Does contagion explain the clustering of happiness among friends?

A few months ago a couple of researchers - James Fowler and Nicholas Christakis -published some findings that were reported around the world in the popular media under the headline: “Happiness is contagious”. At the same time another article cast doubt on these findings by claiming that similar social network effects could be detected for acne, height and headaches.

Having thought about it, the headline “Happiness is contagious” seems to me to have about the same news content as “Influenza is contagious”. We don’t need research to tell us that we obtain pleasure from associating with happy people. But once the headline has grabbed our attention we may feel a desire to read on to find out why the item was considered newsworthy.

Why is there controversy over the research finding that happiness is contagious? The study (‘Dynamic spread of happiness in a large social network’, BMJ, 338, Jan ’09) actually claims to be providing evidence for something more substantial than the emotional contagion in which the mood of one person fleetingly influences the mood of others. The results, based on surveys of a large social network (the Framingham heart study), suggests that if you have a friend who lives within a mile who becomes happier, this increases the probability that you will also become happier. Similar effects were also noted with regard to spouses, siblings and next door neighbours.

The authors claim that their results show that “changes in individual happiness can ripple through social networks and generate large scale structure in the network, giving rise to clusters of happy and unhappy individuals” (p338).

However, in his comment on the Freakonomics blog, Justin Wolfers suggested the most likely reason a person might become happier when a friend becomes happier isn’t because happiness is contagious, but because friends tend to share similar interests and to be influenced by similar things.

Justin Wolfers is right a lot of the time and he might be right about this. It seems to me, though, that the observed tendency for the happiness of friends to increase at the same time could be attributable to more complex processes than either direct emotional contagion or the influence of some factor that is independent of their own actions – such as the football team they support in the national league winning more frequently. It is possible, for example, that their increase in happiness could be an outcome of their involvement in some voluntary community activity.

Recent research on empathy and collective action may be relevant to the clustering of happiness of people involved in social networks. An important characteristic of voluntary collective action is the need to place trust in volunteers who promise to participate for motives other than personal reward for effort. Research by Paul Zak, using the trust game (explained in detail here), indicates that a sense of being trusted results in release of oxytocin (OT) and that increased OT results in more trustworthy behavior. In a recent paper (“Empathy and collective action”) Paul Zak and Jorge Barraza note that release of OT potentiates the release of dopamine (making prosocial behavior more rewarding) and causes synaptic serotonin to rise (reducing anxiety and helping people to sustain altruistic collective action over an extended period of time).

Zak and Barraza suggest that this brain circuit promoting altruistic collective action is stimulated when volunteers do things like spending time together to build empathy, exchanging gifts, sharing meals and sharing adventures. They also cite evidence that people who volunteer to help others report higher levels of happiness.