Why Black Teens Join Gangs

Journal of Blacks in Higher Education (Online) (Sep 15, 2014).

1. Full text
2. Abstract/Details
Hide highlighting
Abstract
TranslateAbstract

A new study led by Dexter R. Voison, a professor in the School of Social Service Administration at the University of Chicago,

identifies factors that lead to gang membership by young African American females.

Full Text
TranslateFull text
Turn on search term navigation

dexter_web_wborderA new study led by Dexter R. Voison, a professor in the School of Social Service Administration at the

University of Chicago, identifies factors that lead to gang membership by young African American females. The study also

finds that African American girls who are involved in gangs are at serious risk of substance abuse and other health risks.

The data showed that low self-esteem, emotional problems, trauma history, low parental monitoring, friends who

engage in risky behavior (e.g., skipping school, selling drugs, having sex), housing instability, and poor neighborhood quality

all contributed to a higher likelihood of these girls being involved in a gang. Those girls who did join a gang were more likely

to be involved in substance abuse, have casual sexual partners, and not test for HIV infection.

"Adolescents who have strained relationships with positive family and community members and have displaced housing may

find a sense of belonging with gangs," Professor Voisin said. "It may be that the gangssatisfy the need for social connections

and survival for these teens. At the same time, there are certain behaviors and norms within some gangs that are associated

with increased social and health risk factors for their members."

Professor Voison recommends that "child welfare and other youth service providers should target anti-ganginterventions for

African American females who report low self-esteem, emotional regulation problems, have housing instability and poor

relationships with their parents."

Professor Voisin received a bachelor's degree from St. Andrews University in Laurinburg, North Carolina. He holds a master

of social work degree from the University of Michigan and a Ph.D. from the Columbia University School of Social Work.

The study, "Correlates of Gang Involvement and Health-Related Factors Among African American Females With a Detention

History," was published in the journal Child Youth and Services Review. It may be accessed here.

* Email
*

Related: University of Chicago

Word count: 314

Copyright BruCon Publishing Company Sep 15, 2014

Why black teens join gangs. (2014). Journal of Blacks in Higher Education (Online), Retrieved from
http://search.proquest.com/docview/1624980647?accountid=37714

Gang involvement and membership among homeless and runaway

youth
Yoder, Kevin AAuthor Information ; Whitbeck, Les BAuthor Information ; Hoyt, Dan RAuthor Information . Youth
and Society 34.4 (Jun 2003): 441-467.

1. Abstract/Details
2. References 38
Hide highlighting
Abstract
TranslateAbstract

The present study documented the extent of gang involvement and gang membership in a sample of 602 homeless and

runaway youth from four midwestern states. The study also compared gang members, gang-involved youth who were not

members, and nongang youth on several dimensions including sociodemographic characteristics, family background, school

experiences, street experiences and exposure, emotional problems, alcohol and drug use, and other delinquent and deviant

behaviors. Findings indicated that a significant number of these youth were gang members (15.4% of the sample) or

involved in gangs (32.2% of the sample). Youth gang members and gang-involved youth reported more family legal

problems, had been suspended from school more, ran away at a younger age, used more alcohol and drugs, were exposed

to more deviant peers, and attempted suicide more than did nongang youth. In addition, youth gang members reported less

parental monitoring, more severe abuse, more street victimization, and more deviant subsistence strategies than did

either gang-involved or noninvolved youth.

Yoder, K. A., Whitbeck, L. B., & Hoyt, D. R. (2003). Gang involvement and membership among homeless and runaway youth. Youth
and Society, 34(4), 441-467. Retrieved from http://search.proquest.com/docview/226982269?accountid=37714

GANG MEMBERSHIP AND VIOLENT VICTIMIZATION*

Peterson, DanaAuthor Information ; Taylor, Terrance JAuthor Information ; Finn-Aage EsbensenAuthor Information
. Justice Quarterly : JQ 21.4 (Dec 2004): 793-815.

1. Full text
2. Full text - PDF
3. Abstract/Details
4. References 44
Hide highlighting
Abstract
TranslateAbstract
 Extant gang research supports an enhancement effect of membership on delinquency; that is, while delinquent youths may

be attracted to gangs, it is also true that gang membership increases delinquency among youths and that while delinquency

levels decrease after gang membership, they do not decrease to nongang levels. In this paper, we build on this research,

examining the relationship between youth gangmembership and violent victimization in a general sample of adolescents. We

find that gang member victimization rates are higher than nongang member rates, not only during membership, but before

and after as well. Thus an enhancement model of gang membership appears to best fit both offending and victimization

rates. This effect of gang affiliation on victimization goes beyond gang members' involvement in violent offending; violence

and gang status equate with cumulative disadvantage in terms of violent victimization. Additionally, contrary to gang youths'

perceptions, gangs appear to offer no protective value to gangmembers; we find no differences in violent victimization

between youths who joined gangs for protection and those who joined for other reasons, either before or after joining.

[PUBLICATION ABSTRACT]

Full text
TranslateFull text
Turn on search term navigation

Headnote

Extant gang research supports an enhancement effect of membership on delinquency; that is, while delinquent youths may

be attracted to gangs, it is also true that gang membership increases delinquency among youths and that while delinquency

levels decrease after gang membership, they do not decrease to nongang levels. In this paper, we build on this research,

examining the relationship between youth gangmembership and violent victimization in a general sample of adolescents. We

find that gang member victimization rates are higher than nongang member rates, not only during membership, but before

and after as well. Thus an enhancement model of gang membership appears to best fit both offending and victimization

rates. This effect of gang affiliation on victimization goes beyond gang members' involvement in violent offending; violence

and gang status equate with cumulative disadvantage in terms of violent victimization. Additionally, contrary to gang youths'

perceptions, gangs appear to offer no protective value to gangmembers; we find no differences in violent victimization

between youths who joined gangs for protection and those who joined for other reasons, either before or after joining.

Delinquent youth gangs have received substantial attention by scholars and the general public during the past 70 years.

Early works-such as those conducted by Thrasher (1927) in the 1920sfocused primarily on descriptive accounts of the

nature of gangs and gang activities. Since that time, however, examinations of delinquent youth gangs have

increased in complexity and their ability to influence public policy. Indeed, the need for studies of delinquent

youth gangs has increased substantially as the number of cities reporting gangproblems grew nearly tenfold between the

1970s and late 1990s (W. Miller, 2001).

The study of criminal victimization has a shorter-but informative-history in the field of criminology. Studies of groups at higher

risk of victimization and of factors related to victimization risk have been undertaken. One key finding from these studies is

that being involved in a delinquent lifestyle increases the risk of personal victimization (Esbensen & Huizinga, 1991;

Lauritsen, Sampson, & Laub, 1991; Loeber, Kalb, & Huizinga, 2001; Shaffer & Ruback, 2002).
With a few notable exceptions, however, the relationship between youth gang membership and risk

of adolescent victimization has rarely garnered researchers' attention. The exceptions include studies on the link

between gang membership and homicide victimization (Curry, Maxson, & Howell, 2001; Rosenfeld, Bray, & Egley, 1999),

the victimization experiences of female gang members (J. Miller, 1998, 2001; Miller & Brunson, 2000), and one study linking

the two (Miller & Decker, 2001).

This relative paucity of research is puzzling given what is known about the linkages between gangs and delinquency and

between delinquency and victimization. Research has consistently shown the connection between gang membership and

involvement in delinquent activities (Battin, Hill, Abbott, Catalano, & Hawkins, 1998; Bjerregaard & Smith, 1993; Esbensen &

Huizinga, 1993; Esbensen & Winfree, 1998; Huff, 1998; Thornberry, Krohn, Lizotte, & Chard-Wierschem, 1993). Research

has also demonstrated the connection between involvement in a delinquent lifestyle and being at increased victimization risk

(Esbensen & Huizinga, 1991; Lauritsen et al., 1991; Loeber, Kalb, & Huizinga, 2001; Shaffer & Ruback, 2002). Given these

connections, it may be logical to expect that gang members, who are more likely to be involved in a delinquent lifestyle,

would also be at increased risk of victimization. These linkages appear particularly important in light of many gang members'

expectations that gang membership will lessen the risk of victimization.

The literature in these areas begs two questions. First, what is the nature of the relationship between being ina gang and

being victimized? second, do youths who join gangs for protection achieve their objective? With respect to the first question,

two studies have reported on gang status and homicide victimization using police data (Curry, Maxson, & Howell, 2001;

Rosenfeld, Bray, & Egley, 1999), and two other studies have reported on this link using qualitative data (J. Miller, 1998,

2001; Miller & Brunson, 2000). The question remains, however, to what extent these studies can inform the nature

of gang status and victimization among adolescents in general. With respect to the second question, of how well

expectations of protection are met through gang membership, again, qualitative research reveals that both violent offending

and victimization are often normative features of gang life (see, e.g., Decker & Van Winkle, 1996), providing some evidence

for the conclusion that gangs do not protect their members from being victimized by others.

In this paper, we examine the relationship between gang membership, reasons for joining the gang(specifically, for

protection), and victimization rates across time. Data allowing exploration of these questions are relatively rare in the field of

criminology, but two aspects of the current study make such an exploration possible: longitudinal panel data and the

transitory nature of adolescent gang membership.

RELEVANT LITERATURE

Gangs, Delinquent Activities, and Victimization Research has consistently illustrated the higher levels of

involvement in delinquency of gang members relative to nongang members. Self-report studies in Rochester, Denver, and

Seattle have all found that gang members account for a disproportionate amount of crime, especially crimes of

violence. In Rochester, gang members comprised 31% of the sample, but committed approximately 69% of all violent

offenses and 82% of the more serious violent crimes of aggravated assault, robbery, and sexual assault (Thornberry, Krohn,
Lizotte, Smith, & Tobin, 2003, pp. 49-50). Similarly, only about 15% of the Denver Youth Survey sample was gang-involved,

but those youths committed approximately 79% of all serious violent offenses (Huizinga, 1997). In Seattle,

too, gang members accounted for 85% of all robberies, even though they comprised only 15% of the sample (Battin et al.,

1998).

But what can account for the higher levels of delinquency among gang members? According to Thornberry and his

colleagues (1993), explanations may be grouped into three categories: selection, facilitation, and enhancement. Selection

explanations propose that gang members are involved in high levels of delinquency prior to joining the gang, and that these

levels continue when they are active in the gang and after they leave it. The selection model leads to an expectation that an

underlying "criminal trait" is responsible for gangmembers being more highly involved than nongang

members in delinquency before, during, and after their involvement with the gang (see, e.g., Gottfredson & Hirschi, 1990).

Conversely, facilitation models suggest that the gang environment itself, rather than an individual trait, is the primary reason

for gang members' increased involvement in delinquency. Under the facilitation approach, levels of delinquency are low

before individuals join the gang, increase markedly as they enter the gang, and then drop back to pre-gang low levels after

they leave the gang. Unlike the selection model, the level of delinquent involvement exhibited by gang and nongang

members differs only when gang members are active in the gang. The final model, enhancement, may be viewed as a

combination of selection and facilitation. According to this explanation, gang members may exhibit extensive

involvement in delinquency before joining a gang, increase their delinquency even more upon entering the gang, and then

decrease their level of delinquency after exiting the gang. The enhancement approach suggests that gang members are

more extensively involved than nongang members in delinquent activities during all three periods, but the differences are

greatest during the periods of gang involvement.

Research examining these contrasting explanations of higher rates of offending for gang than nongang youths has provided

support for both the facilitation and enhancement models of offending. Thornberry and his colleagues (1993, 2003) found

evidence supporting a facilitation effect of gang membership on delinquent behavior in their study of Rochester youths.

Conversely, research conducted by Esbensen and Huizinga (1993) in Denver and by Hill and his colleagues

(1999) in Seattle found evidence for an enhancement effect. That is, while all three found evidence that gang membership

increased the level of youths' involvement in delinquent activities, Thornberry and colleagues found that gang and nongang

youths were not substantively different before becoming involved in gangs, while both Esbensen and Huizinga and Hill et al.

found that gangmembers were more involved in delinquency even before joining gangs and remained so after leaving

the gang.

Given the empirically supported relationships between victimization and delinquency and between delinquency

and gang membership, we propose that gang members will have higher rates of violent victimization than nongang youths

both during and after gang membership. The weight of the evidence also leads us to hypothesize that gang members will

have higher rates of victimization than nongang youths prior to joining their gangs as well.
One somewhat unexpected outcome from these earlier investigations into the effect of gang membership on illegal activity

was the finding that youths transition in and out of gangs rather quickly. Contrary to the popular conception that "once

you're in, you're in for life," the majority of gang members were found to be inthe gang for 1 year or less. In Rochester, for

example, 66% of female and 50% of male gang members were members for 1 year only; 28% of both gang girls

and gang boys were 2-year members; 4% of girls and 14% of boys were members for 3 years; and only 7% of boys (and no

girls) were members for 4 consecutive years (Thornberry et al., 2003, p. 39). This transitory nature of

youth gang membership allows for examination of the effect of gang membership on both offending and victimization rates.

Gang Membership and Victimization Risk

Providing further support for the hypothesis that gang membership should be positively related to victimization are recent

qualitative studies of the lives and experiences of youth gang members as well as research on the gang-drug connection.

One area of inquiry explores the relationship between gangmembership and violence. Involvement in violence is often an

important component of the gang lifestyle. Indeed, early research by Thrasher (1927) found that violence, ranging from "play

group roughhousing" to "gang warfare," was a defining attribute of gangs. According to Decker and Van Winkle (1996, p.

117), violence remains an integral part of gang members' lives that separates them from nongang members. Specifically,

violence "strengthens the bonds between existing members, increases the stake of prospective or fringe

members in the gang, and serves as a means by which nongang youth come to join the gang" (p. 68).

This violence may take a variety of forms, each potentially resulting in an increased likelihood of victimization

for gang members relative to their nongang peers. Members may be required to participate in violent initiation rituals when

entering or exiting a gang. Approximately two-thirds of the gang members in Decker and Van Winkle's sample, for example,

reported being "beat in" as part of their initiation process (p. 69). While the process varied from gang to gang, interviewees

often recounted a process whereby prospective members were expected to fight against several current gang members,

who were arranged in a line or a circle.

Gang members may also be subjected to harsh discipline from members of their own gangs for violating gang rules.

Padilla's (1995) research on one gang organized around drug sales uncovered the use of violence by the gang to sanction

members for violations of collective rules, referred to as "Vs" (p. 57). Although this work was focused on the activities of

the gang, rather than gang members, the process Padilla outlined is similar to one Decker and Van Winkle (1996) describe,

under which violators are expected to walk through a line of other gang members who take turns beating on the

transgressor.

Gang members are also involved in other types of activities that increase their vulnerability to predatory victimization by

others. Previous research has found, for example, that gang members are more likely than non-members to be

involved in drug selling activities (Esbensen & Winfree, 1998; Howell & Decker, 1999; Howell & Gleason, 1999; Huff, 1998;

Maxson, 1995). Padilla (1995) provides one such example in the Diamonds, a gang organized around street-level drug

dealing. Members of the Diamonds were responsible for moving marijuana and cocaine on the streets for higher-level
distributors. While others (e.g., Howell & Gleason, 1999; Klein, 1995; Maxson, 1995) debate the extent to which gangs are

primarily organized around drug-selling behavior or the extent to which gangs control drug sales in a city, extant research

leaves little doubt that gang members are often more involved in drug selling than their nongang peers, whether they are

acting for their own or the gangs' profit.

It is reasonable to conclude that the increased involvement of gang members in drug sales may translate into greater

likelihood of victimization. Jacobs (2000) has documented the increased risk of victimization for individuals involved in such

activities. While Jacobs' work did not examine gang members per se, his findings suggest factors relevant for both gang and

nongang individuals. For example, street-level drug dealers have features making them appealing potential targets of

robbers. They are often in possession of desired goods (e.g., drugs or money) and are less likely to report victimizations to

authorities. Even if they do report the victimizations, police may not be particularly inclined to follow up. Thus, the potential

payoff of robbing a dealer, coupled with the reduced risk of being apprehended and sanctioned, makes robbers view street-

level drug dealers as desirable targets.

Additionally, gang members may be targets of retaliation from rival gangs. Sanders (1994) suggests that an informal code

exists among gang members regarding drive-by shootings. Targeting members of rival gangsfor such shootings is an

acceptable practice, although shooting at "innocents" deliberately is generally prohibited. These processes make violence a

routine part of gang life (Decker & Van Winkle, 1996, p. 117), stark evidence of which is that of 99 gang members

interviewed in the early 1990s in Decker and Van Winkle's St. Louis study, 28 had died a violent death by mid-year 2003

(Decker, 2003).

While prior research has focused primarily on the role of violent offending by gang members, we suggest several

reasons why gang members should be at higher risk of victimization than their peers. To date, however, only one study has

examined victimization rates among gang and nongang youths in a general sample of adolescents: Savitz, Rosen, and Lalli

(1980) examined a sample of approximately 1,000 African American and white boys residing in Philadelphia. Their findings,

contrary to the ethnographic studies and those examining homicide rates, showed no statistically significant differences

between gang and nongang youths in terms of fear of victimization or actual victimization experiences. These findings,

however, may be tied to period or cohort effects, because the study was conducted long before the late 1980s' and early

1990s' elevations in the rates of youth violence and gang involvement.

The majority of these studies suggest differences between gang and nongang youths in terms of victimization experiences.

Specifically, research suggests that gang members are victimized more often than nongang members, despite frequent

claims by gang youths that they joined their gangs for protection. In our study, we further the inquiry into

youth gang membership and youths' victimization experiences with three questions. First, what is the nature of the

relationship between being in a gang and being victimized? second, do youths who are victimized seek out gangs for

protection, or does gang membership increase risk of victimization? And, third, do youths who join gangs for protection get

that protection?
CURRENT STUDY

Data to address these questions are drawn from two studies, one cross-sectional and one longitudinal, that were part of a

multisite evaluation of a school-based youth gang prevention program, the National Evaluation of the Gang Resistance

Education and Training (G.R.E.A.T.) program (see Esbensen & Osgood, 1999; Esbensen, Freng, Taylor, Peterson, &

Osgood, 2002; or Esbensen, Osgood, Taylor, Peterson, & Freng, 2001). The primary goal of the original studies was to

determine the effectiveness of the G.R.E.A.T. program. However, questions designed to elicit information about a wide

variety of relevant issues, such as school environment, peer and family relationships, delinquency and victimization, as well

as questions to allow for theory testing were also included.

The cross-sectional study was conducted in spring 1995; 5,935 8th-grade students in 42 public middle schools in 11 cities1

across the United States filled out anonymous, group-administered, self-report questionnaires. For the longitudinal study,

approximately 3,500 students (6th graders in one city, 7th graders in the other five) attending 22 public middle schools in 6

U.S. cities2 were selected to complete confidential, group-administered pre- and post-test and annual surveys between

1995 and 1999.

Sites for both studies were chosen based on three criteria: (1) the existence of a viable G.R.E.A.T. program, (2)

geographical and racial/ethnic diversity, and (3) cooperation of local law enforcement agencies and school districts. For

more detailed information on the procedures used for sampling and survey administration, consult Esbensen et al. (2001 or

2002).

In all sites in the longitudinal study, active parental consent3 was obtained prior to the year-one follow-up survey.4 A

modified Dillman (1978) method resulted in a total response rate of 67% (n = 2,045) from parents of the original 3,568 study

participants, with 57% of all parents allowing their child's continued participation inthe study.5

Sample Limitations

This public school-based sample has the usual limitations associated with school-based surveys (e.g., exclusion of private

school students; exclusion of truants, sick, and/or tardy students; and potential under-representation of "high-risk" youths).

We also acknowledge that attrition in our sample presents additional limitations because the youths of interest

(i.e., gang members) may be more likely to drop out of the sample over time. Our findings, therefore, can be considered a

conservative estimate of the relationship between gangmembership and violent victimization; that is, findings of a significant

relationship among self-reported gangmembers still attending school may actually underestimate the true relationship

between victimization and gang involvement, if "more serious" gang members tend to drop out of our sample.

The few other researchers who have examined similar questions about the relationship of gang status to gang member

behaviors and experiences using quantitative data have faced similar limitations. Thornberry and his colleagues (2003), for

example, in their analysis of data from the Rochester Youth Development Study, restricted many of their analyses to
male gang members, given the limited number of female gang members in their sample; additionally, some analyses

of gang membership and delinquency were conducted with respondents numbering as low as 10.

Our data, though limited, offer an opportunity to explore questions of gang membership and victimization that is not

present in many other extant data sets. With the cautionary note that some findings reported in this paper may be

considered exploratory, we present intriguing information that we hope can be the starting point for more detailed analyses

using larger samples of youth gang members.

MEASURES

Gang Definition

The debate over how to define gangs and how to determine who is a gang member has a long and ongoing history

(Bjerregaard, 2002; Decker & Kempf-Leonard, 1991; Esbensen, Winfree, He, & Taylor, 2001; Klein, 1995; Maxson & Klein,

1990; Winfree, Fuller, Vigil, & Mays, 1992). Recent work by Esbensen and his colleagues (Esbensen, Winfree, He, & Taylor,

2001) compared attitudes and behaviors of nongang and gang youths using several increasingly restrictive definitions

of gang status. While attitudes and behaviors became more antisocial as the gang definition moved from least restrictive

(self-definition only) to most restrictive (a core member of an organized, delinquent gang), the largest differences were found

between those who ever claimed gang status and those who never did. Based on these findings and to maximize

our gang member sample, we chose a single item ("are you now a gang member?") to identify gang members, recognizing

that by using this method, we may conservatively underestimate the level of victimization in the gang member sample.

Reasons for Joining the Gang

Respondents were presented a list of reasons commonly given by gang members for joining their gangs. Youths were asked

to indicate, by "selecting all that apply," whether they had joined their gangs for fun, for protection, for money, because a

friend was in the gang, because a sibling was in the gang, to get respect, to fit in better, or because they were forced to join.

They were also given the opportunity, in an open-ended response, to provide any additional reasons for joining (these

responses were few and are not presented here).

Violent Victimization

Violent victimization was measured by asking youths to report the number of times during the previous 6 months (previous

12 months in the Year 1 survey) they had been hit by someone trying to hurt them (simple assault), had someone use a

weapon or force to get money or things from them (robbery), and been attacked by someone with a weapon or by someone

trying to seriously hurt or kill them (aggravated assault).

Analyses were conducted using each of the individual victimization items and a composite violent victimization index

comprised of the three violent victimization experiences.6 The skewness of self-reported frequency data presents analysis
problems, but various approaches can be used to remedy this problem, including transforming the data using the natural log,

truncating at the 90th percentile (Nagin & Smith, 1990), or truncating the high-frequency responses according to some

conceptual reasoning. We chose to truncate scores on individual victimization items and the composite measure at 12. Our

premise is that experiencing these types of victimization numerous times in a 6- or 12-month period constitutes high-

frequency victimization. In support of this, frequency data indicated that few respondents experienced more than 12 of each

type of victimization during the designated period. By truncating at 12, we are able to examine these high-frequency

victimization experiences without sacrificing the detail of open-ended self-report techniques.

ANALYSES AND RESULTS

Sample Descriptions

Table 1 presents descriptive statistics for the cross-sectional and the longitudinal study samples. In both, males made up a

larger proportion of the gang than nongang youths, but females were still a fairly signficant proportion

of gang members in both samples (38% of the cross-sectional and 35% of the longitudinal samples). Gang members in both

samples were racially diverse, with whites comprising one-quarter of the cross-sectional gang member sample and almost

one-third of the longitudinal gang member sample.

Enlarge this image.

Table 1. Characteristics of Samples

The average age of students (8th-grade) in the cross-sectional sample was almost 14, while the longitudinal sample

averaged 12 years old during the first survey collection period (7th-grade in five sites, 6th-grade in the sixth) and 16 years

old during the final survey. In both samples, gang members were slightly older than nongang youths, and a greater

proportion of gangthan nongang youths reported living in single-parent homes.

Cross-sectional Sample Results

Prevalence rates lay the groundwork for the subsequent longitudinal analyses. We first examined a cross-section of 8th-

grade youths residing in11 U.S. cities. Using t-test of means comparisons of victimization rates for gang and nongang

youths, our findings reveal that gang members reported more extensive victimization experiences than nongang youths for

measures of assault, robbery, aggravated assault, and total victimization. The magnitude of these differences, however,

varies substantially with the type of victimization examined. For example, while a majority of all youths reported having

experienced some type of victimization in their lives (i.e., ever-prevalence), only 9% and 7% of nongang members reported

ever having been the victim of aggravated assault and robbery, respectively (see Figure 1). In comparison, 43%

of gang members reported having been the victim of aggravated assault and 25% reported having been the victim of

robbery. Smaller differences were found between the two groups in terms of simple assault (48% of nongang members,

66% of gang members) and total violent victimization (51% of nongang members, 75% of gang members).
These differences held when victimization was measured in terms of annual prevalence and annual frequency as well. As

shown in Figure 2, a much lower proportion of nongang than gang youths, for example, reported having been

victimized in the past year by assault (43% compared to 60%, respectively), robbery (7% vs. 21%), and aggravated assault

(8% compared to 38%).

While these descriptive data conclusively point to differences in victimization between gang and nongang youths, cross-

sectional data are inadequate for exploring the temporal relationships between victimization and gang membership. Does

victimization, for example, push a youth into gang membership (e.g., for protection)? Does gang membership expose a

youth to greater risk for victimization? Or does some combination of these two occur? To answer these questions, we turn

now to the longitudinal data that allow for correct temporal ordering in the analyses.

Enlarge this image.

Figure 2. Self-Reported Violent Victimization (Annual Prevalence) by Gang

Longitudinal Sample Results

Using data from the longitudinal study, we compare victimization rates of gang and nongang youths over time to test the

selection, social facilitation, and enhancement models that have been applied to the gang membership-delinquency

relationship. We then explore whether gang members' reasons for joining their gangs are related to their levels of

victimization.

Length of gang membership is an important factor in the feasibility of these analyses; in a 5-year study of adolescentsaged

12 to 17, for example, it would not be possible to examine how gang membership influences youths' victimization or how

victimization may be related to gangjoining if adolescent gang members tend to remain members throughout a majority or all

of their teen years. Consistent with the findings of other contemporary gang researchers using similar adolescent samples

(e.g., Thornberry et al., 2003), the majority of gang members (69%) in our sample reported being gang affiliated for one year

only (results not shown in table format).7 The next most consistent pattern was gang membership for two consecutive years

(22%). Very few youths (7%) were gang members for more than two consecutive years, and only one reported

membership in all five years of the study.

These findings indicate that our data allow our next sets of analyses, in which we examine the relationship

between gang membership and violent victimization by comparing victimization rates of gang and nongang youths across

the study period. In each of the five study years, the comparison is between youths who were gang members in the current

year and those who never were during the study's five years; victimization rates in the years preceding and following the

year of gang membership are reported.

Victimization of Gang and Nongang Youths

We used t-test of means comparisons to explore differences between gang and nongang youths' levels of violent

victimization in each of five years (see Table 2). Each row of Table 2 represents a comparison between

current gang members and nongang youths. In the entire first row of Table 2 (Year 1), the comparison is between youths

who were gang members in Year 1 and youths who were not (in that or any subsequent year). Each cell across that row is

the comparison of Year 1 gang members' to Year 1 nongang members' victimization rates in each study year. Thus, the cell

representing row "Year 1" and column 'Tear 5" compares the victimization rates in Year 5 (1999) of youths who reported

current gang membership or nongang membership in Year 1 (1995). Row two of the table represents the yearly victimization

rates of youths who were gang members in Year 2 compared to youths who were not in Year 2 (or any other year), and so

on. The means reported on the diagonal of the table represent the comparison of current gang members to nongang

members during the year of gang membership. Means in cells to the left and right of the diagonal are victimization

rates in the year(s) before and after gang membership.

Enlarge this image.

Table 2. Violent Victimization Rates of in

For youths who were gang members in Year 1, violent victimization rates were significantly higher than those for nongang

youths in that year and in each of the following years. Youths who were gangmembers in Year 2 also had significantly higher

victimization rates than nongang youths in the year prior to joining their gang, in the year of gang membership, and in the 3

years following gang membership. This same pattern is present for youths who were gang members in Years 3, 4, and

5: Gang members were victims of violence at a higher rate than were youths who were never gang affiliated in each of the

years preceding and in the year(s) following gangmembership, and they tended to experience the most victimization in the

year of membership. Figure 3 clearly depicts these relationships over the five study years.

Enlarge this image.

Figure 3. Violent Victimization Rates of gang

Consistent with prior findings regarding delinquent behavior, this group of findings provides evidence for an enhancement

effect of gang membership on victimization. Rates of violent victimization are higher for gang members than for nongang

youths both before and after their gang membership, and, as with offending patterns found in other research, it is (with the

exception of Year 5 gang members) during the period of gang membership that victimization rates are at their peak (for Year

2 gang members, victimization levels in Years 2 and 3 are nearly identical).

Given the ties between delinquency and victimization and the finding of an enhancement effect of gang membership on

offending, these results were expected. But what about the notion of protection some youths articulate as a reason for

joining their gangs? Are there differences among gang youths in levels of victimization and, if so, are these tied to youths'

reasons for joining gangs? We address these questions in the next section.

Victimization by Reasons for Joining Gangs

Gang members often purport to have joined their gangs for protection from various types of personal victimization. Here we

explore the possibility that those who join "for protection" have higher rates of victimization before joining and lower rates

once actually in their gangs compared to those who report joining for other reasons. If this hypothesis is correct, youths who

joined their gangs for protection should have experienced high rates of victimization prior to gang membership, propelling

them into gangs; if gangsindeed offer the protection these youths seek, then these youths' rates of victimization should drop

once they join their gangs.

A descriptive account of reasons gang members in this sample gave for joining is presented in Table 3. Reasons offered

appear fairly stable over time, with an interesting exception in Year 5, when only 28% reported joining their gangs for

protection. These percentages were much higher in all previous waves; in fact, approximately half of

all gang members in Years 1 through 4 cited protection as a reason they had joined their gangs.

Enlarge this image.

Table 3. Reasons for Joining Gang

To assess the extent to which reasons for gang joining were related to victimization rates, we compared gang members who

reported joining for protection to those who did not. Our results indicate that it appears not to be the case that youths

who join for protection have higher victimization rates prior to gang joining than youths who do not join for this reason, nor

are their victimization rates lower once in the gang compared to youths who join for other reasons (see Table 4).

As shown in Table 4, our t-test of means comparisons revealed no stable pattern of differences between these two groups

of gang members. Only one significant difference (at p < .10) between the groups was found on either the individual

victimization items (results not shown) or on the composite violent victimization measure: Of youths who

reported gang membership in Year 3, significant differences in victimization between those who joined for protection and

those who did not were found for Year 1.8

Enlarge this image.

Table 4. in

Delinquency, Gang Membership, and Victimization

Given the empirically supported relationship between delinquency involvement and victimization experiences, it is possible

that the relationships we have reported here between gang membership and victimization are conditioned by delinquency

involvement.9 That is, it may be the greater involvement of gang than nongang youths in delinquency, rather

than gang member status itself, that is the reason for their greater levels of victimization. We addressed this possibility in two

ways: first, we conducted a series of OLS regression analyses, regressing violent victimization on gang membership,

controlling for a composite measure of violent delinquency (the same items that appear in the violent victimization index).

Next, we grouped our sample into four categories to compare levels of victimization among them. We created the groups

using current gang membership (no, yes) and past 6-months' violent delinquency (no, yes); the resultant groups were
(1) gang-violent, (2) gang-nonviolent, (3) nongang-violent, and (4) nongang-nonviolent. Similar to Battin and her colleagues'

(1998) work on delinquency rates among groups of youth, we then examined violent victimization rates in each year for

these four groups using one-way Analysis of Variance (ANOVA). Bonferroni post-hoc tests were conducted to determine

whether significant differences invictimization were present between specific pairs of groups.

Regression analyses indicated that within each year of gang membership, gang member status had an independent main

effect on violent victimization, and these effects held when controlling for youths' involvement in violent delinquent behavior

(results not reported in table form). Standardized coefficients revealed that delinquency was the more powerful

predictor in each model (except in Year 3), but gangmembership remained a significant contributor to the explained

variance in violent victimization. In Year 3 (the year in which most of the sample had transitioned from middle to high

school), gang membership was the greater influence on victimization.

For the second set of analyses, victimization in each of the five years was examined for each year's current gang member-

violence grouping. Table 5 presents victimization means by group for each year and post-hoc significance tests

results. In each study year, the pattern of victimization across the groups is the same. Gangmembers who were also violent

offenders experienced the highest levels of violent victimization; next were nongang youths who were violent, followed

by gang members who were not violent, and, finally, by nongang, nonviolent youths. In the first two years, statistically

significant differences were found between all pairs except between gang-nonviolent and nongang-violent. In Years 3 and 4,

differences were significant between all pairs except gang-nonviolent/nongang-violent and gang-nonviolent/nongang-

nonviolent.10 It appears, thus, that both violence and gang membership are important contributors to youths' victimization,

but it is the combination of the two that is most dangerous.

Enlarge this image.

Table 5. Victimization by Gang

DISCUSSION AND CONCLUSION

Our data produced similar results about youth gang membership stability-or rather, the lack thereof-as reported in two other

longitudinal studies of adolescents, the Denver Youth Survey (Esbensen & Huizinga, 1993) and the Rochester Youth

Development Study (Thornberry et al., 1993): The majority of gang members in our sample reported membership for only

one year. Because youths move in and out of gangs quickly, we are able to make comparisons of victimization rates

of gang to nongang youths before, during, and after gang membership. Our 5-year panel data thus allow a test of whether a

selection, social facilitation, or enhancement model best fits the relationship between the period of youth gang membership

and youths' experiences of violent victimization (simple assault, aggravated assault, and robbery).

We found that an enhancement model best fits the pattern of the relationship between violent victimization

and gang membership for youths aged 12 to 16. Gang members had higher levels of victimization than nongang youths prior

to their gang involvement, their level of victimization generally was greatest during their gang membership, and, while
victimization decreased after youths left the gang, their levels of victimization remained significantly higher than those of

nongang youths (differences between gang and nongang youths' victimization rates were statistically significant in all years).

We were also interested in determining whether gang members who joined their gangs for protection (approximately half of

all gang members in each year but Year 5) actually received that protection, as evidenced by lower rates of victimization

after joining the gang. Victimization rates across the study years (including before, during, and after gang membership) were

similar among gang members regardless of their reasons for gang joining. Thus joining for protection does not appear to be

linked to higher rates of violent victimization prior to joining for youths who cite protection as a reason compared to

other gang youths, nor does joining for protection offer any clear reduction in subsequent violent victimization.

Prior work in this area has focused on the relationship between homicide victimization and gang affiliation and/or has been

based on qualitative research. The research reported here provides an additional dimension to this body of knowledge, by

replicating earlier work with quantitative analyses and by expanding the population to which these findings hold through the

use of a general sample of youths in public schools. We recognize the limitations that attrition poses in our sample and

accordingly offer our findings as exploratory but not without merit.

Our findings on gang membership's relationship to victimization provide strong evidence that prevention efforts are

necessary to lessen youths' experiences as victims (and as perpetrators) of violence. General prevention programs, offered

prior to middle school, may heighten adolescents' awareness of the increased victimization risks they would face

as gang members. Equally important, given the elevated rates of victimization among youths who join gangs compared to

those who do not, is the need to reduce violence against youths and to intervene promptly with youths who have been

victimized to provide alternatives to gang joining.

In conjunction with those reported by qualitative researchers, the findings reported here emphasize the need to dispel the

myth that gangs provide a safe haven. Gangs do not provide protection for their members. Infact, youths

who join gangs experience greater violent victimization while in their gangs than they do either before they join or after they

leave. Not only does this have implications for prevention of youth ganginvolvement, as stated above, but it also highlights

important junctures for assisting youths to leave gangs. Decker and Lauritsen (1996, p. 114) note that the experience of

violence was cited by current and ex-gangmembers in St. Louis as the top reason for gang members leaving the gang.

These authors suggest, then, that a critical intervention point is when gang members themselves or their friends or family

members have just been victimized by gang-related violence (for example, contacting youths in hospital emergency rooms

or police stations). If adolescents are most likely to be victimized by violence during gang membership, it is worthwhile to

confront them with this reality at the time of their victimization and to provide them with an avenue by which to leave

their gangs. Additionally, our finding that gang members' victimization rates remain elevated after they leave

the gang suggests that we should devote attention to assisting ex-gang members inmaking safe choices about their

behaviors and in avoiding what might be considered "residual victimization" connected to their prior gang member status.
Finally, the connection between victimization and delinquency cannot be ignored. The extant literature has established that

delinquents and victims are often one and the same. We too find violent delinquency to be associated with violent

victimization (for both nongang and gang youths); adding gang membership to the equation, however, exacerbates the

situation: Gang members who engage in violent offending experience higher levels of violent victimization than all other

youths, even nongang youths who engage in violence. This link suggests that general delinquency prevention and

intervention efforts can also serve to reduce both violent victimization and youth gang membership.

Footnote

Portions of this research were presented at the 2002 Annual Meeting of the Western Society of Criminology inSan Diego,

California and the 2003 Annual Meeting of the Academy of Criminal Justice Sciences in Boston, Massachusetts. This

research was supported under award #94-IJ-CX-0058 from the National Institute of Justice, Office of Justice Programs, U.S.

Department of Justice. Points of view in this document are those of the authors and do not necessarily represent the official

position of the U.S. Department of Justice. The authors extend their appreciation to Editor Donna Bishop and the

anonymous reviewers for their constructive comments.

1 Phoenix, Arizona; Torrance, California; Orlando, Florida; Pocatello, Idaho; Will County, Illinois; Kansas City, Missouri;

Omaha, Nebraska; Las Cruces, New Mexico; Philadelphia, Pennsylvania; Providence, Rhode Island; and Milwaukee,

Wisconsin.

2 Phoenix, Arizona; Lincoln and Omaha, Nebraska; Las Cruces, New Mexico; Portland, Oregon; Philadelphia, Pennsylvania.

3 Parents returned signed forms indicating whether or not their child could participate in the study, and only those students

whose parents had indicated that their child was allowed to participate were included in future data collection efforts.

4 Passive parental consent (i.e., parents returned consent forms only if they wanted their child excluded from the survey)

was obtained in five of the sites covering the pre-tests. In one of the sites, active parental consent was required prior to the

pre-test data collection efforts. Active parental consent was obtained prior to post-test data collection in a second site and

prior to the first year follow-up surveys in the four remaining sites.

5 This attrition occurred in spite of rigorous attempts by the researchers to gain parental consent. Active consent efforts

were based on a modified Dillman (1978) method and included three direct mailings to parents of the selected students, with

phone calls made after the second mailing to encourage parents to return the forms. Researchers collaborated with the

schools involved, having teachers distribute and collect forms intheir classrooms and offering incentives such as pizza

parties for any classroom in which 70% or more of the students returned their forms. Some teachers offered incentives of

their own, such as extended lunch periods or extra credit points. see Esbensen, Miller, Taylor, He, and Freng (1999) for a

discussion of the effects of attrition on the sample.

6 Results from the composite measure analyses are reported in this paper; for individual item analyses, similar results were

obtained.
7 Although the pattern held for the whole sample, the percentages reported here are based on analysis of the "complete

data" sample, those youths who had completed a questionnaire in each of the five study years (n = 965). Of these, 59

reported membership at some time during the 5-year period.

8 In another analysis, gang members who had ever reported joining their gangs for protection across the five study years

were compared to gang youths who never reported protection as a reason for gang joining. Again, only one significant

difference was found, and this was in Year 3 in which youths who had joined for protection had higher rates of victimization

than other gang members (results not shown in table format).

9 We thank the anonymous reviewer who queried us about this possibility.

10 Means are reported for all groups in Year 5, but because there were only five gang-violent youths, significance tests are

not reported for comparisons with this group.

References

REFERENCES

Battin, S. R., Hill, K. G., Abbott, R. D., Catalano, R. F., & Hawkins, J. D. (1998). The contribution of gangmembership to

delinquency beyond delinquent friends. Criminology, 36, 93-115.

Bjerregaard, B. (2002). Self definitions of gang membership and involvement in delinquent activities. Youth & Society, 34,

31-54.

Bjerregaard, B., & Smith, C. C. (1993). Gender differences in gang participation, delinquency, and substance abuse. Journal

of Quantitative Criminology, 4, 329-355.

Curry, G. D., Maxson, C. L., & Howell, J. C. (2001). Youth gang homicides in the 1990s. OJJDP Fact Sheet, March 2001

#03. Washington, DC: U.S. Department of Justice, Office of Justice Programs, Office of Juvenile Justice and Delinquency

Prevention.

Decker, S. H. (2003). Personal communication with first author, June 2.

Decker, S. H., & Kempf-Leonard, K. (1991). Constructing gangs: The social definition of youth activities. Criminal Justice

Policy Review, 5, 271-291.

Decker, S. H., & Lauritsen, J. L. (1996). Breaking the bonds of membership: Leaving the gang. In C. R. Huff

(Ed.), Gangs in America (2nd Ed.) pp. 103-122. Thousand Oaks, CA: Sage Publications.

Decker, S. H., & Van Winkle, B. (1996). Life in the gang: Family, friends, and violence. Cambridge: Cambridge University

Press.
Dillman, D. A. 1978. Mail and telephone surveys: The total design method. New York: John Wiley.

Esbensen, F.-A., Freng, A., Taylor, T. J., Peterson, D., & Osgood, D. W. (2002). Putting research into practice: The National

Evaluation of the Gang Resistance Education and Training (G.R.E.A.T.) Program. In W. L. Reed and S. H. Decker (Eds.),

Responding to gangs: Evaluation and, research (pp. 138-167). Washington, DC: U.S. Department of Justice, National

Institute of Justice.

Esbensen, F.-A., & Huizinga, D. (1991). Juvenile victimization and delinquency. Youth & Society, 23, 202-228.

Esbensen, F.-A., & Huizinga, D. (1993). Gangs, drugs, and delinquency in a survey of urban youth. Criminology, 31, 565-

589.

Esbensen, F.-A., Miller, M. H., Taylor, T. J., He, N., & Freng, A. (1999). Differential attrition rates and active parental

consent. Evaluation Review, 23, 316-335.

Esbensen, F.-A., & Osgood, D. W. (1999). Gang resistance education and training (GREAT): Results from the national

evaluation. Journal of Research in Crime and Delinquency, 36, 194-225.

Esbensen, F.-A., Osgood, D. W., Taylor, T. J., Peterson, D., & Freng, A. (2001). How great is G.R.E.A.T.?: Results from a

longitudinal quasi-experimental design. Criminology and Public Policy, 1, 87-118.

Esbensen, F.-A., & Winfree, L. T., Jr. (1998). Race and gender differences between gang and non-gang youth: Results from

a multisite survey. Justice Quarterly, 15, 505-526.

Esbensen, F.-A., Winfree, L. T., Jr., He, N., & Taylor, T. J. (2001). Youth gangs and definitional issues: When is

a gang a gang, and why does it matter? Crime & Delinquency, 47, 105-130.

Gottfredson, M. R. & Hirschi, T. (1990). A general theory of crime. Stanford, CA: Stanford University Press.

Hill, K. G., Howell, J. C., Hawkins, J. D., & Battin-Pearson, S. R. (1999). Childhood risk factors

for adolescentgang membership: Results from the Seattle Social Development Project. Journal of Research in Crime &

Delinquency, 36, 300-322.

Howell, J. C., & Decker, S. H. (1999). The youth gangs, drugs, and violence connection. OJJDP Juvenile Justice Bulletin.

Washington, DC: U.S. Department of Justice, Office of Justice Programs, Office of Juvenile Justice and Delinquency

Prevention.

Howell, J. C., & Gleason, D. K. (1999). Youth gang drug trafficking. OJJDP Juvenile Justice Bulletin. Washington, DC: U.S.

Department of Justice, Office of Justice Programs, Office of Juvenile Justice and Delinquency Prevention.
Huizinga, D. (1997). Gangs and the volume of crime. Paper presented at the Annual Meeting of the Western Society of

Criminology. Honolulu, Hawaii.

Huff, C. R. (1998). Comparing the criminal behavior of youth gangs and at-risk youths. NIJ Research in Brief. Washington,

DC: U.S. Department of Justice, National Institute of Justice.

Jacobs, B. A. (2000). Robbing drug dealers: Violence beyond the law. New York: Aldine de Gruyter.

Klein, M. W. (1995). The American street gang. New York: Oxford University Press.

Lauritsen, J. L., Sampson, R. J., & Laub, J. H. (1991). The link between offending and victimization among adolescents.

Criminology, 29, 265-292.

Loeber, R., Kalb, L., & Huizinga, D. (2001). Juvenile delinquency and serious injury victimization. Juvenile Justice Bulletin.

Washington, DC: U.S. Department of Justice, Office of Justice Programs, Office of Juvenile Justice and Delinquency

Prevention.

Maxson, C. L. (1995). Street gangs and drug sales in two suburban cities. NIJ Research in Brief. Washington, DC: U.S.

Department of Justice, Office of Justice Programs, National Institute of Justice.

Maxson, C. L., & Klein, M. W. (1990). Street gang violence: Twice as great or half as great? In C. R. Huff

(Ed.), Gangs in America (pp. 71-100). Newbury Park, CA: Sage Publications.

Miller, J. (1998). Gender and victimization risk among young women in gangs. Journal of Research in Crime and

Delinquency, 35, 429-453.

Miller, J. (2001). One of the guys: Girls, gangs and gender. New York: Oxford University Press.

Miller, J., & Brunson, R. K. (2000). Gender dynamics in youth gangs: A comparison of male and female accounts. Justice

Quarterly, 17, 801-830.

Miller, J., & Decker, S. H. (2001). Young women and gang violence: Gender, street offending, and violent

victimization in gangs. Justice Quarterly, 18, 115-140.

Miller, W. B. (2001). The growth of youth gang problems in the United States: 1970-98. Washington, DC: U.S. Department of

Justice, Office of Justice Programs, Office of Juvenile Justice and Delinquency Prevention.

Nagin, D. S., & Smith, D. A. (1990). Participation in and frequency of delinquent behavior: A test for structural differences.

Journal of Quantitative Criminology, 6, 335-365.

Padilla, F. (1995). The working gang. In M. W. Klein, C. L. Maxson, & J. Miller (Eds.), The modern gang reader (pp. 53-61).

Los Angeles: Roxbury.

 Rosenfeld, R., Bray, T. M., & Egley, A. (1999). Facilitating violence: A comparison of gang-motivated, gang-affiliated, and

nongang youth homicides. Journal of Quantitative Criminology, 15, 495-516.

Sanders, W. B. (1994). Gangbangs and drive-bys: Grounded culture and juvenile gang violence. New York: Aldine de

Gruyter.

Savitz, L., Rosen, L., & Lalli, M. (1980). Delinquency and gang membership as related to victimization. Victimology: An

International Journal, 5, 152-160.

Shaffer, J. N., & Ruback, R. B. (2002). Violent victimization as a risk factor for violent offending among juveniles. Juvenile

Justice Bulletin. Washington, DC: U.S. Department of Justice, Office of Justice Programs, Office of Juvenile Justice and

Delinquency Prevention.

Thornberry, T. P., Krohn, M. D., Lizotte, A. J., & Chard-Wierschem, D. (1993). The role of juvenile gangs infacilitating

delinquent behavior. Journal of Research in Crime and Delinquency, 30, 55-87.

Thornberry, T. P., Krohn, M. D., Lizotte, A. J., Smith, C. A., & Tobin, K. (2003). Gangs and delinquency indevelopmental

perspective. New York: Cambridge University Press.

Thrasher, F. M. (1927). The gang: A study of 1,313 gangs in Chicago. Chicago: University of Chicago Press.

Winfree, L. T., Jr., Fuller, K., Vigil, T., & Mays, G. L. (1992). The definition and measurement of 'gang status': Policy

implications for juvenile justice. Juvenile and Family Court Journal, 43, 29-37.

Peterson, D., Taylor, T. J., & Finn-Aage Esbensen. (2004). GANG MEMBERSHIP AND VIOLENT VICTIMIZATION*. Justice Quarterly :
JQ, 21(4), 793-815. Retrieved from http://search.proquest.com/docview/228204314?accountid=37714

Criminal business organizations, street gangs and 'wanna-be' groups:

A Vancouver perspective
Gordon, Robert MAuthor Information . Canadian Journal of Criminology 42.1 (Jan 2000): 39-60.

1. Full text
2. Full text - PDF
3. Abstract/Details
Hide highlighting
Abstract
TranslateAbstract

The reasons why individuals become involved with organizations, gangs, and groups include economic and ethnic

marginality, material gain, the attraction of supportive peer groups, and flight from abusive family circumstances. Policy and

program responses must recognize the differences between organizations, gangs, and groups, and the different factors

underpinning initial and continued membership.

Full Text
TranslateFull text
 Turn on search term navigation

Headnote

Criminal business organizations, street gangs and 'wanna-be' groups: A Vancouver perspective(1)

Headnote

Le Greater Vancouver Gang Study (Etude des gangs du Vancouver metropolitain) comporte une recherche detaillee sur 128

membres de gangs. Cette etude a pour objectif d'amasser des donnees (ex., age et origine ethnique); de verifier l'exactitude

des classifications et des definitions des gangs et des groupes; de determiner pourquoi des individus sejoignent et

des gangs. La recherche a concentre son attention sur trois groupes principaux., les associations criminelles d'affaires;

les gangs de rue et les groupes de "wanna-be". les raisons pourquoi it y a des vagues d'activites de gangs et de groupes

demeurent diffIciles et preciser, mais la recherche a pu identifier les principales raisons pourquoi des particuliers se melent

aux associations criminelles, aux gangs et aux autres groupes. Ces raisons sont: la marginalite economique et ethnique,

l'acquisition de biens, les appuis qu'on trouve aupres des membres d'un groupe, et la fuite devant les situations familiales

violentes. Pour que des politiques et des programmes puissent etre efficaces, its doivent faire la difference entre les

associations criminelles, les gangs de rue et les autres groupes marginaux et distinguer quels sont les facteurs qui sous-

tendent l'adhesion initiate au groupe et la continuation de cette participation.

Headnote

The Greater Vancouver Gang Study involved research on and with 128 known gang members. The goals included

developing profile data (e,g., age and ethnicity); testing classifications and definitions of gangs and groups; and

determining why individuals become involved with gangs. The research focussed upon three main groups: criminal business

organizations; street gangs; and wanna-be groups. The causes of the waves of street gang and wanna-be group

activity in Vancouver remain elusive, but the research identifies the main reasons why individuals become involved with

organizations, gangs, and groups. These reasons include economic and ethnic marginality, material gain, the attraction of

supportive peer groups, and flight from abusive family circumstances.

Headnote

To be effective, policy and programme responses must recognize the differences between organizations, gangs, and

groups, and the different factors underpinning initial and continued membership.

Introduction

Those interested in the "gang" phenomenon in Canada are frequently surprised by the absence of a body of research that is

distinctively Canadian and publicly available. The first recorded piece of work was a study of

juveniles in street gangs in Toronto in the 1940's undertaken by Rogers (Rogers 1945), and there is an occasional reference

to street gangs and youth gangs in other publications (e.g., Ley 1975). Some research was also conducted by Joe and

Robinson on street gangs in the "Chinatown" area of Vancouver in the late 1970's (Joe and Robinson 1980). Apart from this,

no organized and systematic research was undertaken, and nothing was published, until quite recently.
The same appears to be true for the research and literature on criminal organizations. Beare's (1996) recently published

book on organized crime (or criminal enterprises) in Canada is the first major contribution to the field. A host of important

definitional, policy, and legislative issues are canvassed (especially with respect to money laundering) and help pave the

way for new empirical research. Dubro's (1985; 1992) journalistic accounts of the Mafia, and of Asian organized

crime, in Canada are illuminating, and the nature and form of Canadian organized crime is usefully clarified in a recently

released report of a pilot survey of organized criminal activity conducted by the Canadian Centre for Justice Statistics

(1999). Wolf's (1991) ethnography of biker gangs in Alberta still stands as the only insider's account of what, for some

analysts, is a criminal organization (e.g., Beare 1996; Canadian Centre for Justice Statistics 1999) but which, in Wolf's view,

is a subculture of rebels - a brotherhood of outlaws - for whom crime is a casual rather than entrenched activity. It is clear

that the debate over the nature of biker gangs will continue for some time and that an accurate picture is often blurred by

regional and other variations. As Wolf (1991: 268) points out, there are tremendous differences between clubs and between

chapters, and those involved in the "full gamut of crime remain a distinct minority". This is supported by one of the

findings in the report of the Centre for Justice Statistics (1999: 19) which acknowledges that "among all the motorcycle

associations, only 1% fall into the underworld of organized crime".

Several pieces of Canadian gang research were coincidentally released in 1993. Three reports, each written for either the

federal or the provincial levels of government (Fasiolo and Leckie 1993; Gordon 1993; Mathews 1993), and an unpublished

M.A. thesis (Young 1993) examined different aspects of the gang phenomenon. Inaddition, Kennedy and Baron (1993) and

Baron and Tindall (1993) published their studies of groups of street youth in part to understand better an apparent

increase in youth violence. All of this work provided not only the answers to some pressing policy questions (e.g., to what

extent are Canadian street gangs dominated by the members of visible ethnic minorities?) but also the building blocks for

more comprehensive research such as the Greater Vancouver Gang Study.

Fasilio and Leckie's work (1993) on the coverage of gangs and gang activity by the Canadian media made an extremely

valuable contribution to what was, at the time, a lively debate over news media amplification of the nation's

street gang "problem." The authors undertook a content analysis of major national news magazines and the major daily

newspapers published in urban centres such as Vancouver, Winnipeg, Toronto, and Montreal between July and October

1992. They found that the media characterized gangs and gang activity as widespread, a significant threat to society, and as

a relatively new phenomenon. Gangs were depicted as a subject of "growing social concern" and the product of an ailing

society; themes which, according to Young (1993), have been trumpeted by Canadian newspapers during every wave of

urban street gang activity since at least 1945.

Fasilio and Leckie noted the absence of any historical reference or perspective in the stories and a focus that accentuated

polarization along ethnic lines: "Asian gangs" were a particular menace. One consequence, the authors argued, was an

unjustified increase in the fear of gangs and gang activity, and a similarly unwarranted concern over the extent to which

immigrants who were members of visible ethnic minorities were responsible for gang related crimes. The
reasons why gangs emerged were largely ignored apart from some vague references to the consequences of immigration

policies and practices. Generally, in the authors' view, the media were contributing to a moral panic.

Gordon's (1993) preliminary study of gang members in British Columbia correctional centres provided something of a

counterfoil to the news media's pre-occupation with "Asian gangs" and "youth gangs." The files of 41 male inmates (adult

and youth) who were identified as active gang members by the police and corrections personnel were examined and the

inmates were asked to participate in interviews. Ten adults and 24 young offenders were interviewed, the goal being to

obtain information about the activities and membership of gangs in the province (especially the Greater Vancouver area)

and the reasons why young people, inparticular, became involved with gangs. A secondary objective was to design and test

some research definitions and instruments for use in a later study (i.e., the Greater Vancouver Gang Study). The majority of

incarcerated gang members were young adult males (the mean age was 19), rather than young offenders. This finding was

of importance in combating the false perceptions created by the news media's use of the term "youth gangs." Also of

importance was the finding that the largest single group of incarcerated gang members were individuals of European ethnic

origin born in Canada (40 percent). The next largest group were individuals of Asian ethnic origin (34 per cent), the majority

having been born in Canada. There were very few gang members of East Indian, aboriginal, and Hispanic origin, and only

one black gang member in the sample. Overall, 32 per cent of incarcerated gang members were not born in Canada; a

proportion higher than the proportion of immigrants in the provincial population and, therefore, the expected percentage (22

percent). This finding lent support to an hypothesized relationship between immigration and gang membership. As Gordon

and Nelson (2000) have pointed out, however, the actual numbers of gang members in the study who were not

born in Canada (14) is extremely small, especially when compared with the size of the immigrant population in the province

as a whole (approximately 750,000 people).

The gang members who were interviewed indicated that they were not coerced or otherwise pressured into

joining gangs and that the process of becoming involved was a gradual, rather than abrupt event. There was a slow drift

towards involvement that was often initiated by a close relative or a friend who was already a member of a gang or who

knew a gang member. A prospective member would spend time on the periphery of the group but might eventually be

drawn In, undergo an initiation ceremony, and then become a full member. As Gordon (1993; 1995) points out, the process

was neither strange nor surprising and was similar to the experiences of a young person joining a legitimate "gang" such as

the boy scouts or a baseball team.

According to Gordon (1993), the availability of choices is a key to understanding a person's involvement ingangs. If an

individual has no access to, or is not encouraged to join, a mainstream group, an "illegitimate" group may be chosen

instead, an hypothesis reflecting the seminal work of Cloward and Ohlin (1960) that was supported, in part, by the research

data. Nearly one half of the subjects had backgrounds filled with domestic problems including drug, alcohol, and physical

abuse in the family coupled with school records of truancy, fighting, suspensions, and expulsions. The picture was one of

troubled individuals from troubled backgrounds. When these problems were related to the reasons why the subjects became

involved with gangs (as stated ininterviews), it was clear that the choice of an illegitimate, rather than legitimate, group was a
function of the absence of alternatives. The reasons why other subjects became involved with gangs were not so clear.

Some became involved primarily to continue associations with friends, to make money, and for the relief of boredom,

considerations that could not be satisfied by membership in a legitimate "gang".

Mathews' (1993) preliminary research in Toronto also involved interviews with youth gang members. The study was

intended to allow young people a voice in the emerging debate on gangs and groups. Youth involved in gang activity were

approached by experienced social services and police personnel and asked to participate in the study. Twelve young people

aged from 14 to 21 years agreed and were interviewed. Interviews were also conducted with social workers, police officers,

school officials, the parents of gangmembers, and some victims of gang activity.

Mathews recognized that the term "gang" can be misleading and that the term "gang/group" is a better way of describing the

different kinds of gatherings of young people. Analysts and policy-makers should view the phenomenon along a continuum

ranging from groups of friends who spend time together and who occasionally get into trouble, to more serious, organized

criminal groups or gangs. Mathews also developed an exploratory "gang/ group involvement cycle" and a multivariate causal

model. The model accounts for the process of joining, remaining with, and leaving a gang by reference to several variables

including individual vulnerability factors that can drive a young person into a gang/group (e.g., an abusive family

environment), situational factors, the larger social context, and the nature of the responses to the youth's behaviour.

Young's (1993) study of the history of gangs in Vancouver filled a significant gap in Canadian gang research by providing

the first socio-historical analysis of youth gang activity. Young scrutinized each daily edition of a major Vancouver

newspaper for an 85-year period and, when articles reporting gang activity were found, these were cross checked against

stories in other newspapers. Young argues that the first significant wave of street gang activity occurred between 1948 and

1959: the era of the "Hoodlum Gangs". These street gangs included the "Alma Dukes" and the "Vic Gang", and the

characteristics and activities of these gangs were similar to those of street gangs in more recent times. The second wave of

street gang activity occurred between 1970 and 1975 and consisted of two, seemingly unrelated clusters of gangs: the

"Park Gangs" (e.g., the Riley Park Gang and the Clark Park Gang), and the "Chinatown Gangs" discussed by Joe and

Robinson (1980). The Chinatown gangs of the early 1970's included some gangs that may have eventually evolved into

criminal business organizations. The third, and most recent, wave of street gang activity began in 1985 and began to

dissipate in the mid-1990's.

Young attempted to explain the wave like pattern during the post 1945 period by focusing on two independent variables:

inward migration; and unemployment rates. A preliminary analysis of these variables indicated a slight positive relationship

between inward migration rates and reported fluctuations in gang activity. Unemployment rates and

fluctuations in gang activity were inversely related but Young cautions against the use of these results, beyond the

development of hypotheses for future research.

The Greater Vancouver Gang Study

Method

The Greater Vancouver Gang Study was designed to build upon the existing but limited pool of knowledge about gangs and

similar groups in Canada, and to inform policy and practice in British Columbia. The Study had four main goals: to develop

profile data on known gang members (e.g., age, gender, and ethnicity); to test the validity of the classifications and

definitions of gangs and groups proposed and discussed in preliminary studies, and to determine why individuals became

involved with and left gangs. This third goal included identifying the processes of involvement and disengagement, with a

particular emphasis on the psychological and sociological factors that propel individuals into gangs. The fourth goal was to

examine the organization, composition, and activities of gangs, from the perspectives of those involved with, and those

working with, gangs.

The research subjects were all adults and youth on the caseloads of corrections personnel in the Greater Vancouver

area in January 1995, and those who were added to their caseloads during the following six months, who were identified by

these personnel as being involved with "gangs". The group included individuals sentenced to terms of imprisonment and

probation or those undertaking other forms of supervised community service. The main portion of the research took place

during the first six months of 1995 and involved reviewing and extracting information from client files, interviewing clients

who were willing to participate, and discussing individual cases with probation officers and police officers. The purpose was

to obtain as many perspectives on a client as possible. A total of 128 subjects were included in the research and 33 agreed

to be interviewed, in some cases with the assistance of a social worker or probation officer who spoke the subject's first

language. The interviews were conducted in probation offices, "natural settings" in the community (e.g., coffee shops),

and in correctional centres. In addition, there were 11 family visits and interviews each conducted by a qualified and

experienced family therapist assisted by an experienced social worker who spoke the languages of the families.

Although the subjects were the entire population of "gang members" known to the provincial Corrections Branch at the time,

they were not the entire population of gang members in the Vancouver area. The extent to which the subjects are

representative of the larger gang population is impossible to determine (no "gangcensus" was available), and the proportion

of the larger gang population represented by the research subjects is similarly impossible to ascertain. The estimated

numbers of gangs and gang members varies considerably at any given moment, and from year to year, and even the best

estimates offered by authoritative sources such as the police are only ever educated guesses. Communication breakdowns,

rivalries, and territoriality affecting the police departments in the Greater Vancouver area also means that no single police

unit has a comprehensive overview of the situation.

The status of the gang members involved in the research could not be determined with certainty. Police and probation

officers offered their best opinions but it was not always clear whether a subject was a core member of a gang, an associate,

or a peripheral (or wanna-be) member. In addition, the gang members involved in the research were those who had been

caught, charged, and convicted, and this must be borne in mind when interpreting the results. Further challenges were

posed by the perennial problem of defining gangs.

Defining "gangs"

One of the initial problems for the Study was finding an accurate and widely accepted definition of a "gang", and a

"gang member". Similar difficulties are reported consistently in research reports and other literature on gangs, regardless of

the location and objectives of the research, and the methods used by the researchers (e.g., Goldstein and Huff 1993;

Mathews 1993; Klein, Maxson, and Miller 1995; Spergel 1995; Curry and Decker 1998). A valid and reliable answer to the

question "what is a gang?" has been elusive and depends upon who is asked. Different police departments and individual

police officers have different conceptions and definitions (formal and informal), as do probation officers, corrections centre

staff, researchers, policymakers, and the members of "gangs" and criminal groups.

As the research proceeded, it became clear that small groups of offenders were being referred to as "gangs" when the

members of these groups did not see themselves that way. Perhaps predictably, the primary offenders seemed to be the

news media who used a range of inaccurate terms to describe groups (e.g., Asian gangs, and youth gangs), thereby

generally distorting and amplifying the "gang" problem. In one case, a daily newspaper actually constructed a name for a

small group of young offenders who, for several weeks, had been bullying high school students into parting with cash and

possessions, and then began to refer to this group as a "gang". The name given by the newspaper the "Back Alley Boys" -

reflected both the chosen scene of crime (alleyways at the rear of houses) and the membership of the group (teen-age

males), but it was an imposed name.

Another example of the problem of construction involved both the media and the police. The so-called "626 gang" was a

group of eight youth and young adults who committed a series of armed robberies of banks, stores, and credit unions in the

Greater Vancouver area, over a four month period in 1992. The group was named the "626 gang" by the police because

they consistently used stolen Mazda 626 automobiles to drive to and escape from the scenes of their robberies. When

members of the "gang" were interviewed it became clear that they had not chosen the "gang@ name and did not see

themselves as a "gang"; they were simply a group drawn together by friendship and a common interest in crime that decided

to try its luck at armed robbery.(2)

Attaining a common and accurate understanding of a phenomenon (or phenomena) being studied, and about which policy

and programming are being developed, is important. In the absence of common and widely accepted conceptions and

definitions research, policy development, program development, and even police operations, both within and between

jurisdictions, will be confused and confusing. The Greater Vancouver Gang Study took a grounded approach to the

development of gang classifications and definitions. A typology was developed and tested prior to the main research and a

great deal of time and effort was expended on this component. Six main types of groups that seemed to attract the label

"gang" were identified: youth movements; youth groups; criminal groups; wanna-be groups; street gangs; and criminal

business organizations.(3) The research concentrated on the last three types.

Criminal business organizations are organized groups that exhibit a formal structure and a high degree of sophistication.

Organizations are comprised primarily of adults, including older adults. They engage incriminal activity primarily for
economic reasons and almost invariably maintain a low profile, which is a characteristic that distinguishes them most clearly

from street gangs. Organizations may have a name, examples being Lotus, The Flying Dragons, and The Big Circle Boys

(Dai Huen Jai). There were 24 members of criminal business organizations in the research, representing both the

organizations named above and three other organizations.

Street gangs are groups of young people, mainly young adults, who band together to form a semi-structured organization

the primary purpose of which is to engage in planned and profitable criminal behaviour or organized violence against rival

street gangs. They can be distinguished from other groups (especially wanna-be groups) by a self-perception of the group

as a gang; a name selected and used by gang members; and some kind of identifying marks such as clothing and colours.

The members will openly acknowledge gangmembership because they want to be seen as gang members by others.

Street gangs will tend to be less visible but more structured, better organized, and more permanent than wanna-be groups.

Examples include Persian Pride, Midnight Rockers, and the quintessential Vancouver street gang of the 1980's and 1990's:

Los DiabLos. There were 35 members of street gangs in the research, representing both the gangs named above and ten

other gangs.

Wanna-be groups are clusters of young people who band together in a loosely structured group to engage inspontaneous

social activity and exciting, impulsive, criminal activity including collective violence against other groups of youths. A wanna-

be group will be highly visible and its members will openly acknowledge their "gang" involvement because they want to be

seen by others as gang members. The group will have a local gathering area and a name, selected and used by its

members, which may be a modified version of the name of either a local or an American street gang. The group may use

clothing, colours, or some other kind of identifying marks. The group's name, meeting ground, and colours may fluctuate.

Examples include a suburban wanna-be group known as Los Cholos that was active in Greater Vancouver during the early

1990's. There were 25 members of wanna-be groups in the research, representing Los Cholos and six other groups.

The categories and definitions are not perfect, partly because of inter-relationships between the groups that tend to blur

boundaries. Some street gangs were allied to criminal business organizations, one example being the link between the

street gang known as Los Diablos and the criminal business organization known as Lotus (Gordon 1998). During the late

1980's and early 1990's, Los Diablos acted as the drug retail and enforcement arm of Lotus, some key members of Los

Diablos also being Lotus associates. Some criminal groups are affiliated with criminal business organizations, many of which

have a cell-like (rather than pyramidal) organizational structure, each cell being potentially independent. Some wanna-be

groups may have been created (in part) by former street gang members and may exhibit more gang-like characteristics than

other wanna-be groups. Despite these overlaps and connections, the typology and its associated definitions reflected a

reality that was recognized and endorsed by a variety of criminal justice practitioners and policymakers.

Results: Why individuals become involved with organizations, gangs and groups
There is no single and simple answer to the frequently posed question, "why do individuals become involved with gangs?"

This, in part, is because the question itself is not terribly useful. It makes more sense to ask why people become involved

with criminal business organizations, or with street gangs, or with wanna-be groups. The answers differ.

