Todd Walker had to do something. In his fifteen years as a youth football coach in Oakland, California, he watched kid after kid get into trouble. Many ended up on the streets and became victims of violence. “I was losing too many kids to homicide,” Walker lamented. “Instead of going to graduations, I was going to funerals.” As Walker knew, violence strikes the young disproportionately. In 2006, homicide was the second leading cause of death among youths ages ten through twenty-four, second only to accidents. Among African American youth, homicide was the number one killer. Often, the perpetrators of these violent acts are other teens. The number of violent crimes committed by juveniles has been falling since 1995, but in recent years it has begun to creep up again.1
Walker wondered how he could get through to these kids. What would convince them to stay off the streets and away from crime and violence? Maybe, he thought, the answer is to scare the living daylights out of them. Walker founded a program called Restoring Inner City Peace (R.I.P.) that shows teens, very graphically, where they might end up if they start selling drugs and hanging out with the wrong people. He takes groups of young people to local cemeteries and funeral homes and cajoles them into touching a gurney used to roll bodies into the mortuary. “When you’re dead, we strap you on here, stick a tag on your toe,” he tells them. “Then we put you in the freezer.” The kids hear from embalmers who explain how they prepare bullet-ridden bodies for funerals and from former convicts who talk about life in prison. Is the message sinking in? Tyris Bamba, a fifteen-year-old participant in the program, thought so. “It was a reminder of where I don’t wanna be,” Tyris said. “Now I think I’m more conscious of the friends I hang out with, and the places that people want to go to—where you know bad things are gonna happen. I just stay away.”2
Walker is one of those admirable people who act on their good intentions, doing something about the problems they see instead of simply complaining about them. He even went so far as to take a second job delivering bodies to a local funeral home so that he could take his kids there and show them the gurneys and coffins. But there’s just one problem with his “scared-straight” approach: it doesn’t work. In fact, it probably does more harm than good, increasing the likelihood that Walker’s kids will commit crimes in the future.
The term “scared straight” was popularized in 1978 by an award-winning documentary of the same name that depicted a program run by a group of inmates at Rahway State Prison in New Jersey. The inmates, all of whom were serving sentences of at least twenty-five years, wanted to do something good for the community and hit upon the idea of meeting with at-risk kids and showing them what prison was really like, with the idea that this “shock experience” would scare the kids away from crime. With the cooperation of the prison superintendent, the local police chief, and a juvenile court judge, the Juvenile Awareness Project was born, in which small groups of teens, typically between the ages of twelve and eighteen, visit the prison and meet with the “lifers.” In a typical session, they are subjected to a verbal harangue from the inmates, including graphic descriptions of rape and murder within the prison walls. After the documentary was aired on television, more than thirty similar programs were founded in and outside of the United States.
In 1999, another television documentary was aired, called Scared Straight! 20 Years Later, which further extolled the virtues of the intervention. The kids who had taken part in the original visit to Rahway State Prison, now in their thirties, were interviewed, and most said that the program had helped them. Arnold Shapiro, the producer of both the original and the new documentary, was not surprised. Since the original film was shown, he noted, “I can’t tell you how many people have come up to me and said, ‘I was a juvenile delinquent, and when I saw this, I stopped, I changed.’ ”3
In chapter 2, in which I talked about how to evaluate interventions, I suggested that we institute a don’t ask, can’t tell policy, in which we stop relying on testimonials about whether a program works. People often don’t know how much a program has helped them; after all, they don’t know how well-off they would be if they hadn’t participated in it. Instead, we need to conduct rigorous tests of social programs such as Scared Straight. The only way we can tell what effect they have is by randomly assigning kids to a group that goes through the program or to a control group that does not. Several studies have done just that, and the results are remarkable in both their consistency and direction: not only do scared-straight programs fail to reduce the likelihood that kids will commit crimes, they actually increase criminal activity. A review of seven experimental tests that measured how likely participants and nonparticipants were to commit crimes, in time periods ranging from three to fifteen months after a scared-straight intervention, found that the kids who attended the interventions were more likely to commit crimes than were kids in the control groups in every single study. The increase in criminal activity among the scared-straight kids ranged from 1 percent to 30 percent, with an average of 13 percent.
