For good or ill, behaviors among humans pass not only by genes, but by language. A judgmental way to put this is that behaviors and ideas are "contagious." Pathological homesickness, apotemnophilia, multiple personality disorder, and even Ursuline convents in seventeenth-century France have been posited to arise from contagion.
In particular, suicide is widely accepted as a contagious behavior. The posited contagion even has a name - the "Werther effect." Belief in the media contagion of suicide is so strong and pervasive that "media guidelines" - a form of voluntary censorship - are widely observed in reporting on suicide.
There is evidence that many behaviors other than suicide are similarly "contagious," however. Violence against others, in particular, is well-studied in its relation to media contagion. The harm of violence, especially homicidal violence like murder and murder-suicide, is much greater than that of suicide. And the evidence in favor of "violence contagion" is stronger than that of suicide contagion. Why, then, are reports of suicide voluntarily censored, while reports of violence are not?
Suicide Contagion: The Evidence
The evidence for suicide contagion through the media is almost entirely ecological. The studies that provide the basis for the phenomenon of suicide contagion are somewhat questionable. Many suffer from lack of control for important variables; those that are controlled suffer from problems with the control groups or small sample size. Some ecological studies have indicated that the suicide contagion phenomenon is real; others have contradicted those findings. Even among studies that find a correlation between suicide rate and media reports of suicides, the correlation is often much weaker than the correlation of the suicide rate with other factors, such as the unemployment rate.
Of course, there is a more abstract critique of ecological-level data in general. A criticism of David Phillips' ecological data on suicide contagion and fatal aircraft incidents goes as follows:
Phillips asserts that the statistically significant increase in aircraft fatalities can be explained due to suicide, as well as a "consciously or unconsciously" induced motive on the part of the pilot to also murder some person or persons. What Phillips does, in effect, is impute suicidal motives to some deceased persons on the basis of the statistically significant increases in accidents. Such a jump is conceptually unwise because it is based on a tautology: the statistical increase is the basis for defining some cases as suicide, but these cases are also used to explain the increase. [Citations removed.]
At any rate, a major problem with suicide contagion research is a lack of empirical evidence at an individual level. The one case-controlled study that I am aware of fails to demonstrate any link between hearing media reports of suicide and making a suicide attempt - and, in fact, demonstrates that hearing a media report of suicide has a significant protective effect against suicide attempts.
The study authors interviewed 153 people, ages 13-34, who were "victims" of nearly lethal suicide attempts and who had been treated at local emergency rooms in the Houston, Texas, area. A control group of 513 subjects was similarly interviewed. The conclusion? Not only did the study fail to demonstrate any sort of "suicide contagion," but, as mentioned above, the authors note a statistically significant protective effect when a subject heard a news report of suicide within 30 days prior to the suicide attempt or had a friend or acquaintance make a suicide attempt. That is, the ER suicide-attempt group was actually less likely than the control group to be aware of a recent media report of a suicide, or to have experienced the suicidal behavior of an acquaintance! The suicide attempt of a parent or relative had no statistically significant effect on suicidal behavior, whereas the usual "suicide contagion" sources had a statistically significant protective effect - the opposite of what the suicide contagion model predicts.
The evidence for violence contagion is much stronger than that for suicide contagion. But whereas suicide censorship is widely accepted, censorship of other-directed violence in media stories is rare.
Violence contagion is demonstrated by the same type of ecological study as suicide contagion. In addition, unlike the suicide case, there is a body of laboratory evidence suggesting that exposure to violent stimuli increases aggressive behavior. However, despite both sources of evidence, the theory that media reports of violence "cause" real-life violence is not at all universally accepted. And the idea that the media should voluntarily self-censor with regard to reports of violence is much less widely accepted than self-censorship of reports of suicide, despite greater evidence for a causal link in the former case.