In the case of criminal business organizations, answers begin to appear once the basic characteristics of their memberships

are examined. Organization members included in the Study tended to be older males (mean age 27.8 years) who were

better educated than the members of street gangs. They were less likely to be economically dis-advantaged, primarily

because of the lucrative nature of their work with an organization. Criminal business organizations engage primarily in the

supply of illegal goods and services (e.g., drugs), areas of commerce in which significant profits are made. Organization

members were more likely to have migrated to Canada than, for example, street gang members (21 out of the 24), and more

likely to be members of a visible ethnic minority (21 out of the 24). Sixteen of the 24 were of Vietnamese or Chinese ethnic

origin.

Members of criminal business organizations and the corrections personnel responsible for their cases (e.g., probation

officers) indicated that there was a social and cultural bond which attracted individuals to organizations and that addressed

their sense of ethnic and cultural marginality in a predominantly EuroCanadian environment. The importance of a shared

language and a sense of belonging accounted for their continued involvement with organizations. Members consistently

stated that, on arriving in Canada, there was a lack of resources and employment opportunities available to them. They had

few marketable skills, and experienced significant difficulties in obtaining rewarding, legitimate employment because of

significant language barriers. In order to overcome their economic marginality and achieve the economic status valued by

the larger Canadian society, organization members were drawn into illegal activities. In short, and perhaps not surprisingly,

membership of criminal business organizations (in Greater Vancouver) is ethnically shaped and meets the economic and

social needs of both organization members and their families.

Particularly telling were the responses given to questions about leaving (as opposed to joining) criminal business

organizations. Members were asked about the process of leaving, and any associated problems such as fear of retaliation.

Two subjects stated that they had considered leaving but could not afford to do so, and the loss of significant income was a

factor for most individuals. Organization members reported earning from $2,500 per week to up to $30,000 per month (tax-

free) and members, police officers and probation officers all agreed that membership of a criminal business organization

was a lucrative endeavour. The majority of subjects had minimal work skills and could not expect to earn a comparable

income in a legitimate field. As one organization member put it, "what else could I do and make this kind of cash... I speak

little English, and legal work means no money" (Gordon and Foley 1998: 34).

Understanding why individuals became involved with street gangs proved to be less straightforward. Thirty-five

street gang members were included in the research (only one was female) and they tended to be much younger than the

members of criminal business organizations (the mean age of street gang members was 18 years). Education levels were

lower and they were more likely to be from economically disadvantaged backgrounds. More than 60 percent were
born in Canada, although 85 percent were members of visible ethnic minorities a significant over -representation.(4) These

included individuals of Indo-Canadian (Fijian), Hispanic, Iranian, Chinese, and Vietnamese ethnic origin.

Discussions with street gang members and their probation officers indicated that individuals became involved with gangs for

a variety of interconnected reasons. Although the prospect of material gain cannot be discounted as an explanation, the

members of gangs were more likely to become involved as a result of peer group attraction. They wanted to belong to a

friendly, supportive group that included their friends or close relatives and this included a desire to be with individuals from

the same cultural and ethnic group; gangmembers felt ethnically marginalized.

Many street gang members wanted to escape from, and find rewarding alternatives to, exceedingly unpleasant family lives.

Their recent domestic histories and current circumstances reflected common characteristics that transcended, in particular,

ethnicity. Most had physically abusive parents resulting in a lack of bonding, fear, and frequent absences from home (to

avoid abuse). Parents seemed unable to control or otherwise influence the behaviour of their children without recourse to

ineffective and extreme physical discipline which, in some cases, had resulted in child protection interventions by the (then)

Ministry for Social Services. Another common factor was poverty, resulting from a combination of single parenthood, an

absent or unemployed father, additional dependant siblings, and an inability to obtain employment due to a combination of

parental commitment to the dependants and a lack of English. In the case of some immigrant and ethnic minority families,

these problems were compounded by isolation from the larger, surrounding community, and even from individuals and

families within that community who shared the same ethnic background (Gordon and Foley, 1998: 52-53).

The relationships among variables such as ethnicity, family circumstances, and peer group attraction could not be measured

precisely but, arguably, they are closely related.(5) A young person or young adult experiencing abuse or neglect in his

family - and this seemed to be present in the case of most street gang members - will probably be strongly attracted to a

welcoming, supportive, and accessible alternative which simultaneously improves his material well being. The choice of

alternative is likely a product of availability and opportunity and, for some young males living in the inner urban areas of

Vancouver in the late 1980's and early 1990's, a viable alternative existed in the form of street gangs. The peer group

attraction that is evidently responsible for young males becoming involved with street gangs is probably no different than the

attraction of other socially legitimate groups such as hockey teams and the boy scouts. There is a common tendency on the

part of adolescents and young adults to group together and the main difference with respect to street gangmembers

appeared to be their choice of "gang". This choice is, to some extent, a function of the opportunities

available in communities.

Understanding why individuals became involved with wannabe groups was clearer than in the case of street gang members.

Twenty-five wanna-be group members were included in the research (only one was female) and they tended to be much

younger than the members of criminal business organizations and street gangs(the mean age of wanna-be group members

was 16.8 years). About 60 percent were born in Canada, and 68 percent were members of visible ethnic minorities.
Although there was an over -representation of subjects from visible minorities, (4) the largest single ethnic group found

amongst wanna-be group members was Canadian born individuals of European descent (28 percent).

Individuals involved with wanna-be groups tended to be involved in less serious and less organized forms of crime

shoplifting, thefts from vehicles, and minor assaults. The members of Los Cholos, for example, spent most of their

time in the suburbs of Burnaby and Coquitlam bullying other adolescents. Los Cholos members wore purple clothing and

their primary mission in life appeared to be persuading other teenagers to part with any items of purple clothing they

happened to be wearing. The group did not have any particular objectives (or significant organization) and would

occasionally attack adolescents and groups of adolescents for no apparent reason other than as a response to a comment,

a glance, or an appearance.

Wanna-be group members tended to fit the profile of hard core young offenders. They came from economically

disadvantaged circumstances and exhibited behavioral and scholastic problems at school due, primarily, to the distractions

of their adverse family circumstances. A disproportionate number attended alternate schools and other learning centres.

They became involved with wanna-be groups primarily because of their distressing family backgrounds, most of which were

characterized by physical and sexual abuse, as well as neglect. The Case of "B" provides a good illustration.

B was a 17 year old Canadian born subject of Aboriginal ethnic origin. He had a severe conduct disorder, severe anger

management needs, and fetal alcohol syndrome. He was permanently removed from his home at age three. Prior to this he

experienced physical and sexual abuse and was frequently neglected by his mother who had significant substance abuse

problems (drugs and alcohol). B had been placed in more than 20 foster homes before the age of three. His adoptive

parents divorced at which point he became unable to control the anger he felt and expressed. According to B's probation

officer, his involvement in a wanna-be group seemed to meet his need to belong to a family.

In many cases, the wanna-be group was a replacement for the families that the members did not have, or that had rejected

them. The group satisfied a variety of unmet emotional needs, especially the need for attachment - for a sense of belonging.

Discussion

The Greater Vancouver Gang Study was designed to expand the existing pool of knowledge of gangs and similar

groups in Canada. The Study explored several possible explanations for "gangs" (inclusively defined), some of which were

not supported by the data. The idea that the involvement of individuals in "gangs" was a function of their experiences with

racism was not supported. Another hypothesis - that the involvement of individuals in "gangs" was a function of the

messages and role models transmitted by the news and entertainment media - was also not supported. When the leaders

and core members of criminal business organizations and street gangs were asked, in interviews, about the possibility that

their behaviour had been shaped by the media (or words to that effect) most dismissed the idea as a joke.

There was some evidence to suggest that the entertainment media may shape the behaviour of individuals involved with

wanna-be groups but only as an intervening, rather than an independent, variable. For example, police and probation
officers reported that after the release of the popular movie "Colors" the behaviour and activities of young people who were

already involved with gangs began to reflect the images portrayed in the film. Graffitti began to appear on walls and fences,

wanna-be group members began to dress in "colours", and individuals and groups began to use the language - especially

the idiom - of the actors in the film (e.g., the term "homes"). Other films may have had a similar impact, but "Colors" appears

to have been the primary influence, for a while.

It is clear that membership of a visible ethnic minority is related to involvement in criminal business organizations,

street gangs and, to a lesser extent, wanna-be groups. It is also evident that ethnicity is possibly more significant than

whether or not a person was born in Canada (i.e., whether or not he is an immigrant). The problem lies in determining the

role that ethnicity plays.

It is possible that the data used in the Greater Vancouver Gang Study were skewed or inaccurate in some way. In particular,

it is possible that individuals from visible ethnic minorities are more easily caught and prosecuted, or are singled out for

attention by a consciously or unconsciously biased criminal justice system. By definition, such individuals are more visible

than individuals of Euro-Canadian ethnic origin, and their ethnicity may be (falsely) associated with "gang" involvement,

primarily because of media-generated and other images. Alternatively, it is only individuals from visible ethnic minorities who

are caught and prosecuted, while Euro-Canadian gang members remain undetected.

This is an important consideration and there is probably some negative labeling taking place on the part of some individuals

working within the criminal justice system. However, police and probation officers in the Greater Vancouver area are

generally sensitized to biases of this kind and labeling, alone, cannot account for the disproportionate representation of

individuals from visible ethnic minorities who are identified as "gang" involved. In addition, there is concrete evidence in the

form of police surveillance photographs and other hard data to confirm the involvement of such individuals in organizations

and street gangs.

Another possibility is that the disproportionate number of individuals from visible ethnic minorities is a product of the lack of

rewarding (materially and emotionally) economic opportunities for these individuals. This is an especially compelling

explanation in the cases of those who were not born in Canada and who face significant language and skills barriers when

they try to enter the legitimate job market. It may well be that experiencing the frustrations of attempting, but failing, to

achieve material success in a highly materialistic society, results in a rejection of conventional ways of succeeding and an

embracing of the unconventional ways. This may be especially prevalent amongst adolescents and young adults who are

bombarded with achievement messages and who are raised on the idea of instant gratification for minimal effort (the so-

called "MacDonald's Syndrome").

While the idea of "impatience" may be dismissed as unsubstantiated, there is ample evidence from the current research to

underscore the importance of economic and, equally importantly, ethnic marginality as significant causal factors. Above all

else, it is the economics of immigration and of ethnicity that seem to account best for a person's involvement in criminal

business organizations and this should be explored in future research.

Conclusion

There are significant differences between criminal business organizations, street gangs, and wanna-be groups, with

implications for policy and programme development. Dealing with a criminal business organization like Lotus requires a

different strategy from one aimed at a street gang such as Los Diablos, or a wanna-be group such as Los Cholos. Unlike

street gangs and wanna-be groups which tend to appear and disappear in waves (Young 1993), criminal business

organizations are relatively constant (Dubro 1992). They are comprised of adult males of primarily Chinese or Vietnamese

ethnic origin who migrated to Canada, and are involved in the provision of illegal goods and services. Clearly, organizations

and their members are quite different from street gangs. And street gangs and organizations are quite different from wanna-

be groups - a highly visible, noisy, and suburban phenomenon consisting of adolescent males from a variety of ethnic

backgrounds who engage in random acts of collective violence and theft, some of which seemed quite pointless to all but

the participants.

Although the development of specific recommendations for new policy and programming was not within the mandate of the

Greater Vancouver Gang Study, the work did produce insights. It is clear that careful definition and classification of

organizations, gangs and groups, and of their members, is critical to any effective policy and programming, but this may be

hard to achieve. Policy and programming developments must address the needs of individuals from visible ethnic minorities

who have recently migrated to Canada and who experience ethnic and economic marginality. The lack of language, and a

lack of both money and the means to obtain money and material goods legitimately may result in individuals

clustering in supportive groups where they are understood and can make money, albeit illegally. Whatever is being

provided in the way of settlement services for immigrants is not reaching some individuals and families. This may be a

function of the length of time during which services are provided for newly arrived immigrants, and especially those who are

refugees.

Lastly, existing community-based anti-gang programming should be continued as an important preventative strategy, but

must be expanded during the onset of a wave of street gang and wanna-be group activity. Anti-gang programming appears

to be most effective when it is aimed at the supply of new gang and group members, rather than existing and well-

established street gang members. Programs in high schools can reduce fear and intimidation, dry up the source

of gang personnel, and help generate a broader, negative perspective of gang membership, especially amongst

younger adolescents. A great deal is accomplished once gang membership is defined as "uncool" by the adolescent sub

culture (Danesi 1994).

Further research into Greater Vancouver's street gangs started in May 1999. Phase 2 of the Greater Vancouver Gang Study

is concentrating on the emergence, growth, activities, and eventually decline of three street gangs from the most recent

period of gang activity: Los Diablos; the East Vancouver Saints; and Gum Wah. In each case, key issues of ethnicity and

immigration will be considered (e.g., the impact of ethnic marginality) along with an examination of the relationships among
the gangs, criminal business organizations, and wannabe groups. A more detailed analysis of the factors responsible for the

beginning of a wave of street gang (and wanna-be group) activity will also be undertaken.

Footnote

Notes

Footnote

1. The research associated with this article was funded by the British Columbia Ministry of the Attorney-General and the

Department of Justice, Canada. The author gratefully acknowledges the support and encouragement of these government

agencies and of the British Columbia Inter-Ministry Committee on Youth Violence and Crime. The principal researcher was

Sheri Fabian, M.A. Major contributions were made by Jacquelyn Nelson, Ph.D., Lynda Fletcher-Gordon, M.S.W., Sergeant

Jim Fisher of the Vancouver Police Department, and Mike White of the British Columbia Ministry for Children and Families.

Footnote

2. Other examples of police constructed "gang" names surfaced during the research. These included the "The Goof Troop",

a name given by the police to a wanna-be group; "The Ethnic Viets", a name given by the police to a conglomerate of

criminal groups affiliated with a criminal business organization; and, "The Pin-Heads", a name used by the police to identify

the members of a criminal group led by one Mr. Pin.

3. Youth movements are extensive national, and often international, social movements characterized by a distinctive mode

of dress or other bodily adornments, a leisure time preference, and other distinguishing features. The vagaries

of adolescent fashion and other larger social and economic developments tend to determine the life spans of these

movements. To the extent that birds of a feather will flock together, adolescents who subscribe to a movement often

accumulate in groups and may be erroneously referred to, usually by the media, as a "gang". Examples include, the "zoot-

suiters" (the 1940's and 1950's), the "mods and the rockers" (the 1960's); "skin-heads" (the 1970's and 1980's); and

"punkers" (the 1980's and 1990's). Youth groups are sometimes referred to as "social gangs" insofar as they are comprised

of small clusters of young people who "hang out" together in public places such as shopping malls, fast food outlets, and

large convenience stores. They are often quite visible, noisy, and energetic and can seem intimidating. At one time, in the

Vancouver area, these groups were referred to as "Mallies" because of their frequent appearance inlarge shopping malls.

Criminal groups are small clusters of friends who band together, usually for a short period of time (no more than one year),

to commit crime primarily for financial gain. They can be composed of young people and/or young (and not so young) adults

and may be mistakenly, or carelessly, referred to as a gang.

Footnote

4. More than three quarters of the research sample were members of visible ethnic minorities (78.0 percent). According to a

1994 projection for the Vancouver Metropolitan Area, approximately 59.0 percent of the general population of the Area were

members of visible ethnic minorities (Gordon and Foley 1998, and Gordon and Nelson 1993).
5. The relationships among economic disadvantage and marginality, ethnicity and other variables will be examined more

thoroughly during Phase 2 of the Greater Vancouver Gang Study, begun in the Spring of 1999.

References

References

References

Baron, S. and D. Tindall

1993 Network structure and delinquent attitudes within a juvenile gang, Social Networks 15: 255-273.

Beare, M.

1996 Criminal Consparacies: Organized Crime in Canada. Toronto: Nelson Canada.

References

Canadian Centre for Justice Statistics

1999 Organized Crime Activity in Canada, 1998. Ottawa: Statistics Canada.

Cloward. R.A. and L.E. Ohlin

1960 Delinquency and Opportunity: A Theory of Delinquent Gangs. Glencoe, II: Free Press.

Curry, D. and S. Decker

1998 Confronting Gangs: Crime and Community. Los Angles: Roxbury.

Danesi, M.

1994 Cool: The Signs and Meanings of Adolescence. Toronto: University of Toronto Press.

References

Dubro, J.

1985 Mob Rule: Inside the Canadian Mafia. Toronto: MacMillan.

Dubro. J.

1992 The Dragons of Crime: Inside the Asian Underworld. Markham: Octopus Publishing.

Fasilio, R. and S. Leckie

1993 Canadian Media Coverage of Gangs: A Content Analysis. Users Report 1993-14. Ottawa: Ministry of the Solicitor

General.

Goldstein, A. and C.R. Huff (eds.)

1993 The Gang Intervention Handbook. Champaign: Research Press.

Gordon, R.M.

1993 Incarcerated Gang Members in British Columbia: A Preliminary Study. Victoria: Ministry of Attorney General.

References

Gordon, R.M.

1995 Street gangs in Vancouver. In J. Creechan and R. Silverman (eds.), Canadian Delinquency. Toronto: Prentice Hall.

Gordon, R.M.

1998 Street gangs and criminal business organizations: A Canadian perspective. In K. Hazlehurst and C. Hazlehurst

(eds.), Gangs and Youth Subcultures: International Explorations. Transaction: New Brunswick.

Gordon, R.M. and S. Foley

1998 Criminal Business Organizations, Street Gangs and Related Groups in Vancouver: The Report of the Greater

Vancouver Gang Study. Victoria: Ministry of Attorney General.

Gordon, R.M. and J. Nelson

1993 Census '93: The Report of the 1993 Census of Provincial Correctional Centres. Victoria: Ministry of Attorney General.

References

Gordon, R.M. and J. Nelson

2000 Crime, ethnicity and immigration. In R. Silverman, James J. Teevan, and Vincent F. Sacco (eds.), Crime inCanadian

Society (6th edition). Toronto: Harcourt Brace.

Joe, D. and N. Robinson

1980 Chinatown's immigrant gangs. Criminology 18: 337 - 345. Kennedy, L. and S. Baron

1993 Routine activities and a subculture of violence: A study of violence on the street. Journal of Research inCrime and

Delinquency 30: 88 - 112.

References

Klein, M., C. Maxson. and J. Miller (eds.)

1995 The Modern Gang Reader. Los Angeles: Roxbury Publishing.

Ley. D.

1975 The street gang in its milieu. In G. Gappert and H. Rose (eds.), The Social Economy of Cities: Vol. 9 Urban Affairs

Annual Review. Beverly Hills: Sage Publications.

Mathews, F.

1993 Youth Gangs on Youth Gangs. Ottawa: Department of Justice.

Rogers, K.H.

1945 Street Gangs in Toronto: A Study of the Forgotten Boy. Toronto: Ryerson Press.

References

Spergel, I.

1995 The Youth Gang Problem: A Community Approach. New York: Oxford University Press.

Wolf, D.R.

1991 The Rebels: A Brotherhood of Outlaw Bikers. Toronto: University of Toronto Press.

Young, M.

1993 The History of Vancouver Youth Gangs: 1900-1985. Burnaby: Simon Fraser University, School of Criminology.

Unpublished M.A. thesis.

AuthorAffiliation

Robert M. Gordon

School of Criminology,

Simon Fraser University,

Burnaby, B.C.

Word count: 7623

Copyright Canadian Criminal Justice Association Jan 2000

 Gordon, R. M. (2000). Criminal business organizations, street gangs and 'wanna-be' groups: A vancouver perspective. Canadian
Journal of Criminology, 42(1), 39-60. Retrieved from http://search.proquest.com/docview/216092237?accountid=37714

ANTISOCIAL BEHAVIOR AND YOUTH GANG

MEMBERSHIP: SELECTION AND SOCIALIZATION
Gordon, Rachel AAuthor Information ; Lahey, Benjamin BAuthor Information ; Kawai, Eriko; Loeber, RolfAuthor
Information ; et al. Criminology 42.1 (Feb 2004): 55-87.

1. Full text
2. Full text - PDF
3. Abstract/Details
Hide highlighting
Abstract
TranslateAbstract

We examine whether gang membership is associated with higher levels of delinquency because boys predisposed to

delinquent activity are more likely than others to join. We use 10 years of longitudinal data from 858 participants of the

Pittsburgh Youth Study to identify periods before, during and after gangmembership. We build on prior research by

controlling for ages and calendar time, by better accounting for gang memberships that occurred before the study began,

and by using fixed effects statistical models. We find more evidence than has been found in prior studies that boys

who join gangs are more delinquent before entering the gang than those who do not join. Even with such selective

differences, however, we replicate research showing that drug selling, drug use, violent behaviors and vandalism of property

increase significantly when a youth joins a gang. The delinquency of peers appears to be one mechanism of socialization.

These findings are clearest in youth self-reports, but are also evident in reports from parents and teachers on boys' behavior

and delinquency. Once we adjust for time trends, we find that the increase in delinquency is temporary, that delinquency

falls to pre-gang levels when boys leave gangs. [PUBLICATION ABSTRACT]

KEYWORDS: youth gangs, delinquency, fixed effects, random effects

Full text
TranslateFull text
Turn on search term navigation

Headnote

KEYWORDS: youth gangs, delinquency, fixed effects, random effects

We examine whether gang membership is associated with higher levels of delinquency because boys predisposed to

delinquent activity are more likely than others to join. We use 10 years of longitudinal data from 858 participants of the

Pittsburgh Youth Study to identify periods before, during and after gangmembership. We build on prior research by

controlling for ages and calendar time, by better accounting for gang memberships that occurred before the study began,

and by using fixed effects statistical models. We find more evidence than has been found in prior studies that boys

who join gangs are more delinquent before entering the gang than those who do not join. Even with such selective

differences, however, we replicate research showing that drug selling, drug use, violent behaviors and vandalism of property

increase significantly when a youth joins a gang. The delinquency of peers appears to be one mechanism of socialization.
These findings are clearest in youth self-reports, but are also evident in reports from parents and teachers on boys' behavior

and delinquency. Once we adjust for time trends, we find that the increase in delinquency is temporary, that delinquency

falls to pre-gang levels when boys leave gangs.

Gangs are a major social problem. In recent decades, for example, gang-related homicides have become a larger proportion

of the total. In Chicago, one study estimates the increase to be from less than 10 percent in1965 to more than 25

percent in 1994 (Howell, 1999). Concerns about gangs have also expanded geographically. Once seen as restricted to a

small number of the nation's largest cities, gang problems are now recognized in many mid-size cities and nonmetropolitan

areas (Howell, 1998). And, even though gangmembership during adolescence is short lived, there is a strong correlation

between it and delinquent behavior (Thornberry et al., 2003). Thus, even a brief spell of participation may expose a youth to

a substantially heightened risk of delinquency. To address the problem, it is important to know if the association

between gang membership and delinquency is causal. If so, what mechanisms operate? Does the delinquency associated

with gang membership persist after a boy's brief participation?

Gang researchers have speculated about these selection and socialization processes. For instance, selection may operate

when new gang members are more likely to be drawn from among boys1 who already engage inantisocial behaviors. On the

other hand, as gang members, boys may be exposed to new opportunities to participate in criminal activities. At the

extreme, if the participation of gang members in criminal activity is entirely explained by selection, then

eliminating gangs would lead to no reduction in crime. Rather than adopting one extreme, we follow earlier research that has

considered selective and socialization processes as complementary rather than competing (Thornberry, Lizotte, Krohn,

Farnworth and Jang, 1994; Thornberry, Krohn, Lizotte and ChardWierschem, 1993).

LITERATURE REVIEW AND RESEARCH QUESTIONS

PRIOR RESEARCH

Cross-sectional studies have documented that gang members have higher delinquency than nonmembers (Curry and

Spergel, 1992; Dukes, Martinez and Stein, 1997; Henry et al, 2001; Huff, 1996). Longitudinal studies of boys provide further

evidence that at least some aspects of delinquent behavior increase during the period of gang membership relative to the

period before, and that these antisocial behaviors then decline after the boy leaves the gang (Esbensen and Huizinga, 1993;

Thornberry et al., 1993; Thornberry, 1998). Thornberry and colleagues (1993) were the first to use longitudinal data to test

the selection versus socialization hypotheses. They found support for a social facilitation model, documenting higher

delinquency among gang members when the youth were active in the gang but no significant differences between nongang

youth and future or former gang members. Later research built on this study, replicating the heightened delinquency when

boys were active in gangs, but also finding some evidence of selection, with future gang members showing greater

delinquency than nongang youth in the year prior to entering a gang (Esbensen and Huizinga, 1993; Thornberry et al.,

2003). Recent studies also suggest that one aspect of delinquencydrug sales-persists after boys exit gangs (see

review in Thornberry, 1998).

This research made an important contribution to literature on gangs, being the first to examine gangmembers along a

developmental trajectory before, during and after gang membership. Still, limitations inthese initial studies can be addressed,

especially using additional waves of data now available from longitudinal studies of delinquent youth. In particular, the

earliest studies (Esbensen and Huizinga, 1993; Thornberry et al., 1993) relied on simple descriptive means, proportions and

individual offense rates, without adjusting for measured or unmeasured characteristics of boys that might bias the

association between gangmembership and delinquency. More recent research has used standard regression models to

adjust for measured controls and random effects models to adjust for unmeasured stable characteristics of boys

(Thornberry, 1998; Thornberry et al., 2003). We extend these recent developments by estimating fixed effects models to

adjust for unmeasured biases (see Method below). This approach allows us to better separate selection from socialization

effects.

In addition, prior studies have defined gang participation in relation to waves of the study rather than specific to the boys'

ages, and did not control for age. Because the periods before, during and after gang entry are correlated with age, age-

related changes in youth behavior might be mistaken for changes associated with periods of gang membership. That is, if a

delinquent behavior generally increases with age, and boys tend to be older when they participate in gangs, then failure to

control for age would lead to an overestimate of the effect of gangs. Likewise, if boys are older when they

are gang members, but a delinquent behavior generally decreases with age, then omitting an age control would understate

the gang effect. Similarly, these studies only roughly control for year-to-year time trends in delinquency, and, within a

longitudinal study, the periods of time when boys are active gang members or have left a gang tend to be later in calendar

time.

These studies also do not consider gang memberships that occur prior to the study. The boys analyzed ranged in age from

11 to 15 when the studies began. Prior analyses of the Pittsburgh Youth Study indicate that gangentry begins as early as

age 9 and that nearly 5 percent of all youth and 20 percent of African American youth have entered a gang by age 15

(Lahey et al., 1999). This has two potential consequences. First, boys whose only gang affiliation occurred before the study

began will be miscoded as never having been gang members. Such miscoding would bias against identifying selection

effects if these former gang members have higher delinquency levels than boys who truly never participate. Second, boys

who participate in gangs both before and during the study, and are not gang members at the first study wave, will be

miscoded as before gang, rather than after gang. To the extent that antisocial behavior is higher after than

before gang membership, this approach underestimates the socialization effect (any increase in delinquent activity from the

period before to during membership). Similar consequences are possible in some prior studies that combine two semi-

annual interviews into yearly measures, coding boys as being in a gang if they report being a member at either interview.

We deal with these issues by using retrospective reports of gang membership before the first study wave, measures

of gang membership and delinquency at each study wave, and controls for age, seasonal and calendar-time trends.
In longitudinal studies, the most examined mechanism through which gang participation might socialize boys into antisocial

activity appears in prior research that has considered boys' association with delinquent peers along with

their gang participation (Battin, Hill, Abbott, Catalano and Hawkins, 1998; Battin-Pearson, Thornberry, Hawkins and Krohn,

1998). Such research builds on the sizable body of literature documenting an association between affiliation with delinquent

peers and youth delinquency (Fergusson, Swain-Campbell and Horwood, 2002; Warr, 2002). We consider this potential

mechanism as well, though unlike prior longitudinal studies we link specific kinds of peer delinquent activity drug sales, drug

use, violent delinquency and property delinquency-to these same specific kinds of antisocial activity on the part of the study

youth. Such specificity may be important, given that prior research suggests that gang involvement has a greater influence

on drug sales and violent delinquency than property delinquency (Battin et al., 1998; Thornberry et al, 1993).

CURRENT STUDY

In some sense, the question of whether gang participation is associated with antisocial activity is a tautology. Two commonly

defined types of youth gangs are the street gang and drug gang, in both of which delinquency is requisite, in the former as

members protect their territory from outsiders and in the latter as members sell drugs and protect their territory from rivals

(Klein, 1997). Becoming a member of a gang may also bring the boy into a history of intragang and intergang disputes.

Simply by association, the boy may be a target against whom others may initiate violence (Howell, 1999). Yet, it can be

fruitful to consider why the strength of this association may vary (Warr, 2002 discusses similar and additional mechanisms

for peer influences on delinquency). Boys who join gangs that as a whole engage in more crimes would be expected to

show a greater increase in their own criminal activity. Once a part of the gang's social network, boys gain greater access to

any existing information within the network regarding when, where, and how to engage incriminal activity. The members of

the group can become role models and mentors to the boy in increasing his repertoire of criminal activity. In addition, as the

network members become a featured part of the boy's larger social network, the norms and approval of the group may

become increasingly relevant for shaping his behavior. Peers are therefore an important socializing agent within gangs.

Several selective processes require careful attention as researchers study how gang participation relates to boys' antisocial

activity. In the principal selective process related to gang entry already noted, boys with a tendency toward antisocial activity

are more likely to become gang members. Research has documented this relationship, showing that boys who demonstrate

higher levels of delinquency early in life or who are on a trajectory of increasing criminality are more likely to enter a gang.

For example, Lahey and colleagues (1999) documented such selection using a smaller number of participants and waves of

the Pittsburgh Youth Study than we analyze here. Selection out of gangs may be relevant as well. That is, boys who have

less of a tendency toward antisocial activity who do become members of delinquent gangs may be more likely to exit. Study

attrition may be a third source of selection. Prior research suggests that boys with more delinquent tendencies are more

likely to drop out of research studies (Thornberry, Bjerregaard and Miles, 1993). If this selective attrition is particularly true of

core, long-term gang members, then the extent to which gangs affect behavior may be overstated. In other words, these

boys may be those most attracted to gang membership because of the benefits of associated gang activities. If data were
observable for these attriters, less of an increase over previous behavior might be evident and they might be more likely to

continue with criminal activity after leaving the gang.

The present study attempts to deal with these issues. Our central research question is whether youths' delinquent activity

rises when they enter gangs and falls when they exit. Given that such an association is observed, our interest is in why it

occurs, with the potential for changes in the delinquent activity of the boy's peers as a central possible socialization

mechanism. Our analysis begins with a descriptive portrait of youths' behavior by gang status. To the extent that delinquent

activity and gang membership share common causes, however, this description provides a biased estimate of the causal

relationship between gang membership and delinquency. We thus use statistical models to comprehensively adjust for

biases associated with preexisting characteristics of boys. We also control for age, season and calendar year and account

for gang memberships that occurred before the study began.

DATA AND METHODOLOGY

SAMPLE

The Pittsburgh Youth Study (PYS) follows three samples representative of all boys who were enrolled in the first, fourth or

seventh grades (called the youngest, middle and oldest samples, respectively) in public schools in Pittsburgh,

Pennsylvania in 1987-1988 (see Loeber et al., 1998 for a detailed description of the PYS). The current study includes 858

boys, 503 from the youngest sample as well as 355 boys from the second cohort of the oldest sample. Because a key

question was omitted from the gang questionnaire that was administered to the first cohort of the oldest sample, those

participants are excluded from this study. The boys described themselves ethnically primarily as African American (487) or

white (351), with the remainder choosing other (10), Asian (6), Hispanic (3) or American Indian (1). At the initial interview

(Wave S), nearly all of the boys in the youngest sample were ages 6 or 7 (53 percent and 41 percent respectively); the

boys in the second cohort of the oldest sample were primarily ages 12 through 14 (35 percent age 12, 41 percent age 13, 21

percent age 14). The latest available interview point for the current paper is in the 10th year of follow-up, when the

boys in the youngest sample were ages 16-17 and the boys in the oldest sample 22-24. Initially, about 85 percent of boys

agreed to participate in the study, with subject retention being above 80 percent in each follow-up wave.

The PYS was designed to provide a sizable sample of boys who engage in serious antisocial behavior and to maintain a

comparison group of nonantisocial boys. Based on an initial screening assessment completed by parents, teachers and the

youth themselves, boys were divided into two groups thought to be at high and low risk for offending. High-risk boys were

oversampled, and sampling weights are available to make estimates representative of first, fourth and seventh graders from

the Pittsburgh public schools in the late 1980s. Although estimating descriptive statistics that represent the population

requires weighted analyses, unweighted regression models are unbiased and efficient when models are correctly specified

(that is, in this context, a correctly specified model does not omit relevant predictors, including potential interactions between

predictor variables and sample stratification variables; DuMouchel and Duncan, 1983). We present unweighted models.

Results are highly consistent in weighted models (details available from the authors).
As expected for an urban public school sample, the boys' families had relatively limited economic resources. When the study

began, the median family income in 1998 dollars was $16,500. The majority of primary caregivers in the study had either not

completed high school (18 percent) or had completed high school but had received no further education (49 percent). And,

at the time of the first interview, 49 percent of the boys lived in single-parent families, 37 percent lived with both biological

parents and 14 percent lived with one biological parent and one nonbiological parent figure.

Pittsburgh is an example of a city with a recently emergent gang problem, in contrast to cities like Chicago and Los Angeles

with their longer traditions of gang activity and research. Gang activity and homicides inPittsburgh escalated in the early

1990s, peaking just prior to the middle of the decade, and then falling through the late 1990s (Copeland, 2001; Kelly and

Ove, 1999; Mamula, 1997). Consistent with these reports, our data show a peak in gang activity around 1993-1994.

DATA COLLECTION WAVES

As shown in Appendix A, data available for this paper covered a 10year period including fifteen waves of data for the

youngest sample and fourteen for the oldest sample. Initially, both grade samples were reassessed at 6-month intervals.

However, due to budgetary reasons, in the third and fourth years of the study the follow-up intervals were increased to 1

year, with the youngest and oldest grade samples alternating in the field. The first wave was the initial screening

assessment (denoted Wave S following the terminology of the study) and later waves are denoted alphabetically from A

(with SS used to denote the wave between R and T).

MEASURES

ASSESSMENT OF GANG MEMBERSHIP

Scholars debate the definition of gangs (Ball and Curry, 1995). Similar to other recent longitudinal studies of

youth gangs and delinquency, we assess gang membership using boys' self-reports. Little research has been conducted on

the validity of self-reported gang membership. However, a recent study of Chicago youth found that though self-reports

of gang membership do not coincide exactly with police identification of gangmembers, self-reported gang involvement does

correlate with official records of delinquency, controlling for boys' race/ethnicity and their own and official reports of prior

delinquency (Curry, 2000).

In the Pittsburgh Youth Study, data were collected about gang membership for the first time in Wave D for the oldest sample

(when the boys were ages 14-16) and in Wave H for the youngest sample (when the boys were ages 10-11). At these

waves, boys were asked if they were currently, or had ever been, a member of a gang. Follow-up questions collected

information about the boy's age at gang entry and exit and characteristics of the gang. During later waves, the boys were

asked about membership in gangs since the last interview with follow-up questions asking about the gang's characteristics

and whether the boy was still a member at the time of the interview.
We use the gang questionnaire to categorize boys into four gang status groups at each study wave: (1) those who had not

entered a gang by the last observation point (Wave T or U) and those who had entered a gang by the last observation point

and the current study wave is (2) before, (3) during or (4) after gang participation. Those who had entered and exited

a gang before the first study wave were coded as "after" gang participation at the first interview. When a boy missed any

interview after the first gang assessment point, we have no information about his gang status for the time between the prior

interview and the missed interview. This means that if a boy dropped out of the study, joined and exited a first gang, and

then rejoined the study, we would mistakenly identify him as not yet entering a gang. We thus censor boys at the first point

of missing data. If the boy was interviewed at later waves, those data are excluded from our models. Results are highly

consistent when we include all available waves of data (details available from the authors).

ANTISOCIAL BEHAVIOR MEASURES: PEERS

The Peer Delinquency Scale (PDS; Loeber et al., 1998) elicits information about the antisocial behavior of the boy's close

friends. The PDS has high internal consistency and has been related to the youth's level of delinquency (Loeber et al, 1998).

Each boy reported how many of his friends engaged in particular antisocial behaviors (twelve behaviors for the youngest

sample before Wave G, fifteeen for all other waves and for the oldest sample). We define four types of peer delinquency,

relating to drug selling, drug use, violent delinquency and property delinquency, drawing on behaviors that were asked about

across all waves (see Appendix B). The response structure changed across time for the youngest sample (from "none, one,

some or all" of the boy's friends to "none, few, half, most or all" of the boy's friends). For consistency across waves and with

the oldest sample, we collapse responses into three categories (0=none of the boy's friends, 2=all of the boy's friends, 1=all

other responses).

ANTISOCIAL BEHAVIOR MEASURES: YOUTH

Boys in the oldest sample completed the forty-item Self-Reported Delinquency Scale (SRD; Elliot, Huizinga and Ageton,

1985). The SRD has been shown to be reliable and has been validated against official crime records (Huizinga and Elliott,

1983). For the first seven waves of data collected from the youngest sample, the study staff developed an ageadapted

measure, the thirty-three-item Self-Reported Antisocial Behavior Scale (SRA; Loeber et al., 1998). In later waves, the SRD

was administered to the youngest sample. In the SRA, boys reported whether they engaged in each activity 0=never, 1-

once, 2=twice, or 3=more than twice in the reference period. In the SRD, boys reported the exact number of times they

engaged in the activity in the reference period. For consistency, we downcode the SRD reports at 3 (that is, if a boy reported

on the SRD that he engaged in the activity 4 or more times in the reference period, the value was receded to 3). We also

convert all reports to 6-month frequencies (by dividing reports for 1-year reference periods in half before top coding).

Appendix B lists our four measures of the boy's self-reported antisocial behaviors, which are consistent with the peer

measures as well as prior research (Thornberry et al, 1993). We also define a measure of aggression to cover all available

waves of SRA/SRD data because only a subset of the items we use to measure violent delinquency are available in the

SRA.
Adapted versions of Achenbach's widely used measures of youth problem behaviors were also administered to the youth

(Youth SelfReport, YSR), his caregiver (Child Behavior Checklist, CBCL) and his teacher (Teacher Report Form, TRF;

Achenbach, 1991a, 1991b, 1991c). Each informant rated the 112 items on the checklist, plus several additional items added

for the study, as 0=not true, 1=somewhat true, or 2=very true of the boy's behavior since the last interview. In the current

paper, we define four categories of delinquent activity to parallel the PDS and SRD measures (see Appendix B).

ANTISOCIAL BEHAVIOR: GANG

Boys were asked separately open-ended questions about the activities of each gang as well as closed-ended questions

about the activities of their most recent gang. Their responses to these questions were used to create delinquency variables

parallel to those created for youth and peer delinquency. In particular, drug sales was based on three dichotomous codes

tapping pushing or selling drugs. Drug use included three codes about doing drugs, getting high, and drinking. Violent

delinquency included ten codes regarding fighting, using weapons in fights, providing protection, robbing or strongarming

people, having forced sex, terrorizing people, attacking someone to hurt them, and shooting or killing people. Property

delinquency included seven codes including robbery, stealing, vandalism and joyriding.

STATISTICAL MODELS

We are interested in the degree to which being a gang member is associated with high levels of delinquency, above and

beyond the levels of delinquency that we would expect because boys predisposed to delinquent activity may be more likely

to enter, and stay in, gangs. One way to address this question would be to control for measured differences between boys

who do and do not participate in gangs. However, any study has only a limited number of such controls. We therefore use

statistical models to adjust for preexisting characteristics of boys, both those that are measured in the study and those that

are not.

By separating gang members along the pre-gang, during-gang, and postgang timelines we partially adjust for selective

differences between boys who do and do not enter gangs. This approach is limited, however. The pregang category is

restricted to the desired set of boys who will later enter gangs, but the during-gangcategory mixes boys who will

leave gangs and boys who will stay in gangs. Thus, the contrast of the post-gang to the during-gang category may be

inaccurate. In addition, even the simple during-gang to pre-gangcontrast will only match the fixed effects approach

(described below) when we restrict the estimation sample to the waves immediately before and after a boy enters a gang.

Two major alternative models are available to adjust for unmeasured preexisting differences: random effects and fixed

effects (Wooldridge, 2002). Both separate the error term from a standard regression model into person-specific and time-

specific components. The first differs across individuals but remains the same within each individual across time. The

second differs across both individuals and time. In the random effects approach, the person-specific component is modeled

as a draw from a distribution. In the most common fixed effects approach, the person-specific component is modeled with a

dummy variable for each person in the sample. These dummy variables can be estimated because each person has at least
two data points. Intuitively, the coefficients for these dummy variables allow the intercept to shift up or down depending on

the collective effect of any characteristics of the boy that are stable across time. These controls remove between-person

variance, thus the analyses examine change within an individual. (Indeed, for two waves of data, the fixed effects model is

identical to a first difference model that predicts change in the outcome variable based on change in the predictor variable).

The precision with which the person-specific effects are estimated increases when more time points are available for each

person and estimates are more precise when more change occurs on the variables of interest. The fact that our person-

specific effects are based on at most ten waves of data, and relatively few boys are observed to enter and exit gangs, will

reduce precision, making it harder to detect significant effects.

The random effects and fixed effects models both have additional advantages and limitations. A major limitation of the fixed

effects model is that time-constant characteristics of individuals cannot be included as predictors because they are collinear

with the person dummy variables. In our case, this limitation prevents us from using the fixed effects model to examine

selection by contrasting boys who never join gangs with boys who will later join gangs (because boys in the "never gang"

category do not change gang status across the study waves). Still, it does allow us to examine socialization by contrasting

boys who do join gangs across time, before, during and after they enter gangs. The random effects model has a different

limitation: It makes the assumption that the person-specific component of the error term is uncorrelated with the predictor

variables included in the model. This assumption is unlikely to hold. We expect that variables we do not measure that may

predict delinquency, such as a genetic tendency toward thrill-seeking or abuse and neglect during early childhood, are also

correlated with gang status. This expectation is precisely why we desire a statistical adjustment for these unmeasured

variables. The fixed effects model allows for such correlation between the person-specific error component and the predictor

variables. The Hausman specification test evaluates this assumption (Wooldridge, 2002). We focus on the fixed effects

approach but also summarize results from random effects models.

An important advantage of both models is that the person-specific characteristics they adjust need not be measured in the

study. Such characteristics would include stable demographic characteristics, such as the boy's race, ethnicity and his

parents' age and educational level when he was born. They would also include the boy's experiences prior to the initial study

wave-such as changes in his parents' marital status or his exposure to violence and abuse during early childhood - because

these remain constant during the study period. They would also include any stable biological or genetic characteristics. Both

models, however, are limited by the fact that they do not adjust for characteristics that change over time but are unmeasured

(for example, change in local policing policies, entry of a new gang into the neighborhood). Our models control for seasonal,

calendar-year and age-related trends to capture some of these time-varying correlates.

Our outcomes are sums of items from youth and peer delinquency scales. We thus estimate negative binomial regression

models (Long, 1997). The negative binomial regression model is appropriate for count outcomes that are skewed right,

improving on the Poisson regression model by allowing the conditional variance to exceed the conditional mean. Strictly

speaking, the negative binomial (and Poisson) apply to counts of discrete events (for example, the number of times the boy

carried a hidden weapon). Although our outcomes are instead sums of small-range (for example, none to three) items
capturing different events (carried a hidden weapon, hit people, threw rocks or bottles at people and so on), the negative

binomial is expected to perform well because of the positively skewed, integer nature of the outcomes. The negative

binomial regression model also has the advantage of a simple factor change interpretation: We can interpret a unit

change in the predictor variable as affecting the expected count by a factor equal to the exponential of the coefficient.

We estimate negative binomial models with and without fixed effects, referring to the latter as "standard" models. The

process by which fixed effects estimates are calculated differs in the negative binomial regression case in contrast to the

least squares approach, with important implications for the precision of the estimates. In particular, the fixed effect negative

binomial regression model is estimated as a conditional fixed effect, with the joint probability of the boy's delinquency scores

at each individual wave being conditional on the probability of his total delinquency across the waves (Hausman, Hall and

Griliches, 1984). The conditioning occurs by placing the sum of each boy's outcome variable across waves in the

denominator. As a result, boys with all zeros on the outcome variable do not contribute to the log likelihood (that is, are

excluded from the estimation sample). This reduced sample size can result in large standard errors. The omission of many

cases also raises intuitive concerns that estimates may be biased.

Ordinary least squares, while less desirable than negative binomial regression in terms of functional form, provides an

alternative estimation approach in which cases without variation on the outcome variable can contribute to the fixed effects

estimates and in which the properties of the estimates do not require large samples. The problem of heteroscedasticity

common to least squares estimation with count outcomes can be dealt with by using the general Huber-White correction for

heteroscedasticity (Wooldridge, 2002). In addition, Allison and Waterman (2002) recently critiqued the conditional fixed

effects negative binomial regression model and evaluated an alternative unconditional fixed effects negative binomial

regression model. We focus on the standard and conditional fixed effects negative binomial regression models in the tables

and text, and comment on the consistency of results based on ordinary least squares and unconditional negative binomial

fixed effects models.

We estimate the unconditional fixed effects negative binomial regression model using SAS 8.2. All other results

presented in this paper were calculated using Stata 7.

RESULTS

GANG PARTICIPATION

Among the study participants, 165 boys were observed to participate in gangs during the course of the study-134

participated in a single gang, 26 in two gangs and 5 in three gangs). Most spent only a brief period inthe gang. Nearly 85

percent were observed to exit the first gang they entered. And nearly three-quarters who exited were observed to

be in the gang for only one or two waves of the study (48 percent one wave, 25 percent two waves). Boys self-

defined gangs and reported a range of gang activities. At the wave when they entered the gang, nearly all boys (99 percent)

reported that the gang they entered engaged in violent delinquency. Of the other three delinquent activities, 59 percent
reported drug selling, 51 percent property delinquency and 21 percent drug use. Fully 12 percent of boys reported that

their gang engaged in all four of the activities and 35 percent reported three of the four activities.

Enlarge this image.

Figure I. A descriptive portrait of youth-reported activity by gang

Average levels of delinquency by boys' gang status for each of the seven self-reported measures are presented in Figure 1.

Within each chart, a selection effect is evident if the second bar is higher than the first, indicating higher levels of

delinquency among boys who will later join gangs than among boys who do not join gangs ("before gang" versus

"never gang"). Socialization is evident if boys report greater delinquent activity when they are gang members (third bar) than

before (second bar) or after (fourth bar) they participate in a gang. If this socialization effect is temporary, then levels of

delinquency "after gang" should return to "before gang" levels. Persistent delinquency following gang participation would

indicate a more permanent change, such as through the boy's maintenance of a new peer networks or taste for delinquency.

The statistical significance of these differences was tested by estimating a standard negative binomial regression

model in which three dummy variables indicate the four gang statuses (data not shown). The charted means in Figure 1 are

predictions from these models.

In these descriptive results, there is evidence of selection for violent and property delinquency, but not for drug sales and

drug use. That is, the average level of drug-related delinquency of future gang members does not differ significantly from the

average level of drug-related delinquency among boys who never join gangs(compare the "never gang" and "before gang"

bars in the top row of Figure 1). In contrast, boys who will later join gangs exhibit significantly higher levels of violent and

property delinquency even before they enter a gang than boys who never join gangs.

For all outcomes we also see a socialization effect of gang participation. In all seven charts in Figure 1, the third bar

"during gang" is significantly higher than the second bar "before gang." This indicates that, among gang members, drug

selling, drug use, violent delinquency and property delinquency all increase on average during the period of

active gang membership relative to the period before gang entry. For violent and property delinquency, this elevation is

temporary. Once the boy exits the gang, his delinquency returns to statistical equivalence with (or falls significantly below)

pre-gang levels (compare "after gang" and "before gang" bars in second through fourth rows of Figure 1). In contrast, drug

selling and drug use both persist after the boy exits the gang. That is, the "after gang" level of drug-related delinquency

remains significantly higher than the "before gang" level in the two charts in the top row of Figure 1.

FIXED EFFECTS MODELS OF YOUTH-REPORTED DELINQUENT ACTIVITY

We present negative binomial models predicting both boys' drug sales and drug use (see Table 1) and boys' violent and

property delinquency (see Table 2). We summarize both standard and fixed effects negative binomial models to highlight the

way our results change when we adjust for preexisting characteristics of the boys in the latter models. We account for the

clustering of waves within boys in the standard model using the Huber-White approach (Wooldridge, 2002).
These models include three dummy indicators to capture the four categories of gang status ("never" in a gang;

ever in a gang and at a wave "before," "during" or "after" participation). We then conduct additional tests to estimate four key

contrasts.2 The first contrast ("before versus never") captures selection: Do boys who will later join gangs ("before") report

higher levels of delinquency than boys who do not join gangs ("never"). The next contrasts capture socialization. Does a

boy's delinquency increase when he enters a gang ("during versus before") and decrease when he exits ("during versus

after")? The final contrast further refines the evidence of socialization: Does a boy's delinquency decline sufficiently after he

exits a gang such that it returns to his pre-gang level of delinquency ("after versus before")? Because these hypotheses are

directional, one-sided significance levels are reported. All of these contrasts are conducted such that the coefficient will be

positive if consistent with the hypothesis.

Enlarge this image.

Table 1. Standard and Fixed Effects Models of Boys' Self-Reported Drug Sales and Drug Use

Enlarge this image.

Table 2. Standard (STD) and Fixed Effects Models of Boys' Self-Reported Aggression, Violent Delinquency and Property Delinquency

All the models presented also control for age, seasonality and calendar-time trends through cohort-specific dummy variables

for study wave and linear and quadratic terms for the boy's age in months. The use of wave dummies, rather than a linear

measure of calendar time, allows for the relationship between time and delinquency to take on an arbitrary shape. For

example, a general increase over time may take on a stair-step shape if increases happen in one season and

plateau inanother. Sudden spikes associated with particular events are also allowed. By being cohort-specific, the wave

dummy variables allow these relationships to differ for the older and younger samples. Within cohort and wave, boys'

ages inmonths vary. The additive and quadratic terms capture any additional time-trends associated with increasing age.

The most consistent finding is that evidence of socialization remains once we control for time-trends (standard models) and

for time-trends and preexisting characteristics (fixed effects models). For all outcomes, the "during versus before" contrast is

significantly positive, indicating that boys report more delinquency when they are active in the gang than they did before they

entered the gang, above and beyond general age-, seasonal- and time-trends in the data and above and beyond preexisting

characteristics of boys that are correlated both with gang entry and delinquency (see "during versus before" rows in Tables 1

and 2). For example, when boys are active in gangs they engage in over 3 times more drug selling than they did before they

entered the gang (exp[1.14]=3.13).

We also see that the evidence of selection for violent and property delinquency remains when we control for time-

trends in the standard negative binomial regression model (see significant "before versus never" contrasts in the "standard"

columns of Table 2). However, adjusting for time-trends reveals selection for drug-related delinquency not evident in the

simple descriptive means (see Table 1). This is because drug-sales and drug-use increase with age, not peaking until late

adolescence, and generally increase with calendar time. Because boys who are "before gang" are younger and at earlier

study waves, their drug-related delinquency appears similar to the average drug-related delinquency across all study waves
of the "never gang" youth until we adjust for these time-related trends. In contrast, violent and property delinquency

generally decline across age and calendar time. Thus, the "before versus never" coefficients in Table 2 for violent and

property crimes, though still significant, are somewhat smaller than those that do not control for time-trends (the coefficients,

not shown, underlying Figure 1).

The findings for the persistence of delinquency after a gang exit also differ in the multivariate versus the descriptive models.

Whereas there was evidence of persistence in drug-related delinquency in the descriptive statistics displayed in Figure 1,

this is not the case in Table 1. The "during versus after" contrasts in Table 1 are positive and significant, indicating that boys

report more drug-related delinquency while they are active members than they do after they leave the gang. In fact the

decrease in drug-related delinquency upon exit is fairly comparable to the increase on entry: For drug sales, boys report an

increase of just over 3 times (exp[1.14]) when they enter a gang and a decline of just under 3 times (exp[1.07]) after they

leave the gang. Indeed, the "after versus before" contrasts for drug sales and drug use in Table 1 are not significant,

indicating that the boys' level of drug-selling and drug use have declined enough to equal the levels they had reported

before they entered gangs. Similar to the bias described above for the selection ("before versus never") findings, this reflects

the fact that the "after gang" periods occur when boys are older and later in calendar time when general time-trends reveal

greater drug-related delinquency.

A similar bias is evident for aggression and violent delinquency, though in the opposite direction. Whereas the descriptive

means in Figure 1 revealed no persistence in violent delinquency after a gang exit, the standard models that adjust for time-

trends do indicate persistence (significant "after versus before" contrasts in the "standard" columns of Table 2). We also see

the greatest difference between the standard and fixed effects models for violent delinquency, indicating a biasing effect of

failure to control for preexisting characteristics of boys that are correlated with both gang participation and violence. The

"after versus before" contrasts are significant for violent delinquency in the standard but generally not the fixed effects

models. Although the "during versus before" contrasts for violent delinquency are significant in both models, they also reflect

this bias by being smaller-one-half to one-quarter the size-in the fixed effects versus the standard models. This indicates that

the increase in violent delinquency upon gang entry is overestimated before we account for unmeasured preexisting

characteristics of boys. Once these confounds are controlled, the increase and decrease in violence associated

with gang entry and exit are fairly comparable, leaving little evidence of persistence once the boy leaves a gang.

Although replicating the generally strong association between current gang participation and elevated delinquent activity,

these findings differ somewhat from prior research. Prior longitudinal studies have found that this association is strongest for

violent delinquency and that drug-related activity is most likely to persist after a gang exit (see Thornberry, 1998 and

Thornberry et al, 2003). Our data are more consistent with these prior studies before we add controls for time trends. That

is, the post-gang periods are later in calendar time and age when drug selling and drug use had increased and violent and

property delinquency were lower among Pittsburgh youth. We also see more evidence of selection into gangs than have

prior studies, perhaps due to accounting for gang memberships that occurred before the study began as well as due to

adjusting for these general time trends.

EXPLORING SOCIALIZATION MECHANISMS

We estimated additional models to examine whether peer delinquency (the proportion of the boys' friends who commit

delinquent acts) is a potential socialization mechanism. Table 3 presents fixed effects models that predict peer drug sales,

drug use, violent delinquency and property delinquency, controlling for age-, seasonal- and calendar-time trends and

preexisting characteristics of boys. The pattern of significant effects generally mirrors the findings for the boys' own activities.

All four outcomes increase significantly when the boy enters a gang ("during versus before"). All also significantly decline

after an exit ("during versus after"), although there is some evidence of persistence in peers' drug use and violent

delinquency ("after versus before").

These findings suggest that changes in peer delinquency may be one mechanism through which boys' own delinquency

changes when they participate in gangs. We thus added peer delinquency to the fixed effects models predicting youth

delinquency within domain (for example, controlling for peer drug sales when predicting youth drug sales). The socialization

coefficients are uniformly reduced suggesting that peer delinquency is a partial mechanism (compare Table 4 with the "fixed

effects" columns in Tables 1 and 2).3

Enlarge this image.

Table 3. Fixed Effects Models of Peer Delinquency

In particular, the coefficients measuring the increase indelinquency associated with gang entry ("during versus before") are

between 13 percent and 46 percent less. Additionally, the coefficients measuring the decrease in delinquency associated

with gang exit ("during versus after") are between 13 percent and 65 percent less. With the exception of Youth Self-Report

property delinquency, however, all of these coefficients remain significant, suggesting that a mechanism in addition to peer

delinquency operates. In addition, generally, peer delinquency explains the gang-entry increase and the gang-exit decrease

comparably enough such that the "after versus before" coefficients continue to be insignificant. The exception is with

aggression. For this outcome, we capture more of the mechanism of a decline in aggression upon leaving a gang than the

mechanism for a rise in aggression upon joining a gang, thus the "after versus before" coefficient is significant when peer

delinquency is controlled.

CONSISTENCY OF FINDINGS IN ALTERNATIVE MODELS

We reestimated the fixed effects models from Tables 1 and 2 using OLS rather than a negative binomial specification and

found that the results of twenty of the twenty-one significance tests are the same. (The exception is that none of the "after

versus before" contrasts are significant in OLS, whereas one is just significant in the negative binomial specification).

Enlarge this image.

Table 4. Fixed Effects Models of Boys' Self-Reported Delinquency Controlling for Peer Delinquency
We also reestimated the models with the unconditional, rather than conditional, fixed effects negative binomial regression

model. These results are consistent with those reported inTables 1 and 2 although they reveal more evidence of

persistence in delinquency (the "after versus before" contrast is significant at one-sided p<.05 for drug sales and onesided

p<.025 for aggression) and, in terms of property delinquency, the results for Youth Self-Report are more similar to the results

for SelfReported Delinquency (in the unconditional model of Youth Self-Report, one-sided p<.025 for "during versus before"

and one-sided p<.05 for "during versus after").

Finally, we reestimated the models with random rather than fixed effects and tested the random effects assumption. The

random effects models, like the fixed effects models, show evidence of socialization. The "during versus before" and "during

versus after" contrasts are significant for all outcomes. However, specification tests suggest that the random effects model is

not adequate for this data. Before we control for time variation by age, season and year, the random effects assumption is

rejected by the Hausman test for all outcomes. When we add controls for time-trends, the assumption is rejected for violent

delinquency (SelfReported Delinquency and Youth Self-Report; details available from the authors). In addition, if the random

effects models adequately adjust for preexisting characteristics of boys, then gang boys should not differ from nongang

boys in these models before they enter gangs. In the language of our tables, the "before versus never" contrast should be

insignificant. However, we find that these contrasts remain significant for all outcomes in the random effects models.

PARENT AND TEACHER REPORTS

Our analyses to this point have been based entirely on youth selfreports. Thus we cannot distinguish

whether gang membership is associated with actual changes in behavior or with other factors such as changes in the social

desirability the boy perceives for reporting delinquent activity when he is involved in a gang. To help examine this issue, we

reestimated the models using parent and teacher reports of the boys' delinquent activities. As with youth self-reports, the

findings for these informants are subject to various interpretations. However, generally, whereas boys' reports might

overestimate gang effects, parent and teacher reports might underestimate them (for example, we might see less of an

association between gang status and delinquent activity because parents and teachers are less aware of the boys'

activities). In addition, because parents and teachers were not interviewed for the latter half of the study period among the

oldest sample, we lack data for the late adolescent and young adult period from these informants. Significant findings based

on parent and teacher reports might therefore be interpreted as strong evidence of socialization pathways.

We estimated the fixed effects models for parents reports (Table 5) and for teacher reports (Table 6) twice, once without

controls for peer delinquency and once with controls for peer delinquency. We begin with the models without peer

delinquency to identify any evidence of socialization and then examine whether peer delinquency might be a socialization

mechanism. Both parents and teachers report that youths' property delinquency increases when they enter a gang and that

it remains elevated after they exit the gang(significant "during versus before" and "after versus before" coefficients in the

"property delinquency/without peer controls" columns). Both parents and teachers also report an elevation in youths' violent

delinquency, though parents appear to see it sooner. In particular, parents report an increase associated with the
boys' gang entry (significant "during versus before" coefficient in the "violent delinquency/without peer controls"

column in Table 5). But teachers do not report such an increase until after the boy has exited the gang(marginally

significantly negative "during versus after" coefficient and significantly positive "after versus before" coefficient in the "violent

delinquency/without peer controls" column of Table 6). Neither parents nor teachers report an association

between gang participation and drug use. Regarding mechanisms of socialization, parents' reports of violent delinquency

increasing when boys join a gang is most explained by youth reports of changes in peers' violent delinquency (see "during

before before" coefficient in "violent delinquency/with peer controls" column of Table 5), The associations with property

delinquency decrease slightly for parent reports and unexpectedly increase for teacher reports once youthreported peer

delinquency is controlled (see "during versus before" and "after versus before" coefficients in "property delinquency/with

peer controls" columns of Tables 5 and 6).

Enlarge this image.

Table 6. Fixed Effects Models of Boys' Delinquent Activity: Teacher Reports

DISCUSSION

SUMMARY OF FINDINGS

We examined the degree to which being a gang member is associated with high levels of delinquency, above and beyond

expected levels (given that boys predisposed to delinquent activity may be more likely to enter and stay in gangs). We built

on prior research by controlling for boys' age and calendar time, by accounting for gang participation before the study began

and by using fixed effects statistical models to comprehensively adjust for all preexisting characteristics of boys. We found

more evidence than prior studies have that boys who join gangs are more delinquent before entering the gang than those

who do not join. Even with such selective differences, however, we replicate prior findings of a substantial increase in drug

selling, drug use, violent delinquency and property delinquency when boys are active gang members. Additional analyses

suggest that the mechanisms through which gangs have their effects include changes in delinquency of the boys' peers.

That is, peer delinquency is elevated during periods of gang memberships and explains a portion of the increase in boys'

delinquency when they enter gangs and the decrease when they exit.

In initial descriptive associations, we replicate prior evidence of drugselling persisting after boys exit gangs. However, this

association is explained by periods after gang membership being correlated with later years when drug-selling had generally

increased among Pittsburgh youth. After we adjust for age, season, calendar year and preexisting characteristics of boys,

we find little or no evidence of persistence in drugrelated, property or violent delinquency after a boy exits a gang.

Associations with parent and teacher reports of boys' delinquent activity further substantiate these findings. Both indicate

increases in boys' property delinquency when they are gang members. Parents, and to a lesser extent teachers, also report

increased violent delinquency, and these associations are explained by youth reports of changes in his peers' violent

criminality.
CONCLUSIONS AND LIMITATIONS

The fixed effects models we employ do not adjust for time varying covariates associated with both delinquency

and gang membership. Thus, we cannot rule out the fact that some third variable predicts both gang participation and

increased delinquency in a particular time period, rather than there being a causal effect of gangs. Our models do adjust for

any such variables correlated with the boys' age, with season and year, or with peer delinquency. Future research

on gangs should focus on measuring potential time-varying confounds and causal mechanisms. Although not using a fixed

effects approach, Thornberry and colleagues updated their earlier models with controls for family poverty level, parental

supervision, commitment to school, negative life events, youth violence and delinquent peers in the year prior

to gang membership. They still find a strong positive association between a dummy variable for current gang membership

and self-reported violence with these time-varying controls (Thornberry, 1998; Thornberry et al., 2003).

In this study, we saw that peer delinquency was elevated during a period of gang membership and explained some of the

association between gang membership and youth delinquency. One way to illuminate this finding in future research would

be to use more detailed data about boys' peer networks. Such studies could examine, for example, whether some

youth join gangs simultaneously with existing friends, and both show elevations in delinquency, or whether some youth add

new delinquent friends, and eliminate prior non-delinquent friendships, during periods of gang membership. Likewise, such

studies could examine how this mechanism operates upon gang exit: Do boys stop associating with delinquent youth, and

return to old friends, or do they and their long-term friends drop out of gangs simultaneously? Another fruitful direction for

understanding this mechanisms would be to continue research that has examined whether there is a different effect of

associating with delinquent peers inside versus outside of a gang structure (for example, Battin et al., 1998; Thornberry et

al., 2003).

We should emphasize that there is also an issue of causal direction. Does a decrease in peer delinquency lead to a

decrease in the boy's delinquency, or does a decrease in the boy's delinquency lead to a decline in his peers' delinquency?

The direction suggested above is sensible, but the other is plausible as well. For example, a boy might exit a gang and

reduce his own delinquency because the opportunities for delinquency are now diminished. Because his delinquency is

reduced, his associations with delinquent boys may decline. Given that some effect of gang membership remains, even after

controlling for peer delinquency, future research should also examine additional mechanisms. For example, Hagedorn

(1994) emphasized the importance of using organizational theories to understand gangs and to recognize heteroegeneity

across gangs in their structures.

Our results do not provide evidence of persistence in the boy's delinquency after he leaves a gang, including drug selling

that had been shown to persist in prior studies. The fact that drug selling does not remain elevated afterwards, once time

trends are controlled, should not minimize the importance of its elevation among active gang members. The increased use

of custodial sentences for drug offenses means that even a brief increase in drug selling during a short period

of gang membership could be substantively meaningful (Blumstein and Beck, 1999). In addition, boys report that increased
violence and peer drug use persist after they leave gangs, indicating that some effects of gang participation may not be

transient. Of course, our failure to find evidence of persistence is susceptible to a Type II error. This potential is amplified by

the fact that relatively few boys are observed to exit a gang more than once. Additional waves of data might provide more

post-gang data points and identify additional gang exits. Understanding why many boys leave gangs, while some stay, is

also an important direction for future research, which would also benefit from larger samples of gang exits.

We find evidence of socialization effects of gangs based on parent and teacher reports, though significant coefficients are

seen for fewer outcomes and contrasts than is the case for youth reports. We cannot distinguish whether the stronger

evidence in self-reports reflects that youths are more aware of their own behavior than parents or teachers, or that they

overstate their delinquency during times when they say they are gang members. Future research might further examine the

degree to which associations hold across multiple informants. For example, though peers may have their own biases (for

example, susceptibility to rumors), they may have greater information about adolescents' activity than would parents and

teachers. Along these lines, recent research in Canada found that classmates reported gang members to be more

aggressive than nongang youth (Craig, Vitaro, Gagnon and Tremblay, 2002).

Using more comprehensive controls for preexisting characteristics of boys as well as controls for age, season and calendar

year, this study further substantiates earlier research on a substantial elevation in delinquency when boys are

active gang members. Limitations notwithstanding, these results reinforce the need for interventions aimed at delinquent

youth gangs.

Footnote

1. Although participation of girls in gangs is an important topic of study, our data source in this paper comprises boys

exclusively. Some of the theoretical ideas laid out may apply to girls as well, although empirical studies with girls would be

required to examine the degree to which mechanisms operate similarly for boys and girls.

2. The additional two contrasts can be calculated based on the presented coefficients. "During versus never" is the sum of

the "during versus before" and "before versus never" coefficients. "After versus never" is the sum of the "after versus before"

and "before versus never" coefficients. All of these coefficients are positive and significant for the models reported in Tables

1 and 2. These coefficients are not presented in the tables because they combine selection and socialization, and thus are

of lesser theoretical interest.

3. The sample size is reduced when we control for peer delinquency due to missing values on these variables. We

estimated the models in Tables 1 and 2 on this reduced sample size to evaluate whether the difference incoefficients

between Table 4 and Tables 1 and 2 is merely due to differences in the samples. The percent reduction in coefficients is

highly similar when we use these new results: 27 percent to 48 percent reduction in the "during versus before" coefficients

and 13 percent to 65 percent reduction in the "during versus after" coefficients when we control for peer delinquency.

References
REFERENCES

Achenbach, Thomas M.

1991a Manual for the Child Behavior Checklist/4-18 and 1991 Profile. Burlington: University of Vermont, Department of

Psychiatry.

1991b Manual for the Teacher's Report Form and 1991 Profile. Burlington: University of Vermont, Department of Psychiatry.

1991c Manual for the Youth Self-Report and 1991 Profile. Burlington: University of Vermont, Department of Psychiatry.

Allison, Paul D. and Richard P. Waterman

2002 Fixed-effects negative binomial regression models. Sociological Methodology 32:247-265.

Ball, Richard A. and G. David Curry

1995 The logic of definition in criminology: Purposes and methods for defining gangs. Criminology 33:225-245.

Battin, Sara R., Karl G. Hill, Robert D. Abbott, Richard F. Catalano and J. David Hawkins

1998 The contribution of gang membership to delinquency beyond delinquent peers. Criminology 36:93-115.

Battin-Pearson, Sara R., Terence P. Thornberry, J.David Hawkins and Marvin D. Krohn

1998. Gang Membership, Delinquent Peers, and Delinquent Behavior. OJJDP Juvenile Justice Bulletin (October). U.S.

Department of Justice, Office of Justice Programs, Office of Juvenile Justice and Delinquency Prevention.

