Search and Homophily in Social Networks

We study the formation of social ties among heteogeneous agents in a model where meetings are governed by agents' directed search. The aim is to shed light on the important issue of homophily (the tendency of agents to connect with others of the same type). The essential contribution of the model is to provide a basic microfoundation for the opportunity/meeting biases that, as the literature highlights, are a crucial element of the phenomenon. Under the assumption that search is more effective in large pools, the equilibrium is characterized by a threshold in terms of group size: large groups only search among similar agents while smaller groups search in the whole population. This threshold behavior is consistent with the empirical evidence observed in a range of social environments such as high school friendships and interethnic marriages. And assuming that search is subject to small frictions, it also generates the bell-shaped form of the so-called Coleman index observed in the data. Other implications of the model supported by the evidence concern the pattern of cross-group ties among small groups, the linearity of excess homophily for large groups, and the positive effect on it of overall population size.Homophily, search, social networks, segregation.