On August 4, 2021, Senators Richard Blumenthal, a Connecticut Democrat, and Marsha Blackburn, a Tennessee Republican, wrote to Mark Zuckerberg asking whether Facebook's research had ever found that its platforms and products can have a negative effect on children's and teens' mental health or well-being, and whether the company would release it. That March, at a House hearing, Zuckerberg had said he did not think the research on screen time and children's mental health was conclusive, and a moment later that the research that we have seen is that using social apps to connect with other people can have positive mental health benefits. Facebook answered the senators with a six-page letter that did not include its own studies. On September 14 The Wall Street Journal began quoting them.
The line that traveled furthest was a slide headline from a presentation Facebook's researchers shared internally in November 2019, called Hard Life Moments. It said, in capitals, that we make body image issues worse for 1 in 3 teen girls. The company's own researchers wrote it, and the Journal quoted it. The subtitle beneath it, which traveled less, said that teens also generally thought that IG made things better or had no impact.
The survey behind the slide asked 22,410 Instagram users in the United States, Japan, Brazil, Indonesia, Turkey and India, aged 13 to over 65, whether they had been through any of six hard experiences, drawn at random from a longer list, in the past 30 days. Those who had were asked, about one of them, what effect Instagram had had. Among the teen girls who reported trouble with body image and got that question, 32.4 percent said Instagram made it worse, 45.5 percent said it made no difference and 22.1 percent said it made it better. Facebook later said that fewer than 150 teen girls, across all six countries, answered it. On the same chart, for 11 of the 12 problems listed, more of the girls who had them said Instagram helped than said it hurt. Body image was the one exception.
The Journal's article did say that the slide described girls who already had the problem, and it reported that the researchers called the causality of some findings unclear and some samples small. The headline above the article said Facebook Knows Instagram Is Toxic for Teen Girls.
On September 29, the evening before its global head of safety was due to testify, Facebook published that presentation and a second one, with notes of its own on every slide. The notes said many samples were small, the questions asked users to judge cause and effect for themselves, and the research could not establish causation. Of its own researchers' headline, the company wrote that in this context, the causal language in the title is myopic. The second presentation, an October 2019 study of teens in the United States and Britain, carried a chart that showed 6 percent for the United States and 13 percent for Britain on its line for teens who said a wish to kill themselves had started on Instagram. The Journal had described those bars as shares of teens who reported suicidal thoughts. Facebook's note recalculated the same answers against every teen in the survey and got about 1 percent, or 16 respondents. What the 6 and 13 are percentages of is the point in dispute, and the chart does not settle it.
At the hearing on September 30, Antigone Davis told senators that this research is not a bombshell. It is not causal research. Blumenthal answered that it was a bombshell: powerful, gripping, riveting evidence that Facebook knows of the harmful effects of its site on children and that it has concealed those facts and findings. When Senator Amy Klobuchar of Minnesota asked whether Davis had known of the finding that Instagram worsens body image issues for one in three teenage girls, Davis tried to put the missing group back: the findings were actually of teen girls who already expressed having that issue. Five days later, questioning the whistleblower Frances Haugen, Senator Ted Cruz of Texas cited the suicide figures as 13 percent of British users and 6 percent of American users.
The short version kept its place. On January 31, 2024, at a Senate Judiciary hearing on child sexual exploitation, Senator Josh Hawley of Missouri read Zuckerberg the 1-in-3 headline, then a second slide in the shortened wording the Journal had printed, about teens blaming Instagram for increases in the rate of anxiety and depression. The slide itself, from the October 2019 study, says rates and ends with among teens. It rests on focus groups with 40 teenagers, recruited around themes of body image, low mood and loneliness, with a baseline group besides.
Whether Instagram harms teenagers at all, the slides could not say, and the wider science has not settled it. In May 2023 the U.S. Surgeon General said there was not yet enough evidence to determine if social media is sufficiently safe for children and adolescents. That December a committee of the National Academies of Sciences, Engineering, and Medicine found that the published research did not support the conclusion that social media causes changes in adolescent health at the population level. A survey of users' own impressions could not answer the question either way, and its authors had written causal headlines over it regardless. The company holding the survey had kept it inside and did not send it when senators asked, so its first public reading came from copies taken out of the company, without the context only Facebook could have supplied.
A percentage is a claim about a group. It is true only of the people it counted.