The program at Rahway State Prison (now called East Jersey State Prison) still exists, and has now served more than fifty thousand at-risk kids over a thirty-year period. Let’s do the math—using the 13 percent figure, we can estimate that the program has caused 6,500 kids to commit crimes they otherwise would not have committed. Scared-straight programs thus merit my bloodletting award: they do more harm than good.4
Why does taking part in a scared-straight program make kids more likely to commit crimes? The story-editing approach suggests an answer, by asking how these programs change kids’ interpretations of why they should stay out of trouble. The problem is that the programs provide kids with external motivation—wanting to avoid the horrors of prison—that can, paradoxically, undermine their internal motivation to take the straight path. A number of years ago, Daniel Lassiter and I demonstrated this phenomenon in a two-session experiment with college students. At the first session, we gave the students twenty trivia questions and told them that the answers were written on a blackboard, hidden behind a piece of cardboard. Because the participants took the quiz alone, they could have easily cheated by moving the cardboard and peeking at the answers, but none of them did, indicating that they had little intrinsic motivation to cheat in this situation. After all, it was a just a trivia quiz with no consequences for being right or wrong.
There were two conditions at this first session, one of which was just as I’ve described—those students took the quiz and decided not to peek at the answers. The second group received a strong external reason not to cheat—the experimenter told them that the data would be ruined if they peeked at the answers and that if they did they wouldn’t receive their credit for doing the experiment. Remember, no one was interested in cheating in the first place—and no one in either condition did—so this threat was superfluous.
A week or so later, the participants took part in what they thought was an unrelated study in a different location, where they took a very difficult achievement test that contained questions similar to those on the Graduate Record Exam. The experimenter instructed participants to spend about a minute on each question and then to move on to the next question without returning to previous ones. Because this was a difficult, diagnostic test, we expected that participants would be more tempted to cheat in this situation. And indeed, many did, by returning to earlier questions after moving on to later ones.
The question was whether the condition that the students had been in at the first session influenced how much they cheated in the second session. The participants who had received the strong superfluous threat not to cheat on the trivia quiz, we reasoned, might have attributed their lack of cheating during the first session to the threat rather than to their intrinsic honesty. That is, they might have said to themselves, “The reason I’m not peeking at the answers is because I might get caught and lose my experimental credit.” As a result of attributing their compliance to the threat, rather than to their lack of desire to cheat in the first place, they might have underestimated how intrinsically honest they were. If so, they might be more tempted to cheat at the second session, because they have come to view themselves as people who are not very honest. This is exactly what we found: the students who had received the superfluous threat at the first session spent more than four times as long reworking their answers on the second test than did the students who had not received the superfluous threat.5
How might this explain why scared-straight programs backfire? Some of the teens who take part in these programs probably have little interest in hanging out with gang members or selling drugs in the first place. But because they are considered “high risk,” they are put in the program and subjected to hardened convicts screaming in their faces that they will be raped and maimed in prison if they become delinquents. They receive strong threats not to do something they don’t want to do in the first place, which, as we saw, can have paradoxical effects. These participants get the message, “Hmm, maybe I am tempted by a life of crime if these convicts are going to such extreme measures to talk me out of it.” That is, just as with the college students in our cheating study, strong external threats at one point in time can increase interest in the forbidden activity at a later point in time. In this manner, scared-straight programs redirect teens’ narratives down the wrong path, making them more likely to give in to temptations and pressures from their peers in the ensuing months.
If this explanation is right, then it should be the case that scared-straight programs are especially likely to backfire with kids who were not interested in criminal activity in the first place, because these kids are most at risk of having their narratives changed from “I’m not very interested in a life of crime” to “Maybe it would be cool to hang out with the guys on the corner.” And, indeed, the original scared-straight program at Rahway State Prison was especially likely to backfire among kids who did not have a criminal record at the outset.
