.

Julia Roberts, as Vivian Ward:

Price just went up to 10.

Richard Gere, as Edward Lewis:

All right, okay. All right, you win, I lose-

Luigi Pistaferri:

So, the story of a very wealthy man meeting almost randomly the very poor girl, and then eventually, essentially, marrying, I think. That's what the movie suggested. That is an example that I use to show that the probability of that happening in the data is extremely low, although it's not zero.

Shankar Vedantam:

But of course, we call it a Cinderella story precisely because it's unusual.

Luigi Pistaferri:

Yeah, it's a sort of underdog stories. We all love underdog stories, but that doesn't make them very real. So, they're very unusual and rare occurrences. So unusual and rare that we make movies about.

Shankar Vedantam:

How did you come to be studying assortative mating? Why were you drawn to this topic? What interests you in it?

Luigi Pistaferri:

I study wealth inequality and why wealth inequality and wealth concentration is rising over the last 20 to 30 years, and one thing that is often cited as an element that may have contributed to a rise in wealth concentration is the marriage market. So, people with wealth marrying other people with wealth exacerbate the extent of wealth inequality or wealth concentration that you see in a society.

Shankar Vedantam:

Some time ago, Luigi spotted some very unusual data that allowed him to quantify the extent of assortative mating in society. The data came from Norway. Why Norway? Well, in most countries, it's hard to find precise data on the socioeconomic backgrounds of people. Most of us don't broadcast our bank balances, and the ones who tell us how rich they are are often narcissists who might be lying. But in Norway, banks directly report to the government about how much citizens have in their accounts.

Luigi Pistaferri:

And then the other thing that is important is that Norway has a wealth tax, unlike the US, unlike many other countries. So, the tax authority collects very detailed information on the assets you own, and pretty much all the assets you own. So, it could be housing, it could be the value of your stock market portfolio. It could be the bonds, it could be the value of private businesses, et cetera. So, you have a very exhaustive information on the assets that people own, and also the debt that they have.

Shankar Vedantam:

Luigi and his colleagues collected 11 years' worth of tax data. It included the income and assets of over five million people. The rich, the poor, and everyone in between. Next, they looked at marriage records.

Luigi Pistaferri:

So, what we did, first of all, was to look at the wealth of the assets that people have years before they get married. One advantage of this data is that you observe people before they enter formal marriage relationships, so you know eventually if you're going to marry X, you know how much income and wealth that X person has before the marriage is observed.

Shankar Vedantam:

What did you find first in terms of who was marrying whom in Norway?

Luigi Pistaferri:

So, we find that if you rank people according to the wealth they have before marriage, then someone who has a wealth of 100 will tend to marry, with a high probability, someone with a wealth of 100. Similarly, someone at the bottom will tend to marry someone who is similar to them. In that sense, we find strong evidence of assortative mating.

Shankar Vedantam:

I just want to spend a moment talking about how this actually unfolds in real life. I mean it's possible that people in Norway are actually asking one another, "How much money do you have in your bank balance?" And basically saying, "Okay, you're on my list, you're off my list." But I suspect it doesn't happen at that level where someone's saying, "Show me your bank balance and I'm going to decide whether to offer my hand in marriage or ask for your hand in marriage." I don't think it happens at that level.

How is this happening if those conversations are not transparent and on the surface? How is this matching actually taking place do you think, Luigi?

Luigi Pistaferri:

It's interesting you make this comment because, actually in Norway it does happen that people have a lot of information, because in Norway tax records are actually public.

Shankar Vedantam:

Oh, I see. They're not just shared with the government, they're actually public.

Luigi Pistaferri:

Yeah, so if you have access to the internet and you have a social security number registered with the Norwegian tax authority, you can go on the web, enter the name of the person you're looking for, and that would spit out information about assets, the income, the taxes they pay, the year they're born, where they live and so forth.

So, you have a pretty... If you want... I mean the interesting thing is that the government at some point realized that this was a little bit too intrusive, even by Norwegian standards, and they changed the rules. The rules is now, yes, you can still search, but the person you search will know who's searching for them. So, this discouraged almost immediately a lot of the searches. Your neighbor, your colleague, and something like that.

