Dynamic spread of happiness in a large social network: longitudinal analysis over 20 years in the Framingham Heart Study

This study investigates whether human happiness is a collective phenomenon that spreads through social networks. While previous research has identified various individual-level determinants of happiness—such as income, genetics, and life events—little was known about how the emotional states of one's social contacts influence an individual's own happiness over time.

The researchers conducted a longitudinal social network analysis using data from the Framingham Heart Study, tracking 4,739 individuals between 1983 and 2003. By utilizing detailed records of social ties—including family, friends, spouses, neighbors, and coworkers—the team mapped the network structure and measured happiness using a validated four-item scale. They employed statistical models to distinguish between the spread of happiness (induction) and the tendency for similar people to associate with one another (homophily).

The study reveals that happiness clusters within social networks, with happy and unhappy individuals forming distinct groups. Crucially, this influence extends up to three degrees of separation: an individual's likelihood of being happy increases significantly if a friend, a friend's friend, or a friend's friend's friend becomes happy.

Physical proximity and the nature of the relationship are key moderators. The effect is strongest among mutual friends living within a mile of each other, while coworkers show no significant influence on one another's happiness. The impact of a happy contact also decays over time, suggesting that emotional contagion is a dynamic, transient process. Furthermore, the researchers found that network centrality—being well-connected within one's local social circle—independently predicts future happiness, suggesting that the structure of one's social life is as important as the specific individuals within it.