Maybe there is another reason why scared-straight programs fail—they don’t deliver a strong enough message. What if we got even tougher with at-risk teens? We could send them off to “boot camps” modeled after basic training in the military, where they learn a thing or two about discipline, hard work, and respect for authority. Wouldn’t that hammer some sense into them?
This approach has been tried without success, though that hasn’t stopped teen boot camps from proliferating. Boot camps became popular in the 1990s and there are now more than fifty such programs in the United States and other countries. Most are for first-time juvenile offenders who are assigned to boot camps instead of prison or other detention facilities. The juveniles typically attend for two to six months, during which time they wear uniforms, follow strict schedules, participate in military-style drills, engage in physical exercise, and perform assigned duties and chores. Although there are very few good studies of their effectiveness, those that have been conducted show that boot camps don’t work to reduce criminal recidivism. It’s easy to think of reasons why boot camps don’t work—maybe they are too punitive, maybe they catch kids too late in the game, or maybe they focus too much on discipline and too little on providing counseling or giving the kids useful skills such as job training. Or, like the scared-straight approach, maybe they don’t pay enough attention to teens’ narratives about who they are and why they might want to avoid a life of crime.6
Instead of getting tough with kids, maybe we should provide them with love, understanding, and positive role models. Such was the belief of Harvard physician and ethicist Richard Clarke Cabot, who in 1935 set in motion one of the most ambitious projects ever conceived to prevent juvenile delinquency. This project, which came to be known as the Cambridge-Somerville Youth Study after the cities in Massachusetts in which it was conducted, targeted a group of more than five hundred underprivileged boys. The researchers did virtually everything they could think of to add positive influences to the boys’ lives. The boys—who entered the program at the age of ten and a half, on average—were each assigned a counselor. The counselors established relationships with the boys and their families, and referred the boys to a variety of services as needed, including academic tutoring, psychotherapy, family interventions, and recreational activities. The boys received a great deal of individual attention over a five-year period; indeed, every possible problem area in the boys’ lives was addressed. The counselors also helped the boys’ families with such things as employment assistance and day care arrangements.
To the great credit of the researchers, they randomly assigned half the boys to a treatment condition that received all or most of the benefits just described or to a control group that received none of these benefits, thereby allowing an experimental test of the intervention. This project remains one of the most ambitious interventions ever tested with an experimental design in the social sciences. Unfortunately, it also ranks as one of the most spectacular failures. A thirty-year follow-up revealed that men who had been in the treatment program were more likely to have died at a young age, more likely to have committed repeated crimes, more likely to show signs of alcoholism, and more likely to have had serious mental illnesses than were the men in the control condition. Moreover, the longer the boys in the treatment group had participated in the program, the more likely they were to have negative life outcomes.7
What happened? One possibility is intimately related to the story-editing approach. In chapter 4, we saw how vulnerable children are to having their feelings and behavior labeled by adults. This can be a good thing if adults provide a positive label, such as “You are responsible kids who do not litter.” But by being selected for the Cambridge-Somerville study, kids were labeled by researchers, teachers, and counselors as “potential delinquents.” The researchers may have inadvertently given the kids the wrong story prompt, redirecting their narratives from “I’m a good kid” to “I’m a troublemaker,” or at least a potential one.
Another problem with the study, according to some experts, is that it brought the at-risk teens together on many occasions, such as on overnight camping trips. Some of the boys were also sent to summer camps, where they had a ready audience of other teens. This can be a problem, the argument goes, because teens learn from one another and reinforce negative patterns of behavior. In fact, the kids in the Cambridge-Somerville study who did the worst were those who went to summer camp two or more times.
