There’s a tremendous amount of controversy over whether social media use is the culprit behind recent dramatic increases in the rate of self-harm among adolescent girls in the United States and the United Kingdom. Jean Twenge, a psychology professor at San Diego State University, and her colleagues have argued yes: depressive symptoms, suicidal ideation, and suicide attempts are all positively correlated with the amount of social media use. Moreover, the first generation to experience adolescence after iPhone saturation is markedly more depressed and more suicidal than their predecessors. Change between cohorts is usually gradual, with big changes showing up between generations, but not usually in only five years. Others have concluded that the topic should not be sensationalized without more rigorous research, particularly given that social media use is a terribly weak predictor of depressive symptoms—predicting 0.36% of depressive symptoms among females and not having a significant effect at all among males.
I have found that thinking about this issue as a family researcher makes me sympathetic to Twenge’s far-from-conclusive evidence. First, family process can buffer between risk and harm. Even if every teen with a smartphone was at increased risk of depression, not all will experience depression. This became very clear to me after Thanksgiving dinner when my brother-in-law alluded to some “mutual coercion” that his college-aged son’s friends practice: when they go out to eat together, they put their cell phones in the middle of the table. The first person to pick up his phone picks up the check for all; if all the phones stay out of reach, it’s Dutch treat. I was impressed that a group of college freshmen agreed that paying attention to each other is more important (or at least more urgent) than whatever may be going on elsewhere, but it is also significant that my nephew’s family knew about the practice.
My nephew has both the relationships and the relationship styles that make it less likely that he will experience whatever harm may be associated with social media use. He seems to illustrate patterns that Reynaldo Rivera, David Santos, Marc Grau, and I found in data from the 2010 EU Kids Online Project. Children with two parents in the household were less likely than other children to use the Internet in problematic ways (e.g., to go without eating or sleeping because of the Internet). However, the advantage associated with having two parents in the household was completely explained by differences in “relational lifestyles.” What this means is that children in two-parent homes were more likely to take input from their families when making decisions, rather than deciding in solitude or having difficulty making decisions. Thus, the lower rates of Internet abuse among children in two-parent families wasn’t a direct result of having two parents (e.g., more accountability or more intense monitoring), but was an indirect result of children in two-parent families having more adaptive (and more interactive) modes of decision-making.
While Twenge comes down strongly on the side of limiting screen time, others put much more emphasis on identifying the conditions under which risk translates into harm. Time exposed to social media may not be among the most important factors. Instead, the family and social relationships that contextualize use may matter more than the screens themselves. Even if my nephew spent excessive amounts of time on Facebook or the Internet when he was alone, he has the social relationships that might keep that risk from translating into harm.
Second, being a family researcher helps me see that even if social media explained only a tiny fraction of depressive symptoms, it could still be important because of the research on divorce and children’s college completion. Divorce doesn’t seem to matter for the college completion chances of the child of two high school dropouts, probably because that kid was unlikely to finish college whether or not the parents stayed together. In contrast, divorce reduces college completion rates among those who would have had a reasonably good chance of making it through college without family disruption. With this kind of reasoning, a small overall effect could be consistent with an important effect in some subgroups.
Finally, even if social media had a universally negative effect on teens—increasing depressive symptoms in all of them—only a subset would “cross over” into being clinically depressed. I have taught many children of divorce at Georgetown University who have subsequently graduated. That’s because many individuals do well even in groups that are disadvantaged.
In sum, I do not agree with other critics that social media consumption predicting only a tiny fraction of all depressive symptoms means that Twenge’s hypothesis is wrong. I do agree that better research on why depression and suicide are increasing is imperative. In the meantime, I am a strong advocate for parents being involved in their children’s lives and their technology consumption.
Laurie DeRose is a Research Assistant Professor at the Maryland Population Research Center at the University of Maryland at College Park, where she has served since its inception in 2001. She is also Director of Research for the World Family Map project.
Editor’s Note: The views and opinions expressed in this article are those of the author and do not necessarily reflect the official policy or views of the Institute for Family Studies.