clear Conditions Associated With an Increase in Violence Committed By or Against Juveniles

There are certain situational conditions that appear to be associated with an increase in juvenile violent offending. For example, incidents of juvenile violence occur more frequently in certain locations and at certain times of the day. Further, the existence of gangs in a community may influence the level of, and be a significant factor in the motivation for committing, violent crimes.


Few researchers have examined the specific locations of homicides committed by juveniles. In one previous study conducted in metropolitan Detroit, Goetting (1989) observed that 19% of juvenile homicides took place in the residence of the victim, 17% occurred in the home of the offender, and 15% occurred in another residence. Thus, about 50% occurred in a place other than a private residence. The locations of victimizations (juvenile violence) in DC showed a clear association with schools. That is, a disproportionate share of juvenile victimizations occurred in or near schools. The Los Angeles homicide study found that 73% of homicide incidents involving a juvenile took place in a public place: street (47%), vehicle (19%), or parking lot (7%). Twenty-three percent were drive-by shootings.

Time of Day

Several studies looked at the time of day during which juvenile violence occurs. In 1993, almost half of all juvenile homicide victimizations in DC occurred between 7 a.m. and 5 p.m.; only 22% occurred between 10 p.m. and 2 a.m. The DC juvenile violence study found that the victimization patterns for all violent crimes during the school year were different from the victimization patterns during the summer break. During the school year, victimizations peaked at 3 p.m., whereas during the summer, victimizations were highest at 10 p.m. and peaked again at 1 a.m. In the Milwaukee study, 28 out of 35 juvenile homicide victimizations (80%) occurred between 4 p.m. and midnight. In contrast to the findings on juvenile homicide victimization in DC, homicides involving juveniles as victims or perpetrators in Los Angeles most frequently occurred late at night, with the peak hour being from 10 p.m. to 11 p.m. Almost half (46%) of these incidents occurred between 10 p.m. and 6 a.m.

These findings suggest that juvenile violence frequently occurs during times when juveniles are less likely to be supervised (e.g., after school or late at night). The DC survey found that the vast majority of boys (75%) spent the afterschool hours unsupervised by any adult for 1 or more days each week. Almost half (48%) were in settings where the absence of an adult prevailed every day after school. The relatively few youth who were in a supervised setting after every school day reported lower rates of delinquency than those youth with fewer afterschool hours supervised by adults (40% of "good kids" were with some adult every day after school, compared with 20% of robbery offenders).

A more important factor than actual adult supervision may be the mere knowledge by a caretaker of where his/her children are after school. The DC survey found that only 9% of the "good kids" reported that their primary caregiver rarely or never knew where they were after school, compared with 15% of the "fighters," 18% of property offenders/drug dealers, 22% of property offenders, 30% of drug dealers, and 33% of robbery offenders.

Further compounding the lack of parental supervision of the boys in the three DC neighborhoods is the fact that teachers and other school staff appear to have given up trying to provide constructive guidance. The vast majority (76%) of boys had been suspended from school at least once; more than half (57%) of the "good kids" had been suspended at least one time. Furthermore, the schools expelled more than 20% of all boys in the three neighborhoods. Although suspensions and expulsions may be justified from the school authorities' point of view, common sense would indicate that simply releasing large numbers of adolescents out into the community, unsupervised during school hours, is likely to raise the number of delinquent acts committed.

In fact, boys who reported being suspended also reported committing, on average, more than three times the number of delinquent acts in the past 6 months as boys who reported no suspensions. The few who at the time of the interview said that they were not in school reported committing, on average, more than four times as many delinquent acts in the prior 6 months as those who said they were attending school at the time of the interview. However, the high level of delinquent activity could simply be an effect of being suspended, since there is more time to get in trouble.


Research on gangs and gang crime is complicated by the fact that vast differences exist in how local and State agencies define gangs and gang membership and how crimes are classified as gang crimes. Moreover, no national-level data exist on the juvenile proportion of gang members or the volume of gang crime (Snyder and Sickmund, 1995). However, some studies have begun to examine the issue of juvenile gang violence at the local level, including several of the juvenile violence studies.

The Los Angeles homicide study found that gangs played a role in almost 80% of the adolescent homicide incidents examined. Two general dimensions of gang involvement, gang member participation and gang motivations, were used to determine the role of gangs in these homicides. Out of 302 incidents in which the motive (or possible motive) for the homicide could be determined, 34% involved at least one current gang member and a clear gang motive. An additional 45% involved at least one current gang member and had either a possible gang motive or other type of motive.