Blumstein, Alfred and Allen J. Beck

1999 Population growth in U.S. prisons, 1980-1996. In Michael Tonry and Joan Petersilia (eds.), Crime and Justice: Prisons,

Vol. 26. Chicago: University of Chicago Press.

Copeland, Dave

2001 Explanations vary over year's increased rate of homicide. Pittsburgh Tribune-Review (July 30, 2001).

Curry, G. David

2000 Self-reported gang involvement and officially recorded delinquency. Criminology 38:1253-1274.

Curry, G. David and Irving A. Spergel

1992 Gang involvement and delinquency among Hispanic and African-American adolescent males. Journal of

Research in Crime and Delinquency 29:273-291.

Craig, Wendy M., Frank Vitaro, Claude Gagnon and Richard E. Tremblay

2002 The road to gang membership: characteristics of gang and nongang members from ages 10 to 14. Social

Development 11:53-68.

Dukes, Richard L., Ruben O., and Judith A. Stein

1997 Precursors and consequences of membership in youth gangs. Youth and Society 29:139-165.

Elliott, Delbert S., David Huizinga and Suzanne S. Ageton

1985. Explaining delinquency and drug use. Beverly Hills, Calif.: Sage Publications.

Esbensen, Finn-Aage and David Huizinga

1993 Gangs, drugs, and delinquency in a survey of urban youth. Criminology 31:565-589.

Fergusson, David M., Nicola R. Swain-Campbell, and L. John Horwood

2002 Deviant peer affiliations, crime and substance use: a fixed effects regression analysis. Journal of Abnormal Child

Psychology, 30:419-430.

Hagedorn, John M.

1994 Neighborhoods, markets, and gang organization. Journal of Research in Crime and Delinquency 31:264-294.

Hausman, Jerry, Bronwyn H. Hall, and Zvi Griliches

1984 Econometric models for count data with an application to the patents-R&D relationship. Econometric 52:909-938.

Henry, David B., Patrick H. Tolan and Deborah Gorman-Smith

2001 Longitudinal family and peer group effects on violence and nonviolent delinquency. Journal of Clinical Child

Psychology 30:172-186.

Howell, James C.

1999 Youth gang homicides: a literature review. Crime and Delinquency 45:208-241.

1998 Youth Gangs: An Overiew. Juvenile Justice Bulletin (August) Washington, D.C.: U.S. Department of Justice, Office of

Juvenile Justice and Delinquency Prevention.

Huff, Ronald C.

1996 The criminal behavior of gang members and nongang at-risk youth. In C. Ronald Huff (ed.), Gangs inAmerica (2nd

edition). Thousand Oaks, Calif.: Sage Publications.

Huizinga, David and Delbert S. Elliott

1986. Reassessing the reliability and validity of self-report measures. Journal of Quantitative Criminology, 2, 293-327.

Kelly, Jack and Torsten Ove

1999 Youth violence found to spring from gangs. Pittsburgh PostGazette (Sunday, September 5).

Klein, Malcolm W.

1997 The American Street Gang: Its Nature, Prevalence, and Control. Oxford: Oxford University Press.

Lahey, Benjamin B., Rachel A. Gordon, Rolf Loeber, Magda Stouthamer-Loeber and David P. Farrington

1999 Boys who join gangs: a prospective study of predictors of first gang entry. Journal of Abnormal Child Psychology

27:261-276.

Loeber, Rolf, David P. Farrington, Magda Stouthamer-Loeber, and Willmet B. Van Kammen

1998 Antisocial Behavior and Mental Health: Explanatory Factors in Childhood and Adolescence. Mahwah, N.J.: Erlbaum.

Long, J. Scott

1997 Regression Models for Categorical and Limited Dependent Variables. Thousand Oaks, Calif.: Sage Publications.

Mamula, Kris B.

1997 Police alliance curbs murder, drugs, despair. Pittsburgh Tribune-Review (November 23, 1997).

Thornberry, Terence P.

1998 Membership in youth gangs and involvement in serious and violent offending. In Rolf Loeber and David Farrington

(eds.), Serious and Violent Juvenile Offenders: Risk Factors and Successful Interventions. Thousand Oaks, Calif.: Sage

Publications.

Thornberry, Terence P., B. Bjerregaard and William Miles

1993 The consequences of respondent attrition in panel studies: a simulation based on the Rochester Youth Development

Study. Journal of Quantitative Criminology 9:127-158.

Thornberry, Terence P., Marvin D. Krohn, Alan J. Lizotte and Deborah Chard-Wierschem

1993 The role of juvenile gangs in facilitating delinquent behavior. Journal of Research in Crime and Delinquency 30:55-87.

Thornberry, Terence P., Alan J. Lizotte, Marvin D. Krohn, Margaret Farnworth and Sung Joon Jang

1994 Delinquent peers, beliefs, and delinquent behavior: A longitudinal test of interactional theory. Criminology 32:47-83.

Thornberry, Terence P., Marvin D. Krohn, Alan J.Lizotte, Carolyn A. Smith and Kimberly Tobin

2003 Gangs and Delinquency in Developmental Perspective. Cambridge: Cambridge University Press.

Warr, Mark

2002 Companions in Crime. Cambridge: Cambridge University Press.

Wooldridge, Jeffrey M.

2002 Econometric Analysis of Cross Section and Panel Data. Cambridge: MIT Press.

AuthorAffiliation

RACHEL A. GORDON*

University of Illinois at Chicago

BENJAMIN B. LAHEY

University of Chicago

ERIKO KAWAI

Chicago, IL

ROLF LOEBER & MAGDA STOUTHAMER-LOEBER

University of Pittsburgh

DAVID P. FARRINGTON

Cambridge University

Gordon, R. A., Lahey, B. B., Kawai, E., Loeber, R., & al, e. (2004). ANTISOCIAL BEHAVIOR AND YOUTH GANG MEMBERSHIP:
SELECTION AND SOCIALIZATION. Criminology, 42(1), 55-87. Retrieved from
http://search.proquest.com/docview/220708441?accountid=37714

Gang involvement and delinquency in a middle school population

 Curry, G DavidAuthor Information ; Decker, Scott HAuthor Information ; Egley, Arlen, Jr. Justice Quarterly : JQ 19.2
(Jun 2002): 275+.

1. Full text
2. Full text - PDF
3. Abstract/Details
Hide highlighting
Abstract
TranslateAbstract

The relationship between self-reported gang involvement and self-reported delinquency has been confirmed in a number of

studies. However, there have been fewer studies of the relationship between self-reported ganginvolvement and officially

recorded delinquency. This article examines variation in self-reported ganginvolvement, operationalized as three distinct

categories--no involvement, gang involvement but not membership, and gang membership--and its relation to both self-

reported and officially recorded delinquency for a population of middle school youths.

Full text
TranslateFull text
Turn on search term navigation

Headnote

GANG INVOLVEMENT AND DELINQUENCY IN A MIDDLE SCHOOL POPULATION*

Headnote

The relationship between self-reported gang involvement and self-reported delinquency has been confirmed in a number of

studies. However, there have been fewer studies of the relationship between self-reported ganginvolvement and officially

recorded delinquency. This article examines variation in self-reported ganginvolvement, operationalized as three distinct

categories-no involvement, gang involvement but not membership, and gang membership-and its relation to both self-

reported and officially recorded delinquency for a population of middle school youths.

The relationship between self-reported gang membership and self-reported delinquency has been widely reported by

researchers and repeatedly substantiated by empirical research (Spergel, 1990). This body of research has consistently

found a positive correlation between self-reported gang membership and self-reported delinquency (Battin, Hill, Abbott,

Catalano, & Hawkins, 1998; Esbensen & Huizinga, 1993; Esbensen & Winfree, 1998; Fagan, 1990; Howell, 1994; Huff,

1996; Thornberry, 1998; Thornberry, Krohn, Lizotte, & Chard-Wierschem, 1993). Similarly, the relationship between self--

reported delinquency and officially recorded delinquency (Elliott, Huizinga, & Morse 1987; Farrington, 1989; Hartstone &

Hansen, 1984; Hawkins et al., 1998) has been systematically explored. The relationship between gang membership and

officially recorded delinquency has not been studied as thoroughly, however.

In the past decade, a considerable body of new evidence has been produced in support of this relationship. Using samples

of high school students and dropouts from Chicago, Los Angeles, and San Diego, Fagan (1989, 1990) found that youths

who identified themselves as gang members had higher levels of self-reported delinquent behavior than did their nongang

counterparts. From the longitudinal Denver Youth Survey, Esbensen and Huizinga (1993) concluded that gang members
self-report two to three times as much delinquency as do nongang members. In a comparable longitudinal study for a

representative sample of Rochester, New York, youths, Thornberry et al. (1993) reported that gang-involved youths were

significantly more likely to report involvement in violence and other forms of delinquency. By following youths over time, the

Rochester researchers revealed that gang membership is a transitional process, with delinquent activity increasing during

periods of self-reported gang membership and declining afterward. In another analysis of the Rochester data, Bjerregaard

and Lizotte (1995) found that self-reported gang members were twice as likely to report carrying firearms and three times as

likely to report selling drugs.

Using longitudinal data on a sample of Seattle youths, Battin et al. (1998) discovered that gang members reported

significantly more acts of violence and general delinquency than did nongang members. The researchers determined that

differences in self-reported delinquency were also reflected in juvenile court reports of delinquency. In addition, they

controlled for differences in having delinquent friends to reveal that gang membership is associated with delinquency above

and beyond the propensity to have delinquent friends (see also, Thornberry, 1998).

Despite this research, several important questions remain about the relationship between gang membership and

delinquency. The first has to do with levels of membership. It is clear that operationalizing gangmembership as a simple

dichotomy (i.e., member, nonmember) has many virtues from a measurement standpoint. However, this parsimony may

miss important substantive variation in levels of membership, especially with younger adolescents, whose membership may

be more transitory and reflect patterns of age-- graded friendships. For middle school adolescents (ages 12-14), their

involvement in gangs may reflect their diverse constellation of friends, some of whom may be gang members and some of

whom may not (Decker & Van Winkle, 1996; Hagedorn, 1998). Thus, it is possible to be involved in gang activities as a

consequence of one's friendships with gang members but not claim membership.

Although previous studies uncovered a relationship between self-reported gang membership and delinquency, research has

yet to examine whether the same relationship holds for those who are involved with gangmembers without claiming

membership themselves. We hypothesize that, given gang members' higher levels of involvement in crime, associating

with gang members has a stronger effect on self-reported involvement in crime than associating with delinquent nongang

peers would have. The longitudinal work of Battin et al. (1998) and Thornberry et al. (1998) provide the basis for our

speculation in this regard. This is one of the key issues this article examines. As dependent variables, we used both self--

reported and official measures of delinquency. It is a well-established principle in measurement theory that the strongest

indication of external validity occurs when two measures of the same trait from different measurement systems are related

(Campbell & Fiske, 1959). Measuring the level of delinquency by both self-reports and official records provides such a test

and follows Battin et al.'s (1998) practice. In this article, we use referral to juvenile court as the measure of official

involvement in delinquency.

GANG PROBLEMS IN ST. LOUIS

The most recent gang problem in St. Louis emerged in the late 1980s and the early 1990s. As with many U.S. cities, St.

Louis had been the scene of reported gang problems during other periods dating back more than a century. In the 19th

century, gangs of German and Irish youths fought in neighborhoods just north of what was then downtown. In his study

of gangs in early 20th-century Chicago, Thrasher (1927) referred to gangsexisting at that time in St. Louis. In the years

following World War II, the city experienced the first appearance of African American gangs, with gangs battling for the same

neighborhoods as their Euro-American predecessors and distinguishing their rivalries with the same colors of blue and red.

Still, when W. Miller (1975) studied St. Louis in the 1970s, he concluded that the city did not have a gang crime problem. By

the late 1980s, St. Louis was identified by a national study (Spergel & Curry, 1993) as having an "emerging" gangproblem.

By 1991, St. Louis made its first contribution to a national tabulation of gang statistics. In that year, the St. Louis Police

Department reported 33 gangs and 8 gang homicides. During the mid-1990s, gang-related homicides had grown to be a

significant portion of all homicides in St. Louis (Rosenfeld, Bray, & Egley,1999). By this time, Decker and Van Winkle (1996)

had undertaken their field study of 99 active gangmembers that would reveal gangs in St. Louis to be disorganized yet

extremely violent.

THE STRUCTURE OF GANGS AND PATTERNS OF GANG INVOLVEMENT

Theoretical patterns of gang involvement are closely related to theories about gang structure. Hypothetical models

of gang structure range from highly structured, hierarchical organizations to loosely structured manifestations of collective

behavior. At one end of this continuum are the observations of Sanchez-Jankowski (1991), who portrayed gangs as

hierarchical organizations with clearly established leaders and rational goals and actions. Taylor (1990) described some of

the gangs he found in Detroit similarly and labeled them "corporate" gangs. Padilla (1992) examined the transformation of a

Chicago youth gang into a more rationally organized and efficient drug-selling enterprise.

These highly structured gangs are in contrast to the gangs observed by other researchers. For Thrasher (1927, p. 202),

the gang was, above all, a primary group based on informality and face-to-- face association. The conflict gangs observed

by Spergel (1964) were neither "permanent" nor "stable." The gangs that Short and Strodtbeck (1965, p. 196) studied were

"very fluid," and in the view of the members interviewed by Suttles (1968, p. 176), the gang was a "happy coincidence" and

"spontaneous cohesion." More recently, Hagedorn (1998) depicted Milwaukee gangs as dynamic, age-- graded

subgroups in which the informal patterns of association varied over time. Participation in and the level of organization of St.

Louis gangs were portrayed by Decker (1996) as increasing and decreasing in conjunction with patterns of violence. Decker

referred to gangs as forms of collective behavior, rather than complex organizations. In a study of organizational

structure in four gangs identified by the police as potential organized crime groups in Chicago and San Diego, Decker,

Bynum and Weisel (1998) found that only one of the gangs displayed any kind of sustained organizational structure, and

that gang fell short of what could be called "organized crime."

Different theoretical processes of how youths become gang members are associated with the highly organized and the

loosely organized perspectives of gang organization. Rationally organized gangs with well-defined structures suggest
rational processes for admitting or recruiting members. As Sanchez Jankowski (1991) reported, becoming a member of a

well-organized gang required two explicit decisions. The individual decided to become a member, and the gang decided to

permit membership. Gang members joined gangsprimarily for material gain but also for recreation and commitment to

community. On the other side of the process, gangs recruited members as needed to strengthen existing resources or to

gain organizational control of new resources. In this model, the line between a member and a nonmember was distinct.

Under the theoretical model of loosely organized gangs, the nature of membership is quite different. From Spergel's (1964),

research in New York, "Identification as a member of a gang ... was not a stable, permanent, once-and-for-all social fact (p.

66) .... One youngster might, for a relatively short period of time, 'hang around' a group and be recognized almost

immediately as a member of that group. Another individual would be present a great deal of the time with the group and be

regarded clearly as a non-member or as a member of another group" (pp. 66-67). Vigil (1988), for Los Angeles barrio gangs,

and Decker and Van Winkle (1996, p. 67), for emerging St. Louis gangs, portrayed the process of joining a gang as gradual,

sometimes taking years. According to Decker and Van Winkle (p. 68), "On average, members of our sample heard about

their gang while they were twelve, started hanging out with gang members at thirteen, and joined before their fourteenth

birthday. This suggests a gradual process of affiliation rather than one of active recruitment." From her ethnographic study of

female gang members in Columbus, Ohio, and St. Louis, J. Miller (2001) reached a similar conclusion of

gradual gang involvement for girls.

The model of loosely organized gangs with gradual processes of affiliation is conducive to the identification of different levels

of gang membership. Thrasher (1927) spoke of the inner circle, the rank and file, and fringe members. Vigil (1988) classified

members as regular, peripheral, and temporary. Hagedorn (1998, p. 90) noted a division between "main groups" and

"wanna be's" in Milwaukee gangs, but added, "A 'wanna be' this week may be in the 'main group' next week."

How youths come to be involved in gangs has important implications for variation in gang behavior. When gangs are highly

organized and membership is well defined, it would be expected that one would find members and nonmembers.

When gangs are loosely structured manifestations of collective behavior, one should find, in addition to members and

nonmembers, a population of youths who are neither completely members nor nonmembers, but

something in between. In this study of St. Louis, where gangs have been observed to be loosely organized, it should be

possible to identify youths with differential levels of ganginvolvement somewhere short of gang membership. In an earlier

analysis, Decker and Curry (2000) found that distinguishing between levels of gang involvement was

important in understanding youths' gang-related perceptions and experiences. In this article, we explore the relationship

between different levels of ganginvolvement and delinquency.

RESEARCH DESIGN

The design is one that has been frequently used in studies of delinquency (Curry & Spergel, 1992; Esbensen & Huizinga,

1993; Esbensen & Winfree, 1998; Fagan, 1990). A population of students attending middle school classes was surveyed

within a classroom setting and asked about their involvement in a variety of behaviors. In cooperation with the St. Louis
Public Schools, three middle schools were selected on the basis of their proximity or distance from concentrations

of gang homicides. Two of these three schools were in poor neighborhoods with considerable gang activity. The third

was in a predominantly middle-class neighborhood with little or no gang activity, and the majority of students in this school

did not reside in the surrounding community. Two additional schools that were included in the study were the St. Louis

Public Schools "Tri-A" Academies. These two academies are reserved for students with disciplinary or legal problems. By

combining these three types of schools (from poor neighborhoods with gang problems, from a middle-class neighborhood

with few gang problems, and from the Tri-A Academies), we believe that the respondents reflect the general character of

students in this urban district.

The survey was administered in the spring semester of the 1995-96 school year. The response rate among students

attending school was high, as is inevitably the case, when students and teachers are offered an alternative to regular class

activities. Participating schools were paid $20 per completed survey for the purchase of educational supplies. Passive

parental consent was used. Only 17 parents requested that their children not participate in the survey. In one classroom, all

students collectively refused to participate; the research assistants reported that students (who may have

been gang members) announced that decision for their classmates. The greatest detriment to participation was school

attendance. Although we had access to school records and were able to code each instrument with students' identification

numbers, school officials could not tell us whether students were absent or had changed schools. In other school

studies in which absentees could be clearly identified (Curry & Spergel, 1992), absentees were more likely to have school

disciplinary and delinquency records than were attending students.

Measurement

Gang membership was measured by the student's answers to the questions, "Do you belong to a gang?" and "Have you

been a member of a gang in the past?" We treated affirmative answers to either question as evidence of ever having been a

member of a gang, following the practice of Esbensen and Winfree (1998) for

identifying gang membership in samples in which most respondents are younger than age 15. In this study, no additional

criteria were imposed on the self-reporting of gang membership. This follows the recommendation of Esbensen, Winfree,

He, and Taylor (2001, p. 124), who found self-nomination to be a "particularly robust measure of gang membership capable

of distinguishing gang from non-- gang youth." Intheir study, along a continuum of an increasingly restrictive definition

of gang membership, the greatest difference in gang behaviors and attitudes was observed between youths who had never

been gangmembers and youths who self-reported only current or prior gang involvement. In another study, Esbensen and

Huizinga (1993) required that respondents also answer one of two items about their group's illegal activity. Since our main

concern was to examine the relationship between the level of gang membership and delinquency, we did not impose any

requirement for illegal activity on our operational definition of gangmembership. This measurement strategy avoided the

potential tautology of establishing the relationship between membership and delinquency in our operational definition (Ball &

Curry, 1995; Short, 1998).

Gang involvement was measured by four of eight criteria used by Curry and Spergel (1992) to measure ganginvolvement

other than self-reported membership in a gang. Curry and Spergel found that self-reported measures of gang-related

behavior, independent of membership, were significantly related to levels of delinquency. The specific four measures we

used are shown in Table 1. Self-reported delinquency was measured by answers to a number of items asking if youths had

committed a particular delinquent act. Most of these items were taken from other instruments used to study self-- reported

delinquency. Officially recorded delinquency was measured by linking students' files to referrals to the St. Louis City Family

Court for either a status offense or a delinquency offense.1 All court referrals for delinquency were captured for each student

who completed a survey.

FINDINGS

Prevalence of Gang Membership

All the students at two of the middle schools and one of the academies were African American. At the third middle school,

the majority of students in the sample (62.1%) were white. Of the 533 respondents to the survey, 80 (15.0%) reported being

either currently or formerly a gang member. The prevalence of gangmembership is somewhat higher than the 10.6%

reported in the multisite, 11-city survey of youths conducted by Esbensen and colleagues (see Esbensen & Winfree, 1998).

This difference is most likely the result of the use of a more restrictive definition in the latter study. Elsewhere, using a similar

classification procedure as the present survey, Esbensen et al. (2001) reported that 16.8% of the sample responded as ever

having been a gang member.

Enlarge this image.

Table 1.

There were differences associated with gender and race. With regard to gender, 18.1% of the boys and 11.5% of the girls

reported ever being a gangmember (a statistically significant difference at the .05 level). These observed prevalence levels

of gang membership are slightly higher than the 13.6% for boys and 7.8% for girls reported by Esbensen and Winfree (1998)

and are, inpart, likely to be the result of the difference in classification procedures and the extent of gang problems in the

study sites. With regard to race, overall, only five white students (8.2%) identified themselves as ever having

been gang members. The expected values for white students were too small to compute a chi-square comparing the

prevalence of gang membership for whites and African Americans. The overall percentage of African American respondents

who reported ever having been gang members was 15.9%, but at the predominantly white school, 28.6% of the African

Americans did so (a statistically significant difference at the .05 level).2

Prevalence of Gang Involvement

Treating any one of the gang-related behaviors in Table 1 as being gang involved, it is possible to examine the proportion of

youths who reported never being a gang member who reported some level of ganginvolvement (see Table 2). Of the 453

respondents who reported never having been a gang member, 260 (57.4%) reported at least one kind of gang involvement.
With regard to gender, 59.1% of the boys and 55.6% of the girls who were never gang members reported some level

of gang involvement. This difference is not statistically significant and suggests that knowledge of and

involvement in gang behavior is quite high, even for nonmembers. With regard to race, 60.1% of the African American

youths and 37.5% of the white youths who were not gang members reported some level of gang involvement (a statistically

significant difference at the .01 level). The one Latino in the survey reported that he had friends who were gang members.

The finding that the majority of nonmembers reported some level of gang involvement underscores the pervasive effect

of gangs among middle school students in these neighborhoods. It suggests that the influence of gangs on nongang

members may be greater than previous studies have estimated.

Gang Involvement and Self-reported Delinquency

Separating respondents into gang members, nongang members, and nonmembers who have some level

of gang involvement gave us an opportunity to expand what we know about the status of being involved with yet not a

member of a gang. Given the empirically established relationship between gang membership and delinquency, any status

that is defined by the social space between gang member and nonmember should be differentially associated with

delinquency. In Table 3, nonmembers with gang involvement are labeled "ganginvolved."

The respondents were asked about their participation in a range of delinquent behaviors, including serious, minor, and drug

offenses. Table 3 shows the relationships between level of gang involvement and self-reported delinquency. For each of the

12 self-- reported delinquent offenses, the differences (as measured by gamma) across the three levels of gang involvement

are statistically significant at the .001 level. For all 12 self-reported delinquency items, the prevalence rate for gang-involved

youths is both greater than that of youths who were not gang involved and lower than that of those who self-

reported gang membership.3

Enlarge this image.

Table 2.

For all but one of the measures, the paired comparisons between nonmembers with no gang involvement and nonmembers

with gang involvement is statistically significant at the .05 level (at least). The exception is stealing something worth more

than $50, for which 3.6% of nonmembers not ganginvolved and 8.1% of the gang-involved nonmembers stand significantly

apart from the 30% reported by gang members. A smaller number of the self-reported delinquency items differentiate

between gang-involved nonmembers and gang members. Most notable, the prevalence rate for using violence against

another person among gang-involved youths (55.8%) is more than double that of youths who reported no gang involvement

(25.4%) and approaches that of self-reported gang members (65.0%).

Enlarge this image.

Table 3.

Gang Involvement and Victimization

The relationship between offending and victimization has been well established in research on delinquency (Lauritsen,

Sampson, & Laub, 1991) and has received support in the gang literature (Decker & Van Winkle, 1996; Moore, 1991). Decker

(1996) reported that more than a dozen of his 99 respondents had been killed within five years of the conclusion of his

study. These results show that gang members have violent victimization rates that exceed those of nonmembers.

Victimization was measured by self-reports of being threatened with a gun, shot at, or injured by gunshot. Table 3 illustrates

that gang membership is as strongly associated with victimization as offending. But at least in one case, having been shot

at, gang-involved nonmembers are more than three times as likely to report affirmatively than nonmembers with

no gang involvement. Those who reported ever having been a gang member were more than four times more likely than

nongang members with no gang involvement to have been victimized in each of the listed events. For example, 12.5% of

the gang members reported having been injured by gunshot, over four times that of non-gang-involved nonmembers (1.6%)

and almost three times the rate reported by gang-- involved nonmembers (4.2%).

Taken together, these findings are congruous with the contagion of violence purported to be associated with gang activity

(Decker, 1996). Initial acts of gang-related violence arouse and expand the willingness of others to engage in retaliatory

violence. Within this spatial concentration, gang members may often assume the role of victim, perpetrator, or witness to

violence. From the method of measurement used in this survey, it is impossible to conclude with certainty

that gang members in this sample were necessarily involved in gang-related violence. However, it is noteworthy that the

association among gang membership, delinquency, and victimization was observed and categorically supported among a

sample of young adolescent gangmembers.

Gang Involvement and Officially Recorded Delinquency

While recent research has continued to support the link between gang membership and self-reported offending, the link

between gang membership and officially recorded delinquency is not as frequently documented. From a survey of

young adolescents in Chicago, Curry (2000) found that among those who self-reported both gang involvement and

delinquency, over half were arrested for at least one delinquent act in the subsequent five years, and over half of those who

were so arrested were apprehended for involvement in a gang-motivated offense. Thus, these findings suggest that there is

a considerable overlap between youth-based surveys and official records regarding adolescents' gang membership and

delinquency.

From an analysis of the data of referrals to the St. Louis Family Court, it was possible to examine the relationship between

self-- reported gang membership and referrals to the court. All court referrals between September 1993 and December 31,

2000, were checked for matches with the respondents to the survey on child's name, date of birth, and address listed in the

school records. Respondents with either status-offense referrals or delinquency referrals between 1993 and 2000 were

identified. Although the survey data were essentially cross-sectional, it was possible to add some degree of chronological

structure to the analysis by classifying each referral as being before or after the administration of the survey.
Table 4 compares the court referral records for respondents who identified themselves as gang members, nonmembers with

some gang involvement, and nonmembers with no gang involvement. As should be expected, rates of referral prior to the

survey (when the respondents were younger) were appreciably lower than in the five years following the survey. The status-

offense referral rates for each of the three gang-involvement groups before the survey was administered were not

significantly different. In fact, the nonmembers who reported gang involvement in the survey had a slightly lower rate of

referrals for status offenses prior to the survey. The referral rates for delinquency before the survey varied in the ordinal

direction found for self-reported delinquency, increasing from nonmembers with no gang involvement to nonmembers

with gang involvement to self-reported members. The difference across the three categories was statistically significant at

the .05 level, but the differences between consecutive pairs of categories were not statistically significant.

Table 4 reveals that the same rank-order relationship holds between the three levels of gang involvement and officially

recorded delinquency. Nonmembers with gang involvement were more likely to have been referred for status offenses than

were nonmembers with no gang involvement. Nonmembers with gang involvement were more likely to have been referred

for delinquency after the survey than were nonmembers with no ganginvolvement. Just over one-third of the gang-involved

youths (36.9%) were referred to the juvenile court for a delinquent offense, well below the prevalence rate for gang members

(48.8%), yet appreciably greater than that of youths who were not gang involved (24.9%). The differences between groups

within each type of referral after the survey are statistically significant. On the whole, these results provide strong evidence

of an intermediary status between gang membership and nonmembership, whereby even a minimal amount of

involvement in gang-related behaviors clearly separates these youths from their uninvolved peers in terms of self-reported

and officially recorded delinquency.

Enlarge this image.

Table 4.

The proportions referred in the five years following the survey were significantly different at the .001 level. The proportion of

non-- gang youths referred was four times that referred prior to the survey, as was the proportion of gang-involved

nonmember youths. While the increase in the proportion of gang members who were referred after the survey was only

three times the proportion of members referred prior to the survey, the amplification of the greater initial proportion

resulted in a much larger gap between the nongang and gang-involved youths. This difference in officially recorded

delinquent offending underscores the importance of self-reported gang membership as a predictor of delinquency

(Esbensen et al., 2001; Esbensen & Winfree, 1998).

These results unequivocally demonstrate that increased involvement in gang-related behaviors for adolescents significantly

enhances the risk of delinquency and victimization. This finding highlights both the significance of gang involvement for

offending and the role of engaging in gang behavior, short of full-fledged membership, in enhancing offending. The

increased levels of offending associated with engaging in gangbehaviors are quite dramatic, underscoring the need to

address both gang-involved and marginally gang-involved youths.

CONCLUSION

This article has demonstrated that the relationship between self-reported gang membership and delinquency found in other

cities holds for St. Louis.4 Since the survey was conducted in middle schools, it is safe to conclude that the relationship

between gang membership and delinquency emerges in early adolescence. Whereas most previous surveys have shown

the relationship between self-reported gang involvement and self-reported delinquency, this study also found a statistically

significant relationship between self-reported gang membership and referral to the juvenile court for delinquency.

In addition to identifying a distinction between gang members and nonmembers with respect to delinquency, the survey

identified an interstitial status between full-fledged gang membership and nonmembership. Nonmember gang involvement

was measured using a set of criteria developed by Curry and Spergel (1992) instudying Chicago youths. The St. Louis

Police Department Gang Unit uses the designation "gang associate" to identify suspects who are known to have some

association with gangs but are not known to be members. We think that this designation may be useful in research. It is

comparable to labels used in other research, such as "peripheral members," "marginal members," and "wanna be's." The

significance of this gang-associate status is demonstrated by its strong relationship to delinquency. For a range of measures

of self-reported and officially recorded delinquency, the prevalence of offending for this group was shown to be between that

of gang members and nonmembers who were not involved in gang behavior. The delinquent involvement of this group of

associate members lies along a scale between full-fledged members and nonmembers.

A longitudinal survey design is required to discover if identified gang associates are in a transitional phase that

would, in some cases, lead to full gang membership or if their association with gangs would fade. Still, one may draw

conclusions about the social structure and dynamics of gang involvement. While the study found that self-

perceived gang membership is an important predictor of delinquency, it demonstrated that marginal gang involvement is

associated with a level of risk of delinquency and violence that is greater than that experienced by nonmembers.

IMPLICATIONS FOR RESEARCH AND POLICY

This study joins others (Esbensen & Winfree, 1998) in showing that the relationship between gangmembership and

delinquency emerges in early adolescence. The findings also indicate that engaging in gangbehaviors is strongly and

consistently associated with involvement in a variety of forms of delinquency and victimization. Thus, associating

with gang members ("hanging out") and engaging in gang behaviors (adopting gang mannerisms) both have important

consequences for delinquency, even for nongang members. It is not necessary for an individual to self-

report gang membership to experience the delinquency- enhancing effects of gang membership. This finding supports the

need for early prevention and intervention programs with youths and families, especially in neighborhoods with high levels

of gang membership or in families with older siblings who belong to gangs. This finding also underscores the need

for gang intervention programs that deal with actively involved gang members, as well as individuals who are peripheral to

the gang. Minimizing the influences of gang members on peripheral members or nonmembers should pay dividends for both

delinquency prevention and intervention efforts. We also believe that this finding should be particularly relevant for
programming directed at young girls who may not (yet) be active gang members, but who may associate with such

individuals either through familial or associational ties.

The image of the gang that is supported by these findings is one of a loosely structured organization. As was noted in prior

studies in St. Louis (Decker, 1996; Decker & Van Winkle, 1996), gang activity can be best viewed as a form of collective

behavior, with many youths perceiving themselves, at least at some point in time, as gang members. However, other youths

who do not consider themselves ever to have been gang members continue to associate with gangs and gang members

and are differentially involved in the overall patterns of gang-related crime and delinquency. This finding may also be true for

individuals who once were gangmembers and have cut their formal ties with the gang, but still maintain friendships

with gang members and occasionally engage in gang behavior with them.

Suppression approaches that treat gangs as organized crime groups have been shown to be inappropriate responses to

most gangs (Decker et al., 1998). The more loosely structured nature of gang organization calls for more comprehensive

responses that require a range of coordinated prevention, intervention, and suppression activities. This finding seems

particularly appropriate for adolescents of the age range in our study. The emphasis in gang programming over the past

decade has been on suppression (Decker & Curry, 2000; Spergel & Curry, 1993). Such approaches have gang members as

their focus and hence may miss a significant number of gang-involved youths who may not be identified as gang members

or may not identify themselves as such, particularly those who are younger adolescents. Suppression efforts appear to be

operating at maximum capacity, that is, there appears to be little excess capacity in law enforcement efforts to

address gangs. The findings about suppression highlight the need for programs that focus on the broad range of

differentially involved gang youths and their range of different problems and needs.

Footnote

1 These data were provided through the cooperation of the St. Louis Family Court.

Footnote

2 The source of this difference remains unknown to us, but it may be that some African American students perceive

themselves differently at an integrated school or feel the need to affiliate with a gang (for protection) in such a setting.

Footnote

3 From a longitudinal study of Seattle youths, Battin et al. (1998) concluded that the effect of gangmembership on self-

reported delinquency remained statistically significant even when having delinquent friends was controlled. To examine the

possibility that our measure of gang involvement may simply be a proxy for delinquent friends, we controlled for four

variables related to information about friends' delinquency. These variables were positively related to self-reported and

officially recorded delinquency as well as gang-involvement status. In a series of logistic regression analyses (17), we

regressed each of the last eight self-reported delinquency measures, the three victimization measures, and the two

postsurvey measures of family court referrals on gang involvement (as a three-value categorical variable) and the report of

delinquent friends. For only one of the self-reported delinquency measures, stealing something worth less than $50, was the
coefficient for delinquent friends statistically significant when gang involvement was not. In the others,

either gang involvement alone was a statistically significant predictor of the outcome variable or both ganginvolvement and

having delinquent friends were statistically significant predictors. Therefore, ganginvolvement had an independent effect on

delinquency beyond that attributed to delinquent friends. The table of results is available from the authors.

Footnote

4 It should be noted that this was not a random sample, so it is not feasible to generalize to other populations.

References

REFERENCES

References

Ball, R. A., & Curry, G. D. (1995). The logic of definition in criminology: Purposes and methods for defining "gangs."

Criminology, 33, 225-245.

Battin, S. R., Hill, K G., Abbott, R. D., Catalano, R. F., & Hawkins, J. D. (1998). The contribution of gangmembership to

delinquency beyond delinquent friends. Criminology, 36, 93-115.

Bjerregaard, B., & Lizotte, A. J. (1995). Gun ownership and gang membership. Journal of Criminal Law and Criminology, 86,

37-58.

Campbell, D., & Fiske, D. (1959). Convergent and discriminant validation by the multi-trait-multi-method matrix.

Psychological Bulletin, 56, 81-105.

Curry, G. D. (2000). Self-reported gang involvement and officially recorded delinquency. Criminology, 38, 100-118.

Curry, G. D., & Spergel, I. A. (1992). Gang involvement and delinquency among Hispanic and African-

American adolescent males. Journal of Research on Crime and Delinquency, 29, 273-291.

Decker, S. H. (1996). Collective and normative features of gang violence" Justice Quarterly, 13, 243-264.

Decker, S. H., Bynum, T., & Weisel, D. (1998). A tale of two cities: Gangs as organized crime groups. Justice Quarterly, 15,

395-425.

Decker, S. H., & Curry, G. D. (2000). Addressing key features of gang membership: Measuring the involvement of young

members. Journal of Criminal Justice, 28, 473-482.

References

Decker, S. H., & Van Winkle, B. (1996). Life in the gang: Family, friends, and violence. New York: Cambridge University

Press.
Elliott, D. S., Huizinga, D., & Morse, B. (1987). Self-reported violent offending: A descriptive analysis of juvenile violent

offenders and their offending careers. Journal of Interpersonal Violence, 1, 472-514.

Esbensen, F.-A., & Huizinga, D. (1993). Gangs, drugs, and delinquency in a survey of urban youth. Criminology, 31, 565-

587.

Esbensen, F.-A., & Winfree, L. T. (1998). Race and gender differences between gang and nongang youths: Results from a

multisite survey. Justice Quarterly, 15, 505523.

References

Esbensen, F.-A., Winfree, L. T., He, N., & Taylor, T. J. (2001). Youth gangs and definitional issues: When is a gang a gang,

and why does it matter? Crime and Delinquency, 47, 105-130.

Fagan, J. (1989). The social organization of drug use and drug dealing among urban gangs. Criminology, 27, 633-669.

Fagan, J. (1990). Social processes of delinquency and drug use among urban gangs. In C. R. Huff (Ed.), Gangs in America

(pp. 183-219). Thousand Oaks, CA: Sage.

Farrington, D. P. (1989). Early predictors of adolescent aggression and adult violence. Violence and Victims, 4, 79-100.

Hagedorn, J. M. (1998). People and folks: Gangs, crime, and the underclass in a rustbelt city (2nd ed.). Chicago: Lakeview

Press.

Hartstone, E., & Hansen, K. V. (1984). The violent juvenile offender: An empirical portrait. In R. A- Mathias, P. De Muro, & R.

S. Allinson (Eds.), Violent juvenile offenders (pp. 83-112). San Francisco: National Council on Crime and Delinquency.

Hawkins, J. D., Herrenkohl, T., Farrington, D. P., Brewer, D. Catalano, R. F., & Harachi, T. W. (1998). A review of predictors

of youth violence. In R. Loeber & D. P. Farrington (Eds.), Serious and violent juvenile offenders: Risk factors and successful

interventions (pp. 106-146). Thousand Oaks, CA: Sage.

Howell, J. (1994). Gangs fact sheet. Washington, DC: U.S. Department of Justice, Office of Juvenile Justice and

Delinquency Prevention.

Huff, C. R. (1996). The criminal behavior of gang members and nongang at-risk youth. In C. R. Huff

(Ed.), inGangs in America (2nd ed., pp. 75-102). Thousand Oaks, CA: Sage.

Lauritsen, J. L., Sampson, R. J., & Laub, J. H. (1991). The link between offending and victimization among adolescents.

Criminology, 29, 265-292.

Miller, J. (2001). One of the guys: Girls, gangs, and gender. New York: Oxford University Press.

References
Miller, W. B. (1975). Violence by youth gangs and youth groups as a crime problem in major American cities. Washington,

DC: U.S. Government Printing Office.

Moore, J. (1991). Going down to the barrio: Homeboys and homegirls in change. Philadelphia: Temple University Press.

Padilla, F. M. (1992). The gang as an American enterprise. New Brunswick, NJ: Rutgers University Press.

Rosenfeld, R., Bray, T. M., & Egley, Jr., A. (1999). Facilitating violence: A comparison of gang-motivated, gang-affiliated,

and nongang youth homicides. Journal of Quantitative Criminology, 15, 495-516.

Sinchez-Jankowski, M. S. (1991). Islands in the street: Gangs and American urban society. Berkeley: University of California

Press.

Short, J. F. (1998). The level of explanation problem revisited: The American Society of Criminology 1997 presidential

address. Criminology, 36, 3-36.

Short, J. F., & Strodtbeck, F. L. (1965). Group process and gang delinquency. Chicago: University of Chicago Press.

Spergel, I. A. (1964). Racketville, slumtown, haulburg. Chicago: University of Chicago Press.

Spergel, I. A. (1990). Youth gangs: Continuity and change. In M. Tonry & N. Morris (Eds.), Crime and justice: A review of

research, Vol. 12 (pp. 171-275). Chicago: University of Chicago Press.

Spergel, I. A., & Curry, G. D. (1993). The National Youth Gang Survey: Research and development process. In A. Goldstein

& C. R. Huff (Eds.), Gang intervention handbook (pp. 359-400). Champaign-Urbana, IL: Research Press.

Suttles, G. (1968). The social order of the slum. Chicago: University of Chicago Press. Taylor, C. S. (1990). Dangerous

society. East Lansing: Michigan State University Press.

References

Thornberry, T. (1998). Membership in youth gangs and involvement in serious and violent offending. In R. Loeber & D. P.

Farrington (Eds.), Serious and violent juvenile offenders: Risk factors and successful interventions (pp. 147-166). Thousand

Oaks, CA Sage.

Thornberry, T., Krohn, M.D., Lizotte, A. J., & Chard-Wierschem, D. (1993). The role of juvenile gangs infacilitating delinquent

behavior. Journal of Research in Crime and Delinquency, 30, 55-87.

Thrasher, F. (1927). The gang: A study of 1,313 gangs in Chicago. Chicago: University of Chicago Press.

Vigil, J. D. (1988). Barrio gangs: Street life and identity in southern California. Austin: University of Texas Press.

AuthorAffiliation
 G. DAVID CURRY**

SCOTT H. DECKER***

ARLEN EGLEY, JR.****

University of Missouri-St. Louis

Curry, G. D., Decker, S. H., & Egley, A.,Jr. (2002). Gang involvement and delinquency in a middle school population. Justice
Quarterly : JQ, 19(2), 275-275+. Retrieved from http://search.proquest.com/docview/228162850?accountid=37714

"Can I Make Any Difference?" Gang Affiliation, the School-to-

Prison Pipeline, and Implications for Teachers
Gass, Kayla M; Laughter, Judson CAuthor Information . The Journal of Negro Education 84.3 (Summer 2015): 333-
347.
Abstract
TranslateAbstract

This article reports on one teacher's year-long project seeking to disrupt the school-to-prison pipeline through class dialogue

around gang affiliations. The review of literature defines the school-to-prison pipeline and its connections to gang affiliation.

The study engaged methods of qualitative action research with seven male participants between the ages of 15 and 16 in a

high school English classroom where they engaged indialogue around gang affiliation and their experiences of criminalized

schools. Findings provided some hope that teachers might play a role in disrupting the school-to-prison pipeline but that

strong student-teacher relationships are necessary. Implications describe how other teachers wanting to disrupt the school-

to-prison pipeline might use curricula that spark discussions around gangs.

Full Text
TranslateFull text
Turn on search term navigation

Headnote

This article reports on one teacher's year-long project seeking to disrupt the school-to-prison pipeline through class dialogue

around gang affiliations. The review of literature defines the school-to-prison pipeline and its connections to gang affiliation.

The study engaged methods of qualitative action research with seven male participants between the ages of 15 and 16 in a

high school English classroom where they engaged indialogue around gang affiliation and their experiences of criminalized

schools. Findings provided some hope that teachers might play a role in disrupting the school-to-prison pipeline but that

strong student-teacher relationships are necessary. Implications describe how other teachers wanting to disrupt the school-

to-prison pipeline might use curricula that spark discussions around gangs.

Keywords: School-to-prison pipeline; students and gangs; action research

After one of my students correctly answered a question that everyone else was struggling with, I eagerly approached him to

give him a high five or fist bump, which was apparently not hip enough for him. Instead, he suggested we do a different kind

of handshake involving a lot of different twists, finger combinations, and intricate patterns. By being included in this
handshake, I felt like after two short weeks my students were beginning to accept me and were welcoming me into their

personal lives. Little did I know that I had just participated in a gang affiliated handshake; the class erupted in laughter and I

was told I was gangbangin' and officially a part of the clique, aware of their secret knowledge (Researcher Journal, 2

October).

In this situation, what should a teacher do? Clearly the students found this moment funny and interactive, an invitation into

their lives through a display of extra-curricular language and action. However, a teacher might be concerned that

handshakes and signs representing gang affiliation might be used by others to criminalize students within a school setting

(Butler et al., 2012; Gregory, Skiba, & Noguera, 2010; Noguera, 2003; Skiba, Rausch, & Ritter, 2004). The question, What

should a teacher do?, when gang issues and identities enter the classroom only scratches the surface: Are my students

involved in gangs? How might that involvement (real or perceived) affect their education? How might that involvement (real

or perceived) affect the way disciplinary policies push them into the school-to-prison pipeline? How might I disrupt the

school-to-prison pipeline? Can I make any difference?

Over the course of a year, the researcher wrestled with these questions, engaging action research to explore how to disrupt

the school-to-prison pipeline through a curriculum purposively designed to generate dialogue

around gangs and gang affiliations. This report of research opens with a review of literature defining the school-to-prison

pipeline and its connections to gang affiliation. The study engaged methods of qualitative action research to record and

analyze the researcher's experiences as a high school English teacher indialogue around gang affiliation

with adolescent male students experiencing criminalized schools. Findings provided some hope that teachers might play a

role in disrupting the school-to-prison pipeline but that strong student-teacher relationships are necessary. Implications

describe how other teachers wanting to disrupt the schoolto-prison pipeline might use curricula that spark discussions

around gangs.

DEFINING GANG

There is no internationally accepted definition of the word gang (Howell & Griffiths, 2016), but even in this article it has

already been used it as if a reader will intuit its meaning. Current federal law defines a criminal street gang as an

ongoing group, club, organization, or association of five or more persons that has as one of its primary purposes the

commission of... a federal felony involving a controlled substance...; a federal felony crime of violence that has as an

element the use or attempted use of physical force against the person of another; and a conspiracy to commit an offense.

(18 USC 521 [a, c])

While researchers are not always specific about drug crime or violence in using the word gang, the historical development

of gangs from the 1980s to the present must recognize the effect of the international drug trade on local gangs (Toldson,

2012). Other labels used in the research include adolescent gangs (Mallett, 2015), street gangs (Toldson, 2012), and

delinquent peer groups (Barnert et al., 2015). In this article, the word gangconnotes all of these possibilities because the
participants in this study used the word gang to describe such groups. When referring to students with gang affiliations, the

students claimed those affiliations; there were no assumptions that the participants were involved in delinquent or criminal

activity.

LITERATURE REVIEW

Research on the school-to-prison pipeline often focuses on disproportionate application of criminalizing discipline policies

(Butler et al., 2012; Toldson, 2012; Toldson & Morton, 2011), what has been termed the discipline gap (Gregory, Skiba, &

Noguera, 2010); however, the students at the center of these policies are often raced, classed, and gendered in ways that

make a future prison term (seemingly) inevitable (Gregory et al., 2010; Lea & Abrams, 2015; Mallett, 2015; Snapp et al.,

2015). Similarly, research on gangs often focuses on who instead of why; perceived gang members are catalogued by

ethnicity, age, and socioeconomic status rather than engaged as individuals responding to social systems that see them

through deficit paradigms (Barnert et al., 2015; Graham & Erwin, 2011; Lea & Abrams, 2015).

Historically, gang affiliation (whether actual or perceived) has been seen as a precursor to future incarceration and closely

related to criminalizing discipline policies. This review begins by defining the school-to-prison pipeline, including its historical

association with gangs, and then synthesizes research examining students and the ways and reasons they join gangs. Each

of these syntheses subsequently provided a foundation for the curriculum work described.

The School-to-Prison Pipeline

In the United States, the prison population continues to grow due to "tough on crime" policies (Mallett, 2015; Meiners &

Winn, 2010) like mandatory minimum sentencing and three-strikesand-you're-out laws (Butler et al., 2012; Davis, 2003;

Mauer & Chesney-Lind, 2003). While crime statistics do not directly implicate K-12 education, they are historically related:

The PIC [Prison Industrial Complex] signifies a particular relationship to education in the United States, as heavy

investment in a carceral state often parallels disinvestment in social welfare programs, including higher education. In the last

two decades, most states increased allocations for corrections exceeding increases for education. (Meiners & Winn, 2010,

p. 272)

In short, "priorities to incarcerate compete against priorities to educate" (Toldson & Morton, 2011, p. 2). Coincidental with the

rise in funding for incarceration are several school-disciplinary policies, such as zero tolerance and out-of-school

suspensions, that both mirror "tough on crime" policies and lead directly to increases in the potentiality of incarceration

(Butler et al., 2012; Gregory, Skiba, & Noguera, 2010; Mallett, 2015; Skiba et al., 2002).

Such disciplinary policies are the foundation of the school-to-prison pipeline: "policies and practices inschools that make it

more likely that students face criminal involvement with the juvenile courts than attain a quality education" (Mallett, 2015, p.

1). These policies and practices are associated with "lower academic achievement, less college readiness, increased
likelihood of becoming a push-out, and racial disparities which impede students from participating in the college pipeline"

(Valles & Villalpando, 2013, p. 262).

From its beginnings, the school-to-prison pipeline has been closely related to gang affiliation. Zero-tolerance policies first

emerged in the 1980s as part of federal drug and weapons policies at a time when emerging international drug markets

increased funding for and power of local gangs (Toldson, 2012). These tough on crime approaches perpetuated a myth of

juvenile gang violence, leading to the passage of the 1994 Gun-Free Schools Act, zero-tolerance discipline policies, and a

rise in the number of police officers in schools. While these policies quickly expanded to include subjective offenses (Triplett,

Allen, & Lewis, 2014), the first policies, in Kentucky and California, specifically called for "student expulsion for drug

and gang-related activity" (Mallett, 2015, p. 5).

When coupled with the social connotation that crime has a Black, male face (Armour, 2000; Toldson, 2012; Triplett, Allen, &

Lewis, 2014), the expansion of zero-tolerance policies for subjective offenses disproportionately affects Black male students:

Research findings indicate that minority students are often referred to the office for subjective offenses, and that racial

discipline disparities are more likely to be found in minor, subjective offense categories. (Triplett, Allen, & Lewis, 2014, p.

354)

For example, a fight between White students might result in an adult-facilitated mediation while a fight between Black

students might mean being arrested by a resource officer (Toldson, 2012).

Gangs in Schools

Criminalizing discipline policies in school are strongly associated with gang affiliation, following patterns along demographic

lines of race, gender, home life, and socioeconomic status (Butler et al., 2012; Dukes, Martinez, & Stein, 1997; Katz, 1996;

Thornberry et al., 1993). In analyzing research on gangs, three themes emerged: (a) types of students targeted or more

likely to join gangs, (b) the reasons motivating students to join gangs, (c) and the effects of schools on gang affiliation.

Throughout each of these themes, the relationships between the criminalizing school discipline policies of the school-to-

prison pipeline and gangaffiliation continue to develop and deepen.

Targeted students. In research on students and gangs, membership was often conceived at three levels: nongang youth,

transient gang members, and stable gang members of local or national gangs (Howell & Griffiths, 2016). Nongang youth are

those who identify themselves as totally separate from a gang identity; they are not associated with a gang, although they

may have social encounters with them at school or around their neighborhood. Transient members are on the fringe;

sometimes they are active participants and at other times they disassociate themselves. Stable gang members consistently

identify themselves as an active participant consistently involved in a gang.

By using these three levels as a framework, Thornberry and colleagues (1993) were among the first to examine patterns as

to which students stood a greater chance of being recruited into a gang. The best predictors as to who was usually targeted
as a potential gang member were ethnic background, gender, home life or living arrangements, and socioeconomic status.

Any reader must note that these are correlational findings and not intended to communicate any level of causation; however,

research has repeatedly confirmed that Native American, Black, and Latino males from a lower socioeconomic status with

little to no stability in living arrangements are most likely to join a gang (Dukes, Martinez, & Stein, 1997; Gregory, Skiba, &

Noguera, 2010; Katz, 1996; Lea & Abrams, 2015). Yoder and colleagues (2003) also emphasized that gang members are

"more likely to have abusive and dysfunctional families of origin" and may be "homeless and runaway youth" (p. 442).

Thornberry and colleagues (1993) also established three models for how a student becomes a member of a gang by looking

at correlations between membership and behavior:

* The selection model "posits that gangs recruit their members from adolescents who are already delinquent, or at least

have a high propensity for delinquency. These individuals are likely to engage in delinquency and drug use regardless of

their membership in gangs" (p. 57). Therefore, gangs recruit based on existing delinquent behaviors of individuals;

* The social facilitation model proposes that individuals change their behaviors after joining a gang and are influenced

by gang members' behavior: "If the social facilitation model is accurate it implies that gangmembers will differ from nongang

members in terms of delinquency and drug use only when they are active members of a gang" (p. 58);

* The enhancement model combines the two former models and supports the notion that "group processes revolving around

such dimensions as status, solidarity, and cohesion are likely to increase the level of delinquency for gang members." (p. 59)

In short, gang affiliation is associated with delinquency, but when and how that delinquency is expressed remains under

contention. Similarly, youths experiencing delinquency are also more likely to be affected by disciplinary policies associated

with the school-to-prison pipeline.

Reasons motivating gang membership. The main reasons students seek out gang membership are to find a sense of

belonging within a group, to find acceptance from peers, to combat low self-esteem and negative social interactions, and

because they lack an adult role model (Barnert et al., 2015; Graham & Erwin, 2011; Yoder, Whitbeck, & Hoyt, 2003).

Looking at data on targeted youth, most criteria described an outsider or other. As such, many youth join gangs to have a

group that accepts them as normal, as an insider.

However, gang membership is not an overnight occurrence. Several factors begin during adolescence that position youth to

seek out gang affiliation. One is social interaction leading to aggressive behavior at an early age. When youth continually

encounter negative social interactions with others, they value companionship and protection but may seek a group that

values aggressive behaviors:

Children with aggressive behaviour tendencies during childhood may be more likely than prosocial children to seek out a

social context, such as a gang, where their antisocial behaviour is supported . . . For stable gangmembers, early behaviour
problems may have contributed to their rejection by a prosocial peer group and acceptance by a similarly deviant peer

group. (Craig et al., 2002, p. 55)

Affiliating with a group who honors and reflects similar behavior patterns as your own solidifies the group, making the group

and the behaviors both stronger and more identifiable.

Debate remains about adult relationships as predictors of gang affiliation, since some research indicated that other factors

contribute more (Ryan, Miller-Loessi, & Nieri, 2007; Winfree, Bckstrm, & Mays, 1994). However, Wang (1994) claimed

that interactions with adults, and particularly role model adults, were the single best predictor of gang membership:

There was a tendency for gang members to name fewer role models than did their nongang peers; among the African-

American students, nongang youth were three times more likely to mention a parent or teacher as a role model compared

to gang members . . .the absence of either a parent or teacher role model was the best predictor of gang membership. (p.

279)

Such research found a distinct difference: nongang youth were much more inclined to consider a parent or teacher a role

model while gang members were most likely to name a rapper (Taylor et al., 2003). While there is conflicting research on the

impact of role models on gang affiliation, nongang youth seem more likely to name someone they know.

The effects of school. In classrooms today, many teachers come face-to-face with future and current gangmembers. A major

concern for teachers encountering gangs is the safety of all students (Astor, Meyer, & Behre, 1999). The research

consistently emphasized the access and promotion of firearms within gangculture: "Firearms appear to be

prevalent in organized gangs. They are present when the gang meets, and its members frequently carry and shoot their

weapons" (Bjerregaard, 2002, p. 49). Going to school does not mean that a gang identity is suddenly absent. When fights

occur, "gang youth were more likely to endorse the use of guns or fists to settle disputes" (Taylor et al., 2003, p. 505).

However, the rise of juvenile violence is a myth (Mallett, 2015); in fact, the last twenty years have seen steep

declines in violent crime (Puzzanchera & Hockenberry, 2010). Nevertheless, schools have responded to these perceptions

throughout those decades with an increase in policies now associated with the school-to-prison pipeline: more metal

detectors, more closed-circuit television surveillance, more police officers in schools, more zero-tolerance approaches, and

more expulsions (Butler et al., 2012; Mallett, 2015; Toldson, 2012). Inshort, responses to just the perception of juvenile

delinquency may in fact be driving that delinquency (Lea & Abrams, 2015). Newer research seeking out the perspectives of

alienated students has begun to speak to such injustice.

For example, Mallett (2015) pointed out both the myth of juvenile violence and how the responses to that myth have created

a self-fulfilling prophecy of delinquency:

These environments, for any student, can produce negative reactions, fears, or worries about their school; however, in some

schools students may feel resentment and negative feelings toward the surveillance and oversight itself. . . It may be these

security measures themselves that were contributing to the disorder. (p. 4; see also Addington, 2009; Sipe, 2012)

Newer research has also sought the perspectives of those who have been incarcerated on how schools played a role in their

delinquency (Barnert et al., 2015; Lea & Abrams, 2015; Toldson, 2012). When schools do not feel safe and students are not

offered a pathway to academic success, the school-to-prison pipeline seems more and more inevitable.

Some participants in this research describe delinquency as an unavoidable choice and see school as a place where their

first risk-taking behaviors were enacted (Lea & Abrams, 2015). The failure of something better sets students on the school-

to-prison pipeline:

Participants expressed the belief that when home and school fail to meet their inner needs, male and female youths spend

more time on the streets. "Blinded" by gangs, drugs, and other negative influences in their neighborhoods, adolescents . . .

become trapped in a pattern of delinquency and incarceration. Once youths have a history of arrest, the threshold for

rearrest becomes much lower. (Barnert et al., 2015, p. 1369)

Despite these seemingly unavoidable choices, the research also demonstrated that adult relationships held potential for

disrupting a move toward gang membership (Ryan et al., 2007). School violence and safety were linked to student-teacher

relationships: "Parents and school personnel can be partners in helping students feel and be safe at school, thus creating a

positive learning environment" (p. 1067). If gang members tend to lack positive relationships with adults, efforts and

education designed to help teachers engage gang members inmeaningful ways might decrease the power of gangs through

protective factors like academic achievement and prosocial peers. (Barnert et al., 2015).

However, there might be a sense of denial among adults in general, and teachers specifically, about the prevalence

of gangs; it may be easier to ignore a problem as large as gang youth or displace the blame on a different area or different

group of people (Pryor & McGarrell, 1993). Such findings were recalled by incarcerated youth, who felt like teachers could

be a positive influence but give up too soon (Barnert et al., 2015; Toldson, 2012). When these research strands are taken

together, there emerges a strong relationship between the school-to-prison pipeline and student gang affiliations. This

correlation indicates that addressing one might help to mitigate the other and that teachers might be in a position to effect

such changes.

METHODS

During the 2012-2013 academic year, the researcher completed a year-long student teaching internship in an English

Language Arts classroom. During this internship, she worked with the coauthor to design and complete the action research

(AR) project reported. As a methodology for unmasking and unmaking systemic oppressions that surround the act of

teaching itself, AR (Hinchey, 2008; Reason & Bradbury, 2006) was most appropriate methodologically for a project designed

to disrupt the school-to-prison pipeline through curriculum development.

Participants and Setting

This study was conducted in an urban characteristic (Milner, 2012) high school in a mid-sized, southeastern city. The

classroom studied began with ten students, all males aged 15-16 years old, who were placed in an English I "Repeaters"

course because they had failed the end-of-course exam their freshman year. By the spring semester and the implementation

of the curriculum discussed, only seven students remained due to alternative placements based on disciplinary charges or

residential changes. Considering these seven male students, four were Black and three were White; two were receiving

special education services. The four Black students claimed gang affiliations; none of the White students

claimed gang affiliations. Every student had been placed on in-school suspension; two of the seven were removed from the

school in the middle of the spring semester. Two students, brothers, had a father incarcerated during the fall semester and

then released during the spring. All seven received free or reduced lunch.

Procedures

When given the opportunity by her mentor teacher to choose a text for the class, the researcher selected Autobiography of

My Dead Brother (Myers, 2006), a multimodal text she thought might engage the students more readily in reading, both for

its Lexile level (for more information about matching readers with text, see https://www.metametricsinc.com/lexile-

framework-reading/) and relevant subject matter. The first step toward engaging this text was an anticipation guide,

discussed in more detail later; the written responses to this guide served as a foundation for several weeks of discussions

while the book was read out loud in class. After finishing the text, the students produced graphic representations extending

scenes from the book.

The unit was designed to promote discussion and interpersonal relationship as much, if not more than, academic content

and resulted in a deeper relationship between teacher and students. In addition to purposively designed curriculum, strong

student-teacher relationships were heightened by the small class size and the fact that the class lasted both semesters, as

opposed to other classes that lasted only one. The researcher specifically sought to develop strong student-teacher

relationships by following Milner's (2007) contentions for positive classroom environments: teachers and students

* envision life beyond their present situation;

* come to know themselves in relation to others;

* speak possibility and not destruction;

* care and demonstrate care; and

* change their thinking to change their actions.

Data Sources
This study focused on hearing the voices of students, so their ideas and opinions played a significant

role inunderstanding gangs in school culture; the choice of data sources reflects this emphasis. Data sources included

* observational field notes of classroom interactions,

* student journals and a questionnaire,

* an anticipation guide and focus group discussions, and

* member checking. Data from the sources were analyzed using qualitative microanalysis.

Observational field notes were recorded during and after class, documenting comments and conversations that

mentioned gang members, gang symbols, or gang affiliations. Personal reflections from journals, informal writing

assignments, and a questionnaire were included because some students more willing to write about their experience with

and knowledge of gangs rather than discussing it aloud. Data were also drawn from a series of qualitative focus groups

building on methods designed to engage the students in discussing gangs and gang affiliation; these class dialogues

developed from the anticipation guide described. Because of the high-risk stakes involved in such discussions, the focus

groups were anonymous and only possible because of the close relationship developed with the students. The focus groups

were not audio- or video-recorded because such a protocol would likely have been too invasive. In addition to the focus

groups, a series of member-checking discussions were facilitated to present on-going data analysis and garner feedback.

FINDINGS

After an entire semester developing strong student-teacher relationships, data beyond observational field notes began to be

captured and analyzed. Responses to an initial questionnaire were minimal and incomplete because of the personal nature

of the questions, so a curriculum was developed that proved more effective.

Questionnaire

A questionnaire (see Appendix A) asked questions regarding gangs in school and in students' lives. Although this

questionnaire was anonymous, only six were completed and with minimal responses. The students were unwilling to

participate because they thought the activity was going to be used for "snitching," or giving the administration or school

resource officer information that the students did not want others to know. Within the six completed questionnaires, four

students identified a role model: two claimed their fathers, one listed his uncle, and one noted the rap artist Lil Wayne. The

one student who mentioned Lil Wayne also claimed that he was currently in a gang or had been approached to join one,

was aware of gangs present at the school, had friends or family in a gang, and did not believe education prevented gangs,

stating "nothing can stop them [gangs]."

Another student, who stated his role model was his dad "because he tells me right from wrong, and I trust him and love him,"

denied being in a gang or ever being approached to join one and responded nothing more than "I don't know" to the rest of
the questions, adding, "It might" when asked if teachers and administrators knowledge of gangs and violence could

prevent gangs from forming or becoming stronger.

These limited responses might support research claims that lacking an adult role model was one of the main reasons

students seek out gang membership (Barnert et al., 2015; Yoder, Whitbeck, & Hoyt, 2003). Taylor and colleagues' (2003)

contention that the most frequent role model for the gang youth was a rapper was validated because the student most

involved with gangs in and out of school identified Lil Wayne as his role model.

Focus Groups

While direct, personal questions proved somewhat ineffective, the curriculum for the spring semester was designed to

generate conversations around gang issues through the study of Walter Dean Myers's (2006) Autobiography of My Dead

Brother. Results from the anticipation guide (Appendix B) formed the basis for a series of focus group dialogues. While there

were only seven participants by the beginning of the spring semester, this group size allowed for stimulating dialogue and

more trustworthy responses than might be expected from a larger group (Vella, 2008); one student wrote his name on the

anticipation guide but refused to fill in any answers. The findings presented span several days of class but have been

organized according to the statements (numbered 1 through 10) presented in the anticipation guide.

The police. The focus groups considered the presence of the police in the students' lives in response to Statement 1. When

it came to police protection and motives, the students explained how there are "good cops" and "bad cops." When asked,

"What makes a cop good?" they described cops who really listen to you and get your side of the story; they do not assume

you are always up to trouble. "Bad cops" were the ones who try to trick you into trouble or assume that you are guilty before

you have had the chance to explain yourself.

Much of the distrust of police officers pertained to being stereotyped, confirming previous research findings (Lea & Abrams,

2015; Mallett, 2015). One Black student explained that when he walks down the street with two or more of his friends who

are also Black, the cops slow down, start asking questions, and make the assumption that they are in a gang. A White

student responded by stating how he believed that this same event happens regardless of race: "A group of guys is

immediately labeled as a gang." He went on to say that, because the media and society focus on gangs, particularly all-

male gangs, whenever a group of young men are walking down the street as a unit it is assumed that they are a gang.

The students then transitioned into a discussion about how police stereotype them like they do cars. They explained that if a

car is lifted, has big rims, and a nice paint job, cops suspect that the driver is involved indrugs because he has money to

spend customizing his car. The cop then has the right to pull the driver over solely based on suspicion, which is how the

students feel while walking down the street: that their appearance and who they are associated with automatically labels

them as gang members.

Adult role models. Responses building on Statement 2, the navet of adults in relation to illegal behaviors, indicated the

class was split between remaining neutral and strongly agreeing that adults were unaware of students
participating in gang activities. Also, when discussing Statement 3, three students strongly agreed that lacking a role model,

lacking familial support, and the need for protection were prime incentives for joining a gang. Such responses confirmed the

reviewed research highlighting the importance of supportive adult role models, strong family units, and feelings of safety

(Barnert et al., 2015; Ryan, Miller-Loessi, & Nieri, 2007; Taylor et al., 2003; Yoder, Whitbeck, & Hoyt, 2003).

Similarly, three students agreed with Statement 7 that schools tended to ignore gangs or gangrelated violence, supporting

the contention of Pryor and McGarrell (1993). These results reflected an opinion several of the students shared that

regarded adults as ignorant of student behaviors related to drugs, gangs, and violence.

Gang relationships and personal safety. Statements 4, 5, and 6 led to several discussions indicating that the reliability and

consistency of gang protection varied as students explained how gangs were different than having a group of

friends. Gang members are not necessarily friends even if they are considered "brothers." The students explained how the

relationship was not one of companionship or commonality, but one of structure and order. One student explained that

being in a gang is like having a job: you are employed by someone of higher rank, your boss, and you are given a task with

the expectation that it will be completed. Another student compared gang relationships to the military, explaining that

although everyone has a common goal or job there are still varying hierarchies of power making it impossible to be "friends."

You must take orders simply because someone of higher rank has told you to do so.

When discussing how friends and peers affected their actions (Statement 8), the students shared mixed feelings. They

explained that knowing someone for a long time built trust and loyalty that almost no situation could sever. One student

noted, "If I have known someone my entire life, I have their back no matter what." Another student said, "But sometimes

people change, even friends. So if they killed someone or something like that, then no. I'm no longer down with that." When

discussing how relationships within gangs impacted loyalty, they explained that there is no choice within a gang. Similar to

the notion that gang members are not the same as having friends, the students expressed that once you are in a gang there

is no opinion or independence; you act as one unit regardless of how you feel about a situation.

In terms of personal safety, individual responses to Statement 9 were compared with other prompts to see if there were

relationships between how safe a student felt in relation to a specific context. The two students who answered "agree" or

"strongly agree" when asked if they worried about their safety also believed that schools ignored gangs and gang-related

violence and felt neutral about police protection. The three students who "disagreed" or "strongly disagreed" that their safety

was a concern had mixed responses to both schools ignoring gang-related activities and peers affecting behaviors. The

three students who did not worry about their safety felt "neutral" or "disagreed" with the statement that gang members offer

protection no matter the circumstances, whereas other students who worried about their safety "strongly agreed"

that gangs offer solid protection.

These results were intriguing because the students who seem to distrust gangs also possessed a stronger sense of safety

and autonomy. The students who fully trusted the protection of gangs, however, felt more insecure in terms of safety. This

led to discussion about gangs and personal safety in which the researcher looked for answers to several questions: Do
these students believe the idea that gangs are indestructible? Are they targeted as potential gang members who fully

believe in the solidarity of gang culture? How might that belief make them question their safety? In contrast, the students

who seemed more secure doubted the solidarity of gang protection. Does this doubt encourage them to seek out safety on

their own, without the gang's help? Are they distrusting of the hierarchies of power within a gang, making them more aware

of how to protect themselves?

