Other recent research: Nation-state formation | Ethnic exclusion and wars | Ethnic boundary making

Boundaries in social networks: Beyond and below racial homophily

This project puts the boundary making approach to work on the social network data that resulted from a co-operative project with Nicholas Christakis, Jason Kaufman, and Kevin Lewis from the department of sociology at Harvard. We have gathered information on the Facebook.com friendship networks of an entire college freshman class. The project has been featured on the front page of the New York Times and many other news outlets in the US and Europe. A first paper describing the dataset has appeared in Social Networks. The data is publicly available here.

In social network research as in sociology in general, race has long been considered one of the most important principles of group formation in U.S. society. A second paper, co-authored with Kevin Lewis, suggests that past scholarship may have exaggerated the actual degree of racial homophily. Exponential random graph modeling of the Facebook data allow us to demonstrate that racial homophily is partly explained by 1) homophily on lower, ethnic levels of categorical differentiation; 2) differences in sociality and thus network size across racial categories; and 3) balancing mechanisms such as the tendency to befriend somebody who considers me a friend (reciprocity) or who is a friend of my friend (triadic closure). In the final step of analysis we put racial closure further in perspective by showing that members of other attribute categories, including students who attended elite boarding schools, originate from certain states of origin, pursue particular academic interests, or display distinct cultural tastes over-privilege friendships with each other just as much as members of the same racial category. This paper is now under review and available upon request.