This evidence is circumstantial, of course, but other studies have confirmed that giving at-risk teens the opportunity to hang out with and impress their peers produces negative outcomes. In one program, for example, at-risk teenagers met in groups once a week for twelve weeks. The researchers were trying to instill prosocial goals and teach the kids about self-regulation, but in fact the intervention had unintended negative effects: compared to teens randomly assigned to a control condition that did not meet in groups, those who met in groups were more likely to engage in delinquent behavior and to use tobacco products. If you want at-risk teens to act out and become even more deviant, it turns out that one of the best things you can do is to arrange for them to hang out together on a regular basis.8
This is probably another reason why scared-straight programs do more harm than good, and why I think we need to worry about Todd Walker’s approach of taking groups of teens to the funeral home. These kids are at precisely the age where the opinion of the sixteen-year-old standing next to them matters a lot more than anything the old dude screaming in their face has to say. And remember the Dollar-A-Day programs we encountered in chapter 5, in which teen moms are paid to avoid getting pregnant again? One reason these programs don’t work might be that they bring at-risk teens together.
What if we gave at-risk kids a lot of attention without bringing them together with their peers? We could assign them an adult mentor who meets with them alone rather than with other at-risk kids. As it happens, this is exactly what the Big Brothers Big Sisters of America program does, and it works. This organization has its roots at the turn of the last century in New York City. In 1902, a group of women began mentoring girls who appeared in Children’s Court, and this eventually became the Big Sisters organization. In 1904, a court clerk named Ernest Coulter began matching volunteer mentors with boys who appeared in his courtroom, and this eventually became the Big Brothers organization. The two groups spread rapidly to other cities throughout the United States, and, in 1977, merged to become Big Brothers Big Sisters of America. Today, there are chapters of Big Brothers Big Sisters in all fifty states and in twelve countries around the world.
The goal of the program is to foster a close relationship between a mentor and a child. Mentors are carefully screened and trained and then matched with a disadvantaged teen. The mentor meets individually with his or her mentee, typically for three to five hours a week, for an average time period of two and a half years. A case manager supervises and develops a plan for each mentoring relationship: for example, he or she might set goals related to school attendance and academic performance. But there are no required activities; rather, the emphasis is on the mentor and mentee finding things to do that they both enjoy and developing a close relationship. In a way, the program embodies Richard Clarke Cabot’s ideal of doling out love and friendship to at-risk kids while avoiding the parts of the Cambridge-Somerville Project that may have done it in (e.g., providing opportunities for kids to hang out with other at-risk teens).
The Big Brothers Big Sisters program was evaluated with a scientifically sound study in the 1990s. In eight American cities, kids ages ten through sixteen were randomly assigned either to get mentors right away or to be placed on a waiting list (the latter group got mentors after the evaluation ended; the time they had to wait was about equal to the normal waiting period for the program). After eighteen months, the youths in the mentoring group were doing better academically, had better relationships with their parents, were less likely to have started using drugs or alcohol, and were less likely to have hit someone than were the kids in the control group. These beneficial effects were especially evident among the youths who were in mentoring relationships for a year or longer.9
The Center for the Study and Prevention of Violence (CSPV) at the University of Colorado has done an excellent job of reviewing the literature and publicizing programs that have been shown to work, according to a strict set of criteria. In addition to the Big Brothers Big Sisters program, the CSPV identified a variety of other effective interventions. These programs target kids of different ages and use different approaches, though it is important to note that all avoid the problem of bringing at-risk teens together. It is worth describing the ones that work, because some (though not all) incorporate story-editing principles.
The Nurse-Family Partnership Program intervenes at the earliest point, namely, before children are born. Disadvantaged women who are pregnant with their first child are assigned a nurse who visits them at their homes two to four times a month, beginning when the women are pregnant and continuing for two years after the child is born. The nurses focus on prenatal care and health, the development of the child once he or she is born, and the personal development of the mother. In addition to advice on health care, for example, the nurses give child-rearing advice; encourage the mothers to develop plans for the future (e.g., educational and employment plans); and advise the mothers on family-planning methods and the timing of additional children. The program has been evaluated with three random clinical trials, each of which has demonstrated positive effects on prenatal health and early child development.
