{"id":31332,"date":"2026-10-19T04:54:00","date_gmt":"2026-10-18T20:54:00","guid":{"rendered":"https:\/\/csccm.org.cn\/?p=31332"},"modified":"2026-10-19T05:42:29","modified_gmt":"2026-10-18T21:42:29","slug":"bmj%e5%8f%91%e8%a1%a8%e8%bf%b0%e8%af%84%ef%bc%9a%e5%ad%a6%e4%b9%a0%e6%be%b3%e5%a4%a7%e5%88%a9%e4%ba%9a%e7%a4%be%e4%ba%a4%e5%aa%92%e4%bd%93%e5%b9%b4%e9%be%84%e9%99%90%e5%88%b6%e6%94%bf%e7%ad%96","status":"publish","type":"post","link":"https:\/\/csccm.org.cn\/?p=31332","title":{"rendered":"[BMJ\u53d1\u8868\u8ff0\u8bc4]\uff1a\u5b66\u4e60\u6fb3\u5927\u5229\u4e9a\u793e\u4ea4\u5a92\u4f53\u5e74\u9f84\u9650\u5236\u653f\u7b56"},"content":{"rendered":"\n

Editorial<\/strong><\/p>\n\n\n\n

Learning from Australia\u2019s social media age restriction policy<\/h1>\n\n\n\n

Amrit Kaur Purba<\/h3>\n\n\n\n

BMJ<\/em>\u00a02026;\u00a0393\u00a0doi:\u00a0https:\/\/doi.org\/10.1136\/bmj-2026-100017<\/a>\u00a0(Published 24 June 2026)Cite this as:\u00a0BMJ<\/em>\u00a02026;393:e100017<\/h3>\n\n\n\n

Evidence suggests that legislation alone is not enough<\/p>\n\n\n\n

Governments globally are considering age restrictions for social media amid concerns about online harms and mental health.1<\/a>2<\/a>3<\/a> In June 2026 the UK government announced that it would bring in legislation, explicitly modelled on Australia\u2019s approach, to ban social media platforms from offering services to under 16s, with protections expected by spring 2027.4<\/a> Yet the first prospective evaluation of a national social media age restriction has reported largely null findings, and interpreting them is not straightforward. In a linked study (doi:10.1136\/bmj-2026-363695<\/a>), Barnes and colleagues report limited evidence of short term effects of Australia\u2019s Social Media Minimum Age Act on adolescents\u2019 social media use.5<\/a> The findings will inevitably shape debates in countries considering similar policies. The central question, however, is not whether the policy worked. It is whether the policy was implemented with sufficient fidelity to allow that question to be answered at all.<\/p>\n\n\n\n

Public health policies can fail because interventions are ineffective, but they can equally fail because implementation is incomplete, compliance is limited, or populations adapt in unexpected ways. Distinguishing between these explanations is a central interpretive challenge, given their very different implications for policy and practice.<\/p>\n\n\n\n

In an observational study, Barnes and colleagues followed 436 Australian adolescents aged 12 to 17, surveyed shortly before the Act took effect in December 2025 and again around three months later. The study documented striking implementation challenges. More than 85% of adolescents under 16 continued to use restricted platforms at three months\u2019 follow-up, predominantly through their own accounts, and a third reported experiencing no age verification measures across platforms subject to the Act.5<\/a> This may reflect platforms\u2019 reliance on self-declared age rather than more robust verification methods recommended by regulators in the UK and Europe.5<\/a> What these figures collectively describe is a partially implemented policy, one in which the mechanism intended to restrict access was not reliably activated. Whether the intervention was delivered with sufficient fidelity - that is, consistent and effective implementation of age assurance measures to allow reliable inferences about its effects - remains doubtful.<\/p>\n\n\n\n

