Published online 26 June 2008 | Nature | doi:10.1038/news.2008.918


The secret of happiness: grinning on the Internet

People who post smiley photos on social networking sites attract happy friends.

computer social networkingLook happy online, and you'll have cheery friends.Anne-Marie Palmer / Alamy

How happy you are is influenced by your social links to people you've never heard of and never met.

That's the conclusion of a US study looking at the spread of happiness and depression across social networks.

Happy people cluster together, the research suggests. And the opposite also seems to be true — so if you are miserable, you are more likely to have miserable friends.

The effect holds in both the real and virtual worlds. People who put smiling photos on their profiles for social networking sites such as Facebook tend to link to one another. Frowners do likewise.

But it's not just direct contact that counts. The link is significant to three degrees of separation — that is, your own emotional state is connected to that of your friends' friends' friends.

"Your happiness depends on the happiness of individuals beyond your own social horizon," says sociologist Nicholas Christakis of Harvard University in Cambridge, who carried out the study. "You can understand happiness by studying individuals, but that only gets you so far. There's more to be learned by studying the group."

Grin to win

Christakis and his colleague James Fowler, a political scientist at the University of California, San Diego, drew their network of emotional states using data from the Framingham Heart Study. This project has monitored the health of the people of Framingham, Massachusetts, since 1948. Christakis presented the latest findings on 25 June at the International Workshop and Conference on Network Science in Norwich, UK.

As well as gathering information on their physical health, the study collected data on people's emotional states, to see whether depression and heart disease are correlated.

Understandably enough, emotions are contagious over short distances and times. Waitresses who smile at their customers get bigger tips, and students with depressed room mates are more likely to get depressed themselves.

The Framingham data allowed Christakis and Fowler to show that the same holds over long time spans, and in large networks of social interaction.

"At one level it's not surprising — if your partner is really happy today, then you feel better," says network researcher Albert-LászlóBarabási of Northeastern University in Boston, Massachusetts. "But it's surprising how systematic and measurable the effect is."

Friends like these

As well as social proximity, geographical distance is important — your friends are more influential the nearer they live. However, the effect was not seen in the workplace — perhaps because the things that make you happy, such as a promotion, might disgruntle your passed-over colleagues.

Christakis and Fowler also looked at the photos on the Facebook pages of a network of 1,700 students at an unnamed university in the northern United States.

Smiley students clustered together. They were also closer to the centre of the network, and had more friends than those who didn't smile. The researchers have preliminary evidence that there are similar patterns for other Facebook properties, such as privacy settings.

Besides emotions, Christakis and Fowler have seen similar patterns in the Framingham data for obesity and smoking. Both factors show what Christakis calls "three-degrees of influence", in that the effects of connections peter out beyond contacts that are twice-removed.


"The regularities are intriguing," he says. "We think there are limits to how far influence can stretch, and we seem to be finding that across a broad range of phenomena."

The findings have big implications for public-health policy, he says, in how we measure cost-effectiveness — as interventions aimed at individuals have knock-on effects — and how we target healthcare.

"People are slowly understanding that there's a huge effect here, but we're till a step away from how we use that," says Barabási. "But if we know of these factors as individuals, then we can alter them, not necessarily by cutting friendships, but by being aware of their influence." 

Commenting is now closed.