Teachers. It was reaffirming to see the majority of students agreed or strongly agreed with Statement 10 that teachers were

a major source of encouragement, affecting students' beliefs that they were capable of completing high school and

becoming successful. However, a measure of responsibility comes from this response. If this response were demonstrated

to be consistent across contexts, then a teacher may be required to envision and promote academic and personal

accomplishments more heavily than content (Barnert et al., 2015; Ryan, Whitbeck, & Hoyt, 2007).

Member-Checking

While reading Myers's novel aloud as a whole class, the researcher prompted the students into discussion around how the

characters in Autobiography of My Dead Brother faced similar realities as the students: a life of danger with or without the

protection of a gang. The ensuing conversations provided a space for member-checking the findings described earlier.

The students explained how one might "represent" or "claim" a gang without being a gang member: throwing gang signs,

struttin', dressing in a similar fashion, and associating with gang members represents a part of the culture while not

necessarily binding a student to that gang. One student, thinking aloud, explained that what makes gangs so powerful was

that there was no real way to escape (Lea & Abrams, 2015). Friends and family members in gangs surrounded these

students on a daily basis whether or not they were in a gang. Another student, expressed,

I chose not to live my life that way, but it seems like it chooses people anyway. Even if I'm not in a gang, but my friend is,

and we're just talking in the hall, people are going to see me and think I'm in a gang too. But what am I supposed to do? Just

not have any friends?

The dangers and lifestyles of gang membership affected more than those officially in a gang. The lines become blurred

when relationships exist outside of the group as gang identities become based on associations rather than a membership list

(Mallett, 2015).

Rise, a character in Myers's novel, starts to interact with gangs when he begins selling drugs. When asked, "What causes

Rise to start selling drugs and begin associating with gangs?" all the students responded with, "Money and power." When

asked if money and power were the most common reasons for joining a gang, they explained that although money, power,

and protection are usually all part of the reason why kids join gangs, those are not the main causes. Money, power, and

protection are benefits to being in a gang but not the reasons for joining.
The students believed the number one reason youths join gangs is because they lack a role model. While speaking about

role models, several students referred to the lack of a father figure. One student explained,

There is only such much my momma can do. Yeah, she can raise a baby and a little boy, but not having my dad around

when I'm becoming a man . . . it don't work like that. I don't have that male figure to look up to, so I go and find one.

Another student agreed: "Yeah, the reality is most of our dads are either dead or in jail."

DISCUSSION

In reflecting on the findings of this study, and the amount of time spent with these students, a primary implication comes to

the fore: Teachers can make a difference if they take time to get know their students (Barnert et al., 2015; Ryan, Whitbeck,

& Hoyt, 2007). When the subject of gangs was first discussed, Kayla pulled up a chair and told her students, "Okay, y'all are

my teachers. Y'all lead the conversations." The willingness to engage students as people with valuable knowledge (Milner,

2007) was an important first step toward upending a system that had already labeled these participants as failures in a

Repeaters' Class. Throughout the year, Kayla developed a space where students experienced academic success and felt

like experts, a crucial goal for teachers wanting to disrupt the school-to-prison pipeline (Mallett, 2015).

In responding to Statement 10 of the Anticipation Guide, five students agreed or strongly agreed that

"Teachers/parents/other adults have encouraged me and led me to believe that I am capable of passing high school and

being successful." However, these teachers/parents/other adults must be in real relationships with the students. This

involves standing up to systems that label students through deficit perspectives, systems that expect certain outcomes

based on imposed racial, class, behavioral, or academic labels (Gregory, Skiba, & Noguera, 2010). The school-toprison

pipeline begins when schools see students as more likely to be incarcerated than achieve academically. Over the course of

Kayla's time with her students, they often discussed how school reminded them of being incarcerated: Failure in one class

resulted in a label that restricted their placement in other classes, restricting where they could move within the school. Even

cinderblock walls and the imposition of a schedule over which they had no control reminded the students of prison. The

reality of in-school suspension reified the idea that isolation was an effective disciplinary policy (Toldson, 2012).

This was the most important feature of Kayla's curriculum and day-to-day interactions with her students. The reasons that

students join gangs are myriad and cross all demographic categories. However, a teacher's rejection of the school-to-prison

pipeline begins when he or she offers students a place to succeed, develops ways for students to feel in control, and

facilitates student voices to reject the labels put on them by others; these steps allow the multitude of factors and influences

driving gang affiliation to be recognized and addressed, factors and influences that will be different with each student.

It is not a big step to include knowledge of and conversations about gangs and the school-toprison pipeline as important

pieces of a teacher's conceptual repertoires of diversity (Milner, 2010). While such work in the classroom might not seem to

have a direct impact on the illegal activities or delinquent behavior often associated with gangs (Bjerregaard, 2002), it does

provide an alternative, a response to what too many students see as an unavoidable choice (Lea & Abrams, 2015). It
sounds clich, but maybe teaching a student the value of picking up a pencil will help prevent him from picking up a gun

(Taylor et al., 2003).

While there is a need to attack the school-to-prison pipeline as a hegemonic system, there is much that can be done in the

classroom by an individual teacher. The school-to-prison pipeline works when students do not see school preparing them for

anything but prison. On paper, Kayla's students were failures; they were not going to graduate from high school. While they

may still not graduate, Kayla was able to carve out at least one classroom where they felt successful, included, and

knowledgeable. Kayla may not have had any impact on whether or not students choose to participate in gangs, but she did

have the opportunity to provide a space where students chose to participate in readings, activities, and discussions

reflecting their own lives.

Study Limitations and Future Research

A primary limitation of this study is the small number of participants; however, this small number was necessary for

developing the close teacher-student relationships allowing for student voices to be solicited and heard. As such, our

findings or implications do not represent all students but place these participants' voices alongside the growing body of

research engaging other small groups around these issues, research such as that conducted by Barnert and colleagues.

(2015, N = 20) and Lea and Abrams (2015, N = 17). Also, this study included only male participants and is not intended to

represent the growing population of female students in gangs or on the school-toprison pipeline. As such, these voices are

again placed alongside the voices of participants in studies looking specifically for female voices (Barnert et al., 2015) and

for populations set on the school-to-prison pipeline because of sexual orientation (Snapp et al., 2015).

Future research should continue to seek out and hear this diverse symphony of voices. The continued compilation of small-

scale qualitative studies will seek answers to questions like, Why? and So what do we do now? that are not addressed by

larger quantitative inquiry only addressing Who? In addition, research can continue to shift school pipelines away from

prison and in more promising directions, such as success incomputing sciences (Charleston, Charleston, & Jackson,

2014), in college (Ward, Strambler, & Linke, 2013), and in PhD programs (Davis-Maye, Davis, & Jones, 2013).

It may be discouraging for teachers to learn that schools are not always meritocratic institutions where every student can

rise above imposed labels; it may offer no relief to learn that these labels should not be used as excuses for student failure

(Milner, 2013). But coming to such realization is an important step toward disrupting the school-to-prison pipeline. Students

may choose gangs because they offer things that schools may not: safety, support systems, an accessible success ladder,

acceptance within a group, and relevant curricula. Disrupting the school-to-prison pipeline begins by creating classrooms

that do not mimic prisons. As demonstrated by Kayla, this is something a teacher can do by offering a place where students

can succeed, feel in control, and are not bound by imposed labels. Will a relevant curriculum presented in a classroom that

engages students as real, knowledgeable people prevent every student from joining a gang or going to prison? Of course,

this is an impossible question. However, these are important first steps.

References

REFERENCES

Addington, L. A. (2009). Cops and cameras: Public school security as a policy response to Columbine. American Behavioral

Scientist, 52, 1426-1446.

Armour, J. D. (2000). Race ipsa loquitur: Of reasonable racists, intelligent Bayesians, and involuntary

Negrophobes. In Delgado, R., & Stefancic, J. (Eds.), Critical race theory: The cutting edge (pp. 180-193). Philadelphia, PA:

Temple University Press.

Astor, R. A., Meyer, H. A., & Behre, W. J. (1999). Unowned places and times: Maps and interviews about violence in high

schools. American Educational Research Journal, 36, 3-42.

Barnert, E. S., Perry, R., Azzi, V. F., Shetgiri, R., Ryan, G., Dudovitz, R., Zima, B., & Chung, P. (2015). Incarcerated youths'

perspectives on protective factors and risk factors for juvenile offending: A qualitative analysis. American Journal of Public

Health, 105, 1365-1371.

Bjerregaard, B. (2002). Self-definitions of gang membership and involvement in delinquent activities. Youth & Society, 34,

31-54.

Butler, B. R., Lewis, C. W., Moore III, J. L., & Scott, M. E. (2012). Assessing the odds: Disproportional practices and

implications for educational stakeholders. The Journal of Negro Education, 81, 11-24.

Charleston, L. J., Charleston, S. A., & Jackson, J. F. L. (2014). Using culturally responsive practices to broaden

participation in the educational pipeline: Addressing the unfinished business of Brown in the field of computing sciences. The

Journal of Negro Education, 83, 400-419.

Craig, W. M., Vitaro, F., Gagnon, C., & Tremblay, R. E. (2002). The road to gang membership: Characteristics of

male gang and nongang members from ages 10 to 14. Social Development, 11, 53-68.

Davis, A. (2003). Are prisons obsolete? New York: Seven Stories Press.

Davis-Maye, D., Davis, D. J., & Jones, T. B. (2013). Who's got next: SOTS's KEMET Academy as a model to improve the

community college to PhD pipeline. The Journal of Negro Education, 82, 243-254.

Dukes, R. L., Martinez, R. O., & Stein, J. A. (1997). Precursors and consequences of membership in youth gangs. Youth &

Society, 29, 139-165.

Graham, A., & Erwin, K. D. (2011). "I don't think Black men teach because how they get treated as students": High-achieving

African American boys' perceptions of teaching as a career option. The Journal of Negro Education, 80, 398-416.
Gregory, A., Skiba, R. J., & Noguera, P. A. (2010). The achievement gap and the discipline gap: Two sides of the same

coin? Educational Researcher, 39, 59-68.

Gun-Free Schools Act of 1994, 20 U.S. Code 7151.

Hinchey, P. H. (2008). Action research. New York: Peter Lang.

Howell, J. C., & Griffiths, E. (2016). Gangs in America's communities (2nd ed.). Thousand Oaks, CA: Sage.

Katz, S. R. (1996). Where the streets cross the classroom: A study of Latino students' perspectives on cultural identity in city

schools and neighborhood gangs. The Bilingual Research Journal, 20, 603-631.

Lea, C., & Abrams, L. (2015). "Everybody takes a road": Perspectives on the pathway to delinquency among formerly

incarcerated young men of color. Manuscript submitted for publication.

Mallett, C. A. (2015). The school-to-prison pipeline: A critical review of the punitive paradigm shift. Child

and Adolescent Social Work Journal. doi:10.1007/s10560-015-0397-1.

Mauer, M., & Chesney-Lind, M. (Eds.). (2003). Invisible punishment: The collateral consequences of mass imprisonment.

New York: New Press.

Meiners, E. R., & Winn, M. T. (2010). Resisting the school to prison pipeline: The practice to build abolition democracies.

Race Ethnicity and Education, 13, 271-276.

Milner, H. R. (2007). African American males in urban schools: No excuses-teach and empower. Theory into Practice, 46,

239-246.

Milner, H. R. (2010). What does teacher education have to do with teaching? Implications for diversity studies. Journal of

Teacher Education, 60, 118-131.

Milner, H. R. (2012). But what is urban education? Urban Education, 47, 556-561.

Milner, H. R. (2013). Analyzing poverty, learning, and teaching through a critical race theory lens. Review of

Research in Education, 37, 1-53.

Myers, W. D. (2006). Autobiography of my dead brother. New York: Harper Tempest.

Noguera, P. (2003). Schools, prisons, and social implications of punishment: Rethinking disciplinary practices. Theory into

Practice, 42, 341-350.

Pryor, D. W., & McGarrell, E. F. (1993). Public perceptions of youth gang crime: An exploratory analysis. Youth & Society,

24, 399-418.
Puzzanchera, C., & Hockenberry, S. (2010). Juvenile court statistics, 2010. Pittsburgh, PA: National Center for Juvenile

Justice.

Reason, P., & Bradbury, H. (Eds.). (2006). Handbook of action research. Thousand Oaks, CA: Sage.

Ryan, L. G., Miller-Loessi, K., & Nieri, T. (2007). Relationships with adults as predictors of substance use, ganginvolvement,

and threats to safety among disadvantaged urban highschool adolescents. Journal of Community Psychology, 35, 1053-

1071.

Sipe, P. (2012). Newjack: Teaching in a failing middle school. In Bahena, S., Cooc, N., CurrieRubin, R., Kuttner, P., & Ng,

M. (Eds.), Disrupting the school-to-prison pipeline (pp. 32-41). Cambridge, MA: Harvard Educational Review.

Skiba, R. J., Michael, R. S., Nardo, A. C., & Peterson, R. (2002). The color of discipline: Sources of racial and gender

disproportionality in school punishment. Urban Review, 34, 317-342.

Skiba, R. J., Rausch, K., & Ritter, S. (2004). Children left behind: Series summary and recommendations. Education Policy

Briefs from Indiana Youth Services, 2(4).

Snapp, S. D., Hoenig, J. M., Fields, A., & Russell, S. T. (2015). Messy, butch, and queer: LGBTQ youth and the school-to-

prison pipeline. Journal of Adolescent Research, 30, 57-82.

Taylor, C. S., Lerner, R. M., von Eye, A, Bobek, D. L., Balsano, A. B., Dowling, E., & Anderson, P. M. (2003). Positive

individual and social behavior among gang and nongang African American male adolescents. Journal

of Adolescent Research, 18, 496-522.

Thornberry, T. P., Krohn, M. D., Lizotte, A. J., & Chard-Wierschem, D. (1993). The role of juvenile gangs infacilitating

delinquent behavior. Journal of Research in Crime and Delinquency, 30, 55-87.

Toldson, I. A. (2012). Insecurity at Black schools: When metal detectors do more harm than good. The Journal of Negro

Education, 81, 303-306.

Toldson, I. A., & Morton, J. (2011). A million reasons there're more Black men in college than in prison: Eight hundred

thousand reasons there's more work to be done. The Journal of Negro Education, 80, 1-4.

Triplett, N. P., Allen, A., Lewis, C. W. (2014). Zero tolerance, school shootings, and the postBrown quest for

equity in discipline policy: An examination of how urban minorities are punished for White suburban violence. The Journal of

Negro Education, 83, 352-370.

Valles, B. G., & Villalpando, O. (2013). A critical race policy analysis of the school-to-prison pipeline for Chicanos. In Lynn,

M., & Dixson, A. D. (Eds.), Handbook of critical race theory in education (pp. 260-269). New York: Routledge.
Vella, J. (2008). On teaching and learning: Putting the principles and practices of dialogue education into action. San

Francisco, CA: Jossey-Bass.

Wang, A. Y. (1994). Pride and prejudice in high school gang members. Adolescence, 29, 279291.

Ward, N. L., Strambler, M. J., & Linke, L. H. (2013). Increasing educational attainment among urban minority youth: A model

of university, school, and community partnerships. The Journal of Negro Education, 82, 312-325.

Winfree, L. T., Bckstrm, T. V., & Mays, G. L. (1994). Social learning theory, self-reported delinquency, and youth gangs: A

new twist on a general theory of crime and delinquency. Youth & Society, 26, 147-177.

Yoder, K. A., Whitbeck, L. B., & Hoyt, D. R. (2003). Gang involvement and membership among homeless and runaway

youth. Youth & Society, 34, 441-467.

AuthorAffiliation

Kayla M. Gass Oak Ridge High School

Judson C. Laughter University of Tennessee, Knoxville

AUTHORS

KAYLA M. GASS teaches ninth-grade English at Oak Ridge High School in Oak Ridge, Tennessee. JUDSON C.

LAUGHTER is an Assistant Professor of English Education at the University of Tennessee, Knoxville.

All comments and queries regarding this article should be addressed to jud.laughter@utk.edu

Appendix

(ProQuest: Appendix omitted.)

Word count: 8101

Copyright Howard University Summer 2015

Gass, K. M., & Laughter, J. C. (2015). "Can I make any difference?" gang affiliation, the school-to-prison pipeline, and implications
for teachers. The Journal of Negro Education, 84(3), 333-347. Retrieved from
http://search.proquest.com/docview/1770939474?accountid=37714

The violent gang and the construction of masculinity amongst

socially excluded young men
Baird, Adam. Safer Communities 11.4 (2012): 179-190.

1. Full text
2. Full text - PDF
3. Abstract/Details
4. References 69
Hide highlighting
 Abstract
TranslateAbstract

Purpose - The aim of this paper is to help in understanding the relationship between the construction of the male identity and

how social violence may be "reproduced" (using the concept of habitus after Pierre Bourdieu), in poor and socially excluded

contexts. The paper aims to inform debate and policy making. Design/methodology/approach - The paper draws on

empirical data collected in 2008, in the form of life-history interviews with male youths - including members and non-

members of gangs - from two poor and very violent neighbourhoods in Medelln, Colombia's second largest city. Findings -

Masculinities alone do not account for urban violence, but they play an integral role why violence is reproduced. In socio-

economically excluded contexts the gang becomes an attractive vehicle for "doing masculinity" for boys and young men.

Youths who did not join gangs tended to have family support to develop a "moral rejection" of gangs, crime and violence

during childhood, which contributed to them finding non-gang pathways to manhood. Youths who joined gangs were less

likely to develop this "moral rejection" during childhood, often due to family problems; and were more likely to admire

older gang members, and perceive the gang as an attractive pathway to manhood. Research limitations/implications - As the

sole researcher a limited number of 32 individuals were interviewed. Originality/value - There is a lack of research on

masculinities and gangaffiliation in the UK and across the globe. This paper provides new conceptual ideas for

understanding whyyoung men make up the vast majority of violent gang members, whilst providing an original data set from

a very violent urban setting.

Full Text
TranslateFull text
Turn on search term navigation

Introduction

This paper draws on original empirical data collected in 2008, in the form of life-history interviews with male youths -

including members and non-members of gangs - from two poor neighbourhoods in Medelln, Colombia's second largest city.

It seeks to bring us closer to understanding the relationship between masculinities - that is the construction of the male

identity ([20] Hearn, 1996, pp. 203-4) - and how social violence may be "reproduced" in poor and socially excluded contexts.

There is a generalised lack of research into masculinities and gang affiliation, not just in the UK, but across the globe. Whilst

noting this absence, this paper also highlights some of the emergent work around masculinities, socio-economic exclusion

and gangs. It then argues conceptually that whilst masculinities alone do not account for urban violence, they play an

integral role in understanding why such violence is reproduced, using the concept of habitus after French sociologist [6]

Bourdieu (1977). These claims are supported by presenting a "snapshot" of quotes from youths' life-history interviews,

analysing what this tells us about the interaction between masculinity, socio-economic exclusion and the reproduction of

violence.

Medelln in perspective

Much of our knowledge about gangs emanates from the USA where systematic empirical research has been conducted

since the 1920s ([54] Thrasher, 1927; [12] Cloward and Ohlin, 1960; [57] Yablonsky, 1997). Until the 1990s
rigorous gang research in Europe was largely absent, bar perhaps James Patrick's A Glasgow GangObserved ([34] Patrick,

1973). Consequently, much European scholarship has used US research as a window on gang life ([26] Klein et al. ,

2001); in more recent years gang research has expanded across the world, including multi-country comparisons (Covey,

2003; [26] Klein et al. , 2001; [19] Hagedorn, 2008; [15] Dowdney, 2007; [51] Small Arms Survey, 2010). Understandably

there is much contemporary debate about how gangsarise ([46] Rodgers, 2010), particularly regarding policy-oriented

responses to gang-related crime and violence ([50], [51] Small Arms Survey, 2008, 2010). Whilst comparative approaches

to gang research can illuminate, pointing to commonly shared traits - such as being urban, male dominated, and violent or

criminal - the range of contexts in which gangs arise also make such comparisons problematic. The very definition of the

"gang" is itself contested ([40] Pitts, 2008).

This poses an initial challenge for this paper. Medelln has an extremely violent recent history, involving a confusing array of

illegal armed groups, often linked to the broader armed conflict in Colombia. Armed groups have been active in Medelln's

poor neighbourhoods since the 1950s ([28] Medina Franco, 2006; [41] Ramrez, 1979), and since the late 1980s the city has

been notoriously violent; in 1991 achieving the highest per capita homicide rate in history, at 381 per 100,000 inhabitants

([52] Surez Rodrguez et al. , 2005, p. 203)[1] . This was brought about by a cocktail of gangs, paid assassins - or sicarios -

drug cartel violence, urban militias linked to left-wing guerrilla groups, and right-wing paramilitary death squads interlinked

with state violence. In2008 when the data were collected for this paper, right-wing paramilitary groups linked simultaneously

to the state's security forces had taken over, through force or coercion, most of the gangs in the city, and at street level

"gang member" and "paramilitary" became largely synonymous in the eyes of the community.

Given these chronic levels of violence ([36] Pearce, 2006b), Medelln presents a particularly intense face of urban violence

when compared to the UK, or across Europe and the USA. Despite the caveats posed by contextual differences,

comparative elements remain: young, poor men in Medelln persist at the heart of gang life, a demographic which is

consistent amongst gangs across the globe. As such, learning from the role that masculinities play in violence

reproduction in Medelln is relevant to other contexts.

Theorising gangs and masculinities

In the 1950s [33] Parsons (1963, pp. 302-5) stated that in Western societies the "occupational system" "is the most

important competitive process in which the individual must achieve his status", where the family's status depends on the

occupation of one member, the husband and father. Gang researchers in the USA noted this, stating the gang could act as a

"vehicle" for achieving manhood, particularly for youths "cut off from the possibility of manhood for a prolonged period",

arguing that the "ganging process" provides symbolic evidence of the "urge for manhood" (Bloch and Niederhoffer in [57]

Yablonsky, 1997, pp. 171, 172). Similarly, [12] Cloward and Ohlin (1960, p. 48) wrote that "aberrant behaviour may result"

where the individual encounters obstacles to "growing up": "This problem of adjustment arises from efforts to conform to

cultural expectations under conditions in which conformity is hampered or excluded".

These contextual blockages to manhood were reflected years later by [10] Bourgois' (2003, pp. 293, 295) study of Puerto

Rican drug gangs in 1990s New York, where men with business acumen and diverse progeny commanded "respect".

However, these masculinising process were undermined for excluded young men: "The former modalities of male respect

are no longer achievable within the conjugal household... one can discern the gender-specific form of the experience of

social marginalization in the Puerto Rican diaspora".

Whilst more recent comparative studies of gangs across the globe have raised the question of masculinities ([14] Covey,

2003; [19] Hagedorn, 2008; [15] Dowdney, 2007), it has rarely been investigated or theorised insignificant depth. Of these

multi-country studies, in his chapter in "The Eurogang Paradox", [25] Kersten (2001) goes furthest by focusing on the idea of

hegemonic masculinity to explain how it is related to gangmembership in Germany, concluding that class and masculinity

interact, conceptualising hooliganism as a form of masculine, working class rebellion.

In the UK, only a few have focused on masculinities in relation to gangs. [31] Muncie (2009) noted that men are responsible

for 92 per cent of violent crime, where young men in economically deprived areas abandoned by the state, unleashed

"maverick" and extreme forms of masculinity. [40] Pitts (2008, pp. 94-5) refers to masculinity and menace, where the

"glittering prizes" of the drug gang life-style and respect are "indissoluble" from masculinity and gender identity for

disadvantaged young men.

Whilst these authors recognise the role of gender construction in gang violence, researchers rarely target masculinity as a

rubric for in-depth research, and empirically based work to that effect is particularly rare. This is perplexing because at a

global level young men are consistently the protagonists of violence, suggesting there is a relationship between being "male"

and violence. In 2002 the WHO reported that "Males accounted for three-quarters of all victims of homicide, and had rates

more than three times those among females. The highest homicide rates in the world - at 19.4 per 100,000 - were found

among males aged 15-29 years" ([27] Krug et al. , 2002, p. 6). Medelln is no exception to this trend. From 5,450 homicide

victims in 1990, 95 per cent of these were men and 65 per cent were between the ages of 15 and 29 ([42] Revista

Planeacin Metropolitana, 1991, p. 3), data which has remained unerringly constant over the last two decades[2] . Arguably

a focus on masculinities in relation to urban violence deserves more of our scholarly attention.

Further, whilst recognising there is less research into female gang membership ([5] Batchelor, 2009), very little is understood

about how girls, women and femininities relate to "masculinisation" - the culturally located process of "becoming a man" -

and how this in turn can be related to gang membership and violence.

Much can be learned from emergent literature on masculinities and violence in Latin America since the 1990s. Not least, to

avoid the reductive stereotypes of machismo and to recognise the complex meanings of masculinity to nuance our research

([16], [17] Gutmann, 1996, 2001; [55], [56] Viveros Vigoya, 2001, 2002). We should also note, as [44] Rodgers (2006, p.

286) does, that machismo alone cannot account for violence: "ultimately wider socio-cultural norms and structures such as a

long history of violence or machismo can only be seen as contributing to rather than determining the institutional

development of Nicaraguan pandillerismo [gang activity]". Nevertheless, amidst a range of masculinities,

hegemonic macho types are those most reproduced in Latin America, grounded in gender inequalities within societies in the

region, and have been consistently linked to violence ([18] Gutmann and Viveros Vigoya, 2005, p. 118).

Whilst the emasculating effects of deindustrialisation and shifting masculinities have been documented by a number of

scholars in the West, in some cases referred to as the "crisis of masculinity" ([12] Cloward and Ohlin, 1960; [71] Cleaver,

2002; [19] Hagedorn, 2008), only [4] Barker (2005) has focused concertedly on the emasculating effects of poverty,

exclusion, inequality, found in Latin America's periphery, and how this is related to gang membership. In the Global South

outside of Latin American, a notable body of research into masculinities, exclusion and gangs in Cape Town, was conducted

by [24] Jensen (2008).

Although literature on urban violence, particularly in Latin America, increasingly links socio-economic exclusion to

violence[3] , conceptually this needs to be linked more rigorously to masculinities to develop our understanding of urban

violence. It is the way that deprived socio-economic conditions interact with "masculinisation" that can cast light upon the

(re)generation of violence. What is it about "becoming a man" incontexts of exclusion that lead some youths, but not others,

into violence? What processes lead new generations of young men to replenish the ranks of gangs? This approach should

be considered complimentary to literature that, quite correctly, has begun to understand the connectedness between

violence and social exclusion.

Masculinities and violence "reproduction"

Conceptualising violence is a slippery task. In recent years scholars have come to understand violence as a multi-causal,

socially generated "continuum" or "reproduced" phenomenon[4] . If urban violence is a reproduced phenomenon, focusing

on masculinities can cast light upon the way in which violence is reproduced in urban contexts of socio-economic exclusion.

To this end, this paper argues that male youths are disposed - that is they have a tendency - to reproduce the prevailing

versions of masculinity they are exposed to growing up; and in doing so in contexts of exclusion and violence, some youths

reproduce that violence.

To analyse this the concept of habitus was used (b6 b7 b8 b9 Bourdieu, 1977, 1992, 2001, 2005). Although this and related

concepts will not be covered in detail in this paper[5] , habitus disposes youths to make their own transition to a culturally

valued, and thus common and reproduced, gendered self; expressing agency and dynamism in doing so, though their

subjectively held experience of masculinisation. In short boys are disposed to become men in a way that reflects the older

men around them. Whilst, of course, this reproduction of practice is imperfect, allowing for multiple identities, agency and

social change, masculine habitus ([13] Coles, 2009) helps explain the generalised intergenerational transmission, or

continuum, of male behaviour. This helps us explain the reproduction of "masculine" social practice.

As early as 1960, [12] Cloward and Ohlin (1960, p. 48) have referred to the "ganging" process, that the gangsatisfies the

"deep-seated" urge for manhood, or "compulsive masculinity", in societies that do not provide adequate opportunities of

preparation for adolescents' transition into adulthood ([57] Yablonsky, 1997, p. 172). Habitus helps us grasp why this "deep-
seated" urge for manhood is reproduced, whilst structural constraints and opportunities shape how masculinity plays

out in practice the social world. Therefore, the reproduction of masculinity can be linked to the reproduction of urban

violence in contexts of socio-economic exclusion. Conceptually this helps us explain why the vast majority of gangs are

comprised of male youths and why this is reproduced across generations. Let us explore these concepts using the empirical

data.

Life-history interviews: reproducing masculinities and gang membership

I worked as a participant observer at two community organisations in 2008 to facilitate interviews with local male

youths in the poor the north-eastern and north-western precincts of Medelln. Two sets of youths were interviewed to map

out their life histories - 17 youths (average age 21.5) who had violent histories via their membership to armed groups -

principally gangs linked to paramilitary structures - and 15 youths who not only avoided joining gangs, but worked with the

local community organisation CoVida [6] , developing values that rejected violence and crime[7] (average age 23.4). The

broader aim of the research was to provide an understanding of why some youths from similarly deprived socio-economic

backgrounds became violent, whilst others took a markedly different path in life ([3] Baird, 2011). Such data are relatively

scarce, as few researchers have conducted these types of ethnographies in dangerous contexts, particularly when studying

armed actors. A number of challenging issues arose in the data collection which are not covered here but have been

outlined in a previous article on methodology ([2] Baird, 2009).

Whilst there may be multiple reasons for a youth joining a gang, contexts of poverty and exclusion, which also have a high

concentration of gangs, present particular challenges for boys and male youths. To understand these, first of all we need to

grasp what it means to be a man in peripheral Medelln today according to youths living there. As mentioned above

hegemonic or macho types of masculinity are widespread in Latin America, and are synonymous with social status, respect,

money, sexual access to women and often violence; but there is contextual definition, or localised nuances, to masculine

identity in Medelln. Hernando and Sammy, who worked at CoVida , were asked to outline what was locally understood as

"being a man".

Hernando :

Hernando : ...Paisa culture [in Medelln] is really machista and patriarchal... Whichever man leaves the rigid cultural norms

of what a man and a woman are is a marica [softie/poof/homosexual].

Sammy :

Sammy : Here you notice particularly the strong difference between men and women. Being a man is to be strong, a brute,

bringing home money, a protector, skilful, a womaniser, a chauvinist, macho, having power, being respected. Being a

woman is the inverse of being a man... being weak, fragile, not having power, not having status, to be subordinated ... .
As boys become men, habitus disposes them to reproduce these local versions of "what it is to be a man"[8] . The youths

interviewed were disposed to "masculinise" and "live up" to existing versions of masculinity, even if they were not

necessarily conscious of the fact that this behaviour was shaped by their masculine habitus [9] .

However, "doing" masculinity depends on the options available for each individual, which are termed here, in a slightly

grandiose fashion as "masculinisation opportunities". What constitute "masculinisation opportunities" varies dramatically

according to the socio-economic conditions in which a youth is brought up, shaping each youth's pathway to manhood. Such

processes are of course nuanced, and interact in complex ways with agency and happenstance so we should be wary of

drawing overly "neat" conclusions, or categorising de factocausal factors of youth violence. However, in general, socio-

economic exclusion is emasculating. Gangaffiliation has thus been conceived as a response to emasculation by some: [25]

Kersten (2001) refers to the "down and out" turned into rebellious hero, who is feared, earning respect

through gang membership; Jensen refers to the agterbuurte gang member in Cape Town, who becomes a "bad mother

fucker" through heroic identification with the gang to invert a sense of masculine disempowerment ([24] Jensen, 2008, p.

92); [4] Barker (2005) draws similar conclusions. These conceptions of "protest masculinity" ([1] Adler, 1928), contribute to

exaggerated or hyper-masculine sub-cultures in gangs which are violent, risky and macho ([4] Barker, 2005, pp. 71, 82; [57]

Yablonsky, 1997, p. 172; Hume IN [35] Pearce, 2006a, p. 70).

What is argued here is that emasculation reduces licit "masculinisation opportunities". This makes it a struggle for male

youths to "become a man" according to popular expectations. Such contexts have a particular gendered affect on boys and

male youths because they require the contestation of emasculation, to stand on the shoulders of [10] Bourgois' (2003, p.

295) ideas of the gender-specific effects of marginalisation.

For many of these youths this contestation was a burden. The youths at CoVida were asked to reflect on growing up in their

neighbourhood. In particular, Pelicorto's reference to "frustrated dreams" reflects the absence of licit (masculinisation)

opportunities in his neighbourhood, which contrasts starkly with the "excellent option" represented by "role

models" in street gangs.

Pelicorto :

Pelicorto : There aren't any heroes [for youths], just unfortunate imaginaries [or role models]... of what the "good life" is. It's

not about what makes someone feel good as a human being, it's about owning things... young men need to show off... Life

is about... partying... good clothes, to have a motorbike... Youths have reduced their whole lives to that ... .

They just finish school and out to work... they want to get a job and then at some time think about leaving the neighbourhood

because it's so poor... But that just becomes a load of frustrated dreams because the obstacles they face in this city mean

they can't be fulfilled. When you feel really frustrated you begin to look at the options in front of you and of course you begin

to find other options.

Adam : And an option is to join an armed group?

Pelicorto : Exactly... The other problem is that [youths] who live here have role models who have taken up those offers

[from gangs] and today... they live well. The rule of law is non-existent... So at the end of the day they think "well... This

seems to be an excellent option" ... .

Pepe :

Pepe : I think that they [gangs] are models that reproduce themselves and they reproduce with great ease and efficiency

[because] one of the reference points here that is latently constructed is that of the "cacique " or "boss" [the leader of the

local gang]... who enjoys strong economic solvency... but also he's got respect, recognition, power. So of course the young

lads round here say "fuck me, this is the ticket!"... It's also seen as the easy route... they are given a gun, and a gun is

already a big deal [his emphasis]... a very resounding symbol... The armed actor, the head of the gang, or the person who

has been getting involved with armed groups... enjoys significant status and recognition.

The following interviews were with youths with a history of gang involvement, which seek to understand who they admired,

their role models, what they understood as successful or productive ways of "being a man", and how that might be achieved;

and in particular how this might have been related to their decision to join the gang ([29] Moser, 2009).

Mauricio & Andres :

Mauricio : [Drug Baron] Pablo Escobar's henchmen used [to go to Andrs'] house [because they were friends of his father].

So you see, they [became] his role models.

Andrs : ...I was always being set that example in my youth... I always saw guns... so you begin to think you're a big

somebody with all this stuff around.

Adam : And what were the examples of men like?

Andrs : Of course, the motorbike, the gun, pretty girls, the money... So then as time went by I was... 15 or 16, I got

involved in that lifestyle... They set me the example... to want a motorbike, to be like them ... .

Notes:

Notes : Yeah bro, you know what happens? The same thing that happens to all of us youngsters; picking up a gun for the

first time means putting on the big trousers, so [you're] gonna want announce this to the whole world by killing someone.

Adam : Explain more about putting on "big trousers"... what when a kid picks up a gun at 16 years old now he thinks he's a

man?

Notes : That's it, that story is as old as the hills... if not you're a bottler and a pussy ... .

Carritas :
Carritas : ...women are also participants in this [war], they're the ones that provoke everything. If you're in love with a sweet

girl and you're not in a gang and you don't have a motorbike, you don't have nothing. She's not even gonna look at you... On

the other hand, if you pass by in a nice car with the boss the women are looking, you know. So the women also play their

part.

The youths in gangs that I spoke to admired the gang leader - the boss, duro or cacique - above all in their neighbourhood.

They were very clear about why they admired the boss and what they valued as being a "successful" man. These localised

signifiers of "doing masculinity" or masculine capitals[10] , included access to material goods such as fast motorbikes,

expensive clothes and trainers; but were also symbolic such as "respect" and "fear" generated by being a "tough

guy" in the gang, and sexual access to the most coveted women in the neighbourhood. Whilst some youths cited self-

defence or protection from other gangs as a reason to join their local gang, these same youths put more significant

emphasis on their reason for joining the gang to gain respect, nice clothes, access to women, et cetera. Thus,

the gang became a pathway to manhood, serving as both a reputational and economic project, a mechanism to obtain

status, respect, notoriety, sexual access to women, and to access the relative "riches" of gang crime. Such "life-style"

motivations to join gangs have been highlighted by a number of gang researchers ([40] Pitts, 2008, pp. 94-5; [47] Rubio,

2008), but here are conceptualised as masculine capitals.

The accumulation of masculine capitals through membership to the gang gave these youths what they perceived as a

dominant position of status in the local neighbourhood. The youths at CoVida confirmed that most young men in their

neighbourhood looked up to the duros . This meant gang membership stood out as a "masculinisation opportunity" or "the

best" pathway to manhood for many youths growing - as Pepe said, a lot of youths think "fuck me, this is the

ticket!" Gang membership and violence emerged as seemingly powerful tools to obtain these capitals; lamentably, to secure

them gangs depended on the deployment or threat of violence to control their criminal economic interests, such as

racketeering and drugs sales. In fact, many of these youths demonstrated significant agency in joining gangs. Gangs are not

necessarily for the weakest or most desperate youths, they can be seen as the "best" livelihood opportunity; the best way for

ambitious young men to accumulate masculine capital and to achieve status when the other options around them are

unattractive, undignified or few and far between.

Whilst using violence can be a masculine act in itself, none of the youths interviewed said that they wanted

to join the gang because it was an opportunity to "use violence", but rather to access material goods, women, "respect" and

the like. This is a polemical suggestion and not one intended to romanticise the gang as an "emancipatory project",

but gang violence seen from this perspective can be understood as an externality or by-product of youths' masculine

assertion in conditions of emasculation. Arguably, upon entering gangs, youths were not be striving "to be violent" they were

striving "to be men", driven by the less than conscious dispositions of the masculine habitus . Gang membership was a

mechanism "instrumentalised" or "operationalised" to accumulate masculine capital, and this process contributes to the

continuum of violence.
[10] Bourgois (2003, p. 319) has highlighted that Puerto Rican drug gangs in New York are epiphenomena - or symptoms -

of systematic exclusion, such as poverty, inequality and blocked access to opportunities for the less well off. Certainly then,

as Harvey states, the notion of space in the city is not just the territorial environment, but fundamentally it is about social

relations ([46] Rodgers, 2010). Arguably, gangs and their violence are reproduced by young men living in urban contexts of

socio-economic exclusion precisely because gangs are instrumentalised to mitigate, contest, or deflect the effects of

emasculation. As such, we can argue from a "masculinities perspective" that gang violence is prompted by urban exclusion,

and therefore cannot be understood in isolation from the political economy of the modern city.

Explicitly in terms of reproduction, youths' masculine habitus , the dispositions to become men, will force their way to the

surface. In the absence of licit opportunities, and the possibility of gang affiliation there is a certain "masculine logic" to

joining, and hence reproducing the gang. In the narratives of the youths in gangsthey clearly sought to emulate the

prototype gang boss, and the gang subculture became a fertile place for learning to use and reproduce violence. If we

examine urban violence through this lens, it is unsurprising that in conditions of exclusion and gang presence, male youths

are the protagonists of violence reproduction. Infact as [4] Barker (2005) raises, we should be asking why more youths are

not members of gangs. Ergo, we can begin to understand the reproduction of violence as a gendered social practice -

in poor communities where social and economic opportunities are limited boys are more likely to join gangs.

Finally it should be noted that the linkages between masculinisation and violence reproduction become compounded

where gangs exist over a number of generations in a given area. This has been the case inMedelln. They have become

structure or "institution-like"[11] , embedding themselves in communities across generations, compounding ontologically

"what it means to be a successful man", and crucially how to achieve that "success". This improves the efficacy at which

they can draw up successive generations of young men into their ranks, which reinforces the reproduction of violence.

However, how these "institutions" perform in the social world to shape agents behaviour depends upon the particular

historical and local contexts in which they function. Gang "structures" in Chicago, London, Medelln or the Cape Flats may

generate significantly different activities by gang members. However, although gangs may simultaneously express facets of

both welfare and predation, gangs generally have a negative impact on communities[12] .

Why do some youths join gangs whilst others do not when they both live in contexts of exclusion, poverty and

violence?

Analysing the empirical data, both youths with violent histories and those that joined the community development

organisation CoVida sought to "masculinise". Both did so reflecting traditional and contextual elements of masculinity, and

they coveted masculine forms of status, respect, dignity and esteem. These youths' dispositions to "masculinise" often

reflected each other. However, how they pursued prevailing masculine values varied considerably given the influences upon

them as they grew up. They took divergent routes to manhood because the opportunities available to "masculinise" for

youths at CoVida were in fact markedly different to those of "violent" youths (for futher discussion see Baird, 2012).
Despite living in similarly poor neighbourhoods, the family upbringing of these youths was very different with a complex

range of experiences. The youths at CoVida viewed their childhood upbringing and home life in a much more positive light

than youths who ended up joining gangs. Influences at home and the support and affective relationships they experienced

acted as "buffers" which prevented youths from entering gangs intwo significant ways: first of all a commonality that

all CoVida youths shared was their rejection of violent and criminal groups such as gangs by the time they reached

adolescence. This was instilled in the youths when they were children by influential older family members; frequently parents

- both mothers and fathers - but at times grandparents, elder siblings or the extended family such as uncles or aunts.

Second, they were usually supported at home with their schooling and a range of "alternative" interests to hanging out on

the street corner - such as participating in church groups, youth clubs or socialising with youths who liked different music to

the norm. These two factors - the genesis of a "moral" rejection of gangs and the support for alternative interests, had a

significant influence upon the socialisation spaces these youths chose inadolescence, and subsequently their pathways to

manhood. Adolescent socialisation spaces, hence acquaintances and friendships, proved fundamental for youths' rejection

or integration into gangs.

Of the youths who joined gangs, half had older family members, normally brothers, in gangs. These family members were

male role models, shaping what these youths understood as "doing masculinity". They also provided exposure and

socialisation into armed groups, and hence "opportunities" to develop their masculine identity. The majority of these youths

also had significant problems at home and little domestic support for extra-curricular activities or developing alternative

interests to the street corner. When these youths came into contact with the ubiquitous street gangs on their doorsteps, they

did not reject them as a possible masculinisation opportunity as the youths who joined CoVida had done. They had not

developed the same moral rejection of gangs and violence, nor alternative interests and socialisation spaces to the "street

corner".

Whilst tendencies emerged from the data, plotting a series of clear-cut "causal factors" would be problematic. The social

world and individuals are complex and they will not fit neatly into a set of explanatory factors. As they grew up each youth's

decision-making and consequent social action was shaped by a complex of context, agency, opportunity and happenstance.

These youths' upbringings were neither entirely "positive" nor "negative"; violent fathers could also be caring fathers;

"violent" youths' older brothers in gangs sometimes tried to stop them following in their footsteps; and youths

at CoVida could also display macho traits.

Where youths rejected violence and gangs, they had the possibility to encounter and enter CoVida at a later stage in their

lives, although often entering this organisation came through "chance" exposure to it. Some of these youths may also have

chosen other non-violent pathways if they had not come into contact with the community organisation. However, once

youths had entered CoVida they found "positive" male role models inthe older youths and men there. The organisation also

became a key socialisation space through work and friendships, and thus was central to the development of their masculine

identity and behaviour.

It should be highlighted that most youths in the communities where the research took place, neither joined the gang or

community organisation, demonstrating that there are many different pathways to manhood incontexts of exclusion and the

majority do not involve joining a gang[13] . Unfortunately, the implications of this are that only a minority of youths need to

opt to join the gang for high levels of urban violence to continue. This is a serious challenge for those seeking to interrupt

violence reproduction.

Conclusions

These arguments do not claim to be hard and fast rules, rather they strive to challenge and develop our understanding of

masculinities in relation to gang affiliation and the reproduction of urban violence. The social world is rich and complex. We

should not overestimate how much "reality" of the social world we can capture and we should be shrewd in any prescriptions

we may offer to counter gang violence. However, it would be remiss not to recognise that male youths' pursuit of manhood

significantly informs their social actions, including their decisions to join gangs. Boys will strive to become men and so

reproduce existing masculine practices, and in their attempts to do so, some of them will reproduce violence. Grasping these

gendered and reproductive dynamics can enhance structural explanations of urban violence, namely those that consider the

effects of socio-economic exclusion.

The challenge is to tackle armed groups as "masculinisation opportunities" and "standard bearers" of masculinity, and to

provide alternative options that confer capital and dignity upon male youths where gangsand socio-economic exclusion are

prevalent. We can think of this in terms of generating more possibilities for young men in these contexts to achieve dignified

masculine livelihoods. Beyond this we may need to challenge power relationships and the dominating nature of traditional

masculinities themselves, but that issue is beyond the remit of this paper.

Interviews

Andrs (20 July 2008) Interview, Medelln.

Aristizabal (15 July 2008) Interview, Medelln.

Carritas (16 July 2008) Interview, Medelln.

El peludo (3 June 2008) Interview, Medelln.

Hernando (21 June 2008) Interview, Medelln.

Mauricio (20 June 2008) Interview, Medelln.

Notes (16 July 2008) Interview, Medelln.

Pelicorto (10 June 2008) Interview, Medelln.

Pepe (11 April 2008) Interview, Medelln.

Sammy (3 June 2008) Interview, Medelln.

This research was assisted by a grant from the Drugs, Security and Democracy fellowship program administered by the

Social Science Research Council and the Universidad de Los Andes with cooperation and funding provided by the Open

Society Foundations and the International Development Research Centre.

Footnote

1. In a city of 1.6 million a staggering total of 6,349 homicides were recorded that year, and in total between 1986 and 1993

there were 33,546 homicides ([70] Marquz Valderrama and Ospina, 1999, p. 14). For comparative purposes, Perlman

(2008, p. 52) refers to Rio de Janerio being one of the most violent cities in the world where in 2004. The homicide rate then

was 37.7 per 100,000, a tenth of the homicide rate in Medelln in1991. In 2008, the rate in Medelln then began to climb

again to 45.6.

2. In a longitudinal study between 1990 and 2002, the vast majority of homicide victims in Medelln were consistently young

males between 15- and 34-year old, a trend which remained unchanged in 2009 ([52] Surez Rodrguez et al. , 2005; [22]

Hylton, 2010).

3. These can be understood as "structural" factors. [61] Galtung (2002) has further described factors such as inequality,

exclusion, racism or sexism as forms of "structural violence". [30] Moser and van Bronkhorst (1999) add that political

legacies provided the foundations for present day violence in Latin America and the rapid rise in urban violence is

interrelated with factors such as rapid urbanisation, poverty, exclusion, weak governance, the increasingly organised nature

of crime and increased drug proliferation. "Multiple marginality" ([30] Moser and van Bronkhorst, 1999; [14] Covey, 2003).

4. [49] Scheper-Hughes and Bourgois (2004) list a number of conceptual explanations of violence - cycles, mirrors, spirals,

mimetic, and reproductive - arguing that "we can rightly speak of... a continuum of violence ([49] Scheper-Hughes and

Bourgois, 2004).

5. French Sociologist Pierre Bourdieu's "thinking tools" from his Theory of Practice , were used to analyse the empirical data.

One of these "tools" is habitus , which explains how individuals are disposed to reproduce existing social practices, which is

used to understand the reproduction of masculinity, or the masculine habitus . "The habitus, the durable installed generative
principle of regulated improvisations, produces practices which tend to reproduce the regularities immanent in the objective

conditions of the production of their generative principle ... " ([6] Bourdieu, 1977, p. 78). Masculine habitus has also been

used by [13] Coles (2009).

6. CoVida is a pseudonym.

7. Youths engaged in "positive" social activities have also been referred to as "prosocial" ([62] Daly and Wilson, 1997; [63]

Barker, 2000).

8. Of course, not every youth reproduces the same version of "masculinity" ad infinitum , but habitus does lead to a

generalised continuum of "maleness".

9. Habitus works at a "less than conscious" level ([6] Bourdieu, 1977, pp. 78-9; [23] Jenkins, 1992, p. 76). Habitus is "a set of

basic, deeply interiorized master-patterns; mental habit; cultural unconscious ... " ([53] Swartz, 1997).

10. Field and capital after [6] Bourdieu (1977). Capital is used here in reference to "masculine capital", the localised assets

or signifiers of identity. These capitals confer male identity in a relational field, linked to male esteem and dignity. This type

of capital is distinct to [29] Moser' (2009) use of "perverse social capital" inreference to gangs.

11. See [43] Rocha and Rodgers (2008, p. 91) and [19] Hagedorn (2008, xi(7), pp. 18-9). [46] Rodgers (2010, p. 4) argues

that disorder and ontological insecurity play indirect roles in the emergence of gangs as forms of "social sovereignty".

12. See [37] Pearlman (2009) gangs tend to have a greater negative impact with increased involvement indrug dealing. [45]

Rodgers (2007) and as [29] Moser (2009) have noted the "perverse social capital" of gangs is based on force, violence,

illegal activities.
 13. One complexity of this study is that a small percentage of youth actually joined gangs, but an even smaller percentage

joined community development organisations. "Prosocial" and "violent" youths were divergent from the "normal"

youths in terms of their social practices.

Baird, A. (2012). The violent gang and the construction of masculinity amongst socially excluded young men. Safer
Communities, 11(4), 179-190. doi:http://dx.doi.org/10.1108/17578041211271445

Youth in terroristic groups, gangs, and cults: The allure, the animus,
and the alienation
Levine, Saul, MDAuthor Information . Psychiatric Annals 29.6 (Jun 1999): 342-349.

1. Full text
2. Full text - PDF
3. Details
Hide highlighting
TranslateFull text
Turn on search term navigation
Since time immemorial, youth (pradolescents, adolescents, and young adults) have "manned the barricades" in rebellions

and revolutions. Perceived as "cannon fodder" by some, as idealistic warriors by others, and as violent thugs or dangerous

terrorists by still others, young people have been both perpetrators and victims of violent confrontations and insurrections

over the millennia. And, as we enter the next millennium, this phenomenon, at least, has remained immutable and constant:

young men (especially) have always filled the roles of the "deliverers" of defiance, danger, and destruction.

We have been only too aware, especially during the past two decades, of the prevalence of guns and violence in America

and other societies.1"8 We have too often been confronted dramatically by the horror of violence in our schools, streets, and

homes. Whereas the frequency of these occurrences waxes and wanes (it is on the wane right now), we should never

assume that this is a phenomenon of the past.9 Although youth involved instreet crimes and violence are

commonplace in our consciousness (one has only to watch television or read newspapers), we are all not as aware of young

people involved in social, political, ethnic, or other ostensibly ideologic acts of violence.

From the Intifada in the West Bank to the cultural revolution in China, from the Hutu-Tutsi conflagration inRwanda to the

Jihads in the Middle East, and from the brutality in the Balkans to the fury of the French Revolution, youth have only too

gladly joined the fray, with energy, commitment, and zealotry. Adolescents and even children have been inextricably

involved in terrorist activities in Iraq, Northern Ireland, India, Turkey, Indonesia, Sri Lanka, and all points in between.

Youth are remarkably susceptible to ideologic and passionate solicitations. They want to believe, with every sinew of their

existence.10"13 Elsewhere we (and others) have written of the propensity of young people to feelings of alienation,

demoralization, and low self-esteem.14"17 Further, we have elucidated that, in the trajectories of adolescents and young

adults, four crucial elements, arbitrarily designated as the four Bs (being, belonging, believing, and benevolence), serve as

intrinsic needs at that stage, as monitoring milestones of one's life, and, much later in the life span, as cornerstones for

assessing (in retrospect) the worth and mearimgfulness of one's life.1318'19

What is terrorism? This seemingly recent but in reality ageless phenomenon is made up of a number of components. It

is in part determined by the perceived threat in the "eye of the beholder." If an individual (or a group, a community, or a

society) concludes that a serious threat exists to his or her stability, security, and locus of control and that physical harm is

an intrinsic part o that threat, then this individual may well be terrorized (ie, in a state of abject fear or terror).

The terrorist, on the other hand, is an individual, or more often a group of individuals, who, in a spirit of dedicated self-

righteousness, threatens and /or enacts violence against a perceived and demonized target - an individual, a group, an

institution, or a society. There is usually a cabal, or secretive planning group, that provides protection, support, validation,

encouragement, rationale, justification, and internal pressure (centripetal). This phenomenon is seen along a spectrum of

intensities and dimensions. Adolescents thrive ingroups,20,23 and all groups exert some of the above characteristic

influences on their members. However, it is only within groups that have intense ideologic rationales and belief systems, and

charismatic, controlling, and committed leadership, that the propensity and susceptibility exists for violent expressions

toward perceived "enemies."7,16'20

In studying hundreds of members of groups popularly labeled as "cults," we found similarities and commonalities across all

cults (discussed below).15,21,22 However, the cults that threatened or planned physical action had zealous and persuasive

leaders who embodied ideologic principies and anger at targeted enemies. These leaders evinced suspiciousness and

mistrust of outsiders (especially organized law enforcement), expectations of persecution (or even assault), and certitude

about having been wronged or harmed. There is total dedication to their mission of retribution: redress of wrongs is the sine

qua non of terrorist rationale, and eradication of designated villains and their institutions is their ultimate modus operandi.

REASONS YODNG PEOPLE FIND INTENSE IDEOLOGIC GROUPS ALLURING

Critical Period/Existential

Young people have begun their lifelong quest for meaning in their lives.23 Many are brirnrning with enthusiasm, energy, and

idealism. However, many others are beset with existential and identity-related questions and are feeling dispirited

(demoralized), marginal (alienated), and selfdeprecatory (low self-esteem).12"" If, at this crucial juncture (or crtica! period),

they are exposed to the rninistrations and solicitations of individuals and groups who seemingly offer "antidotes" to their

melancholic miasma, they may be particularly susceptible to those invitations. They see a way out, either ideologically

(ostensibly) or psychologically (de facto). If they are attracted to the premise and possibilities of the group and soon discover

that they are energized and enhanced, there is no gainsaying a quantum leap in their perceived quality of life and,

especially, in the way they feel about themselves and their circumstances.

These groups offer to their members16,20,21 sense of:

1. being - they feel intrinsically better, more worthwhile, more esteemed, and more fulfilled;
2. belonging - they feel that they are an integral member of a group, a community, of respected, caring, like-minded

individuals;

3. believing - they feel that they have a "raison d'tat," an important meaning to their lives beyond subsistence and

materialism; and

4. benevolence - they feel that they are contributing to the betterment and the well-being of their fellow human beings.

In this context, we can understand the commitment engendered by these groups and why it is so difficult for a young

individual to: (a) want to leave; (b) extricate himself, if he does wish to leave; and (c) succumb to the desires (or threats) of

parents and loved ones that he quit the group, which, paradoxically, he views as the font of his happiness and

fulfillment in life.

Psychodynamic Factots

Clearly, there are multivariate psychodynamic, relational models and styles of families that could pertain to the allure of

these groups. Some examples include:

1. The young group member might be unconsciously flexing his "muscles" of independence, autonomy, and separation-

individuation in an overt display of rebellion against parental or other authority, leaving his mundane, middle-class pursuits to

demonstrate the superiority of this flagrantly different path.

2. He might paradoxically be joining a group that ostensibly fulfills all his parental teachings of egalitarianism and

antimaterialism and yet subtly "sticks it to them" for their hypocritical, upper-middle-class lifestyle choices (in this manner, he

has his cake and eats it, too).

3. In an authoritarian-dominated household, the young member has a family-energized need for manifesting polemic and

power, and succeeds in a demonstration of ultimate exemplification of supplanting his father's dominance and leadership.

4. The young person "escapes" his parental domination and uses the group as a vehicle of emancipation. He asserts this

autonomy and independence, and has found a legitimate, or rational, way of "escaping."

5. He has been ingrained with ideologic messages in his politicized family and has adopted them "en face et en masse," with

a total, unequivocal, ideologic embrace.

Perhaps more important is the fact that all human beings have a "dark side," borne of the confluence of biogenetic and

psychosocial factors. Given the "right" context and circumstances, we all can manifest acts of terror that, ordinarily, we

would eschew and abhor.24"26

Personality Needs/Psychopathology
Depending on such things as temperament, intelligence, mood, genetics, and social circumstances, some young people

have a propensity to a particular behavior. Some primitive and primal needs might be fulfilled by the group. For example,

some young individuals join groups as a vehicle for expression of their anger at traditional authority. Others join because of

an inordinate need for peer acceptance, and communal support. Still others have their needs for security and clarity fulfilled

by the diminution of perceived ambiguity, nebulousness, and drifting. Some members appreciate the rules, order, and rituals

imposed by the group's hierarchy and protocol. Some require the stimulation of the group's social and ideologic activities to

keep themselves "activated," or even "high." And, of course, some join for reasons other than ideology: given their young

ages, they may well feel a heightened sense of erotic arousal in a welcoming group of peers.

As DSM-IV is currently organized, most pathological entities represent "spectrum disorders" (ie, pathology is based on

enumeration of symptoms and severity over time), and yet many normal individuals manifest some degrees of the traits and

symptomatology, in lesser frequency and confluence. However, there are many DSMIV disorders that predispose to overt

manifestations of aggressive and even violent behavior, even at a very young age.27 Individuals with impulsive disorders,

substance abuse disorders, hypomanic behavior, paranoid ideation, thought disorders, psychoses, psychopathy, and other

psychiatric conditions have certainly been members of various intense groups.21 Blatant psychopathology that infringes on

others or severely limits the individual's performance of duties or day-to-day functioning and living usually ensures some

form of extrusion from the group. In some instances, the group can act as a "cover," a structured expression of the inner

turmoil, and even a therapeutic community for some severely disturbed individuals.21

Ideology/Tzeitgeist

Although most young people need to believe and belong, when radical groups afford them this opportunity, a small number

will gravitate to them for their dissatisfaction with the status quo. Those who do join intense ideologically based groups,

whatever their individual personalities and psychodynamics, fervently believe inthe absolute righteousness and the moral

superiority of their "cause clbre." They are convinced that the ultimate goals of their community of true believers are

warranted and worthy, no matter what the methods used to attain them (the ends do justify the means). In interviews with

hundreds of cult and other intense group members, it became obvious that although there were many rationales offered for

joining, usually along ideologic lines (ostensible), the psychological (de facto) factors prevailed in most

instances.15,16,21,22

History and contemporary life are replete with myriad examples of "rationally" induced conflict and conflagration, whether

they be quasi-religious, quasi-political, quasi-racial, quasi-territorial, or quasi-ethnic innature. We can seemingly always

"rely" on human beings to bear vengeful internecine feelings and actions toward one another.19,24,26,28'29

Unfortunately, some young terrorists have been "bred in the bone." If their families and, even more so, their schools have

inculcated them from an early age with hate and demonization of perceived enemies (eg, Iraq, Taliban, and Iran), it is to be

expected that at least some will end up in radical movements. The availability and the use of guns has made gun violence

almost commonplace in many communities. In addition, contemporary, instantaneous communication, especially via
television and the Internet, makes propaganda inthe service of radical beliefs available. In sophisticated advertising, violence

may be extolled and glorified. And, young people now can easily access information and materials for weaponry, explosives,

and other means of destruction.

Psychological

An individual in his late teens or early 20s might have areas of personality and pursuits that are well developed and mature,

but other traits and behaviors that are immature or stagnant. Individuals tend to be more labile and malleable at this age. As

a result, under conditions of severe stress, for example, one might regress to earlier, immature behaviors.29

But, in general, this is the most healthy, vigorous, robust, and energetic period of a person's life. Young adults are at their

peak physically (eg, cardiovascular functioning, strength, and respiratory functioning). They are functioning

maximally in cognition, reasoning, memory, and abstract thought.

When this enormous potential is not realized, a group and an ideology, which can captivate and harness this energy, proves

to be a boon to the drifting individual. There is also good evidence that no matter what the ideologic rationale for joining

these groups, the committed members feel better - their self-esteem is enhanced, their symptoms of anxiety or depression

are diminished, their "vegetative" functions (eg, appetite and sleep) and their sense of physical health are improved, and

their perceptions of their competence, locus of control, social acceptance, and optimism are all reinforced.13,15'16,21 Also,

there is sometimes seen a less appealing feature of a kind of "bliss" - overt happiness without the reciprocity of warm

feelings in others. A variant on this is a smug self-satisfaction that can easily be taken as abject arrogance by outside

observers. Some feel more aggrandized, or even notorious or glamorous (in the popularized traditions of Carlos, Che

Guevara, Mao, and Pancho Villa).

In addition, these individuals feel special, elitist, and privileged to serve (a) with their peers, (b) under their leaders, (c) the

tenets of their belief system, and (d) their fellow citizens. They believe all of this with a tenacity and a passion that can

border on zealotry, intolerance, and, occasionally, danger to others.

Gangs

There are characteristic elements in most societies that predispose certain young people to join groups that are popularly

perceived as particularly nefarious and dangerous. There are, of course, street gangs, which used to be a phenomenon

largely restricted to inner-city, congested areas in large cities. However, more recently, with instantaneous communication

and transmission or cultural ethos, even rural, relatively sparsely populated areas are no longer immune.

These groups are usually homogeneous in race, gender, socioeconomic class, ethnicity, language, immigration status, and

values. The factors that contribute to the development of particularly virulent gangs are listed inTable l.1"3,6,7'25
Gangs vary in the degrees of violence and antisocial activities that they inflict on each other, on competing gangs, or on the

public at large. Some are extraordinarily dangerous, brutal, and corrupt, whereas others are not dissimilar to relatively pacific

groups of young men (especially). Some engage in egregiously illegal activities (eg, shootings, armed robbery, drug

distribution, "protection," or loan sharking). Others fight over their demarcated territory, over their women, or over a

perceived slight to their honor. Although gangs differ from cults and political terrorists, all of these groups share common

social and structurai characteristics. Table 2 contains a list of the common features of radical groups that can captivate the

imagination and energy of young peopie.13,15'16,22

Enlarge this image.

Factors That Contribute to the Development of Particularly Virulent Gangs

Enlarge this image.

Common Features of Radical Groups That Can Captivate the Imagination and Energy of Young People

CONTINUUM OF RHETORIC

It is clear that there have always been, and will always be, radicals at the extreme poles of any dimension of human opinion,

belief, and behavior. There are standard deviations from the mean along a bell-shaped distribution curve for any human

ideology and endeavor. Just as in the above discussion of symptoms and psychopathology, it is the degree to which certain

behaviors are espoused or pursued and the extent to which others' Lives are compromised that determine the imposition of

the label "dangerous."

For every group labeled as "terrorist" in nature, there are perhaps hundreds or even thousands that share many of the

commonalities mentioned above and yet are civic minded, accepting of others, and acceptable to the community. Wherein

lies the difference? Many (most) groups evince a strong set of values that they are proud of and that they promulgate. It is

when the believer becomes more zealous that discomfort or a potential threat to security is perceived. Those heartfelt

precious values can evolve into a more all-encompassing belief system that is more systematic and that has a uniform prism

that filters the many human phenomena being witnessed.

This author is not sure that the adage "the pen is mightier than the sword" is valid, but there is no doubt that words can

influence behavior, engender fear, inflame passions, and instigate violence. A true believer is one whose perception of

reality is determined by the all-embracing nature of his belief system.11"13 He might adopt a zealous stance, upholding with

fervor a set of beliefs that may well be construed and interpreted by others as peace loving, forgiving, tolerant, and

antithetical to zealotry. The "mesmerization" of group members to become violent "defenders of the faith" is usually

dependent on many factors, which have already been alluded to (eg, leadership, group pressure, personality, or Tzeitgeist).

Often, individual members can adopt a terroristic stance independent of the inflammatory rhetoric of leaders or writings.

Although groups can and do provide a viable source of the four Bs, there is a propensity for some to escalate along the

following continuum:
values - ideals - beliefs - zealotry - violence

Even lofty ideals can gradually presage intimidation and violence, given the "right" inflammatory confluence of factors or

social and psychological "chemistry." Combinations of such things as radical, inflammatory leadership, passionate polemics,

perceived wrongs, prejudices, and personal needs coalesce to foment a culture of confrontation: constructive dialogue is cut

off and the rationale for violence is given a "civilized" facade.

The young people who are increasingly willing to use physical threats, coercion, extortion, torture, and murder might well

have been, a few years earlier, exponents of peaceful persuasion and proselytization (eg, Patty Hearst). On the other hand,

we have found that those who are the most prone to violence have had historically consistent, developmental trajectories;

physical mtimidation has long been an integral part of their lives.

FATE OF YOUTHFUL TERRORISTS

The terrible effects of terrorism are abject and multifold, regardless of whether the violent mtimidations and acts are

"effective." Tragically, there are victims and there are escalating retributive actions. Violence invariably begets more

violence. The ripple effects on the friends, families, acquaintances, and communities of the victims are palpable and

poignant, both at the time of the acts and for generations thereafter.

What happens to the legions of recruits in radical groups throughout the world? If we conclude that the average member of

an intense, activist group is young, then new members must be supplanting older ones who drop out. This is clearly

seen in studies of cults, where the "half-life" is less than 2 years (ie, by the end of the second year of membership,

approximately 95% of the members have left the group).21 Members become disillusioned. They may be resentful of the

personal loss of individuality, the rigid hierarchy, the hypocrisy of the leadership, and the double standard in lifestyle for

leaders and followers. They may become homesick, longing for family, friends, school, or creature comforts.

It is typical for most zealous individuals to lose their dedicated, focused energy and to "burnout." Youthful ideologues,

revolutionaries, zealots, provocateurs, and terrorists tend to dissipate their passionate zeal over time and redirect their

energies to personal pursuits in fiscal, familial, social, and political directions. There are ample examples of older politicians,

leaders, and statesmen who, in their youth, espoused radical philosophies and civic disobethence, openly called for the

overthrow of tyrannic governments, or even engaged in violent acts. Some served time in jail for their actions, but many

others did not. Some have even become pillars of their communities.

What happens to the others? Incarcerations and death take their toll on the diehard zealots. Others fade away, resuming

mundane, middleclass existences.

IMPUCATIONS for our youth, critical CHALLENGES TO SOCIETY

Many young people are susceptible to the overtures of groups that offer clarity, simplicity, communality, and an overriding

belief in the righteousness of a cause. This propensity of youth has been widely used in preventive, therapeutic, and

correction programs for substance abusers, antisocial behavior, and other disorders. But, when it comes to offering most of

our younger citizens constructive, socially sanctioned outlets for their energy, idealism, and commitment, we have decidedly

not been as assiduous or even concerned. Clearly, no matter what is done, as long as there are (a) real or perceived

inequities or inflicted injuries in society, (b) charismatic and controlling leaders with a mission to radically alter structures or

right wrongs, and (c) youth who are wayward, disordered, or searching, there will be the advent of dangerous and

destructive individuals and groups.25

Studies have shown that early intervention programs can reduce the incidence of violent propensities and actions.30,31

However, we must attempt to inspire our youth, through education and example, to idealistic goals and pursuits so as to

harness their latent energy and desires.19 There have been some successful government-sponsored programs around the

world (eg, Youth Corps, Vista, Katimavik, and Youth Aliya), but these have been on a relatively small scale. Churches,

temples, scouts, volunteer organizations, youth groups, and others have all filled this breach. However, as long as we as a

society avoid values and idealism, choose to pursue materialism and acquisitiveness, and increase the gap between the

haves and the have-nots, the more we are tilling the soil for anarchistic or destructive individuals or groups.

We must recognize that even in the "best" societies, angry, resentful, and potentially explosive individuals and groups are

prevalent. Conflicting ideas can be encouraged and civil disobethence tolerated. However, when violence is a probability,

active preventive intervention is a necessity. Young people who are clearly dangerous, regardless of their justified rationale

or their psychodynamic background, must be confronted and controlled.

We must work toward a more egalitarian society and must recognize that there will always be discrepancies inwealth and

opportunities. Humanistic and community values, overriding social concerns, and respect for diversity are laudable and,

indeed, vital goals. But, we must also educate and demonstrate that intolerance, mtimidation, and especially violence are

universally forbidden and will be punished expeditiously and severely. We must be vigilant and alert to possible clues to

zealotry and violence by individuals and /or groups, and must treat all terrorists as pariahs and criminals.

