Boys Become Criminals by Talking About It First

Photo: Warner Bros. Pictures

So what would you do if your girlfriend got pregnant? Shoot her?”
“No, punch her in the stomach, real hard.”

This conversation occurred in an observation room at Oregon Social Learning Center. Tom Dishion and his colleagues were trying to learn more about why some kids become delinquent. He and many other behavioral scientists knew that most adolescents who get in trouble do so with other adolescents. Delinquency is a group enterprise. But Dishion took the research a step further. He wanted to see if he could actually observe the social influence processes that motivate kids to defy adult expectations and engage in criminal acts. So he asked young men who were participating in a longitudinal study of delinquency to bring a friend into the lab and have a series of brief conversations about things like planning an activity or solving a problem with a parent or friend.

What he found was startling. The conversations these young men had about deviant activities provided direct and strong reinforcement of deviant behavior. Even though these thirteen-and fourteen-year-old boys knew their conversations were being observed and recorded, some of them talked quite freely about committing crimes, getting drunk, taking drugs, and victimizing girls. Even more surprising was the fact that the amount of this kind of talk predicted whether individuals engaged in delinquent behavior well into adulthood.

You might think this occurred simply because adolescents who are already involved in problem behavior tend to talk more often about deviance. Maybe this topic was a by-product of their delinquent lifestyle and didn’t influence their delinquent behavior. But that was not the case. It is true that young people’s levels of problem behavior predict their future problem behavior. However, even when Dishion’s team controlled statistically for the influence of prior deviant behavior, the level of deviancy talk in these thirty-minute videotaped discussions predicted adult antisocial behavior two years later. The conversations escalated their deviant behavior.

Why would deviant talk lead to deviant behavior? And why did some kids talk about deviance while others didn’t? The answer Dishion reached is the most interesting thing about his research: If you want to understand why people do things, look for the reinforcers. Dishion and his colleagues coded not only the deviant talk of these kids, but also the reactions of their friends. He simply coded deviant and nondeviant talk and two possible reactions to each statement: pause or laugh. They found that the more laughs a boy got for what he said, the more he talked about that topic. In pairs where most of the laughs followed deviant talk, there was a great increase in the deviant talk. In statistical terms, 84 percent of the variance in deviant talk related to the rate of laughter for deviant talk. That is huge.

Even more interesting was the fact that the rate of reinforcement for deviant talk strongly predicted later delinquency. Boys whose friends approved of their talk of delinquent and violent acts were more likely to engage in these acts. Dishion called interactions like this “deviancy training.” In subsequent research he showed that simply letting at-risk kids get together raised the level of their misbehavior.

A colleague of Dishion’s, Deborah Capaldi, studies violence between men and women. She wondered if the deviancy training that Dishion discovered influenced how boys treat girls. When the boys in Dishion’s study were seventeen or eighteen years old, the researchers invited them back to have another conversation with a friend. One thing they asked them to talk about was what they liked and disliked about girls they knew. They coded these conversations in terms of how often they talked in hostile ways about girls. (The example at the beginning of this excerpt comes from one of those conversations.) When the young men were twenty to twenty-three years old, Capaldi got data from them and their girlfriends or wives about how often the men were physically violent. Sure enough, those who had talked approvingly about violence toward women years earlier were, in fact, more aggressive toward their partners.

A further analysis of the interactions of these boys when they were sixteen or seventeen revealed that the men who were most violent at twenty-two or twenty-three were the ones who not only had received reinforcement for talk of deviant behavior, but also had a friend or friends with whom they had coercive interactions. In addition to assessing the deviancy training I described above, this study coded how much the two friends were coercive toward each other. In this case, coercion was defined as engaging in dominant or dismissive behavior, using profanity, and being abusive to the other person. Even when the researchers controlled statistically for the teens’ antisocial behavior as children and the quality of parental discipline, those who talked about deviance the most and were coercive toward their friend were most likely to engage in violence as adults.

Despite all the evidence about how peers influence each other’s deviant behavior, our society routinely deals with delinquency by bringing troubled youth together. Tom Dishion found this out the hard way. He did a randomized trial of the impact of a parenting intervention (an early version of the Family Check-Up) versus a program to teach kids good self-management and study skills. He randomly assigned families of at-risk youth to either participate in the parenting intervention or not, and randomly assigned the youths to participate or not participate in the self-management program. He expected that the best outcomes would be for those who received both programs.

The parenting program worked as expected, but to Dishion’s surprise, the self-management program led to increases in youth smoking. When he coded videotapes of the interactions of youth in self-management training groups, he discovered that those who were talking about deviant activities were most likely to get others’ attention. This was exactly the process he showed in the research I described earlier: peer social approval reinforces deviant behavior.

Additional evidence has accumulated that raises serious questions about many of the things our society does in efforts to deal with at-risk or delinquent youth. We routinely put students who are doing less well academically on a different academic track than those who are doing well. In the process, we stigmatize them and bring them together, where they become friends and have little contact with students who embrace prosocial norms and behaviors. If at-risk kids have trouble in school, they are often transferred to alternative schools where the student body consists almost entirely of kids who experiment with drugs and other forms of risk taking. If they are arrested, they are locked up with other problem youth or treated in groups of adjudicated youth.

If left to their own devices, young people who are at risk for problems typically reinforce each other’s deviant tendencies. Thus, families and schools need to be sensitive to the importance of peer relationships in young people’s development. And while peer influences are especially important in adolescence, ensuring that adolescents have friends who support their positive social development ideally begins in early childhood, when children learn to regulate their emotions and cooperate with others.

From The Nurture Effect: How the Science of Human Behavior Can Improve Our Lives & Our World by Anthony Biglan, PhD. Reprinted with permission: New Harbinger Publications, Inc. Copyright © 2015 Anthony Biglan, PhD.

Boys Become Criminals by Talking About It First