Research Watch: What's to Blame for Youth Violence?

January 2001
Executive summary, full report, and related press information available free at www.surgeongeneral.gov/library/youthviolence, or at (800) 789-2647

The first-ever surgeon general's report on youth violence was recently released by Dr. David Satcher, a Clinton appointee who still holds his position in the Bush administration. The report hardly made a ripple in the public debate, but what stands out as especially noteworthy were the press reports regarding what was not in the report.
In a press conference, Dr. Satcher was asked why media violence was barely mentioned in the report; he responded that the media is not a major influence on youth violence. As someone who has read dozens of studies and reports about the impact of media violence on children and society, I was surprised to hear this. It sounded eerily like other comments on the topic over the previous weeks: a report on ABC's "20/20" claiming that media violence does not cause violence and may actually be good for kids, and a scathing Youth Today column by Mike Males ("Unreality Bites," November 2000) claiming that media violence was taking attention away from more important policy issues.

But what about the voluminous stack of research reports on the impact of media violence on youth? When a TV news magazine claims that TV violence is not dangerous, I don't take it too seriously, but I was beginning to feel outnumbered. More importantly, how would youth workers respond to these reports? This Research Watch is an opportunity to take a careful look and try to figure out what is going on. What we will find is relevant to all media coverage of research on youth, not just the topic of violence.

Totally Bogus?

On "20/20," Jonathan Freedman, a psychology professor at Toronto University who happens to receive funding from the Motion Picture Association of America, claimed that research does not support the notion that media violence causes aggression. He trashes reports by the American Psychological Association, the American Academy of Pediatrics and others that claim that more than 1,000 studies prove the case against media violence, saying: "There aren't over a thousand studies. There are about 200 studies, give or take a few, depending on which ones you count."

Isn't 200 enough? (In the interest of disclosure, readers should know that this writer co-authored some of those studies.) Actually, the most important issue is not the exact number of studies, but their quality.

There are dozens of well-designed studies that show that TV, movies and other media affect what viewers believe and how they behave. This is true of many different kinds of attitudes and behaviors — positive and negative — but many studies conclusively show a statistical link between watching violent programs and behaving aggressively. And, of course, billions of dollars have been spent on media advertising because it is well established that even brief messages can be powerful in shaping behavior.

There are very few studies, however, of whether exposure to media violence causes criminal behavior.

The seminal study of media violence and criminal behavior (rather than aggressive behavior) is by Leonard Eron, Ph.D., and Rowell Huesmann, Ph.D. In 1960, Dr. Eron began studying aggression and the TV viewing habits of 875 third-graders in upstate New York. The researchers tracked some of those children until they were 30. Their conclusion: The eight-year-old children who watched more violent TV programs were more aggressive. They also found the kids who scored higher on aggression when they were eight were far more likely to be arrested as adults, to have moving traffic violations, and to abuse their children.

ABC trashed this research through the words of science writer Richard Rhodes. Rhodes stated that Huesmann testified at a congressional hearing that it was possible to predict whether someone would have been arrested by the age of 30 according to how much violent television they had watched when they were eight. This "electrified the committee," said Rhodes, but is "a totally bogus finding."

Like beauty, "totally bogus" is in the eyes of the beholder. The use of the word "predict" was probably misunderstood because its statistical meaning is different from what most of us mean when we use the word. In this study, there was a statistical analysis showing that children who watched more violent television were more likely to be arrested as adults — that watching TV violence predicts later criminal behavior. However, that does not mean it is possible to use that information to accurately predict exactly which children will grow up to be criminals or child abusers.

The term "predict" in this case simply means that the kids who watch more violent TV are more likely to be arrested as adults. Since criminal behavior is not very common, the statistical relationship could be caused by a small number of children who watch a great deal of violent TV and grow up to be criminals. It is not possible to accurately predict exactly which children will be influenced, because many children who watch violent TV are not arrested as adults — the research only tells us that watching violent programs increases the likelihood of arrest as adults. It does not tell us anything else about those who are or are not arrested, or who become child abusers.