Democratic, progressive societies have little choice but to exist in suspended balance, as it were, between tolerance of

conflict on the one hand and intolerance of violence on the other. Our youth must be made aware of their parents' antipathy

toward any form of terrorism. However, at the same time, a spectrum of outlets must be available for young people for

ideologic expression, communal membership and commitment, and altruism toward fellow citizens.

References

REFERENCES

1. Violence among youth. Am] Orthopsychiatry. 1996;66.

2. Block CR, Block R. Street Gang Crime in Chicago. Washington, DC: U.S. Department of Justice; 1993.
3. Callahan CM, Rivera FR Urban high school youth and handguns. JAMA. 1995;267:3038-3042.

4. Earls F, Carins R, Mercy J. The control of violence and the promotion of nonviolence in adolescents. In: Promoting the

Health of Adolescents: New Directions for the Twenty-First Century. London: Oxford Press; 1993.

5. KiEios M. International correlations between gun ownership and rates of homicide and suicide. Can Med Assoc J,

1993;148:1721-1723.

6. Kleck G. Point Blank: Guns and Violence in America, Gainesville, FL: Florida State University Press; 1993.

7. Jankowski MS. Islands in the Street: Gangs and American Urban Society. Berkeley, CA: University of California Press;

1992.

8. Reiss D, Richters J, eds. Children and Violence. New York: The Guilford Press; 1993.

9. Wilentz S. Bombs bursting Ln air, still. New York Times Magazine. June 26, 1995:41^13.

10. Adelson J. The development of ideoogy in adolescents. In: Drogastin S, Elder G1 eds. Adolescence in the Life Cycle.

New York: Wiley; 1976.

11. Hoffer A. The True Believer. New York: Harper and Row; 1951.

12. Levine S. The role of beliefs in adolescents. In: Michelson W, Levine S, Spina A, eds. Child in the City: Changes and

Challenges, vol. 2. Toronto: University of Toronto Press; 1979.

13. Levine S. Adolescents, believing and belonging. In: Adolescent Psychiatry, vol. 2. Chicago: University of Chicago Press;

1979:41-53.

14. Levine S. Alienation as an affect in adolescents. In: The Adolescent and Mood Disturbance. New York: International

Universities Press; 1982.

15. Levine S. Alienated Jewish youth and religious seminaries. In: Adolescence, vol. 19. New York: Libra Publishers;

1983:183-199.

16. Levine S. Radical Departures: Desperate Detours to Growing Up. San Diego: Harcourt, Brace Jovanovich; 1984.

17. Levine S. The myths and needs of contemporary youth. Annals of American Society of AdolescentPsychiatry.

1987;14:48-62.

18. Friend D. The Meaning of Life. New York: Harper and Row; 1996.
19. Levine S. The Tao and Talmud of adolescent trajectories: being, belonging, believing, benevolence. Annals of American

Society of Adolescent Psychiatry. In press.

20. Braungart R. Youth and social movements. In: Drogastin S, Elder G, eds. Adolescence in the Life Cycle. New York:

Wiley; 1976.

21. Levine S. Cults and mental health: clinical conclusions. Can J Psychiatry. 1981;26:534-539.

22. Levine S. Youth and religious cults: a societal and clinical dilemma. In: Adolescent Psychiatry, vol. 6. Chicago: University

of Chicago Press; 1978:75-89.

23. Erickson E. Identity: Youth and Crisis. New York: Norton; 1968.

24. Arendt H. On Violence. New York: New York Review of Books; 1968.

25. Levine S. The development of wickedness: from whence does evil stem? Psychiatric Annals. 1997;27:617-623.

26. Whiteson L. Creative Violence: An Alternative View of the Role of Violence in Human Affairs. In press.

27. Redlich F. Children Who Hate. Chicago: University of Chicago Press; 1962.

28. Lorenz K. On Aggression. New York: Harcourt, Brace, World; 1968.

29. Rutter M, Rutter M. Developing Minds: Challenges and Continuity Across the Life Span. New York: Basic Books; 1993.

30. Richter J, Martinez, PE. Violent communities, family choices, and children's chances: an algorithm for improving the

odds. Dev Psychopaihol 1993;5:609-627.

31. Zigler E, Taussig C, Black K. Early childhood intervention: a promising preventative for juvenile delinquency. Am

Psychol. 1992;47:997-1008.

AuthorAffiliation

Saul Levine, MD

AuthorAffiliation

Dr. levine is Director, Institute of Behavioral Health, Qiitdren's Hospital and Health Center, San Diego, and Professor and

Director, Division of Child and Adolescent Psychiatry, University of California, San Diego. Address reprint requests to Saul

Levine, MD, 3030 Children's Way, Suite Ul, Sa Diego, CA 92123.

Word count: 4249

Copyright Slack, Incorporated Jun 1999

Levine, S., M.D. (1999). Youth in terroristic groups, gangs, and cults: The allure, the animus, and the alienation. Psychiatric
Annals, 29(6), 342-349. Retrieved from http://search.proquest.com/docview/217049907?accountid=37714

American gangs: There are no children here

Anonymous. The Economist 333.7894 (Dec 17, 1994): 21-23.

1. Full text

2. Abstract/Details
Hide highlighting

Abstract
Translate [unavailable for this document]

Since the late 1800s, teenagers (usually poor male immigrants) have banded together in American cities to act tough and

fight over turf. Where gangs were once regarded with a certain vague romanticism, they now inspire an appalled

fascination. In the cities, they are a source of unremitting fear. In the past decade, trouble has given way to chaos

as gang violence has become more common, vicious, and random. Gangs are a backbone of the drug trade, and their

activities are increasingly lethal. But the connection between gangs, drugs, and the mayhem that plagues so many inner-city

streets is less straightforward than the received wisdom has it - and more intractable. The right question revolves not

around why gangs are getting violent more often but around why the same incidence of gang violence is getting more lethal.

One answer is simple: weapons are becoming more sophisticated, potent, and deadly. The gang crisis is deeply entwined

with the US' most intractable social failure: the entrenchment of its underclass.

Full Text
Translate [unavailable for this document]

FOR several sweltering days in summer, of America could talk about was Robert Sandifer. A member of one of Chicago's

notorious street gangs, the Black Disciples, Sandifer was suspected of having sprayed bullets from a semiautomatic pistol

into a group of teenagers on the city's poverty-stricken South Side. One young girl was killed. Days later, so was Sandifer--

by his own gang. This much was harrowing but unexceptional. What shocked the country was the fact that Sandifer, a 23-

times convicted felon, was all of 11 years old.

Among gang members, shock does not come so easily. Ask some of the older ones around Chicago--the ancient warriors

still alive after 30--about Sandifer, and they may tell you instead about a killing a month earlier in a west-side ghetto

controlled by a gang called the Four Corner Hustlers. Early one July night the gang's founder, Walter "King" Wheat, was

sitting in his dull-orange Oldsmobile when a shirtless youngster rode by on a bicycle and blew his head off. The boy, as it

happened, was a member of the very gang his victim helped launch in 1968.
It was not, however, a gang that Wheat and his generation recognised any longer. Such old-timers recall, almost wistfully,

that the Four Corner Hustlers used to live by a code. True, they were violent. True, they were criminals (mainly burglary and

extortion). But there was loyalty, too. Children were protected and innocent neighbours were not hassled. Then things

changed. The gang grew huge and indiscriminately murderous. Internal warfare raged. And suddenly the Hustlers'

membership was dominated by wild adolescents who, inthe words of one gang vgeteran, "would just as soon smoke you as

look at you."

Walter Wheat and Robert Sandifer are part of the same story. Since the late 1800s teenagers (usually poor male

immigrants) have banded together in America's cities to act tough and fight over turf. But where gangswere once regarded

with a certain vague romanticism--think of "West Side Story"--they now inspire an appalled fascination. In newspapers, on

television and in rap music, gang culture has merged into popular culture. Two famous Los Angeles gangs, the Bloods and

the Crips, are household names across America.

In the cities they are also a source of unremitting fear. Gangs, of course, have always caused trouble. But inthe past decade

trouble has given way to chaos, as gang violence has become ever more common, vicious and random. From coast to coast

there are plenty of places where gunfire provides a steady background noise; where schools close when shots ring out;

where children are dressed with care each morning because the wrong colours could make them targets, and put to sleep

under their beds some nights because stray bullets are less likely to find them there.

How did gang violence get so out of control? Politicians, policemen and the press repeat the received wisdom. Gangland,

they say, has "one corporate". In the 1980S the emergence of crack cocaine as the poor urbanite's drug of choice

spurred gangs to turn themselves from loosely organised groups into highly structured, brutally efficient businesses. As cash

began to flow, drug territory became more valuable, and hence more worth killing for. The body count rose.

This view is widely held. It sounds sensible, perceptive. But despite having elements of truth--the reorgangising of gang into

quasi-corporations, for example--it is not the right explanation. Gangs are indeed a backbone of the drug trade. Their

activities are increasingly lethal. But the connection between gangs, drugs and the mayhem that plagues so many inner-city

streets is less straightforward than the received wisdom has it--and more intractable.

Disciples of death

Behind the grim images lie grim statistics. Called everything from "crews" (in Washington, DC) to "posses" (inRaleigh, North

Carolina), gangs operate almost everywhere bar the smallest towns--and there too, sometimes. A survey by the National

Institute of Justice in 1992 found that the police knew of nearly 5,000 gangs with 250,000 members in America's 79 biggest

cities; there are probably many more that do not make it into the police figures.

Every city's gang problem has its own character, but Chicago provides a vivid glimpse of trends being seen across the

country. Police there know of at least 40 big gangs. It is, however, two giant black ones, the Gangster Disciples and the Vice

Lords, and two Hispanic ones, the Latin Kings and the Latin Disciples, that matter most.
These four gang account between them for half of the city's 50,000-odd active gang members. Each has been around for at

least 30 years. Their long histories have often intersected, notably in the mid-1980s when the four banded together into two

rival alliances: the "Folk" (Gangster Disciples and Latin Disciples) and "People" (Vice Lords and Latin Kings). Such unions

are not uncommon. Indeed, during the 1970s much of the growth of this "big four" came through mergers with smaller rivals

seeking strength through size. This expansion quickened in the 1980s when the gangs diversified aggressively into the drug

trade, a business then ripe for expansion into new markets.

The corporate lingo fits. For gangs today are organised (on paper at least) like most companies: with flow charts, finance

committees and boards of directors. In a South Side stronghold of the Gangster Disciples, a 20-year-old "area co-ordinator"

called D-Max says he controls a stretch of turf ten blocks long and five blocks wide. Working for him are 150 "soldiers", ten

of whom are his personal "security staff". Above him are the gang's "regents", and its 12 or so "governors", the middle-aged

leaders who oversee the entire operation--often from prison cells. D-Max gets drugs from the gang, which he sells wholesale

or retail. Most of the profits, plus the "street taxes" paid by dealers who want to work in "his" area, are passed up the chain

of command.

The sums can be vast. Officials guess that the Gangster Disciples rake in some $300m a year from drug sales. Some of

these profits are from marijuana; some from heroin. But the bulk comes from cocaine. If the ganghave always

dabbled in drugs, the cocaine explosion of the 1980s--especially in the cheap, crystallised derivative known as crack--turned

dealing in the stuff into their main function. A 1989 study by the General Accounting Office estimated that the Bloods and

Crips together controlled fully one-third of America's crack market.

With one explosion came another: in violence. According to the same National Institute of Justice survey, America's cities

witnessed more than 1,000 "gang-elated" murders in 1991. What this means is not entirely clear, because different cities

define the term differently. In Los Angeles and some other cities, any murder involving a gang member is considered gang-

related; elsewhere, a homicide is not counted as gang-related unless its motive is somehow tied to a gang's activities.

However it may be defined, no one doubts that gang violence is soaring in city after city. In Chicago (which uses the more

restrictive definition) the police blame gangland slayings for virtually the entire increase inmurders over the past five years.

Last April the city's worst public-housing project, the Robert Taylor Homes, was the scene of a vicious outbreak

of gang warfare. In the space of three days the police received more than 300 reports of gunshots; 13 people were killed. "It

got so bad", recalls one young mother, "that I wouldn't let my kids go outside--even to walk to school."

Warfare this extreme is rare; but the questions it raises are not. After the shooting stopped, police came up with two theories

about its cause. One held that the Gangster Disciples and the Black Disciples were feuding over drug territory. The other

was that a squabble between members of the two gangs over a girl had escalated into a bloodbath. Predictably, the first

theory was reported by the press and repeated by politicians. The second was ignored.
Yet the truth is that, to this day, no one is really sure what sparked the madness at the Robert Taylor Homes. The same

could be said of countless other outbreaks of gang violence. The idea that such instances are linked directly to drug dealing

is both logical--lucrative markets demand vigilant defences--and widely accepted. Gangs are often out to make money. But

the evidence suggests that when bullets fly and bodies fall, something other than the profit motive is usually at work.

A pity about the coat

Often that something can be absurdly trifling. Kody Scott, for years a famously cruel leader of Los Angeles's Eight-Tray

Gangster Crips, and now a maximum-security prisoner (and author of an autobiography entitled "Monster"), recalls that "one

of the biggest wars inside the Crips goes back to when a young girl had a leather coat taken by a guy in my neighbourhood.

It erupted into a full-scale war, with people getting dropped. And today 25 or 30 deaths can be attributed to that one coat

being taken."

Mr Scott's story is far from unique. Asking gang members why they kill each other elicits long and varied litanies of minor

slights and empty symbols. A "homeboy" wearing blue "rags" (bandannas) wanders on to turf where red is the only

acceptable colour. Another drives into rival territory and "throws" his gang's hand sign at a clutch of enemies. A girlfriend

gets a sidelong leer; a younger brother's bicycle is stolen. These stories are scarily simple--and rarely do their narrators

mention drugs.

unsurprisingly, perhaps: why would they incriminate themselves? But gang members are startlingly candid (if unspecific)

about their crimes. They admit to selling drugs; they admit to drive-by shootings. What they say far less often, however, is

that the two have anything to do with each other.

Three academics at the University of Southern California who have spent almost ten years studying the connections

between gang, crack and homicide in Los Angeles have found them to be so weak that they have urged the police to "move

away from gang specialisation in narcotics enforcement." A former Harvard professor, Walter Miller, reckons that only 10%

of violent gang crime in Boston from 1984 to 1994 turned on drug dealing or use. In Chicago, two researchers, Carolyn and

Richard Block, discovered that, of 288 gang-related murders in the city between 1987 and 1990, only eight were drug-

related.

These numbers are not as incredible as they seem. Start with the fact that Hispanic gangs, which account for half of all big-

city gang members, are among the most violent; yet they are not terribly involved in drugs. Last year in San Antonio, a city of

5,000 (mostly Latino) gang members, 1,200 drive-by shootings were reported; police think ten times that many actually took

place. But among the city's main gangs--LA Boyz, Big Time Criminals, Hispanics Causing Panic--drug trafficking is no big

deal.

That could never be said of the Crips or the Vice Lords. But although most black gangs are plainly both violent and

"entrepreneurial", the two sides do not mix much. There is a clear division between "gang-bangers", the hardcore, gun-toting

young men whose life is the gang and whose intimidating images turn up in magazines, and ordinary gang members.
Some gang members have jobs, others go to school, and they deal most of the drugs. "Bangers stand at the corner, hats on,

colours flying, down for the shoot," says Curtis, a teenage Gangster Disciple. "Gang members be in the alley, serving

someone, making dollars."

No doubt these eager young capitalists rely at times on their allies for muscle. No doubt turf battles do at times occur. But if

the claim that street gangs are behaving more like traditional organised-crime operations means anything, it means that

competing factions are able (usually) to coexist peacefully so that everyone can profit. Gang members know that, as one

puts it, "too much shooting scares the customers away." Thus members have a strong incentive to minimise the bullets

flying through their places of business.

Gangbangers do not care much about business. They deride the mere members as "punks". For them, to be ina gang is not

mainly a matter of money, but of power, reputation and status. That is why the most trivial things can cause battles to break

out; why, once the shooting starts, the tit-for-tat can escalate fast into all-out war; and why, once the war is over, few on

either side can give an accurate account of how it started. Violence is an end, not a means.

Talking about one savage skirmish in Los Angeles in the 1980s, Mr Scott gives an account echoed by many gangbangers:

Our war, like most gang wars, was not fought for territory or any specific goal other than the destruction of individuals. The

idea was to drop enough bodies, cause enough terror and suffering so they'd come to their senses and realise that we were

the wrong set to fuck with.

Youth's a stuff will not endure

Bravado and' machismo are nothing new in gangs. Yet somehow they seem to be producing a higher order of havoc. If

drug-dealing is not the reason, what is? What else has made gangs so much more violent?

Maybe the assumption behind the question is not quite right. Lurching from funeral to funeral, it is easy enough to

see why inner-city dwellers think that gangs are getting increasingly violent. Yet the few reliable studies there are of such

things suggest that gang may not be committing a greater number of brutal crimes today than they did in years past; it is

rather that the crimes they commit happen to be taking more lives. The right question, then, revolves not

around why gangs are getting violent more often, for they may not be, but around why the same incidence of gang violence

is getting more lethal.

One answer is simple: weapons are becoming more sophisticated, potent and deadly. As recently as ten years ago,

say gang members, the deadliest arms you could lay your hands on easily were sawn-off shotguns. Now automatic and

semi-automatic assault weapons are de rigueur. In a burnt-out flat in Chicago's East Garfield Park, several Vice Lords

display a collection of AK-47s, Uzis, "street sweepers" and Tec-9s that would make a mercenary quiver.
The effect of such technology on the body count is clear enough. Where once gangbangers could squeeze off only a handful

of shots, one at a time, now they are able to take out literally dozens of enemies in a single spray. And not just enemies:

innocent bystanders too. For Curtis, the young Gangster Disciple, this is the unfortunate part. "You try to just hit your mark,"

he says, "but with these guns, you never know who you gonna get."

Curtis should know. Since joining the gang he has taken part in a dozen drive-by shooting, his first at the age of 13. This

hints at another reason why gangs are getting more lethal. Not only are the guns more deadly, but they are found

increasingly in the hands of children. Indeed, if there is one thing every gang veteran points to when asked about

changes in his crew, it is a demographic shift in which gang leaders are getting older and gang soldiers are getting younger.

Jeffery Haynes, an anti-gang counsellor in Chicago, estimates that 80% of boys aged 13 to 15 in the South Side area where

he works are involved in gangs.

In part this may be due to drugs. As gangs have got deeper into dealing in dope, their leaders have found it useful to rely on

early-teen soldiers--who, if they are arrested, will attract penalties far lighter than those imposed on adult offenders.

Recruiting inner-city children is easy enough, given the lure of quick cash. "I wanted a bankroll in my pocket," recalls Curtis.

"It was either join a gang or get a paper route, and I wanted the big money, quick."

But the sad fact is that recruitment is barely necessary. Some children join gangs mainly out of fear. "In this neighbourhood,"

says Mr Haynes, "if you're in a gang, at least you're safe within your three square blocks of turf. But if you're not in a gang,

everyone still thinks you are, so you're a target wherever you go." Other children join more out of emotional need--for a

sense of belonging, structure, self-esteem. Still others join out of habit: because a big brother did so first. Whatever the

draw, the gangs are always there.

That alone makes them unique. They are strong and thriving institutions in a part of the world where every other institution--

family, school, church--has crumbled virtually to dust. It is, therefore, hardly surprising that children who are poor, ill-

educated and, typically, raised by a single parent (of whom a depressing number are unemployed, alcoholic, drug-addicted

or some mixture of the three) should flock to gangs. But equally unsurprising should be the fact that such children, armed

with the most powerful weapons, have brought a new level of violence to gang life. Many of them say they do not care

whether they live or die. Why should they be any more regarding of the lives of others?

Would that there were some ready way out of this hole. Politicians talk endlessly about gangs, but their proposals offer little

hope. A ban on assault weapons, however well-intentioned, will do little when gang-bangers brag that they buy many of their

best guns from dirty cops. Legalising drugs, for all its other virtues, would probably not make gangs much less violent. And,

so far as anyone knows, no public policy has yet been devised to deal with raw nihilism.

The unduckable truth is that the gang crisis is deeply entwined with America's most intractable social failure: the

entrenchment of its underclass. Until politicians and the public are ready to attack that problem root-and-branch--until they
 are ready, in particular, to make a Herculean effort to improve the life prospects of young black men--gangs will grow more

powerful and more wicked. And, soon enough, even the Sandifer case will not seem shocking.

American gangs: There are no children here. (1994, Dec 17). The Economist, 333, 21-23. Retrieved from

http://search.proquest.com/docview/224168745?accountid=37714

PSYCHOSOCIAL CORRELATES OF SUBSTANCE USE

BEHAVIORS AMONG AFRICAN AMERICAN YOUTH
Wright, Darlene RAuthor Information ; Fitzpatrick, Kevin MAuthor Information
. Adolescence 39.156 (Winter 2004): 653-67.

1. Full text
2. Full text - PDF
3. Abstract/Details
4. References 76
Hide highlighting
Abstract
TranslateAbstract

Cross-sectional data were collected on substance use behaviors and potential correlates in 1,494 African American students

enrolled in grades 5-12 in eight schools in a central Alabama school district. Using a risk and asset framework, self-reported

recent alcohol, cigarette, and marijuana use were analyzed by identifying and measuring levels of influence, including

individual, family, and school. For alcohol and marijuana use, recurrent risk factors were age, being hit by a parent, affiliation

with gangs, and a tolerant attitude of peers toward drug use. For cigarette use, risk factors were peer-oriented: associations

with gangs or cohorts holding lenient attitudes about substance use. For all substances, salient asset factors were academic

achievement and parental monitoring. Findings suggest that efforts to reduce substance use behaviors should be directed

at adolescents in terms of academic achievement and grade level as well as their social environments. For the latter,

peer/family risks and family/school assets should be the foci for programs to minimize the short- and long-term

consequences of these behaviors. Hence, the emphasis should be placed on modeling attitudes, preventing gang and family

violence, encouraging parental supervision, and building positive teacher-student interactions. [PUBLICATION ABSTRACT]

Full text
TranslateFull text
Turn on search term navigation

Headnote

ABSTRACT

Cross-sectional data were collected on substance use behaviors and potential correlates in 1,494 African American students

enrolled in grades 5-12 in eight schools in a central Alabama school district. Using a risk and asset framework, self-reported

recent alcohol, cigarette, and marijuana use were analyzed by identifying and measuring levels of influence, including

individual, family, and school. For alcohol and marijuana use, recurrent risk factors were age, being hit by a parent, affiliation

with gangs, and a tolerant attitude of peers toward drug use. For cigarette use, risk factors were peer-oriented: associations
with gangs or cohorts holding lenient attitudes about substance use. For all substances, salient asset factors were academic

achievement and parental monitoring. Findings suggest that efforts to reduce substance use behaviors should be directed

at adolescents in terms of academic achievement and grade level as well as their social environments. For the latter,

peer/family risks and family/school assets should be the foci for programs to minimize the short- and long-term

consequences of these behaviors. Hence, the emphasis should be placed on modeling attitudes, preventing gang and family

violence, encouraging parental supervision, and building positive teacher-student interactions.

While most youth live within reasonably constructive social settings and experience nominal stress, a significant number are

at risk for adverse outcomes (Collins & Repinski, 1994; Jessor, 1998). Individuallevel variables (e.g., gender and age) and

socio-environmental factors (e.g., family and school violence) constitute multiple predictors of unhealthy behaviors and

choices for youth. Alternatively, avoiding health risks is contingent on the extent of asset, or protective, factors such as

positive self-esteem and academic achievement as well as parental and teacher availability and support. Hence, social

experiences play a crucial formative role in translating the adequacy or inadequacy of environmental resources into

behavioral and psychological outcomes.

This is particularly true for low-income African American adolescents who are more likely than Caucasian and Hispanic

youth to reside in poor neighborhoods (Anderson, 1990; Jargowsky, 1997; Wilson, 1987). Massey and Denton (1993) point

out that residential segregation in American central cities is more pronounced for African Americans than other races. Low-

income African Americans are more likely to be isolated in deteriorating neighborhoods than are poor Hispanic and non-

Hispanic whites. Poor local contexts typically expose young residents to violence and crime while isolating them from

conventional role models and employment opportunities. In these poor communities, both families and schools suffer from

inadequate social and economic resources. As such, homes and schools are often unsupportive and unsafe, increasing the

likelihood of adverse behavioral outcomes for young African Americans.

Low-income African American adolescents may therefore be subject to different social processes and structures

than adolescent subgroups with similar or better economic and social resources. This, in turn, may lead the former to

respond in a culturally unique manner not only to the effects of risks but also to treatment and intervention (Johnson &

Hoffmann, 2000; Myers & Taylor, 1998). For this reason, a major goal of this study was to increase understanding of the role

of risk and protection in predicting negative behavioral outcomes among African American adolescents.

The multi-domain perspective of a risk and asset factors approach emphasizes the development of the adolescent within a

dynamic social context that offers ever-changing challenges and threats (Fitzpatrick & LaGory, 2000). This model may be

used to hypothesize age, situational, and transitional variation in youth susceptibility to substance use. By selecting

variables that are applicable to adolescents in general as well as African Americans in particular, we are able to better

understand what factors correlate with substance use outcomes in the sample. In this way, we are able to hypothesize and

discuss why general youth risk variables were or were not associated with alcohol, cigarette, and marijuana use in an
African American sample. We are also able to see how those variables specific to low-income, minority youth

performed in the analysis (e.g., Myers & Taylor, 1998; Xiaoming, Feigelman, & Stanton, 2000).

Previous literature supports the association between persistent stress resulting from physical abuse in the home and

unhealthy behavioral responses (Chaffin, Kelleher, & Hollenberg, 1996; Prino & Peyrot, 1994). Adolescents seek to relieve

tension through risk behaviors such as substance use that only compound the negative psychological and physical toll. The

nature and type of peer interactions are also consistent and powerful predictors of adolescent substance use (Bailey &

Hubbard, 1990; England & Petro, 1998; Ennett & Bauman, 1993; Farrell & White, 1998; Jessor, 1998; Simons, Wu, Conger,

& Lorenz, 1994; Windle, 2000). Peers, who often exert a more powerful influence on decisions and actions than parents or

other authority figures, are particularly significant during middle adolescence for predicting both normative tasks and risk

behaviors such as alcohol and drug use. Research further indicates that parental influence may be a stronger determinant of

substance use for younger teens (e.g., seventh grade) while peer influences are more significant in later development (e.g.,

ninth grade) (Windle, 2000). This variation in response according to age is also indicated for gender, with adolescent males

reporting more substance use than do females. Pressures to adopt certain attitudes are pathways for peer influence.

Association with deviant peer groups such as gangs or with cohorts holding a permissive attitude about drug and alcohol

use may increase adolescent acceptance of and participation in these risk behaviors.

Protective individual characteristics such as high grade point average are reported as consistent indicators of decreased use

of alcohol, cigarettes, and marijuana (Dekovic, 1999; Resnick et al., 1997; Williams, Stiffman, & O'Neal, 1998). This positive

aspect of academic life is often associated with a belief that challenges or adversities may be overcome. This, in turn, may

increase personal resiliency and capacity for constructive coping and reduced risk behaviors. Minority adolescents who feel

that academic commitment will create opportunities for educational success and occupational advantages may be less

vulnerable to destructive acts than those who believe it will not. Also, high self-esteem is generally recognized as an asset

and has been associated with decreased adolescent substance use. Adolescents with positive self-concepts, as

characterized by feelings of pride and self-worth, are better able to cope with developmental and social challenges of

adolescence including identity formation and establishment of autonomy. Parental monitoring and supportiveness are

important dimensions in regulating adolescent risk behaviors. An appropriate level of monitoring encourages youthful

exploration and learning within the confines of supervision and control, thus reducing conduct problems and delinquency

and increasing psychosocial competence (Brown et al., 1993; Crouter, MacDermid, McHale, & Perry-Jenkins, 1990; Gamier

& Stein, 1998). Besides home, schools represent another social context in which connectedness and perceptions of care

may offset risk (Dryfoos, 1990).

Risk behaviors such as alcohol, cigarette, and marijuana use threaten physical and mental health and may increase the

likelihood of engaging in other adverse behaviors. For example, besides the physiological detriments, cigarette use often

precedes marijuana and alcohol use and is concurrent with other risk behaviors such as fighting. Further, the earlier

an adolescent initiates alcohol and marijuana use, the more likely a pattern of indiscriminant drug abuse, especially more
serious health compromising substances such as cocaine or heroin, will occur (Davis, 1999; Inciardi, Horowitz, & Pottierger,

1993; U.S. Department of Health and Human Services, 1991).

Behavioral outcomes such as these are the result of relationships with significant others like parents and friends (Bandura,

1977; Hawkins & Weis, 1985). Tendencies regarding risk behaviors, for example, are learned actively; that is, family and

peers influence perceptions and behaviors toward or against specific outcomes. Adolescents interacting with significant

others holding positive attitudes about and engaging in substance use are at greater risk for perceiving and acting in a

similar way. Likewise, parental and peer disagreement with and discouragement from drinking, smoking cigarettes and/or

marijuana may provide an effective shield against initiating such behaviors.

As indicated earlier, poor social contexts typically exacerbate normal risks associated with childhood, adolescent, and

adulthood transitions, making youths from disadvantaged neighborhoods more likely to engage in these risk behaviors

(Brewster, Billy, & Grady, 1993; BrooksGunn et al., 1993; O'Connor, Tilly, & Bobo, 2001). Minority adolescents are a

population at risk for behavioral problems because of their generally low SES and their limited access to health services and

programs. Health clinics and after-school programs, which could offer critical support for unattended youths, are often

underfunded or nonexistent. Further, the actual use of health services, relative to level of need for such care, may be

hindered or restricted by ineffectual programs that have been designed with a particular subset of the population in mind,

usually middle-class white adolescents. The inability of African American youth and their families to identify programs as a

reliable means of reducing risk behaviors affects the course and pattern of treatment for minority youth. As such,

understanding the role of race is essential for successful creation and implementation of intervention strategies.

Understanding etiological beliefs is important not only as a way of understanding how youth construct surrounding contexts

but also because such beliefs may be inextricably linked to attitudes towards others, definition and framing of social

problems, endorsement of public programs, and own behaviors (Neighbors, 1990). A multi-domain perspective of a risk and

asset factors approach emphasizes the development of the adolescent within a dynamic social context that offers ever-

changing challenges and threats (Fitzpatrick & LaGory, 2000; Fitzpatrick, 1997). By testing the applicability of a risk and

asset model to this African American sample, this paper seeks to fill knowledge gaps in the psychosocial aspects of health

risks among minority adolescent populations. This should shed light on why some African American youth are able to avoid

negative behavioral outcomes while others adapt through harmful and destructive behavior such as substance use.

To this end, the principal focus of this study was to address the question: How are various risk and asset factors related

to adolescent substance use behavior in an African American sample? This was accomplished by: (1) identifying domains of

risks and assets (individual, family, peer, and school), and (2) measuring the strength of direct relationships between risks,

assets, and substance use outcomes. The analyses focused primarily on the cross-sectional relationships between

sociodemographic, risk factors, asset factors, and substance use behaviors.

METHODS
Sample

The sample consisted of 1,494 students enrolled in a central Alabama school system (grades 5-12) who completed an

annual Student Health Risk Assessment Survey during May 2002. Ninety-one questionnaires from non-African Americans

were excluded from the analysis.

The survey instrument was a 62-item self-administered questionnaire designed to collect responses for assessing various

related behavioral, attitudinal, and psychosocial dimensions. Questions included a battery of demographic, psychological,

and social correlates operating in multiple spheres of influence (e.g., individual, family, peer, and school). Student

participation was voluntary. A letter of consent was sent home by the school district explaining the purpose of the study. The

sample was 52% female and had a median age of 14 years, with 45% from high schools (grades 9-12), 22.1% from middle-

schools (grades 7-8), and 32.5% from elementary schools (grades 5-6). The response rate for this sample was

approximately 65%. No data were collected for students absent on the day the survey was administered.

Measures

Behavioral outcomes. The dependent substance use variables, alcohol, cigarettes, and marijuana, were measured by the

frequency students reported using these substances during the 30 days prior to the survey. These questions are similar to

those used in the Centers for Disease Control and Prevention's (1997) Youth Risk Behavioral Surveillance Survey. The

variables related to alcohol and marijuana use are six-category, ordinal variables where 0 = none; 1 = 1-2 times; 2 = 3-9

times; 3 = 10-19 times; 4 = 20-39 times; 5 = 40 or more times. The smoking variable is a five-category, ordinal variable

where 0 = none; 1 = less than 5 days; 2 = 5-15 days; 3 = 16-25 days; 4 = 25 days or more. While subject to debate, some

have argued that ordinal variables with five or more categories can be treated as interval variables (e.g., Demaris, 1992).

Nevertheless, our use of ordinal-level variables in this regression analysis calls for some caution when interpreting the

results.

Sociodemographic factors. The variables age and gender were selected as critical sociodemographic factors, with age

operationalized using school type/grade level: 1 = elementary (grades 5-6), 2 = middle (grades 7-8), and 3 = high (grades 9-

12). Gender is dichotomized with 0 = male.

Risk factors. Risk variables were identified at family- and peer-level spheres of influence. One family risk factor indicates

degree of physical assault (slap, punch, kick, or beaten) by parent or resident adult from 0 = never to 4 = all the time. This

five-category, ordinal variable was used to assess exposure to violence within the family experienced by the respondent.

There are two peer risk factors. Questions about whether the respondent is a gang member, has been asked to join a gang,

or has friends that are members (0 = no, 1 = yes) were summed into a single variable with higher scores indicating

increasing gang connectedness or negative peer influence. A measure of peer attitudes about the use of drugs and alcohol

is a five-category, ordinal variable with scores from 0 = strongly disagree to 5 = strongly agree.
Asset factors. Protective factors, or assets, were identified at individual-, family-, and school-level spheres of

influence. In investigating the association between individual asset factors and behavioral outcomes, an academic

achievement variable was used as a subjective assessment of grades mostly received in school ranging from 1 = D's and

F's to 7 = mostly A's. Self-esteem is a ten-item scale with scores ranging from 1 = strongly agree to 4 = strongly disagree (M

= 16.50, SD = 3.32). The questions focus on self-report information about satisfaction, pride, worth, respect, and attitude

toward self. An ordinal parental monitoring variable assesses if parents/guardians know where respondent goes with friends

ranging from 0 = never to 4 = all the time. Finally, a dichotomized school asset variable asked students if there is someone

at school who cares about them or what happens to them (0 = no, 1 = yes).

RESULTS

Regression Models: Assets and Risks by Behaviors

Alcohol use. Table 1 presents results of a regression of recent alcohol use on sociodemographic, family and school risk, and

asset variables. In Model 1, both age and gender were significant (p < .01), with older students and males reporting greater

alcohol use in the past 30 days. In Model 2, risk variables were introduced. These variables were all significant and

remained so for the complete model (p < .01). In Model 3, asset variables were included. Grades received in school (p < .05)

and parental monitoring (p < .01) were significant. Interestingly, while age remained significant for the entire model, gender

not only lost significance with the addition of the risk variables, but also changed direction with the inclusion of the asset

variables. The total variance accounted for by the complete model was 19%.

Cigarette use. Table 2 presents results from a regression of cigarette use on sociodemographic, risk, and asset variables.

Once again, age and gender were significant in the first model (p < .01). Older students and males again reported greater

cigarette consumption in the last 30 days. This time, however, age lost significance with the addition of risk variables and

gender with the asset variables. Among risk variables, affiliation with gangs and peer attitudes were significant (p < .01).

Similar to alcohol use, significant asset variables were grades received in school (p < .05) and parental monitoring (p < .01).

The total variance accounted for by the complete model was 16%.

Marijuana use. Table 3 presents the final set of results with marijuana use regressed on sociodemographic, risk, and asset

variables. As before, in Model 1, older students and males were more likely to report recent substance use. Age remained

significant for the complete model (p < .01) while gender lost significance with the addition of the asset variables. Both family

abuse (p < .05) and deviant peer (p < .01) risk variables were significant. Similar to the other substances under study,

grades received in school (p < .05) and parental monitoring (p < .01) were significant. Although not statistically significant,

greater self-esteem in this regression was associated with more marijuana use over the past 30 days rather than less. Since

self-esteem and other personality factors are known to be influential in motivating or inhibiting behavior (e.g., Shedler &

Block, 1990), these results do not necessarily imply that it is not important for the students in this sample. More likely, self-

esteem was overshadowed by more direct acting factors relating to peer and parental influences. The total variance

accounted for by the complete model was approximately 24%.

Enlarge this image.

Risk/Asset Regression Model of Alcohol Use in

Enlarge this image.

Risk/Asset Regression Model of Marijuana Use in

DISCUSSION

The principal objective of this paper was to explore the associations between hypothesized sociodemographic, risk, asset

factors and self-reported adolescent substance use behavior among an African American sample. This paper looked at both

the individual effects of risk and asset factors as well as the significance of these variables organized into functional

groupings (individual, family, peer, and school).

Age was significant in the alcohol and marijuana use models, losing significance with the inclusion of risk factors for

cigarette use. A trend was noted in the regression results suggesting that substance use behaviors became more frequent

as youth advanced from elementary to high school. This may be explained by more time spent away from home and

increased importance of peer influence with increasing grade level. Gender effects on self-report recent substance use were

more complex. Gender lost its ability to significantly predict substance use with the inclusion of risk and asset factors.

Specifically, risk factors intervened to reduce gender's impact for drinking while assets changed who reported more alcohol

use from male to female. It appears that, for both smoking and marijuana use, assets tempered gender's overall

significance. The complexity of the relationships between gender and assets warrants further study on the possible

interactive effects between these sets of variables. This also draws attention to the need to possibly implement gender-

specific preventive and intervention efforts to offset risks.

A predictive family-level risk variable was exposure to violence within the family, which may prompt adolescents to seek

relief from associated tension and stress through substance use, especially alcohol and marijuana (Resnick et al., 1997).

The contribution of unconventional peer attitudes and actions was also consistent across all behavioral risk variables in the

analyses. This suggests that knowledge regarding a student's association with peers tolerant of substance use and/or

actually involved in risk behaviors is useful in predicting recent drinking, smoking, and marijuana use for that individual.

These students may be modeling behaviors of friends or selecting friends with similar beliefs and habits. That is, peer

associations may be the antecedents of risk behaviors or they may serve to reinforce and perpetuate preexisting behaviors.

The cross-sectional nature of the study precludes delineating between these alternatives. Either way, previous investigators

have shown negative peer interactions to be powerful predictors of unhealthy choices and actions (Ennett & Bauman, 1993).

The results support the idea that substance experimentation, especially marijuana use, is often normative inadolescent peer

groups. Those students choosing to abstain are found to be psychologically and socially withdrawn (Shedler & Block, 1990).

So while the effect of self-esteem in tempering risk behaviors was not significant in the overall regression model, it is known

to be an influential motivating or inhibiting factor for behavioral outcomes. This may also hold true for the lack of significance
for perceptions of care at school. Having a teacher, counselor, coach, or support personnel that is perceived as caring and

reliable enables students to connect more readily to their schools. More likely, these assets were overshadowed by more

direct acting factors relating to peer and parental influences.

Parental monitoring and grades received in school were the only asset variables to significantly predict substance use

behaviors. The results from this sample of African American students support recent findings that low levels of perceived

monitoring in an urban, low-income African American sample were related to increased substance use (Xiaoming et al.,

2000). The prediction that higher academic achievement is associated with decreased substance use behaviors was

supported by the results. Students reporting higher grades received in school tended to have lower self-reported incidence

of recent alcohol, cigarette, and marijuana use. This finding supports previous research indicating that those students

reporting lower grades also indicate greater use of alcohol, cigarettes, and marijuana (Dekovic, 1999).

Limitations of Study

Data were obtained using a self-report methodology, which can be subject to bias. However, there is growing evidence that

such measures of adolescent risk behaviors are generally reliable and valid (Brener et al., 1995). Although information was

not available regarding non-participants, it included those who were not able to obtain parental permission as well as those

suspended, absent, or no longer students at the schools. Even though there was no a priori justification to assume that there

was a differential distribution of behaviors or other factors between those who did and did not obtain parental permission, it

is reasonable to assume that those in the latter categories were likely to have had increased rates of health-risk behaviors

making the conclusions a conservative estimate for the entire school population.

As with any cross-sectional study design, causality between independent variables and outcomes cannot be inferred from

the reported associations. Temporal ordering of variables necessitates longitudinal analysis. Such studies may be useful for

examining the long-term effects of individual and social variables on behavioral outcomes and facilitates exploration of how

risk and asset variables differentially relate to risk behaviors over time. Also, since regression analyses were performed on

several ordinal-level variables, caution should be exercised in the interpretation of these findings.

Because of the circumscribed demographic characteristics of the sample of students used in this study, generalizability may

be limited. However, the intent of the study was to evaluate a risk and asset theoretical model for use with low-income

African American students. This subgroup of adolescents may be at particular risk for negative behavioral outcomes. As

such, programs and policies will require research focusing on this target population to increase success with regard to

participation.

Implications for Practice

Adolescent risk and asset models should be specific to the cultural/ ethnic group. The notion of differential outcomes of risk

and asset variables related to race may account for why some variables were nonsignificant or weak predictors of risk

behaviors while other factors were important contributors to the explanatory power of the model. For example, self-esteem
and the perception that someone at school cares about the student were not significant predictors of substance use for this

sample.

Evaluating health and risk outcomes and their potential correlates within various domains of influence gives us an idea of

where the emphasis should be placed for possibly reducing adolescent substance use behaviors for minority youth. The

consequences for high self-reported incidence of alcohol, cigarette, and marijuana use are multi-domain as well. For the

individual there is the potential for injury, alcoholism, progression to hard drugs as well as lung cancer, emphysema, and

loss of potential for success. Within the family, the result is often conflict over the behaviors. For the community, there is

reduced neighborhood capabilities, increased police and treatment costs, and diminished effectiveness of schools. As such,

the foci for intervention programs to reduce personal, community, and social consequences and long-term costs of health-

compromising behaviors should extend beyond the individual (student) to include peers, family, school, and neighborhood.

The results of this analysis suggest that the emphasis for intervention programs should be on modeling attitudes,

preventing gang participation and family violence, encouraging parental monitoring and guiding of their children's behaviors,

facilitating positive teacher-student interactions, and building positive school environments. Suggested further research that

would expand on the findings in this analysis include prospective longitudinal studies to better assess temporal trends and

causality and comparing across school systems having different proximity of community resources and risk areas to homes

and schools.

Aside from the relevance of these results for development of substance abuse prevention strategies within school systems,

they may also be applicable in clinical circumstances for efficient screening of risk and protective factors when considering

the likelihood of alcohol, cigarette, and marijuana use among minority adolescents. The data suggest that clinicians should

investigate for indications of child abuse, exposure to deviant peer influences, and grade level as risk factors and extent of

parental monitoring and scholastic performance as protective factors for adolescent substance misuse. Simultaneously,

these factors may provide a useful qualitative clinical indicator of an adolescent's resiliency to such health and safety risks.

Because of the cross-sectional nature of the data, this study was not suited for evaluating the applicability of the self-

medication hypothesis among this sample of adolescents (e.g., Hall & Degenhardt, 2000; Markou, Kosten, & Koob, 1998;

Swendsen et al., 2000). However, such constructs admittedly provide an additional useful context for interpretation of results

of substance abuse risk factor assessments which, in clinical rather than research settings, necessarily must consider a

wider scope of findings and associated temporality issues.

References

REFERENCES

Anderson, E. (1990). Street wise: Race, class, and change in an urban community. Chicago: The University of Chicago

Press.
Bailey, S. L., & Hubbard, R. L. (1990). Developmental variation in the context of marijuana initiation among adolescents.

Journal of Health & Social Behavior, 31, 58-70.

Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ: Prentice-Hall.

Brener, N. D., Collins, J. L., Kann, L., Warren, C. W., & Williams, B. I. (1995). Reliability of the Youth Risk Behavior Survey

Questionnaire. American Journal of Epidemiology, 141, 575-580.

Brewster, K., Billy, J., & Grady, W. (1993). Social context and adolescent behavior: The impact of community on the

transition to sexual activity. Social Forces, 71, 713-740.

Brooks-Gunn, J., Duncan, G. J., Kato, P., & Sealand, N. (1993). Do neighborhoods affect child and adolescentdevelopment?

American Journal of Sociology, 99, 353-395.

Brown, B. B., Mounts, N., Lamborn, S. D., & Steinberg, L. (1993). Parenting practices and peer group

affiliation in adolescence. Child Development, 64, 467-482.

Carvajal, S. C., Wiatrek, D. E., Evans, R. L., Knee, R. C., & Nash, S. G. (2000). Psychosocial determinants of the onset and

escalation of smoking: Cross-sectional and prospective findings in multiethnic school samples. Journal of Adolescent Health,

27, 255-265.

Centers for Disease Control and Prevention. (1997). 1995 Youth Risk Behavior Surveillance Survey. Atlanta: Centers for

Disease Control and Prevention.

Chaffin, M., Kelleher, K., & Hollenberg, J. (1996). Onset of physical abuse and neglect: Psychiatric, substance abuse, and

social risk factors from prospective community data. Child Abuse and Neglect, 20, 191-203.

Collins, W. A., & Repinski, J. (1994). Relationships during adolescence: Continuity and change in interpersonal

perspective. In R. Montemayor, G. R. Adams, & T. P. Gullotta (Eds.), Personal relationships during adolescence (pp. 7-36).

Thousand Oaks, CA: Sage.

Crouter, A. C., MacDermid, S. M., McHale, S. M., & Perry-Jenkins, M. (1990). Parental monitoring and perceptions of

children's school performance and conduct in dual- and single-earner families. Developmental Psychology, 26, 649-657.

Davis, N. J. (1999). Youth crisis: Growing up in the high-risk society. Westport: Praeger.

Dekovic, M. (1999). Risk and protective factors in the development of problem behavior during adolescence. Journal of

Youth and Adolescence, 28, 667-685.

Demaris, A. (1992). Logit modelling: Practical applications. Quantitative applications in the social sciences, No. 86. New

York: Sage.
Dryfoos, J. (1990). Adolescents at risk: Prevalence and prevention. New York: Oxford University Press.

England, E. M., & Petro, K. D. (1998). Middle school students' perceptions of peer groups: Relative judgments about group

characteristics. Journal of Early Adolescence, 18, 349-373.

Ennett, S. T., & Bauman, K. E. (1993). Peer group structure and adolescent cigarette smoking: A social network analysis.

Journal of Health & Social Behavior, 34, 226-236.

Farrell, A. D., & White, K. S. (1998). Peer influences and drug use among urban adolescents: Family structure and parent-

adolescent relationships as protective factors. Journal of Consulting and Clinical Psychology, 66, 248-258.

Fitzpatrick, K. M. (1997). Fighting among America's youth: A risk and protective factors approach. Journal of Health & Social

Behavior, 38, 131-148.

Fitzpatrick, K. M., & LaGory, M. (2000). Unhealthy places: The ecology of risk in the urban landscape. New York: Routledge.

Garnier, H. E., & Stein, J. A. (1998). Values and the family: Risk and protective factors for adolescent problem behaviors.

Youth and Society, 30, 89-120.

Hall, W., & Degenhardt, L. (2000). Cannabis use and psychosis: A review of clinical and epidemiological evidence.

Australian New Zealand Journal of Psychiatry, 34, 26-34.

Hawkins, D., & Weis, J. (1985). The social development model: An integrated approach to delinquency prevention. Journal

of Primary Prevention, 6, 73-97.

Inciardi, J. A., Horowitz, R., & Pottierger, A. E. (1993). Street kids, street drugs, street crimes. Belmont, CA: Health.

Jargowsky, P. A. (1997). Ghettos, barrios, and the American city. New York: Russell Sage Foundation.

Jessor, R. (Ed.). (1998). New perspectives on adolescent risk behavior. New York: Cambridge University Press.

Johnson, R. A., & Hoffman, J. P. (2000). Adolescent cigarette smoking in U.S. racial/ethnic subgroups: Findings from the

National Education Longitudinal Study. Journal of Health & Social Behavior, 41, 392-407.

Markou, A., Kosten, T. R., & Koob, G. F. (1998). Neurobiological similarities in depression and drug dependence: A self-

medication hypothesis. Neuropsychopharmacology, 18, 135-174.

Massey, D. S., & Denton, N. A. (1993). American apartheid: Segregation and the making of the underclass. Cambridge, MA:

Harvard University Press.

Myers, H. F., & Taylor, S. (1998). Family contributions to risk and resilience in African American children. Journal of

Comparative Family Studies, 29, 215-229.

Neighbors, H. (1990). The prevention of psychopathology in African Americans: An epidemiological perspective. Community

Mental Health Journal, 26, 167-179.

O'Connor, A., Tilly, C., & Bobo, L. D. (Eds.). (2001). Urban inequality: Evidence from four cities. New York: Russell Sage

Foundation.

Prino, C. T., & Peyrot, M. (1994). The effect of child physical abuse and neglect on aggressive, withdrawn, and prosocial

behavior. Child Abuse and Neglect, 18, 871-884.

Resnick, M. D., Bearman, P. S., Blum, R. W., Bauman, K. E., Harris, K. M., Jones, J., Tabor, J., Beuhring, T., Sieving, R. E.,

Shaw, M., Ireland, M., Beahnger, L. H., & Udry, J. R. (1997). Protecting adolescents from harm: Findings from the National

Longitudinal Study of Adolescent Health. Journal of the American Medical Association, 278, 823-832.

Shedler, J., & Block, J. (1990). Adolescent drug use and psychological health: A longitudinal inquiry. American Psychologist,

45, 612-630.

Simons, R. L., Wu, C., Conger, R. D., & Lorenz, F. O. (1994). Two routes to delinquency: Differences between early and late

starters in the impact of parenting and deviant peers. Criminology, 32, 247-276.

Swendsen, J. D., Tennen, H., Carney, M. A., Affleck, G., Willard, A., & Hromi, A. (2000). Mood and alcohol consumption: An

experience sampling test of the self-medication hypothesis. Journal of Abnormal Psychology, 109, 198-204.

U.S. Department of Health and Human Services. (1991). Healthy People 2000: National health promotion and disease

prevention objectives. NIH Publication Number (PHS) 91-50212. Washington, DC: U.S. Government Printing Office.

Williams, J. H., Stiffman, A. R., & O'Neal, J. L. (1998). Violence among urban African American youths: An analysis of

environmental and behavioral risk factors. Social Work Research, 22, 3-13.

Wilson, W. J. (1987). The truly disadvantaged: The inner city, the underclass and public policy. Chicago: University of

Chicago Press.

Windle, M. (2000). Parental, sibling, and peer influence on adolescent substance use and alcohol problems. Applied

Developmental Science, 4, 98-110.

Xiaoming, L., Feigelman, S., & Stanton, B. (2000). Perceived parental monitoring and health risk behaviors among urban

low-income African American children and adolescents. Journal of Adolescent Health, 27, 43-48.

AuthorAffiliation

The authors would like to thank Mark LaGory and Michael Windle for their insightful comments. In addition, we would like to

thank the Bessemer City School system and Mary Fitzpatrick, Safe and Drug-Free Schools Coordinator, for making

available the data used in this study.

 Please direct all correspondence and requests for reprints to Darlene R. Wright, Department of Sociology, Samford

University, Birmingham, AL 35229; Phone: (205) 726-4029; E-mail: drwright@samford.edu or kfitzpat@uab.edu.

Copyright Libra Publishers Incorporated Winter 2004

Wright, D. R., & Fitzpatrick, K. M. (2004). PSYCHOSOCIAL CORRELATES OF SUBSTANCE USE BEHAVIORS AMONG AFRICAN

AMERICAN YOUTH. Adolescence, 39(156), 653-67. Retrieved from

http://search.proquest.com/docview/195935875?accountid=37714

A systemic analysis of the dynamics and organization of urban

street gangs
Ruble, Nikki M; Turner, William LAuthor Information . The American Journal of Family Therapy 28.2 (Apr-Jun 2000):
117-132.

1. Full text
2. Full text - PDF
3. Abstract/Details
Hide highlighting
Abstract
TranslateAbstract

Street gangs function as ongoing, open social systems in relation to their surrounding sociocultural context. To better

understand street gangs and provide effective intervention plans, they must be viewed from this systematic and holistic

perspective.

Full Text
TranslateFull text
Turn on search term navigation

Headnote

Street gangs function as ongoing, open social systems in relation to their surrounding sociocultural context.

Fundamentally, gangs are comparable to some family systems. In fact, most gangs do consider themselves to be `families. "

The same concepts used to describe family systems may also be applied to street gangs. Street gangs usually exhibit a

highly complex organization, structure, process, and functionality. All of these aspects intertwine within and around

the gang to form a web of interconnectedness and continuity. To better understand street gangs and provide effective

intervention plans, they must be viewed from this systematic and holistic perspective.

Street gangs have been documented in cities in the U.S. throughout most of the country's history (Spergel, 1990). In some

cities, gangs have been credited with an alarming share of violent crime, most notedly homicide. Victims of the serious and

often random violence that has come to be known as a common feature of street gangs have retreated within their

communities and neighborhoods in fear-afraid to let their children walk to school, go to the corner store, or even play outside

of their own homes (Conly, Kelly, Mahanna, & Warner, 1993). For years, social scientists, police officials, and popular media

have all struggled to understand the essence of street gangs (Sanders, 1994). Perhaps street gangs can be best understood
through the lens of a systemic approach by paying careful attention to their systemic dynamics, functions, and organizational

structures.

Street gangs can be defined as groups of youths and young adults with varying degrees of cohesion and structure, who

have regular contact with one another, ways of identifying their group, and rules of behavior within the system (Conly et al.,

1993). Gangs serve numerous functions for their members, including providing a source of status, identity, cohesion, self

esteem, and a sense of belonging (Harris, 1994). Furthermore, gangmembers experience an intense emotional closeness

born of children growing up together on the street and turn to each other for support and often for the mere function of

survival (Lyon, Henggeler, & Hall, 1992).

There are at least three distinct types of gangs. The first is referred to as a social gang. This gang is a relatvely permanent

group that "hangs out" at a specific location (i.e., store, street corner, abandoned building). Social gangs are not likely to

participate in serious delinquent activitiy and will engage in physical violence only if members are attacked and retaliation is

unavoidable. The group stays together because of mutual attraction among its members, rather than through a need for

protection or esteem. These gang members tend to have the closest association with the norms and values of

society in general (Fagan, 1989). A second type of gang is the delinquent gang. This gang type is structurally cohesve and

is often organized around the pursuit of monetary gain derived from illegal activity. Gang survival is dependent upon each

member's precise execution of his/her assignment and the provision of back-up support when necessary. The

accomplishment of these group enterprises may be viewed by gang members as their method of getting ahead in the world

(Fagan, 1989). A third gang type is the violent gang. The primary purpose of this gang is to obtain the power and associated

emotional gratification that violent activities can bring to youths (Hardy, 1996). Both leaders and followers tend to

overestimate the importance, size, and power of their group. These gangs tend to have a highly structured hierarchy of

leaders and followers. The violent gang is also characterized by intragroup violence. Gang members may be verbally violent

with one another, but often hostility and aggression will take on a physical form (Lyon et al., 1992).

Although in some gangs a particular type may be easily distinguished, there is often a blending of gang types within any

given street gang. Moreover, researchers and others who study gangs tend to be more concerned with

categorizing gang types, than are the gang members themselves.

No one has developed a satisfactory count of the number of gang members or gangs nationwide. Attempts to do so have

been hampered by the variation of the way gangs have been defined from one site to another (Conly et al., 1993). However,

it has been reported that Los Angeles County has documented the most gangmembers and gangs throughout the U.S.

(Harris, 1994). Harris (1994) found that there are at least 600 gangswith an estimated 100, 000 gang members

operating in Los Angeles County. Homicides committed by gangmembers currently account for approximately 35% of the

homicides in Los Angeles County annually <Rogers, 1993). Throughout the country, some of the most well-

known gangs include the "Grips," the "Bloods," the "Vicelords," and the "Latin Kings." The majority of these

particular gangs are located in large cities, such as Los Angeles, Chicago, New York, Miami, and Milwaukee (Clark, 1992;

Hagedorn & Macon, 1988; Sanders, 1994).

DEMOGRAPHICS

Age

The ages of gang members range widely. However, most gangs studied report members falling between 10 and 30 years of

age, with the majority being between 14 and 24 years old (Huff, 1989; Winfree, Backstrom, & Mays, 1994).

Some gang members have been found to be as young as 8 years old. These younger members are often seen as

expendable and are used as drug couriers (Borringer, 1995).

Gender

Gang behavior, especially gang-related crime, tends to be largely a male phenomenon (Winfree et al., 1994). Although

street gangs are predominately male, recent research reveals that the number of all female gangs is rising rapidly (Clark,

1992; Harris, 1994). There may be a number of females associated with a particular gang, but they usually serve as

auxiliaries or branches of male gangs. They are often expected to support the male gang completely, but they are rarely

able to become official members (Moore, 1991). Females associated with male gangs are typically used to carry weapons,

provide alibis, act as spies and lures, and provide sex for male members (Winfree et al., 1994).

Ethnic and Racial Composition

The ethnic and racial composition of gangs seem to be overrepresented by minority group members. Most gangs tend to be

racially exclusive (Clark, 1992). Gangs are often divided into four main racial groups: African-American origin, Asiatic origin

(Chinese, Japanese, Korean, Taiwanese, Samoan, and American Indian), European origin (English, Italian, Irish, Slavic,

Russian, and German), and Hispanic origin (Mexican, Puerto Rican, Panamanian, and others from Spanish-speaking

countries) (Miller, 1975). Overall, African-American and Hispanic gangs are the most

dominant gangs represented in the gang population (Conly et al., 1993).

Location

Contrary to popular belief, the location of gang activity varies widely. There are three main areas that gangsusually exist.

The primary location for gang activity is in the inner cities, especially areas in decay. Gangstend to form in the shifting,

changing, or transitional neighborhoods of the larger cities, such as the "projects." These areas tend to be characterized by

social disorganization and rapid population shifts (Conly et al., 1993). Another area conducive to the formation of gangs are

stable slums. Stable slums, for example, South Central Los Angeles, are characterized by areas where population shifts are

slow, permitting patterns of behavior and traditions to develop over a number of years (Siegel & Senna, 1985). A third

place gangs originate is suburban and rural areas (Conly et al., 1993). Slums and ghettos have shifted from the inner city or

ring city, to suburban areas-that is, to formerly middle-class areas that are now in decay. In the past 10 years,

many gangs have claimed this turf as their own and have begun to take over some suburban housing projects. The fact that
there is less intensive police patrol in these areas has contributed to the shift of gangs out of the cities and slums to the

outer edges of certain districts (Siegel & Senna, 1985).

GANGS AS SYSTEMS

From a systems perspective, street gangs can be viewed as ongoing, open, social systems. As is true of street gangs, social

systems function as a whole in relation to their sociocultural context (Broderick, 1990). Each member of the gang affects and

is affected by other members and by the context of which they operate. Inline with systems' thinking, street gangs are

fundamentally comparative to family systems. The same concepts that are used to describe family systems, such as

hierarchies, subsystems and suprasystems, entropy and negentropy, boundaries, communication, and homeostasis

(Broderick, 1990; Minuchin, 1974; Walsh, 1982), can also be applied to the street gang. Because most gangs do consider

themselves to be a kind of family, it is both critical and prudent that they be viewed and studied from the family systems

perspective (Vigil, 1988.) For many gang members, their gang family is the only family that they have. The

street gang functions as a "surrogate family," providing its members with affection, understanding, recognition, loyalty, and

emotional and physical protection (Morales, 1992). This sense of companionship and belonging fills a void that has been left

empty by the gang member's previous family relationships (Hardy, 1996). For some gang members, they join because they

do not have another family; for others, they join to escape a dysfunctional situation in their families of origin (Vigil, 1988).

Once a youth has gained membership with a gang, he/ she is fully accepted as a family member. Many gang members

report that they are willing to die or kill for their gang, expressing their ultimate love and loyalty to their gang family (Belitz &

Valdez, 1994).

Hierarchies

From a systems perspective, gangs, like families, have an organized structure within which their members assume roles and

carry out certain responsibilities. Both are organized based on hierarchies (Haley, 1976). Infamilies, parents are in the

executive position of leadership in the family, and children assume the subordinate position. In the leadership role, parents

have the responsibility of making decisions that will affect the safety and security of all family members (Minuchin, 1974).

Most street gangs are organized in much the same way. Joe (1994, p. 396) states, "Gangs are interconnected and

hierarchically organized according to age and experience." Most researchers and practitioners agree that gangs usually

consist of a set of leaders, peripheral members, and recruits (Conly et al., 1993). The hierarchy and structure of most

street gangs can be described by beginning with the hardcore gang member or the "OG," which stands for "original

gangster." This type of gang member is heavily involved in the gang, making it central to his or her life. The "OG" is often a

violent criminal and he or she is very committed to gang activities. The original gangster has usually been part of the

street gang for a long period of time. Because of his/her experience, the "OG" is often looked up to by other, less

experienced gang members. Novice gang members may look to the "OG" for guidance, support, and protection as they

navigate their way through the first stages of their gang membership. The "associate" gangmember usually knows

people in the gang, but does not participate in all gang activities. He or she participates in some gang activities, but is less
likely to get involved with the more serious, sometimes deadly or illegal activities of the gang. The "wannabe" is often

infatuated with gang behavior. Wannabes are usually associated, but not committed, to the gang. This type of gang affiliate

is often in middle school or just slightly older (Jackson, as cited in Conly et al., 1993).

Street gangs may have one leader or multiple leaders and multiple cliques, each with a slightly different interest and

responsibility in the gang (Conly et al., 1993). Leadership in the gang has been described as (a) multiple, where there is

more than a single leader at any one time; (b) informal, where the choice of a leader is not structured; (c) situational, where

the leadership role may only be in certain situations; and (d) functional, where situated leadership is based on a particular

function (Sanders, 1994). Whatever the leadership structure may be, it is certain that proven leaders have specific, known

qualities. One member of the "Latin Kings" stated, "Leaders must be daring, dangerous, and not afraid" (Hagedorn & Macon,

1988, p. 93).

Subsystems

Any given system consists of smaller systems called subsystems. Family subsystems have their own interdependence and

mutual influence among their members (Nichols & Everett, 1986). There are three main subsystems in the average family:

the spousal or couple, the parental, and the sibling subsystem (Minuchin, 1974). The spousal/couple subsystem refers to the

bond between partners. The parental subsystem (often called the executive subsystem) refers to the responsibilities and

duties involved in parenting. The sibling subsystem refers to the relationship and interactions among siblings (Minuchin,

1974 ). Subsystems are most visible in the street gang in the form of cliques. Most cliques seem to be age-graded, although

some cliques have mixed-age membership (Hagedorn & Macon, 1988). Cliques may form within the hierarchical structure of

the gang-producing subsystems among leaders, associates, and wannabes. A parental clique may form among the leaders

of the group. The parental clique may act as the "parents" of the gang. The leaders of the gang act as a parental subsystem

by "watching over" gang members and activities, and providing guidance, direction, and protection for the new and/or less

experienced gang members. Spousal or couple subsystems may be seen in gangs (especially male gangs) within the

relationships that form between gang members and the female auxiliaries that "hang" with the gang. Often these females

are expected to support their male companions as "spouses," but are rarely allowed to become official members of

the gang. Sibling subsystems often form among the cliques that consist mainly of associate and wannabe members. These

less experienced members often join together as "siblings" to provide one another the support needed as they begin to gain

experience in the gang. One study found that in Los Angeles, gangs that contain about 100 to 125 members produce cliques

with an average of 30 to 40 members each. These cliques have fluid membership and some sort of connection to the other

cliques in the gang (Hagedorn & Macon, 1988).

Suprasystems

All systems are embedded within larger systems called suprasystems (Nichols & Everett, 1986). Families are surrounded by

suprasystems in the form of friends, extended family, neighborhoods, communities, schools, the workplace, and an overall

social and ethnic environmental element (Broderick, 1990). The suprasystems that are most likely to affect and be affected
by the street gang are the community or neighborhood of which the gang is a part and the law enforcement agencies that

patrol the location where the gang frequently operates. Researchers who work with Los Angeles Chicano gang members

state, "The word for gang and for neighborhood is identical. `Mi barrio' refers equally to `my gang' and `my neighborhood"'

(Hagedorn & Macon, 1988, p. 134).

The social balance between a gang and its community seems to be a delicate one. In some communities, gangs are

afforded a certain amount of community tolerance. This community tolerance generally exists for three reasons:

1. gang members may be family members and neighbors of community residents;

2. community residents identify with the economic and social challenges that gang youths face;

3. a gang may help to establish some degree of order in its community, for example, by protecting local businesses from

attacks from rival gangs (Jankowski, 1991).

Law enforcement agencies (such as police officers) interface with gangs on a regular basis. Sanders (1994, p. 178) states

that police officers "stand as society's institutional interface in the social construction of gangactivities." However, interviews

reveal that gang members often have an intense dislike for police officers who use unnecessary "strong-arm" tactics when

arresting or questioning them. Gang members may feel vengeful when police officers behave unprofessionally, and it is

likely that they will seize the first opportunity to "get even." When police officers demonstrate personal concern

for gang members (i.e., asking how they're doing when they see them on the streets), in some cases, a mutual respect will

develop between the officers and some gang members. Veteran police officers and gang members may know each other on

a first-name or nickname basis, and may demonstrate a great deal of respect for one another (Huff, 1989).

Entropy

A system that is either too open or too closed will probably be dysfunctional. At either extreme, the system can be described

as being in a state of entropy. Entropic systems are often moving toward maximum disorder and disintegration. In these

disorganized systems, the members may use their energy thoughtlessly or in a random manner. When energy is

used in such a way, the system is in chaos and out of balance. By allowing in either too much information or not enough

information, the identity and survival of the system may be threatened. In families, for example, tasks may be pursued in a

conflictual or haphazard manner causing nothing to ever get done. When this happens, a sense of coherence or order

seems to be lacking. The absence of organization may be undermining the ability of the family members to successfully

complete their tasks. The movement of the system at this point is toward entropy (Becvar & Becvar, 1982).

Gangs depend on organization for the survival of the system. Some gangs are highly organized, while others have very little

organization (Sanders, 1994). A gang that has little organization may naively use all of its energy and then begin to decay.

Entropy is likely to occur in newly formed gangs, who do not yet understand the structure and functions that keep

the gang system alive. Gangs that operate in a disorganized state may allow too much or not enough information to enter
the system. For example, if the gang does not have clear external boundaries and allows unknown outsiders to enter

the gang, they may open themselves up to a vulnerable state. In this state, some of the core rules and boundaries that hold

the gang together may be threatened and the system may begin to disintegrate.

A gang that is moving toward entropy will soon be dissembled unless new information is introduced into the homeostatic

system. A gang that is in a state of decay requires negentropy to build it back up. Negentropy is a movement away from

disorder. Negentropy can be described as the introduction of new energy into the static system. The system is allowing

information to come in and is permitting change that is appropriate, while it is screening out information and avoiding

changes that would threaten the survival of the system (Becvar & Becvar, 1982). New energy may enter the gang system

when new members are initiated into the gang or when the gang experiences a change in leadership. Gang violence also

seems to add an enormous amount of energy into the gang. For example, drive-by shootings and gangbangs are a

continuing source of power for the gang (Sanders, 1994). When gang members participate in violent activities,

the gang system is strengthened by the bond that forms among the members. The more members collaborate to pull off

violent activities for the "good" of the gang, the stronger this bond becomes. This new energy, whether it be positive or

negative behavior, may allow the gang to reconstruct its system and resist the breakdown enforced by entropy.

