Bans Are What Politicians Reach for When They Stop Thinking

Sir Keir Starmer says banning under-16s from social media is “standing up for children”. A libertarian like me might be expected to say that championing youngsters by taking away their options is inherently contradictory. Yet I understand the impulse. As a father, I too worry about doomscrolling, pornography, bullying, grooming, and self-harm content online. Anyone who has watched a child hooked to a screen gets the appeal of support for pulling the plug too.

But the fact that something can be harmful is not itself a good argument for banning it. This is where economists think more carefully than politicians. Almost all activities in life have “good” and “bad” features. Cars cause fatal accidents. Excess food makes us obese. Playgrounds result in injuries. Conversations hurt our feelings. We accept, though, that banning these activities would come at an unacceptable cost, especially relative to less intrusive ways of mitigating the harms.

Politicians’ impulses tend to try to mandate “good” things and ban “bad” ones. The prime minister thus asks whether we can really say social media is “always” safe. We can’t, of course. But cars are not safe when driven at 90mph past a school, or medicines when taken at excessive doses. Sensibly, we choose speed limits, licences and prescriptions rather than bans, because cars also get us places and medicines save lives.

The assumption in banning under-16s from using social media platforms, then, is that, unlike these other things, the optimal use is zero. Really? A 15-year-old watching YouTube tutorials on coding is not like one scrolling through anorexia content. A teenager using X to follow politics is not equivalent to being hounded by anonymous online bullies. A socially awkward child deepening friendships on Snapchat is not suffering “screen time”. A moment’s thought shows social media produces more than harm. Any ban restricts activity that is benign or fruitful too.

In fact, the costs of a ban go beyond kids. To keep under-16s off platforms, companies must verify who is not under 16, requiring age-gating for everyone. Each individual check may feel minor in isolation. In aggregate, the time costs and loss of privacy population-wide will be large. 

Would a ban’s benefits outweigh these costs? Starmer acknowledges that, as in Australia, large numbers of children will find technological workarounds, including VPN use (virtual private networks) or migrating to less visible online spaces, so reducing the benefits. There will also be a cat-and-mouse regulatory game as new platforms incorporate social media-like features, before inevitable calls to extend the ban to them.

Now, one could still justify the policy if the net harms of social media were widespread and large, and more targeted regulation impossible. Yet both of those assumptions seem highly questionable too.

The assumed negative effects on mental health and wellbeing from social media use are smaller and more contested than politicians imply. Before the pandemic, Amy Orben and Andrew Przybylski, then Oxford academics, analysed three huge datasets, covering 355,358 adolescents in the UK, US, and Ireland. They found the association between screen time and wellbeing was negative, but that digital technology explained, at most, 0.4 per cent of variation in wellbeing. A tiny effect.

Broader reviews of very heavy social media use do find links with depression, anxiety, body image concerns and cyberbullying but, again, usually with only small to moderate effects, and only for some users.

Is this strong enough to ban social media use for all under-16s? I would argue not. Not least because there are more targeted ways to regulate harm, including crackdowns on online grooming, parents using technology to restrict their children’s screen functions, or headteachers enforcing phone-free schools.

Starmer would say that parents in polling and the consultation overwhelmingly support a broader ban. I will let therapists explore why so many parents prefer politicians telling their children “no”. But parents’ outsourcing preferences are, again, weak justifications for restricting everyone else’s liberties.