Adolescents also continued to access platforms through others\u2019 accounts or through accounts containing false age information.5<\/a> Whether circumvention differs across adolescent populations is unclear. If some groups are better positioned to bypass restrictions than others, digital regulation could widen existing health inequalities - an unintended consequence that future evaluations should explicitly examine.6<\/a> Implementation challenges should not, however, automatically be interpreted as policy failure. The findings highlight the importance of distinguishing the delivery of a policy from its effectiveness - a distinction well recognised in tobacco control and gambling regulation, in which implementation fidelity is a precondition for effectiveness.7<\/a>8<\/a>9<\/a>10<\/a><\/p>\n\n\n\n

The linked study used a regression discontinuity design - one of the strongest quasi-experimental approaches available when randomisation is impossible.11<\/a> Such designs estimate the effect of legal exposure to a policy, not the effect of actual restriction of social media access. That distinction is crucial. Where compliance is limited and circumvention widespread, null findings cannot distinguish between an ineffective policy, an insufficiently implemented one, or a study unable to detect modest effects under real world conditions. The implementation findings may therefore be more informative than the causal estimates, because implementation is the mechanism through which any policy effect must operate.<\/p>\n\n\n\n

Digital policy interventions operate within complex adaptive systems.12<\/a>13<\/a>14<\/a> Platforms, users, and technologies adapt in response to regulation, generating feedback loops and unintended consequences that complicate evaluation. In the linked study, younger adolescents in particular reported increased use of messaging apps and outdoor recreation following implementation of the Act.5<\/a> These early signals are worth tracking. Substitution towards alternative platforms could dilute or redirect harms rather than eliminate them.15<\/a><\/p>\n\n\n\n

Evaluating such interventions also requires consideration of how effects vary by platform, content, and engagement pattern.16<\/a>17<\/a>18<\/a> A platform specific intervention such as the Act may produce effects that aggregate use measures cannot capture, including harms that shift rather than disappear as adolescents migrate to unrestricted environments. Different platforms carry different risk profiles, and which platforms adolescents use determines which harms they are exposed to. A systematic review including more than one million young people suggests that social media may influence outcomes beyond mental health, including alcohol use, vaping, and gambling.19<\/a> Future evaluations should focus on outcomes beyond mental health to capture the full effects of the Act.<\/p>\n\n\n\n

No single study design will resolve the questions that this policy raises. In a policy debate too often driven by assertion rather than evidence, rapid collection of real world data such as this is an essential contribution, but capturing wider effects will require systems approaches that track substitution, adaptation, and unintended consequences across populations and over time. Future evaluations should examine implementation fidelity, compliance, circumvention, and platform migration.20<\/a> Methodological pluralism - combining natural experiments, longitudinal studies, qualitative research, and objective digital measures - will be needed.<\/p>\n\n\n\n

Australia\u2019s experience shows that legislating a restriction is not the same as enforcing one: when age assurance relied on self-declared age, most adolescents continued to access restricted platforms. Countries now adopting similar measures - including the UK, which has committed to comparable restrictions and has tasked its regulator with defining effective age assurance before implementation - will need those mechanisms in place from the outset, rather than retrofitted once circumvention is already widespread. As governments across Europe, North America, and elsewhere consider similar approaches,4<\/a>21<\/a> Australia\u2019s experience suggests that implementation may matter as much as legislation, and that lesson may prove as consequential as any headline result.<\/p>\n","protected":false},"excerpt":{"rendered":"

Editorial Learning from Australia\u2019s social media age re […]<\/p>\n","protected":false},"author":3,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[24,23],"tags":[],"_links":{"self":[{"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/posts\/31332"}],"collection":[{"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=31332"}],"version-history":[{"count":1,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/posts\/31332\/revisions"}],"predecessor-version":[{"id":31333,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=\/wp\/v2\/posts\/31332\/revisions\/31333"}],"wp:attachment":[{"href":"https:\/\/csccm.org.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=31332"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=31332"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/csccm.org.cn\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=31332"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}