Youth Gangs and Crime
In this Summary, NYGC reports neither the total number of gang homicides nor a national estimate of gang homicides for 1998. There are many sources of error in gang homicide data. First, compiling national gang homicide data through surveys of law enforcement agencies involves asking agencies to provide NYGC with a service they may not routinely provide for local assessment and policymaking. These agencies compile crime data for investigative purposes, not for the purpose of preparing survey responses. Second, whether law enforcement agencies count only gang-motivated homicides (which grow out of a gang function) or gang-related homicides (in which a gang member need only be involved in some capacity) can make a big difference in the resulting homicide count. Third, gang definitions vary significantly among law enforcement agencies and State statutes. Fourth, law enforcement officials view of gang crime as a public safety concern can affect homicide reporting. A major reason for underreporting of gang homicide data is that many jurisdictions are unable or choose not to identify the offense as gang-related. For a variety of reasons, officials in some localities deny the existence of gang problems altogether; in others, public policy does not regard gang problems as a major concern. Finally, an analysis of the 199698 National Youth Gang Survey data revealed that a few large cities tend to report erratic numbers of gang homicides (Curry, Howell, and Maxson, forthcoming).
Taking these issues into account, Curry and colleagues analyzed gang homicides reported in the National Youth Gang Survey during 199698. The analysis, which was conducted for NYGC, will be published shortly as an OJJDP Fact Sheet (forthcoming). Part of the analysis is summarized here. Curry and colleagues included in their analysis cities with populations of 25,000 or more that reported gang problems and gang homicides over the 3-year period from 1996 to 1998. Remarkably, 99.1 percent of the cities with populations of 25,000 or more responded to one or more of the three surveys, and nearly three-fourths responded to all three surveys. (Large cities that reported erratic numbers of gang homicides were excluded from the analysis.)
The analysis found that more than one-half (55 percent) of the cities that experienced gang problems in at least 1 year during 199698 did not report any gang homicides (see table 28), but almost all (93 percent) of these were cities with populations less than 200,000. Of the cities that experienced gang problems in the 3-year period, nearly one-half (45 percent) reported a gang homicide in at least 1 year.
The overwhelming majority of the cities with populations of 25,000 or more that reported gang homicides in the 199698 period (88 percent) reported a maximum of 110 homicides in any given year of the 3-year period (table 28). Ten percent (45 cities) reported a maximum of 1150 homicides, and 2 percent (8 cities) reported a maximum of more than 50 homicides.
Among the 237 cities that reported a gang problem and provided a homicide statistic in all 3 years, 49 percent reported a decrease in gang homicides over the 3-year period, 15 percent stayed the same, and 36 percent reported an increase. In a comparison, Curry, Howell, and Maxson (forthcoming) found that among 408 cities that reported gang homicides (in several surveys) during both the early and middle 1990s, a decrease was observed in 32 percent of the cities and an increase was observed in 29 percent. The remainder (39 percent) stayed the same at both points in time. Thus, the gang homicide picture appears to have been more dynamic in the latter part of the decade than in the early to middle 1990s.
Since the first reports of youth gang activities were made, gangs typically have been associated with criminal acts (Sante, 1991). In 1998, survey respondents were asked to indicate the proportion of youth gang members who were involved in the following offenses in their jurisdiction: aggravated assault, robbery, burglary/breaking and entering, motor vehicle theft, larceny/theft, and drug sales. These crimes were selected because they are serious and/or violent crimes commonly associated with gang activity.
Readers are advised that this question was asked differently in the 1998 survey than in the 1996 and 1997 surveys. The 1998 question asked what proportion of gang members were involved in various crimes, producing a measure of the prevalence of gang member involvement in criminal activity. In the two prior surveys, respondents were asked to what degree gang members were involved in various crimes. Responses to that question provided a measure of the incidence of gang member criminal activity.
In 1998, a larger proportion of gang members were involved in drug sales than in any other criminal activity measured in the survey (see table 29). Slightly more than one-fourth (27 percent) of the respondents estimated that most/all of their gang members were involved in drug sales; the next highest level of involvement was in larceny/theft (17 percent), followed by burglary/breaking and entering (13 percent), aggravated assault (12 percent), motor vehicle theft (11 percent), and robbery (3 percent).