Contagion and Moral Responsibility
I believe that the insistence that suicide is media-contagious, but violence is not, is not rational, but is a consequence of the differential attributions of moral responsibility in cases of suicide versus other-directed violence. Suicide is seen as an irrational act; the actor, as the story goes, is not in control of himself, certainly not sane, and is therefore vulnerable to external effects.
On the other hand, the idea that violent acts like homicides are attributable to media suggestion is generally seen as a pathetic excuse. Perpetrators of violence are perceived as much more morally responsible for their acts than suicides; despite evidence to the contrary, idea contagion is psychologically ruled out as a cause of violence, but not of suicides (though there are exceptions to this line of thinking).
Is political corruption contagious? Adultery? Prostitution? Drug abuse? Such questions are rarely even studied. Obesity certainly appears to be contagious. If so, should we censor reports of these topics to avoid a contagion effect? To do so would seem ludicrous and counter-productive, not to mention contrary to our political ideals. But the censorship of suicide goes unchallenged.
Moral Responsibility and Willingness to Censor
The more an actor is seen as the agent of his actions, the less outside influences are seen as affecting his actions. Therefore, in cases where moral responsibility is strongly attributed to an actor, outside influences are unlikely to be taken seriously as a cause of his actions - and, therefore, it is not necessary to censor these "outside influences" (such as media reports).
It is my belief that the widespread voluntary censorship of reports of suicide - from use of politically correct language to pervasive norms of message content - are the result of the modern trend to exculpate suicides from moral responsibility and redefine suicide as an act of insanity. There is, however, little evidence that suicides are any less morally responsible for their actions than murderers. Certainly, many other behaviors are media-contagious - but they are not censored, nor are many of them even studied.
I think that one possible explanation is that, at a deep level, people understand that suicide is just not that bad compared to actual acts of violence - despite hysterical language describing suicide as "self-murder." We want to exculpate people from acts to which we are sympathetic. While we often refuse to define acts outside of societal norms as "not wrong," we may nonetheless refuse to attribute full moral responsibility to these acts. However, this sort of sympathy backfires in our society. People who are "not responsible for their actions" must be "protected," often in painful and dehumanizing ways; and society is responsible for their "protection," often to the detriment of freedom.
Think of the children.
On a largely unrelated note, could this be the stupidest news story about suicide of all time?
1. Altheide, David. "Airplane Accidents, Murder, and the Mass Media: Comment on Phillips." Social Forces 2:593-596 (Special Issue, 1981).
2. Bollen, Kenneth, and David Phillips. "Imitative Suicides: A National Study of the Effects of Television News Stories." American Sociological Review 47:802-09 (1982).
3. Coalition of Law Abiding Sporting Shooters. "Ideas Kill: Science Shines a Light on Port Arthur Deaths." Retrieved from http://www.class.org.au/ideas-kill.htm on 04/06/2009.
4. Elliot, Carl. "A new way to be mad." The Atlantic, December 2000.
5. Gunter, Barrie. "Media Violence: Is There a Case for Causality?" American Behavioral Scientist 51:1061 (2008).
6. Jones, Marshall, and Elizabeth Rapley. "Behavioral Contagion and the Rise of Convent Education in France." Journal of Interdisciplinary History 31.4:489-521 (2001).
7. Mercy, James, Marcie-jo Kresnow, Patrick W. O’Carroll, Roberta K. Lee, Kenneth E. Powell, Lloyd B. Potter, Alan C. Swann, Ralph F. Frankowski, and Timothy L. Bayer. "Is Suicide Contagious? A Study of the Relation between Exposure to the Suicidal Behavior of Others and Nearly Lethal Suicide Attempts." (American Journal of Epidemiology 154:2 (2001).
8. Phillips, David. "The Impact of Mass Media Violence on U.S. Homicides." American Sociological Review 48:4:560-568 (1983).
9. Stack, Steven. "Divorce, Suicide, and the Mass Media: An Analysis of Differential Identification, 1948-1980." Journal of Marriage and the Family 2:553-560 (1990).