Friendship, Communication, and Vote Agreement of Congressmen in the United States

The digital traces of congressmen on Twitter allow researchers to observe how these officials connect and communicate with each other in a direct and unobtrusive manner. Using data from Twitter and from other sources (e.g., roll-call vote data), this paper employs the exponential random graph modeling to map the friendship and communication networks of congressmen on Twitter, to uncover the generative mechanisms that underlie such networks, and to reveal the interrelationship among friendship, communication, and vote agreement. The friendship and communication networks of congressmen on Twitter demonstrate a high degree of political polarization. Four homophily attributes, namely, same-party, shared public concern, same-state, and same-chamber, play significant roles in generating ties in friendship and communication networks. Reciprocity and triadic closure are two endogenous mechanisms that operate in both friendship and communication networks. Outdegree is a significant endogenous mechanism that only affects the formation of ties in the communication networks of congressmen than in their friendship network. The magnitudes of homophily attributes are bounded by reciprocity and triadic closure. The friendship ties of congressmen on Twitter facilitate their communication in the social media website, whereas both friendship and communication ties on Twitter increase the likelihood of vote agreement between congressmen.