Although a 199798 comparison is not possible because the form of the question changed in 1998, it can be seen that the prevalence pattern reported in 1998 differs in several important respects from the incidence pattern reported in 1997. In 1997, survey respondents estimated that the incidence rate was highest for drug sales (29 percent), followed by larceny/theft (28 percent), aggravated assault (28 percent), motor vehicle theft (27 percent), burglary/breaking and entering (26 percent), and robbery (13 percent).
Comparison between the two surveys suggests that the 1998 prevalence and 1997 incidence rates are comparable for drug sales but that the 1997 incidence rates appear to be much higher than the 1998 prevalence rates for other offenses. The 1997 incidence rates are about twice the prevalence rates for all other offenses except robbery. In the case of robbery, the 1997 incidence rate is four times higher than the 1998 prevalence rate. This comparison suggests that a very small proportion of gang members are involved in robbery but that those who are involved are very actively involved.
Table 29 shows that prevalence of youth gang member involvement in crime varied by area type in 1998. When most/all and some responses are combined, this table indicates that gang member involvement in aggravated assault and robbery was highest in large cities. In the past, youth gang crime has been associated almost exclusively with large cities. However, in 1998, survey respondents in suburban counties reported the highest levels of gang involvement in motor vehicle theft, larceny/theft, and drug sales and also reported relatively high levels of gang member involvement in aggravated assault and robbery. Large cities and suburban counties reported far more gang member involvement in violent crimes than did small cities and rural counties.
Clearly, serious youth gang crimes are no longer concentrated only in the Nations large cities. Surprisingly, the highest proportion of gang members involved in burglary/breaking and entering was reported in rural counties, followed closely by suburban counties, then large cities. Gang members in small cities and rural counties also were reported to be extensively involved in drug sales and property crimes. Very small proportions of gang members were reported to be involved in robbery and aggravated assault in small cities and rural counties. More than 8 in 10 respondents in small cities and rural counties estimated that few or none of their gang members were involved in robbery. Low levels of involvement in small cities and rural counties were also reported for aggravated assault (6 in 10 respondents) and motor vehicle theft (7 in 10).
Table 29 shows a fairly high level of gang member involvement in serious and violent crimes in rural counties compared with small cities. Again, when most/ all and some responses are combined, a larger proportion of gang members in rural counties than small cities were reported to be involved in every offense except larceny/theft.
Table 30 shows the estimated proportion of youth gang members involved in crime in 1998, by region. Comparing the percentages in each region with the national average reveals some important regional differences. When most/all and some responses are combined, gang members in the Midwest were involved at a level above the national average in only one offense: drug sales. Somewhat lower levels of involvement in the Midwest than elsewhere were reported for aggravated assault and motor vehicle theft.
Gang members in the Northeast had much higher than average involvement in robbery and somewhat higher than average involvement in aggravated assault and drug sales. Their involvement was much lower than the national average for motor vehicle theft (lowest level of all regions) and larceny/theft.
Gang members in the South had much higher than average involvement in burglary/breaking and entering (highest level of all regions). Involvement in this region equaled or exceeded the national averages for all other offenses.
The West shows a different pattern. There, gang member involvement in burglary/breaking and entering and larceny/theft was similar to the national averages, but involvement was much higher in motor vehicle theft (highest level of all regions), somewhat higher in aggravated assault and robbery, and lower in drug sales (lowest level of all regions).
In sum, significant regional differences were reported with respect to the prevalence of gang member involvement in various serious and violent crimes. Northeastern gang members had higher than average involvement in robbery, aggravated assault, and drug sales. In the Midwest, gang member involvement was more than the national average in only one offense: drug sales. Gang members in the West were reported to have the highest level of involvement of all regions in motor vehicle theft and the lowest level of all regions in drug sales. Southern gang members were characterized by much higher than average involvement in burglary/breaking and entering (highest level of all regions) and high levels for all other offenses.