Contagious Happiness

Maybe ten years ago, I was made aware of a book entitled Contagious Emotions, which was about the codependent nature of depression and how friends and relatives of someone suffering from depression could help the depressed person break free without themselves getting “infected.” Ever since then, I’ve enjoyed making reference to the concept of “contagious emotions” and watching peoples’ reaction when I did so.

And now, there’s this report in last Friday’s British Medical Journal demonstrating that emotions—at least to the extent they involve happiness-- are, well, contagious.

Co-authors James Fowler and Nicholas Christakis used records generated by the long-running Framingham Heart Study to do extensive social network analysis. Using responses to a subset of questions from the CES-D that were administered to participants several times throughout that study’s period, Fowler and Christakis were able to trace reported happiness levels of study participants and people in their social networks.

Here’s what they found:

• “Clusters of happy and unhappy people are visible in the network, and the relationship between people’s happiness extends up to three degrees of separation (for example, to the friends of one’s friends’ friends.)”
• Furthermore, people who are surrounded by lots of happy people, as well as people who are “central in the network” (a term of art in social network analysis that characterizes the type of person who in the old days would have been labeled as a BMOC), are more likely to have future happiness.
• And finally, Fowler and Christakis found that geographic proximity was a major factor in the “transmission” of happiness: The closer you’re physically located to a happy person, the greater the likelihood you’ll be happy, too. (But the big exception here is the workplace. “Effects were not seen between coworkers.” And there were some puzzling findings regarding “coresident spouses.” See below for more on these two points.)

Press accounts of this study include ringing endorsements by several big names in the area of behavioral research. The Washington Post quotes Martin Seligman, the “founding father” of positive psychology, (“pathfinding … totally original … the findings are striking”) and Ed Diener, co-author along with Sonja Lyubomirsky of many of the reports that form the underpinning of Lyubomirsky’s recent book The How of Happiness (“extremely exciting … interesting, provocative and important). The Harvard Crimson refers to an e-mailed statement (“stunning”) by Daniel Gilbert, author of Stumbling on Happiness and TED conference speaker. And finally, the Gray Lady reports the tempered approbation of Nobel laureate Daniel Kahneman, who calls the study an “extremely important and interesting work” but has some difficulty accepting (in the absence of independent replication) one of the study’s conclusions, namely that the “happiness effect” was not as strong for co-resident spouses as it was for next door neighbors (8% versus 34%).

So this study is a big deal. And as the authors point out, it has major relevance for public health

There is of course a tradition of community approaches to mental health, but this longstanding concern is now being coupled with a burgeoning interest in health and social networks. More generally, conceptions of health and concerns for the wellbeing of both individuals and populations are increasingly broadening to include diverse "quality of life" attributes, including happiness. Most important from our perspective is the recognition that people are embedded in social networks and that the health and wellbeing of one person affects the health and wellbeing of others. This fundamental fact of existence provides a conceptual justification for the specialty of public health. Human happiness is not merely the province of isolated individuals.

But before we all hop on the “happiness is contagious” bandwagon, we need to take note of some soft spots in the research.

For one, there’s the puzzling finding that’s left Daniel Kahneman scratching his head. How can it be that having a happy next door neighbor has a more powerful effect on your own happiness than having a happy spouse? (The authors hypothesize that this may be due to the fact that people are more likely to take “happiness cues” from people of their own gender—and they note that in their study sample all spouses were opposite-gender.) As Dr Kahneman has pointed out, unless another study independently replicates this finding, it will be tough to accept it as being valid.

In a similar vein, it’s difficult to make sense of the finding that there was no “happiness effect” in the workplace. Fowler and Christakis write in the paper that this “suggest[s] that the social context might moderate the flow of happiness from one person to another” In other words, if happiness is a contagion, then there’s something powerful in the immune system of the workplace that counteracts it there. But what might that be? The paper itself doesn’t say, but the New York Times article says “Professor Fowler believes inherent competition at work might cancel out a happy colleague’s positive vibe.” But there are no indications of any plans—by Fowler and Christakis or any other researchers—to test out that hypothesis. So, in the absence of further illuminative research into why the workplace—where millions of people spend at least a quarter of their time in close proximity to each other—is immune to the contagious spread of happiness, you can count me along with Daniel Kahneman as someone who needs a bit more convincing.

These two (non-trivial) issues aside, there’s the underlying question of whether and to what extent correlation implies causation when you use social network analysis as applied to health outcomes. In the same issue of the British Medical Journal as the Fowler-Christakis paper, there’s another paper, by Federal Reserve Bank of Boston economist Ethan Cohen-Cole and Yale professor Jason Fletcher, that addresses just that.

What they found: “Current empirical methods used to estimate causal social network effects might detect implausible network effects, including "contagion" in headaches, skin problems, and height between adolescent friends.” In other words, it’s possible to apply the social network analysis methodology used by Fowler and Christakis to other data sets and to conclude that, if your friends are tall, you’re more likely to be tall yourself—an implausible (to put it mildly) example of the “contagion” effect.

The Harvard Crimson does a nice job putting the Cohen-Cole and Fletcher study in context:

“Our study certainly does not refute their happiness paper, but it just suggests some caution that if you don’t take care to control for other factors, that you might be finding contagion where none exists,” said Jason M. Fletcher, a professor of public health at Yale.

Fletcher co-authored a study suggesting that perceived network effects could be erroneous. Using the same statistical methods as the happiness study, his study found that characteristics like acne, headaches, and height are contagious among adolescents, indicating that the methods used in the happiness study can produce spurious results.

“There’s no such thing as a social contagion in height,” Fletcher said.

Fletcher and his co-author, B. Cohen-Cole ’95, an economist at the Federal Reserve Bank of Boston, suggested that the happiness study could be biased because happy people are often friends and that their good moods are not necessarily influenced by each other.

“Friends select people to be their friends based on similar characteristics,” said Fletcher, “and potentially happy people choose to be friends with other happy people.”

He added that friends are often exposed to the same environment, including similar levels of crime, risk, and weather, and that those external variables could influence happiness more than a friend’s mood.

In light of these criticisms, both research groups plan to continue probing into the field of happiness with future studies.

“The whole point of science is that you want to capture a great idea but then retain healthy skepticism,” Fowler said.

“Capture a great idea, but retain healthy skepticism”—It sounds a lot like another wise maxim that got a lot of play a quarter of a century ago….