Australia plans social media ban for under-16s

A furious scientific debate has erupted over whether digital technologies like smartphones and social media really are doing harm to the young people who use them, as governments around the world act to limit teenagers' access to these platforms.

The controversy, which was ignited by a recent influential book blaming phones for rising youth anxiety, has revealed deep uncertainties in the research evidence even as policymakers from Arkansas to Australia push ahead with sweeping bans and restrictions.

A timeline of the scandal
In March, social psychologist Jonathan Haidt of New York University published a science-pop book entitled The Anxious Generation. He blames increasing mental illness in young people over the past 15 or so years on the emergence of mobile phones and, by extension, social media.

One of the first reviews in Nature on Haidt's book by Duke University psychological scientist Candice Odgers reminded readers of a common criticism of many expert readers: though social media has been associated with bad outcomes, we don't know whether it causes those bad outcomes.

In April, some experimental studies on social media use in which researchers get people to cut back their social media use reveal benefits, said Haidt.

In May, psychologist Christopher Ferguson of Stetson University published a "meta-analysis" of dozens of social media experiments and found, all things considered, cutting down on social media does not make a difference to mental health.

Later in August, Haidt and his colleague Zach Rausch published an argument through a blog saying the Ferguson's method is flawed. They claimed that on the same analysis conducted in a different way it shows the impact really of social media on one's mental health.

Not much later on, one of us (Matthew B. Jané) published his own blog post pointing out that Haidt and Rausch's re-analysis had flaws but also the original Ferguson's meta-analysis was problematic. This post was also arguing properly that re-analysing Ferguson's meta-analysis doesn't provide any convincing evidence for influence in the field of mental health via social media.

Challenging Ferguson, Haidt and Rausch rewrote their post. In September and October they returned with two further posts, which pointed out more serious errors in Ferguson's work.
Jané concurred with the mistakes that Haidt and Rausch found, and has set out to rebuild Ferguson's database-and analyses-from scratch.

The discussion and additional work continues. Another group of researchers has just published a criticism (as a preprint, and thus without external review by other experts) that strongly disagrees with Ferguson, based on essentially the same poor methods used in Haidt and Rausch's first blog post.

The evidence is diverse – but not very strong
Why so much fuss? Part of the reason is that experiments where researchers get people to reduce their social media use produce varied results: some show a benefit, some show harm, and some show no effect.

However, the bigger issue, in our view, is simply that the evidence from these experimental studies just isn't very good.

One of the experiments incorporated within Ferguson's meta-analysis involved some German Facebook users not using the social media platform for two weeks, while others continued as normal. Participants then had to self-report their mental health and life satisfaction.
People who were asked to use Facebook less did report spending less time on the platform. There wasn't a detectable difference in depressive states, smoking behavior, or life satisfaction at any point in time between the two groups. The difference in self-reported physical activity was very small.

Another classic experiment involved 143 undergraduate students and randomized them into either reduce their usage of Facebook, Snapchat and Instagram for ten minutes a day for a month, or to not change anything. Then, the authors asked participants to report their anxiety, depression, self-esteem, autonomy, loneliness, fear of missing out and social support.

Most measures of mental health and wellbeing were not significantly different at the end of the month between the two groups, except that people who reduced their social media use tended to have small reductions in self-reported loneliness, while depression scores improved modestly among participants reporting high depression at baseline.

Big questions that existing social media experiments can't answer
Such studies answer very narrow, specific questions. They cannot possibly answer the big question of whether long-term reduction in social media use benefits mental health.

They look at specific platforms, not overall social media use. For another, most experiments do not really define "social media". Facebook is clearly social media, but what about messaging services such as WhatsApp, or even Nintendo's online gaming platform?

In addition, few use interventions or outcomes that could be measured objectively. They simply ask people—often undergraduates—to reduce their use of social media and then report how they feel. The obvious biases run the gamut, not the least of which is that the way people say they feel depends on whether they were told to do something in their life or not.

In the medical assessment of a drug's impact on mental health, it is not unusual to use a placebo-a substitute that should have no biological effect on the participant. Placebos are a very effective way in the attempt at avoiding bias since it ensures that the participant does not know whether the person he was tested with was the drug or not.

For social media reduction studies, placebos are practically impossible. You cannot trick a participant into thinking he is reducing social media when he is not.
 
Individual changes and a social problem
Moreover, all such studies work at the level of changing the individual behavior. However, social media is inherently social. If one college class uses Instagram less, then that might have nothing to do with their mental health even if Instagram is bad, since everyone around them is still using it as much as they ever did.

Finally, none of the studies included teens. As things stand, there is simply no valid evidence that encouraging teens to use social media less has any impact on their mental health.

Which brings us back to the obvious question. Does reducing social media improve teen mental health? With the existing evidence, we don't think it's possible to know.