The risk that an adolescent will become involved in violent offending and/or be a victim of violence varies based on a variety of factors, including individual characteristics, family characteristics, peer and school influences, neighborhood environment, and daily activities. Although there is no formula for determining exactly who will become violent (or a victim of violence), it is clear that some individuals are at greater risk than others. This section identifies those factors that are associated with an individual's increased risk for involvement in juvenile violence. For purposes of this report, a juvenile is defined as an individual less than 18 years of age.

Individual Factors

Individual factors refers to the broad range of individual characteristics that may be related to behavioral patterns in a variety of ways. These factors include demographic characteristics, such as gender, race, and age, and physiological and psychological characteristics. The focus of this section is on demographic predictors of violent behavior.

Juveniles At Risk of Becoming Violent

In general, the most powerful demographic predictors of individual violent criminality are gender and age. Boys in late adolescence and young men are much more likely to be serious high-rate offenders than girls or older men (Chaiken, 1998b). Further, studies using official record data have consistently found greater involvement of African-Americans in violent offending than of Caucasians (LaFree, 1995). Overall, the research findings from the projects included in this report confirm these patterns.

Gender. Violent offenses are overwhelmingly committed by males. In the DC study of juvenile violence, of the 2,686 juveniles charged with the 4 most serious person offenses (homicide, rape, robbery, and aggravated assault), 82% were males. The SC homicide study found that 88% of the juveniles who committed homicide between 1992 and 1994 were male. Not surprisingly, males were also more likely to display early signs of aggressive behavior, specifically in the form of bullying. The SC bullying study reveals that males were significantly more likely than females to report bullying their peers and twice as likely as females to engage in physical actions to bully others.

The Causes and Correlates research findings indicate that, in general, a greater percentage of males are involved in serious violence than females (Tatem-Kelley et al., 1997). This is consistent with past research findings indicating that violence is more prevalent in males. However, females reported considerable involvement in serious violence. In the Denver sample, the prevalence of serious violence among females ages 13 to 15 was more than half that of males the same age. The difference was even less in the Rochester sample. In fact, at age 13, 18% of females reported the commission of serious violence compared with 16% of males. Thus, females appeared to be increasingly involved in violent behavior.

Age. Results from the violence studies indicate that many juveniles involved in violent behavior begin this behavior by age 15. In the DC juvenile violence study, of the 2,686 juveniles charged with the 4 most serious person offenses,¹ almost 40% were 15 or younger. In the SC homicide study, the mean age at the instant offense of youth in the homicide group was 15.8. Youth in the assault and battery group averaged 15.6 years and youth in the other serious offenses group averaged 15.1 years.

Past research has generally indicated that rates of violence among males tend to peak at ages 15 to 17 and then decline (Tatem-Kelley et al., 1997). The Causes and Correlates research has not documented a decline in males' self-reported involvement in serious violence in late adolescence. However, females did show an expected age curve with prevalence rates peaking in midadolescence and declining thereafter.

Race. Both the DC juvenile violence study and the SC homicide study found that African-Americans were disproportionately arrested for violent offenses. African-Americans account for approximately 65% of the total population in the District of Columbia, yet the DC juvenile violence study found that, of the 2,686 juveniles charged with the 4 most serious person offenses, 98% were African-American. Whereas the total population in the State of South Carolina is approximately 30% African-American, the SC homicide study found that 82% of the juvenile homicide offenders referred to the State solicitor were African-American, 16% were Caucasian, and 2% were other races. African-Americans were somewhat overrepresented in the homicide and assault and battery groups compared with the other serious offender group.

In the Causes and Correlates study, prevalence rates were examined by age and ethnicity. In Denver and Rochester, three ethnic groups were included: Caucasians, African-Americans, and Hispanics. Because there were virtually no Hispanics in the Pittsburgh sample, only Caucasians and African-Americans were studied in that sample. With only one exception (18-year-olds in Rochester), prevalence rates were higher among minority groups than among Caucasians at each age and site.

Juveniles At Risk of Becoming Victims of Homicide/Violence

Most commonly, studies have revealed that juvenile homicide victims are of the same race and gender as their perpetrators and that the most likely victims of juvenile homicide are acquaintances, followed by strangers, and then family members (Melton et al., 1998). The most recent national data indicate that in 1995, 54% of victims were acquaintances, 36% were strangers, and 10% were family members (Sickmund et al., 1997).

In terms of gender, the DC juvenile violence study found that between 1993 and 1995, 88% of juvenile homicide victims were male. The results further show that in 1993, 57% of juvenile assault victims were male and in 1994, 59% were male. Regarding the age of juvenile victims, between 1993 and 1995, 10% of juvenile homicide victims in DC were 11 years of age or younger. Nearly 69% were 16- or 17-year-olds.

