What is the Contagion Effect?
Suicide contagion is when exposure to modeled suicide behavior compels others to imitate that behavior. It falls under the umbrella of behavioral contagion which is the tendency of behaviors to spread through a group via modeling and imitation. According to behavioral contagion theory, people have a preexisting motivation to engage in a behavior, but this desire is offset by internal restraints to avoid the behavior. However, when this behavior is modeled by someone else, it lessens other peoples’ internal restraints and makes them more likely to engage in the behavior (Gould). With behaviors that are seen as taboo, the modeling of the behavior, especially by someone who is seen as a role model, makes the behavior seem more acceptable (withgoodreasonradio.org).
Social Learning Theory states that imitative learning is influenced by how much one relates with the model- gender, age, characteristics-, rewards the model receives for the behavior, and consequences the model faces. These three factors play a role in how strongly a person is impacted by a modeled suicide (Estrada). The degree to which some is influenced also depends on their preexisting motivations and tendencies. Speaking of media portrayals of suicide, Dr. Aguirre says, “for someone who has been thinking about suicide, it brings it to the forefront” (Gould). In this sense, it is important to note that media cannot cause a person to commit suicide, but it might push someone who is already standing on the precipice over the edge.
One of the first accounts of serious suicide contagion was the deaths of several young men in the UK following the publishing of a book called The Sorrows of Young Werther (1774). These men dressed and acted like the main character of the book, Werther, and took their lives in the same manner. After this cluster of suicides, the book was banned (Pirkis and Blood). In 1974, David Phillip coined the term the “Werther Effect” referring to copycat suicides triggered by a media portrayal of suicide or suicidal behavior (Devitt).
Today, suicide contagion from media is a prominent concern, especially among youth. Most research has found a significant influence of news reports on suicide and celebrity deaths on suicide rates. This has been witnessed many times throughout history with some of the most notable deaths being those of Marilyn Monroe, Kate Spade, Anthony Bourdain, and Robin Williams (Patel). However, the impact of fictional suicide stories, as well, has been a growing question- amplified recently with the release of Netflix’s 13 Reasons Why (Devitt, VanderWeele).
Nonfictional Media and Suicide Contagion
Many studies have found strong connections between suicide widely discussed in the news or other forms of media and increased suicide rates (hhs.org). Clear correlations have been found between celebrity deaths and an increase in suicide rates using methods of time series analysis (Devitt). For example, suicide rates increased by 12 percent directly after the death of Marilyn Monroe in 1962 (Stack). In the days following more recent celebrity suicides, more people looked up how to kill themselves (Mazurek and Ruckstuhl). Gould of Columbia University says that suicide contagion is the cause of at least 5 percent of youth suicides a year (Patel). Well media coverage of nonfictional suicides certainly cannot cause suicide, it clearly can negative influence people at risk, epically when discussed inappropriately.
The US Department of Health and Humans Services and the World Health Organization have been leaders in developing and implementing guidelines for reporting on suicide. It is suggested that reports be brief, explanations should not be oversimplified, and the victim should not be glorified. The reasoning for having brief reports in that prolonged exposure to suicide content increases the changes of contagion. Oversimplified explanations can over empathize acute negative events in a person’s life and neglect the complex factors that lead to suicide. Glorifying the victim not only gives the idea that suicide is okay, but that it also can elevate one’s status (hhs.org).
Nevertheless, results on the effects of media stories of suicide on contagion are still have some notable discrepancies. It is possible that these discrepancies are due to differences in the type of story presented. Overall, there are much weaker findings on the effect of fictional studies than nonfictional suicides. Furthermore, people who identify more with the person who committed suicide are more likely to imitate the behavior. Still, there is variation in the degree to which nonfictional stories impact viewers (Stack). Stack applied logistic regression techniques to 419 findings from 55 studies to try to explain these variations in terms of variations in social learning. He found stories about the death of a celebrity were 5.27 times more like to produce contagion effects, whereas, stories focusing on the negative effects of suicide were 99% less likely to cause copycat behavior. This positive effect of factual and non-glorifying suicide coverage has been noticed by other researchers as well, and it has been termed the Papageno effect (Gauld et al.). These stories might encourage people to ask for help. Suicide prevention hotlines have noted a significant increase in calls the days following recent celebrity suicides (Mazurek and Ruckstuhl). Additionally, Stack found that T.V. stories were 79% more likely to cause contagion than print stories, and stories about females were 4.89 times more likely to cause imitation than stories about males (Stack).
