Amidst years of gridlock and polarization on technology policy issues, social media regulation is emerging as a rare area of bipartisan consensus. The idea is so popular that a policy group created to help shape the next Democratic Presidential platform—“Project 2029”— views “online internet safety” as its starting point, analogizing the idea to this generation’s tobacco moment.
The plan—“Kids Over Clicks”—rests on two empirical claims. First, social media is a major cause of declining adolescent mental health; second, that platform design is inherently addictive. While public academic figures like Jonathan Haidt have come out in support of this plan, a closer look at the evidence shows that neither claim is as well-established as current talking points suggest.
Wobbly Evidence On Social Media & Mental Health
Researchers have produced a large volume of research. But as a recent report from the American Consumer Institute explains, across all study types—correlation, longitudinal, and even experimental—the evidence is mixed and weak. Even the positive studies linking social media use to clearly defined benefits—such as well-being, improvements in physical and mental health, and strengthening relationships —are mixed and weak. Taken together, the only real conclusion is we still don’t understand how social media impacts well-being.
Correlational studies are weak because, by definition, they cannot establish causality and most, if not all, the findings fall in a range that is statistically considered of low significance. Longitudinal studies, which analyze the effects of social media over time, suffer from self-reporting bias, confounding issues, and reverse causation—all of which make them less reliable for isolating social media’s specific contribution. And experimental and causal studies tend to focus on adults, not kids, and often fail to effectively isolate the role of pre-existing mental health conditions.
In many cases, studies acknowledge the lack of concrete evidence. For instance, a study from the journal Social Science & Medicine – Mental Health analyzed 32 studies that assessed the impact of social media restrictions on well-being, including anxiety and depression, and concluded: “small observed effect sizes suggest that restriction is likely not the most effective method of improving subjective well-being in today’s digital age.”
The Kids Over Clicks plan from Project 2029 therefore turns on a faulty premise: despite what the critics may say, social media has not precipitated the decline in teen mental health many assert as fact.
More Wobbly Evidence On Social Media & Addiction
The plan also heavily leans on addiction claims—which provide the single, strongest piece of evidence supporting mental health problems from social media use. The claims spring from a recent court ruling in California that found that Meta and Google platform designs—on Instagram, Facebook, and YouTube—use inherently engaging features like infinite scroll, autoplay notifications, and variable reward structures to keep users on their screens.
To be clear, companies do have an incentive to maximize engagement and encourage users to spend time on their apps. But the science isn’t settled, the claims are poorly contextualized, and the evidence is often misunderstood—sometimes intentionally. Unlike the many studies where academics have tested the effects of social media and poor mental health outcomes, finding mixed results, clinical addiction has been studied far less. Academics acknowledge this gap, but public figures have treated that research gap as evidence of addiction and as settled science.
Associations between problematic social media may exist in some narrow cases, but psychiatry does not recognize social media as addictive. And the authoritative handbook published by the American Psychological Association (APA), the Diagnostic and Statistical Manual of Mental Disorders, most recently the DSM-5, has still not labeled social media use as addictive.
The imbalance between the political narrative and ongoing research points to a disconnect between critical evidence and the broader public perception of harm, which the Kids Over Clicks exploits as a critical plank of Project 2029.
The current bipartisanship on teen online safety has given this plan more momentum. However, bipartisan agreement is no substitute for strong empirical evidence. If the priority is to address teens’ mental health, at the very least, as lawmakers advance proposals like these, they should ensure that the evidence supporting them is as robust as the policies themselves.
Pablo Garcia Quint is a technology policy analyst with the American Consumer Institute, a nonprofit education and research organization. Follow Pablo on X @gq_pablo