The DC juvenile violence study found that the majority of juvenile victims of violence are African-American. All but one of the juvenile homicide victims between 1993 and 1995 were African-American. The DC juvenile violence study also found that during 1993, 95% of youth victims of all nonfatal violent crimes were African-American -- 1,476 as compared with 79 Caucasian youth. In 1994, 94% were African-American.

Demographic Characteristics of Participants in Homicide Incidents Involving Juveniles

Unlike the DC juvenile violence and SC homicide studies, which focused on juvenile offenders and juvenile victims as exclusive categories, the Los Angeles homicide study looked at victims and offenders of homicide incidents involving juveniles as one group. Thus, the Los Angeles sample includes both juveniles and adults. Of the 311 homicide incidents, 30% involved a juvenile victim but only adult suspects, 46% involved only adult victims but a juvenile suspect, and 24% involved juveniles as both victims and suspects. This finding is interesting because it indicates that adults are frequently involved in violence by and against juveniles. Slightly less than one-quarter (24%) of the homicide incidents involving juveniles were "kids against kids."

The Los Angeles homicide study found that of the 1,248 individuals designated by law enforcement as victims or suspects in 311 homicide incidents involving juveniles, 92% were males and 96% were minority (58% Hispanic and 28% African-American). Thus, in terms of gender and race, the Los Angeles homicide study echoes results from the other studies showing substantial involvement of minority males in juvenile violence. The mean age of all victims was approximately 23 years and the mean age for offenders was approximately 18.5; median ages for victims and offenders were 17 and 18, respectively.

Additional Individual Factors

A number of additional individual-level factors beyond the demographic characteristics of gender, race, and age are linked with subsequent violent activity (Hawkins et al., 1998). These factors include hyperactivity and risk-taking behavior, aggressiveness, early initiation of violence (by age 12-13), and involvement in other forms of antisocial behavior. These factors are beyond the scope of most of the present studies.² However, some did look at criminal history factors.

Criminal History Factors

Research on the careers of serious and violent offenders suggests that early onset of delinquency and violent behavior predicts more serious and chronic violence among youth (Hawkins et al., 1998; Thornberry, Huizinga, and Loeber, 1995). The Causes and Correlates projects in Denver, Pittsburgh, and Rochester examined the ages of onset of serious delinquency³ for juvenile offenders in urban areas and found that most males who eventually became persistent serious offenders had committed their first serious nonviolent offense by age 14 -- 85% in Pittsburgh, approximately 65% in both Denver and Rochester (Stouthamer-Loeber et al., 1997). The average age of first referral in the SC homicide study was 14 years for youth in the homicide group, 14.2 years for youth in the assault and battery group, and 14.1 years for youth in the other serious assault group.

Relatively few studies have examined patterns of delinquent and criminal activity among juvenile homicide offenders in particular; most focus on violent offenders as a whole. The SC homicide study compared juvenile homicide offenders with other serious juvenile offenders on the offense type of first referral to the State solicitor. For youth in the homicide group, their first referral most typically was for an offense against persons (33%). In every instance, this offense against persons was the target homicide offense. The next most frequent type of referral was an offense against the public order (31%) (e.g., driving under the influence), followed by property offenses (17%), other offenses (11%) (e.g., blackmail or extortion, driving with a suspended license), and status offenses (8%).

In comparison with the homicide group, the most frequent first referral for youth in the assault and battery group was also for an offense against persons (40%, all of which were the target offense of assault and battery with intent to kill), followed by property offenses (22%), other offenses (17%), public order offenses (15%), and status offenses (7%). The other serious offense group differed, however, in that the most frequent first referral was for a property offense (32%), followed by offenses against persons (28%), public order offenses (27%), other offenses (8%), and status offenses (5%). For the majority of the group, their first referral was for the target offense.

These findings suggest that juveniles who are referred for homicide and those who are referred for assault and battery with intent to kill are similar in that both groups lack official juvenile justice records that could be used to identify them before they are involved in fatal or near-fatal offending. However, this does not eliminate the possibility that they display other problem behaviors that could be detected in other systems (e.g., schools, social services).

Neighborhood Factors

In addition to individual factors, contextual factors contribute to an adolescent's risk of violence. Such contextual factors include family, school, peers, and community and neighborhood factors (Hawkins et al., 1998). Several studies cited in this report examined neighborhood factors associated with juvenile violence.

The DC juvenile violence study found that one census tract, the Douglas neighborhood of Southeast DC, emerged as a high-risk zone for the three nonfatal violent crimes. It was the highest risk tract for rape and assault of juveniles in 1993 and 1994 and for juvenile robbery victimizations in 1993. The tract has a poverty rate of 41%, compared with 17% for the District as a whole. In addition, 86% of households were single-parent (female-headed) households.