Overall research suggests that glamorized, drawn-out, oversimplified, and overly detailed stories on nonfictional suicides, increase suicide rates in individuals already at risk and who relate to the victim being discussed (hss.org, Gauld et. al., Stack). On the other hand, stories that discuss the negative impacts of suicide can enough people to ask for help and can make loved one more aware (Mazurek and Ruckstuhl, Stack). Similar correlations as these are more ambiguous between fictional suicide stories and suicide rates (Devitt).
Fictional Media and Suicide Contagion
Evidence regarding the relationship between fictional suicides and contagion is inconclusive (Devitt). Some findings suggest a significant impact on suicidal behavior, while others find no association. Gould reviewed 29 studies conducted world-wide since 1974 and found that 15 produced data supporting imitation of suicide. The remaining studies either had mixed results or no support for imitation. Importantly, the nature of suicidal behavior did not impact whether imitation was seen or not (Gould). However, as with nonfictional stories, identification with the character was a strong predictor for imitation (Gauld et. al.).
Pirkis and Blood systematically reviewed 11 ecological studies and found strong and consistent correlations between T.V. broadcasts featuring suicide and suicidal behavior rates (Gauld et. al.). Ostroff et al (1985) and Ostroff and Boyd (1897) found an increase in overdoses after the release of a film showing a teenage couple in a suicide pact. “The first two people to present after the film were an adolescent couple who had attempted suicide by overdose in a pact modeled on that in the film” (Pirkis and Blood). Gould and Shaffer (1986) observed the number of adolescents in New York hospitals for suicide attempts in the two weeks following the broadcasting of four different movies. Three of the movies depicted adolescent suicide and the fourth was an adult grappling with suicide. An increase in both patients and completed suicides was seen in the two weeks after the showing of the first three movies as compared to the two weeks prior. An increase in patients was seen in the two weeks after the fourth movie (Pirkis and Blood). A small-scale study in the UK in 1986 found increases in people presenting to emergency departments for overdoses in the days after the airing of an EastEnders episode where a character overdosed (Pirkis and Blood, Devitt). Hawton et al (1999) studied the number of patients presenting to emergency departments for overdoses and explicitly asked if they were influenced by a television show or movie. 20% reported yes. Another 17% reported having seen a show with a deliberate overdose, but that is did not influence their behavior (Gould). Biblarz et al (1991) examined the attitudes and perceptions of healthy college students randomly assigned to view content with teenage suicide, violence or neither. No conclusive findings were found (Gould). Another study examined the philological reactions in term of heart rate and breathing rate in groups of people who had attempted suicide, were thinking about suicide, and who were not suicidal. Heart rate and breathing rate increased the least in people who had attempted suicide. This suggests that such content influences people with different characteristics differently (Gould). A cross sectional survey of high school students by Martin (1996) looked at T.V. viewed by the students and suicide ideation and attempts exhibited. A correlation was found between viewing of content depicting suicide and suicidal behaviors only when the person also had real-life experience of suicide suggesting the real-life exposure has a much stronger negative impact than fictional exposure (Gould).
Some researchers have suggested that seasonal fluctuations are a confounding factor possibly leading to increases in suicide rate rather than the showing of a particular film or show. This possibility was illustrated by a 1993 UK study. There was an increase in young female suicide attempts after the showing of an episode of Casualty where a girl killed herself the first time it was shown in January. However, when is was repeated six months later, the increase was not seen. Girls who attempted suicide after the first showing were asked if they were impacted by the show and 4% said yes (Pirkis and Blood). Other studies on the same show were conducted in different locations, and results differed by location. It was suggested that any increase or decrease may have just been random fluctuations, but no concrete conclusions could be made (Pirkis and Blood). A German study found a significant increase in suicides of young males following two different showing of an episode that were separated by a year. Of these men, the one who survived their suicide attempts were asked if they were influenced by the show. Answers to this question varied (Pirkis and Blood).
