Social Econ Blog

For my facebook obsessed readers, here's a big sudy of facebook social networks from 10 schools in Texas which I played a small role in developing (basically one of the authors contacted me to find out how I got my facebook data and I told him):

This paper documents the structure and composition of social networks on university campusesand investigates the processes that lead to their formation. We use a large dataset that identifies students in one another's social network on campus and link these data to university records on each student's demographic and school outcome characteristics. The campus networks exhibit common features of social networks, such as clusteredness. We document the factors that are the strongest predictors of whether two students are friends. Race is strongly related to social ties. In particular, blacks and Asians have disproportionately more same race friends than would arise from the random selection of friends, even after controlling for a
variety of measures of socioeconomic background, ability, and college activities. Also, two students are more likely to be friends if they share the same major, participate in the same campus activities, and, to a lesser extent, come from the same socioeconomic background. Next, we develop a model of the formation of social networks that decomposes the formation of social links into effects based upon the exogenous school environment and effects of endogenous choice arising from
preferences for certain characteristics in one's friends. We use student-level data from an actual social network to calibrate the model. Our model generates many of the characteristics common to social networks. We simulate network structures under alternative university policies. We find that changes in the school environment that affect the likelihood that two students interact have only a limited potential to reduce the segmentation of the social network.