The Wall Went Up. Nobody Was at the Gate.
Australia passed a law in 2025 that banned social media for anyone under sixteen. The idea was straightforward: draw a line, keep kids off platforms that were allegedly wrecking their mental health, and call it done. The news cycle treated it like a foregone conclusion — another victory for evidence-based policy.
It didn't work. Not even a little.
A new study published in the BMJ tracked 436 Australian teenagers before and after the ban went into effect. Three months in, more than 85 percent of those under sixteen were still on the restricted platforms. Most through their own accounts. The age verification? You typed in a birthday and got waved through. The platforms didn't really enforce anything. As the researchers put it, with characteristic understatement: policy implementation by the platforms did not seem to have prevented adolescents from using them.
The wall went up. Nobody was at the gate.
The U-Shaped Curve Nobody Expected
Here's where the story gets interesting — and uncomfortable for anyone who wanted a simple narrative.
A separate study in JAMA Pediatrics followed over 100,000 Australian adolescents across three years. The researchers weren't looking at enforcement. They were looking at outcomes. What they found was a U-shaped correlation between social media use and well-being.
Moderate users had the best outcomes. Both heavy users AND kids who didn't use social media at all fared worse.
And there was a gender dimension that caught my attention. For boys specifically, having no social media at all became increasingly problematic as they moved through adolescence. By the late teens, the risks of total abstinence actually exceeded the risks of heavy use. That's a finding that should make anyone who supports blanket bans pause and think.
The takeaway isn't that social media is harmless. It's that the relationship between usage and well-being is far more nuanced than policymakers — and parents — tend to assume.
The Study That Found Almost Nothing
Then there's the University of Manchester research, which followed 25,000 eleven-to-fourteen-year-olds across three school years. The result? Essentially zero effect of social media use or gaming on anxiety and depression symptoms.
They even checked the distinction that dominates parental worry — mindless scrolling versus active chatting. Neither pattern drove mental health difficulties. The type of use didn't matter.
I'll be honest: this finding surprised me more than the enforcement failure. We've built an entire anxiety industry around screen time, and a study this large finding essentially nothing is... humbling. It doesn't prove social media is good for kids. But it does suggest we've been dramatically overestimating the scale of the problem.
The Causation Problem We Can't Ignore
All of these studies share a critical limitation: they're correlational. We don't know which direction the causation runs.
Does social media make kids worse off? Or do kids who are already struggling use social media more as a coping mechanism? Candice Odgers, one of the field's most rigorous researchers, suggests the reverse causation is very real: struggling kids use social media more precisely because they aren't finding enough support in their daily lives.
This matters. A lot. Because if the causation runs the other way, then banning social media doesn't fix the underlying problem. It just removes a coping mechanism from kids who need it most.
We simply don't know yet. And that uncertainty should make us humble about how aggressively we regulate.
Why Blanket Bans Miss the Mark
Here's what I think, and why I think it:
Banning social media for teenagers is like trying to put the genie back in the bottle. For a sixteen-year-old, these platforms aren't just entertainment to be rationed. They're where a substantial part of social life happens now. The group chat lives there. Plans are made there. A kid figures out where they stand with their peers there.
Cutting them off doesn't return them to some pre-digital idyll. It removes them from the room where their friends are.
Social media certainly can cause problems for some people. I've seen family members who are terminally online in ways that seem to wind them up. I've watched nieces and nephews make negative social comparisons between themselves and the highly curated world their peers present. But this is a conditional problem — dependent on the particular person, their broader life context, and how they use these platforms.
Conditional problems don't yield to unconditional solutions.
What We Should Do Instead
The researchers who actually study this for a living typically recommend something less soundbite-able than a blanket ban: regulate specific design features that hook kids rather than evicting them entirely.
This means targeting the algorithms, the infinite scroll, the notification systems — the things that are actually designed to be addictive. Not the kids.
It's a more complicated policy conversation. It requires technical expertise and sustained political will. But it's also the approach that matches the complexity of the actual problem.
Australia's ban was a bold experiment. The results so far suggest it was the wrong experiment. Not because social media is harmless, but because blanket bans are too blunt an instrument for a finely textured problem. We need smarter regulation, not bigger walls.