The oldest of the three studies followed the children until they were fifteen years old. Among children born to poor, unmarried women, the program succeeded in preventing adolescent delinquency. At age fifteen, the kids whose mothers had been part of the program were less likely to have run away, less likely to have been arrested or convicted of a crime or parole violation, and less likely to have abused alcohol or drugs, compared to the kids whose mothers had not been part of the program.10
Why has this program been so successful when other home-visitation programs have not? In chapter 4, for example, we saw that the Healthy Families America program, designed to prevent parents from abusing their children, doesn’t work. Although there are many differences between the Nurse-Family Partnership Program and Healthy Families America, two in particular stand out, both of which involve story editing. First, the Nurse-Family Partnership Program attempts to increase mothers’ self-efficacy, which is their belief that they have what it takes to carry out desirable actions. Specifically, moms are told that they have the ability to bring about desired changes in their own lives and their children’s lives, which probably helps to break a self-fulfilling cycle of negative thinking. In this respect it is similar to the interventions discussed in chapter 4 that do reduce child abuse. Second, the Nurse-Family Partnership Program has some similarities to the skill-based treatment program in the Netherlands (also discussed in chapter 4), in that it attempts to get mothers to attend more closely to their children’s needs and to foster secure mother-infant attachment relationships. Its success may thus be due in part to the fact that it incorporates some of the best features of other proven interventions.
The Promoting Alternative Thinking Strategies (PATHS) program targets young children in a classroom or preschool setting. The curriculum, which involves half-hour lessons taught three times a week, focuses on increasing and improving children’s self-control, emotional understanding, self-esteem, social relationships, and interpersonal problem-solving skills. Studies that randomly assigned classrooms to receive PATHS or not found that children in the PATHS classrooms exhibited superior self-control, had fewer conduct problems, tolerated frustration better, and better recognized other people’s emotions.11
As its name implies, the Olweus Bullying Prevention Program (OBPP) is designed to prevent bullying in schools. It targets kids at a wide range of ages, namely, those in elementary school through junior high school. The program, which is coordinated by a committee of school administrators, teachers, counselors, students, and parents, begins with an anonymous student questionnaire to gauge the extent of bullying in the school, followed by a conference day at which the results of the survey are presented to students, teachers, and parents. Teachers set firm guidelines against bullying and intervene swiftly when they see it in the classroom or the playground. They meet with bullies, their victims, and parents to come up with a plan to end the bullying. The program has been evaluated in at least four countries with positive results—in schools in which the program was implemented, there was a decrease in students’ reports of bullying.12
What about mental health services? Two forms of family therapy have been shown to be effective, both of which target teenagers who are exhibiting conduct problems. Multisystemic therapy (MST) targets teenagers who are at risk for being placed outside of the home. A therapist sees the juvenile and his or her family in their home for several hours per week, typically for four to five months. A major focus of MST is to teach parents and other caregivers better parenting skills. Another focus is to reduce the juveniles’ contact with deviant peers and increase their contact with peers who are better role models. That is, the therapy specifically targets the problem we encountered earlier with other interventions, namely, that contact with peers increases deviant behavior, especially like-minded peers. One recent study, for example, randomly assigned juvenile sex offenders to receive either MST or therapy provided by the juvenile probation department. A year after beginning the treatment, those who received MST were significantly less likely to be engaging in delinquent behavior in general and sexually deviant behavior in particular. The more the kids’ parents improved their parenting skills in the MST group, and the less time the teens spent with at-risk friends, the better the outcomes. There have been at least fourteen experimental tests of MST involving more than thirteen hundred families; these studies have shown that MST improves family functioning, increases school attendance, reduces criminal activity, and reduces the use of alcohol and drugs.13
Functional family therapy (FFT) also targets teens that have conduct disorders, but takes less time (a total of eight to twenty-six hours, depending on the severity of the case) and can be delivered by people with a variety of training (such as probation officers and mental health professionals) in a variety of settings (including the home and juvenile court). It has three phases: engagement and motivation, in which the leader builds trust with teens and their families and targets areas to change; behavior change, in which leaders try to improve parenting skills, family communication, and conflict management; and generalization, in which the leader attempts to maintain positive changes, in part by connecting the family to community resources. Experimental tests of FFT have yielded encouraging results: the criminal recidivism rate among teens who got FFT was 26–73 percent lower than it was among teens who did not get FFT. The intervention even reduced the arrest rate of the siblings of the treated teens.14
What if teens have already committed offenses that put them at risk for incarceration? The Multidimensional Treatment Foster Care (MTFC) program is an alternative to incarceration in which teens are placed in foster families. The families receive twenty hours of training in ways to manage a teenager who has conduct problems, and once the child is placed in a family there are a lot of services available. There are weekly meetings of foster parent families, individual therapy sessions for the teens, and group therapy sessions for the teens’ parents or other caregivers. The teen’s case manager monitors his or her progress at school, maintains daily contact with the foster family, and administers an individual program that each teen is expected to follow. Psychiatric care and medication are arranged as needed. Teens typically spend between six and nine months with the foster family.
