The Social Structure of Acquaintanceship Networks in Chile

What makes some people develop larger networks than others, and consecutively benefit from the material and non-material resources nested in those networks (Lin, 2002)? To respond this question, we track down and systematize previous work on social homophily and primary networks, as well as develop some new insights following Blau's (1977) macro-structural approach. Our main claim is that people who belong simultaneously to social groups that partially overlap at the societal level, and therefore are exposed to social divisions that reinforce each other, will have less contact with members of other social groups, which in turn, will decrease the size of their acquaintanceship networks. We test this claim using data from a national probability sample that included twenty-five items of the form "How many X's do you know?" where "X" represent a sub-population of interest (also known as aggregate relational data). With these measures we estimate the size of respondent acquaintanceship networks using the scale-up estimator developed by McCarty and colleagues (2001). While controlling for an ample set of demographic variables, we find support for our social group configuration hypothesis. For example, we show that Chileans who simultaneously identify as Catholics, lean towards right-wing positions and have a high educational level, all three traits that are positively correlated, have significantly smaller networks that other Chileans who have a crossed group membership configuration.