The SC study of community social disorganization and crime examined rates of juvenile violence in 264 rural counties (in 4 States) with total populations ranging from 560 to 98,000. The juvenile populations included in the analyses ranged from 50 to 11,000. The study found that juvenile violence was consistently associated with rates of family disruption, ethnic heterogeneity, and poverty. Juvenile arrest rates for violent crimes displayed a curvilinear relationship to population size such that per capita arrest rates went up with increases in juvenile population in the range from 50 to 4,000. Beyond this level, increasing juvenile population had little impact on arrest rates for violent offenses other than robbery.

The Milwaukee homicide study examined the spatial distribution of homicide victimization, both adult and juvenile, in census tracts of Milwaukee. The researchers found that the majority of victimizations from 1989 to 1993 were concentrated in the most deprived census tracts of the city, labeled "dangerous neighborhoods." Homicides of juveniles ages 13 to 17 were even more concentrated. Eighty-five percent of the 34 juvenile victimizations during this time period occurred in 19 "dangerous neighborhoods," with more than half taking place in just 4 neighborhoods. Three of these neighborhoods were among the most disadvantaged in the city's African-American community, as measured by neighborhood stress levels and economic opportunity scores.

Additional analyses using life history information obtained on the juveniles interviewed in the Milwaukee homicide study revealed that all 11 of the juvenile gang-related offenders resided in extreme poverty areas, 10 lived in single-parent households, and 6 experienced serious household violence. Overall, out of 29 juvenile homicide offenders (including 4 groups -- gang-related, drug-related, robbery-related, and other) interviewed, 90% lived in a single-parent household and 45% reported serious incidents of household violence. These findings indicate the extent to which juvenile homicide offenders live in disadvantaged homes and neighborhoods.

Prevalence of Violence Among Youth in High-Risk Neighborhoods

Twenty years of research repeatedly has shown that in any city or neighborhood a small percentage of offenders are responsible for committing a large proportion of the crime that occurs there (Chaiken, 1998b). Two violence studies, the DC survey and the Los Angeles survey, specifically explored the prevalence of violence among adolescent males in high-risk neighborhoods.

The data collected in the DC survey support prior research findings that a small group of offenders are responsible for a large percentage of violent crime. Among all boys interviewed, 7% were responsible for committing 36% of all the reported delinquent acts. This small number of youth committed close to one-fourth (21%) of all juvenile assaults, close to half (44%) of all drug deals, and close to half (44%) of all property crimes committed by the entire group of boys in the 6 months prior to this study.⁴

A substantial number of studies also demonstrate that few youth make it through adolescence without doing something that could get them into trouble, but most are not seriously involved in crime. Relatively few of the boys (22%) interviewed in these DC neighborhoods failed to self-report any acts that would be considered criminal. But even these "good kids," in the 6 months before the study, committed, on average, more than one act that could be considered a juvenile offense, such as running away or underage drinking.

The Los Angeles survey found that 30% of the boys interviewed from high-risk neighborhoods reported committing at least one violent offense in the 6 months prior to the interview. The most common offenses were throwing bottles or rocks at people (15%), being in a gang fight (9%), and hitting someone with the intent to hurt them (13%). With regard to victimization, the Los Angeles survey found that 34% reported at least one violent victimization within the 6 months prior to the interview. The most common types of victimization were having objects thrown at them (21%) and being hit (13%). Eight percent of the youth reported being attacked with a weapon. As the Los Angeles survey shows, some youth were both offenders and victims (19%). Only 11% were offenders but not victims and 15% were victims but not offenders. The majority (55%) were neither victims nor offenders.

Summary

Although the results of these studies cannot be generalized to the total population of juveniles, these individual snapshots appear to be consistent with findings from past research. Overall, juvenile violence is committed primarily by males and often occurs intraracially among minority males. While some younger adolescents do commit violent offenses, the majority of juvenile offenders and victims are 16- and 17-year-olds. An examination of neighborhood factors indicates that many violent juvenile offenders live in disruptive and disorganized families and communities. However, as the surveys with the children living in high-risk neighborhoods show, the majority of youth who live in such environments are not involved in serious delinquency.

1 Includes homicide, rape, robbery, and aggravated assault.

2 The Causes and Correlates studies have examined these factors in great detail. Information on these results can be found in the book Serious and Violent Juvenile Offenders: Risk Factors and Successful Interventions (Loeber and Farrington, eds., 1998).

3 Includes aggravated assault, robbery, rape, gang fights, burglary, theft over $50, arson, auto theft, fencing, forgery, and fraud.

4 Analysis of the DC survey data identified six varieties of criminal behavior among the adolescents interviewed: (1) "good kids" who committed no criminal acts (22%), (2) "fighters" who committed only assaults (19%), (3) drug dealers who sold drugs and committed occasional assaults (5%), (4) property offenders who do not commit robbery or drug dealing (32%), (5) property/dealers who commit property offenses and drug dealing but not robbery (16%), and (6) robbery offenders, the 7% of boys responsible for committing 36% of all the reported delinquent acts.

Juvenile Violence Research	OJJDP Report to Congress