MTFC has been evaluated in a number of experiments and has been found to have beneficial effects for both boys and girls. In most studies, teens are randomly assigned either to the program or to the normal course of care in their state, which typically involves group residential facilities. Those who get MTFC are less likely to run away, less likely to be arrested, and less likely to use hard drugs than are those in group residences. One recent study confirmed that an important component of MTFC’s success is the fact that case managers monitor the contact the teens have with other teens and do their best to minimize it, thus limiting the amount of time teens spend with their delinquent peers.15
There are other ways to reduce violence that involve story editing more directly. In chapter 5, we saw that teens who feel alienated from their community are most at risk to become pregnant or get someone pregnant. It seems reasonable to assume that alienated, disengaged kids are also more likely to go down the path to delinquency. If so, then the same story-editing interventions that reduce teenage pregnancy—getting kids involved by having them do volunteer work—might also reduce delinquent behavior, including violence.
There is evidence that this is indeed true from two of the programs we encountered in chapter 5, Teen Outreach and Reach for Health (RFH). Not only were students in Teen Outreach less likely to become pregnant, they were significantly less likely to fail a course or be suspended from school, suggesting that the benefits of the program extend beyond preventing teenage pregnancy. Furthermore, eighth graders who participated in the RFH program became less violent. On a survey administered six months after the program, these students reported that they got into fewer fights, were less likely to have threatened someone, and were less likely to have a knife or gun than were students in the control group. This same reduction in violence was not found among seventh graders who were in the program, possibly because they engaged in less extensive volunteer work than the eighth graders did.16
Here is how one of the eighth graders in the Reach for Health program described the experience of volunteering in a nursing home: “The elders benefit because they get to talk to people. Some of them are lonely. And they tell us some great stories. This one guy was telling me all about his experiences in World War II. I love going there and I know that they love us being there.” Or consider this report from a teacher who helped supervise the volunteer work: “We take this one kid who is one of the most violent kids in our school. He’s always getting into fights. And then you see him at this nursing home. There’s this patient who is paralyzed in one arm from a stroke. She can’t feed herself. And for three hours, this boy sits next to her and just feeds her. One spoon after another. And he looks straight into her eyes. You can see how much he loves her.”17
Now, as I’ve suggested at several points in this book, we can’t rely on testimonials alone to evaluate a program. But we know from good experimental studies that volunteering reduces teen violence, and these quotes may provide clues as to why—teens change their view of themselves as a result of their volunteer work. Someone who thinks, “Mrs. Johnson is looking forward to my visit next week, and I’m looking forward to seeing her” is probably a lot less likely to get into a fight and risk getting suspended than someone who feels no connections at all.
In 1982, James Wilson and George Kelling published an influential article in The Atlantic magazine entitled “Broken Windows: The Police and Neighborhood Safety.” The major premise of the article was that appearances matter when it comes to preventing crime. Minor signs of disorder, such as broken windows and graffiti, signal to people that a neighborhood is deteriorating and that it is permissible to break the law—leading to an increase in more serious crimes. A number of police departments took this to heart by adopting “zero tolerance” approaches to petty crimes, such as drinking in public, graffiti, and vandalism, with the assumption that preventing highly visible minor crimes would decrease people’s tendency to commit more serious crimes.