A Stew of Risk Factors

Regardless of the strengths or weaknesses of this one longitudinal study, there are dozens of studies showing that exposure to media violence increases the likelihood of violent behavior. Some studies found children imitating unusual aggressive behaviors that they had just seen on a TV program, while others are based on parents' measures of children's total TV viewing over a period of several weeks, linked to teachers' ratings of the child's general aggressiveness or cooperative behavior in school or on the playground. Note the distinction between research on violent behavior (which is relatively common) and arrests (which are relatively rare and therefore more difficult to predict unless you have many thousands of adults in your sample.)

When the surgeon general explains that media violence is not a major cause of youth violence, or when Mike Males complains that there is too much attention to media violence compared to other causes of violence, they are correct that there are other causes of violence that are probably more important. However, most "risk factors" for violence — the factors that increase the likelihood that a youth will be violent — are not strong predictors by themselves. It is the combination of risk factors that tends to lead to violence. Media violence, like other risk factors, may have a very strong impact on some children, and no apparent impact on others. Unfortunately, we don't yet know which children will be affected, and which won't.

In his remarks, the surgeon general specified that violence is influenced by the availability of guns. However, the report itself takes a developmental perspective: how personal characteristics interact with the social context, from prenatal factors to adolescence.

The report specifies that its most important message is that youth violence is not an intractable problem. We have the tools to reduce or prevent violence, the report says, but we waste most of our resources on programs that do not work or may not work, instead of focusing on those that are proven to be effective.

The report describes the risk factors that increase the likelihood that a youth is violent and the protective factors that decrease the likelihood. They point out that risk factors do not necessarily cause violent behavior — they may just be correlated.

Risk factors include factors that are relatively unchangeable, such as being male, being hyperactive and having a low IQ, as well as those that can potentially be changed, such as exposure to TV violence, antisocial attitudes, substance use, poverty, gang membership, and abusive or neglectful parents.

Protective factors also include factors that can't be changed (being female, having a high IQ) and those that can be changed (positive social interactions, perceived sanctions for transgressions, parental monitoring, and school recognition for involvement in conventional activities).

The Cost of Curbing Violence

The report includes a table of the comparative costs and benefits of prevention and intervention programs — a purely economic approach to the issue. Several programs seem to save at least 60 cents for every dollar spent: the Perry Preschool Program, the Seattle Social Development Project, Prenatal and Infancy Home Visitation by Nurses, and Big Brothers Big Sisters of America. Listed as costing much more than they save: the Syracuse Family Development Research Program, the Quantum Opportunities Program, and boot camps.

The big winners from a purely monetary perspective were three types of community-based programs for adolescent juvenile offenders: multi-systemic therapy, functional family therapy, and multidimensional treatment foster care. For example, the latter was estimated as saving the taxpayer more than $14 for every dollar spent compared to the costs of treatment in a regular group home.

These comparisons indicate that it is easier to save money by improving programs for juvenile offenders than by prevention efforts aimed at a more general population of "at-risk" children. That's because the kinds of childhood and youth behaviors that are linked to later violence are difficult or impossible to change in the programs that were evaluated: There were no evaluations of programs aimed at preventing broken homes, teaching parents to provide love and support to their children, improving parents' use of discipline, or restricting children's exposure to violent media.

If we are to believe the surgeon general's conclusions, we probably would want to focus on providing more and better parenting programs in schools throughout the country, then evaluate them to make sure they work. Unfortunately, the report does not provide the detailed research information necessary to judge the accuracy of its conclusions; it is a broad overview of previously published research that provides no data for the discerning reader.

The bottom line is that there are many factors — at home, at school, and in the community — that can increase or decrease the likelihood that a youth will become violent. Media violence is one of them. In a logical world, this factor would be relatively easy to change, compared to poverty, family relationships and school failure. In the U.S., those changes have not been a public policy priority — leaving the burden on youth workers and parents to grapple with the problem, mostly without the kinds of programs that have proven to be effective.

Diana Zuckerman, Ph.D., is executive director of the National Center for Policy Research for Women and Families, based in Washington, D.C. Contact: cpr4wandf@aolcom.

Zuckerman, Diana. "What’s to Blame for Youth Violence?" Research Watch review of "Youth Violence: A Report of the Surgeon General." Youth Today, March 2001, p. 14.

©2000 Youth Today. Reprinted with permission from Youth Today. All rights reserved.