A social media announcement for a social media age

This morning the Prime Minister announced the government will be banning under 16s from accessing certain social media apps from Spring 2027. I’m sympathetic to the idea of a social media ban for children, even though the evidence base on social media harms for children and young people is overall quite weak. But how you implement a ban is a really key question that the government have clearly failed to properly consider before hurtling into their announcement. It’s all very ‘something must be done’ regardless of whether it will be effective in achieving its objectives, and regardless of the unintended consequences. How very 2026.

There are two key questions for government policymakers:

• Is a social media ban the right kind of policy intervention at all?

• What would an effective ban look like, and how could one be implemented without introducing other costs, including age-verification for everyone wanting to access large swathes of the internet?

Banning social media for all under 16s is a big intervention. The argument for doing so is that accessing social media does young people significant harm. The psychologist Jonathan Haidt argued in The Anxious Generation that using smartphones and other similar devices can cause social deprivation, sleep deprivation, undermine children’s attention spans and become addicted. For some children it will also act as a gateway to harmful content, such as violent porn and extremist ideology.

But the evidence base is pretty weak. A government-published review of the evidence from earlier this year found studies show a small but consistent correlation between the time young people reported they spent on social media and self-reported mental health. Correlation is not causation, however, and this could be explained by young people with poorer mental health being drawn to spend more time on social media. That said, it is self-evident that access to social media means young people coming into contact with harmful content, such as on self-harm, suicide and porn.

The truth is we simply don’t know the long-term population-level effects of young people having unfettered access to social media apps and smartphones. That doesn’t necessarily mean we should wait for evidence about whether there is harm - which could be a very long time in the making - before acting, particularly if we think the benefits of social media to children are marginal.

Many people have a strong gut reaction that social media is not good for young brains because we can instinctively feel that social media’s addictive, polarising and dumbing-down qualities aren’t very good for us as adults. There’s no turning the clock back, but how many of us see social media as a net positive in our lives? I also suspect its educational benefits for children are easily over-exaggerated.

If you accept the case for intervention, there are two possible approaches. One is to try and make the open internet a safer place for children through regulation. The arguments in favour of this approach - favoured by campaigners like the NSPCC - are that it’s better to give children a taste of the world they will inhabit as adults, and the burden should fall on tech companies to make the world safe for children, rather than taking away children’s access to these platforms.

I think this approach is naive as to what governments and state regulators can effectively make multibillion dollar corporations do. Regulation is necessarily a blunt instrument, social media (let alone AI) is complex, and legal duties to users are difficult to define operationally. I am therefore sceptical that well-meaning but imperfect regulation can make social media platforms safe for children to roam unsupervised. I also don’t buy the argument that children need to be using open-access social media platforms in their early teens in order to understand how to use it as adults.