Would social media ban for children work here? Australia offers lessons.

Australia became the first country to prohibit those under age 16 from using major social media platforms in December 2025. The aim was to limit the potentially harmful cognitive, social, and mental health effects on children.

So far, the results do not look promising. 

Researchers studying the bans early effectiveness found nearly 75 percent of Australia’s 14- to 15-year-olds were not complying, in large part because they perceived so few others were doing so, according to a new National Bureau of Economic Research working paper. 

Those who did adhere were seen as out-of-the loop socially, while violators faced no penalties. 

In this edited conversation, Cass Sunstein, Robert Walmsley University Professor and one of the paper’s co-authors, explains why such a ban is probably not the cure-all officials envision and suggests some tweaks to ramp up compliance.

---

Why does it appear the ban isn’t working?

One reason why it’s not working, certainly not as hoped, is that there’s no strong incentive for young teenagers to comply.

They don’t face punishment if they don’t comply, and they don’t get a reward if they do comply. So, from the standpoint of a 14- or 15-year-old, this may be background noise.

Another reason they don’t comply is that there’s potential exclusion from a group because most people are not complying. 

The third is that non-compliers give a certain signal that they’re part of the uncool group, and that’s not what people want to give.

Was it surprising to see such poor compliance just months after the ban was put in place?

The results did surprise me. The magnitude of violation is very high. I’m not shocked by it, but I wouldn’t have expected it.

The disparity between the power of social norms and the power of the legal restriction is not amazing but is noteworthy. It’s observed in some contexts, but not in others. The fact that it’s so clear here is striking.

Teens said what has kept them on social media was not wanting to feel disconnected from friends and a general fear of missing out, or FOMO, a social dynamic you say is understudied?

Some members of the team, including me, have been working simultaneously on the power of the fear of missing out. What we’re looking at is why for Instagram and TikTok, potentially for smoking, for getting tattoos, for a wide range of things, people will buy — with time or money — something whose existence they deplore.

We find that among many college students in the U.S., a lot don’t like the existence of TikTok in their community, and a lot don’t like the existence of Instagram either. And if they could push a button and get rid of them, they would. But so long as they exist, they’re going to use them.

This shows the intense power of inclusion in a group, even when the thing that is necessary for inclusion is unwelcome to the people who use it or buy it. So that’s a simultaneous research project.

And while we don’t know that in Australia people wish social media would go away, we do know that young people are going to stay on so long as other people are staying on, even if the law wants them off, and that’s very noteworthy. 

The paper is exemplary of an understudied phenomenon, which is the extent to which people do things because other people are doing them in circumstances in which they wish other people weren’t doing those things. But it’s really hard to create a different equilibrium, even if the law is on the side of the different equilibrium.

For a ban to succeed, the percent of people abiding by it has to reach a certain tipping point before staying off social media becomes the new normal. What’s that tipping point and how might it get pushed?

We have their self-report, which says in the vicinity of three-quarters have to be off in order for participants who are now on to get off. Now that’s self-report and not as good as actual behavior, but there’s good reason to think from our data that things would tip if three-quarters were off. 

We observe this with respect to smoking in the U.S. — the percentage is way down. There’s also been a reduction in drinking. For some earlier generations, if you didn’t drink, you looked like you’re not interested in fun or are a very frightened person. 

If Australia could get the number up high, somewhere north of 70 percent, then you could get the number of users way, way down. 

There are ways to move incrementally toward there. There’s separate data, not in the paper, suggesting that if people learn that there’s an emerging norm, that knowledge tends to create a self-fulfilling prophecy. 

I’m interested in the possibility of Australia or some country having an educational campaign that invokes an emerging norm in favor of not being on or not being on so much. That would be great.

A widely held perception that all the cool kids were still using social media shaped teens’ behavior. The idea was those who complied with the ban were implicitly labeling themselves as uncool?

Isn’t that a striking finding? 

One thing that we’re all interested in is the social meaning of action. If I buckle my seat belt these days, the social meaning is “I’m doing what you should do.” At one point, if you buckled your seat belt, the meaning was, “You think the driver is reckless or you yourself are a coward.” None of us is in control of the social meaning of our action.

So, if you’re off social media, you label yourself as a certain type. That’s just observed, but our data shows that the kids who stay off social media are seen as part of the unpopular kids. And for Australia, that’s a very serious problem.

What are some reforms that could keep teenagers off these platforms?

One would be a public education campaign designed to encourage teenagers to stay off. It could point to the benefits of not spending your time staring at a screen. It could point to the existence of the law. It could point to the increasing numbers of people who are staying off, if that’s true. It could point to the advantages of seeing your friends in person. 

A second kind of approach would give incentives to teenagers to stay off. That would be a new policy, but it could be that teens who stay off would get rewards like free concert tickets or discounts on certain things, and if they’re not that expensive, but have more symbolic value than economic value, it could be that they’d have a real effect. We don’t know, but it could be part of a package of responses. 

A third thing that could be done would be to tweak the policy. About two-thirds said they want time limits on their use rather than prohibitions. That would have some disadvantages as well as advantages, but it’s at least worth considering. That might be something that would reduce the FOMO problem.

Is this a problem that could be solved using strategies other than blanket bans?

I don’t believe, and my co-authors don’t believe, that this is an insoluble problem. 

If Australia wanted to change the numbers, it has a host of strategies. It could have a campaign designed to try to alter the social meaning of not being on social media. 

It could use young validators who would say, “ I’m off and happier.” It could use cool young validators who kind of look the part say, “I’m off and I’m happier.” It could appeal to parents saying, “Keep your kids off,” and it could have some slogan or some account of why that’s a good thing for parents to do. It could try to tackle the collective action problem directly by saying, “Let’s stay off together,” or something like that. 

It could alter the policy from a flat ban to time limits. It could enlist the social media platforms more aggressively than it has so far. The social media platforms are subject to fines if they don’t take reasonable steps. It looks as if that’s a soft enforcement policy. 

So now that we have some real clarity, not perfect, but real clarity, about why kids are not complying, we know what Australia might do, or what American states might do, or what any country might do if it wants to reduce the number of young people on social media.