The “broken-windows” theory adopts a classic story-editing approach because it attempts to alter people’s interpretations of their environment by changing relatively minor story prompts (e.g., graffiti on public buildings), under the assumption that doing so will lead to big changes in behavior (reduced crime). Whether it really works has been a matter of debate. Supporters point to the fact that when New York City adopted this approach in the 1990s, the crime rate dropped dramatically. Critics point out that the crime rate was dropping throughout the country at this time, even in cities that did not adopt the broken-windows approach.
As we have seen, there is nothing like a good experiment to settle the question, and researchers in the Netherlands obliged by experimentally manipulating the existence of public signs of lawlessness and observing whether this influenced people’s criminal behavior. In one study, for example, they put a five-euro bill in an envelope that had a cellophane window and placed the envelope halfway out of a mailbox, so that the bill was clearly visible to passersby. Then, from a hidden location, they observed how many passersby stole the envelope. When there were no visible signs of lawlessness, only 13 percent of the passersby gave into temptation and pocketed the envelope. But when the researchers added signs of lawlessness—graffiti painted on the mailbox or litter scattered on the ground below—this percentage doubled.
This study shows that signs of disorder, such as graffiti and litter, can cause people to reinterpret a situation as one in which other kinds of disorder are permissible, such as stealing money. People are highly sensitive to social norms (information about what other people are doing and what they approve of), and subtle indicators of these norms can have dramatic effects on people’s behavior (we will encounter this lesson again in the next chapter). When there are signs that lawlessness is the norm, people are more likely to act lawlessly.
But does it work for police to clean up signs of disorder and enforce laws against petty crimes? There is evidence that it does. Researchers identified thirty-four high-crime areas in Lowell, Massachusetts, and randomly assigned half of them to receive the broken-windows strategy of policing. The police eliminated signs of disorder and petty crime in these areas by cleaning up vacant lots, improving lighting, finding shelters for homeless people, making arrests for drug dealing and drinking in public, and increasing foot and car patrols. During the next year, there were significantly fewer serious crimes, such as robbery and assault, in the areas that received the broken-windows policing strategy than there were in the areas that did not. Signs that lawlessness is the norm—such as vandalism and public drunkenness—do appear to breed more serious crimes.18
I have the greatest admiration for the Todd Walkers of the world, people who see a problem in their communities and devote their lives to solving it. Walker’s Restoring Inner City Peace program required a great deal of time and effort to put into place and it certainly meets the “common sense” criterion. How can it hurt to scare some sense into at-risk kids by taking them to funeral homes? Well, we’ve seen that it can hurt in at least two ways, by giving at-risk teens more of an opportunity to hang out together and by encouraging the teens to think, “Maybe I am attracted to a life of crime if it takes visits to a funeral home to scare me out of it.”
What might Walker and other inspired people do, instead of trying to scare kids? It turns out that there are several existing programs that work, such as Big Brothers Big Sisters of America. Matching kids with effective, caring mentors through this organization, or others like it, will do more good than taking them to a funeral home. Involving the kids in volunteer work in their communities is another effective approach.
There is a larger lesson here for all parents, namely, that we should care a lot about who our kids are hanging out with. It is hardly newsworthy, of course, that peer pressure can be very strong among adolescents. I think we sometimes forget, however, how powerful these influences are. (They certainly were a surprise to the researchers who designed programs to help at-risk teens, only to find that the programs backfired by bringing the teens together.) As parents, we can’t control everything our teens do, nor should we try to. It’s not like we get to pick our kids’ friends, especially when they get to middle school and beyond. But by encouraging kids to engage in some activities and not others, we can influence the pool of teens from which our kids will pick their friends. This is one way that volunteer work can help—by exposing kids to like-minded peers who are interested in helping others. Volunteering also helps kids, as we have seen, by fostering the narrative that they are effective people who have a stake in their community.
Finally, we saw that people are sensitive to signs of what other people are doing and what they approve of, and that such things as graffiti and litter can signal to people that criminal behavior is permissible. Keeping one’s neighborhood clean and orderly will encourage others to frame it as a place where they ought to behave well.