Boundaries

The systems concept of boundaries is very important to the structure and function of the street gang. Infamilies, boundaries

are characterized by the physical and emotional barriers that distinguish individuals and families and regulate the amount of

contact occurring among them (Minuchin, 1974). Boundaries delineate where one system stops and another system begins

(Nichols & Everett, 1986). Boundaries regulate membership in the gang (i.e., who is in and who is out), and

information in the gang (i.e., what is or is not done or talked about within the gang or outside of the gang). Street gangs tend

to have rigid boundaries, which denotes that boundaries are relatively impermeable. Gangs typically do not approve of

active members having friends, especially close friends, outside of the gang (Winfree et al., 1994). Members tend to

establish group boundaries and consider those outside the group as "pousers." Those who have not been formerly initiated

into the gang are considered outsiders and are not able to cross the boundaries that surround the gang. Gang members are

very careful not to discuss gang activities and plans with outsiders for fear that they may be enemies and/or spies for a

rival gang, and thus a threat to the gang system (Kennedy & Baron, 1993).

Street gangs base much of their structure on the fact that their boundaries are impermeable. Gangs take pride in having

selective membership practices and codes of secrecy. Members are selected into a gang only after they demonstrate the

necessary skills required, and the ability to be loyal and committed to the gang. Most gangs have highly rigorous initiation

processes which involve perspective members demonstrating their ability to understand the unwritten rules of the gang. The

initiation process usually entails a physical beating by several other gang members. The beating must be endured without

complaint and without fighting back. The slightest whimper or other expressed signs of pain could result in rejection of

membership. The initiation process can be considered a prerequisite to weed out the weak and uncommitted (Vigil, 1988).
Within the gang, unwritten rules inform members of appropriate norms and behaviors expected among the gang and of the

traditional attitudes felt about other gangs in the area. Unwritten rules center on covert behaviors and beliefs such as:

(a) appropriate gang colors and attire;

(b) roles and duties delineated to members;

(c) ownership of turf and how that turf is to be protected;

(d) which gangs are rival gangs and why those gangs have become enemies.

The gang member who fails to live up to the gang's requirements (i.e., the member who does not defend his honor) incurs

immediate loss of status, removal from the gang, physical harm, or even the risk of being killed (Kennedy & Baron, 1993).

Communication

All behavior is regarded as communication, transmitting interpersonal messages (Walsh, 1982). Within a family system,

communication occurs continuously through both verbal and nonverbal interactions (Minuchin, 1974).

Street gangs communicate both between and across systems with the use of their own unique language, which is mostly

nonverbal and symbolic. There is a distinct emphasis on symbols and colors (Vigil, 1988). For example, in Los Angeles, the

street gang known as the Crips wear blue, usually in the form of bandannas, as their identifying color. Conversely,

the gang known as the Bloods, use the color red to signify their gangaffiliation (Clark, 1992). In large gangs, hand signals

are used to identify themselves to one another and to other gang members. For instance, holding up three fingers on the

right hand signifies membership with the Latin Kings. The three fingers represent the tips of a king's crown (Hagedorn &

Macon, 1988). Tagging (more commonly known as graffiti) is one of the most widely used types

of gang communication. Gang names, the street names of their members, and various self laudatory and rival-deprecatory

comments make up typical tagging messages. Some messages are simple and crude, while others are elaborate and

artistic. The terms "gangster," "gang," "gangbanger," and "mob" are used frequently in gang graffiti, revealing something

about the tagger's self-identity. Crip and Blood graffiti often include the characters C/K or B/K, which stand for Crip-killer or

Blood-killer. When a gang wants to challenge a rival gang, they will paint an X over their rival's tagging message, and place

their own name on the wall (Sanders, 1994). The number "187," which is the Los Angeles Police Department's code for

murder, is also used extensively by gang taggers. By placing the number "187" next to another gangster's name on a

building or wall, the message is clearly a foreshadowing for a gang hit. Some graffiti messages are quite complex. For

example, the message, "Cyco 33 OG 8-ball" has a variety of messages within it. "Cyco" is the nickname of the tagger. The

number "33" refers to the tagger being a member of a 33rd Street gang. "OG" stands for original gangster. To use the letters

"OG," the gangster is probably someone who has killed for the gang. Last, "8-ball" is a code meaning someone who deals

cocaine (Borringer, 1995). All of the types of communication used by gangs help to maintain the boundaries within and

around the gang. The communication style of colors and symbols are especially important in maintaining the rigid

membership boundaries that surround the gang.

Within a gang system, members may communicate with one another by using similar symbolic messages. For example,

hand signals are often used to send brief messages to one another, especially in a life-threatening situation when verbal

communication is too timely and dangerous. Gang members may also use a number of code words and slang language to

communicate with one another. This type of language helps to maintain the boundaries around the system, and protect

the gang's secrets from pousers in gangs that have a large membership (Borringer, 1995) .

Homeostasis/ Morphogenesis

The concept of homeostasis assumes that systems attempt to maintain stability in the midst of frequent internal and external

changes to the system. Homeostatic systems are able to accept change, but only within certain limits (Minuchin, 1974).

More appropriately, gangs like family systems, are morphogenic in nature, meaning that they tend to change in response to

their environment in order to maintain continuity within their system. In families, members may contribute to this homeostatic

balance through a mutually reinforcing feedback loop. Family members seek to regulate tension and to restore the family

equilibrium or homeostasis (Walsh, 1982).

Street gangs may be either homeostatic or morphogenic in their structure and activities. Homeostatic gangsmay adapt to

change, but they are not hypersensitive to environmental stimuli. These gangs tend to be more rigid in their structure, and

thus they are more resistant to change. Homeostatic gangs are not likely to endure over time. They tend to come onto

the gang scene with lots of strength, but eventually they surrender to the impact of entropy. Gangs that are in an extensive

homeostatic state are often too rigid to allow negentropy to enter their system, and as a result, they begin to decay and

ultimately fizzle out.

Some gangs tend to be morphogenic instead of homeostatic in their systemic nature. The morphogenic gangis more

receptive to change. These gangs are very sensitive to environmental stimuli, and they may readily allow new information to

enter the system to resist the breakdown enforced by entropy. The morphogenic gang tends to gain strength over time.

Examples of morphogenic gangs include the Crips and the Bloods of Los Angeles. These gangs have withstood the test of

time by having both flexible and rigid boundaries when appropriate. Morphogenic gangs learn from their experiences, adjust

to new information, and may eventually reach a morphogenic state characterized by ultra stability.

CLINICAL IMPLICATIONS

Utility of Systems Concepts

Applying systemic concepts to street gangs allows for an acknowledgment that gangs, as all systems, have multiple and

complex levels. All levels of street gangs must be reviewed before prevention and intervention strategies will be effective.

Viewing gangs from this holistic and systemic perspective provides a number of benefits for both researchers and

practitioners:
1. Family systems thinking explains not only why, but also how gangs function as a whole. This allows for some of the

negative and blameful language to be shifted away from gang life, and for more empathetic ideas to be introduced into the

study of street gangs. For example, if clinicians and researchers understood the systemic functioning of street gangs,

interventions could be organized to address all aspects and dimensions of the life of a gang member.

2. Understanding gangs as systems helps to explain the function that they serve for their members based on a holistic

perspective. The functionality of gangs may be viewed in terms of a breakdown in the home, neighborhood, and/or

community. Therefore, programs and interventions aimed toward gang members should encompass a multilevel approach,

inclusive of individual, family, community, and cultural influences.

3. Systems thinking enables outsiders to understand the structure and hierarchy of gang organization. Understanding power

balances and cliques within the gang can be helpful when implementing intervention plans. For example,

because gang leaders and original gangsters have such a high level of influence over other gang members, an intervention

plan that targets the upper level of the hierarchy may have a ripple effect for the rest of the gang.

4. Knowledge of street gangs from a systems' perspective provides information on how members can replace dysfunctional

systems with more appropriate, working systems. Appropriate systems may be developed by providing alternative resources

and activities in place of repetitive gang activities. For example, intervention programs that teach gang members job skills

and provide job placement will redirect gang members into more positive activities, instead of just lecturing to members that

they should not participate in negative gangactivities.

5. Understanding the systemic cohesion involved in gang membership provides knowledge about how gangsoperate as

"family systems." Sensitivity toward gang loyalty and codes of secrecy allows for a conducive environment for intervention

plans. Thus, gangs should be viewed as family systems who care for, protect, and support one another. Although violent

activity may accompany certain aspects of gang life, there are other parts of gang membership that are not often discussed.

The cohesion, connectedness, loyalty, and protective nature of gang organization often receive little or no attention. It is

important that these concepts be recognized, since violence is often a forced product of these underlying assumptions that

hold street gangstogether.

FAMILY THERAPY INTERVENTIONS

Since family dysfunction has often been identified as a correlate to gang involvement, intervention and treatment plans must

focus on the underlying family issues that may have led the gang member to search for another family system. Much of the

therapy process may need to center around the family's history of violence, neglect, and/or abuse. Families must

acknowledge the impact that abuse and/or neglect has had on the family system and on the individual child who has

suffered and/or witnessed the family violence. Understanding the gang member's emotional responses to family traumas

allows for a better understanding of the context and meaning of his/her gang participation (Belitz & Valdez, 1994). Family

therapy should also address the family's dynamics, roles, rules, and means for resolving family conflict (Clark, 1992).
Understanding these concepts within the family system provides a window into the thoughts, emotions, and behaviors that

may have been learned from the youth's family of origin and then later modeled through gang membership and activities.

A primary goal for family therapy is to assist the family in verbally expressing and listening to one another without becoming

aggressive or punitive. Family therapy allows the gang member to express the anger, hurt, resentment, and sense of

rejection at those in the family who may have abused, neglected, or failed to protect the youth (Belitz & Valdez, 1994). This

time for expression forces parents and other family members to actively listen to the emotional expression of the youth with

the support and guidance of a therapist. For many family members, this may be the first time they have heard the emotional

context surrounding the youth's gangmembership. For the youth, this may be the only time he or she feels safe enough to

express those deeply buried feelings and emotions. Concurrently, families need to be taught healthy and effective

communication patterns as the main mechanism for resolving family conflict in a way that is both respectful to individuals

and to the family system. During these therapy sessions, when the lines of communication are finally open for the family, the

issue of forgiveness must be addressed. If parents and family members want forgiveness for events that have taken

place in the past, they must ask for it. Because some of the youth's scars are so deeply embedded, it may take months or

even years for forgiveness to take place. If parents and other family members do not feel a need to ask for forgiveness,

the gang member must be prepared to let go of his or her anger and initiate the healing process without the support of

parents and/or family members.

It is very important during family sessions that the youth's role in the family be explored and redefined. For example, gang-

related activities and behaviors that are violent may be refrained as learned behavior from the youth's family of origin (Belitz

& Valdez, 1994). New and more appropriate beliefs and behaviors must be introduced to the youth and his or her family

during therapy sessions. Furthermore, a discussion on gender roles and societal stereotypes of men and women is also

highly relevant. For many gang members, the only male and female role models they have had are of a dominant and

violent father and a dependent and submissive mother. These concepts and beliefs should be explored by the therapist,

followed by a psychoeducational approach to challenge the learned gender roles adopted by the youth and his or her family.

Family sessions should also explore the sociocultural stressors that are influencing the youth and the family system.

Discussions of the cultural context surrounding the family offers insight and understanding into a number of family dynamics

and the cultural meaning of gang participation. Therapy sessions should focus on identifying cultural strengths and

integrating those strengths into more positive individual and family identifies (Belitz & Valdez, 1994).

If gang youth are still living at home, parents need to be empowered as the decision-making adults in the family. Family

rules need to be established along with the consequences for breaking those rules (Belitz & Valdez, 1994). Structural family

therapy can be used to help redefine the parents as the executive subsystem. Youth need to understand their place in the

family hierarchy and learn the new roles and behaviors that are involved in being part of the child subsystem.

Exercises in the therapy room, such as role plays and family sculptures, can help to reinforce the new hierarchial structure.
Some families may bring multigenerational gang involvement into family therapy. This may complicate family intervention

because family members support one another in their gang activity. For this reason, it is important that

multigenerational gang participation be dealt with directly, keeping in mind the family's systemic dynamics (Clark, 1992).

Supplementary Interventions

Although family therapy should be the main focus of intervention for gang youth, other treatment plans, such as individual

and group work, also can be effective and valuable components of the therapy process. Individual work, in conjunction with

family therapy, allows youth to explore their own identities both within their family system and as independent and

autonomous individuals outside of the system. Individual treatment with gang members should focus on identifying how

basic needs are being met by gang participation. Identifying these needs includes exploring the gang member's yearning for

a family to provide basic dependency needs and the quest for the gang to provide psychological and cultural identity. It

should be noted that asking a gang member to drop out of the gang so that productive therapy can continue is a highly

unrealistic request. Gang members should be encouraged to develop more adaptive ways of meeting needs and

functioning inthe larger world so that gang involvement can progressively become less important. Gang members should

also be encouraged to maintain the positive elements that may be associated with gang membership, such as being strong,

loyal, willing to take risks, and being a survivor. Furthermore, individual treatment may allow for gang members to

philosophically examine the meaning of life and death, their own underlying meaning of gang involvement, the importance of

morality, and the human capacity for positive and negative thoughts and behaviors (Belitz 8r Valdez, 1994).

Group therapy can be a powerful agent of change for many gang members. Group work is most effective after some family

and individual work has already been done. Once the youth has committed to making a change inhis or her life, the group

process can be used as a strong reinforcement agent to achieve these goals. Participating in a group often

allows gang members to receive validation for their feelings and experiences. This helps participants to break down the

artificial barriers that separate them from rival gang members and reduces the alienation they may feel when they are

outside their gang territory. A major focus of group work with gang members is exploring identity issues and how

participating in a gang relates to those issues. The group also facilitates the learning of new social skills, such as providing

and receiving feedback among peers, being assertive in a nonaggressive manner, and providing and receiving emotional

support from peers, family members, and role models. Furthermore, the group process should include an education series

about mentoring and tutoring services, job skills development, and employment assistance. The effectiveness of the group

process is further strengthened if former gang members either help lead the group or come to the group on a regular basis

to offer their own experiences related to gang membership. Current gang members are often empowered and motivated by

former gang members who have now succeeded in more positive endeavors.

In order to conduct a successful group of diverse gang members, some specific rules must be enforced. Some rules may

include maintaining the property where the group is held as neutral territory, no tagging or graffiti on the property, showing
respect for others by not verbally or physically threatening one another, allowing everyone an opportunity to participate, and

attending on a voluntary basis (Belitz & Valdez, 1994).

CONCLUSION

For years, it has been popular belief that street gangs are just a collection of hoodlums who only cause trouble and make

neighborhoods look bad. Little research has been conducted on how gangs operate as systems and what systemic function

the gang serves for the members who are part of that system. As stated earlier, the gang serves as a family for

most gang members. Many individuals who join gangs are looking for the closeness, cohesion, and acceptability that they

have been deprived of elsewhere. A gang may provide members with esteem, stability, and a connection with the world that

they have never experienced. Granted, gangs may also provide danger and violence for their members, but to truly

understand street gangs, those aspects cannot be studied as separate entities. Gangs, as all systems, function based on an

interrelatedness that connects every aspect of gang life together within a complex web of interactions. Successful research

endeavors and prevention and intervention programs must aim to approach street gangs from a holistic and systematic

perspective, taking into account the highly complex structure, process, functionality, and culture that are located within the

core of every street gang.

References

REFERENCES

References

Becvar, R. J., & Becvar, D. S. (1982). Systems theory and family therapy. A primer New York: University Press of America.

Belitz, J., & Valdez, D. (1994). Clinical issues in the treatment of Chicano male gang youth. Hispanic Journal of Behavioral

Sciences, 16, 57-74.

Borringer, S. (1995). Personal communication. Appalachian State University, Boone, NC.

Broderick, C. B. (1990). Family process theory. In J. Sprey (Ed.) Fashioning family theory. New approaches (pp. 171-206).

Newbury Park, CA: Sage.

References

Clark, C. M. (1992). Deviant adolescent subcultures: Assessment strategies and clinical interventions. Adolescence, 27,

283-293.

Conly, C. H., Kelly, P., Mahanna, P., & Warner, L. (1993). Street gangs: Current knowledge and strategies. Washington, DC:

National Institute of Justice.

Fagan, J. (1989). The social organization of drug use and drug dealing among urban gangs. Criminology, 27, 649-(52.
Hagedorn, J., & Macon, P. (1988). People and folks: Gangs, crime, and the underclass in a rustbelt city. Chicago: Lake View

Press.

Haley, J. (1976). Problem-solving therapy. San Francisco: Jossey-Bass.

References

Hardy, K. (1996, March). Adolescents and violence. Workshop presented at the Annual Meeting of the Kentucky Association

for Marriage and Family Therapy, Louisville, KY.

Harris, M. G. (1994). Cholas, Mexican-American girls and gangs. Sex Roles, 30, 289-301. Huff, C. R. (1989).

Youth gangs and public policy. Crime and Delinquency, 35, 524-537.

Jankowski, M. S. (1991). Islands in the street: Gangs and American urban society. Berkeley: University of California Press.

Joe, K. A. (1994). The new criminal conspiracy? Asian gangs and organized crime in San Francisco. Journal of

Research in Crime and Delinquency, 31, 390-415.

Kennedy, L. W., & Baron, S. W. (1993). Routine activities and subculture of violence: A study of violence on the street.

Journal of Research in Crime and Delinquency, 30, 88-112.

References

Lyon, J. M., Henggeler, S., & Hall, J. A. (1992). The family relations, peer relations, and criminal activities of Caucasian and

Hispanic-American gang members. Journal of Abnormal Child Psychology, 20, 439-449.

Miller, W. (1975). Violence by youth gang and youth groups as a crime problem in major American cities. Washington, DC:

U.S. Government Printing Office.

Minuchin, S. (1974). Families and family therapy. Cambridge, MA: Harvard University Press.

Moore, J. W. (1991). Going down in the barrio: Homeboys and homegirls in change. Philadelphia, PA: Temple University

Press.

Morales, A. T. (1992). Latino youth gangs: Causes and clinical intervention. In L. A. Vargas & J. Koss-Chioino (Eds.)

Working with culture: Psychotherapeutic intervention with ethnic minority children and adolescents (pp. 129-154). San

Francisco: Jossey-Bass.

References

Nichols, W. C., & Everett, C. A. (1986). Systemic family therapy: An integrative approach. New York: Guilford.

Rogers, C. (1993). Gang-related homicides in Los Angeles County. 43rd Annual Meeting of the American Academy of

Forensic Sciences. Journal of Forensic Sciences, 38, 831-834.

Sanders, W. B. (1994). Gangbangs and drive-bys: Grounded culture and juvenile gang violence. New York: Aldine De

Gruyter.

Siegel, L. J., & Senna, J. J. (1985). Juvenile gangs in the United States. In Juvenile delinquency: Theory, practice, and law,

(2nd ed., pp. 253-276). St. Paul, MN: West Publishing.

Spergel, I. (1990). Youth gangs: Continuity and change. In N. Morris & M. Tonry (Eds.) Crime and justice: A review of

research (pp. 177-179). Chicago: University of Chicago Press.

References

Vigil, J. D. (1988). Barrio gangs: Street life and identity in Southern California. Austin, TX: University of Texas Press.

Walsh, F. (1982). Conceptualizations of normal family functioning. In F. Walsh (Ed.) Normal family processes (pp. 3-28).

New York: Guilford Press.

Winfree, L. T., Backstrom, T. V., & Mays, G. L. (1994). Social learning theory, self reported delinquency, and youth gangs: A

new twist on a general theory of crime and delinquency. Youth & Society, 26, 147-177.

AuthorAffiliation

NIKKI M. RUBLE

Florida State University, Tallahassee, Florida, USA

WILLIAM L. TURNER

University of Kentucky, Lexington, Kentucky, USA

AuthorAffiliation

Address correspondence to Nikki Ruble, 230 Sandels Building, Florida State University, Tallahassee, FL 32306. E-mail:

nruble@hotmail.com

Word count: 7117

Copyright Brunner-Mazel Publishing Company Apr-Jun 2000

Ruble, N. M., & Turner, W. L. (2000). A systemic analysis of the dynamics and organization of urban street gangs. The American

Journal of Family Therapy, 28(2), 117-132. Retrieved from http://search.proquest.com/docview/230116970?accountid=37714

Urban violence and street gangs

James Diego VigilAuthor Information . Annual Review of Anthropology 32 (2003): 225-242.

1. Full text
2. Full text - PDF
3. Abstract/Details
 4. References 156
Hide highlighting
Abstract
TranslateAbstract

Abstract What causes urban street gang violence, and how can we better understand the forces that shape this type

of adolescent and youth behavior? For close to a century, social researchers have taken many different paths in attempting

to unravel this complex question, especially in the context of large-scale immigrant adaptation to the city. In recent decades

these researchers have relied primarily on data gathered from survey quantitative approaches. This review traces some of

these developments and outlines how frameworks of analysis have become more integrated and multidimensional, as

ethnographic strategies have come into vogue again. For the last couple of decades, either a subculture of violence (i.e., the

values and norms of the street gang embrace aggressive, violent behavior) or a routine activities (i.e., hanging around high

crime areas with highly delinquent people) explanation dominated the discussion. To broaden and deepen the picture, many

other factors need to be considered, such as ecological, socioeconomic, sociocultural, and sociopsychological,

particularly in light of the immigrant experience. A multiple marginality framework lends itself to a holistic strategy that

examines linkages within the various factors and the actions and interactions among them and notes the cumulative nature

of urban street gang violence. Questions that are addressed inthis more integrated framework are: Where did they settle?

What jobs did they fill? How and why did their social practices and cultural values undergo transformations? When

and in what ways did the social environment affect them? Finally, with whom did they interact? In sum, in highlighting the

key themes and features of what constitutes urban street gang violence, this review suggests that the qualitative style that

relies on holistic information adds important details to traditional quantitative data.

Full text
TranslateFull text
Turn on search term navigation

Headnote

Key Words urban studies, integrated and holistic frameworks, gang formation theories

Headnote

Abstract What causes urban street gang violence, and how can we better understand the forces that shape this type

of adolescent and youth behavior? For close to a century, social researchers have taken many different paths in attempting

to unravel this complex question, especially in the context of large-scale immigrant adaptation to the city. In recent decades

these researchers have relied primarily on data gathered from survey quantitative approaches. This review traces some of

these developments and outlines how frameworks of analysis have become more integrated and multidimensional, as

ethnographic strategies have come into vogue again. For the last couple of decades, either a subculture of violence (i.e., the

values and norms of the street gang embrace aggressive, violent behavior) or a routine activities (i.e., hanging around high

crime areas with highly delinquent people) explanation dominated the discussion. To broaden and deepen the picture, many

other factors need to be considered, such as ecological, socioeconomic, sociocultural, and sociopsychological,

particularly in light of the immigrant experience. A multiple marginality framework lends itself to a holistic strategy that

examines linkages within the various factors and the actions and interactions among them and notes the cumulative nature
of urban street gang violence. Questions that are addressed inthis more integrated framework are: Where did they settle?

What jobs did they fill? How and why did their social practices and cultural values undergo transformations? When

and in what ways did the social environment affect them? Finally, with whom did they interact? In sum, in highlighting the

key themes and features of what constitutes urban street gang violence, this review suggests that the qualitative style that

relies on holistic information adds important details to traditional quantitative data.

Urban gang violence has been examined from various sociological and psychological perspectives (Covey et al. 1992,

Decker 1996). Though the violence includes an array of crimes (Zimring 1998), it is the gang conflicts that concern us here:

the turf and drug wars and battles over resources and the drivebys and "counting coup" escapades. Prior to the

1970s, gang violence was still popularly associated with white ethnic enclaves in the cities of the Midwest and East,

and gang incidents were typically brawls involving fists, sticks, and knives. Today, gangs are made up largely of darker-hued

ethnic groups, especially African Americans and Latino Americans, and handguns and other military hardware are the

typical vehicles for the acts of aggression and violent rampages so common in large cities, West and East (Cook & Laub

1998, Sanders 1994, McNulty 1995).

Gangs are now made up, as they were in earlier days, primarily of groups of male adolescents and youths who have grown

up together as children, usually as cohorts in a low-income neighborhood of a city. Yet, only about 10% of youth in most low-

income neighborhoods join gangs (Vigil 1988, Short 1996, Esbensen & Winfree 2001). Those who do so participate

together in both conventional and antisocial behavior (Thornberry 2001). It is the antisocial behavior, however, that attracts

the attention of authorities as well as the general public (Decker & Van Winkle 1996, Curry et al. 1994, Bursik & Grasmick

1995).

Hagedorn (1998) has made a distinction between the periods before and since the 1970s, respectively labeling them

industrial and postindustrial. He maintains that the latter period has brought more violence because of several factors: "the

adoption of economic functions by some urban gangs, the use of violence to regulate illicit commerce, the proliferation of

firearms, the effect of prison on neighborhood gangs, and the effect of mainstream cultural values of money and success

on gang youth with limited opportunities" (Hagedorn 1998, pp. 369-70).

Access to sophisticated weaponry has made violence easier to carry out, as Canada (1995) has illustrated inhis Fist, Stick,

Knife, Gun. Indeed, in 1968 there were about 80 million guns in the United States, and by 1978 the figure had grown to 120

million; it was 200 million by 1990 (Reiss & Roth 1993). There were 50 million handguns, including those inexpensive pistols

often referred to as Saturday Night Specials, in 1991 (Hagedorn 1998). The historical importance of guns in U.S. culture and

the strong political support from groups such as the National Rifle Association (NRA) for continued manufacturing and

relatively unrestricted sales of firearms have worked to facilitate gang acts of aggression and violence (Lott & Mustard

1997). Meanwhile, drug use and abuse have blurred the thinking of, and in a limited way drug sales have increased motives

for, the street youth who perpetrate most of the violence (Blumstein 1995, Klein & Maxson 1994). However, there are also

other factors that need to be considered in comprehending gang violence (Spergel & Curry 1998).
Sociologists and social psychologists have extensively examined youth violence. Many, like Farrington (1996, 1998), have

generally alluded to violence as an aspect of human aggression (Huesmann & Moise 1999). More recently, others,

Messerschmidt (1995) for example, have seen it as an expression of masculinity in working-class culture. In concurrence

with anthropological evidence, most agree that males are more physically aggressive than females (Rohner 1976, Ember

1995). However, in recent years female gang members or females affiliated with male gangs have begun to get

involved in and participate in gang violence (Chesney-Lind 1999). Female gang members are many fewer in number-from

4% to 15% of all gang members (Campbell 1991, 1990; Curry 2001; Miller 2002, 2001), but studies show that a high

percentage of all incarcerated females in the United States belong to gangs (Giordano 1978). Indeed, the arrest rates of

young women recently have increased at a faster pace than for nongang males (OJJDP 1996), and the offenses they are

being charged with increasingly involve violence. Of course, some of that increase is probably due to the fact that the

criminal justice system is less gender-biased in recent years: Females who engage in deviant gang behavior are more apt

today to be treated like their male counterparts rather than being shuttled into counseling sessions, as was likely in the past.

Girls and young women in gangs (usually in auxiliaries of male gangs but sometimes in autonomous groups or even,

rarely, in mixed gender groups) are especially severely impacted by street socialization. Like males they must struggle with

the same forces that generated their street experience but in addition must contend with their own homeboys, who typically

have little respect for them. As gender roles continue to change, however, violence is becoming a more salient aspect of the

role of females in gangs (Campbell 1990).

Like young males, many female youths are subjected to: culture conflict, poverty, and associated family and school

problems. In addition, they are apt to undergo personal devaluation, stricter child-rearing experiences, tension-filled gender

role expectations, and problems with self-esteem stemming from all these forces. Sexual abuse and exploitation

experiences, initially with male relatives and later male street peers, can lead to pent-up rage. Not surprising, some young

females are now channeling that rage into holding their own in the violence of the street gang world. Thus, some females

may take on the persona of a crazy person, as for instance the Chicanas who call themselves Loca and do their best to live

up to it (Dietrich 1998). Male gang members are generally dismissive of "gang girls"; but around these extremely violence-

prone young women, the males show much more respect.

Nevertheless, most aggressive gang behavior is committed by male youths between 14 and 18 years of age (Hirschi &

Gottfredson 1983, Gottfredson & Hirschi 1990, Zimring 1996, Oliver 2003). Erikson (1968) has singled out the psychosocial

moratorium in that age span as a status crisis in the transition from childhood to adulthood. Ambivalence and unpredictability

characterize that period of human development, as teens face their new social and sexual identities and roles with

uncertainty. As a result, they increasingly rely on peers and slightly older male role models as guides. Some teenagers,

especially those living in highly stressful conditions, find it particularly difficult to work through this stage of human

development. Many gangmembers, who characteristically have been raised in marginalized, highly stressful families, have

their social development arrested and remain peer-dependent well into their thirties and even forties. Of importance to note,

sociologists have quantitatively charted a noted increase, high point, and decrease in aggressive, violent actions in that
same 14 to 18 age span (Hirschi & Gottfredson 1983, Gottfredson & Hirschi 1990). Because gang members often are

organized hierarchically by age, older gang members are able to goad younger members (the 14- to 18-year-olds) to carry

out acts of violence against rival gangs (Vigil 1988a,b).

Violence can be defined as a conscious physical act aimed at causing injury, which often includes bodily and psychological

trauma and usually is identified with certain neighborhoods (Tonry & Moore 1998, Oliver 2003). Most current theories are

framed to explain delinquency or aggression and not youth violence (Farrington 1998). Although there is considerable

research still in its infancy focused on biological determinants of youth violence, such as low heart rate, the majority of the

investigations and evidence support social and psychological factors (Raine et al. 1997). Indeed, when biological factors are

invoked they are usually associated with social and psychological explanations.

Let us summarize the anthropological, ethnographic tradition on the subject. Members of the Chicago School conducted

some of the first studies on gangs in the 1920s. These researchers were a group of sociologists and anthropologists who

advocated and regularly practiced urban community studies that involved fieldwork observation and intensive interviewing.

Thrasher (1927) is the initial researcher credited with studying gangs(and also Ashbury 1927 in New York, but his work was

more anecdotal) in detail and establishing the baseline information from which all subsequent studies followed, including

those of today (Hagedorn 1988, 1998). Later, Whyte (1943), Yablonsky (1966), Keiser (1969), Suttles (1968), Dawley (1992

[1973]), and Miller (1973, 1975) added to this ethnographic tradition. Soon after, a renewed interest in gangs emerged

among ethnographically oriented researchers such as Moore (1978, 1991) and Horowitz (1983), and by the late 1980s and

1990s qualitative, anthropological investigations were underway across the nation (Alonso 1999). Anthropologists (e.g., Vigil

1988a, 2002a; Sullivan 1988; Bourgois 1995; Fleischer 1995, 1998) have reworked and reconfigured these perspectives.

Frameworks of analyses are now more holistic (Bronfenbrenner 1979), and methods of research have been broadened as

ethnographic approaches have become more popular, even among sociologists (Hagedorn 1988; Padilla 1992; Anderson

1994, 1997, 1998).

In the more recent past, major explanations of gang violence can be fit (usually) into one of two frameworks: subculture of

violence (e.g., especially Wolfgang & Ferracuti 1967) and routine activities (e.g., Felson 1987) explanations. In view of

recent emphases on multidimensional ways of looking at the subject, I underscore that neither becoming a member of a

subculture of violence nor being present where routine activities of a violent nature occur is sufficient, by itself, to explain

youth participation in gang violence. Let us take each of these two explanations and describe them in more detail.

The subculture of violence construct posits that it is the normative behavioral systems of groups that support, encourage,

and condone violence (Wolgang & Ferracuti 1967). These norms help guide gang members in how and when they react to

real or imagined slights and threats to themselves or fellow gang members, such as hostile stares (called "mad dogging" by

street youths in Los Angeles), a chance encounter with known gangenemies (e.g., when cruising or walking in nongang

territories), or paybacks (i.e., retaliation by consciously seeking gang enemies to attack). Violence is expected or required

under these and other conditions and situations; otherwise the gang member risks being disrespected ("dissed") by
other gang members. Failure to live up to these norms brings a loss of honor, and, as Horowitz (1983) has stated, ones self-

image is tarnished for either not promoting or defending these norms.

Many other researchers have relied on one form or another of the subculture of violence explanation. Some of the most

knowledgeable of gang writers have focused on heart and courage in fighting (Cohen & Short 1958), the lower-class value

of confronting trouble and exhibiting toughness (Miller 1958), or the emphasis on violent behavior in a "conflict subculture"

(Cloward & Ohlin 1960). Focusing on disputes between individuals, Luckenbill & Doyle (1989) have underscored that

challenges to the notion of self result in a series of events (i.e., naming, claiming, and aggressiveness) that leads to the use

of force or violence by members of the subculture.

In contrast to a subculture of violence perspective, a routine activities explanation (Felson 1987) relies on space/time

dimensions (e.g., at the wrong place at the wrong time) to substantiate a pattern of violence. The potential for violence is a

product of opportunity where one spends more time with criminal offenders who are more likely to participate in offending

activities. In short, motivated offenders, suitable targets, and an absence of capable guardians converge in certain times and

places to increase the possibility of a crime (Kennedy & Baron 1993, Felson & Cohen 1980). Risky lifestyles of young males

on the streets increase contact with similarly adventurous males and thus such associations and interactions heighten the

potential for crime and victimization (Sampson & Lauritzen 1990).

A number of other writings reflect the routine activities approach, stressing that it is the " `criminogenic potential' of certain

routines that accounts for their `victimization potential"' (Kennedy & Baron 1993, p. 92). Most assaults take place in contexts

such as bars, parties, and other gatherings. Although ecological factors are paramount in a routine activities explanation,

some writers suggest that rational choice is involved even when certain areas or neighborhoods are known as high-risk

places (Seigal & Senna 1991, Cornish & Clarke 1986). A whole set of spatial and human factors are reflected in violent

episodes.

In sum, both the subculture of violence and routine activities approaches can play a role in broadening our understanding of

the issue of gang violence. Some of the shortcomings in the subculture of violence approach are quite obvious. Little

empirical evidence has been offered on how violent norms are transmitted, and much of the inferences on subcultural values

are derived from the behavior. Moreover, the approach tends to suggest a level of organization in gangs that is not

present in the street gangs that others and I have studied, which are more loosely affiliated (Vigil 1988a, 2002a; Fleischer

1998; Klein & Crawford 1967; Short & Strodtbeck 1965; Suttles 1968). On the other hand, a routine activities explanation

also has its limitations. For example, this perspective typically provides very little ethnographic evidence, relying mostly on

official crime statistics, demographic variables, and victim surveys. Though time/space are crucial, an anthropological

perspective suggests that this type of data must be combined with other factors. Thus, other situations and conditions need

to be considered in the equation.

Two important considerations will provide examples for why a broader picture must be drawn if we are to better

understand gang violence. Embedded in both the subculture of violence and routine activities frameworks are the issues of
street realities and the state of mind of the individual. In previous research (Vigil 1987), I have outlined the importance of

street socialization and a street state of mind known as "locura" (from loco, crazy, and defined as a spectrum of behavior

reflected in a type of quasi-controlled insanity) among Mexican-American gangs and "crazy niggah" among

black gang members. A short elaboration of street socialization is followed by the meaning and import of locura.

Street realities insure that a street subculture emerges among children that are bereft of social control from families, schools,

and law enforcement, which have all failed to maintain or even gain a guiding influence on their lives. Street socialization is

important because some individuals with particularly tarnished, traumatic family and personal backgrounds have had to

spend most of their lives in the streets. This begins in early childhood and reaches a high point when an age/gender identity

crisis erupts during adolescence. It is then that the more group-oriented preteen activities coalesce and merge into that of

the street gang, and in most instances it is a continuous process. As a result, youth who are street socialized dictate the

behavioral and attitudinal traits of the streets. The streets have become the arena for what is learned and expected by

others to gain recognition and approval. Friendship and counsel, protection from street predators, and a wild, adventurous

lifestyle become crucial to street survival. This type of survival through safety in numbers and reliance on friends to "watch

your back" can lead to reckless behavior and an early adoption of locura as a way to negotiate the streets and especially to

gain the support of street peers.

The culmination of all the street experiences is the shaping of a mindset of locura. It is an attitude that is deeply internalized

by some gang members, especially the regular ones who have had particularly traumatic lives and are "crazy-like," but is

equally instrumental as an attitude that can be adopted as circumstances dictate. Thinking and acting loco is like playing

with insanity, moving in and out of crazy, wild events and adventures, showing fearlessness, toughness, daring, and

especially unpredictable forms of destructive behavior. This psychosocial mindset has become a requisite for street survival

and a behavioral standard for identification and emulation. Gang members collectively value locura because it helps

assuage fear and the anxiety associated with the fight-flight (and even the middle ground of fright) dilemma that street

realities impose on a person.

Thus, a complex problem such as gang violence necessitates examining many factors, such as neighborhood effects,

poverty, culture conflict and sociocultural marginalization, and social control, among other gangdynamics. Such a

combinative approach has been defined and described in other works either as multiple marginality (Vigil 1988a, 2002a;

Vigil & Yun 2002) or integrated systems (Elliot 1994, Farrington 1996) or multivariate analysis (Cartwright & Howard 1966).

But as we note, it also allows for the inclusion of street socialization (Vigil 1996, Bourgois 1995) and locura [the

psychological state of quasi-controlled insanity (Vigil 1988b, Yablonsky 1966)]. In short, an eclectic approach

(Bronfenbrenner 1979) is warranted to address the array of facets central to gang violence, and they must be

integrated in ways that show the actions and reactions among them.

Multiple marginality (Vigil 2002a, 1988a; Vigil & Yun 1998, 2002) is an integrated framework that allows for the inclusion of

the subculture of violence, routine activities, street socialization, locura, and other factors; inshort, it combines sociogenic
and psychogenic elements and actions (see Figure 1). Klein (1995), Elliot (1994), and Farrington (1996) utilize

multidimensional frameworks too but interpret the phenomena more from a sociological perspective (Covey et al. 1992).

Multiple marginality addresses the questions of what, where, how, why, and with whom, and it aids in explanations that show

dynamic exchanges and interrelationships. As brief examples of how multiple marginality might shed light on these

questions, let us take each question inturn to examine how they relate to violence. "What" is a query that is easily answered

by defining and describing what constitutes a youth gang, including its violent proclivities. "Where" places the gang in a

certain locale in sections of urban areas that are usually visually distinct and spatially separate from more upscale

neighborhoods and where violence is more likely. "How" requires an explanation of the social mechanisms and

psychological predispositions of gang members in carrying out the gang's most salient activities, violent acts among them.

"Why" is answered when we describe the situations and motivations that shape the thinking and actions of gang members

toward either other gang members or unsuspecting targets for violence. "When" is simply the time and place that is more

likely to trigger gang actions that end inviolence. Finally, "With Whom" carefully outlines the characteristics

of gangs and gang members from different neighborhoods and cohorts who are likely to participate in the violent street

rituals that often lead to injury or death. A more detailed analysis follows with examples from various research enterprises

and theoretical persuasions.

This review on urban violence and gangs emphasizes a multiple marginality framework as a way to integrate the concepts of

subculture of violence, routine activities, street socialization, and locura, while noting other significant factors where

appropriate. There are gangs in many ethnic communities to which this more holistic framework applies (Vigil 2002a), and

some of the assessments may have limited application to other regions of the world (Klein et al. 2001, Hazlehurst &

Hazlehurst 1998, Blanc 1995).

Enlarge this image.

Figure 1

Most important for the purposes of this review is that multiple marginality is more than a laundry list of factors but a model

showing sequential, cumulative linkages among factors. For anthropologists particularly, but also for the

early gang researchers, the quest for an understanding of urban gangs began when researchers started following peasants

and sharecroppers into cities. Thus, immigration and the experiences of immigrants adapting and adjusting to city life form

the basis for all else that follows, including and especially the maladaptation that so often occurs among them (Martinez

2000). In this vein, there are multiple areas in which immigrants and especially their children find themselves betwixt and

between, beginning with where they settle, what jobs they fill, and how and why their social and cultural values and practices

are challenged and typically undermined and revamped. It also takes into account when the social environment shapes

personal identities with whom the individuals interact. As noted above, no more than 10% of the youths

become gang members in most affected neighborhoods, and the most marginalized families and children in each of these

neighborhoods tend to fall into this category.

On a theoretical plane, multiple marginality is, in an essential way, similar to the integrative framework that Farrington

(1996,1998) and Bronfenbrenner (1979) suggest is needed for increased explanatory power. As such, it accounts for the

reciprocal actions and reactions among factors, taking stock of sources and modes of human aggression, all the while

identifying the interrelationships of social and personal development in the context of rapid urbanization and uneven culture

change.

One of the first things to contemplate is that the populations of mostly young ethnic minority populations that contain most of

the gang members in the United States do not live in the more comfortable areas of the cities. They did not fall from grace

and then settle in the worst, most marginal places of the cities. The overwhelming fact is that they started their city

lives in places they could afford, usually rundown, dilapidated, worn-out residences on the East coast where a criminal life

style was already in vogue (Venkatesh 1996). Inthe West, the Los Angeles area primarily, often inferior empty spaces were

all that remained for the newcomers, and thus squatter-like settlements arose. Here is where the ecological factor takes

center stage (Vigil 2002b).

Place, in this instance, breeds certain human traits and adaptations. This factor gave us the concentric circle explanation of

the Chicago School in the early crime and gang research days, a theory that singled out the second circle of the bull's-eye

as where immigrants settled, and crime and vice were highly prevalent as compared to the other inner and outer circles of

the model (Park et al. 1925, Shaw & McKay 1942). Indeed, as noted previously, it is in place that the routine activities theory

builds its case (Felson 1987, Kennedy & Baron 1993). Other writers have also relied on place to explain the propensity for

violence or violent incidents (Kornhauser 1978, Covey et al. 1992, Block & Block 1995). Generally, immigrants

reside in neighborhoods of cities that are visually distinct (e.g., rundown, inferior) and spatially separate (e.g., across the

tracks, freeway, river). In sum, being new to the city and separate from mainstream people and institutions limit the access,

exposure, and identification that one has with the dominant culture and customs, which thus blocks avenues for integration.

Being thwarted and bottled up in this manner, along with residential overcrowding, which I discuss shortly, engenders

frustration and aggression among poor people in general. Space becomes a premium among them; places to congregate,

shop, and cultivate romantic liaisons, among other social outlets and activities, become contentious issues, potential

battlegrounds for conflict.

Is it an accident that most gang violence, drug sales, and turf wars, among other rivalrous issues, occur between gangs from

similar marginal areas? Place alone seems to have an explanatory power beyond most of the other factors because so

many other human habits and adjustments emanate from the demands of place. However, let us turn to some of the other

elements that frame gang violence.

Socioeconomic status has received as much, perhaps more, attention than ecology in explanations for violence, but

certainly both in tandem show the cumulative buildup suggested by multiple marginality. Again, the discussion on status

focuses on youth from various ethnic minority populations, and, as with place, it is the marginality of their status that often is

the source for aggression and violence. For example, they know that they live on the other side of the tracks; have limited
access to entry-level jobs; receive harsh, uneven treatment from authorities, including and especially law enforcement; and

are faulted for their own problems (Vigil 1988a, 2002a).

Competition over resources often sparks aggressive behavior and actions that occasionally flare up and get out of control,

resulting in very violent, unplanned incidents. In recent years, much has been written about how industries and jobs have

vacated the inner city, the places we have mentioned, and set up business indistant suburban environs or foreign nations

(Wilson 1987, Hagedorn 1988, Johnson & Oliver 1991). This development has led many residents of the abandoned

neighborhoods to participate in illicit, illegal, and informal substitute economic arrangements, of which drug trafficking and

competition over markets was a major one (Beckett 1997). Drug trafficking and the conflict for markets, for a time

and in some cities but not all (Howell 1996, Maxson 1998), accelerated and heightened gang violence, leading to economic

explanations for the increase in gang violence.

Within a multiple marginality framework, however, the drug connection just added to an already dismal status dissonance

(Fagan 1999). In fact, quite early in gang research, several social scientists built strong cases for economic theories. The

most well known, of course, was Merton's strain theory, which simply refers to the mismatch between the means of low-

income peoples to reach the status goals established by dominant society. With such structural barriers in place, differential

opportunity paths were sought, such as illicit or illegal ventures that often entailed violence (Merton 1949, Cloward & Ohlin

1960, Kornhauser 1978, Covey et al. 1992, Moore 1978).

Yet, it should be obvious that not everybody in such places or statuses becomes a violent gang member because of where

they are and what they do. Indeed, as noted earlier, the great majority of youths in these situations avoid involving

themselves in gang activities. Nonetheless, they have to learn how to negotiate and navigate through the aura of aggression

and violence that pervades certain places. This variation among low-income peoples in the same neighborhood can involve

a number of things. What, then, leads the minority into gangs and violence? Can one live in the poorest house among poor

houses? Can he have a much bigger family than other families in the area? Can his family be materially poorer than the

others and also have less social capital? And as Anderson (1998) puts it, can poor people be "decent" or "street"

irrespective of what demands are made of them? Finally, because of what is present in these marginal places and the

strains inherent in a marginal status, can these forces lead youth to spend more time in the street?

The discussion to this point has addressed the questions "where" and "what." Let us turn to "how," "why," and "with whom."

It is in these gang research areas that much has been written, but often with such a diffused impact, because what the

authors have to say is dependent on place and status as a basis for their explanations. Social and cultural marginality are

often intertwined; but for our purposes let us separate them for now. Socialization routines are definitely transformed when

immigrants or migrants of low socioeconomic status must adapt to a place in the city. This is particularly the case with social

control institutions, such as family, schools, and law enforcement.

Social control theory accounts for how these dynamic changes unfold, and there are many variations of this explanation

(Hirschi 1969, Goffredson & Hirschi 1990, Vigil 2002a, Covey et al. 1992). Families become stressed when their structure
and function change as they undergo urbanization, but the stress is greatly increased and intensified under the marginal

situations and conditions outlined above (Sampson & Laub 1990, Loeber & Stouthamer-Loeber 1986). For example, in the

face of job discrimination, both parents may have to work. In the absence of affordable childcare, without close friends or

relatives to care for them, children become latch-- keyed and have to fend for themselves during the day. Moreover, cultural

strains on rules and duties between first-generation parents and second-generation children, and attenuated father presence

and/or single-parent (mostly female) households, often add to a family's stress (Vigil 1988a, 2002a).

Schools generally have a poor record with poor, ethnic minority populations and especially so if there is a sharp cultural

contrast between the majority and minority cultures. This becomes painfully obvious when little or no communication or ties

between school and home occurs. This void is exacerbated in the classroom when cultural and language differences

interfere with learning. More important, the current school practices for students who are experiencing learning difficulties

due to poverty and/or cultural differences are narrow and poorly funded. Thus, for the most marginalized segments of the

minority populations, the social control mechanisms commonly inculcated by family and school are all but absent. The

cumulative effects, as an example, become overwhelming at this point: small dilapidated home; large family under crowded

conditions; precarious family structure with attenuated functions, sometimes marked by violence; unaccomodative schools;

hostile and aggressive law enforcement; and, in the absence of home and school socialization, street socialization emerges

and a multiple-aged peer group becomes a substitute (Vigil 1999).

Already among street youth the sources of aggression and violence become obvious, sometimes starting out with when they

are the victims of domestic violence to go along with what the streets have to offer. In short, street socialization leads to a

street subculture, and this is where and how the subculture of violence is learned and practiced (Wolfgang & Ferruti 1967).

This concept is a variation of Sutherland's much earlier work (1934) and essentially maintains that a violent way of life

dominates the streets, and a subcultural group of youth are the carriers that instruct newcomers in the art of street violence.

Street socialization combines elements from social control theory (i.e., family, schools, and so on) and the subculture of

violence concept; it explains how a person becomes exposed to the streets and then learns the gang subculture to

participate in violent acts.

Closely linked to this alteration of the socialization process is the cultural facet of multiple marginality. Enculturation is one

process of learning the culture to which one belongs, and acculturation is another process of learning a new culture. What

we have with many gang members from different cultural backgrounds is erring acculturation in their adaptation to the city

and all that it entails (i.e., place, status, social control and socialization, and so on). It is a core part of marginalization in that

fragmented family values and beliefs, uneven schooling and Anglicization, and culture contact and conflict changes lead

many youth to identification with the street. What I have referred to as "choloization" for Mexican Americans is simply a form

of the cultural marginalization that many other ethnic groups encounter. African Americans, for example, undertook different

waves of migration from the mostly rural South to American cities, north, east, and west throughout the twentieth century.

The creation of a street gang subculture itself is made up of bits and pieces, fragments, of their past culture as they
acculturate to the new, present culture; and in their case, it has been a historical process going back to their abrupt

uprooting from Africa.

Miller's theory of "lower-class culture as a generating milieu" (1958) for street violence addresses some aspects of the

results of cultural marginality. The new street values and beliefs that stem from these changes can work very well on the

streets when surviving violent confrontations is required; oftentimes the threat of violence is used by street youth as a

deterrent for having violence used against them. Thus, although street youth aggression and violence stem from various

sources and motives, it is the street gang subculture and the complexities of its formation and spread that best explain the

level of gang violence today-with drivebys, wanton shootings of innocent bystanders, and other physical harm as

the gang takes on a life of its own (Sanders 1994).

As noted earlier, the most aggressive and violent behavior among gang members occurs during that adolescent status

crisis, between childhood and adulthood. The social psychology of human development, especially the redefinition of

masculinity and capability, is filled with real and symbolic messages. Earlier marginal experiences in these youths' lives lead

to a social identity that, not coincidentally, shapes a personal identity during what Erikson referred to as the "psychosocial

moratorium" (Erikson 1968). This is a time inhuman development when ambiguity and conflict reign in the clarification and

affirmation of one's age and gender identity. Bloch & Neiderhoffer (1958) underscored in their gang study that "becoming a

man" was a strong generator of aggressive and violent behavior. Part of the evidence to this theory was referenced in a

cross-cultural study (Burton & Whiting 1961) that assessed male initiation rites that helped youth begin to think of

themselves as men, particularly in those societies where young males are raised almost exclusively by women but must

abruptly cut the apron strings and become male warriors. It should not escape our attention that most modem-

day gangs have an initiation ordeal that helps test and screen novitiates (Vigil 1988a, 1996); and, equally important, many of

the gang males come from single-parent families, usually female-headed, and then must adjust to street socialization

dominated by experienced gang males.

Humans are so malleable and resilient that, rather than remain lost in this aura of marginality, confused and full of rage, they

reconstitute a subculture and identity using bits and pieces of the past mixed with the present but all shaken by the forces of

a difficult city reality. Key to this street gang identity is learning to act crazy to survive (i.e., "loco" for Latino gangs and "crazy

niggah" for blacks), to be unpredictable, ready for any action, even killing somebody, to show you are "down" for your

homeboys and set or "barrio." Many of the gangmembers, however, have had particularly traumatic lives and, indeed, are

crazy; these are usually known as the most loco among Latinos and "ghetto heroes" among African Americans, according to

Monster Kody's autobiography (Shakur 1993). Suffering from a type of soul death and sense of worthlessness, leading them

to question why anyone else should be worth anything, these crazies can be responsible for most of the ganghomicides or,

at the least, for instigating more conflicts and confrontations. Strangely, sharing this aura of aggressiveness bordering on a

quasi-controlled insanity mindset and behavior makes for a strong street bond. Street gang members look up to one another

and show deference and respect for the locos and ghetto heroes. What has happened is that, either because of their own

soul death, personal traumas, and wish to prove their masculinity during the adolescent passage, or all of these, a
street gang member has joined more than a particular gang but a whole troop of suicidal persons who play a type of street

Russian roulette. The mentality is: A bullet is meant for you or me-if you are the one who survives, you, as would be

expected, amass more stature and respect. In recent years, unfortunately, the bullets that have moved out of the roulette

circle in the streets are marked "To whom it may concern"!

CONCLUSION

In its simplest trajectory, multiple marginality can be modeled as: place/status street socialization -> street subculture ->

street identity. The literature is rich in certain areas, such as social control, subculture of violence, and socioeconomic

factors, but notably lacking is the qualitative information and insight that tell us how and why people become violent if (a)

their family is dysfunctional, (b) they join a violent gang, or (c) they need money. The dynamic processes are

missing in most of these studies, however useful they may be inother regards. Anthropological theorizing and methods can

add significantly to the gang violence literature. As this review has noted, an integrated framework showing actions and

reactions among factors is where the research is going, and the multiple marginality framework is merely another step in that

direction. Taking a holistic perspective is an essential anthropological enterprise, and talking to and watching

people in different settings over a long period of time while they evince different moods and behaviors is also part of that

heritage (Bronfenbrenner 1979). In this manner, the actions and reactions are gathered, the reasons and modes for human

aggression better understood, and the connections between the street social identity and the personal identity are

discerned.

Anthropology and sociology were once one discipline; when they split, sociologists forgot their roots and reached for larger

samples to make more general statements about urban issues. Though street gangmembers are a difficult population to

investigate, as many sociologists already clearly understand, it is not that difficult to integrate intensive, in-depth interviewing

of gang members as part of large-scale surveys. Large numbers count, but it is in the details that human complexity is better

examined and understood.

The contexts of time, place, and people are also an important part of the equation. For example, some gangshave been

around for generations and are found in various large and middle-sized cities in the United States (Klein 1995).

Studies in larger cities show a tradition of an age-graded gang structure (Moore 1978, Klein 1971). This cohorting network

works to maintain the size of the gang as older members "mature out" and the younger gang members, approximately from

the ages of 14 to 18, are more likely to be led into and participate in violent behavior. Attention to female gangs and affiliates

has also increased (Miller 2001, Campbell 1991) and helped broaden the discussion of gender roles (Messerschmidt 1995,

Moore 1991, Moore & Hagedorn 1996). Newer gangs have emerged in different urban enclaves as a result of large-scale

immigration, where increasingly more street youth are opting for this lifestyle (Waters 1999). A few of these less-

rooted gangshave tried to catch up to the older gangs by becoming just as violent, and they are succeeding.

ACKNOWLEDGMENTS
A hearty "thank you" goes to the many community residents and street gang members who helped me better understand the

street life that generates so much violence. Students who helped on this project by digging up references are also

appreciated. Although I am responsible for the construction of this review and what is written and emphasized, I would like to

express a deep appreciation to my colleague, Cheryl Maxson, who offered critical advice and feedback along the way.

Finally, to my long-time friend and co-writer on other projects, John M. Long, goes a heartfelt thanks.

The Annual Review of Anthropology is online at http://anthro.annualreviews.org

References

LITERATURE CITED

References

Alonso A. 1999. Territoriality among AfricanAmerican street gangs in Los Angeles. MA thesis. Dept. Geogr., Univ. South.

Calif.

Anderson E. 1994. The code of the streets. Atl. Mon. May:81-94

Anderson E. 1997. Violence and the inner-city street code. In Violence and Childhood in the Inner City, ed. J McCord, pp. 1-

30. Cambridge, UK: Cambridge Univ. Press

Anderson E. 1998. The social ecology of youth violence. See Tonry & Moore 1998, pp. 65104

References

Ashbury H. 1927. Gangs of New York: an Informal History of the Underworld. New York: Knopf

Beckett K. 1997. Crime and Drugs in Contem

References

porary American Politics. New York: Oxford Univ. Press

References

Blanc CS, with contributors. 1995. Urban Children in Distress: Global Predicaments and Innovative Strategies. Florence,

Italy: UNICEF

Bloch HA, Neiderhoffer A. 1958. The Gang: a Study in Adolescent Behavior. New York: Philos. Libr.

Block CR, Block R. 1995. Street gang crime in Chicago. See Miller et al. 2001, pp. 18699

References

Blumstein A. 1995. Youth violence, guns, and the illicit drug industry. J. Crim. Law Criminol. 86:10-24

Bourgois P. 1995. In Search of Respect: Selling

References

Crack in El Barrio. New York: Cambridge Univ. Press

References

Bronfenbrenner U. 1979. The ecology of human development. Cambridge, MA: Harvard Univ. Press

Bursik RJ Jr, Grasmick HG. 1995. Defining gangs and gang behavior. See Miller et al. 2001, pp.8-13

Burton RV, Whiting WM. 1961. The absent father and cross-sex identity. Merrill-Palmer Q. 7:85-95

References

Campbell A. 1990. Female participation in gangs. See Huff 1996, pp. 163-82

Campbell A. 1991. The Girls in the Gang. New York: Blackwell. 2' ed.

Canada G. 1995. Fist, Stick, Knife, Gun. Boston: Beacon Press

Cartwright D, Howard K. 1966. Multivariate analysis of gang delinquency: ecological influences. Multivar. Behav. Res. 1:321-

37

Chesney-Lind M. 1999. Girls and violence: an overview. See Flannery & Huff 1999, pp. 171-200

References

Cloward RA, Ohlin LE. 1960. Delinquency and Opportunity: a Theory of Delinquent Gangs. New York: Free Press

Cohen AK, Short JF Jr. 1958. Research in delinquent subcultures. J. Soc. Issues. 68:452-62 Cook PJ, Laub J. 1998. The

unprecedented epi

demic in youth violence. See Tonry & Moore 1998, pp.27-64

Cornish DB, Clarke RV. 1986. The Reasoning Criminal. New York: Springer-Verlag Covey HC, Menard S, Franzese RJ.

1992. Ju

venile Gangs. Springfield, IL: Thomas Curry GD. 2001. Female gang involvement. See Miller et al. 2001, pp. 121-33

References

Curry GD, Richard AB, Fox RJ. 1994. Gang Crime and Law Enforcement Recordkeeping. Washington, DC: US Dept.

Justice, Natl. Inst. Justice, Res. in Brief

Dawley D. 1992 [1973]. A Nation of Lords. Prospect Heights, IL: Waveland Press Decker SH. 1996. Collective and

normative fea
tures of gang violence. Justice Q. 13:243-64 Decker SH, Van Winkle B. 1996. Life in the Gang: Family, Friends, and

Violence. New York: Cambridge Univ. Press

References

Dietrich L. 1998. Chicana Adolescents: Bitches, `Ho's, and Schoolgirls. Westport, CT: Praeger

Elliott D. 1994. Serious violent offenders: onset, developmental, and termination. The Am. Soc. Criminol. 1993 Pres.

Address. Criminology 32:1-21

Ember CR. 1996. Universal and variable patterns of gender difference. In Cross-Cultural Research for Social Science, ed.

CR Ember, M Ember, pp. 201-24. Upper Saddle River, NJ: Prentice Hall

References

Erikson E. 1968. Psychosocial identity. In International Encyclopedia of the Social Sciences, ed. D Sills, 7:61-65. New York:

McMillan, Free Press

Esbensen F, Winfree LT Jr. 2001. Race and gender differences between gang and nongang youths. See Miller et al. 2001,

pp. 10620

References

Fagan J. 1999. Youth gangs, drugs, and socioeconomic isolation. See Flannery & Huff 1999, pp. 145-70

Farrington DP. 1996. The explanation and prevention of youthful offending. In Delinquency and Crime: Current Theories, ed.

JD Hawkins, pp. 113-31. Cambridge, UK: Cambridge Univ. Press

Farrington DP. 1998. Predictors, causes, and correlates of male youth violence. See Tonry & Moore 1998, pp. 421-76

Felson M. 1987. Routine activities and crime prevention in the developing metropolis. Criminology 25:911-31

References

Felson M, Cohen LE. 1980. Human ecology and crime: a routine activities approach. Hum. Ecol. 4:389-406

Flannery DJ, Huff CR. 1999. Youth Violence: Prevention, Intervention, and Social Policy. Washington, DC: Am. Psychiatr.

Press

Fleisher MS. 1995. Beggars and Thieves: Lives of Urban Street Criminals. Madison: Univ. Wisc. Press

Fleisher MS. 1998. Dead End Kids: Gang Girls and the Boys They Know. Madison: Univ. Wisc. Press

Giordano PC. 1978. Girls, guys and gangs: the changing social context of female

References
delinquency. J. Crim. Law Criminol. 69(1): 126-32

References

Goffredson M, Hirschi T. 1990. A General Theory of Crime. Stanford, CA: Stanford Univ. Press

References

Hagedorn JM. 1988. People and Folks: Gangs, Crime and the Underclass in a Rustbelt City. Chicago: Lake View

Hagedorn JM. 1998. Gang violence in the postindustrial era. See Tonry & Moore 1998, pp.364-420

Hazlehurst K, Hazlehurst C. 1998. Gangs and Youth Subcultures: International Explorations. New Brunswick, NJ:

Transaction

Hirschi T. 1969. Causes of Delinquency. Berkeley: Univ. Calif. Press

Hirschi T, Gottfredson M. 1983. Age and the explanation of crime. Am. J. Sociol. 89(3):55284

References

Horowitz RB. 1983. Honor and the American Dream. New Brunswick, NJ: Rutgers Univ. Press

References

Howell JC. 1996. Youth Gangs, Homicides, Drugs and Guns. Tallahassee, FL: Natl. Youth Gang Cent.

Huesmann LR, Moise IF. 1999. Stability and continuity of aggression from early childhood to young adulthood. See Flannery

& Huff 1996, pp.73-95

Huff CR. 1996. Gangs in America. Thousand Oaks, CA: Sage. 2 ed.

Huff CR, ed. 2002. Gangs inAmerica III. Thousand Oaks, CA: Sage

References

Johnson J, Oliver M. 1991. Economic restructuring and black male joblessness in U.S. metropolitan areas. Urban Geogr.

12(6):542-62

Keiser RL. 1969. The Vice Lords: Warriors of the Streets. New York: Holt, Rinehart, & Winston

Kennedy LW, Baron SW. 1993. Routine activities and a subculture of violence: a study of violence on the street. J Res.

Crime Delinquency 30(1):88-111

Klein M. 1971. Street Gangs and Street Workers. Englewood Cliffs, NJ: Prentice-Hall Klein M. 1995. TheAmerican

Street Gang. New

York: Oxford Univ. Press

References

Klein M, Crawford LD. 1967. Group, gangs and cohesiveness. J. Res. Crime Delinquency 4:63-75

Klein MW, Kerner HJ, Maxson CL, Weitekamp EGM. 2001. The Eurogang Paradox: Street Gangs and Youth Groups in the

U.S. and Europe. Boston, MA, Dordrecht, The Neth.: Kluwer Acad.

References

Klein MW, Maxson CL. 1994. Gangs and cocaine trafficking. In Drugs and the Criminal Justice System: Evaluating Public

Policy Initiatives, ed. D MacKensie, C Uchida, pp. 136-48. Newbury Park, CA: Sage

Kornhauser RR. 1978. Social Sources of Delinquency. Chicago: Univ. Chicago Press Loeber R, Stouthamer-Loeber M.

1986. Family

factors as correlates and predictors of juvenile conduct problems and delinquency. In Crime and Justice, ed. M Tonry, N

Morris, pp. 29-149. Chicago: Univ. Chicago Press

Lott J, Mustard D. 1997. Crime, deterrence and right-to-carry concealed handguns. J. Legal Stud. 26:1-13

References

Luckenbill DF, Doyle DP. 1989. Structural position and violence: developing a cultural explanation. Criminology 27(3):419-35

Martinez R Jr. 2000. Immigration and urban violence: the link between immigrant Latinos and types of homicides. Soc. Sci.

Q. 81(l):365-74

Maxson CL. 1998. Gang homicide. In Studying and Preventing Homicide, ed. D Smith, M Zahn, pp. 197-219. Thousand

Oaks, CA: Sage

References

McNulty PJ. 1995. Natural born killers? Preventing the coming explosion of teenage crime. Policy Rev. 71:84-95

Merton RK. 1949. Social Theory and Social Structure. Glencoe, IL: Free Press Messerschmidt J. 1995. From patriarchy to

gender: feminist theory, criminology, and the challenge of diversity. In International Feminist Perspective inCriminology:

Engendering a Discipline, ed. NH Rafter, F Heidensohn. pp. 167-88. Philadelphia: Open Univ. Press

References

Miller J. 2001. One of the Guys: Girls, Gangs, and Gender. New York: Oxford Univ. Press

References

Miller J. 2002. The girls in the gang. In Gangs in America III, ed. CR Huff, pp. 175-97. Thousand Oaks, CA: Sage
Miller J, Maxson CL, Klein MW, eds. 2001. The Modern Gang Reader. Los Angeles: Roxbury. 2nd ed.

Miller W. 1973. The molls. Society 11(1):32-35 Miller WB. 1958. Lower class culture as a generating milieu

of gang delinquency. J Soc. Issues 14(3):419-35

Miller WB. 1975. Violence by Youth Gangs and Youth Groups as a Crime Problem in Major American Cities. Washington,

DC: US Dept. Justice

References

Moore JW. 1978. Homeboys. Philadelphia: Temple Univ. Press

Moore JW. 1991. Going down to the barrio: homeboys and homegirls in change. Philadelphia: Temple Univ. Press

Moore JW, Hagedorn JM. 1996. What happens to girls in the gang? See Huff 1996, pp. 20518

References

OJJDP (Off. Juv. Justice Delinquency Prev.). 1996. Female Offenders in the Juvenile Justice System: Statistic Summary.

Washington, DC: US Dept. Justice

Oliver W. 2003. The Violent Social World of Black Men. San Francisco: Jossey Bass. Rev. ed. In press

Padilla F. 1992. The Gang as an American Enterprise. New Brunswick, NJ: Rutgers Univ. Press

References

Park RE, Burgess EW, McKenzie RO, eds. 1925. The City. Chicago: Univ. Chicago Press

References

Raine A, Brennan PA, Farrington DP. 1997. Biosocial bases of violence: conceptual and theoretical issues. InBiosocial

Bases of Violence, ed. A Raine, PA Brennan, DP Farrington, pp. 99-130. New York: Plenum

Reiss A Jr, Roth JA. 1993. Understanding and Preventing Violence. Washington, DC: Natl. Acad. Press

RohnerRP. 1976. Sex differences in aggression: phylogenetic and enculturation perspectives. Ethos 4:57-72

Sampson RJ, Laub JH. 1990. Crime and deviance over the life course: the salience of

References

adult social bonds. Am. Soc. Rev. 55:60927

References

Sampson RJ, Lauritsen JL. 1990. Deviant lifestyles, proximity to crime, and the offender-victim link in personal violence. J.

Res. Crime Delinquency 27(2):11039

References

Sanders WB. 1994. Gangbangs and Drive-bys: Grounded Culture and Juvenile Gang Violence. New York: Aldine de Gruyter

Seigal LJ, Senna JJ. 1991. Juvenile Delinquency. St. Paul, MN: West Publ. 4' ed.

References

Shakur S (a.ka. Monster Kody). 1993. Monster: the Autobiography of an LA. Gang Member. New York: Atl. Mon. Press

Shaw C, McKay R. 1942. Juvenile Delinquency and Urban Areas. Chicago: Univ. Chicago Press

References

Short J. 1996. Personal, gang, and community careers. See Huff 1996, pp. 3-11

Short JF Jr, Strodtbeck FL. 1965. Group Process and Gang Delinquency. Chicago: Univ. Chicago Press

Spergel IA, Curry GD. 1998. The national youth gang survey: a research and development process. See Miller et al. 2001,

pp. 25465

References

Sullivan M. 1988. Getting Paid. Ithaca: Cornell Univ. Press

Sutherland EH. 1934. Principles of Criminology. Chicago: Lippincott

Suttles G. 1968. Social Order of the Slum. Chicago: Univ. Chicago Press

Thornberry TP. 2001. Risk factors for gang membership. In The Modern Gang Researcher, ed. J Miller, CL Maxson, MW

Klein, pp. 32-43. Los Angeles: Roxbury. 2"nd ed.

References

Thrasher ft 1927/1963. The Gang. Chicago: Univ. Chicago Press

Tonry M, Moore MH. 1998. Youth Violence. Chicago: Univ. Chicago Press

Venkatesh SA. 1996. The gang in the community. See Huff 1996, pp. 241-56

Vigil JD. 1987. Street socialization, locura behavior, and violence among Chicano gang members. In Violence and

Homicide in

References

Hispanic Communities, ed. J Kraus, S Sorenson, PD Juarez, pp. 231-41. Washington, DC: Nati. Inst. Mental Health

References
Vigil JD. 1988a. Barrio Gangs: Street Life and Identity in Southern California. Austin: Univ. Texas Press

Vigil JD. 1988b. Group processes and street identity: adolescent Chicano gang members. Ethos 16(4):421-45

Vigil JD. 1996. Street baptism: chicano gang initiation. Hum. Organ. 55(2):149-53

Vigil JD. 1999. Streets and schools: how educators can help Chicano maraginalized gang youth. Harvard Educational Rev.

