Keith N Hampton

The dominant narrative insists that adolescents are living through an unprecedented mental health crisis caused by smartphones and social media. In books, headlines, and policy debates, screens are blamed for luring young people away from the “real world,” fostering anxiety, and eroding healthy development.

Our new longitudinal study suggests that this diagnosis gets the story backward.

Today’s teens are not an “anxious generation” broken by screens. They are a constrained generation using digital platforms to adapt to an adult world that has systematically stripped away their autonomy.

The vanishing spaces and autonomy of childhood:

The narrative of digital harm rests on a fundamental blind spot: it ignores decades of structural changes to childhood and adolescence that took root long before the first smartphone appeared.

1. The Collapse of Unsupervised Free Play: Beginning in the late 20th century, heightened parental fears over traffic, crime, and “stranger danger” drastically curtailed children’s physical range and autonomous outdoor exploration.
2. School Reforms and Academic Enclosure: Educational policies over the past thirty years—sparked by the 1994 Prisoners of Time and solidified by high-stakes testing regimes (in the U.S. this was codified in the 2001 No Child Left Behind Act and the 2010 Common Core Standards)—systematically squeezed out recess, shortened lunches, and reallocated informal social time toward academic instruction and structured peer interactions.
3. The Rise of Intensive Parenting: Driven by rising economic inequality, parenting has shifted heavily toward constant monitoring and adult-directed extracurricular activities designed to advance academic goals rather than foster autonomous peer relationships.

We are by no means the first to identify these trends; see the foundational work in this area by Peter Gray on free play, and research on intensive parenting by Matthias Doepke and Fabrizio Zilibotti. What is surprising is that these trends have been so widely ignored in empirical research on social media, with some popular narratives even claiming that the adoption of mobile phones and social media is responsible for trends like the decline of free play—which clearly started much earlier.

When research on social media studies adolescent well-being without measuring change in adolescent opportunities for informal, adult-free socializing, it suffers from severe omitted-variable bias. It misattributes adolescent distress to social media, rather than recognizing that screen use is an adaptive response to the loss of independence and unstructured time with peers.

What the longitudinal data actually shows:

To test whether social media use harms adolescent well-being once changes in offline social opportunities are accounted for, we studied a cohort of public school students in Michigan over three years—following them from early adolescence (grades 8/9) to late adolescence (grades 11/12).

Using cross-lagged panel modeling (CLPM), we controlled for autoregressive stability, baseline socioeconomic status, and time-varying changes in academic achievement, extracurricular participation, family time, and in-person peer socializing.

Our findings challenge the core tenets of the screen-harm argument:

• Social media use in early adolescence was associated with a subsequent increase in self-esteem for girls. Controlling for baseline self-esteem and participation in other social activities, girls who used more social media early on reported higher self-esteem in late adolescence. For boys, there was no significant relationship over time.
• Social media did not displace in-person friendships. Greater social media use was not associated with a drop in time spent hanging out with friends face-to-face. Instead, for girls, it was associated with spending less time with family—the exact same developmental substitution toward peer orientation that has long characterized healthy adolescence.
• Contextualizing effect sizes: Using internal benchmarking to interpret practical significance, the positive prospective association between social media use and girls’ self-esteem (β = .11) was similar in magnitude—and opposite in direction—to the drop in self-esteem associated with pandemic-era remote learning lockdowns (β = -.13), and comparable to the difference in self-esteem reported between boys and girls from early to late adolescence (β=-.10).

Digital spaces are a necessary “Stand In”:

Adolescents require voluntary, unstructured interaction with peers away from adults to develop social resilience, identity, and healthy coping mechanisms. Decades of research in sociology, communication, and human development have demonstrated that youth culture requires autonomous social territory.

When schools and parenting norms eliminate the physical and social spaces where teens can simply “hang out,” digital media functions as a stand-in—not a replacement. Online platforms do not destroy offline peer interaction; they preserve peer culture under conditions of severe constraint. Again, we are not alone in demonstrating the power that social media and other digital spaces have for helping young people maintain connections; see the pioneering work in the space by danah boyd, Mizuko Ito, amongst others.

Legislative pushes to ban or heavily restrict adolescent access to social media without addressing the deeper structural enclosure of childhood risk compounding the problem. Banning screens does not magically restore public spaces, extended recesses, or help adolescents find time and space for unstructured interactions. Instead, it threatens to sever one of the few remaining social lifelines that many young people have left to connect, build self-worth, and navigate the transition to adulthood together.

The full study—“The Constrained Generation: A Longitudinal Study of Adolescent Self-Esteem and the Misattribution of Harm to Social Media”—is open access at https://journals.sagepub.com/doi/full/10.1177/21582440261505080