Opinion, Berkeley Blogs

Social media reading list: Christakis and Fowler, Connected

By Claude Fischer

[ Connected: The Surprising Power of Our Social Networks and How They Shape Our Lives is part of the 2011 Summer Reading List at UC Berkeley. The theme this year is "Social Media."]

A couple of years ago, social scientists Nicholas Christakis and James Fowler got a lot of media attention for new research suggesting that people can “catch” all sorts of habits – like smoking, overeating, feeling happy, and feeling lonely – from the people they know. More dramatic still, the research suggested that you can catch such habits from the people who know the people you know. Christakis and Fowler have put their findings and those of many other researchers into a fascinating and easy-to-read book entitled, Connections. The subtitle tells you the theme: The Surprising Power of Our Social Networks and How They Shape Our Lives.

The idea that our family, friends, and even acquaintances influence our lives is the central thesis of scholars who study “social networks.” It also underlies much of our everyday behavior. When we worry about who our kids hang out with, it is because we believe that “bad influences” can shape their lives. When we go to a party to “network,” we do so because we believe that who you know, and who they know, can determine career opportunities. Still, the claim in the Christakis and Fowler subtitle is sharply contested.

Certainly, kids who get into trouble tend to be friends with kids who get into trouble; people in powerful jobs tend to hang with people in powerful jobs. But which came first, the connection or the similarity? Critics of social network research argue that the “shaping our lives” claim is overblown: People are like the people they know because they prefer each others’ company (say, kids into drugs seek out other kids with that inclination; opera buffs like to be with opera buffs). Surely, both processes are at work: similar types of people seek each other out and people who spend time together become more similar – they, for example, learn about opera from friends or family, they are introduced to a drug by a classmate. The question, as yet pretty open, is how much each of the two processes contributes to the end result, which is that we tend to be pretty similar to the people in our networks.

Christakis and Fowler gained particular attention by claiming even more power for social networks, arguing that individuals are influenced through network ties by people they may not have even met. Using data from a long-term health study, the authors present evidence suggesting that if the friend of my friend is, say, obese or despondent, I run an elevated risk of becoming that way, too, even if the mutual friend we have in common is not affected, if he stays thin or cheery. This mysterious influence at a distance (Christakis and Fowler are not quite sure how it might happen) has come under sharp critique in the academic blogosphere; some arguing that the finding is a technical error.

It is safe to say that most researchers in this area (your humble correspondent included) are unpersuaded about influence at a distance. But we do share the authors’ conviction, well-explained in this book, that whom we know really matters.