Involvement of youth gangs in drugs has been a major public concern since the crack cocaine epidemic began in the mid-1980's (Klein, 1995). Several issues have been debated and researched, including the extent of gang member involvement in drug sales (Decker and Van Winkle, 1994); the interrelationship of youth gangs, drugs, and crime (Howell and Decker, 1999); the geographical location of most gang drug distribution (Maxson, 1995); the race/ethnicity of gang members involved in drug sales (Esbensen and Winfree, in press); and the ability of youth gangs to manage drug trafficking operations (Moore, 1990).

Until now, national law enforcement data have not been available on the involvement of gang members in drug sales and the extent of gang control or management of drug distribution. The survey results reported below provide new information on these important issues.

Drug sales

On average, respondents estimated that 43 percent of the drug sales in their jurisdictions involved gang members. Averages among area types did not differ substantially, and the variation was not statistically significant. Respondents in suburban and rural counties reported that 46 percent of the drug sales in their jurisdictions involved gang members, followed by 43 percent for small cities and 41 percent for large cities (see figure 18).

Variation in gang member involvement in drug sales by region was found to be statistically significant. The average percentages of drug sales involving gang members in the four regions were as follows: 47 percent in the Midwest, 45 percent in the South, 41 percent in the Northeast, and 38 percent in the West (see figure 19).

The use of averages such as those reported above is somewhat misleading for two reasons. First, more than half (57 percent) of all respondents said their "youth gang" definition included drug gangs (see figure 6). As a result, many respondents included what might be considered purely drug gangs along with traditional youth gangs in their responses to the survey. Second, the responses showed a bimodal distribution; i.e., large proportions of respondents reported either low or high gang member involvement in drug sales in their jurisdictions.

For the analysis below, the proportion of drug sales involving gang members was divided into three levels: low (0-33 percent), moderate (34-66 percent), and high (67-100 percent). Nearly half (47 percent) of gang members were reported to be involved at the low level of drug sales, 26 percent at the moderate level, and 27 percent at the high level (see table 28).

Differences between area types in the level of gang member involvement in drug sales were not statistically significant. Nevertheless, table 28 reveals that the highest proportion of jurisdictions reporting high gang member involvement in drug sales was found in small cities (32 percent) and the lowest was found in large cities (25 percent). In addition, 49 percent of jurisdictions in large cities reported low involvement, while 42 percent of jurisdictions in rural counties reported low involvement.

Table 29 shows considerable variation in gang member involvement in drug sales by region at the low (0-33 percent), moderate (34-66 percent), and high (67-100 percent) levels. These variations were found to be statistically significant. Approximately 32 percent of jurisdictions in the Midwest and 29 percent in the South reported that their gang members were involved in high levels of drug sales. Only 26 percent of jurisdictions in the Northeast and 19 percent in the West reported that their gang members were involved at a high level. These data indicate that slightly more jurisdictions in the midwestern and southern regions reported gang member involvement in drug sales at a high level.

The West and the Northeast had the highest percentage of respondents reporting low gang member involvement in drug sales, with 54 and 51 percent, respectively. Thus, most of the gang member involvement in drug sales in these regions was low.

For jurisdictions that responded to the question regarding gang member involvement in drug sales, the estimated age of gang members based upon unweighted averages was as follows: 21 percent under age 15; 45 percent ages 15-17; 28 percent ages 18-24; and 5 percent over age 24 (see table 30). The average proportion of gang members in the 18-24 and the over 24 age ranges was slightly higher in jurisdictions in which gang member involvement in drug sales was reported at the moderate (34-66 percent) and high (67-100 percent) levels. The average proportion of gang members under 18 was slightly lower in jurisdictions that reported gang member involvement in drug sales at these higher levels. The variation in the ages of gang members by level of involvement in drug sales was found to be statistically significant for all age ranges except gang members under 15.

As noted earlier in this Summary, the race/ethnicity of gang members, based on unweighted averages, was as follows: African-American, 32 percent; Caucasian, 32 percent; Hispanic, 28 percent; Asian, 6 percent; and "other," 2 percent (see table 17). However, agencies that responded to the question regarding gang member involvement in drug sales (see table 31) indicated that gang members in their jurisdictions were 34 percent African-American, 29 percent Hispanic, 29 percent Caucasian, 6 percent Asian, and 2 percent "other."