Shankar Vedantam:

Of course, what happens in Norway doesn't happen in most places. But Luigi says you don't need to access someone's detailed financial information to glean how wealthy they are.

Luigi Pistaferri:

So, you can infer how wealthy someone is by the car that they drive, or the kind of clothes that they wear, or things of that sort. Or the type of education they have or what they can afford, where they go on vacation, et cetera. So there's a lot of signals that display wealth that comes from the things that you're spending your money on.

Shankar Vedantam:

One of the other things that you found in your paper, fascinatingly, was not just wealthy people pooling their wealth together, but what happens to that wealth afterwards. Can you talk about what you found, Luigi?

Luigi Pistaferri:

We find that people who are really good at generating high return for their assets tend to marry people who are very good at making a high return from their assets. There seems to be assortative mating of returns on top of the assortative mating on wealth.

Shankar Vedantam:

Can I just ask you to try and speculate on what could be driving that? Because let's say I show up at a date in a very fancy car, and my date shows up at the date driving a very fancy car, we can both infer that the other person is wealthy. But how in the world would I know that my date actually has a high rate of return on his or her investment, and how would my date figure that out about me?

Luigi Pistaferri:

If you date someone, you typically don't marry a person after one day, after going out one night. I mean, yes, there are the shotgun marriages in Vegas, but we're not talking about that. We're talking about normal people.

Normal people date, sometimes cohabit for a long time, so they learn a lot about what the other person does with his money before they actually merge their wealth. I use the example of Lady Mary in Downton Abbey. Lady Mary at some point has to marry, and because it's a society where she will marry some other rich guy, assortative mating on wealth is guaranteed.

So, the question is, she has two potential suitors. One is a guy who is really good at managing his estate, and the other one is really someone who is an aristocrat. He's very wealthy, but he wastes his money in different ways.

And so, the question that we ask rhetorically is, okay, should she marry the guy who is not very good at managing his wealth? And the answer is, and this is something that we provide evidence for, the answer which kind of person should she pick is, well, it depends on who's going to end up managing the resources of the family after the family's formed.

In those families where, of course, there is a lot of wealth, you have greater incentives to have efficiency. So, to have the portfolio being managed by someone who's really good to start with.

Shankar Vedantam:

What Luigi is saying is that when wealthy people marry other wealthy people, the shared family wealth is invariably managed by the person with a better head for money. Maybe someone with an entrepreneurial streak or a sophisticated investor. Assortative mating ensures that with each generation, wealth not only merges with wealth, but that wealth accumulation accelerates on top of wealth, and this has profound consequences.

Luigi Pistaferri:

When you merge wealth, you have immediately a scale effect. You have accelerating forms of wealth inequality, so this is an explosive form of wealth acceleration. That's exactly what people have seen. If you look at work by economists like Emmanuel Saez and Gabriel Zucman, that is exactly one of the findings they have for the United States in the last 30 years. The degree of concentration, the wealth distribution has been accelerating at a very faster rate.

Shankar Vedantam:

I'm wondering besides your academic work and your research, do you see the effects of assortative mating around you, Luigi? Do you notice that in your social circle? I mean I feel like as I read this paper, I started to think about the people whom I know, and it's remarkable how similar so many of the couples are whom I know. This is whether they're older or younger, people tend to come from the same background. Is that the same for you?

Luigi Pistaferri:

Oh, yeah. I mean we have extreme cases, like professors in the department being married to other professors. So, that's an extreme form of assortative mating on education. In fact, I have students working on that. I think it's very rare for people who get to the top of the distribution of education to be married with someone from high school degree or high school dropout. If you look around you and you find some example, you know that these are very rare exceptions. So, yeah, assortative mating is everywhere.

Shankar Vedantam:

When we talk about wealth inequality, we typically think of systemic causes: the housing market, student debt, the tax system. We rarely think about how our own decisions, decisions that are very personal involving whom we love and whom we marry, could accelerate inequality.

