How Social Media May Harm Boys and Girls Differently
For girls, social media may be more harmful.
Posted May 12, 2020 | Reviewed by Jessica Schrader
Professor Jean Twenge (author of iGen) recently joined Professor Jonathan Haidt (co-author of The Coddling of the American Mind) and colleagues to analyze the effects of social media on adolescent girls and boys.
They suggested that a good way to understand effect sizes is to compare effects across domains. In answer to the question "Does social media have harmful outcomes?" Twenge and Haidt answer, "Compared to what?"
Most people are not surprised to learn that heroin use in adolescence is associated with negative emotional well-being. The more an adolescent uses heroin, the more negative their well-being, on average, as measured by the statistical parameter r, the correlation coefficient.
How does social media compare to heroin use? Gender matters—a lot. The correlation coefficient r relating social media use to negative emotional well-being is substantially larger for social media use than for heroin use in adolescent girls. For adolescent boys, the correlation for heroin use is substantially greater than the one for social media use.
Disclaimer #1: We are talking about emotional well-being, not necessarily life and death. Heroin use is associated with a risk of accidental overdose resulting in death. Social media overdose does not result in death.
Disclaimer #2: Other research by Twenge and her colleague, Keith Campbell, demonstrates that social media is generally not harmful when usage is less than roughly one hour per day. The association between social media and adverse effects begins with usage beyond one hour per day. However, there does not appear to be any safe lower limit on heroin usage. Adolescents should avoid heroin use completely.
Any the notion that "social media is worse than heroin for adolescent girls" should be refined: "In this analysis, use of social media, for adolescent girls, is correlated with adverse outcomes such as anxiety and depression more strongly than heroin use is." And correlation in itself does not prove that one factor causes the other. Even with the disclaimers, that's still a surprising finding.
Twenge, Haidt, and colleagues do not offer any suggestions in this paper for the apparent gender difference. I addressed this question in an essay for The New York Times. The second edition of my book Girls on the Edge, which comes out in August, includes a chapter devoted to answering this question. Briefly, here are some potential explanations:
- Girls are more invested in social media than boys are, on average. Girls post many more photos on social media, and the photos are different. While the boys' photos typically focus on something the boy is doing or something the boy saw, the girl is more likely to post a photo where the emphasis is on how the girl looks. If you don't like Jake's photo of the pretty cheerleader at the football game, Jake doesn't care. But if you don't like Emily's selfie at the game, Emily is likely to take that more personally.
- Boys post a wider range of their lived experience; girls' postings are (on average) more curated and more polished. A boy and a girl both get sick. They both throw up. The boy posts a photo of his own vomit on Instagram. It would be very rare for a girl to do that. Another boy, looking at Brett's vomit, is unlikely to want to be Brett. But girls, looking at other girls' social media, may envy the fun/amazing/funny things the other girls are doing.
- Boys often overestimate how interesting their own lives are to other people. Girls, on the other hand, are more ready to believe that other girls are having more fun than they are.
Put these effects together and you can see why the average girl may be more vulnerable than the average boy to potential negative effects of social media. (The average boy is much more vulnerable than the average girl to the addictive properties of video games, but that's a different topic.)
Jean M. Twenge, Jonathan Haidt, and colleagues, "Underestimating digital media harm," Nature Human Behaviour, 4:346-348, April 2020,