Skip to content

With an AARP membership, there’s always more to discover! Check out your member benefits.


The Author Speaks: Would You Jump Off a Bridge if Your Friend’s Friend Did? Probably.

Interview with Nicholas A. Christakis, coauthor of <i>Connected: The Surprising Power of Our Social Networks and How They Shape Our Lives</i>

If you’ve ever wondered why you have trouble losing weight, consider blaming your best friend’s friend—even if you’ve never met her. That’s the startling suggestion made by Nicholas A. Christakis and James H. Fowler, two researchers who spent years studying social networks, the intricate webs of complex connections that tie us all together. What they found may forever change the way you view your most personal habits, like diet, smoking, voting and how often you smile.

The researchers’ new book, Connected: The Surprising Power of our Social Networks and How They Shape Our Lives, demonstrates how we all can be judged by the company we keep. Christakis, a professor of medicine, sociology and health care policy at Harvard, and Fowler, a political science professor at the University of California, San Diego, studied something called “social contagion.” They wanted to find out why a single town can experience a sudden outbreak of obesity, laughter or violence. Their findings are startling: When one person quits smoking, or gains weight, or finds happiness, it has a ripple effect on his friends, his friends’ friends, and his friends’ friends’ friends. The best comparison might be a school of fish—hundreds of fish make minute course adjustments based on nearby fish, which are reacting to stimuli such as predators or food. Those who don’t swim with the group—or don’t quit smoking as everybody else does—find themselves increasingly marginalized. (Read an excerpt from Connected.)

Christakis and Fowler’s work—funded by the National Institute on Aging and the National Science Foundation—has been described as both obvious and brilliant. And it’s nudging a relatively obscure field of social networks into the mainstream. But their work has oddly discomfiting implications: Does free will exist if someone you don’t even know can make you gain weight or influence your vote?

To better understand their findings, the AARP Bulletin spoke with Nicholas Christakis about why people often act like fish.

Q. Why does your research throw people for a loop?

A. We’re able to show that all kinds of phenomena people think of as deeply personal—such as their body size, or whether they start smoking, or how happy they are, whether they vote, how kind they are—seem to be strongly influenced by the actions of other people. It’s also surprising, even to us, to have shown that we’re influenced not just by the people we know—friends and relatives and coworkers—but by the friends and relatives and coworkers of our friends and relatives and coworkers.

Q. How does that work?

A. Well, let me develop an analogy for you. Imagine right now that you don’t have the flu, but your friend’s friend’s friend has the flu. So that flu is going to spread inexorably to you. The point is that similar kinds of processes operate for social phenomena. And I should stress that not everything spreads in networks, and not everything in networks spreads the same way. Germs, money, ideas, behaviors and emotions all spread, just in different ways.

Q. What does a social contagion look like?

A. People are influenced in a kind of social domino effect. There’s a kind of herding behavior that humans show, like stampeding buffalo.

Q. Buffalo?

A. Yes, we find that large interconnected groups of smokers, who don’t necessarily know one another, quit together at once like a herd of buffalo changing direction. So it would make about as much sense to ask an individual smoker, “Why did you quit?” as it would to ask a single buffalo in a stampede, “Why are you running to the left?” A buffalo might say, if it could speak, “I didn’t choose to run to the left, the whole herd is running to the left.” And in very deep ways, humans behave just like that.

Q. But how can things like happiness or smoking be contagious thirdhand?

A. The idea is that what’s spreading from person to person is a social norm. So your friend says, “Let’s have muffins and beer,” which is a terrible combination, by the way. But you go along with her suggestion, and adopt the eating behaviors of your friend, and in turn pass on those eating behaviors to your other friends, and your friends’ friends. So you can have a cascade of influence from one person to another.

Q. What about body size?

A. The idea is that as your friend gains weight, it changes your ideas of what an acceptable body size is, and you then transmit this same norm to another person.

Q. How so?

A. Let’s say Betty, Sue and Jane are friends. And Jane stops her exercise regimen and starts gaining weight. Sue thinks, well, it’s still the same Jane; so Sue now changes her ideas about what an acceptable body size is. Sue herself doesn’t stop her exercise regimen, she doesn’t gain any weight, but she nevertheless has had her ideas about obesity modified because of Jane’s behavior.

So let’s say Sue’s friend Betty starts eating a lot. In the past, Sue might have said to Betty, you should really watch what you’re eating. But because Sue’s ideas of acceptable body size have been changed, she doesn’t say anything to Betty. So Jane’s behavior affects Betty, without actually affecting Sue—something just changed in her mind.

Q. Lots of research supports the Six Degrees of Kevin Bacon game. But what’s so groundbreaking about your work is that you found a similar rule: the three degrees of influence rule.

A. Studies over the past 50 years have looked at the structure of social networks and found that people are connected by an average of six degrees of separation. So you could make a connection between any two people on the planet by an average of six handshakes. By studying a variety of phenomena, we found that people’s influence only spreads out to about three degrees of separation. So we influence, and are influenced by, our friends’ friends’ friend, but no further.

Q. Is that true online, too?

A. Online networks are the same—but different. The three degrees rule only applies to people who have real connections offline. Then there can be a cascade. But your 500 Facebook friends’ 500 friends are not influencing you.

Q. I have a friend whose friend works in the White House. So theoretically I could make President Obama start smoking again?

A. Yes, that’s correct! If you started smoking, it might encourage your friend to start smoking, which could create a more permissive environment around the president and he might resume his smoking. So let me encourage you not to do that.

Q. OK, deal. I will not start smoking and therefore I will save the life of the president.

A. I know we’re joking, but I should say these things are not determiners. They are risk factors that spread from one to the other.

Q. Yet it raises a lot of questions about free will.

A. What we’ve shown is that much of what people think is completely individualistic or within their control is actually influenced by others. So whether you are happy or not, or gain weight or not, is influenced by the choices and actions of others—sometimes people you don’t even know. On the other hand, we think these observations raise the salience—because if you make positive changes in your life, you don’t just benefit yourself, your friends and your family, you potentially benefit hundreds of thousands of other people.

Q. But how do you know that it isn’t just because the people you choose to be around are more likely to make similar choices in their own lives?

A. They might! We’re not saying that the only reason groups of people behave with synchronicity is network effect. Nor are we saying that birds of a feather don’t flock together—they do! Rich people hang out with rich people. It’s also the case that rich people make each other rich, because you help your friends.

We use a variety of statistical methods to show that there are also influence effects. And what’s amazing to us is why so many people are so resistant to this. It’s very commonsensical, that we are influenced by each other—in fact we spend our lives trying to influence each other!

Q. Call me a romantic, but finding love seems like the best possible reason to have networks.

A. One of the reasons we are genetically programmed to live in networks is because networks distribute information about predators and prey, and information about sexual partners. So the network serves to bring together compatible people.

Q. Networks also spread unfortunate things. It’s now flu season. How can people work against their networks in order to avoid getting the flu?

A. I was recently asked, if emotions spread, how can you keep from catching bad emotions? You cannot. Bad things spread in networks—suicide, violence, misinformation and germs all spread. But good things also spread—happiness, love, kindness, useful information spreads, too. And on balance, the benefits we derive from the good more than compensate us than the harm we get from the bad.

Christie Findlay is deputy editor of Capitol File magazine and former editor in chief of Politics magazine.

Join the Discussion

0 | Add Yours

Please leave your comment below.

You must be logged in to leave a comment.