When we come back, if our personal connections can exacerbate the problem of inequality, surprising new research suggests personal connections might also be the answer to the problem. You're listening to Hidden Brain, I'm Shankar Vedantam.

This is Hidden Brain, I'm Shankar Vedantam. When two people with similar interests, backgrounds, and temperaments get together romantically, we're often inclined to say the couple is made for each other. What is harder to see are the downstream effects of what is known as assortative mating.

Our tendency to connect with others like ourselves can greatly exacerbate inequality over time. This may be one reason why research by the Harvard economist Raj Chetty and his colleagues has found that where you were born can shape your outcomes in life. If you were born in a poor neighborhood and hang out with other people who are poor, it's a lot harder to climb the socioeconomic ladder than if you were a rich person born into a wealthy community. Here is Raj from our episode some years ago called Zip Code Destiny.

Raj Chetty:

Think about upward mobility in the context of the classic rags-to-riches notion. Suppose you're a kid growing up in a low-income family, say the bottom fifth of the income distribution. What are your odds of reaching the top 5th of the income distribution? Making that leap from the bottom to the top. So, it turns out if you look at the US as a whole, that number is 7.5 percent. 7.5 percent of kids who start out in a relatively low-income family will end up reaching the top fifth on average.

Shankar Vedantam:

That doesn't paint a very positive picture of economic mobility in the United States. Recently, however, new research suggests a way out of the trap. To understand what it is, I want to take you back to the early 1980s and the story of a man named Matthew Jackson.

Matthew Jackson:

So, this was actually in late high school and then the first couple of years in college. I was working in warehouses in Chicago, mostly unloading trucks. It was one of the few jobs that you could get without a connection. I needed the money to help pay for college, and so, it certainly wasn't a glamorous job.

I mean, it was what you would expect, so it was a warehouse that sold catalog stuff, and we would ship televisions, household wares, diapers, food products. All kinds of things would come in these trucks and our job was to sit there for hours and wait until a truck came and then get the truck unloaded in half an hour and then wait for the next one. It was hard work physically, but mentally it was just extremely boring.

Shankar Vedantam:

Matthew's managers had, shall we say, a less-than-trusting attitude toward him and his fellow blue-collar workers.

Matthew Jackson:

We had lie detector tests that we were given every couple of weeks, because I think the main thing that the company actually worried about was losing... They lost a lot of money to employee theft. So, it's easy for somebody unloading a truck to just unload something to the side and get away with it.

Shankar Vedantam:

Yeah, the proverbial bag falling off the truck, as they say.

Matthew Jackson:

Exactly. So, it was clearly a position that changed my perspective on work and what I wanted to do with my life.

Shankar Vedantam:

Matthew loved academics and wanted to do work that he found stimulating, but that turned out to be really hard to find.

Matthew Jackson:

And so, when it came to college, paying for a private university, there was a lot more money needed than we could afford. And so, I worked during high school. I worked these trucking jobs, and actually, on weekends I would work a night watchman shift at a local medical center. That was the best-paying job I ever had at that time, which was I'd go in at midnight and then stay there until 8:00 AM and just walk around the building making sure nobody was breaking in.

So, it was jobs like that that were available without connections, and so, they were jobs that paid and helped save money for universities.

Shankar Vedantam:

That must have been pretty deathly boring as well, right? Walking around from midnight to eight o'clock in the morning to make sure no one's breaking into a building.

Matthew Jackson:

Yeah, the worst was a day where you did a shift of unloading trucks and then went to the medical center overnight, and by the next morning, you were just completely wiped out. It was tough.

Shankar Vedantam:

Like many people without connections, Matthew didn't have a way of breaking into the kind of job he knew he would be good at. But that changed one day after he got to college. He was having a chat with his undergraduate advisor Hugo Sonnenschein about his summer plans.

Matthew Jackson:

So, I was having a talk with Hugo one day in his office and he said, "Well, what are you going to do this summer?" And I said, "Well, I'll go back to Chicago and unload trucks." He said, "No, no, no, no, no. We're going to get you a job that you'll actually learn something from." And he literally picked up the phone while I was in his office and called a friend of his in Chicago, Galen Burghardt, who had been a friend of his from his years at University of Massachusetts Amherst.

