If you want to get rich, make rich friends, a Harvard economist found in 2 new studies

• New studies examined how different types of social capital impact someone's upward mobility.
• It turns out that being friends with richer people is the only factor associated with upward mobility.

If you want to earn more, it pays to have friends who are richer than you. That's because being friends with them is an important determinant of your economic mobility, according to two new studies from Harvard economist Raj Chetty.

Chetty and his team examined the social networks of 72.2 million Facebook users ages 25 to 44 to create their own three categories of "social capital." The first is how connected lower-income people are with higher-income people — what they call "economic connectedness."

The next is what they call "social cohesion," which basically gauges whether someone's friend group has cliques, and if friends within that person's network are mutual friends with each other. The third measure is "civic engagement," which looks at whether someone participates in civic organizations like volunteer groups, or trusts them.

The researchers find that, out of those three ways of measuring social capital, the only one actually linked to upward economic mobility is friendships with people from a higher socioeconomic status. In fact, if lower-income kids grew up in areas that have the same economic connectedness as higher-income kids' neighborhoods, their future earnings increase by an average of 20%.

The research comes as upward economic mobility — long viewed as the American Dream — is increasingly out of reach for many. Wealth gains for the lowest earners are being wiped out by inflation. The chasm between the wealthiest and everyone else has only grown, especially as America's highest earners own more and more of the stock market.

People from higher socioeconomic backgrounds tend to have more friends than people coming from a lower-income background

Kids who grow up in areas where there's a higher rate of friendships between people of low and high socioeconomic status "have much higher rates of upward mobility."

For instance, one of the studies contrasts Minneapolis and Indianapolis. In Minneapolis, people from a low-socioeconomic background have a much greater share of friendships with high-socioeconomic people, compared to a far lower share in Indianapolis. Because of that, lower-income kids in Minneapolis make more money in adulthood. By the time they're 35, kids from Minneapolis reach a higher income percentile bracket, making about $34,300 in 2015 dollars compared to $24,700 for kids in Indianapolis.

That also means that, for people living in lower-income neighborhoods, their upward mobility is not impinged upon by simply living in a lower-income neighborhood; it means that they have fewer chances to interact with and befriend higher-income people.

"Changing affordable housing policies, busing policies to get kids to go to the same school, changing school district boundaries, things like that — those may all be very valuable to increase economic connectedness," Chetty said in a press conference. "We think they are, but that's not enough, even if we were to perfectly integrate every school, every neighborhood and so on, we would still have half of the social disconnection between low and high-income people left."

So, if one way to up your economic mobility — or set up your kids for a greater shot at upward movement — is to befriend people from a higher socioeconomic background, then why isn't that always happening? In essence, the researchers find, people with the lowest socioeconomic status make most of their friends by proximity; those are people who live in their neighborhoods.

For instance, people from higher socioeconomic backgrounds make more of their friends in college, which is not accessible for many, especially the poorest Americans. People below the 25th percentile have, on average, $30,575 in student debt — less than some higher earners, but accounting for a greater chunk of their income. But even in college, Americans from a lower socioeconomic background are less likely to befriend their wealthier peers, part of what the researchers call "friending bias."

That friending bias "seems to be determined by the structure of institutions, rather than just personal preferences. The places in which people interact shape to a great degree, the types of people they end up meeting at the friending," Chetty said. For instance, Chetty said, friendships formed in religious institutions are more likely to cut across the class lines.

Another example could be schools rethinking how they track students into different level classes, since that often follows class lines. It might also be rethinking the space itself: "There are other efforts to change the architecture of schools, so that kids are literally coming into contact more or eating in the lunch hall more with kids from different backgrounds."

But, according to Chetty, it's also important to keep in mind more traditional programs aimed at upward mobility, like job training.

"The types of traditional programs that tend to be successful, tend to have a social capital element where we're connecting people to others who can help them — giving them mentoring, giving them networks that are gonna help them get those better jobs — and so I think there's a broader lesson to how we structure policies in the US and beyond."