69(3):27088

References

Vigil JD. 2002a. A Rainbow of Gangs: Street Cultures in the Mega-City. Austin: Univ. Texas Press

Vigil JD. 2002b. Community dynamics and the rise of street gangs. In Latinos! Remaking America, ed. MM Suarez-Orozco,

MM Paez, pp. 97-109. Berkeley, CA: Univ. Calif. Press; Cambridge, MA: David Rockefeller Cent. Lat. Am. Stud., Harvard

Univ.

Vigil JD, Yun SC. 1998. Vietnamese youth

References

gangs in the context of multiple marginality and the Los Angeles youth gang phenomenon. In Gangs and Youth Subcultures:

International Explorations, ed. K Hazlehurst, C Hazlehurst, pp. 117-39. New Brunswick, NJ: Transaction

References

Vigil JD, Yun SC. 2002. A cross-cultural framework to understand gangs: multiple marginality and Los Angeles. See Huff

2002, pp. 161-74

Waters T. 1999. Crime and Immigrant Youth. Thousand Oaks, CA: Sage

Whyte WF 1943. Street Corner Society. Chicago: Univ. Chicago Press

Wilson WJ. 1987. The Truly Disadvantaged. Chicago: Univ. Chicago Press

Wolfgang ME, Ferracuti ft 1967. The Subculture of Violence. London: Tavistock Yablonsky L. 1966. The violent Gang. New

York: Macmillan

References

Zimring FE. 1996. Kids, guns, and homicide: policy notes on an age-specific epidemic. Law Contemp. Probl. 59:261-74

Zimring FE. 1998. Youth Violence. Oxford: Oxford Univ. Press

AuthorAffiliation

James Diego Vigil

 AuthorAffiliation

School of Social Ecology, University of California, Irvine, California 92697-7080; email: vigil@uci.edu

Copyright Annual Reviews, Inc. 2003

James, D. V. (2003). Urban violence and street gangs. Annual Review of Anthropology, 32, 225-242. Retrieved from

http://search.proquest.com/docview/199843684?accountid=37714

ANTISOCIAL BEHAVIOR AND YOUTH GANG

MEMBERSHIP: SELECTION AND SOCIALIZATION
Gordon, Rachel AAuthor Information ; Lahey, Benjamin BAuthor Information ; Kawai, Eriko; Loeber, RolfAuthor
Information ; et al. Criminology 42.1 (Feb 2004): 55-87.

1. Full text
2. Full text - PDF
3. Abstract/Details
Hide highlighting
Abstract
TranslateAbstract

We examine whether gang membership is associated with higher levels of delinquency because boys predisposed to

delinquent activity are more likely than others to join. We use 10 years of longitudinal data from 858 participants of the

Pittsburgh Youth Study to identify periods before, during and after gangmembership. We build on prior research by

controlling for ages and calendar time, by better accounting for gang memberships that occurred before the study began,

and by using fixed effects statistical models. We find more evidence than has been found in prior studies that boys

who join gangs are more delinquent before entering the gang than those who do not join. Even with such selective

differences, however, we replicate research showing that drug selling, drug use, violent behaviors and vandalism of property

increase significantly when a youth joins a gang. The delinquency of peers appears to be one mechanism of socialization.

These findings are clearest in youth self-reports, but are also evident in reports from parents and teachers on boys' behavior

and delinquency. Once we adjust for time trends, we find that the increase in delinquency is temporary, that delinquency

falls to pre-gang levels when boys leave gangs. [PUBLICATION ABSTRACT]

KEYWORDS: youth gangs, delinquency, fixed effects, random effects

Full text
TranslateFull text
Turn on search term navigation

Headnote

KEYWORDS: youth gangs, delinquency, fixed effects, random effects

We examine whether gang membership is associated with higher levels of delinquency because boys predisposed to

delinquent activity are more likely than others to join. We use 10 years of longitudinal data from 858 participants of the
Pittsburgh Youth Study to identify periods before, during and after gangmembership. We build on prior research by

controlling for ages and calendar time, by better accounting for gang memberships that occurred before the study began,

and by using fixed effects statistical models. We find more evidence than has been found in prior studies that boys

who join gangs are more delinquent before entering the gang than those who do not join. Even with such selective

differences, however, we replicate research showing that drug selling, drug use, violent behaviors and vandalism of property

increase significantly when a youth joins a gang. The delinquency of peers appears to be one mechanism of socialization.

These findings are clearest in youth self-reports, but are also evident in reports from parents and teachers on boys' behavior

and delinquency. Once we adjust for time trends, we find that the increase in delinquency is temporary, that delinquency

falls to pre-gang levels when boys leave gangs.

Gangs are a major social problem. In recent decades, for example, gang-related homicides have become a larger proportion

of the total. In Chicago, one study estimates the increase to be from less than 10 percent in1965 to more than 25

percent in 1994 (Howell, 1999). Concerns about gangs have also expanded geographically. Once seen as restricted to a

small number of the nation's largest cities, gang problems are now recognized in many mid-size cities and nonmetropolitan

areas (Howell, 1998). And, even though gangmembership during adolescence is short lived, there is a strong correlation

between it and delinquent behavior (Thornberry et al., 2003). Thus, even a brief spell of participation may expose a youth to

a substantially heightened risk of delinquency. To address the problem, it is important to know if the association

between gang membership and delinquency is causal. If so, what mechanisms operate? Does the delinquency associated

with gang membership persist after a boy's brief participation?

Gang researchers have speculated about these selection and socialization processes. For instance, selection may operate

when new gang members are more likely to be drawn from among boys1 who already engage inantisocial behaviors. On the

other hand, as gang members, boys may be exposed to new opportunities to participate in criminal activities. At the

extreme, if the participation of gang members in criminal activity is entirely explained by selection, then

eliminating gangs would lead to no reduction in crime. Rather than adopting one extreme, we follow earlier research that has

considered selective and socialization processes as complementary rather than competing (Thornberry, Lizotte, Krohn,

Farnworth and Jang, 1994; Thornberry, Krohn, Lizotte and ChardWierschem, 1993).

LITERATURE REVIEW AND RESEARCH QUESTIONS

PRIOR RESEARCH

Cross-sectional studies have documented that gang members have higher delinquency than nonmembers (Curry and

Spergel, 1992; Dukes, Martinez and Stein, 1997; Henry et al, 2001; Huff, 1996). Longitudinal studies of boys provide further

evidence that at least some aspects of delinquent behavior increase during the period of gang membership relative to the

period before, and that these antisocial behaviors then decline after the boy leaves the gang (Esbensen and Huizinga, 1993;

Thornberry et al., 1993; Thornberry, 1998). Thornberry and colleagues (1993) were the first to use longitudinal data to test

the selection versus socialization hypotheses. They found support for a social facilitation model, documenting higher
delinquency among gang members when the youth were active in the gang but no significant differences between nongang

youth and future or former gang members. Later research built on this study, replicating the heightened delinquency when

boys were active in gangs, but also finding some evidence of selection, with future gang members showing greater

delinquency than nongang youth in the year prior to entering a gang (Esbensen and Huizinga, 1993; Thornberry et al.,

2003). Recent studies also suggest that one aspect of delinquencydrug sales-persists after boys

exit gangs (see review in Thornberry, 1998).

This research made an important contribution to literature on gangs, being the first to examine gangmembers along a

developmental trajectory before, during and after gang membership. Still, limitations inthese initial studies can be addressed,

especially using additional waves of data now available from longitudinal studies of delinquent youth. In particular, the

earliest studies (Esbensen and Huizinga, 1993; Thornberry et al., 1993) relied on simple descriptive means, proportions and

individual offense rates, without adjusting for measured or unmeasured characteristics of boys that might bias the

association between gangmembership and delinquency. More recent research has used standard regression models to

adjust for measured controls and random effects models to adjust for unmeasured stable characteristics of boys

(Thornberry, 1998; Thornberry et al., 2003). We extend these recent developments by estimating fixed effects models to

adjust for unmeasured biases (see Method below). This approach allows us to better separate selection from socialization

effects.

In addition, prior studies have defined gang participation in relation to waves of the study rather than specific to the boys'

ages, and did not control for age. Because the periods before, during and after gang entry are correlated with age, age-

related changes in youth behavior might be mistaken for changes associated with periods of gang membership. That is, if a

delinquent behavior generally increases with age, and boys tend to be older when they participate in gangs, then failure to

control for age would lead to an overestimate of the effect of gangs. Likewise, if boys are older when they

are gang members, but a delinquent behavior generally decreases with age, then omitting an age control would understate

the gang effect. Similarly, these studies only roughly control for year-to-year time trends in delinquency, and, within a

longitudinal study, the periods of time when boys are active gang members or have left a gang tend to be later in calendar

time.

These studies also do not consider gang memberships that occur prior to the study. The boys analyzed ranged in age from

11 to 15 when the studies began. Prior analyses of the Pittsburgh Youth Study indicate that gangentry begins as early as

age 9 and that nearly 5 percent of all youth and 20 percent of African American youth have entered a gang by age 15

(Lahey et al., 1999). This has two potential consequences. First, boys whose only gang affiliation occurred before the study

began will be miscoded as never having been gang members. Such miscoding would bias against identifying selection

effects if these former gang members have higher delinquency levels than boys who truly never participate. Second, boys

who participate in gangs both before and during the study, and are not gang members at the first study wave, will be

miscoded as before gang, rather than after gang. To the extent that antisocial behavior is higher after than

before gang membership, this approach underestimates the socialization effect (any increase in delinquent activity from the
period before to during membership). Similar consequences are possible in some prior studies that combine two semi-

annual interviews into yearly measures, coding boys as being in a gang if they report being a member at either interview.

We deal with these issues by using retrospective reports of gang membership before the first study wave, measures

of gang membership and delinquency at each study wave, and controls for age, seasonal and calendar-time trends.

In longitudinal studies, the most examined mechanism through which gang participation might socialize boys into antisocial

activity appears in prior research that has considered boys' association with delinquent peers along with

their gang participation (Battin, Hill, Abbott, Catalano and Hawkins, 1998; Battin-Pearson, Thornberry, Hawkins and Krohn,

1998). Such research builds on the sizable body of literature documenting an association between affiliation with delinquent

peers and youth delinquency (Fergusson, Swain-Campbell and Horwood, 2002; Warr, 2002). We consider this potential

mechanism as well, though unlike prior longitudinal studies we link specific kinds of peer delinquent activity drug sales, drug

use, violent delinquency and property delinquency-to these same specific kinds of antisocial activity on the part of the study

youth. Such specificity may be important, given that prior research suggests that gang involvement has a greater influence

on drug sales and violent delinquency than property delinquency (Battin et al., 1998; Thornberry et al, 1993).

CURRENT STUDY

In some sense, the question of whether gang participation is associated with antisocial activity is a tautology. Two commonly

defined types of youth gangs are the street gang and drug gang, in both of which delinquency is requisite, in the former as

members protect their territory from outsiders and in the latter as members sell drugs and protect their territory from rivals

(Klein, 1997). Becoming a member of a gang may also bring the boy into a history of intragang and intergang disputes.

Simply by association, the boy may be a target against whom others may initiate violence (Howell, 1999). Yet, it can be

fruitful to consider why the strength of this association may vary (Warr, 2002 discusses similar and additional mechanisms

for peer influences on delinquency). Boys who join gangs that as a whole engage in more crimes would be expected to

show a greater increase in their own criminal activity. Once a part of the gang's social network, boys gain greater access to

any existing information within the network regarding when, where, and how to engage incriminal activity. The members of

the group can become role models and mentors to the boy in increasing his repertoire of criminal activity. In addition, as the

network members become a featured part of the boy's larger social network, the norms and approval of the group may

become increasingly relevant for shaping his behavior. Peers are therefore an important socializing agent within gangs.

Several selective processes require careful attention as researchers study how gang participation relates to boys' antisocial

activity. In the principal selective process related to gang entry already noted, boys with a tendency toward antisocial activity

are more likely to become gang members. Research has documented this relationship, showing that boys who demonstrate

higher levels of delinquency early in life or who are on a trajectory of increasing criminality are more likely to enter a gang.

For example, Lahey and colleagues (1999) documented such selection using a smaller number of participants and waves of

the Pittsburgh Youth Study than we analyze here. Selection out of gangs may be relevant as well. That is, boys who have

less of a tendency toward antisocial activity who do become members of delinquent gangs may be more likely to exit. Study
attrition may be a third source of selection. Prior research suggests that boys with more delinquent tendencies are more

likely to drop out of research studies (Thornberry, Bjerregaard and Miles, 1993). If this selective attrition is particularly true of

core, long-term gang members, then the extent to which gangs affect behavior may be overstated. In other words, these

boys may be those most attracted to gang membership because of the benefits of associated gang activities. If data were

observable for these attriters, less of an increase over previous behavior might be evident and they might be more likely to

continue with criminal activity after leaving the gang.

The present study attempts to deal with these issues. Our central research question is whether youths' delinquent activity

rises when they enter gangs and falls when they exit. Given that such an association is observed, our interest is in why it

occurs, with the potential for changes in the delinquent activity of the boy's peers as a central possible socialization

mechanism. Our analysis begins with a descriptive portrait of youths' behavior by gang status. To the extent that delinquent

activity and gang membership share common causes, however, this description provides a biased estimate of the causal

relationship between gang membership and delinquency. We thus use statistical models to comprehensively adjust for

biases associated with preexisting characteristics of boys. We also control for age, season and calendar year and account

for gang memberships that occurred before the study began.

DATA AND METHODOLOGY

SAMPLE

The Pittsburgh Youth Study (PYS) follows three samples representative of all boys who were enrolled in the first, fourth or

seventh grades (called the youngest, middle and oldest samples, respectively) in public schools in Pittsburgh,

Pennsylvania in 1987-1988 (see Loeber et al., 1998 for a detailed description of the PYS). The current study includes 858

boys, 503 from the youngest sample as well as 355 boys from the second cohort of the oldest sample. Because a key

question was omitted from the gang questionnaire that was administered to the first cohort of the oldest sample, those

participants are excluded from this study. The boys described themselves ethnically primarily as African American (487) or

white (351), with the remainder choosing other (10), Asian (6), Hispanic (3) or American Indian (1). At the initial interview

(Wave S), nearly all of the boys in the youngest sample were ages 6 or 7 (53 percent and 41 percent respectively); the

boys in the second cohort of the oldest sample were primarily ages 12 through 14 (35 percent age 12, 41 percent age 13, 21

percent age 14). The latest available interview point for the current paper is in the 10th year of follow-up, when the

boys in the youngest sample were ages 16-17 and the boys in the oldest sample 22-24. Initially, about 85 percent of boys

agreed to participate in the study, with subject retention being above 80 percent in each follow-up wave.

The PYS was designed to provide a sizable sample of boys who engage in serious antisocial behavior and to maintain a

comparison group of nonantisocial boys. Based on an initial screening assessment completed by parents, teachers and the

youth themselves, boys were divided into two groups thought to be at high and low risk for offending. High-risk boys were

oversampled, and sampling weights are available to make estimates representative of first, fourth and seventh graders from

the Pittsburgh public schools in the late 1980s. Although estimating descriptive statistics that represent the population
requires weighted analyses, unweighted regression models are unbiased and efficient when models are correctly specified

(that is, in this context, a correctly specified model does not omit relevant predictors, including potential interactions between

predictor variables and sample stratification variables; DuMouchel and Duncan, 1983). We present unweighted models.

Results are highly consistent in weighted models (details available from the authors).

As expected for an urban public school sample, the boys' families had relatively limited economic resources. When the study

began, the median family income in 1998 dollars was $16,500. The majority of primary caregivers in the study had either not

completed high school (18 percent) or had completed high school but had received no further education (49 percent). And,

at the time of the first interview, 49 percent of the boys lived in single-parent families, 37 percent lived with both biological

parents and 14 percent lived with one biological parent and one nonbiological parent figure.

Pittsburgh is an example of a city with a recently emergent gang problem, in contrast to cities like Chicago and Los Angeles

with their longer traditions of gang activity and research. Gang activity and homicides inPittsburgh escalated in the early

1990s, peaking just prior to the middle of the decade, and then falling through the late 1990s (Copeland, 2001; Kelly and

Ove, 1999; Mamula, 1997). Consistent with these reports, our data show a peak in gang activity around 1993-1994.

DATA COLLECTION WAVES

As shown in Appendix A, data available for this paper covered a 10year period including fifteen waves of data for the

youngest sample and fourteen for the oldest sample. Initially, both grade samples were reassessed at 6-month intervals.

However, due to budgetary reasons, in the third and fourth years of the study the follow-up intervals were increased to 1

year, with the youngest and oldest grade samples alternating in the field. The first wave was the initial screening

assessment (denoted Wave S following the terminology of the study) and later waves are denoted alphabetically from A

(with SS used to denote the wave between R and T).

MEASURES

ASSESSMENT OF GANG MEMBERSHIP

Scholars debate the definition of gangs (Ball and Curry, 1995). Similar to other recent longitudinal studies of

youth gangs and delinquency, we assess gang membership using boys' self-reports. Little research has been conducted on

the validity of self-reported gang membership. However, a recent study of Chicago youth found that though self-reports

of gang membership do not coincide exactly with police identification of gangmembers, self-reported gang involvement does

correlate with official records of delinquency, controlling for boys' race/ethnicity and their own and official reports of prior

delinquency (Curry, 2000).

In the Pittsburgh Youth Study, data were collected about gang membership for the first time in Wave D for the oldest sample

(when the boys were ages 14-16) and in Wave H for the youngest sample (when the boys were ages 10-11). At these

waves, boys were asked if they were currently, or had ever been, a member of a gang. Follow-up questions collected
information about the boy's age at gang entry and exit and characteristics of the gang. During later waves, the boys were

asked about membership in gangs since the last interview with follow-up questions asking about the gang's characteristics

and whether the boy was still a member at the time of the interview.

We use the gang questionnaire to categorize boys into four gang status groups at each study wave: (1) those who had not

entered a gang by the last observation point (Wave T or U) and those who had entered a gang by the last observation point

and the current study wave is (2) before, (3) during or (4) after gang participation. Those who had entered and exited

a gang before the first study wave were coded as "after" gang participation at the first interview. When a boy missed any

interview after the first gang assessment point, we have no information about his gang status for the time between the prior

interview and the missed interview. This means that if a boy dropped out of the study, joined and exited a first gang, and

then rejoined the study, we would mistakenly identify him as not yet entering a gang. We thus censor boys at the first point

of missing data. If the boy was interviewed at later waves, those data are excluded from our models. Results are highly

consistent when we include all available waves of data (details available from the authors).

ANTISOCIAL BEHAVIOR MEASURES: PEERS

The Peer Delinquency Scale (PDS; Loeber et al., 1998) elicits information about the antisocial behavior of the boy's close

friends. The PDS has high internal consistency and has been related to the youth's level of delinquency (Loeber et al, 1998).

Each boy reported how many of his friends engaged in particular antisocial behaviors (twelve behaviors for the youngest

sample before Wave G, fifteeen for all other waves and for the oldest sample). We define four types of peer delinquency,

relating to drug selling, drug use, violent delinquency and property delinquency, drawing on behaviors that were asked about

across all waves (see Appendix B). The response structure changed across time for the youngest sample (from "none, one,

some or all" of the boy's friends to "none, few, half, most or all" of the boy's friends). For consistency across waves and with

the oldest sample, we collapse responses into three categories (0=none of the boy's friends, 2=all of the boy's friends, 1=all

other responses).

ANTISOCIAL BEHAVIOR MEASURES: YOUTH

Boys in the oldest sample completed the forty-item Self-Reported Delinquency Scale (SRD; Elliot, Huizinga and Ageton,

1985). The SRD has been shown to be reliable and has been validated against official crime records (Huizinga and Elliott,

1983). For the first seven waves of data collected from the youngest sample, the study staff developed an ageadapted

measure, the thirty-three-item Self-Reported Antisocial Behavior Scale (SRA; Loeber et al., 1998). In later waves, the SRD

was administered to the youngest sample. In the SRA, boys reported whether they engaged in each activity 0=never, 1-

once, 2=twice, or 3=more than twice in the reference period. In the SRD, boys reported the exact number of times they

engaged in the activity in the reference period. For consistency, we downcode the SRD reports at 3 (that is, if a boy reported

on the SRD that he engaged in the activity 4 or more times in the reference period, the value was receded to 3). We also

convert all reports to 6-month frequencies (by dividing reports for 1-year reference periods in half before top coding).

Appendix B lists our four measures of the boy's self-reported antisocial behaviors, which are consistent with the peer
measures as well as prior research (Thornberry et al, 1993). We also define a measure of aggression to cover all available

waves of SRA/SRD data because only a subset of the items we use to measure violent delinquency are available in the

SRA.

Adapted versions of Achenbach's widely used measures of youth problem behaviors were also administered to the youth

(Youth SelfReport, YSR), his caregiver (Child Behavior Checklist, CBCL) and his teacher (Teacher Report Form, TRF;

Achenbach, 1991a, 1991b, 1991c). Each informant rated the 112 items on the checklist, plus several additional items added

for the study, as 0=not true, 1=somewhat true, or 2=very true of the boy's behavior since the last interview. In the current

paper, we define four categories of delinquent activity to parallel the PDS and SRD measures (see Appendix B).

ANTISOCIAL BEHAVIOR: GANG

Boys were asked separately open-ended questions about the activities of each gang as well as closed-ended questions

about the activities of their most recent gang. Their responses to these questions were used to create delinquency variables

parallel to those created for youth and peer delinquency. In particular, drug sales was based on three dichotomous codes

tapping pushing or selling drugs. Drug use included three codes about doing drugs, getting high, and drinking. Violent

delinquency included ten codes regarding fighting, using weapons in fights, providing protection, robbing or strongarming

people, having forced sex, terrorizing people, attacking someone to hurt them, and shooting or killing people. Property

delinquency included seven codes including robbery, stealing, vandalism and joyriding.

STATISTICAL MODELS

We are interested in the degree to which being a gang member is associated with high levels of delinquency, above and

beyond the levels of delinquency that we would expect because boys predisposed to delinquent activity may be more likely

to enter, and stay in, gangs. One way to address this question would be to control for measured differences between boys

who do and do not participate in gangs. However, any study has only a limited number of such controls. We therefore use

statistical models to adjust for preexisting characteristics of boys, both those that are measured in the study and those that

are not.

By separating gang members along the pre-gang, during-gang, and postgang timelines we partially adjust for selective

differences between boys who do and do not enter gangs. This approach is limited, however. The pregang category is

restricted to the desired set of boys who will later enter gangs, but the during-gangcategory mixes boys who will

leave gangs and boys who will stay in gangs. Thus, the contrast of the post-gang to the during-gang category may be

inaccurate. In addition, even the simple during-gang to pre-gangcontrast will only match the fixed effects approach

(described below) when we restrict the estimation sample to the waves immediately before and after a boy enters a gang.

Two major alternative models are available to adjust for unmeasured preexisting differences: random effects and fixed

effects (Wooldridge, 2002). Both separate the error term from a standard regression model into person-specific and time-

specific components. The first differs across individuals but remains the same within each individual across time. The
second differs across both individuals and time. In the random effects approach, the person-specific component is modeled

as a draw from a distribution. In the most common fixed effects approach, the person-specific component is modeled with a

dummy variable for each person in the sample. These dummy variables can be estimated because each person has at least

two data points. Intuitively, the coefficients for these dummy variables allow the intercept to shift up or down depending on

the collective effect of any characteristics of the boy that are stable across time. These controls remove between-person

variance, thus the analyses examine change within an individual. (Indeed, for two waves of data, the fixed effects model is

identical to a first difference model that predicts change in the outcome variable based on change in the predictor variable).

The precision with which the person-specific effects are estimated increases when more time points are available for each

person and estimates are more precise when more change occurs on the variables of interest. The fact that our person-

specific effects are based on at most ten waves of data, and relatively few boys are observed to enter and exit gangs, will

reduce precision, making it harder to detect significant effects.

The random effects and fixed effects models both have additional advantages and limitations. A major limitation of the fixed

effects model is that time-constant characteristics of individuals cannot be included as predictors because they are collinear

with the person dummy variables. In our case, this limitation prevents us from using the fixed effects model to examine

selection by contrasting boys who never join gangs with boys who will later join gangs (because boys in the "never gang"

category do not change gang status across the study waves). Still, it does allow us to examine socialization by contrasting

boys who do join gangs across time, before, during and after they enter gangs. The random effects model has a different

limitation: It makes the assumption that the person-specific component of the error term is uncorrelated with the predictor

variables included in the model. This assumption is unlikely to hold. We expect that variables we do not measure that may

predict delinquency, such as a genetic tendency toward thrill-seeking or abuse and neglect during early childhood, are also

correlated with gang status. This expectation is precisely why we desire a statistical adjustment for these unmeasured

variables. The fixed effects model allows for such correlation between the person-specific error component and the predictor

variables. The Hausman specification test evaluates this assumption (Wooldridge, 2002). We focus on the fixed effects

approach but also summarize results from random effects models.

An important advantage of both models is that the person-specific characteristics they adjust need not be measured in the

study. Such characteristics would include stable demographic characteristics, such as the boy's race, ethnicity and his

parents' age and educational level when he was born. They would also include the boy's experiences prior to the initial study

wave-such as changes in his parents' marital status or his exposure to violence and abuse during early childhood - because

these remain constant during the study period. They would also include any stable biological or genetic characteristics. Both

models, however, are limited by the fact that they do not adjust for characteristics that change over time but are unmeasured

(for example, change in local policing policies, entry of a new gang into the neighborhood). Our models control for seasonal,

calendar-year and age-related trends to capture some of these time-varying correlates.

Our outcomes are sums of items from youth and peer delinquency scales. We thus estimate negative binomial regression

models (Long, 1997). The negative binomial regression model is appropriate for count outcomes that are skewed right,
improving on the Poisson regression model by allowing the conditional variance to exceed the conditional mean. Strictly

speaking, the negative binomial (and Poisson) apply to counts of discrete events (for example, the number of times the boy

carried a hidden weapon). Although our outcomes are instead sums of small-range (for example, none to three) items

capturing different events (carried a hidden weapon, hit people, threw rocks or bottles at people and so on), the negative

binomial is expected to perform well because of the positively skewed, integer nature of the outcomes. The negative

binomial regression model also has the advantage of a simple factor change interpretation: We can interpret a unit

change in the predictor variable as affecting the expected count by a factor equal to the exponential of the coefficient.

We estimate negative binomial models with and without fixed effects, referring to the latter as "standard" models. The

process by which fixed effects estimates are calculated differs in the negative binomial regression case in contrast to the

least squares approach, with important implications for the precision of the estimates. In particular, the fixed effect negative

binomial regression model is estimated as a conditional fixed effect, with the joint probability of the boy's delinquency scores

at each individual wave being conditional on the probability of his total delinquency across the waves (Hausman, Hall and

Griliches, 1984). The conditioning occurs by placing the sum of each boy's outcome variable across waves in the

denominator. As a result, boys with all zeros on the outcome variable do not contribute to the log likelihood (that is, are

excluded from the estimation sample). This reduced sample size can result in large standard errors. The omission of many

cases also raises intuitive concerns that estimates may be biased.

Ordinary least squares, while less desirable than negative binomial regression in terms of functional form, provides an

alternative estimation approach in which cases without variation on the outcome variable can contribute to the fixed effects

estimates and in which the properties of the estimates do not require large samples. The problem of heteroscedasticity

common to least squares estimation with count outcomes can be dealt with by using the general Huber-White correction for

heteroscedasticity (Wooldridge, 2002). In addition, Allison and Waterman (2002) recently critiqued the conditional fixed

effects negative binomial regression model and evaluated an alternative unconditional fixed effects negative binomial

regression model. We focus on the standard and conditional fixed effects negative binomial regression models in the tables

and text, and comment on the consistency of results based on ordinary least squares and unconditional negative binomial

fixed effects models.

We estimate the unconditional fixed effects negative binomial regression model using SAS 8.2. All other results

presented in this paper were calculated using Stata 7.

RESULTS

GANG PARTICIPATION

Among the study participants, 165 boys were observed to participate in gangs during the course of the study-134

participated in a single gang, 26 in two gangs and 5 in three gangs). Most spent only a brief period inthe gang. Nearly 85

percent were observed to exit the first gang they entered. And nearly three-quarters who exited were observed to
be in the gang for only one or two waves of the study (48 percent one wave, 25 percent two waves). Boys self-

defined gangs and reported a range of gang activities. At the wave when they entered the gang, nearly all boys (99 percent)

reported that the gang they entered engaged in violent delinquency. Of the other three delinquent activities, 59 percent

reported drug selling, 51 percent property delinquency and 21 percent drug use. Fully 12 percent of boys reported that

their gang engaged in all four of the activities and 35 percent reported three of the four activities.

Enlarge this image.

Figure I. A descriptive portrait of youth-reported activity by gang

Average levels of delinquency by boys' gang status for each of the seven self-reported measures are presented in Figure 1.

Within each chart, a selection effect is evident if the second bar is higher than the first, indicating higher levels of

delinquency among boys who will later join gangs than among boys who do not join gangs ("before gang" versus

"never gang"). Socialization is evident if boys report greater delinquent activity when they are gang members (third bar) than

before (second bar) or after (fourth bar) they participate in a gang. If this socialization effect is temporary, then levels of

delinquency "after gang" should return to "before gang" levels. Persistent delinquency following gang participation would

indicate a more permanent change, such as through the boy's maintenance of a new peer networks or taste for delinquency.

The statistical significance of these differences was tested by estimating a standard negative binomial regression

model in which three dummy variables indicate the four gang statuses (data not shown). The charted means in Figure 1 are

predictions from these models.

In these descriptive results, there is evidence of selection for violent and property delinquency, but not for drug sales and

drug use. That is, the average level of drug-related delinquency of future gang members does not differ significantly from the

average level of drug-related delinquency among boys who never join gangs(compare the "never gang" and "before gang"

bars in the top row of Figure 1). In contrast, boys who will later join gangs exhibit significantly higher levels of violent and

property delinquency even before they enter a gang than boys who never join gangs.

For all outcomes we also see a socialization effect of gang participation. In all seven charts in Figure 1, the third bar

"during gang" is significantly higher than the second bar "before gang." This indicates that, among gang members, drug

selling, drug use, violent delinquency and property delinquency all increase on average during the period of

active gang membership relative to the period before gang entry. For violent and property delinquency, this elevation is

temporary. Once the boy exits the gang, his delinquency returns to statistical equivalence with (or falls significantly below)

pre-gang levels (compare "after gang" and "before gang" bars in second through fourth rows of Figure 1). In contrast, drug

selling and drug use both persist after the boy exits the gang. That is, the "after gang" level of drug-related delinquency

remains significantly higher than the "before gang" level in the two charts in the top row of Figure 1.

FIXED EFFECTS MODELS OF YOUTH-REPORTED DELINQUENT ACTIVITY

We present negative binomial models predicting both boys' drug sales and drug use (see Table 1) and boys' violent and

property delinquency (see Table 2). We summarize both standard and fixed effects negative binomial models to highlight the

way our results change when we adjust for preexisting characteristics of the boys in the latter models. We account for the

clustering of waves within boys in the standard model using the Huber-White approach (Wooldridge, 2002).

These models include three dummy indicators to capture the four categories of gang status ("never" in a gang;

ever in a gang and at a wave "before," "during" or "after" participation). We then conduct additional tests to estimate four key

contrasts.2 The first contrast ("before versus never") captures selection: Do boys who will later join gangs ("before") report

higher levels of delinquency than boys who do not join gangs ("never"). The next contrasts capture socialization. Does a

boy's delinquency increase when he enters a gang ("during versus before") and decrease when he exits ("during versus

after")? The final contrast further refines the evidence of socialization: Does a boy's delinquency decline sufficiently after he

exits a gang such that it returns to his pre-gang level of delinquency ("after versus before")? Because these hypotheses are

directional, one-sided significance levels are reported. All of these contrasts are conducted such that the coefficient will be

positive if consistent with the hypothesis.

Enlarge this image.

Table 1. Standard and Fixed Effects Models of Boys' Self-Reported Drug Sales and Drug Use

Enlarge this image.

Table 2. Standard (STD) and Fixed Effects Models of Boys' Self-Reported Aggression, Violent Delinquency and Property Delinquency

All the models presented also control for age, seasonality and calendar-time trends through cohort-specific dummy variables

for study wave and linear and quadratic terms for the boy's age in months. The use of wave dummies, rather than a linear

measure of calendar time, allows for the relationship between time and delinquency to take on an arbitrary shape. For

example, a general increase over time may take on a stair-step shape if increases happen in one season and

plateau inanother. Sudden spikes associated with particular events are also allowed. By being cohort-specific, the wave

dummy variables allow these relationships to differ for the older and younger samples. Within cohort and wave, boys'

ages inmonths vary. The additive and quadratic terms capture any additional time-trends associated with increasing age.

The most consistent finding is that evidence of socialization remains once we control for time-trends (standard models) and

for time-trends and preexisting characteristics (fixed effects models). For all outcomes, the "during versus before" contrast is

significantly positive, indicating that boys report more delinquency when they are active in the gang than they did before they

entered the gang, above and beyond general age-, seasonal- and time-trends in the data and above and beyond preexisting

characteristics of boys that are correlated both with gang entry and delinquency (see "during versus before" rows in Tables 1

and 2). For example, when boys are active ingangs they engage in over 3 times more drug selling than they did before they

entered the gang(exp[1.14]=3.13).

We also see that the evidence of selection for violent and property delinquency remains when we control for time-

trends in the standard negative binomial regression model (see significant "before versus never" contrasts in the "standard"

columns of Table 2). However, adjusting for time-trends reveals selection for drug-related delinquency not evident in the

simple descriptive means (see Table 1). This is because drug-sales and drug-use increase with age, not peaking until late

adolescence, and generally increase with calendar time. Because boys who are "before gang" are younger and at earlier

study waves, their drug-related delinquency appears similar to the average drug-related delinquency across all study waves

of the "never gang" youth until we adjust for these time-related trends. In contrast, violent and property delinquency

generally decline across age and calendar time. Thus, the "before versus never" coefficients in Table 2 for violent and

property crimes, though still significant, are somewhat smaller than those that do not control for time-trends (the coefficients,

not shown, underlying Figure 1).

The findings for the persistence of delinquency after a gang exit also differ in the multivariate versus the descriptive models.

Whereas there was evidence of persistence in drug-related delinquency in the descriptive statistics displayed in Figure 1,

this is not the case in Table 1. The "during versus after" contrasts in Table 1 are positive and significant, indicating that boys

report more drug-related delinquency while they are active members than they do after they leave the gang. In fact the

decrease in drug-related delinquency upon exit is fairly comparable to the increase on entry: For drug sales, boys report an

increase of just over 3 times (exp[1.14]) when they enter a gang and a decline of just under 3 times (exp[1.07]) after they

leave the gang. Indeed, the "after versus before" contrasts for drug sales and drug use in Table 1 are not significant,

indicating that the boys' level of drug-selling and drug use have declined enough to equal the levels they had reported

before they entered gangs. Similar to the bias described above for the selection ("before versus never") findings, this reflects

the fact that the "after gang" periods occur when boys are older and later in calendar time when general time-trends reveal

greater drug-related delinquency.

A similar bias is evident for aggression and violent delinquency, though in the opposite direction. Whereas the descriptive

means in Figure 1 revealed no persistence in violent delinquency after a gang exit, the standard models that adjust for time-

trends do indicate persistence (significant "after versus before" contrasts in the "standard" columns of Table 2). We also see

the greatest difference between the standard and fixed effects models for violent delinquency, indicating a biasing effect of

failure to control for preexisting characteristics of boys that are correlated with both gang participation and violence. The

"after versus before" contrasts are significant for violent delinquency in the standard but generally not the fixed effects

models. Although the "during versus before" contrasts for violent delinquency are significant in both models, they also reflect

this bias by being smaller-one-half to one-quarter the size-in the fixed effects versus the standard models. This indicates that

the increase in violent delinquency upon gang entry is overestimated before we account for unmeasured preexisting

characteristics of boys. Once these confounds are controlled, the increase and decrease in violence associated

with gang entry and exit are fairly comparable, leaving little evidence of persistence once the boy leaves a gang.

Although replicating the generally strong association between current gang participation and elevated delinquent activity,

these findings differ somewhat from prior research. Prior longitudinal studies have found that this association is strongest for
violent delinquency and that drug-related activity is most likely to persist after a gang exit (see Thornberry, 1998 and

Thornberry et al, 2003). Our data are more consistent with these prior studies before we add controls for time trends. That

is, the post-gang periods are later in calendar time and age when drug selling and drug use had increased and violent and

property delinquency were lower among Pittsburgh youth. We also see more evidence of selection into gangs than have

prior studies, perhaps due to accounting for gang memberships that occurred before the study began as well as due to

adjusting for these general time trends.

EXPLORING SOCIALIZATION MECHANISMS

We estimated additional models to examine whether peer delinquency (the proportion of the boys' friends who commit

delinquent acts) is a potential socialization mechanism. Table 3 presents fixed effects models that predict peer drug sales,

drug use, violent delinquency and property delinquency, controlling for age-, seasonal- and calendar-time trends and

preexisting characteristics of boys. The pattern of significant effects generally mirrors the findings for the boys' own activities.

All four outcomes increase significantly when the boy enters a gang ("during versus before"). All also significantly decline

after an exit ("during versus after"), although there is some evidence of persistence in peers' drug use and violent

delinquency ("after versus before").

These findings suggest that changes in peer delinquency may be one mechanism through which boys' own delinquency

changes when they participate in gangs. We thus added peer delinquency to the fixed effects models predicting youth

delinquency within domain (for example, controlling for peer drug sales when predicting youth drug sales). The socialization

coefficients are uniformly reduced suggesting that peer delinquency is a partial mechanism (compare Table 4 with the "fixed

effects" columns in Tables 1 and 2).3

Enlarge this image.

Table 3. Fixed Effects Models of Peer Delinquency

In particular, the coefficients measuring the increase indelinquency associated with gang entry ("during versus before") are

between 13 percent and 46 percent less. Additionally, the coefficients measuring the decrease in delinquency associated

with gang exit ("during versus after") are between 13 percent and 65 percent less. With the exception of Youth Self-Report

property delinquency, however, all of these coefficients remain significant, suggesting that a mechanism in addition to peer

delinquency operates. Inaddition, generally, peer delinquency explains the gang-entry increase and the gang-exit decrease

comparably enough such that the "after versus before" coefficients continue to be insignificant. The exception is with

aggression. For this outcome, we capture more of the mechanism of a decline in aggression upon leaving a gang than the

mechanism for a rise in aggression upon joining a gang, thus the "after versus before" coefficient is significant when peer

delinquency is controlled.

CONSISTENCY OF FINDINGS IN ALTERNATIVE MODELS

We reestimated the fixed effects models from Tables 1 and 2 using OLS rather than a negative binomial specification and

found that the results of twenty of the twenty-one significance tests are the same. (The exception is that none of the "after

versus before" contrasts are significant in OLS, whereas one is just significant in the negative binomial specification).

Enlarge this image.

Table 4. Fixed Effects Models of Boys' Self-Reported Delinquency Controlling for Peer Delinquency

We also reestimated the models with the unconditional, rather than conditional, fixed effects negative binomial regression

model. These results are consistent with those reported inTables 1 and 2 although they reveal more evidence of

persistence in delinquency (the "after versus before" contrast is significant at one-sided p<.05 for drug sales and onesided

p<.025 for aggression) and, in terms of property delinquency, the results for Youth Self-Report are more similar to the results

for SelfReported Delinquency (in the unconditional model of Youth Self-Report, one-sided p<.025 for "during versus before"

and one-sided p<.05 for "during versus after").

Finally, we reestimated the models with random rather than fixed effects and tested the random effects assumption. The

random effects models, like the fixed effects models, show evidence of socialization. The "during versus before" and "during

versus after" contrasts are significant for all outcomes. However, specification tests suggest that the random effects model is

not adequate for this data. Before we control for time variation by age, season and year, the random effects assumption is

rejected by the Hausman test for all outcomes. When we add controls for time-trends, the assumption is rejected for violent

delinquency (SelfReported Delinquency and Youth Self-Report; details available from the authors). In addition, if the random

effects models adequately adjust for preexisting characteristics of boys, then gang boys should not differ from nongang

boys in these models before they enter gangs. In the language of our tables, the "before versus never" contrast should be

insignificant. However, we find that these contrasts remain significant for all outcomes in the random effects models.

PARENT AND TEACHER REPORTS

Our analyses to this point have been based entirely on youth selfreports. Thus we cannot distinguish

whether gang membership is associated with actual changes in behavior or with other factors such as changes in the social

desirability the boy perceives for reporting delinquent activity when he is involved in a gang. To help examine this issue, we

reestimated the models using parent and teacher reports of the boys' delinquent activities. As with youth self-reports, the

findings for these informants are subject to various interpretations. However, generally, whereas boys' reports might

overestimate gang effects, parent and teacher reports might underestimate them (for example, we might see less of an

association between gang status and delinquent activity because parents and teachers are less aware of the boys'

activities). In addition, because parents and teachers were not interviewed for the latter half of the study period among the

oldest sample, we lack data for the late adolescent and young adult period from these informants. Significant findings based

on parent and teacher reports might therefore be interpreted as strong evidence of socialization pathways.
We estimated the fixed effects models for parents reports (Table 5) and for teacher reports (Table 6) twice, once without

controls for peer delinquency and once with controls for peer delinquency. We begin with the models without peer

delinquency to identify any evidence of socialization and then examine whether peer delinquency might be a socialization

mechanism. Both parents and teachers report that youths' property delinquency increases when they enter a gang and that

it remains elevated after they exit the gang(significant "during versus before" and "after versus before" coefficients in the

"property delinquency/without peer controls" columns). Both parents and teachers also report an elevation in youths' violent

delinquency, though parents appear to see it sooner. In particular, parents report an increase associated with the

boys' gang entry (significant "during versus before" coefficient in the "violent delinquency/without peer controls"

column in Table 5). But teachers do not report such an increase until after the boy has exited the gang(marginally

significantly negative "during versus after" coefficient and significantly positive "after versus before" coefficient in the "violent

delinquency/without peer controls" column of Table 6). Neither parents nor teachers report an association

between gang participation and drug use. Regarding mechanisms of socialization, parents' reports of violent delinquency

increasing when boys join a gang is most explained by youth reports of changes in peers' violent delinquency (see "during

before before" coefficient in "violent delinquency/with peer controls" column of Table 5), The associations with property

delinquency decrease slightly for parent reports and unexpectedly increase for teacher reports once youthreported peer

delinquency is controlled (see "during versus before" and "after versus before" coefficients in "property delinquency/with

peer controls" columns of Tables 5 and 6).

Enlarge this image.

Table 6. Fixed Effects Models of Boys' Delinquent Activity: Teacher Reports

DISCUSSION

SUMMARY OF FINDINGS

We examined the degree to which being a gang member is associated with high levels of delinquency, above and beyond

expected levels (given that boys predisposed to delinquent activity may be more likely to enter and stay in gangs). We built

on prior research by controlling for boys' age and calendar time, by accounting for gang participation before the study began

and by using fixed effects statistical models to comprehensively adjust for all preexisting characteristics of boys. We found

more evidence than prior studies have that boys who join gangs are more delinquent before entering the gang than those

who do not join. Even with such selective differences, however, we replicate prior findings of a substantial increase in drug

selling, drug use, violent delinquency and property delinquency when boys are active gang members. Additional analyses

suggest that the mechanisms through which gangs have their effects include changes in delinquency of the boys' peers.

That is, peer delinquency is elevated during periods of gang memberships and explains a portion of the increase in boys'

delinquency when they enter gangs and the decrease when they exit.

In initial descriptive associations, we replicate prior evidence of drugselling persisting after boys exit gangs. However, this

association is explained by periods after gang membership being correlated with later years when drug-selling had generally
increased among Pittsburgh youth. After we adjust for age, season, calendar year and preexisting characteristics of boys,

we find little or no evidence of persistence in drugrelated, property or violent delinquency after a boy exits a gang.

Associations with parent and teacher reports of boys' delinquent activity further substantiate these findings. Both indicate

increases in boys' property delinquency when they are gang members. Parents, and to a lesser extent teachers, also report

increased violent delinquency, and these associations are explained by youth reports of changes in his peers' violent

criminality.

CONCLUSIONS AND LIMITATIONS

The fixed effects models we employ do not adjust for time varying covariates associated with both delinquency

and gang membership. Thus, we cannot rule out the fact that some third variable predicts both gang participation and

increased delinquency in a particular time period, rather than there being a causal effect of gangs. Our models do adjust for

any such variables correlated with the boys' age, with season and year, or with peer delinquency. Future research

on gangs should focus on measuring potential time-varying confounds and causal mechanisms. Although not using a fixed

effects approach, Thornberry and colleagues updated their earlier models with controls for family poverty level, parental

supervision, commitment to school, negative life events, youth violence and delinquent peers in the year prior

to gang membership. They still find a strong positive association between a dummy variable for current gang membership

and self-reported violence with these time-varying controls (Thornberry, 1998; Thornberry et al., 2003).

In this study, we saw that peer delinquency was elevated during a period of gang membership and explained some of the

association between gang membership and youth delinquency. One way to illuminate this finding in future research would

be to use more detailed data about boys' peer networks. Such studies could examine, for example, whether some

youth join gangs simultaneously with existing friends, and both show elevations in delinquency, or whether some youth add

new delinquent friends, and eliminate prior non-delinquent friendships, during periods of gang membership. Likewise, such

studies could examine how this mechanism operates upon gang exit: Do boys stop associating with delinquent youth, and

return to old friends, or do they and their long-term friends drop out of gangs simultaneously? Another fruitful direction for

understanding this mechanisms would be to continue research that has examined whether there is a different effect of

associating with delinquent peers inside versus outside of a gang structure (for example, Battin et al., 1998; Thornberry et

al., 2003).

We should emphasize that there is also an issue of causal direction. Does a decrease in peer delinquency lead to a

decrease in the boy's delinquency, or does a decrease in the boy's delinquency lead to a decline in his peers' delinquency?

The direction suggested above is sensible, but the other is plausible as well. For example, a boy might exit a gang and

reduce his own delinquency because the opportunities for delinquency are now diminished. Because his delinquency is

reduced, his associations with delinquent boys may decline. Given that some effect of gang membership remains, even after

controlling for peer delinquency, future research should also examine additional mechanisms. For example, Hagedorn
(1994) emphasized the importance of using organizational theories to understand gangs and to recognize heteroegeneity

across gangs in their structures.

Our results do not provide evidence of persistence in the boy's delinquency after he leaves a gang, including drug selling

that had been shown to persist in prior studies. The fact that drug selling does not remain elevated afterwards, once time

trends are controlled, should not minimize the importance of its elevation among active gang members. The increased use

of custodial sentences for drug offenses means that even a brief increase in drug selling during a short period

of gang membership could be substantively meaningful (Blumstein and Beck, 1999). In addition, boys report that increased

violence and peer drug use persist after they leave gangs, indicating that some effects of gang participation may not be

transient. Of course, our failure to find evidence of persistence is susceptible to a Type II error. This potential is amplified by

the fact that relatively few boys are observed to exit a gang more than once. Additional waves of data might provide more

post-gang data points and identify additional gang exits. Understanding why many boys leave gangs, while some stay, is

also an important direction for future research, which would also benefit from larger samples of gang exits.

We find evidence of socialization effects of gangs based on parent and teacher reports, though significant coefficients are

seen for fewer outcomes and contrasts than is the case for youth reports. We cannot distinguish whether the stronger

evidence in self-reports reflects that youths are more aware of their own behavior than parents or teachers, or that they

overstate their delinquency during times when they say they are gang members. Future research might further examine the

degree to which associations hold across multiple informants. For example, though peers may have their own biases (for

example, susceptibility to rumors), they may have greater information about adolescents' activity than would parents and

teachers. Along these lines, recent research in Canada found that classmates reported gang members to be more

aggressive than nongang youth (Craig, Vitaro, Gagnon and Tremblay, 2002).

Using more comprehensive controls for preexisting characteristics of boys as well as controls for age, season and calendar

year, this study further substantiates earlier research on a substantial elevation in delinquency when boys are

active gang members. Limitations notwithstanding, these results reinforce the need for interventions aimed at delinquent

youth gangs.

Footnote

1. Although participation of girls in gangs is an important topic of study, our data source in this paper comprises boys

exclusively. Some of the theoretical ideas laid out may apply to girls as well, although empirical studies with girls would be

required to examine the degree to which mechanisms operate similarly for boys and girls.

2. The additional two contrasts can be calculated based on the presented coefficients. "During versus never" is the sum of

the "during versus before" and "before versus never" coefficients. "After versus never" is the sum of the "after versus before"

and "before versus never" coefficients. All of these coefficients are positive and significant for the models reported in Tables

1 and 2. These coefficients are not presented in the tables because they combine selection and socialization, and thus are

of lesser theoretical interest.

3. The sample size is reduced when we control for peer delinquency due to missing values on these variables. We

estimated the models in Tables 1 and 2 on this reduced sample size to evaluate whether the difference incoefficients

between Table 4 and Tables 1 and 2 is merely due to differences in the samples. The percent reduction in coefficients is

highly similar when we use these new results: 27 percent to 48 percent reduction in the "during versus before" coefficients

and 13 percent to 65 percent reduction in the "during versus after" coefficients when we control for peer delinquency.

References

REFERENCES

Achenbach, Thomas M.

1991a Manual for the Child Behavior Checklist/4-18 and 1991 Profile. Burlington: University of Vermont, Department of

Psychiatry.

1991b Manual for the Teacher's Report Form and 1991 Profile. Burlington: University of Vermont, Department of Psychiatry.

1991c Manual for the Youth Self-Report and 1991 Profile. Burlington: University of Vermont, Department of Psychiatry.

Allison, Paul D. and Richard P. Waterman

2002 Fixed-effects negative binomial regression models. Sociological Methodology 32:247-265.

Ball, Richard A. and G. David Curry

1995 The logic of definition in criminology: Purposes and methods for defining gangs. Criminology 33:225-245.

Battin, Sara R., Karl G. Hill, Robert D. Abbott, Richard F. Catalano and J. David Hawkins

1998 The contribution of gang membership to delinquency beyond delinquent peers. Criminology 36:93-115.

Battin-Pearson, Sara R., Terence P. Thornberry, J.David Hawkins and Marvin D. Krohn

1998. Gang Membership, Delinquent Peers, and Delinquent Behavior. OJJDP Juvenile Justice Bulletin (October). U.S.

Department of Justice, Office of Justice Programs, Office of Juvenile Justice and Delinquency Prevention.

Blumstein, Alfred and Allen J. Beck

1999 Population growth in U.S. prisons, 1980-1996. In Michael Tonry and Joan Petersilia (eds.), Crime and Justice: Prisons,

Vol. 26. Chicago: University of Chicago Press.

Copeland, Dave

2001 Explanations vary over year's increased rate of homicide. Pittsburgh Tribune-Review (July 30, 2001).
Curry, G. David

2000 Self-reported gang involvement and officially recorded delinquency. Criminology 38:1253-1274.

Curry, G. David and Irving A. Spergel

1992 Gang involvement and delinquency among Hispanic and African-American adolescent males. Journal of

Research in Crime and Delinquency 29:273-291.

Craig, Wendy M., Frank Vitaro, Claude Gagnon and Richard E. Tremblay

2002 The road to gang membership: characteristics of gang and nongang members from ages 10 to 14. Social

Development 11:53-68.

Dukes, Richard L., Ruben O., and Judith A. Stein

1997 Precursors and consequences of membership in youth gangs. Youth and Society 29:139-165.

Elliott, Delbert S., David Huizinga and Suzanne S. Ageton

1985. Explaining delinquency and drug use. Beverly Hills, Calif.: Sage Publications.

Esbensen, Finn-Aage and David Huizinga

1993 Gangs, drugs, and delinquency in a survey of urban youth. Criminology 31:565-589.

Fergusson, David M., Nicola R. Swain-Campbell, and L. John Horwood

2002 Deviant peer affiliations, crime and substance use: a fixed effects regression analysis. Journal of Abnormal Child

Psychology, 30:419-430.

Hagedorn, John M.

1994 Neighborhoods, markets, and gang organization. Journal of Research in Crime and Delinquency 31:264-294.

Hausman, Jerry, Bronwyn H. Hall, and Zvi Griliches

1984 Econometric models for count data with an application to the patents-R&D relationship. Econometric 52:909-938.

Henry, David B., Patrick H. Tolan and Deborah Gorman-Smith

2001 Longitudinal family and peer group effects on violence and nonviolent delinquency. Journal of Clinical Child

Psychology 30:172-186.
Howell, James C.

1999 Youth gang homicides: a literature review. Crime and Delinquency 45:208-241.

1998 Youth Gangs: An Overiew. Juvenile Justice Bulletin (August) Washington, D.C.: U.S. Department of Justice, Office of

Juvenile Justice and Delinquency Prevention.

Huff, Ronald C.

1996 The criminal behavior of gang members and nongang at-risk youth. In C. Ronald Huff (ed.), Gangs inAmerica (2nd

edition). Thousand Oaks, Calif.: Sage Publications.

Huizinga, David and Delbert S. Elliott

1986. Reassessing the reliability and validity of self-report measures. Journal of Quantitative Criminology, 2, 293-327.

Kelly, Jack and Torsten Ove

1999 Youth violence found to spring from gangs. Pittsburgh PostGazette (Sunday, September 5).

Klein, Malcolm W.

1997 The American Street Gang: Its Nature, Prevalence, and Control. Oxford: Oxford University Press.

Lahey, Benjamin B., Rachel A. Gordon, Rolf Loeber, Magda Stouthamer-Loeber and David P. Farrington

1999 Boys who join gangs: a prospective study of predictors of first gang entry. Journal of Abnormal Child Psychology

27:261-276.

Loeber, Rolf, David P. Farrington, Magda Stouthamer-Loeber, and Willmet B. Van Kammen

1998 Antisocial Behavior and Mental Health: Explanatory Factors in Childhood and Adolescence. Mahwah, N.J.: Erlbaum.

Long, J. Scott

1997 Regression Models for Categorical and Limited Dependent Variables. Thousand Oaks, Calif.: Sage Publications.

Mamula, Kris B.

1997 Police alliance curbs murder, drugs, despair. Pittsburgh Tribune-Review (November 23, 1997).

Thornberry, Terence P.
1998 Membership in youth gangs and involvement in serious and violent offending. In Rolf Loeber and David Farrington

(eds.), Serious and Violent Juvenile Offenders: Risk Factors and Successful Interventions. Thousand Oaks, Calif.: Sage

Publications.

Thornberry, Terence P., B. Bjerregaard and William Miles

1993 The consequences of respondent attrition in panel studies: a simulation based on the Rochester Youth Development

Study. Journal of Quantitative Criminology 9:127-158.

Thornberry, Terence P., Marvin D. Krohn, Alan J. Lizotte and Deborah Chard-Wierschem

1993 The role of juvenile gangs in facilitating delinquent behavior. Journal of Research in Crime and Delinquency 30:55-87.

Thornberry, Terence P., Alan J. Lizotte, Marvin D. Krohn, Margaret Farnworth and Sung Joon Jang

1994 Delinquent peers, beliefs, and delinquent behavior: A longitudinal test of interactional theory. Criminology 32:47-83.

Thornberry, Terence P., Marvin D. Krohn, Alan J.Lizotte, Carolyn A. Smith and Kimberly Tobin

2003 Gangs and Delinquency in Developmental Perspective. Cambridge: Cambridge University Press.

Warr, Mark

2002 Companions in Crime. Cambridge: Cambridge University Press.

Wooldridge, Jeffrey M.

2002 Econometric Analysis of Cross Section and Panel Data. Cambridge: MIT Press.

AuthorAffiliation

RACHEL A. GORDON*

University of Illinois at Chicago

BENJAMIN B. LAHEY

University of Chicago

ERIKO KAWAI

Chicago, IL

ROLF LOEBER & MAGDA STOUTHAMER-LOEBER

 University of Pittsburgh

DAVID P. FARRINGTON

Cambridge University

Gordon, R. A., Lahey, B. B., Kawai, E., Loeber, R., & al, e. (2004). ANTISOCIAL BEHAVIOR AND YOUTH GANG MEMBERSHIP:

SELECTION AND SOCIALIZATION. Criminology, 42(1), 55-87. Retrieved from

http://search.proquest.com/docview/220708441?accountid=37714

Long-Term Consequences of Adolescent Gang Membership for

Adult Functioning
Gilman, Amanda BAuthor Information ; Hill, Karl GAuthor Information ; Hawkins, J DavidAuthor Information
. American Journal of Public Health 104.5 (May 2014): 938-45.

1. Full text
2. Full text - PDF
3. Abstract/Details
4. References 51
Hide highlighting
Abstract
TranslateAbstract

We examined the possible public health consequences of adolescent gang membership for adult functioning. Data were

drawn from the Seattle Social Development Project, a longitudinal study focusing on the development of positive and

problem outcomes. Using propensity score matching and logistic regression analyses, we assessed the effects

of adolescent gang membership on illegal behavior, educational and occupational attainment, and physical and mental

health at the ages of 27, 30, and 33 years. In comparison with their nongang peers, who had been matched on 23

confounding risk variables known to be related to selection into gang membership, those who had joined

a gang in adolescence had poorer outcomes inmultiple areas of adult functioning, including higher rates of selfreported

crime, receipt of illegal income, incarceration, drug abuse or dependence, poor general health, and welfare receipt and lower

rates of high school graduation. The finding that adolescent gang membership has significant consequences in adulthood

beyond criminal behavior indicates the public health importance of the development of effective gangprevention programs.

Full Text
TranslateFull text

Headnote

Objectives. We examined the possible public health consequences of adolescent gang membership for adult functioning.
Methods. Data were drawn from the Seattle Social Development Project, a longitudinal study focusing on the development

of positive and problem outcomes. Using propensity score matching and logistic regression analyses, we assessed the

effects of adolescent gang membership on illegal behavior, educational and occupational attainment, and physical and

mental health at the ages of 27, 30, and 33 years.

Results. In comparison with their nongang peers, who had been matched on 23 confounding risk variables known to be

related to selection into gang membership, those who had joined a gang in adolescence had poorer outcomes in multiple

areas of adult functioning, including higher rates of selfreported crime, receipt of illegal income, incarceration, drug abuse or

dependence, poor general health, and welfare receipt and lower rates of high school graduation.

Conclusions. The finding that adolescent gang membership has significant consequences in adulthood beyond criminal

behavior indicates the public health importance of the development of effective gangprevention programs. (Am J Public

Health. 2014;104:938-945. doi:10.2105/AJPH.2013.301821)

Research has consistently shown that gang membership has proximal adverse conse- quences during adolescence

that, in turn, lead to significant social and economic costs.1-3 Active gang members are much more likely than their nongang

peers to engage in criminal behavior,2,4,5 especially serious and violent offending6; in addition, they are more likely to be

involved in drug use and selling,5 have more difficulties inschool,4 and are more likely to be violently victimized.3 However,

with few exceptions, there is scant research on possible broader, long-term public health consequences

of gang membership.7-9

Levitt and Vankatesh,8 in a 10-year follow- up of a sample of Chicago youths, found that those who reported

being in a gang during adolescence were more likely to be arrested and incarcerated as adults, were more likely to rely on

illegal income, and had obtained less formal education than their nongang peers. When they controlled for background

charac- teristics such as home environment and early school performance, only the relationship

between adolescent gang membership and ille- gal income in adulthood remained significant.

In an ethnographic study, Hagedorn10 con- ducted follow-up interviews of 228 Milwaukee- based founding

male gang members several years after his initial study. Two thirds of the sample did not have a high school diploma or

general equivalency diploma (GED), fewer than 32% were employed, and more than 63% had been incarcerated.

In a quantitative study involving a commu- nity sample, Thornberry et al.9 examined the extent to which gangmembership

negatively affected the timeliness of developmental tran- sitions into adulthood, such as completion of schooling and

establishment of a career. They found that, after control for individual and environmental risks, male respondents who had

been short-term or long-term gang mem- bers in adolescence had a greater likelihood of cohabitation before marriage at the

age of 22 years than male respondents who had not been gang members. Long-term gang members also had significantly

higher rates of unstable employment, school dropout, early pregnancy, and teenage parenthood at the follow-up. Fe-

male gangmembers were more likely to expe- rience untimely or problematic transitions, including unstable employment,
teenage preg- nancy, and early motherhood. Both male and female former gang members were more likely to report adult

arrests.

Recently, Krohn et al.11 found that there was an indirect positive relationship between adolescent gangmembership and

engagement in street crime and arrest in adulthood. Specifically, gang membership was related to precocious transitions

into adulthood that then predicted disrupted family relationships and economic instability. They found that this path

eventually led to criminal behaviors at the age of 30 years. Inour study, we extended these investigations by assessing the

effects of gang membership on adult functioning holistically, examining possible later adult outcomes.

Life course theory12-14 provides a frame- work to understand how gang membership in adolescence may affect illegal

behavior, edu- cational and occupational attainment, and physical and mental health inadulthood. This theory emphasizes

the strong connection be- tween childhood events and experiences inadulthood (trajectories), as well as significant events

that create a disruption in a trajectory (turning points). Although we did not directly test a trajectory change model in this

study, life course theory would suggest that joining a gang during adolescence may initiate a negative developmental

cascade into both criminal and noncriminal domains, including decreased edu- cational and occupational attainment and

poor physical and mental health.15,16 Indeed, Melde and Esbensen reported that "youth who join gangs experience

noteworthy changes in their emotions, attitudes, and behavior,"17(p539) sug- gesting that gang membership may serve as

asignifi cant turning point in an individual's life course.

One such life course theory, the social development model,18 provides specific mechanisms through which these

consequences may operate. This model articulates the mechanisms of social- ization and identifies parallel but separate

causal paths for prosocial and antisocial processes con- sisting of opportunities for involvement, actual involvement, skills,

rewards, bonding, and adop- tion of beliefs. Participation in the prosocial path is seen to increase subsequent positive

outcomes and decrease risk behaviors. By contrast, partic- ipation in the antisocial path is seen to decrease positive

outcomes and increase problem be- haviors. From the perspective of this model, gang membership in adolescence may

serve as a turning point, drastically changing the opportunity structure for young people. As gang-involved youths move

through subsequent cycles of socialization, they are likely to experi- ence reduced prosocial functioning across sev- eral

domains and escalated problem behaviors as they transition into adulthood, even if they are no longer members of a gang.

According to Krohn and Thornberry,

it is reasonable to expect that being a member of a gang during adolescence will be associated with disrupted transitions

from adolescence to adult- hood and, ultimately, will adversely impact life chances.7(p149)

They noted that the challenge is to discover empirically to what extent gang membership contributes to negative

outcomes in adulthood over and above general delinquency and other risk factors in adolescence. This was precisely the

aim of our study.

METHODS
We used longitudinal data from the Seattle Social Development Project (SSDP), which in- cludes a multiethnic sample of

808 fifth-grade students who in 1985 were attending 18 elementary schools serving high-crime neigh- borhoods of Seattle.

Participants have been followed prospectively into adulthood. Of the 808 students, 396 (49%) were female, 381 (47.2%)

were White, 207 (25.6%) were Afri- can American, 177 (21.9%) were Asian American, and 43 (5.3%) were Native Ameri-

can. Of the total sample, about 5% self-identified as Hispanic. More than half of the sample (52%) had participated in free or

reduced-price school lunch programs at some point in the fifth, sixth, or seventh grade, indicating that they came from low-

income families.

Data were obtained from multiple sources, including the youths, their parents or adult caretakers, teachers, and school

records. Sur- vey data were first collected in 1985, when most of the participants were 10 years of age (mean = 10.3, SD =

0.52); data were then col- lected in the spring of each year through 10th grade and again in12th grade. As adults, the

participants have been interviewed every 3 years. Outcomes were assessed at the ages of 27, 30, and 33 years (retention

rates at these follow-up points were 95%, 93%, and 92%, respectively).

Measures

Gang membership. Gang membership was measured prospectively from ages 13 to 18 years with the question "Do you

belong to a gang? " followed by "What is the name of the gang?" (to distinguish gangs from informal peer

groups). In addition, the survey administered in adulthood included a retrospective gangmembership item ("Have you ever

belonged to a gang? "). There were some inconsistencies in reporting across time. Sensitivity analyses revealed that those

who ever reported that they had joined a gang, even if they responded negatively at a different time, were more similar to

those who consistently reported being in a gang than those who were never in a gang with respect to demographic and risk

factors. Therefore, respondents were classified as having joined a gang (and assigned a code of 1) if they ever reported

having done so, either prospectively or retrospectively. This approach yielded a value of gang membership (a code of 0 or 1)

for all study respondents.

Control variables. We used 23 control vari- ables (as shown in Table 1), identified both empirically and theoretically as

influencing self-selection into gangs, to calculate propensity scores and match the gang and nongang groups. Included were

21 variables in 5 do- mains (individual, family, school, peer, and neighborhood) that have consistently been shown in this

sample and others to be pre- dictive of gang membership,19 -23 as well as demographic variables also known to be re-

lated to gang membership. All control variables were measured as an average of the fifth- and sixth-grade responses, before

the age of gang membership onset for the majority of those who joined (see Hill et al.21 for a full description of how each of

these variables was operation- alized).

We included an additional control for each of the models predicting general health, de- pression, and anxiety to further

ensure that any observed relationships between gang mem- bership and these adult outcomes were not caused by the

presence of the outcome vari- ables (i.e., health, depression, anxiety) prior to ganginvolvement. For each of the other

outcomes (e.g., illegal behavior, educational achievement, substance use), pregang levels of these variables (or close
proxies) were included in the propensity score given that they were hypothesized to predict selection into adoles-

cent gang membership. However, the same was not true for general health, depression, and anxiety. Therefore, the model

predicting poor general health in adulthood included a retro- spective assessment of lifetime poor health, measured at the

age of 21 years, as an additional control. Individuals who answered yes to the question "Would you say you have been

sickly alarge part of your life?" were assigned a code of 1, and all other responses were coded as 0. Early levels of anxiety

and depression (measured in fifthgrade)wereassessedwithitemsfromthe Teacher Report Form of the Child Behavior

Checklist.24

Adult outcomes. We investigated outcomes in 3 broad domains relevant to public health (assessed in 2002, 2005, and 2008,

when re- spondents were aged 27, 30, and 33 years, respectively): illegal behavior, educational and occupational

attainment, and physical and mental health. If respondents endorsed an outcome at age 27, 30, or 33 years, they were

assigned a code of 1; otherwise, they were assigned a code of 0.

With respect to illegal behavior, self- reported crime was a measure of whether the participant reported committing at least

one property or violent offense in the preceding year at age 27, 30, or 33 years. Receipt of illegal income was a measure of

whether respondents reported receiving income from drug sales or other illegal activity in the pre- ceding year (coded as 1)

or did not report receiving such income (coded as 0). Similarly, incarceration was assessed by asking individ- uals whether

they had spent any time in jail or prison in the preceding year (affirmative responses were coded as 1, and all other

responses were coded as 0).

High school graduation (not including GED) was self-reported and coded as 1 (and coded as 0 otherwise). We chose to

exclude GED as an indicator of high school graduation because research has shown that those who graduate from high

school have significantly better life outcomes than those who take an equivalency exam.25 Welfare receipt was a measure

of receipt of public assistance (coded as 1) in the past year at age 27, 30, or 33 years.

We used a self-reported measure of general health. This variable was dichotomized into excellent or good health (coded as

0) and fair or poor health (coded as 1) at ages 27, 30, and 33 years. Mental health was assessed with vari- ables indicating

whether participants met the criteria of the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition26 (DSM-IV)

for major depression or generalized anxiety disorder (coded as 1 if yes and 0 otherwise) at age 27, 30, or 33 years.

