A teenager who quits social media gives up more than just that. She often becomes the only one missing from the group chat where plans get made.
The value of social media depends on who else is using it, which means that most teenagers are unlikely to want to cut back alone. Any solution requires collective action.
That sounds, on its face, like a strong argument for a ban. If individual exit is too costly to be self-sustaining, the law can do what individuals cannot — coordinate everyone so that they exit at once. In December 2025, Australia became the first country to try barring anyone under 16 from holding accounts on Facebook, Instagram, TikTok, Snapchat and six other major platforms. The ban, which is enforceable under the law, was instituted to protect children from, among other things, access to pornography and violent material.
We are sympathetic to Australia’s goals. Four months in, we conducted two systematic surveys to learn about how Australian teenagers are responding. The collective-action logic that makes a ban appealing in theory is the same logic that appears to be making this one fail in practice, at least so far.
To succeed, a ban has to push compliance past a tipping point — staying off becomes the new normal and nonusers no longer suffer from FOMO. The Australian ban hasn’t done that, and under the current design it won’t.
Both of us have long been interested in social norms, the limits of law and how people can, or cannot, coordinate to achieve shared goals. Why, exactly, has the Australian approach failed to achieve its purposes? About a dozen countries are now considering similar laws, which have also been enacted or are under consideration in numerous American states. What might be done instead?
Let’s start with enforcement. The Australian law requires “reasonable steps” from social media platforms to keep those under 16 off. If they do not take such steps, the platforms are potentially subject to fines. But individual teenagers face no penalty at all. In fact, 78 percent of the teenagers we surveyed believe there are no personal consequences for violating the ban. They are right.
Most teenagers also describe circumvention as easy: false birth dates, a parent’s account, a sibling’s login, a VPN. Any one of these works. So whether the ban succeeds depends almost entirely on whether teenagers choose to comply.
Only about 27 percent of 14- and 15-year-olds do. Roughly two-thirds report using a banned platform in the past week. That is nowhere close enough to the tipping point we’d need to see.
We asked Australian teenagers directly: what share of your peers would need to stop using social media before you would? The average answer was about 70 percent. And the average understates the problem. Even if compliance hit that share, fewer than 75 percent of teenagers would have reached their own personal threshold. The tipping point is not where the average sits; it is where enough teenagers with low thresholds have flipped to pull the rest along.
To make things worse: When we asked teenagers whether peers who comply with the ban are more popular or less popular than those who do not, 45 percent said less popular. Only 4 percent said more.
This is the cigarette story in reverse. Smoking rates did not plummet in the United States and other nations because of taxes alone. They plummeted because the social meaning of smoking flipped. With the help of public-health campaigns, a lot of people quit, and smoking no longer became important to maintain social connection.
A close parallel for today’s teenagers: Members of Generation Z drink far less than the generations before them, even though the legal drinking age has not changed since 1984. The age limit clearly did not do that. Norms around drinking did. Laws aimed at youth behavior succeed when norms move with them, and stall when norms move against them.
In Australia, the norm is not to act consistently with the social media ban. There is good reason to worry that, at least now, the same story would hold for the United States, the United Kingdom and many other nations.
A more effective approach would need to act on social norms and incentives directly. In addition to working with social media platforms to improve enforcement, Australia should consider three reforms.
First, the ban should be supplemented with tools meant to change norms. For example, Australia might adopt public information campaigns to publicize the potential harms of social media use and improve attitudes toward those who stay off the relevant platforms.
Second, consider offering real rewards to teens. The government could subsidize free concert tickets, transit credits or discounts of various kinds. Parents could offer extra allowance conditional on their children not using social media platforms or exceeding screen time limits.
Third, take seriously what teenagers prefer. Seventy-two percent told us they would rather have time limits on their use than no use at all. Time limits are most effective when they are adopted in coordination across a peer group, so that no teenager pays the social cost of cutting back alone.
Australia took a serious and important step in adopting its new law. But nations all over the world, beware: When the value of a product comes from collective participation, a mandate alone might not be enough to change behavior.
Tipping points can be crossed, as the smoking decline shows. The lesson there is that laws are most likely to change behavior when norms move with them. When norms move in the opposite direction, laws can be an exercise in futility.
Leonardo Bursztyn is a professor at the University of Chicago where he directs the Normal Lab. Cass R. Sunstein teaches at Harvard, where he directs the Program on Behavioral Economics and Public Policy.
The Times is committed to publishing a diversity of letters to the editor. We’d like to hear what you think about this or any of our articles. Here are some tips. And here’s our email: letters@nytimes.com.
Follow the New York Times Opinion section on Facebook, Instagram, TikTok, Bluesky, WhatsApp and Threads.