More recent studies have continued to have conflicting results. Much the of the most current research has focused on Netflix’s show 13 Reasons Why. A meta-analysis by Ferguson (2018) failed to find evidence supporting suicide contagion resulting from fictional depictions including 13 Reasons Why. VanderWeele et. al. critic Ferguson’s work saying he did not consider setting-based differences in suicide rates. Well results were ambiguous, when looked at by setting, there were some locations (42%) where a significant (r > 0.10) positive correlation between fictional depictions and suicide rates was evident, and there were other locations (18%) where a significant negative correlation ( r < 0.10) was seen. They suggest that the influences of fictional portrayals of suicide might be contagion or productive based on the setting. Unlike Ferguson, Niederkrotenthaler et al (2019) found strong evidence supporting contagion from 13 Reasons Why. Controlling for both temporal trend and seasonal patterns, they found a 13% increase in adolescent suicides in the 3 months following the release of the show. No similar increase was seen in any other age group (VanderWeele et. al.).
Consistent with 18% of Ferguson’s setting-adjusted results, other studies have found positive effects of fictional suicide stories. Four studies reviewed by Pirkis and Blood found films that lead to increases in public knowledge and encouraged people to ask to help. They say that “appropriate portrayals, for example, those that emphasize negative consequences or alternative courses of action, could actually have a positive, educative effect.” This is consistent with the Papageno effect evident with some nonfictional stories.
Overall, results are inconsistent. Strength of causation is difficult to measure because of numerous confounding factors influencing suicide. Additionally, it is impossible to know if people who died were influenced by media content (Pirkis and Blood). There are also locational variations in correlations between the showing of fictional suicides and increase in suicide rates (VanderWeele). This might be due to differences in education, health resources, and socioeconomic status. The impact may also be influence by target audience and platforms such as with 13 Reasons Why. The show targeted adolescents which is a vulnerable group, and the availability on Netflix may make it easier for adolescents to watch away from parent supervision, It also allows for binging of the program which may make that material more overwhelming and give adolescents less of an opportunity to process and think critically about the content between showings. It is also possible that increases in suicide rates are inflated by second hand- contagion effects. Maybe a few individuals were influenced by the media content and then other people were influenced by the behavior of those individuals. It is difficult to know what primary effects are of seeing media content. Nevertheless, it is important to note that there is likely some influence going on that is impacting people’s health. Anything that can increase suicidal behavior should be taken seriously by health professional, researchers, and the public.
Devitt, P. (2017). 13 Reasons Why and Suicide Contagion. Scientific American. Retrieved from https://www.scientificamerican.com/article/13-reasons-why-and-suicide-contagion1.
Estrada, M. (2019). Psychology 101: Learning. [Lecture]. University of Rochester.
Gauld, C & Wathelet, M & Medjkane, F & Pauwels, N & Bougerol, T & Notredame, C. (2019). Construction and Validation of an Analytical Grid about Video Representations of Suicide (“MoVIES”). International Journal of Environmental Research and Public Health, 16. DOI: 10.3390/ijerph16152780.
Gould, M. (2006). Suicide and the Media. The New York Academy of Sciences. 200-224. Retrieved from http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.530.4169&rep=rep1&type=pdf.
Marzurek, M & Ruckstuhl, L. (2018). Mental health professional fear ‘contagion’ effect. Metro. Retrieved from https://www.mcleanhospital.org/media/mental-health-professionals-fear-contagion-effect.
Patel, Sima. (2018). Suicide contagion: What we know and what we don’t. abc News. Retrieved from https://abcnews.go.com/Health/suicide-contagion/story?id=55751220.
Pirkis, J & Blood, W. (2001). Suicide and the Media. Crisis, 22, 155-162.
Stack, S. (2005). Suicide in the Media: A Quantitative Review of Studies Based on Nonfictional Stories. The American Association of Suicidology Suicide and Life-Threatening Behavior 35, 121-133.
The Contagion Effect. (2017). With Good Reason Virginia Humanities. Retrieved from https://www.withgoodreasonradio.org/2017/07/the-contagion-effect.
VanderWeele, T & Mathur, M & Chen, Y. (2019). Media Portrayals and Public Health Implications for Suicide and Other Behaviors. JAMA Psychiatry, 76, 891-892.
What does “suicide contagion” mean, and what can be done to prevent it?. (2019). U.S Department of Health and Human Services. Retrieved from https://www.hhs.gov/answers/mental-health-and-substance-abuse/what-does-suicide-contagion-mean/index.html.