And said, "I've got this undergraduate, he needs some work. He lives in Chicago." Galen was working at the Chicago Mercantile Exchange and Galen said, "Yeah, we can figure out something."

Shankar Vedantam:

When Matthew returned to Chicago, he started working at the Chicago Mercantile Exchange. All of a sudden, he realized that his math skills could be put to use in the field of economics. The summer job proved to be transformative.

Matthew Jackson:

It was just a incredible change in terms of the kind of work I was doing, and I was actually doing research for them, which eventually became a paper that I wrote and played into the research I did as a PhD student. It was really life-changing in many ways.

Shankar Vedantam:

That chance conversation with his advisor changed the trajectory of Matthew's life. He went on to become a respected economist, eventually finding his way to Stanford University.

Matthew Jackson:

I think it's easy for us to underestimate how much our trajectory in life is determined by the connections that we have in terms of friends and family, and how critical those are because it's not like every day these matter, but occasionally they matter, and then they matter in a very big way.

Without that connection, it's possible that I wouldn't have done that work in the summers and I wouldn't have had that experience, and it wouldn't have opened my eyes to research possibilities. I wouldn't have done a PhD. It put me on a different road. Without that, I probably would have gone back and worked in a warehouse another summer, and then who knows what would have happened after that.

But those kinds of connections, they're infrequent in terms of the changes that they make, but they can be big changes at those points in time. Whether or not you get your foot in the door or not is really so vital, and it really depends on our networks.

Shankar Vedantam:

So, another element of the story that jumps out at me is that Hugo and Galen went to school together, and of course, they didn't go to school together in order that one day Hugo would introduce you to Galen. They didn't plan 25 years ahead, so this was a chance... The fact that they were going to school together was probably driven by chance as much as anything else, but that chance encounter that they had ended up playing a pivotal role in your life.

Matthew Jackson:

Yes, yeah, and I think that's one thing that is important to understand about our networks is that they're not necessarily optimized, in the sense that connections that I make give me information that can help my students and can help my colleagues. When I'm thinking about, "Oh, should I go to this conference? Am I going to learn something new?" I'm usually thinking about this for myself. I'm not thinking, "Oh, maybe I'll learn this thing, which I can actually tell so-and-so about."

And so, the full value of these kinds of interactions and the importance of reaching out. The value of that to the greater society isn't something that's necessarily governing people's choices when they're deciding, oh, should I go and do this, or should I go and meet this person, or does it make sense to call this person? Do I follow through with this friendship and so forth?

So, there's a lot of things that we do that aren't optimizing on that. These serendipitous relationships matter a lot and can have a lot of consequences, but they're not something that's fully accounted for when we form our relationships. So, that means that somehow, we're all under-forming and under-broadening our networks and enriching ourselves.

Shankar Vedantam:

What Matthew was pointing out is that social networks have ripple effects. I may think that my choice of friendship influences only me and my friends, but in fact, those choices can profoundly shape the lives of other people down the road. When we come back, if our friendships and romantic attachments can unintentionally be the source of division and inequality, Matthew's research has found that they can also be drivers of connection and opportunity. You're listening to Hidden Brain, I'm Shankar Vedantam.

This is Hidden Brain, I'm Shankar Vedantam. Some of the biggest problems in the world today are the result of mismanagement and corruption. But there are also many serious problems that stem from the internal architecture and predispositions of the human mind. One of those powerful predispositions is that we gravitate toward people who are similar to us in some way. It's a phenomenon known as homophily. At Stanford University, Matthew Jackson says this happens so regularly in daily life that most of us don't even realize we do it.

Matthew Jackson:

If you walk into a conference or a room or a meeting where there's people that you don't know, one of your natural tendencies is to start trying to find common ground. You ask people questions, and then once you find something in common, you really zero in on that.

If you find that you both came from the same state or hometown or that you both went to the same school at some point or that you both know somebody in common, those become points where you feel much more comfortable because now you've got something to talk about and something in common. It helps you understand the other person, it helps you understand their perspective, it reduces uncertainty.

