The Law as Lab

Britain is moving to keep under-16s off social media, and — as New Scientist reports — researchers are scrambling to turn that policy into a measurable experiment before enforcement kicks in. Nobody planned it as science. But a law dropped onto a population is an intervention whether you intended one or not. Flip a switch on a few million teenagers and you've got a natural experiment: no ethics board, no consent forms, no control group that anyone designed on purpose. The universe handed us a trial and we ran it by accident, on children, because we couldn't wait for the data.

Which is the quietly hilarious truth about how a species this clever actually governs itself. Most of our biggest decisions are intuitions wearing the costume of certainty. We feel that something is hurting kids — the anxiety charts, the doomscroll, the blue glow at 2 a.m. — and we act on the feeling, because the feeling is loud and the evidence is slow. The inconvenient part: "social media harms adolescents" is still, scientifically, a contested claim. The correlations are real. The causation is a swamp. Plenty of careful researchers think the panic outpaced the proof.

So we get the full comedy of governance: do the thing to everyone, then find out whether the thing was worth doing.

And the finding-out is genuinely hard. How do you measure the effect of an absence? You can't ethically assign a control group of kids who are allowed to keep scrolling. You have to compare against the past, against other countries, against the teenagers who quietly route around the ban with a VPN and a fake birthday. Some kids will get their evenings back. Some will migrate to whatever the ban forgot to name. Some will be exactly as anxious as before, for reasons that had nothing to do with their phones. Untangling those threads, after the fact, out of a system you've already perturbed — that's the job the researchers signed up for.

Coherenceism has a phrase for this: alignment over force. A ban is force. It compels. What the studies are really asking is whether the force produced alignment — kids whose lives actually got lighter — or merely compliance, the brittle kind that shows up as a workaround and a lie about your age. Force can change a number on a dashboard without changing anything that matters. The only way to know which one you got is to measure, carefully, after the fact, and be willing to hear an answer you won't like.

It's worth sitting with how backwards this is, and how unavoidable. We are matter that learned to legislate before it learned to measure. We'll keep doing it — the harms feel too urgent to wait, and they might be. But the honest version of the move is the one Britain's researchers are quietly attempting: pass the law, yes, and then have the humility to check your work. The experiment is already running. The least we can do is read the results.