The social media hypothesis blames social media for the deterioration of teenagers’ mental health. However, many randomized controlled trials (RCTs) show that encouraging heavy social media users to quit for a short time has only small effects on mental health in the short term. This new preprint argues that these studies fail to convincingly show that social media is not to blame. Why? 3 reasons. 1.) The social media hypothesis concerns the effects of joining social media; RCTs identify the effect of quitting. 2.) The social media hypothesis concerns the cumulative effect of using social media for years; RCTs identify the effect of quitting social media for weeks. 3.) The social media hypothesis concerns a large-scale change—the widespread adoption of social media among teenagers. RCTs, in contrast, test small-scale effects of individuals quitting while most peers remain online. These local experiments may not capture the broader cultural and psychological impact of universal social media use. When social media use is ubiquitous, it can fundamentally alter the broader culture and social milieu, even affecting the mental health of teenagers who abstain from social media entirely.