Psychologically, it's just much more comfortable to be in a situation where you understand the other person and you know what their background is and their viewpoint, than meeting somebody who you don't really understand at all and might have very different perspectives than you. That's a harder situation to be in. And so, we naturally tend to those things, even though that's not necessarily the best thing in our long-term interest.

Shankar Vedantam:

So, I want you to talk a little bit about that, which is that these forces are in some ways such natural forces of feeling saying, "I want to be comfortable around other people, and someone who basically shares my outlook or likes the same sports team or comes from the same geographic background or has the same racial background or income background. I feel more comfortable around this person, and so, I'm going to spend time with them." That very natural thing ends up playing this really consequential role when multiplied across an entire society.

Can you talk about how this homophily essentially ends up segregating groups of people and sealing off aspects of society or access to income mobility for people who are especially poor?

Matthew Jackson:

Yeah, so I think one thing that's pretty amazing in the U.S. is that there's a lot of heterogeneity. Different parts of the country have very different mobility patterns. You're much more likely to advance if you live in say the Upper Midwest than if you live in the Deep South. There are places where you see connections across class lines, but there's a lot where you don't.

There's a huge amount of homophily by economic class, and that tends to be in areas where you have very low mobility. When you look at it in terms of how predictive are parents' income of children's income, the U.S. doesn't do so well. The chance that you look like your parents is fairly high. That means that for some reasons, children, their future is heavily determined by the conditions that they're born into, and the situations that they're born into.

Shankar Vedantam:

At a fundamental level, there's something about that that rankles, doesn't it? Which is that I think many of us imagine that we should not be constrained by the circumstances of our birth, by the accident of birth. We think that that should not determine what happens in terms of the outcomes in life. But really, what you're saying is, accidents of birth are playing an enormous role in the likelihood of people's economic and social success in a country like the United States.

Matthew Jackson:

Yes, definitely. In fact, I think when we think about the philosophy of it and the moral aspects of it, when people object to inequality, a lot of what we're objecting to is unequal opportunities because it's saying you were just born into circumstances, no matter how hard you worked, you were going to be behind.

That's different than saying, "Oh, so-and-so happened to get lucky," or, "So-and-so is really talented." Obviously, Michael Jordan can play basketball at a level that other humans can't, and we don't envy him... Well, maybe we envy him, but we don't say he doesn't deserve what he earned from that. That's very different from saying somebody who was born in a community and never got a chance because of where they started.

Shankar Vedantam:

As they try to figure out how something works or how to fix a problem, scientists sometimes look for what's called positive deviance. Let's say a virus is spreading. You can study people who fall sick to figure out the nature of the pathogen, but you can also study people who don't fall sick, the positive deviants, and ask what their secret is.

If you can figure it out, the insight can help you keep other people from falling sick. In this case, the word deviant is not being used as a pejorative, but as something of an accolade. Matthew and other researchers knew from the data that most kids who were born poor don't break into the ranks of the rich, but a small number do. What was their secret?

Matthew's own life had been transformed by a chance conversation and the fact that his college advisor was friends from his college days with someone who worked at the Chicago Mercantile Exchange. Matthew had a hunch the same thing on a massive scale might be connected to differences he was seeing in income mobility. To test this idea, he turned to Facebook.

Matthew Jackson:

Facebook was really valuable on several dimensions. So, we were actually able to look at 84% of the population of the age group we were looking at. Secondly, we can see all the friendships of these people and really see who are their high school friends, who are their college friends, who are their work friends. So, we can actually classify where their friendships are formed, where their networks look like.

We can also begin to see a lot of demographics about the people, so we can estimate their income level and get information about exactly where they live and so forth. So, just putting all that information together is something that's pretty rare to have, and it allowed us to see how those social connections actually translate into the economic outcomes.

Shankar Vedantam:

When you look at this pattern of friendships, what do you find in terms of homophily?

Matthew Jackson:

Yeah, so the main thing that pops out in terms of the homophily that we were looking at was, take somebody who's below median income and ask what fraction of their friends are above median income? If it was a completely well-mixed random world, 50% of their friends should be above median income. It's just over 38% on average.

