After its release in 2017, the Netflix series “13 Reasons Why” spurred controversy over concerns that its portrayal of a teenage girl’s suicide could increase suicide contagion among adolescents.
Though a much-publicized 2019 study found a contagion effect among boys, a subsequent reanalysis of that data by the Annenberg Public Policy Center (APPC) of the University of Pennsylvania concluded that, to the contrary, the series had no clear effect on teen suicide.
Now, in a pair of commentaries published in PLOS ONE, the original authors challenged the APPC reanalysis and APPC research director Daniel Romer defended his critique.
“We stand by our reanalysis,” Romer said. “There is no reason or evidence to suggest that the show had an effect before it was even released. And as the authors of the study acknowledged, one would expect the show to have a strong effect for female adolescents, which was not found.”
The debate over ’13 Reasons Why’
In their 2019 paper, Bridge et al. claimed to find an increase in suicide in 10- to 17-year-old boys over as long as a 10-month period, starting the month before Netflix released the series. But an APPC reanalysis of that data, published early in 2020, failed to detect any reliable increase in suicide in girls and an increase for boys one month before and one month after the release in April 2017. Another study using the same methodology also found an increase for males in March and April and no significant effect on females in April, consistent with APPC’s findings.
In their new PLOS commentary, Bridge et al. responded that Netflix “was actively broadcasting advertisements and series’ trailers” in March 2017 “that targeted youth and encouraged them to watch this dramatization of an adolescent girl’s suicide.” But Romer finds “considerable evidence” that the show “did not create concerns about contagion until April,” citing other independent analyses that focused on April as the point at which Google searches and crisis-line discussions began to rise. The study of crisis lines, for instance, “found no change in trend the month before the release and a sharp decrease” shortly after the release of the series.
“Thus,” Romer argued, “there was no evidence that the series produced anywhere near the attention that would have been required to produce contagion in the month prior to its release and, if anything, the series coincided with a decline in crisis conversations that followed its release.”
Romer said that if one were to predict a contagion effect from the series, it would be for young females. In his reanalysis, Romer found a modest but statistically unreliable increase in suicide in April among girls that was unique to that month.
“Unfortunately, looking at aggregate monthly suicide rates is not a very sensitive method for detecting either the harmful or helpful effects of media depictions of suicide,” Romer said. In a separate study, Romer and colleagues found that viewing the second season of “13 Reasons Why” may have had beneficial effects on some young viewers and harmful effects on other viewers. These opposing effects make it difficult to determine whether the potentially harmful effect for some female adolescents was counterbalanced by beneficial effects for others, he said.
In his original reanalysis, Romer said the Bridge study failed to account for ongoing trends in adolescent suicide, in particular a strong rise in 2017. In their new commentary, Bridge et al. defended the use of their analytic model, but Romer responded that their model “seriously underestimated the upward trend in overall suicide …”
Understanding media effects
Romer said it is important to gain a better understanding of how shows like “13 Reasons Why” affect vulnerable audiences so that the television producers can develop entertaining and helpful programming without creating adverse effects on viewers.
In his current commentary, Romer concluded, “In view of our still limited knowledge about how these events affect vulnerable audiences, we should resist drawing bold conclusions about effects that defied predictions about both the gender of the victims and the time when the effect should appear.”