Cooperative Behavior Cascades in Human Social Networks

This study investigates whether cooperative and uncooperative behaviors spread through human social networks via social contagion. While theoretical models suggest that network structure influences the evolution of cooperation, empirical evidence has been scarce due to the difficulty of separating causal influence from homophily—the tendency for similar people to associate with one another. To overcome this, the authors analyzed data from a series of public goods games where subjects were randomly assigned to groups, ensuring that any observed correlation in behavior between connected individuals was due to social influence rather than selection bias or shared environmental factors.

The researchers found that cooperative behavior cascades through social networks. When a focal individual (the "ego") interacts with others (the "alters"), the alters' contribution levels significantly influence the ego's future contributions. This influence is not limited to direct interactions; it extends up to three degrees of separation, meaning an individual's behavior affects people they have never met. This cascade effect persists for multiple periods, and the authors estimate that each additional unit contributed to the public good in the first period is effectively tripled over the course of the experiment due to the subsequent influence on others.

These results provide experimental evidence that social contagion plays a critical role in the evolution of cooperation. By demonstrating that behavioral imitation can create cascades, the study suggests that small changes in social norms or institutional rules can have outsized impacts on group behavior. This multiplier effect may help explain how altruism and cooperation emerge and are maintained in structured populations, offering a new perspective on the coevolution of human social networks and cooperative strategies.