If you look at people who are above median income, it's above 70% of their friends are above median income. So, there's a huge gap between what fraction of people's friends are above median income depending on whether you're above or below yourself. And so, that level of homophily is pretty strong.

Shankar Vedantam:

How did you measure people's income using Facebook data?

Matthew Jackson:

We inferred it. We have locations on people, and then we have a lot of information about people: what phone they use, if they did go to college, where did they go to college, and so forth. We used 22 variables, I think, in the end, and a machine learning algorithm that could basically estimate them. You could put that together with census data. Census data goes down to a block level, so you have a very fine grade of where somebody's located, and the interesting thing about that is that if you're looking in the U.S. and you look down to a block level, you pretty much can nail somebody's income. So, putting that census data together with various variables we could observe with people and the locations, we could build a model that said, "Look, if you tell us these characteristics about a person, we can tell you pretty much what their income level is."

Shankar Vedantam:

So, at one level, what you're finding in some ways using this very large sample set is essentially a large-scale replication of something that we've already known on a small scale, which is that people tend to cluster together. People who are poorer tend to have more friends who are poor and fewer friends who are rich. People who are richer tend to have more friends who are rich, fewer friends who are poor.

But you didn't just stop there. You actually went somewhere further with the study. Walk me through what you did now with all of these connections that you've established that people have or don't have.

Matthew Jackson:

Yeah, so the base question that we started with was, what is it that predicts whether somebody who's born into the poor half of the distribution could end up in the top half of the income distribution? And in particular, there's lots of different things about a person's network and community structure that you could imagine would matter. Things that are commonly referred to as social capital. We know financial capital matters, if you have wealth, that helps you. Human capital, if you have parents who are educated, that's going to help you. But social capital. What is it about your network that matters?

Shankar Vedantam:

Three possibilities jumped out.

Matthew Jackson:

One would just be, okay, does the community all know each other? So, we can look at that in the Facebook data and you can just look at, okay, if I have two friends, are they friends with each other? And you can see how tightly-knit communities are in terms of that clustering, so that's something we can measure directly.

Shankar Vedantam:

If poor kids from tight-knit communities were the ones who made their way into upper-income groups, the solution to improve income mobility would be to encourage greater community cohesion. Possibility two: a higher degree of generosity and altruism in the community. If you come from an especially kind place, maybe there are people in groups who can be stepping stones to help a poor kid succeed. The nationwide Facebook data contained the answer about people in different communities.

Matthew Jackson:

You can see on their Facebook page, are they part of a volunteer organization? Are they part of some NGO or charity? When you look at things like the kinds of trust or volunteering rates and so forth, those are often associated with how well functioning are the local organizations, and do we have good institutions in our area. So, those are things that we often think would help people out in terms of economic mobility.

Shankar Vedantam:

There was a third possibility to explain why poor people in some places were better able to transform their economic station in life. These places had more connections between rich and poor people. They had bridges between groups that counteracted the effects of homophily. In these places, people were more likely to talk to people who came from different backgrounds than themselves. Poor people and rich people counted one another as friends or acquaintances. Matthew called this economic connectedness.

Matthew Jackson:

You can find communities where you have no connectedness, but you have very strong clustering, or you can find places where people are very civically minded, but people don't cluster. You can find each of these in different combinations across the country. In the end, it turns out that the only thing that really was a strong predictor was this economic connectedness of whether or not there was friendships between the poor people and the wealthier people.

Shankar Vedantam:

Matthew was blown away by the strength of the data.

Matthew Jackson:

I think one of the surprising things to us, at least to me, was how strongly the economic connectedness ended up predicting somebody's mobility. That means if you're in a community where you have strong economic connectedness, you're very likely to be in a community where the poor have a chance of growing up to have above-median income or is rising.

That correlation also withstood all kinds of other things, so you could put in whatever variables you think of: how poor is the area, how unequal is the area, is it urban/rural? What's the percent of White versus Black versus Hispanic? You can look at all kinds of different variables, and none of those really had nearly the same kind of power that this did. I think one striking thing was how powerful that is as a predictor.

