Political Inequalities Start at Home: Parents, Children, and the Socialization of Civic Infrastructure Online

Abstract

We use a two-wave panel survey of parent-child dyads in the United States to connect online democratic divides with the unequal socialization of political interest in the home. We test a model connecting parent socioeconomic status to the amount of political communication in the home and the subsequent development of youth political interest over the course of an election cycle. We develop the theoretical concept of online civic infrastructure to foreground how interest-driven social media use in adolescence may shape future opportunities for civic and political engagement by building network connections and opening up flows of communication that carry news, political information, and opportunities for mobilization.

Keywords:

• political interest
• socialization
• social media
• youth participation

Acknowledgments

We are grateful to Kristen Landreville and the anonymous reviewers for their comments on the manuscript.

Funding

This research was supported by grants from the Donald W. Reynolds Journalism Institute and the Journal Foundation.

Supplemental Material

Supplemental data for this article can be accessed on the publisher’s website at https://doi.org/10.1080/10584609.2017.1333550.

Notes

^1. There are significant differences between the W1 and W2 means for political expression on social media (p = .001), family communication (p = .031), and political interest (p < .001). The difference between the two waves for political customization has a p-value of .10.

^2. Our models are both autoregressive and cross-lagged. We regress each dependent variable on itself to determine the stability of each key variable over time (autoregressive) as well as looking at the effects of our mediators on the key constructs in the second wave of the data (cross-lagged).

^3. This is an unexpected finding, and one to which we hesitate to assign a substantive interpretation. We suspect the sign flip is a problem of multicollinearity. Although the variance inflation factor (VIF) diagnostics are within acceptable limits, the bivariate correlation between Family Political Communication at W1 and W2 is .71.

Additional information

Funding

This research was supported by grants from the Donald W. Reynolds Journalism Institute and the Journal Foundation.

Notes on contributors

Kjerstin Thorson

Kjerstin Thorson is Assistant Professor, College of Communication Arts & Sciences, Michigan State University.

Yu Xu

Yu Xu is a PhD student, Annenberg School for Communication and Journalism, University of Southern California.

Stephanie Edgerly

Stephanie Edgerly is Assistant Professor, Medill School of Journalism, Media, Integrated Marketing Communucations, Northwestern University