The alarming increase in social inequality in the US corrodes the national ethos as the “land of opportunity” (Grusky and Hill 2018). Embedded in that imagery is the ideal of a nation where rich and poor can co-mingle at work, at play, and at school. The causes and consequences of such cross-SES relationships are the focus of the proposed research. Integrating network dynamics with models of social stratification provides untapped potential for our theoretical and empirical understanding of social inequality. Building on research examining network ties that bridge social strata (Blau 1977; Blau and Schwartz 1984; Bourdieu 1980; Burt 1992; Lin 1999; Sewell, Haller, and Portes 1969; Simmel 1908; Wright and Cho 1992), we propose to study the consequences of adolescent friendship dynamics that bridge socioeconomic boundaries for long-term socioeconomic attainment of disadvantaged youth.
Past research shows that homophily in SES-stratified networks reinforces socioeconomic stratification (DiMaggio and Garip 2012; Manzo 2013; Tsvetkova, Wagner, and Mao 2018). Cross-SES friendships counter this stratifying effect of homophily by providing access to resources that are otherwise unavailable to lower-SES students. Moreover, in homophilous networks stratified by SES, cross-SES ties connect not only unequal nodes but also bridge unequal clusters. The paradoxical effect of stratified homophily is that the stratification of entire clusters makes ties between clusters more consequential in attenuating the stratifying effect of homophily. Cross-SES ties give the disadvantaged access not only to the instructional, cultural, and social capital of advantaged individuals but also to their advantaged networks. A deeper understanding of the contextual dependencies, underlying mechanisms, and long-term consequences of cross-SES friendship for socioeconomic attainment is the overall objective of the proposed research. The proposal has three primary goals:
1. To disentangle the effects of higher-SES peers on socioeconomic attainment into an “achievement” effect of educational spillover and a “background” effect of peer origin indicative of cultural and social capital, and into pre- and post-graduation effects.
2. To examine three important aspects of the ecological context of cross-SES friendship: (i) the effect of neighborhood and school racial and SES composition on the prevalence of cross-SES friendship, (ii) the effect of network duration, size, density, homophily, and segregation on the returns to cross-SES friendship, and (iii) teasing apart direct peer effects, indirect ripple effects on adjacent students, and contextual effects of cross-SES friendship.
3. To leverage newly available data and advanced modeling techniques to overcome three data limitations: First, most individual-level data on social inequality is obtained from surveys that rely on respondents’ estimates of peer features rather than direct measures. Second, most surveys with network component measure respondents’ direct connections (ego networks), not the complete network structure, which limits the ability to measure ecological effects. Third, network data is rarely longitudinal, making it difficult to tease apart the causal direction.
Until recently, data had not been collected long enough to study long-term network effects. The newly released Add Health wave (Harris and Udry 2020) opens up unprecedented opportunities to study the causes and consequences of cross-SES ties, not only because of the sheer number of observations, but more importantly, by tracking adolescent respondents far enough into adulthood to examine the long-term socioeconomic consequences, and by including direct measures of peer attributes and longitudinal data on friendship network structure. Using these data, we propose to build a relational model of attainment that disentangles (i) informational, cultural, and social capital effects of peers, (ii) pre- and post-graduation effects, (iii) ecological effects of network duration, size, density, homophily, and segregation, and (iv) direct, indirect, and contextual peer effects.
Intellectual merit
Social network analysis and stratification research have progressed in relative isolation, creating untapped potential on both sides. Network analysis has evolved with insufficient attention to the processes that unfold on the edges, in particular regarding the consequences of diffusion processes for social stratification. Stratification research has concentrated on social-psychological determinants of attainment with insufficient attention to individuals’ embeddedness in social networks.
To be sure, social-psychological attainment models have also included network processes, such as aspirations of significant others (Morgan 2005; Sewell et al. 1969), and social capital inherent in personal networks (Bourdieu 1980; Burt 1992; Granovetter 1973; Lin 1999). Work on educational attainment has likewise included network processes, such as spillover of educational achievement (for a survey, see Sacerdote 2011), and contextual effects of SES composition of classrooms and schools (Van Ewijk and Sleegers 2010; Perry and McConney 2010). However, these efforts at building relational models of attainment have not been integrated and were unable to fully address concerns about causality due to the intractability of the network endogeneity problem (Mouw 2003): Do ties to higher SES peers promote socioeconomic attainment, do attainment aspirations motivate tie formation, strength, and duration to higher SES peers, or do shared environments shape both socioeconomic attainment and network structure?
Newly available data and advances in causal inference for networked data now make it possible to attack the endogeneity problem head-on, and in doing so, bridge the disciplinary divide between social network analysis and stratification research. Building towards a relational model of attainment, the proposed research seeks to disentangle five processes identified as highly relevant in previous research:
- The consequences of cross-SES ties for socioeconomic attainment from the confounding effects of causal priors that influence both tie formation and socioeconomic attainment (network endogeneity and shared environment), and from individual determinants affecting life chances.
While we know that homophily in an SES-stratified world reinforces stratification, we lack evidence that heterophily improves long-term socioeconomic attainment. - Informational, cultural, and social capital effects of peer relationships.
Splitting peer effects into an “background” effect of peer origin and an “achievement” effect of educational spillover, we separate informational from cultural and social capital mechanisms of peer relations - Pre-graduation and post-graduation effect of higher-SES peers.
Separating pre- and post-graduation effects will show the extent to which peer effects on life chances are processes within school, and whether peers continue to influence their friends when effects on educational attainment have terminated. - Ecological effects of network duration, size, density, homophily, and segregation on the prevalence and returns to cross-SES friendship.
This is especially important for two reasons: (i) the high levels of racial and socioeconomic segregation of U.S. schools and communities, and (ii) the possibility that individual effects of higher-SES peers and ecological effects have countervailing effects at the individual level. - Direct individual-level effects of higher-SES peers from indirect ripple effects of cross-SES ties on adjacent students and from contextual effects of SES classroom/school composition.
Separately measuring direct and indirect effects makes it possible to examine interactions between.
Broader impact
Unpacking the contextual dependencies of adolescent network dynamics will inform social policy by providing important insight into potentially conflicting individual-level and group-level effects of cross-SES friendships. Moreover, measuring the effects of cross-SES relations on access to cultural and social capital will point to broader opportunities for interventions that extend beyond family and school. It is often assumed that the limits of solidarity end at family boundaries. Coleman’s (1966) landmark study concluded that resources provided by peers are more important for school achievement than are the resources provided by the school. Measuring the long-term effects of cross-SES friendship beyond spillover in school, we take Coleman’s analyses one step further, to examine the extent to which cross-SES friendship acts as a counteracting force against the self-reinforcing mechanisms of segregation and stratification. More generally, the effectiveness of peer-directed policy interventions depends decisively on the accurate identification of causal structures, the primary goal of the proposed research.
This project is funded by NSF. A shorter summary of the project can be found here.