Australia’s social media minimum age moves from law to long-term health surveillance
Australia’s Online Safety Amendment (Social Media Minimum Age) Act 2024 has triggered one of the first large-scale attempts to measure what happens when a government forces a hard cutoff on adolescent access to social platforms. The policy requires age-restricted services to take reasonable steps to stop Australians under 16 from creating or keeping an account, and the evaluation is built around a blunt practical question: what happens when digital access is removed at scale from early adolescents and the families around them?
The study is not a clinical trial in the usual sense, but its design borrows heavily from epidemiologic methods. Investigators plan to follow 4121 young people aged 10 to 16 years, along with their parents or caregivers, through a mixed-methods framework that combines repeated surveys, smartphone-based tracking, linked and unlinked administrative data, interviews, focus groups and open-ended responses. The protocol matters because policy effects in this age group are unlikely to show up in a single outcome. Social media use can shape sleep, attention, peer connection, school behavior, exposure to harmful content and family conflict, all of which can shift in different directions once account access changes.
One sentence captures the scale of the intervention: preliminary estimates suggest that about 1,194,000 young people may be affected by account deactivation.
Measuring adherence as a behavioral exposure, not a legal label
The evaluation frames “adherence” as a participant’s experience of account removal or deactivation and any attempt to bypass or circumvent those restrictions. That choice is important. In public health terms, the exposure is not simply whether the law exists, but whether a child actually loses access, whether that loss is partial or complete, and whether workarounds emerge through new accounts, altered ages or platform migration. The protocol also makes clear that children and parents are not being treated as the regulated parties under the law; the platforms are.
This distinction will shape how the data are interpreted. A teenager whose account disappears entirely is likely to experience a different psychosocial response than one who retains access through informal circumvention. For that reason, adherence is being treated as a key analytic variable in the planned causal inference work. The protocol names targeted maximum likelihood estimation, or TMLE, as the primary strategy for estimating the effect of adherence on outcomes such as psychosocial well-being and distress. In practical terms, that approach is meant to reduce bias while handling the messy reality of nonrandom behavior changes across families and age groups.
Five waves of family data will capture change over time
The core design is longitudinal. Parent-child surveys will be collected in five waves, allowing investigators to examine both early disruption and longer-term adaptation after implementation. Longitudinal follow-up matters because the consequences of losing social media access may not appear immediately. Some young people may feel relief if online conflict falls away; others may experience social exclusion, anger or a search for substitutes. Parents may report less concern about harmful exposure, or they may confront new battles over screen time, messaging apps and peer norms.
Passive smartphone tracking adds another layer of detail. With opt-in consent, the study can observe digital behavior beyond self-report, which should improve exposure measurement for time spent on devices and patterns of engagement. Linked administrative health and education data will help the evaluation move beyond attitude alone. Those records may offer a more objective view of school participation, service use or other downstream markers of well-being, although the protocol does not claim specific endpoints beyond the broad categories already named.
The analytical plan is ambitious. Descriptive analyses will examine frequencies, average values, percentages, factor analysis and reliability testing across survey waves. The investigators also anticipate using logistic regression, multilevel modeling and a random-intercept cross-lagged panel model to separate between-person differences from within-person change. In a policy experiment of this size, those distinctions matter because the most informative signal may lie in how the same child changes over time rather than in how one cohort compares with another.
Psychological strain, digital habits and family dynamics sit at the center
The planned outcomes extend well past screen time. Investigators will assess digital practices, physical well-being, psychological well-being, social well-being, family effects and broader community attitudes. That scope reflects a basic feature of adolescent development: social media is not only a communication tool, it is part of identity formation, peer maintenance and daily regulation of mood and belonging. Removing access could alter sleep timing, reduce exposure to disturbing content or increase face-to-face contact. It could also intensify loneliness or conflict if peers remain online while one child is cut off.
The protocol’s qualitative component is designed to capture those lived experiences. Interviews, focus groups and thematic analysis of open-ended survey responses should expose how families interpret the law, what workarounds appear, and which harms or benefits feel most immediate. Structured coding will help organize those narratives, but the real value lies in identifying mechanisms that numbers alone can miss. A parent’s report of calmer evenings, for example, carries a different policy meaning if it appears alongside a teenager’s account of isolation from friends.
Ethics, consent and policy oversight drive the rollout
The study received ethical approval from the Australian Institute of Family Studies Human Research Ethics Committee under application approval number 2025/06. It also complies with the National Statement on Ethical Conduct in Human Research and the Australian Institute of Aboriginal and Torres Strait Islander Studies Code of Ethics for Aboriginal and Torres Strait Islander Research. Consent was obtained from participants at baseline and again in each subsequent wave, and only parents and young people who provided consent are eligible for participation.
Those safeguards are not incidental in a study of this kind. Research on minors, digital behavior and family dynamics can expose tensions around privacy, autonomy and coercion, especially when the policy under evaluation is already controversial. The protocol’s consent structure and ethics framework are meant to keep the evaluation anchored in voluntary participation even as the policy itself imposes a compulsory age threshold on platforms.
The findings will be released through public reports, peer-reviewed publications and stakeholder engagement activities. Policy makers in Australia will be the immediate audience, but the implications are broader because several countries are watching closely to see whether age-based social media regulation can be enforced without creating new harms. The study is built to answer a difficult question with empirical rather than ideological tools: when access is restricted for children under 16, what changes first, what changes later and what kinds of family and community effects follow?