Shankar Vedantam:

In other words, correct me if I'm wrong, but if I was a poor person in the United States and I wanted to ask myself how can I extricate myself from my economic station and move to a better station in life? And I had to choose between living in a community that was very cohesive, where everyone knew one another, people in some ways bonded together very tightly, or in a community that had very good social service organizations and charities.

Or I lived in a community where I could actually be friends with people who were wealthier than I am. What I'm hearing you saying is that my odds would be the highest of actually changing my social station if I was in that last community, if I had friends from different economic strata.

Matthew Jackson:

Yeah, overwhelmingly. In fact, how well people know each other within your community and whether you have well-functioning civic organizations and connections to those are actually sometimes even negatively correlated with the outcome in terms of mobility.

I think part of what's actually happening here is that in those poorer communities then you see more of that cohesiveness because people are helping each other out, and you see more volunteering, or you might see more volunteer organizations appearing because they're needed, and they provide vital help and informal support and services that you don't have otherwise. So, they're still useful in some ways, but they're not helping poor kids grow up to have a chance at being not constrained and end up where their parents were.

Shankar Vedantam:

How much of a difference does economic connectedness make in a community?

Matthew Jackson:

Yeah, so the difference in outcome if you take a typical poor kid and give them the same friendships network that a typical rich kid would have, that predicts a 20% increase in their lifetime income. It's very hard to find any variable that has that kind of magnitude of change in lifetime income.

Shankar Vedantam:

What do you think is actually happening at a granular level here when people have these cross-SES, cross-socioeconomic status friendships? What exactly is happening? Are they just getting contacts? Are they getting leads to jobs? What do you think is actually changing their life outcomes?

Matthew Jackson:

I think as with many things it's a combination that matters. There's those vital opportunities that might be rare but can matter. It could be 10 years down the line that I have a former high school friend and I'm out of a job and I can call them up and say, "Oh, I know you work at this bank, could you get me a job there?" And they can say yes.

So, it could be something like that, but it could also be just the information that we have when we're in high school, and am I seeing other kids go to college? Did I see what it took for them to go to college? Do I see them actually getting jobs afterwards and what do I see them doing? What do I envision myself doing? A lot of that is formed by that community, and I think when you put all these things together, especially at an early age, they just amplify themselves, right?

So, the more you study, the more you can study, the more chance you have of advancing. The more friends you're going to make who are also taking classes that are preparatory for going to college and so forth. It all plays into itself and then it continues throughout your lifetime, in terms of those friendships then giving you opportunities and everything else.

I think just all of us can try to think a little bit about what is it that's limiting our networks? What can we do? Where are there opportunities for us to step out a little bit from what we normally do and broaden those horizons and bridge across whatever gap it is that might matter?

Shankar Vedantam:

It's natural to want to spend time with people who are like us. It doesn't make you a bad person to join a book club or a birdwatching group and find others who share the same interests, backgrounds, and temperaments. But this very human impulse magnified on the scale of a nation has the propensity to create sealed economic caste systems where it becomes very hard for people who are poor to change the trajectory of their lives. When we reach out to form connections with people in different economic stations, it has the potential to transform our lives, their lives, and the lives of strangers in the far future.

Hidden Brain is produced by Hidden Brain Media. Our audio production team includes Brigid McCarthy, Annie Murphy Paul, Kristin Wong, Laura Kwerel, Ryan Katz, Autumn Barnes, and Andrew Chadwick. Tara Boyle is our executive producer. I'm Hidden Brain's executive editor.

For today's unsung hero, we're bringing you a story from our sister show, My Unsung Hero. Today's story comes from Sanaa Kerroumi. Some years ago, Sanaa was living in France, working on a PhD. One cold winter weekend, she and her friends took a break and spent the day at Disneyland Paris. They stayed as long as they could, until it was finally time to catch the last tram home. One by one, all of Sanaa's friends got off at their stops until she was the only one left.

