Is Facebook harmful? To some, the answer is clear. But we should proceed with caution.
“Facebook’s products harm children, stoke division, weaken our democracy and much more,” Facebook insider turned whistleblower Frances Haugen recently alleged before a Senate subcommittee. To back up her claim that Facebook is harmful to children, Haugen referenced internal Facebook research showing that 13.5% of teenage girls said suicidal thoughts became more frequent and 17% of girls said that their eating disorders worsened after using Instagram (which is owned by Facebook). “Facebook, over and over again, has shown it chooses profit over safety,” Haugen asserted during the subcommittee hearing. Republican Sen. Marsha Blackburn concurred, stating that “It is clear that Facebook prioritizes profit over the well-being of children and all users.”
It would not be surprising, some might think, if Instagram is harmful to adolescent females in the ways reported by Facebook’s internal investigation. Instagram provides a bombardment of idealized images that many may feel they fall short of. Facebook’s algorithms, moreover, direct users’ focus toward attention-grabbing content that may lead to or reinforce unhealthy habits or behaviors. Given that Facebook itself — which has a financial incentive/bias toward showing that its products are not harmful — has found a link between Instagram use and negative mental health consequences, it may seem obvious to conclude that Instagram harms young women in the ways suggested.
Caution in drawing definitive conclusions based on Facebook’s research, however, is warranted.
In a recent editorial, Temple University psychology professor Dr. Laurence Steinberg, who has researched and written extensively on adolescent development, pointed out that Facebook’s investigation did not follow generally accepted standards of academic research for testing causal hypotheses.
Causality is best established through controlled, randomized trials. In such a setting, a researcher has control over a treatment that is randomly assigned to participants. To determine whether engagement on a social media site has negative consequences, for example, a researcher conducting an experiment could randomly assign some participants to engage on the site for a certain period of time (the treatment group) and others to refrain from using the site (the control group) for the same period of time. A comparison of the treatment and control groups would then allow for an assessment of whether there is a connection between engagement in the social media site and mental health or behavioral outcomes.
Absent experimental control, even though there may be an association between two variables, such as Instagram use and mental health, the association may not be causal (that is say, Instagram and mental health would appear to be related but it might not be that use of Instagram causes young adult teens to feel depressed or more self-conscious about their looks). It could be that another factor is responsible for a perceived connection. For example, as Steinberg points out, familial conflict could both increase social media use and lead to depression, creating the impression that there is a relationship between social media use and mental health when familial conflict is the real driving force behind the observed association.
It could also be that causality runs in the opposite direction of the expected way. In relation to social media and mental health, it could be that those who are more depressed tend to use social media more, rather than the use of social media causing adolescents to feel depressed. Again, there would be an observed association between social media use and mental health, but not in the assumed causal way.
Recent concern over Facebook is similar in some ways to earlier “moral panics” such as those over television, certain types of music, and video games, among other things, as New York Times opinion columnist Farhad Manjoo has pointed out. A seemingly obvious culprit is fingered as having a disproportionate impact on producing or contributing to a societal ill. The history of such panics suggests, again, that caution is warranted in definitely concluding that social media use in general or use of Instagram in particular is responsible for certain mental health or behavioral outcomes.
This is not a defense of Facebook. It is certainly possible and perhaps even likely that Facebook has the negative effects alleged by Haugen. But the point remains that definitive conclusions should not be drawn from Facebook’s internal research alone.
Instead, Facebook’s investigation should be considered as one piece of evidence, based on a non-experimental research design, in a broader effort to sort out how social media affects our lives. When it comes to ourselves and our children, we should be wary of social media sites and cognizant of their potentially negative effects.