Prior research demonstrates that adolescents who join street gangs are more involved in delinquent behavior than are adolescents who are not involved in gangs. This is especially true for serious and violent offenders (Thornberry and Burch, 1997). The Rochester Youth Development Study (RYDS) examined the proportion of delinquent acts that gang members committed. The RYDS sample was divided into two groups: (1) gang members -- youth who reported being in a gang at some point prior to the end of high school, and (2) nonmembers -- youth who reported no involvement in gangs prior to the end of high school. Although only 30% of the RYDS sample were gang members, the results indicated that gang members were responsible for 65% of the delinquent acts, twice as many acts as one would expect, given their share in the population. Even more striking is the fact that gang members accounted for 86% of all the self-reported serious delinquent acts and 69% of all the violent acts.

The Los Angeles survey of youth living in high-risk areas found that a surprisingly low number of interviewed youth (8%) said they had been gang members at some point in their lives. However, the majority reported at least one of the following: being warned by parents about gangs in the neighborhood, frequent talk about gangs around the neighborhood, or frequent gang activity, and/or gang rivalries close by. In the Los Angeles survey, 36% of youth reported that there was pressure on neighborhood youth to join gangs. Of those who reported current or previous gang membership, 52% described themselves as a leader or one of the top people in the gang. This group reported that their first sustained contact with gang members was, on average, at the age of 12 and that they had become full members at the age of 13. Despite the low level of gang membership reported by the Los Angeles sample, 36% of the most recent violent offenses committed by these youth were against gang members, and 51% of the most recent victimizations experienced by the overall sample were committed by gang members.

Similarly, the DC survey found that only 15% of boys in the high-crime-rate neighborhoods had ever joined a gang. There was some greater proclivity to associate with gangs among more delinquent boys. Less than 9% of "good kids" and "fighters" had ever joined a gang while one-third of the robbery offenders said they had joined a gang. In addition, 30% of the drug dealers reported membership in a gang, at some point. However, as previous research found in other cities, gang membership in these DC neighborhoods was temporary and relatively short, with the typical interval of gang membership being 1 to 2 years.

Motivation Underlying Juvenile Homicides

Previous studies have examined the motives of juvenile perpetrators of homicides and have recognized a distinction between homicides related to crime and homicides related to conflict (Bailey, 1996; Cornell, 1990; Cornell, 1993; Loper and Cornell, 1996). For example, Loper and Cornell (1996) analyzed the FBI Supplemental Homicide Reports for 1984 and 1993 and observed that 59% of the homicides were committed by boys in the course of another criminal act (such as robbery or rape); 43% were committed in the course of conflict. Research by Cornell and colleagues indicated that juvenile males were more likely than both females and adult males to commit homicides in the course of another criminal act (Cornell, 1993; Loper and Cornell, 1996). Additionally, crime-related homicides appeared to be committed more often by youth having an extensive history of delinquent activity (Cornell, 1990).

Two of the OJJDP violence studies examined the motives behind juvenile homicides. The Milwaukee homicide study examined four primary types of homicide: gang-related, drug-related, robbery-related, and other argument-related. Of the 29 juvenile homicide offenders interviewed, 11 were involved in gang-related incidents, 5 in drug-related incidents, 7 in robbery-related incidents, and 6 in other argument-related incidents.

The Los Angeles homicide study examined the circumstances surrounding the homicide incidents involving juveniles, specifically the motives underlying the offenses. This study found that gang rivalry was the reported motive in 33% of the incidents and was the probable motive in an additional 15% of the incidents. Commission of another crime was reported as the motive in 14% of the incidents. A drug-related motive was reported in only 6% of incidents.

The study further examined the circumstances surrounding those homicides having different types of motives (e.g., gangs, drugs, other crimes, arguments, and other motives) to identify variation, if any, in the nature of the homicides. The results showed that gang-motivated homicides were distinctively different from homicides having other motives, in several ways. First, gang-motivated homicides were slightly more likely to occur during the late hours of the night. Second, gang-motivated homicides were more likely to take place in an open setting than were homicides having other motives. Third, nearly all gang-motivated homicides (and all drug-motivated homicides) involved firearms.

The Los Angeles homicide study found that 25% of the incidents involved an altercation that escalated in intensity. This suggests that there was direct victim involvement in the conflict in at least one-fourth of the incidents.


The pattern of juvenile violence appears to differ in DC and Los Angeles. In DC, most of the violence occurred in association with school, either on or near school premises, and usually in the afterschool hours. However, homicides involving juveniles in Los Angeles often occurred late at night, in a public place, and with the involvement of gang members. It is not known why such different patterns exist. The pattern of homicides involving juveniles in Los Angeles resembles the pattern of juvenile violence found in DC during the summer months -- with violence frequently occurring late at night rather than after school. Two possible factors may be the difference in weather patterns between the two cities and the existence of year-round schools in Los Angeles, which means that some unsupervised youth are out of school year-round.

Although few boys surveyed in high-risk neighborhoods in DC and Los Angeles were involved in gangs, the existence of gangs was well-known to the youth interviewed in both cities. Consistent with prior research, gang members in these communities self-reported higher levels of delinquency than nongang members.

Previous Contents Next

Juvenile Violence Research OJJDP Report to Congress