Further examination of gang member involvement in drug sales by race/ethnicity showed that the highest average proportions of Caucasian, Hispanic, and Asian gang members were in jurisdictions reporting low gang member involvement in drug selling and that the highest average proportion of African-American gang members was in jurisdictions reporting high gang member involvement in drug selling. When the level of gang member involvement in drug sales was taken into account, the differences among the average proportions of African-American, Hispanic, Caucasian, and Asian gang members were statistically significant.

Drug distribution

Some respondents indicated that youth gangs control or manage a substantial proportion of drug distribution in their jurisdictions (see table 32). The estimated percentages for each level of gang control of drug distribution (excluding those who responded "do not know") were as follows: none, 6 percent; less than one-fourth, 41 percent; less than half, 23 percent; more than half, 29 percent; and all, 1 percent.⁷ Nearly half (47 percent) of the respondents who provided estimates said that gangs "control or manage" less than one-fourth of all drug distribution in their jurisdictions.

The level of gang control of drug distribution did not appear to vary substantially by area type. Approximately 31 percent of large cities and 30 percent of small cities and suburban counties indicated that gangs control or manage more than half or all of the drug distribution in their jurisdictions (see table 32). In contrast, 21 percent of the respondents in rural counties said that gangs control more than half or all of the drug distribution. The relationship between drug distribution and area type was not found to be statistically significant.

Regionally, there was somewhat more variation in the level of drug distribution controlled or managed by gangs in comparison with a variation by area type. The relationship between the level of gang control of drug distribution and the region was statistically significant. As table 33 shows, the proportion of respondents that reported that gangs control or manage more than half or all of the drug distribution in their jurisdictions was higher in the Midwest (36 percent) and Northeast (33 percent) than in the South (27 percent) or West (25 percent). Approximately 52 percent of the respondents in the West and 50 percent in the South said that gangs control either less than one-fourth or none of the illegal drug distribution, compared with 43 percent in the Midwest and 40 percent in the Northeast.

In jurisdictions that responded to the question regarding gang-controlled drug distribution, the estimated age of gang members, based on unweighted averages, was as follows: 21 percent under age 15; 46 percent ages 15-17; 28 percent ages 18-24; and 5 percent over age 24 (see table 34). Respondents who said gangs control all of the drug distribution reported a greater proportion of adult gang members (58 percent). In contrast, in jurisdictions in which respondents said gangs controlled none of the drug distribution, a much smaller proportion of adult gang members was reported (20 percent). Thus, the average proportion of adult gang members was higher in jurisdictions that said gangs controlled more of the drug distribution. The variation between levels of gang-controlled or gang-managed drug distribution was found to be statistically significant in all age ranges.

The ages of gang members involved at some level (all, more than half, less than half, or less than one-fourth) in drug distribution varied by population size. The highest average proportion of gang members involved in some drug distribution was in the 15-17 age range in all population categories. More generally, the average proportion of juvenile gang members involved in some drug distribution decreased as population size increased (see table 35). Conversely, the average proportion of adult gang members increased as population size increased. The variation in the ages of gang members involved in gang-controlled drug distribution was statistically significant when population was taken into account.

In jurisdictions that responded to the question regarding gang control or management of drug distribution, the race/ethnicity of gang members, as based on unweighted averages, was as follows: African-American, 33 percent; Caucasian, 30 percent; Hispanic, 29 percent; Asian, 6 percent; and "other," 2 percent (see table 36). As observed with gang member involvement in drug sales, the average proportion of African-American gang members was higher in jurisdictions that reported a greater degree of gang control of drug distribution; the proportion increased from 18 percent in jurisdictions that reported no gang control of drug distribution to 59 percent in jurisdictions that reported gang control of all drug distribution. All other racial/ethnic groups, except "other," were skewed toward low gang control of drug distribution.

In sum, the greater the gang control of drug distribution, the greater the average proportion of African-American gang members reported in those jurisdictions. The opposite was true for Hispanics, Asians, and Caucasians. Additionally, the variation in race/ethnicity of gang members was statistically significant when level of drug distribution was taken into account.