School Diversity and Social Capital: Can Researchers Measure Youth Social Networks Using Administrative Data?

Peer networks, particularly those that yield cross-class and cross-racial social connections, are a critical factor driving youth educational outcomes and economic mobility, yet network data are notoriously expensive and difficult to collect. We seek to develop and test a strategy to measure students’ social networks using administrative data, which is typically available in most school districts and state agencies. Specifically, we will determine how a measure of network strength and diversity, computed from administrative data, performs compared to the “gold-standard” approach in social networks research, which consists of lengthy surveys that ask participants to self-report their ties from a roster of all potential connections. We examine what predicts strong and weak cross-class and cross-race ties across four high schools in the metropolitan Detroit area with different racial and socioeconomic demographics. Our primary aims are to: (1) Examine how accurately proxies for adolescent social network ties (strong and weak), constructed from state administrative data, predict self-reported ties. We will also explore how these predictions vary across schools with different demographics, and for different types of social ties, including cross-class friendships (economic connectedness) and cross-race friendships. (2) Develop student-level measures of economic connectedness, using weighted factors from state administrative data, for schools with different racial and socioeconomic characteristics. Our goal is to develop a new, more scalable measure of students’ social capital, which can be used by other researchers to explore the causes and consequences of various forms of social capital.

School Diversity and Social Capital: Conditions That Foster Bridging Ties in Schools

Social networks can drive upward mobility for marginalized youth, yet cross-class ties remain difficult to form due to structural constraints, organizational barriers, and homophily. Using merged administrative and social network survey data from 948 11th-grade students, and interviews with 32 students, across five diverse high schools in metropolitan Detroit, we examine the mechanisms through which bridging ties form across social classes. Using exponential random graph models, we found that shared extracurricular activities and courses are the strongest predictors of cross-class friendship formation, while residential proximity plays a minimal role in most schools. Interviews further illuminated mechanisms through which school activities created “third spaces” for cross-class interaction. Findings suggest schools can foster cross-class ties through intentional programming and course scheduling, countering the effects of residential segregation, with important implications for educational equity and social mobility.