Suicide ideation among adolescents appears widespread in certain Micronesian communities and is popularly expressed in recent songs composed locally and aired on Micronesian radio stations, and in graffiti adorning T-shirts and high school walls. A number of young boys who attempted suicide reported that they first saw or heard about it when they were 8 or 10 years old. Their suicide attempts appear in the spirit of imitative or experimental play. One 11-year-old boy, for example, hanged himself inside his house and when found he was already unconscious and his tongue protruding. He later explained that he wanted to "try" out hanging. He said that he did not want to die, although he knew he was risking death. Such cases of imitative suicide attempts by boys as young as five and six have been reported recently from Truk. Several cases of young adolescent suicide deaths recently in Micronesia were evidently the outcome of such experiments. Thus as suicide grows more frequent in these communities the idea itself acquires a certain familiarity if not fascination to young men, and the lethality of the act seems to be trivialized. Especially among some younger boys, the suicide acts appear to have acquired an experimental almost recreational element.

There is something very chilling about this passage. Suicide isn't supposed to be trivialized like this. But the truly chilling thing about it is how familiar it all seems. Here we have a contagious epidemic of self-destruction, engaged in by youth in the spirit of experimentation, imitation, and rebellion. Here we have a mindless action that somehow, among teenagers, has become an important form of self-expression. In a strange way, the Micronesian teen suicide epidemic sounds an awful lot like the epidemic of teenage smoking in the West.

Teenage smoking is one of the great, baffling phenomena of modern life. No one really knows how to fight it, or even, for that matter, what it is. The principal assumption of the anti-smoking movement has been that tobacco companies persuade teens to smoke by lying to them, by making smoking sound a lot more desirable and a lot less harmful than it really is. To address that problem, then, we've restricted and policed cigarette advertising, so it's a lot harder for tobacco companies to lie. We've raised the price of cigarettes and enforced the law against selling tobacco to minors, to try to make it much harder for teens to buy cigarettes. And we've run extensive public health campaigns on television and radio and in magazines to try to educate teens about the dangers of smoking.

It has become fairly obvious, however, that this approach isn't very effective. Why do we think, for example, that the key to fighting smoking is educating people about the risks of cigarettes? Harvard University economist W. Kip Viscusi recently asked a group of smokers to guess how many years of life, on average, smoking from the age of twenty-one onward would cost them. They guessed nine years. The real answer is somewhere around six or seven. Smokers aren't smokers because they underestimate the risks of smoking. They smoke even though they overestimate the risk of smoking. At the same time, it is not clear how effective it is to have adults tell teenagers that they shouldn't smoke. As any parent of a teenage child will tell you, the essential contrariness of adolescents suggests that the more adults inveigh against smoking and lecture teenagers about its dangers, the more teens, paradoxically, will want to try it. Sure enough, if you look at smoking trends over the past decade or so, that is exactly what has happened. The anti-smoking movement has never been louder or more prominent. Yet all signs suggest that among the young the anti-smoking message is backfiring. Between 1993 and 1997, the number of college students who smoke jumped from 22.3 percent to 28.5 percent. Between 1991 and 1997, the number of high school students who smoke jumped 32 percent. Since 1988, in fact, the total number of teen smokers in the United States has risen an extraordinary 73 percent. There are few public health programs in recent years that have fallen as short of their mission as the war on smoking.

The lesson here is not that we should give up trying to fight cigarettes. The point is simply that the way we have tended to think about the causes of smoking doesn't make a lot of sense. That's why the epidemic of suicide in Micronesia is so interesting and potentially relevant to the smoking problem. It gives us another way of trying to come to terms with youth smoking. What if smoking, instead of following the rational principles of the marketplace, follows the same kind of mysterious and complex social rules and rituals that govern teen suicide? If smoking really is an epidemic like Micronesian suicide, how does that change the way we ought to fight the problem?

The central observation of those who study suicide is that, in some places and under some circumstances, the act of one person taking his or her own life can be contagious. Suicides lead to suicides. The pioneer in this field is David Phillips, a sociologist at the University of California at San Diego, who has conducted a number of studies on suicide, each more fascinating and seemingly improbable than the last. He began by making a list of all the stories about suicide that ran on the front page of the country's most prominent newspapers in the twenty-year stretch between the end of the 1940s and the end of the 1960s. Then he matched them up with suicide statistics from the same period. He wanted to know whether there was any relationship between the two. Sure enough, there was. Immediately after stories about suicides appeared, suicides in the area served by the newspaper jumped. In the case of national stories, the rate jumped nationally. (Marilyn Monroe's death was followed by a temporary 12 percent increase in the national suicide rate.) Then Phillips repeated his experiment with traffic accidents. He took front-page suicide stories from the Los Angeles Times and the SanFrancisco Chronicle and matched them up with traffic fatalities from the state of California. He found the same pattern. On the day after a highly publicized suicide, the number of fatalities from traffic accidents was, on average, 5.9 percent higher than expected. Two days after a suicide story, traffic deaths rose 4.1 percent. Three days after, they rose 3.1 percent, and four days after, they rose 8.1 percent. (After ten days, the traffic fatality rate was hack to normal.) Phillips concluded that one of the ways in which people commit suicide is by deliberately crashing their cars, and that these people were just as susceptible to the contagious effects of a highly publicized suicide as were people killing themselves by more conventional means.

The kind of contagion Phillips is talking about isn't something rational or even necessarily conscious. It's not like a persuasive argument. It's something much more subtle than that. "When I'm waiting at a traffic light and the light is red, sometimes I wonder whether I should cross and jaywalk," he says. "Then somebody else does it and so I do too. It's a kind of imitation. I'm getting permission to act from someone else who is engaging in a deviant act. Is that a conscious decision? I can't tell. Maybe afterwards I could brood on the difference. But at the time I don't know whether any of us knows how much of our decision is conscious and how much is unconscious. Human decisions are subtle and complicated and not very well understood." In the case of suicide, Phillips argues, the decision by someone famous to take his or her own life has the same effect: it gives other people, particularly those vulnerable to suggestion because of immaturity or mental illness, permission to engage in a deviant act as well. "Suicide stories are a kind of natural advertisement for a particular response to your problems," Phillips continues. "You've got all these people who are unhappy and have difficulty making up their minds because they are depressed. They are living with this pain. There are lots of stories advertising different kinds of responses to that. It could be that Billy Graham has a crusade going on that weekend — that's a religious response. Or it could be that somebody is advertising an escapist movie — that's another response. Suicide stories offer another kind of alternative." Phillips's permission-givers are the functional equivalent of the Salesmen I talked about in chapter 2. Just as Tom Gau could, through the persuasive force of his personality, serve as a Tipping Point in a word-of-mouth epidemic, the people who die in highly publicized suicides — whose deaths give others "permission" to die — serve as the Tipping Points in suicide epidemics.