Sources of Variation in Social Networks

Abstract

What explains the large variation in the number of contacts (degree) that different participants of social networks have: age, randomness, or some unobservable quality measure? The model presented in Jackson and Rogers (2007), which emphasizes age as the main source of variation, successfully fits the degree distributions observed in real-world social networks, but is inconsistent with the relationship between age and degree that is also observed. After documenting this fact, I extend the Jackson-Rogers model to allow for individuals to vary in some unobservable quality measure. This quality measure embodies individuals' abilities to attract new contacts. Within-cohort variation in degree identifies variability in quality. I find that heterogeneity in age and quality are both important in explaining the variation in degree that is observed in social networks.

JEL Classification: D85