Online and divided: Social media's effects on American life

Background: Social media is deeply embedded in American political and social life, yet evidence about its consequences remains fragmented across single outcomes, specific platforms, and predominantly younger populations. It therefore remains unclear whether its associations extend broadly across social life or are concentrated in particular domains and operate through distinct communicative and psychological mechanisms.Aims: This study examines the relationships between social media engagement and five domains of American social life— political polarisation, institutional trust, subjective wellbeing, racial attitudes, and religious practice— and tests the mediating roles of online political talk, cross-cutting exposure, and anticipated loneliness without internet access.Methodology: Data were drawn from the ۲۰۲۴ General Social Survey (N= ۳,۹۸۶), with model-specific subsamples resulting from its split-ballot design. Descriptive statistics, Spearman correlations, multivariate ordinary least squares regressions with demographic controls, and observed-variable path analyses were employed.Findings: Given the cross-sectional design, the findings are interpreted as associational rather than causal. Greater social media engagement was associated with higher ideological extremity (β= ۰.۰۵۸, P= ۰.۰۴۴) and greater anticipated loneliness without internet access (β= ۰.۳۷۵, P < ۰.۰۰۱). Path analysis indicated that the relationship between social media engagement and polarisation was primarily channelled through online political talk (indirect effect: β= ۰.۲۳۰, P < ۰.۰۰۱). Cross-cutting exposure also produced a smaller positive indirect effect (β= ۰.۰۳۴, P= ۰.۰۲۶), suggesting that encountering opposing views online may intensify rather than reduce polarisation. Social media engagement had no significant direct association with happiness but showed a significant negative indirect association through anticipated loneliness (β= −۰.۰۱۸, P < ۰.۰۰۱). No significant associations emerged for institutional trust or religious practice, while only a small negative association was found for endorsement of innate explanations of racial inequality. Overall, social media’s associations with American social life appear selective and mechanism-dependent rather than universal.Conclusion: They are concentrated in politically expressive interaction and psychological dependency, shifting analytical attention from exposure alone to platform affordances that reward adversarial participation and continuous connectivity. These findings support a domain-sensitive and mechanism-centred approach to digital media effects while underscoring the need for longitudinal and experimental verification.

کلیدواژه ها:

general social survey ، Path analysis ، political polarisation ، Psychological wellbeing ، social media engagement