Large correlational studies are very consistent in finding that teens who are heavy users of social media are more likely to be depressed than light users. But does heavy social media use cause depression, or does depression cause heavy social media use?

The gold standard for answering this question is a random assignment experiment. For example, researchers could randomly assign some people use social media their normal amount (the control group) while others stop using social media or cut back on their use (the experimental group). Many studies have taken this approach (for a list of these studies, click here). When these experiments last two weeks or more, they nearly always find that those who cut back or eliminated social media use are happier and less depressed.

These studies are actually even stronger evidence than they might first appear – they ask average users to cut back to light or no use. In the correlational data, the largest differences are instead between average users and heavy users. If experiments asked heavy users to cut back their use for several weeks, effect sizes would likely be even higher.

One set of researchers took advantage of a natural experiment: Facebook rolled out at different times across college campuses. Sure enough, mental health issues rose among students when their campuses adopted Facebook.

There’s also the question of whether we’re considering individuals or groups. At the group level, when trying to explain why teen depression rose so much since 2012, it seems more clear that social media use (and/or the prevalence of smartphones) cause depression, and not the other way around. If depression caused social media and smartphone use, you’d have to argue that an unknown factor caused teen depression to increase, which then led teens to buy smartphones and start using social media more. That seems pretty unlikely – the technology came first and grew in popularity, and then teen depression rose.