Since its release on 31 March 2017, the television show 13 Reasons Why, which features the graphic suicide of a teenage girl, has been the source of controversy as to whether it increased real-life teen suicides. The show has artistic merit and is well acted. However, the subject matter, based on a novel of the same name, is likely to spark anxiety. In 2017, the National Association of School Psychologists released a statement warning of the potential dangers of the film.
People often argue that viewers—particularly young viewers—are likely to imitate the behaviors they see depicted in fiction. This has led to controversy over whether violent video games promote violence in real life; whether thin actresses promote eating disorders; and whether smoking in movies can be blamed for real-life lung cancer deaths. Evidence increasingly suggests that fictional media has little detrimental impact on viewers. But what about a suicide-themed show? No one thinks watching 13 Reasons Why would cause a healthy teen to kill herself, but what about vulnerable kids, who may already be considering suicide?
It’s a tricky question. Like many emotive issues, the idea of suicide contagion through fiction—sometimes called the Werther effect after Johann Wolfgang von Goethe’s 1774 novel, The Sorrows of Young Werther—is often taken as self-evident. But evidence in its favor has always been shaky and oversold. Through a meta-analysis of studies exploring suicide-themed media, including heavy metal music, I found considerable differences between study outcomes. Overall, the evidence does not support the contention that suicide-themed media causes suicides in real life.
The evidence against 13 Reasons Why has also been mixed. Some studies have suggested that watching the show could have positive or nuanced effects, such as fostering more dialogue between parents and youth, or might reduce suicidal ideation and increase empathy in those who watched the entire show, but have inverse effects on those who stopped watching before the end. But, last summer, two separate studies, analyzing the same national Centers for Disease Control data, concluded that 13 Reasons Why was associated with a rise in teen suicides.
However, as I have explained elsewhere, the results of the first study were a mess. No effects were found for girls (the demographic most likely to watch the show and most closely comparable to the main character), and results for boys were inconsistent. The suicide rate among boys was already high before the show. The authors attribute this to advance publicity for the show—a rationalization to attempt to force the inconvenient data to confirm their hypothesis. Most months of 2017 didn’t show a rise in suicides; a few did; some showed a decline. Overall, this pattern suggests that 13 Reasons Why didn’t contribute to teen suicides. Boys were more likely to have been influenced, if at all, by the suicides of several high-profile male performers during the period. A second study, using the same data, provided remarkably different results—in itself, this is a clear red flag. This second study found a consistent increase in suicides among both boys and girls.
Neither study effectively controlled for increasing rates of suicide overall. Suicides have been increasing for years, among all age groups. In fact, both the highest suicide rate, and the largest increase in suicides involve middle-aged adults—not exactly the demographic most likely to view 13 Reasons Why. The 13 Reasons Why hypothesis considers teen suicides outside the context of increasing suicide rates among the entire population. The two time-series studies of 13 Reasons Why failed to adequately control for year-to-year increases in suicide through 2017, nor did they acknowledge that teen suicides always tend to spike in April, the month the series was released. It’s likely that 13 Reasons Why was released during what is always a bad month for teen suicides. Had the show been released during a month with lower suicide rates, we might instead be debating (just as spuriously) whether the show decreased the number of suicides.
A new, third study, using exactly the same database as the previous two, has provided a third set of results. This study attempted to control for yearly and seasonal patterns in teen suicide and found that the release of 13 Reasons Why had no statistically significant effect on suicides among either teen boys or girls. This suggests that both the first two studies were flawed. The author of the third study has suggested that some vulnerable young people may have been negatively affected by the show—although some may have experienced positive benefits. This is a reasonable hypothesis, but lacks data to support it. I suspect it conflicts with the study’s own outcomes.
These three studies looked at exactly the same data and came up with three entirely different sets of results. As social scientists, this should give us pause. Each person can now pick the study she prefers, which is not terribly scientific. This fits a pattern I observed during my meta-analysis of previous studies. Scholars can analyze the same data in different ways and produce diametrically opposed results—not just different conclusions or interpretations, but different results.
Population time-series studies can be illuminating, although they are by nature correlational, not causal. For instance, I suspect we’d see a significant correlation between Beyoncé’s yearly income and the global temperature of the Earth, but we wouldn’t say that Beyoncé is making the world hotter, at least not literally. Single-event time series studies, such as studies involving the release of a show, an event that only occurs once, are even trickier. In real life, coincidences abound. Nicolas Cage movies are correlated with swimming pool deaths in the US. But Nicolas Cage movies do not actually cause drownings—even though I might choose death over watching some of his movies. These studies all reflect the tendency of social scientists to leap at questionable data, particularly when it helps them appear virtuous in a moral cause.
In the late 1980s and early 1990s, heavy metal artists Ozzy Osbourne and Judas Priest were each sued for causing suicide. Neither lawsuit was successful, but these concerns reflected both popular folk wisdom and the contemporary scholarly consensus on the supposed evils of heavy metal. More recent scholarship suggests that, if anything, heavy metal music has a positive impact on fans. Few people today think that Osbourne or Judas Priest are dangerous influences. But we’re applying the same narratives to 13 Reasons Why.
Unfortunately, Netflix caved to pressure, after the first two studies came out. They removed the graphic suicide theme from the first season—long after most people had presumably already seen it. As I noted at the time, this did little to appease critics. It just caused Netflix to lose credibility as defenders of free speech. Had they been patient and waited for reanalysis of those problematic first two studies, they might have been more reluctant to engage in self-censorship.
Suicide is complex. For the most part, its origins lie in biological and genetic predispositions to depression, stressful life environments and substance abuse and the absence of coping mechanisms and social support. To date, there is no firm evidence linking fictional media to increases in suicidal behavior. These population studies give us no reason to believe that calculus has changed. It’s possible that the show might have had idiosyncratic effects. That is, the show might have saved some young people, by decreasing their suicidal ideation, while increasing suicidal ideation in others. Idiosyncratic effects can be difficult to tease out of social science data. But claims that the show is detrimental to young people in the aggregate haven’t been confirmed by data and groups such as the National Association of School Psychologists should be more cautious in their future public statements.
Providing suicide hotlines and information about treatment and prevention options to accompany a show like 13 Reasons Why is a great idea. But arguing for censorship or exaggerating the show’s dangers can distract us from more pressing but less sensationalist issues that do cause suicide. Adult pearl clutching may even attract more teens to the show. In general, we need to be more cautious as to how we communicate the effects of fictional media on audiences. These effects tend to be weaker and more nuanced than many believe. That is as true of 13 Reasons Why as it has proven to be of video games, books, movies and heavy metal music.
Image by Nicole Alexander – Vimeo: Capsule MTV Movie Awards June 16 2018