The Case for Letting Teenagers Online

There is a rare and suspicious level of consensus forming around teenagers and the internet. Parents, legislators, and a growing shelf of bestselling books agree that the phones are hurting kids and the fix is to get them off. Age verification, school bans, delayed smartphones, curfews on the apps. When a position becomes this unanimous this quickly, it is worth asking what the agreement is skipping over.

The strongest version of the case, argued most prominently by Jonathan Haidt, points at a genuine correlation between the arrival of the smartphone and a rise in adolescent anxiety and depression. The correlation is real and the timing is striking. But the leap from there to a simple causal story, phones in and wellbeing out, runs ahead of what the evidence can carry. The same period saw a financial crisis, a change in how distress is reported and diagnosed, and a general collapse of unsupervised outdoor childhood that predates the iPhone by decades. Pinning the whole shift on one device is tidy in a way that reality rarely is.

What the ban-first framing tends to erase is everything the messy online adolescence actually provides. For a teenager who is isolated by geography, temperament, disability, or a hostile hometown, the internet has been the one reliable place to find people who share an obscure interest or an unspeakable difficulty. The queer kid in a small town, the aspiring programmer with no local mentor, the teenager whose particular grief has no outlet at school: for all of them the connection has been genuine, and removing it in the name of safety takes something real away.

Growing up has always involved managed exposure to risk. We let teenagers ride bikes in traffic, take jobs, form intense friendships that will hurt them, and travel without an adult, because the alternative, a childhood sealed against all danger, produces adults who never learned to handle any. The internet is a domain where this logic seems to switch off, where the only acceptable amount of risk is none, and where the developmental value of learning to navigate a complicated public is treated as if it did not exist.

The honest position is not that the current arrangement is fine. Recommendation systems tuned for engagement can push vulnerable kids toward genuinely harmful material, and that specific mechanism deserves specific pressure. But there is a large distance between fixing what is actually broken and deciding that the correct dose of the open internet for a sixteen year old is zero. The messy version, in which teenagers make mistakes online and learn from them under some adult attention, is not a failure to be engineered away. It is what growing up has always looked like, moved to a new room.