Facebook Study: You’re Less Like Your Friends Than You Think
If people realized how different they are from their fellow citizens, the country would fall apart in a weekend. Working as a journalist taught me that. I can’t help noticing, whenever I talk to people from different places, cultures and ideologies, how much my current interview doesn’t know about my last one. And whenever people imagine the lives of those they don’t know well, they fill in the blanks with a picture of themselves. Now Sharad Goel at Messy Matters, working with two colleagues, has quantified this phenomenon, using a powerful tool for exploring how people line up with each other. It’s called Facebook.
A big trove of social-science research says that most of us prefer to live, work and play near people who are “similar.” Hence the kinds of political, economic and cultural clusterings that create “red states” and “blue states,” and neighborhoods that are instantly recognizable as yuppie, buppie, gay, blue-collar and so on.
But “similarity” is not a stable trait. It’s a feeling, which changes with circumstances. As the psychologist Kurt Lewin pointed out, on many measures any mother is more similar to other mothers than she is to her partner or to her children. Nonetheless, when it comes to defining family, that similarity in gender, age and life-experience doesn’t count as much as the other similarities she feels with her kids and husband. The profound likeness she has to her spouse and offspring makes family possible. The profound unlikeness she feels to those same people makes “girls’ night out” possible. Both feelings are real and important. Neither is permanent. That’s human nature.
Because similarity is a supple notion, it’s not accurate to say human beings only look for people like themselves, and then join that team. At other times, it works the other way round: First they join the team, then they decide its members are like themselves.
That’s what Goel and his co-authors, Duncan J. Watts and Winter Mason, have documented, using Facebook. (You can read the entire paper, in pdf form, here.) They created a game there, whose players had to describe their views on hot-button political topics (Israeli-Palestinian issues, taxes, universal health care). Then they guessed how their friends would answer the same questions. Meanwhile, Facebook being Facebook, the friends were doing the same.
Results? People way overestimate the extent to which their friends agree with them. And the extent of the error rises with distance: Your guess about your best friend’s views is more likely to be right than your guess about a more casual pal.
That, I think, is the reason big countries don’t fall to pieces: Knowing people are “like us” on one score (we’re all Americans), we just figure they’re like us on others. Churchgoers think people go to church; drivers are shocked that some of us live without cars; bisexuals figure everyone is bi, while gays and straights consider it obvious that bisexuals are just kidding themselves. The other day David Brooks pointed out that most people believe morality comes from God, and I had a Moment. He is, of course, right. But I had to put some conscious effort into realizing that. My default setting, despite years of exposure to reality, is to figure that most people gave up that sort of thinking when they were 12. Because, well, I did. And aren’t most people, you know, like me?