Sanaa Kerroumi:

And I noticed that the tram was super empty, there were just two people in there, two guys, both of them French. And I wear the hijab, so I really stand out. Wherever I go I do stand out.

Shankar Vedantam:

One of the men was in the front of the tram. The other sat a few rows behind Sanaa. She didn't feel like either of them were paying her or her headscarf much attention, so she put on her headphones and looked out the window.

Sanaa Kerroumi:

Something, maybe the switch, the time between music or something like this, when you have the small silence, I heard something, and I turned, and I saw this guy just being very agitated. He was waving his hands and swearing and just pointing in my direction, and I wasn't sure if it was me or the other guy because I didn't do anything that may trigger him to do that.

So, I assumed that he was probably talking to the guy behind me. I took my headphone off, and that's when I heard all those very harmful words that he was saying. That we took the jobs, that we are making Europe uncivilized, that we are everything that is wrong with Europe and France. At that moment, I was like, "Yes, this is about me. This is definitely about me."

I wasn't scared. I was just numb and shocked. I didn't understand why is this happening? It didn't even cross my mind that I would be in danger, but I was like, "Why? Why? Why? Why? What did I do? Why?"

Shankar Vedantam:

Just as Sanaa began to stand, the man who had been sitting behind her rushed to the front of the tram and placed himself between her and the angry man.

Sanaa Kerroumi:

Few seconds he was up front of me, and he kept on pushing the guy to the door of the tram. He kept on saying, "You don't touch her, you don't say those words to her. She's just a student. Everything that's wrong with Europe is you." This phrase I remember it very clearly: "Everything that is wrong with Europe is people like you, not her."

Shankar Vedantam:

The man defending Sanaa kept the angry man from getting close to her.

Sanaa Kerroumi:

At the beginning, he kept pushing him toward the tramway door, and same time yelling at the driver, "Stop the tram, this guy needs to get off at the next station."

Shankar Vedantam:

Once the tram stopped, he pushed the man off the tram and then stood by the doors to make sure he couldn't get back on.

Sanaa Kerroumi:

He stayed there, didn't talk to me. Stayed by the door just watching me. This happens a lot when you are veiled, especially in France. When they see you as veiled, they really don't know how to interact with you. They don't know what are your boundaries? So, I'm assuming that's why he didn't come and talk to me. Although it would have not been a problem at all. I probably would have welcomed it, but I'm pretty sure that that's the reason.

Shankar Vedantam:

Sanaa was too shocked to say anything either. She and the man stood in silence on the tram until her stop came. When she got out, the man followed and made sure that she got into her building safely. Sanaa was still in too much shock to turn around and say goodbye.

Sanaa Kerroumi:

It was the very first time that I encountered something like it in my entire life. I heard stories. Of course, I did. I knew that it happens, but I never assumed that it would happen to me, and at the moment, even what the nice guy did didn't register. I was still so focused on what the bad guy said. Why was he attacking me? What did I do?

And then somehow, I was like, "Ah, but wait, I wasn't alone in this. There was another guy who knew that I didn't do anything wrong, that I didn't deserve the hatred and the violence, and he actually defended me." And then I was like, "Wait, he had no reason to defend me. He wasn't sharing the same beliefs and the same culture, and yet he was there. He actually put himself in danger by doing that. He could have gotten hurt, he could have gotten in so much trouble, and he didn't think about that."

Shankar Vedantam:

Every time Sanaa experienced Islamophobia after that day, she remembered the man who defended her. His kindness and courage gave her the strength to remain in France.

Sanaa Kerroumi:

I don't remember it as the day where I was attacked by a stranger. I remember it as the day where I was saved by a stranger, and that made all the difference, all the difference. And I'm really grateful to that guy for this.

Shankar Vedantam:

Listener Sanaa Kerroumi. Some years after the incident, Sanaa finished her PhD and moved back to her home country of Morocco. When we spoke to her, she was working as an international management consultant. If you would like to help us build more stories like this, please act now. Visit support.hiddenbrain.org and join the hundreds of other Hidden Brain listeners who have signed up to help. Again, that's support.hiddenbrain.org. I'm Shankar Vedantam, see you soon.