Similarly, participants were asked a series of questions about their alcohol and drug use. Participants were assigned a code

of 1 if, during the preceding year, they met DSM-IV26 criteria for either alcohol abuse or dependence at age 27, 30, or 33

years, and they were assigned a code of 0 if they did not meet these criteria. The same coding scheme was used for drug

abuse or dependence.

Data Analysis

Our goal was to determine whether there are unique long-term consequences of adolescent gangmembership for adult

functioning over and above the effects of early childhood risk factors associated with self-selection into the gang lifestyle.
One of the greatest challenges researchers face when examining the effects of a "treatment " received by a group

(e.g., gang membership) is to adequately control for all of the confounding variables that might in- fluence self-selection into

the treatment group. Yanovitzky et al.27 noted that, in the absence of a randomized controlled trial, one of the most effective

approaches to addressing this selection bias is the use of a propensity score analysis.28 Thus, we used propensity score

matching to ensure that the only variable that distinguished the treatment group (i.e., those who had joined a gang in adoles-

cence) from the control group was gang membership itself. The psmatch229 module in Stata (StataCorp LP, College

Station, TX) was used in conducting our analyses.

First, we used multiple imputation to ac- count for missing data in the sample. All vari- ables used in the analyses were

included in the multiple imputation model. The average rate of missing data across all variables was 5.2% (the variables with

the highest percentages of missing data were poor refusal skills and family structure, which were part of the 23-item

propensity score).

Second, we calculated a propensity score, which can be described as the conditional probability of receiving the

"treatment" in each of the 40 imputed data sets for the entire sample according to our 23 control variables. A nearest

neighbor without replacement matching strategy was used to match the 173 participants who had joined a gang in adoles-

cence with 173 participants who had never joined a gang but showed a similar propensity for doing so. Only these 346

individuals were included in subsequent analyses. We tried several matching strategies and selected this approach because

it optimized balance and statistical power while sufficiently reducing bias. An important advantage of propensity score

analyses over multivariate regression analyses is the ability to ensure that the 2 comparison groups are adequately

balanced. That is, in our propensity score analysis, we were able to show that there were no statisti- cally significant

differences between the 2 groups on any of the 23 control variables included (Table 1).

Finally, a series of logistic regression models were estimated in which those who had joined a gang were compared with

those who had not on several measures of adult functioning. As recommended by Graham,30 results from the imputed data

set were combined to esti- mate unbiased parameters and standard errors. Our hypotheses were

that gang membership in adolescence would predict higher rates of illegal behavior, lower educational and occu- pational

attainment, and poorer physical and mental health in adulthood.

RESULTS

Within the SSDP sample, 21.4% of the respondents (n = 173) reported ever having joined a gang. The average age at which

respondents had joined a gang was 14.9 years. No individual reported joining a gang after the age of 19 years (Figure 1).

Prevalence rates and means for each of the control variables among gang and nongang respondents both before and after

propensity score matching are shown in Table 1.

Nationwide, a disproportionate number of gang members are male and members of ethnic minority groups.31 This

disproportionality was also reflected in the SSDP data. Almost 42% of the gang members in this sample were African
American. By contrast, only 21% of those who had never joined a gang were African American (Table 1). Twenty-nine per-

cent of those who had joined a gang were White, 20% were Asian American , and 9% were Native American. Nearly 76% of

those who joined were male, while only 44% of those who had never joined were male. All control variables were related

to gang membership in the expected direction at the P = .1 level. After matching, there were no statistically significant

differences (at the P = .05 level) between groups according to race/ethnicity, gender, or any of the other 21 variables used to

calculate the propensity score (Table 1).

An additional criterion for assessing balance between matched samples is to calculate the standardized mean differences

between the groups. Austin reported that "there is no threshold for standardized differences that has been uniformly

accepted as indicative of meaningful imbalance,"32(p1210) but some re- searchers claim that a difference of less than 10%

indicates a significant reduction in bias.27,33 After matching, only 4 of the 23 variables showed differences above 10%

(absolute values ranged from 10.27% to 12.50%). Because these differences were low (most were well under the

recommended threshold) and were not statistically significant, we believed that we had established an ade- quate control

group.34

Bivariate prematching associations (Table 2, first column) revealed that adolescent gang membership significantly predicted

poorer functioning across all 3 domains in adulthood at the zero-order level: risk taking and illegal behavior (including self-

reported crime, receipt of illegal income, and incarceration), educa- tional and occupational attainment (including high school

graduation and welfare receipt), and physical and mental health (including general health, alcohol abuse or dependence,

and drug abuse or dependence). The remaining 2 indicators of adult functioning (anxiety and depression) were elevated

among those who joined a gang inadolescence relative to non- gang members, but the differences did not achieve statistical

significance. The second column of Table 2 shows the results from logistic regressions estimated with the matched samples.

Illegal Behavior

Adolescent gang membership significantly predicted self-reported crime, receipt of illegal income, and incarceration even

after control- ling for salient risk factors via propensity score matching. Specifically, incomparison with those who had never

joined a gang, those who reported joining a gang in adolescence were (at age 27, 30, or 33 years) nearly 3 times as likely to

report committing a crime in the preceding year, 3.66 times more likely to report receiving income from illegal sources, and

2.37 times more likely to have spent time incarcerated in the preceding year.

Educational and Occupational Attainment

Adolescent gang membership significantly predicted lower rates of high school gradua- tion; those who had joined

a gang in adoles- cence were about half as likely to graduate from high school as those who had never joined a gang but

shared similar risk back- grounds. The association with welfare receipt was also marginally significant (P = .05). Those who

joined a gang were about 1.7 times more likely to be receiving public assistance in adulthood.

Physical and Mental Health

Finally, adolescent gang membership pre- dicted poorer health and mental health in adulthood. Specifically, those who had

joined a gang in adolescence were about 1.7 times more likely to report poor health at age 27, 30, or 33 years than those

who had never joined a gang. They were also nearly 3 times more likely to meet the criteria for drug abuse or

dependence in the preceding year.

Continued Gang Membership in Adulthood

Gang membership was largely an adolescent phenomenon (Figure 1). Sensitivity analyses examining whether the 17

respondents who reported current gang membership at the age of 27 years differed from those who had left gangs revealed

comparable patterns of re- sponses in the 2 groups, with one exception. Analyses excluding the 17 current gang mem- bers

showed that the relationship between adolescent gangmembership and educational attainment was no longer statistically

signifi- cant.

DISCUSSION

Our analyses show that adolescent gang membership has a significant impact on adult functioning across all 3 of the

domains exam- ined in this study. These effects remained even after controlling for individual, family, peer, school, and

neighborhood characteristics. Certain outcomes, including a higher propen- sity for illegal behaviors as well as lower rates of

educational and occupational attainment, are not surprising given the well-established relationship

between gang membership and criminal behavior and lower involvement inconventional institutions. However, less expected

is the significant relationship between adolescent gangmembership and poor physical and mental

health in adulthood. In fact, the results of this study confirm the assertions of life course theory and the idea of develop-

mental cascades discussed earlier. Our analyses elucidate the role that gang membership may play in propelling youths

down a path resulting in negative consequences that appear to "cascade" into 3 broad areas of adult functioning.

Our study is not the first to highlight a relationship between antisocial behavior and poor health outcomes.35-38 This link is

probably a result of higher rates of risk taking in addition to lower rates of activities that promote better health among those

who engage in antisocial behavior.39 Our analyses indicate that adolescent gang membership predicts a stressful and risky

life course well into early adulthood. Future etiological research should focus on whether such out- comes vary according to

sociodemographic characteristics (e.g., race/ethnicity and gender), as well as whether individual, family, peer, school, and

neighborhood risks contribute to these outcomes independently of or in inter- action with gang membership.

Study Limitations

Our data were derived from a Seattle com- munity sample of young people who were adolescents in the 1980s and early

1990s. Caution should be applied when generalizing our results to other geographic areas. Each community and, indeed,

each gang has its own distinct set of variables that may affect long- term consequences. Given the life course the- ory

framework described earlier, however, we believe these processes to be general and find no reason to expect that the

relationships between gang membership and adult func- tioning reported here would differ dramatically in other places.
It also should be noted that our data were self-reported. It is possible that gang member- ship itself was underreported or

overreported in our sample. However, self-reported data have been used extensively in gangstudies to

determine gang membership,14,40-46 and Esbensen et al.47 determined that self- identification is a highly reliable indicator

of gang membership.

Implications for Public Health

To help communities engage in prevention activities effectively, it is important to under- stand the role joining a gang may

play in affecting an adolescent's life course. Indeed, as early as 1993, when the discipline of pre- vention science was only

beginning to gain national attention, Coie et al.48(p1017) noted the need for life course models that "extend their sights

beyond the acute onset of disorder and take account of the whole life cycle, working backward and forward from the initial

clinical outcome."32(p1210) Researchers and the lay community often consider gang membership only in the context of

adolescence and, fre- quently, only interms of involvement in crime and delinquency. We have found

that adolescent gang membership has long-term consequences that extend beyond criminal activity, indicating

that gang membership may have implications for public health beyond public safety.

Our findings suggest that effective gang prevention efforts may result not only in reductions in adolescentproblem behavior

but also in higher adult functioning across multiple domains. It is our hope that the results of this study will provide motivation

for pre- vention scientists to develop, implement, and test effective programs to prevent young people from joining gangs. j

References

References

1. Dukes RL, Martinez RO, Stein JA. Precursors and consequences of membership in youth gangs. Youth Soc.

1997;29(2):139-165.

2. Howell JC. The Impact of Gangs on Communities. Washington, DC: US Department of Justice, Office of Juvenile Justice

and Delinquency Prevention; 2006.

3. Peterson D, Taylor T, Esbensen F-A. Gang mem- bership and violent victimization. Justice Q. 2004;21(4): 793-815.

4. Decker SH, Van Winkle B. Life in the Gang: Family, Friends and Violence . Cambridge, UK: Cambridge Uni- versity

Press; 1996.

5. Hill KG, Lui C, Hawkins JD. Early precursors of gang membership: a study of Seattle youth. Available at:

https://www.ncjrs.gov/pdffiles1/ojjdp/190106.pdf. Accessed January 6, 2014.

6. Huff CR. Comparing the Criminal Behavior of Youth Gangs and At-Risk Youths. Washington, DC: US Depart- ment of

Justice, National Institute of Justice; 1998.

7. Krohn MD, Thornberry TP. Longitudinal perspec- tives on adolescent street gangs. In: Liberman A, ed. The Long View of

Crime: A Synthesis of Longitudinal Research. New York, NY: Springer; 2008:128-160.

8. Levitt SD, Vankatesh SA. An Analysis of the Long- Run Consequences of Gang Involvement. Cambridge, MA: Harvard

University; 2001.

9. Thornberry TP, Krohn MD, Lizotte AJ, Smith CA, Tobin K. Gangs and Delinquency in Developmental Perspec- tive.

Cambridge, UK: Cambridge University Press; 2003.

10. Hagedorn J. People and Folks: Gangs, Crime, and the Underclass in a Rustbelt City. 2nd ed. Chicago, IL: Lake View

Press; 1998.

11. Krohn MD, Ward JT, Thornberry TP, Lizotte AJ, Chu R. The cascading effects of adolescent gang in- volvement across

the life course. Criminology . 2011; 49(4):991-1028.

12. Elder GH. Perspectives on the life course. In: Elder GH, ed. Life Course Dynamics: Trajectories and Transi- tions, 1968-

1980. Ithaca, NY: Cornell University Press; 1985:23-49.

13. Gotlib IH, Wheaton B. Trajectories and turning points over the life course: concepts and themes. In: Gotlib IH, Wheaton

B, eds. Stress and Adversity Over the Life Course: Trajectories and Turning Points . Cambridge, UK: Cambridge University

Press; 1997:1-25.

14. Sampson RJ, Laub JH. Crime and deviance in the life course. Annu Rev Sociol. 1992;18:63-84.

15. Masten AS, Cicchetti D. Developmental cascades. Dev Psychopathol. 2010;22(3):491-495.

16. McLeod JD, Almazan EP. Connections between childhood and adulthood. In: Mortimer JT, Shanahan MJ, eds.

Handbook of the Life Course. New York, NY: Springer; 2003:391-411.

17. Melde C, Esbensen F-A. Gang membership as a turning point in the life course. Criminology . 2011; 49(2):513-552.

18. Catalano RF, Hawkins JD. The social development model: a theory of antisocial behavior. In: Hawkins JD, ed.

Delinquency and Crime: Current Theories. New York, NY: Cambridge University Press; 1996:149-197.

19. Dishion TJ, Veronneau MH, Myers MW. Cascading peer dynamics underlying the progression from problem behavior to

violence in early to late adolescence. Dev Psychopathol. 2010;22(3):603-619.

20. Esbensen F-A, Huizinga D. Gangs, drugs, and delinquency in a survey of urban youth. Criminology. 1993;31(4):565-

589.

21. Hill KG, Howell JC, Hawkins JD, Battin Pearson SR. Childhood risk factors for adolescent gangmembership: results from

the Seattle Social Development Project. JRes Crime Delinq. 1999;36(3):300-322.

22. Huizinga D, Weiher AW, Espiritu R, Esbensen FA. Delinquency and crime: some highlights from the Denver Youth

Survey. In: Thornberry TP, Krohn MD, eds. Taking Stock of Delinquency: An Overview of Findings From Contemporary

Longitudinal Studies . New York, NY: Kluwer Academic/Plenum Publishers; 2003:47-91.

23. Lacourse E, Nagin DS, Vitaro F, Cote S, Arseneault L, Tremblay RE. Prediction of early-onset deviant peer group af

filiation: a 12-year longitudinal study. Arch Gen Psychiatry. 2006;63(5):562-568.

24. Achenbach TM, Edelbrock C. Manual for the Child Behavior Checklist and Revised Child Behavior Profile. Burlington,

VT: University of Vermont Press; 1983.

25. Cameron SV, Heckman JJ. The nonequivalence of high school equivalents. J Labor Econ. 1993;11(1):1-47.

26. Diagnostic and Statistical Manual of Mental Disor- ders, 4th edition. Washington, DC: American Psychiatric Association;

1994.

27. Yanovitzky I, Hornik R, Zanutto E. Estimating causal effects in observational studies: the propensity score approach. In:

Hayes AF, Slater MD, Snyder LB, eds. The Sage Sourcebook of Advanced Data Analysis Methods for Communication. Los

Angeles, CA: Sage Publications; 2008:159-184.

28. Rosenbaum PR, Rubin DB. The central role of the propensity score in observational studies for causal effects.

Biometrika. 1983;70(1):41-55.

29. Leuven E, Sianesi B. PSMATCH2: Stata Module to Perform Full Mahalanobis and Propensity Score Matching, Common

Support Graphing, and Covariate Imbalance Testing. Chestnut Hill, MA: Boston College Department of Economics; 2003.

30. Graham JW. Missing data analysis: making it work in the real world. Annu Rev Psychol. 2009;60:549-576.

31. National Gang Center. National Youth Gang Survey analysis. Available at: http://www.nationalgangcenter. gov/Survey-

Analysis. Accessed January 6, 2014.

32. Austin PC. Goodness-of-fit diagnostics for the pro- pensity score model when estimating treatment effects using

covariate adjustment with the propensity score. Pharmacoepidemiol Drug Saf . 2008;17(12):1202-1217.

33. Normand ST, Landrum NB, Guadagnoli E, et al. Validating recommendations for coronary angiography following acute

myocardial infarction in the elderly: a matched analysis using propensity scores. J Clin Epi- demiol. 2001;54(4):387-398.

34. Riou Franca L, Payet S, Le Lay K, Launois R. Drotrecogin alfa's impact on intensive care workload in real life practice: a

propensity score approach. Value Health. 2008;11(7):1051-1060.

35. Piquero AR, Daigle LE, Gibson C, Piquero NL, Tibbetts SG. Are life-course-persistent offenders at risk for adverse

health outcomes? J Res Crime Delinq . 2007;44(2):185-207.

36. Bardone AM, Moffitt TE, Caspi A, Dickson N, Stanton WR, Silva PA. Adult physical health outcomes of adolescent girls

with conduct disorder, depression, and anxiety. J Am Acad Child Adolesc Psychiatry. 1998; 37(6):594-601.

37. De Genna NM, Stack DM, Serbin LA, Ledingham JE, Schwartzman AE. From risky behavior to health risk: continuity

across two generations. J Dev Behav Pediatr. 2006;27(4):297-309.

38. Shepherd J, Farrington D. The impact of antisocial lifestyle on health: family, school, and police interven- tions can

reduce health risks. BMJ. 2003;326 (7394):834-835.

39. Odgers CL, Robins SJ, Russell MA. Morbidity and mortality risk among the "forgotten few " : why are girls in the justice

system in such poor health? Law Hum Behav. 2009;34(6):429-444.

40. Bjerregaard B, Smith C. Gender differences in gang participation, delinquency, and substance use. J Quant Criminol.

1993;9(4):329-355.

41. Esbensen F-A, Huizinga D, Weiher AW. Gang and non-gang youth: differences in explanatory factors. J Contemp Crim

Justice . 1993;9(2):94-116.

42. Hindelang MJ, Hirschi T, Weis JG. Measuring Delinquency . Beverly Hills, CA: Sage Publications; 1981.

43. Klein M. The American Street Gang: Its Nature, Prevalence, and Control. New York, NY: Oxford University Press; 1995.

44. Savitz L, Rosen L, Lalli M. Delinquency and gang membership as related to victimization. Victimology. 1980;5(2-4):152-

160.

45. Sullivan ML. Are "gang" studies dangerous? Youth violence, local context, and the problem of reification. In: Short JF,

Hughes LA, eds. Studying Youth Gangs. Lanham, MD: AltaMira Press; 2006:15-35.

46. Thornberry TP, Krohn MD, Lizotte AJ, Chard- Wierschem D. The role of juvenile gangs in facilitating delinquent

behavior. J Res Crime Delinq . 1993;30(1): 55-87.

47. Esbensen F-A, Peterson D, Taylor TJ, Freng A. Youth Violence: Sex and Race Differences in Offending, Victimi- zation,

and Gang Membership . Philadelphia, PA: Temple University Press; 2010.

48. Coie JD, Watt NF, West SG, . The science of prevention: a conceptual framework and some directions for a national

research program. Am Psychol. 1993; 48(10):1013-1022.

AuthorAffiliation

Amanda B. Gilman, MSW, Karl G. Hill, PhD, and J. David Hawkins, PhD

About the Authors

The authors are with the Social Development Research Group, School of Social Work, University of Washington, Seattle.

Correspondence should be sent to Amanda B. Gilman, MSW, PhD Candidate, Social Development Research Group,

University of Washington, 9725 3rd Ave NE, Suite 401, Seattle, WA 98115 (e-mail: abg5@uw.edu). Reprints can be ordered

at http://www.ajph.org by clicking on the "Reprints" link.

This article was accepted November 25, 2013.

Contributors

A. B. Gilman developed the research question, con- ducted the analyses, and led the writing of the article. K. G. Hill and J.

D. Hawkins contributed to study conceptu- alization, data collection, and the writing of the article.

Acknowledgments

This project was supported by the National Institute on Drug Abuse (grants R01DA003721, R01DA009679, and

R01DA024411-03-04), the Robert Wood Johnson Foundation (grant 21548), and the National Institute of Mental Health

(grant 5 T32 MH20010).

We are grateful to the anonymous reviewers as well as the editor for their valuable comments on this article. We also thank

the members of the Seattle Social Develop- ment Project (SSDP) analysis team for their input on substantive and

methodological issues and the data collection and data management teams for their tireless work. Finally, we thank the

SSDP respondents, whose continued participation makes this research possible.

Note. The content of this article is solely the re- sponsibility of the authors and does not necessarily represent the official

views of the funding agencies.

Human Participant Protection

All of the study procedures were approved by the University of Washington institutional review board. Written informed

consent was obtained from all study participants.

Word count: 5387

Copyright American Public Health Association May 2014

Gilman, A. B., Hill, K. G., & Hawkins, J. D. (2014). Long-term consequences of adolescent gang membership for adult

functioning. American Journal of Public Health, 104(5), 938-45. Retrieved from

http://search.proquest.com/docview/1524713313?accountid=37714

Being in "Bad" Company: Power Dependence and Status in

Adolescent Susceptibility to Peer Influence
 Vargas, RobertAuthor Information . Social Psychology Quarterly 74.3 (Sep 2011): 310-332.

1. Full text
2. Full text - PDF
3. Abstract/Details
4. References 68
Hide highlighting
Abstract
TranslateAbstract

Theories of susceptibility to peer influence have centered on the idea that lower status adolescents are likely to adopt the

behaviors of high status adolescents. While status is important, social exchange theorists have shown the value of analyzing

exchange relations between actors to understand differences in power. To build on status-based theories of peer influence,

this study analyzes power dependence relations in three adolescent friendship groups. Analyzing adolescent interaction as

social exchange showed how being in a group with balanced power relations insulated adolescents from peer influence,

even when some peers were delinquent or low academic achievers. In contrast, adolescents in groups with unbalanced

power relations were particularly susceptible to peer influence. This study presents an additional way to analyze the peer

influence process and illustrates the importance of applying social psychological theory to cases of micro inequality,

particularly in the context of small groups. [PUBLICATION ABSTRACT]

Full Text
TranslateFull text
Turn on search term navigation

Headnote

Abstract

Theories of susceptibility to peer influence have centered on the idea that lower status adolescents are likely to adopt the

behaviors of high status adolescents. While status is important, social exchange theorists have shown the value of analyzing

exchange relations between actors to understand differences in power. To build on status-based theories of peer influence,

this study analyzes power dependence relations in three adolescent friendship groups. Analyzing adolescent interaction as

social exchange showed how being in a group with balanced power relations insulated adolescents from peer influence,

even when some peers were delinquent or low academic achievers. In contrast, adolescents in groups with unbalanced

power relations were particularly susceptible to peer influence. This study presents an additional way to analyze the peer

influence process and illustrates the importance of applying social psychological theory to cases of micro inequality,

particularly in the context of small groups.

Keywords

peer effects, power dependence, small groups, youth

One of the most consistent findings in social science has been the correlation between the behaviors of adolescents and

their peers. Classic research such as Asch's (1951) laboratory studies of conformity and Coleman's (1966) research on

school composition sparked a vast literature on adolescent peer effects. While this literature established the importance of
peers for adolescent socialization, less is known about the peer influence process. Although scholars have identified factors

rendering adolescents susceptible to the influence of peers, explanations for what makes adolescents more or less

susceptible almost exclusively rely on status differences between adolescents. By status, I am referring to one's

standing in a social hierarchy as determined by respect or deference (Ridgeway and Walker 1995). Scholars have argued

that emulation of peer behavior is highly dependent upon an adolescent's standing in the social hierarchy of his or her peers,

with lower status individuals more likely to adopt the behaviors of higher status individuals (Cohen and Prinstein 2006;

Gerrard et al. 2008).

While status-based approaches have yielded important insights, research has shown that adolescent groups can be diverse

and may not map easily into groups with easily distinguishable status hierarchies. Adolescentfriendship groups can be

comprised of a mix of high/low academic achieving, high/low socially skilled, and delinquent/nondelinquent youth. Research

by Haynie (2001) emphasized the importance of accounting for this diversity in peer groups as 61 percent

of adolescents report friendships with delinquent peers. These complications illustrate the need for additional perspectives

on peer effect susceptibility that can account for being in diverse friendship groups.

In this article, I build on the literature on susceptibility to peer effects through an ethnographic analysis of power dependence

relations in three adolescent friendship groups. While status-based approaches focus on the characteristics of individuals, a

power dependence approach focuses on characteristics of social relations between individuals. In this study, I use power

dependence theory to analyze adolescent peer interaction as social exchange, focusing on how adolescents were

dependent upon one another for valued resources.

In short, power dependence theory refers to the idea that actors are dependent upon one another for valued resources and

that power resides within relations of dependence, not the characteristics of individuals (Emerson 1962). Instead of

analyzing higher or lower status individuals, the unit of analysis was characteristics of social relations between adolescents,

paying great attention to whether group members had exclusive access to a resource important to other group members and

whether group members were engaged in social exchange. Data showed that adolescents in groups with balanced power

relations were less susceptible to peer effect mechanisms than adolescents in groups with unbalanced power relations.

Adolescents in groups with balanced power relations were insulated from peer pressure to use drugs or change educational

behaviors largely because of the shared power among group members. In contrast, the group with unbalanced power

relations had members that were powerless, or completely dependent upon the group for a valued resource. In this

particular group, two adolescents were dependent upon the group for having friends because neighborhood violence greatly

constrained their friendship choices. Instead of staying at home alone, these adolescents chose to spend time with the only

group of non-gang affiliated friends near home and, as a result, participated in the activities of the group that consisted

largely of drug use. This study suggests that power dependence relations and factors constraining access to alternative

friends are important for understanding susceptibility to peer effects.

This study also illustrates the usefulness of social psychological theory for the study of micro-level inequalities. Specifically,

power dependence theory can be a useful framework for studying the peer effects process, as it revealed that adolescents'

place in the power structure of their friendship group alters susceptibility to different peer effect mechanisms. In the study

of adolescent problem behavior, social psychological theory can be a useful tool to complement and complicate cultural

explanations that emphasize the importance of distinct class-based or oppositional cultures (Fordham and Ogbu 1986; Willis

1981). While the subcultural perspective has been useful, it leaves little room for

understanding adolescent behavior indiverse groups that do not have distinct class or oppositional group cultures (although

for an exception, see Harding 2010). I hope to illustrate the usefulness of social exchange and power dependence

theory in social psychology for understanding complex forms of micro inequality at the small group level.

To begin, I review the general literature on peer effects, paying great attention to theories of the peer effects process and

peer effects mechanisms. I then describe my theoretical perspective in understanding susceptibility to peer effects and

conclude by urging policymakers to consider interventions aimed at targeting not just the norms and values of adolescents,

but also wider social problems that may leave adolescents incompromised positions of power among peers.

ADOLESCENT PEER EFFECTS AND MECHANISMS

Research shows that young people become increasingly involved with their peers as they age (Brown and Theobald 1999;

Savin-Williams and Berndt 1990), such that they are seen as an important socialization context alongside families, schools,

and neighborhoods (Blumenfeld 1992; Eccles, Wigfield, and Schiefele 1998; Kindermann 1993). While some researchers

have cautioned against using descriptive studies as evidence of peer effects because of selection bias, or the notion

that adolescents self-select into peer groups (Cohen 1977; Kandel 1978; Newcomb, Huba, and Bentler 1986), psychologists

have shown strong causal evidence for peer effects through random assignment in youth intervention programs, showing

the effect of deviant peers on mental health (Dodge and Sherrill 2006), education (Reinke and Walker 2006), and juvenile

justice (Lipsey 2006). While sociological studies have been unable to disentangle selection from socialization, the

association of peers and adolescent outcomes has been established in a number of domains: having delinquent friends

increases individual offending (Aseltine 1995; Haynie 2001; Matsueda and Anderson 1998); having friends who smoke or

drink increases one's cigarette smoking and alcohol use (Urberg, Degirmencioglu, and Pilgrim 1997); and having friends

with poorer mental health leads to more individual depression (Prinstein 2007; Sameroff and Peck 1998).

While the literature has shown the importance of peers, studies rarely identify the mechanisms behind their effects. By

mechanisms, I refer to modes of influence, or ways by which peer influence is likely to manifest during interactions. In their

recent review, Brown et al. (2008) identified four peer effect mechanisms. The first was peer pressure, defined as direct

attempts to affect certain attitudes or behaviors in another person or group. The second was behavioral display (or

modeling), which occurred when someone displayed an attitude or behaved in a way desired by other people (Dunphy 1963;

Zine 2001). Third was antagonistic behavior, teasing or ridicule that maintained group status hierarchies (Adler and Adler

1995; Eder and enga 2003; Fine 1987). The fourth mechanism was structuring opportunities, or instances where
an adolescent was invited to a situation that facilitated a behavior without imposing it (Brown et al. 2008). For example,

when a peer invites a teen to a party, he or she may be more likely to experiment with drugs or alcohol.

These four mechanisms were not a complete list, but rather an overview of the mechanisms analyzed in this study. Next, I

turn to the central concern of this article, theories of adolescent susceptibility to peer effects.

Status and Models of Susceptibility

Most theoretical models of susceptibility to peer effects focus on status differences between adolescents. For example, the

prototype willingness model (Gibbons et al. 2008) argued that adolescent susceptibility was based on rewards associated

with emulating peer behavior, suggesting that motivations to engage in risky behavior were high when they were associated

with higher status or popular groups (Gerrard et al. 2002; Gibbons and Gerrard 1997). In addition, the deviancy regulation

model argued that desire for positive self-image dictated decisions to conform, especially when the reference group was

perceived as higher status (Blanton, Stuart, and Van den Eijnden 2001).

To include factors beyond status differences, I utilize Brown et al. 's (2008) conceptual model for two reasons. First, it is the

most comprehensive because it includes multiple factors influencing susceptibility. Second, it views peer influence as

dependent on contextual and individual factors, which allows room for analysis of how status or group power structure

changes as adolescents interact across multiple settings. Brown et al. 's (2008) conceptual model (Figure 1) breaks down

susceptibility into three major factors: (1) openness to influence, (2) the status of those exerting influence, and (3)

relationship dynamics. Studies have paid most attention to status differences between adolescents as a key factor, with very

little attention to relationship dynamics.

Literature on openness to influence has focused on differences in relation to status characteristics such as age (Warr 1993,

2002) and gender (Thornberry and Krohn 1997), as well as on family structure or attributes of parents (Berndt 1979; Brown,

Ciasen, and Eicher 1986; Brown et al. 1993; Mounts and Steinberg 1995; Steinberg 1987). Status has also been at the

forefront of susceptibility studies looking at relationship dynamics. Ethnographies of adolescent groups have shown patterns

of influence in groups with clear status hierarchies (Adler and Adler 1995; Dimitriadis 2003; MacLeod 1995). Others have

studied the strength of relationships between adolescents, showing that adolescents may be more easily persuaded to

smoke (Lightfoot 1997) or enroll in math courses (Frank et al. 2008) by weak as opposed to strong ties.

In summary, approaches to adolescent susceptibility to peer effects focus on status differences. In this article, I argue that

peer influence is also based on the power structure of adolescent friendship groups. Inaddition to having high

status, adolescents in groups with unequal power structures wielded power over members by having exclusive access to

resources important to the group. Analyzing power structures inadolescent friendship groups can help create a clearer

understanding regarding when higher or lower status is likely to make adolescents more or less susceptible to particular

peer effect mechanisms.

Overlooking group power structures may blind researchers from additional factors. Within a status framework, protecting

one's self from peer influence appears limited to attaining higher status within the group or changing the values of the group

such that particular behaviors (e.g., smoking or drug use) are no longer seen as higher status. An important and

unanswered question in this framework is: why does the lower status adolescent not leave the group or seek alternative

friendship groups? If one is in a lower status position, whynot leave that position and join a group in which you can be higher

status? Romans (1961) pointed to this question in his earlier formulations of social exchange theory, and Emerson (1972a,

1972b) subsequently developed the notion of power dependence. In the following section, I discuss my theoretical

perspective.

A POWER DEPENDENCE APPROACH

This study contributes to the peer effects process literature by utilizing power dependence theory from social psychology to

analyze the interactions of three adolescent friendship groups. From a microsociological perspective, the peer influence

process is a case of micro-level inequality, with Adolescent A having direct or indirect power over Adolescent B. Research

on susceptibility to peer effects has focused on status inequality between adolescents as an important factor in the peer

influence process. Incorporating power dependence into our understanding of micro inequality between adolescents would

present an additional perspective illuminating additional factors (e.g., constraints on adolescent friendship choices) to take

into consideration when designing interventions.

Generally, Emerson's (1972a, 1972b) theory argued that actors were dependent on one another for valued resources, were

motivated to obtain more of the resources valued and controlled by others, and engaged inrecurring exchanges in which

benefits were contingent on benefits provided. For Emerson (1962:32), "power is a property of the social relation; [and] not

an attribute of the actor." Thus, I do not analyze individual characteristics, but rather the social relations between specific

group members, such that the unit of analysis is the social relation between individual members within three groups.

Through a combination of ethnographic observation and follow-up interviews with group members, I consider power

dependence in the social relations of group members by focusing on whether or not a group member had exclusive access

to a resource important to other group members. This strategy helped identify characteristics of power dependence relations

between adolescents, specifically whether or not power relations were balanced or unbalanced. In two of the three

groups in this study, adolescents were engaged inreciprocal exchange, which helped insulate them from the influence of

their peers as each member had equal power within the group.

In contrast, unbalanced power relations meant relations where one adolescent is more dependent on another than vice

versa for a valued resource. According to Emerson (1962), the actor in the disadvantaged position has numerous strategies

to balance the unequal power relationship. These included reducing their own motivational investment in the relationship,

seeking alternative partners, increasing their exchange partner's motivational investment in the relationship, or withdrawing

from the relationship altogether. In this study, access to alternative friends for the disadvantaged adolescent was particularly

important for understanding peer influence.

To analyze status, I focused on ascribed status characteristics of group members such as gender, educational achievement,

and age. I also focused on status systems emerging from social contexts, such as ability to perform particular tasks. As

Emerson (1962) stressed, power relations change as contexts change; thus, paying attention to both status and power

dependence relations helped describe peer influence processes ineach friendship group.

At this point, it is important to clarify that I apply power dependence theory differently than ways used in the social

psychological literature. While I follow social exchange theory's foundation that power is a structurally determined

phenomena resulting in differential exchange outcomes (Thye 2000), social exchange researchers typically conduct

laboratory experiments with social networks as the object of analysis (Cook and Rice 2006), not ethnographic observations

of friendship groups. In ongoing relationships within groups, it is difficult to discern whether scope conditions of power

dependence theory are met. Nonetheless, the theory provides a useful framework for understanding cases of micro-level

inequality in other subfields of sociology, especially those involving small groups.

Although social psychological theories such as social exchange are often limited to the scope of particular situations or task-

oriented groups, insight from these theories should be applied to adolescent friendship groups, as the dynamic task and

non-task orientations of such groups pose an interesting challenge to social psychologists. While adolescent friendship

groups may not be exclusively task oriented, as illustrated by the work of Adler and Adler (1995), they are sites where status

hierarchies and power dependence relations intersect. Recent research on the intersection of status characteristics theory

and social exchange theories (Thye 2000; Thye, Wilier, and Markovsky 2006) shows the benefits of applying theories

slightly outside of their scope.

METHOD

Data were collected through participant observation and in-depth interviews with members of three friendship

groups in "Puebla," a Chicago neighborhood. Participants were second-generation Mexican American adolescents. Initial

observations occurred at a local community organization youth group and subsequently included additional sites as I

became acquainted with the adolescents. As a volunteer in the youth group, I befriended these adolescents by chaperoning

field trips, teaching a mini sociology course, and frequently attending youth group meetings. Soon after joining the youth

group, cliques became noticeable, and after gaining the trust of key informants in each group, I accessed the social circles

of these adolescents.

Field sites ranged from the community organization and local gym to back alleys, vacant lots, and abandoned bridges where

group members spent their leisure time. The fieldwork is based on a year's worth of observations (approximately 250

hours in total). Months after the initial observations, interviews took place to ask group members specific questions about

field observations: in total, I interviewed 23 adolescents, including each adolescent in the three friendship groups, with some

members interviewed two or three times. My ethnic background and close proximity to the age of my participants

helped in accessing these social circles. Gender, however, was a barrier in accessing female friendship groups; males were

more open and easy to observe than the all-female adolescent groups I encountered in Puebla. As a result, these findings
may be limited to male and mixed gender adolescent groups and not to groups of females or dyadic relations (romantic ties

or best friends). In addition, the specific findings from this study, particularly those in relation to neighborhood violence,

should be carefully generalized. As Stolte, Fine, and Cook (2001) argue, microsociologists aim at generalizing to situations,

not populations. Thus, findings from this study suggest that researchers and policymakers should focus on situations

placing adolescents in disadvantaged positions of power (be it as a result of neighborhood violence, within romantic

relationships, or as a result of pledging in sororities or fraternities).

I define a friendship group as three or more adolescents who voluntarily spent time with one another both within and outside

school and work, or groups that hung out to have fun. Most quantitative studies rely on adolescent self-reports to measure

friendship groups, which raise validity issues as friendship can mean different things to different adolescents. The qualitative

nature of this project allowed me to base my identification of friendship groups on observation, allowing me to confirm

identification of a group with all members and specifically pay attention to whom group leaders invited when planning group

activities. The three friendship groups in this study are described in tables in the Appendix.

To measure power dependence, I paid close attention to exchange relationships between group members, particularly who

made suggestions, led, and received compliance from group members, as well as whether favors or gifts between members

were reciprocated. By observing group relations in the field, my data provide an alternative means to view status and power

dynamics than those typical of laboratory studies. As adolescents in this study ranged from 16 to 18 (juniors and

seniors in high school), they may have been less susceptible to peer effects because of their older stage in adolescence

(Giordano 2003).

Finally, it is important to mention the role of parents in the study. With the exception of Ron and Al in the 12th Street Group,

most parents had a handsoff approach to monitoring their kids. In interviews with parents, very few knew their child's close

friends; largely because these parents worked multiple jobs and long hours to support the family. The few girls in the study

were free to participate in social outings near their homes and would often sneak out of the house to accompany their

friends to parties. On the other hand, boys were often allowed to come in and out of the home as they pleased. The results

of this study were not meant to contest the importance of parenting, but rather highlight mechanisms, in addition to

parenting, that may explain variation in peer effect susceptibility.

THE 12TH STREET GROUP

The 12th Street Group consisted of Ron, Sergio, and Al, boys who were similar on most status characteristics (gender, race,

age) but different in educational achievement and social skills (see Appendix, Table 1). Ron and Al were both honor roll

students at a private high school outside Puebla while Sergio failed a grade inelementary school and struggled to avoid

dropping out. One might suspect that the high achievers would affect the behavior (and perhaps school performance) of the

low achiever, or vice versa. Observations, however, revealed no peer influence for educational behaviors. In the 12th Street

Group, this can be attributed to balanced power relations, as each member had access to a resource valuable to group

members and engaged in exchange.

This was first noticeable at the local youth center, where the group task was publishing a monthly newsletter and where

status was based on academic skills. This was evident when adolescents edited each other's articles.

As Rosa read Sergio's article, she exclaimed in front of the group "No! You spelled this wrong . . . you need to elaborate on

this, you didn't put the organization's name, or where this took place." Sergio responded "yeah yeah." Rosa continued, "You

capitalize LIF because it's an acronym." Sergio disputed her claim. Rosa continued correcting Sergio "it's spelled 'start,' not

"starf!,"' the group laughed, Rosa responded, "Oh my god you guys really are dumb." Sergio responded "what!" Rosa turned

to the youth group coordinator, and asked if it was ok for Sergio to rewrite his entire article.

In the youth center, status was based on writing, reading, and speaking (or academic) skills. The division of labor at the

youth center revealed that Ron and Al were high status in this context. The youth group coordinator's assignment of tasks to

particular adolescents made clear who was trusted with the more prestigious and complicated tasks.

On a busy summer day a group of documentary filmmakers came to visit the group, asking to interview some youth. With a

large number of office duties to complete, the coordinator had to choose carefully whom to assign to which task. As the

coordinator went around the table, Sergio was assigned to make copies, others were assigned to draw posters, while Ron

and Al were assigned to go with the documentary film crew for the interview.

As Ron and Al performed well in both school and the youth group, expectations formed within the group, with Ron and Al

having higher status in relation to educational performance. Conversely, Sergio was lower status, as Ron and Al joined

others hi ridiculing Sergio. Ridicule in this context exemplified the antagonistic behaviors peer effect mechanism (Brown et

al. 2008), where playful teasing maintained the group hierarchy. While Sergio's role within the group's status hierarchy is

maintained through this mechanism, he made no effort to emulate the educational behaviors of others, nor did he leave the

group. In fact, he resisted peer pressure to change his attitude toward schooling. During an introduction to sociology

session, I asked the group how they felt about their prospects for college. While a number of youth discussed high

aspirations, Sergio responded infront of the group of 10 youth and myself:

"It's not like I wake up and say I'm a day closer to my success. I just let life live itself, and luckily I know when to do

something good or bad. Honestly, I can't tell you that I'm going to come out good, but the odds are I'll end up living in the

community or somewhere else, but not like in a good situation." Ron responded, "Don't talk like that man."

This exemplified one of many efforts by Ron and Al to change Sergio's attitude toward schooling, which ranged from inviting

Sergio to study groups and offering to help with homework. Despite these efforts, Sergio's behaviors and attitudes toward

school did not change. This scenario illustrated the limitations of Cohen and Prinstein's (2006) finding that lower status

group members were likely to accede to the opinions of higher status peers. Shadowing the group in contexts outside the

youth group revealed (1) how Sergio eluded this pressure and (2) why Sergio did not opt out of the group.

Conversing in Al's backyard, Ron described how Sergio was the "ladies man" of the group. As they all looked up profiles of

girls on a social Web site, Ron and Al praised Sergio while browsing profiles of girls he dated. At times, Ron and Al teased
Sergio by calling him a "manwhore," but these interactions brought about resistance from Sergio, who ridiculed Ron and Al

for their lack of "success" with girls. This contrasted Sergio's acceptance of ridicule at the youth group. Sergio's higher social

skills placed him in the role of broker between Ron, Al, and the girls they courted.

While on break from a youth group meeting, the group walked to a corner grocery store to buy a drink. Upon entering the

grocery store, Ron's eyes were immediately focused on the young girl behind the cash register. After calling us over, Ron

said to the group, "check out those tig ol bitties" [telling the group to focus on the young girl's breasts]. Ron urged Sergio to

flirt with the girl and introduce him to her.

This interaction illustrated Ron's dependence on Sergio for romantic relationships. While Ron and Al's higher status granted

them privileges at the youth group, it served no use in contexts where the group interacted with girls. These observations

illustrated the balanced power relations within the group, as each member had an important resource valued by the group.

Another example occurred when the 12th Street Group played basketball.

Only four other people were in the gym so we played a four on four full court game. Al, Sergio, Ron, and I played against

four others. Prior to the game, Sergio and Al observed our opposing team warming up and evaluated their talent based on

how they shot the ball. When the game started, Sergio directed Ron to guard the least talented player on the opposing team,

and Ron obliged. During the game, Ron declined numerous scoring opportunities, passing to Sergio or Al. When Ron was

open for a scoring opportunity, Sergio would refrain from passing him the ball. Ron made no complaints about our style of

play.

Ron's deference to Al and Sergio illustrated another status hierarchy within the group. By passing up opportunities and

listening to the commands of Sergio, Ron deferred to higher status members in this specific context. Unlike the youth group

context where Sergio was lower status, the hierarchy was reversed, with Sergio being higher status and Ron being lower

status. In the following section, I discuss how these social relations had implications for susceptibility to peer effects.

Peer Pressure and Educational Behaviors in the 12th Street Group

Interviews showed how balanced power relations insulated group members from adopting each other's academic behaviors.

Ron spoke of his frustration with Sergio's school performance:

Ron: I know Sergio is smart, he's just lazy ... we were just talking about it the other day, it was me, Sergio and Al. I know I

can get into most of the colleges I want, but with Sergio, he's not on the same level as me. His GPA is not as high, and I was

worried because I know he's not going to get accepted to the same colleges as I get accepted to.

Robert: Do you try to encourage him to do better?

Ron: Yeah, I get on his case all the time, but nothing ever seems to come out of it.
When Ron mentioned "getting on Sergio's case," he meant teasing Sergio for his academic shortcomings. Sergio

explained why he put up with Ron's ridicule, explaining "it doesn't bother me, he'll say I'm slow, but I'm like, whatever man."

When asked about why he accepted teasing from Ron and Al, Sergio responded:

Because it's funny, we sit around a lot and get bored, so we just say stupid shit 'cause we ain't got nothing else to do. If it

weren't for these guys, I'd probably be in a gang or something.

As Puebla was plagued by gang violence, finding safe places was difficult for adolescents. Being in the 12th Street Group

granted Sergio the ability to socialize and spend leisure time in a safer environment; thus, power relations between group

members were balanced. Sergio depended on the group to provide social opportunities that did not involve deviant activities.

Conversely, Ron and Al depended on Sergio for romantic relationships and a competitive advantage while playing

basketball at the local gym.

Summary

The 12th Street Group showed that balanced power relations were important insulators from peer pressure and behavioral

display mechanisms. Throughout data collection, I did not observe a member conforming to direct peer pressure of others to

change academic behaviors. Members did not emulate each other's academic aspirations or behavior toward girls. Group

members were insulated from each other's influence because of their balanced power relations, exemplified by each

member having a resource important to group members and the reciprocated exchange between group members.

For power dependence relations to insulate adolescents, the group must engage in interactions across multiple contexts that

do not reinforce the same status hierarchy. For example, had the group only interacted in settings where academic skills

were highly valued, Sergio would have no power in the group because he would not be able to help Ron and Al with courting

girls. This suggested that while status characteristics may form the basis of power and influence in a specific context

(e.g., in basketball), the exercise of power in an adolescent friendship group depended on whether status was similar across

all contexts of interaction.

THE BELL PARK GROUP

The Bell Park Group revealed that power dependence relations and higher status insulated adolescents from different peer

effect mechanisms. This group was mixed on the basis of gender, class, educational achievement, and risk-taking activities

(see Appendix, Table 2). Some were at the top of their high school class with parents who had college degrees from Mexico;

others were below average students from working families. Brought together by geography (living near Bell Park), the group

served a social function. Antonio described the group's formation:

Antonio: It was just whoever we hung out with, when we started hanging out with new people, and if we liked them, he kept

hanging out with us.

Robert: What do you think you guys liked about new members?

Antonio: I don't know . . . they are just like us ... like Ben . . . he's just like me and everybody else . . . it's just . . . you have to

know what you're doing. Like the one thing about it is that everybody has their own mindset. Like if everything goes as

planned, everyone will be fine. Basically you feel comfortable with them, we've actually kicked people out of the group, well

not actually kicked them out, but not called them. We're the kids that party but still maintain their grades.

Antonio's comments showed how a mix of similar senses of humor and achievement orientations brought the group

together. Group interactions took place at school, the youth group, abandoned railroad tracks, and house parties. Tomas

was a low-achieving student (2.3 GPA) whom group members described as the leader. Most attributed Tomas's leadership

to his outspoken personality, often displayed when the group planned and organized activities. The following example

illustrated Tomas's status as leader of the group:

Every time we go out, they're like "Tomas, what are we doing? How are we getting there?" That's always me. Lately I

haven't done anything because . . . because they were like, this movie sucks, well I'm like, you know what, fuck you guys. If

you want to go, you decide. I'm tagging along.

One evening, Tomas refrained from his role as event planner and allowed the group to decide. In the absence of Tomas's

leadership, the group became disorganized. One member would offer an idea only to be rejected by others, resulting in the

group driving around all night and doing nothing else. In the context of planning activities, the group was dependent on

Tomas.

Despite his power in the group, Tomas (a low-achieving student) did not pressure group members to conform to his

academic indifference; in fact, the opposite occurred. As the group sat at their lunch table, the following conversation

ensued.

Tomas and Yesenia (the member with the highest GPA) were conversing about their future plans in front of the group.

Yesenia got on Tomas's case, "Dude, you should go to college, you're smart." Tomas responded, "Nah, I don't want to put

up with all the bullshit, I don't want to leave the city." Yesenia replied, "I've been in Puebla all my life, I want to get out."

Tomas asked "Are you ready to go to a big school?" Yesenia replied "I don't know, but that doesn't matter, you should be

staying after school with us [to get extra math homework in preparation for college]."

At school, Yesenia had higher status, as she was one of the group's high achievers. Her status, however, did little to

influence Tomas's behavior or attitudes toward schooling. When I began shadowing the group, Tomas's GPA had been 2.9,

and one semester after taking a full-time job, his GPA dropped to 2.3. Tomas's parents (both working-class immigrants)

respected his aspirations, whether that was to work or stay in school, but his peers did not share the sentiment. The group

pressured Tomas to work fewer hours or quit his job altogether, yet Tomas refrained from doing so. Despite failing to

conform to the group, Tomas was not isolated or excluded. Power dependence relations within the group helped

explain why this was the case.

Not only was the group dependent on Tomas for leadership, but also for fulfilling a graduation requirement, as graduation

from Chicago Public Schools required performing 40 hours of community service. Group members attended the local youth

group to fulfill their hours. Through his past involvement with the youth group, Tomas occupied a leadership position that left

him responsible for recording student service hours. As the group depended on Tomas for leadership and documenting their

service hours, they accepted his academic behaviors. Like Sergio in the 12th Street Group, Tomas illustrated how power

dependence can insulate youth from peer pressure or behavioral display when the group was involved in multiple tasks.

Low-status Member Antonio

While Yesenia received high status for her academic achievement, Antonio, the lone male high achiever in the group, did

not; in fact, Antonio was lower status across all contexts of interaction. Regardless of group activities or context, Antonio

was the target of jokes and ridicule.

Tomas: Antonio is seen as a little . . . how do you say it ... feminine. We've had times where he's like, why does that have to

be gay or what's not cool about sushi? [laughing] We always make fun of him. The kid says the dumbest things. We'll argue

about something and he'll know it's wrong, and there's evidence proving he's wrong, but he tries to force his opinion on

everyone. We always make fun of him for that

Yesenia confirmed Tomas's observations, "Antonio can be imposing at times, which annoys some of us, but when someone

makes fun of him for it, it's hilarious!" Tomas continued, "When he doesn't get what he wants he throws a fit. He's such a

drama queen. He'll say things like I hate my life, I hate my parents, blah blah blah, he's got a whiny attitude." Despite the

teasing, the group was careful not to alienate Antonio because they heavily depended on him for transportation to social

events (Antonio was the only group member with a car). On the other hand, Antonio was dependent on the group for

instrumental purposes.

Antonio: I don't really like a lot of people. I only associate with people who are doing things with their lives. If I don't see a

benefit in talking to you, I won't talk to you. With [the Bell Park Group], we, for the most part, are people who are doing things

with our lives in our own different ways. Plus, we're all funny, so that's why I hang out with the group.

Within the group, Antonio was closest to Yesenia. These two were the highest achieving students in the group, with GPAs

above 3.5. Antonio's explanation showed he was dependent on the group to be friends with similarly academic oriented

students while maintaining a social life. In his interview, Antonio described why he refrained from hanging out with other

academically oriented students: "Those kids don't have a Ufe, all they do is study." In Puebla, it was difficult for students like

Antonio to be involved in friendship groups composed solely of high-achieving students, as many high-achieving

students in Puebla were not allowed to leave home by their parents because of gang violence. For Antonio, his friendship

with Yesenia (whom Antonio's parents knew well) persuaded his parents to allow him to attend social activities. In addition,

Antonio's car allowed the group to attend social activities (movie theaters, parks, lakefront) outside the resource-deprived

neighborhood.
In contrast to Sergio in the 12th Street Group, Antonio did not have higher status in any context, even when it came to

school and academic achievement. While the group admired and respected Yesenia's achievement, Antonio's was seen as

illegitimate because of the school he attended. Although all group members attended Puebla High School, the school was

actually broken into two separate campuses, one for Arts and the other for Science. Yesenia had been enrolled in the

Science school, while Antonio attended the Arts school. In the group, Yesenia's achievement was given more credit because

the Science campus was perceived as the more difficult of the two. As a result, group members perceived Antonio's

achievement as second-rate.

Thus far, data on Antonio revealed evidence for how balanced power dependence relations insulated him against group

peer pressure or behavioral display mechanisms. Despite being lower status in all contexts, Antonio's powerful

position in the group (by having a car) protected him from peer pressure to change his academic behaviors. After eight

months of fieldwork, however, Antonio's status within the group dramatically changed, revealing that balanced power

relations may only insulate from specific forms of peer influence.

Limits of Balanced Power Relations

In his initial interview, Antonio downplayed teasing from friends, saying, "The person who gets made fun of most varied from

time to time depending on what's going on." Weeks later, group dynamics changed dramatically as Antonio received

acceptances to two prestigious universities. News of Antonio's accomplishment coincided with an annual fundraiser by the

local youth group, and to celebrate Antonio's accomplishment, the group attended the fundraiser. Antonio and Tomas were

the first group members to arrive, and many local community organizers walked over to Antonio, congratulating him on his

accomplishment and asking questions about which school he would choose.

As the remaining group members arrived shortly after, the fundraiser revealed a number of interesting conversations

between low academic achieving members of the group and organizers who congratulated Antonio. Specifically, Tomas,

who in previous interviews and observations expressed little confidence in his ability to attend a four-year college, started

speaking of attending four-year universities and majoring inbusiness. In speaking to the youth group coordinator, Tomas

said, 'Teah, I can get into Southwestern University if I wanted, I have a friend who works on the admissions committee, but

I'm counting on getting into Stewart University instead."

These interactions yielded important insight into the limitations of balanced power relations as an insulator of peer influence.

First, Tomas's change in attitude toward college showed that balanced power dependence relations cannot

insulate adolescents from emulating (or what Brown et al. 2008 call the behavioral display mechanism) the aspirations of

peers during episodes of impression management (Goffman 1959, 1967). At the youth group, Tomas's emulation of

Antonio's educational aspirations helped him avoid the embarrassment of being a low academic achiever in a setting where

academic achievements were being honored. Tomas's behavior after the event confirmed this as the organizational

fundraiser was the only instance throughout the entire data collection process where he expressed high academic

aspirations. This highlighted the importance of differentiating between the emulation of behaviors (like studying) and the
emulation of aspirations for peer effects. The Bell Park Group showed that while balanced power dependence relations may

not insulate adolescents from expressing the similar aspirations of their peers in particular settings, it did insulate group

members from emulating the behaviors of peers in relation to improving academic achievement.

Second, Antonio's accomplishment granted him higher status, which insulated him from teasing (or the antagonistic behavior

mechanism). Prior to this, Antonio's academic achievement was seen as illegitimate, and his acceptance into prestigious

schools legitimated his academic skill. As a result, teasing of Antonio decreased significantly as the group admired Antonio

for his accomplishment. After this change, Ben, the lowest achieving student in the group, with a 1.9 GPA, who regularly

skipped school, became the target of jokes. As the group hung out at a local restaurant, Tomas and Carla teased Ben for his

risky behaviors. Ben shared how he was arrested for driving his parents' car without a license, to which group members

responded 'You're a dumbass! You're an idiot!" When asked why they teased Ben, Carla responded "we all make fun of him

[Ben] because he asks for it, because he's just real ignorant sometimes. He's lazy because he skips [school] a lot, but he's

still [cool]." After this change in the group's status hierarchy, Antonio described the function of teasing and ridicule in the

group:

The teasing we do kind of sets a ranking system within the group, but it depends on the situation. Some people may lead

others, and some may not. Like, Carla knew someone who was throwing a party and a place where she could get in for free,

and she is normally someone who would not plan anything. Andres is the lazy one, they try to get him to do well in school.

Tomas is the athletic one, Yesenia and me are the smart ones. We make fun of each other because it's funny, and it's also a

way of putting people in their place.

Antonio's quote illustrated how higher status insulated adolescents from teasing. Although recipients of teasing

changed in the Bell Park Group irrespective of balanced power relations, the academic behaviors of group members did not

change. This was also the case for Sergio in the 12th Street Group, where teasing and ridicule did not result in behavior

change. This suggested that balanced power dependence relations and status serve as insulators from different types of

peer influence mechanisms.

Summary

Data from the Bell Park Group indicated that status processes and balanced power dependence relations

insulated adolescents from different types of peer effect mechanisms. Balanced power dependence relations insulated

group members from peer pressure to change academic behaviors, but it did not insulate adolescents from emulating the

aspirations of peers in certain contexts. Nevertheless, emulation of academic aspirations within the group did not produce

behavioral changes among group members with respect to academic achievement. While the Bell Park Group illustrated

limitations of balanced power relations, showing that it did not protect against emulation of aspirations, emulation of

aspirations was not enough to produce behavior changes. Although data presented do not provide a strong enough test to

rule out the importance of emulation of aspirations for peer effects on academic achievement, the Bell Park Group (and the
Thomas Group in the following section) illustrated the importance of delineating the role of status and power

dependence in the peer influence process.

THE THOMAS GROUP

The Thomas Group provided a counterexample illustrating how unbalanced power dependence relations made particular

group members susceptible to peer pressure, structuring opportunities, and behavioral display mechanisms. Named after a

local street, the Thomas Group was comprised of seven adolescents, three female and four males. This group was

mixed in relation to class and educational achievement (see Appendix, Table 3). In this analysis I focus on the experiences

of Joe and Hector, who consumed drugs and alcohol largely as a result of the indirect influence of their peers.

David as Higher Status in the Thomas Group

David's interactions with the group illustrated he was leader and higher status in the group. David's status was legitimated

largely by his close friendship with the girls in the group and his high GPA. Whenever David wanted to go somewhere or do

something, the girls quickly agreed, and the three other males followed along. Hector and Joe rarely had a say in the

process, while Esteban went along with whatever his girlfriend (one of the female members of the Thomas Group) wished.

For example, after a youth group meeting, Joe wanted the group to go to Hector's house, but his suggestion was met with

no response from group members. After a few minutes David suggested they walk to Thomas Street, the girls (and Esteban)

agreed and followed David, while Joe and Hector followed moments later.

Observations at the youth group and Rosewood High School illustrated how David was higher status in the group. At school

and the youth group, academic tasks were markers of higher status. At Rosewood High School, where the Thomas Group

attended school, the student with the highest GPA was awarded a new laptop computer. With a 3.7 GPA, David won the

computer, earning the reputation as one of the smartest students. At the youth group, David was also higher status, as the

coordinator would often share David's success story to the group, urging them to follow his example.

Susceptibility in the Thomas Group

Hector and Joe revealed how unbalanced power dependence relations created greater susceptibility to particular peer effect

mechanisms. As lower status members, Hector and Joe were often targets of harsh jokes and had little say in planning

group activities. On some occasions, the group attended parties outside Puebla and left Hector and Joe to plan their own

way of getting to and from group outings. One might argue that the group was dependent on Hector and Joe for

entertainment because they were constantly teased and ridiculed. This was not the case, however, because on many

occasions the group left Hector and Joe with no transportation to the event. If the group were dependent on Hector and Joe,

they would have at least made efforts to ensure they could attend group outings.

Although Joe participated in the group, he was not content with his status. In an interview, he questioned David's higher

status, "He's not smart because he doesn't really use it. He has a good GPA but he doesn't really care about school."
Robert: How do you respond to David making fun of you?

Joe: It's bullshit you know ... I should talk back; say "fuck you!"

Robert: But you don't?

Joe: I know. Because I don't ... I mean, I do care, but I don't want to go through the problems. What can I do? I'll be like

"shut up you little bitch, you fucking faggot!"

Robert: What would you want to do?

Joe: I would of fucking punched him in the head, I'll be like "shut the fuck up!" But I can't.

Robert: You feel powerless when your friends make fun of you?

Joe: Yeah, because I really couldn't do anything about it, I can't change it, so I'm like . . . whatever.

Joe's view of David's status as illegitimate illustrated that Joe's conformity to the group's teasing had little to do with status

inequality and more to do with unbalanced power dependence relations (as expressed by his feelings of

powerlessness). In a comprised position, Joe had nothing to bargain with to the group and the group was not dependent on

Joe. Thus, Joe was faced with deciding between conforming to the group or possibly being excluded. He begrudgingly

chose the former.

Robert: Why do you hang out with David and the group?

Joe: Well . . . (pause). If it wasn't for the other people in the group, I wouldn't really hang out with David.

Robert: Why don't you bring up your dislike of David to the group?

Joe: Are you kidding? They'll be like, "Oh, you don't like David? Why's that?" Blah Blah Blah.

Joe illustrated his concern about exclusion from the group. While he was dependent on the group to hang out with Esteban,

no group member was dependent on Joe, and as a result, he had no power to contest the group's power structure.

Specifically, Joe was afraid of starting a conflict he thought he could not win. Joe did not possess resources (economic or

social) needed by the group. In the following examples, data show how unbalanced power dependence relations can make

the disadvantaged group member more susceptible to peer influence.

Peer Influence in the Thomas Group

Joe's disadvantaged position in the imbalanced power dependence relation contributed to his conformity to all peer effects

mechanisms. For example, on numerous occasions Joe expressed his dislike of techno music and tried to persuade group

members to listen to reggae, but his efforts were consistently rebuffed. One day while the group was hanging out at a
storefront, Joe was listening to reggae on his headphones. As we heard the music from Joe's headphones, David walked

over and yelled "turn that shitty music off!" Joe gave David a dirty look, but then turned off the music. In this example, Joe

gave in to David's direct peer pressure to turn off his music. Later that evening, the group attended a neighborhood house

party that was playing loud techno music. Despite his expressed dislike for this style of music, Joe joined the

group in dancing to this music, emulating the group's musical tastes. In this example, Joe's conformity has taken the form of

the behavioral display peer effects mechanism (imitating the behaviors or attitudes of others).

The most dramatic example of conformity came in relation to smoking marijuana. Prior to joining the group, Joe had

experimented with marijuana in grade school. Yet after hanging out with the Thomas Group, Joe confessed that he smokes

marijuana far more often now.

Joe: Before I would smoke every once in a while with friends, but these guys [the Thomas Group], they smoke all the time.

That's all they do, go out to Thomas Street and light up.

Robert: And you join them?

Joe: Yeah.

Robert: How come?

Joe: I don't know, well. . . . It's like, if I'm around it more and like . . . my friends are doing it, I'll do it. It's just that when I'm

with Esteban and everybody, that's what we do. I'd rather do other stuff, but when we're broke, we just all pitch in what we

got to buy a blunt.

In this example, Joe's conformity operated through the mechanism of structuring opportunities, or the creation of a situation

that facilitates a certain behavior (Brown et al. 2008). Joe's disadvantaged position inthe group's unbalanced power relations

made him more susceptible to the influence of his peers, but it was unclear whether this was the result of his lower status or

lack of power.

Joe's explanation for why he conformed to the group instead of hanging out with other friends revealed the importance of

external factors, which in this case was neighborhood violence. In Puebla, Joe would have to travel through violent streets to

see his other friends. As Joe explains, "I got other friends I went to school with on the other side of the neighborhood, but I

ain't gonna cross 20th Street and the Dragons [local street gang] territory to get there. I ain't getting shot." While Joe's lower

status contributed to his susceptibility, his inability to access alternative friendship groups made him dependent on the

Thomas Group to have a social life.

Hector provided additional support for the importance of power dependence relations. Known as "the stupid one" by group

members, Hector explained his involvement in the group to me at a neighborhood block party. Throughout the night, Hector

drank alcohol by himself and the group only interacted with him to tease him. I asked Hector why he chose to hang out with
a group that referred to him as the "stupid one." He paused and replied, "I don't really know why, it's just that they're always

planning shit, they're always giving me something to do on a Friday night, going out, doing stuff like this. It's different than

what I would prefer to do, but it gives me something to do." Like Joe, Hector revealed that he was dependent on the Thomas

Group to participate insocial activities. Rather than being at home, Hector conformed to the group, taking on his role as a

target of jokes. When asked what he preferred to do, Hector mentioned playing sports or video games with friends. With no

power to negotiate activities with the group, Hector was put into the compromised position of having to choose between

conformity or exclusion.

Hector's compromised position contributed to his consumption of alcohol. In interviews, Hector explained his strong distaste

for beer, but in the presence of the group at parties, Hector drank beer without hesitation. When asked why he drank beer

with the group, he bashfully responded, "I don't know. ... If they [the group] do it I think I should do it. It's a way for me to be

like, cool with them. If I didn't they'd think I'd be judging them or think I'm trying to be better than them or something." This

example illustrated how being in a disadvantaged position in a group with unbalanced power dependence relations can

leave an adolescent vulnerable to the structuring opportunities mechanism. As Hector lacked opportunities to socialize, he

was in a compromised position of choosing between socializing with the Thomas Group or remaining at home.

Joe and Hector's behavior illustrated the importance of context, particularly for accessing alternative friendship groups.

Although their disadvantaged position in the group forced them to choose between conforming or leaving the group, their

lack of access to alternatives contributed to them adopting the behaviors of their peers. Specifically, their access to

alternative friendship groups was constrained by neighborhood violence and lack of resources (not having a car). While

Hector and Joe had friends outside the Thomas Group, it was difficult for them to hang out with these friends due to

contextual factors associated with neighborhood poverty.

Summary

Data from the Thomas Group suggested that disadvantaged adolescents in a group with unbalanced power relations were

more susceptible to peer pressure, behavioral display, and structuring opportunities mechanisms than adolescents in groups

with balanced power relations. The increased susceptibility manifested when adolescents were faced with the compromised

position of choosing between conforming or leaving the group. For example, in the Bell Park Group, Antonio cannot

pressure Tomas to conform to his positive academic behaviors because Antonio was dependent on Tomas for fulfilling his

community service requirement. Similarly, Ron and Al in the 12th Street Group could not change Sergio's academic

behaviors because they relied on him to court girls. In the case of the Thomas Group, members did not depend on Hector

and Joe for anything. Combined with their limited access to alternative friends (stemming from neighborhood violence),

Hector and Joe were powerless to contest the group power structure.

Interactions in the Thomas Group also illustrated that specific peer effect mechanisms result from status inequality among

group members. Lower status explained why Hector and Joe were consistently teased; however, it provided little

evidence in relation to why they adopted peer behavior. Hector and Joe's disadvantaged position in the group's power
relations and their lack of access to alternative friendship groups greatly influenced their decision to consume drugs and

alcohol. This finding speaks to the importance of the characteristics of power relations between adolescents and contextual

features in the peer influence process.

CONCLUSION

Data from three friendship groups showed that status and power dependence relations insulated adolescentsfrom particular

mechanisms of peer influence. Being in a group with balanced power dependence relations insulated adolescents from peer

pressure and structuring opportunities mechanisms. For members of the 12th Street and Bell Park groups, having an

independent source of power in the group (through having access to a resource important to the group) shielded members

from having to conform to peer behaviors. For example, Sergio in the 12th Street Group did not conform to the pressure of

his peers to change his educational behaviors because his peers relied on him for romantic relationships. These interactions

highlighted a missing factor in the peer effects literature, power dependence relations, as it mediated peer influence in these

groups. Future studies should incorporate characteristics of the social relationships of adolescents as an additional

factor in the peer effects process.

These findings also fill some gaps in the literature on susceptibility and peer effect mechanisms. Data illustrated several

limitations with a status-centered approach to studying the peer effects process. Future studies of adolescent peer

effects in laboratory settings would benefit by including measures of power dependence or creating multiple situations

reflecting the multicontextual nature of adolescent relationships.

Status remained a powerful factor, particularly for patterns of teasing and ridicule; however, the effects seemed limited to

specific contexts. For example, in the Bell Park Group, Antonio's newly gained status influenced Tomas to emulate his

behavior. This influence, however, was minimal as Tomas did not emulate Antonio's behavior outside of the community

organization context. If status was a context-specific factor in the peer effects process, future research should pay attention

to features of adolescent environments that modify the salience of status. This study suggests that contexts where higher

status group members receive praise from outsiders may be likely to produce emulation from lower status group members.

This study illustrated the potential contributions of social psychological theory to studies of micro inequality, particularly

within group contexts. In the study of adolescence, social psychological theories can complement subcultura! explanations

of problem behavior, particularly within the inner city. Findings of this study point to the importance of focusing on small

group interaction for better understanding the impact of macro-level factors. In this study, observing groups illustrated the

importance of structural factors such as neighborhood violence and poverty in the peer influence process. If sociologists are

interested in more clearly understanding social problems, whether it be adolescent peer influence, inequality, or culture, the

small group is a context ripe for further exploration.

Finally, this study has a number of policy implications. An exclusive focus on status in the peer effects process leads some

scholars, such as Alien and Antonishak (2008), to argue that the best way to modify peer influence effects may be to change
the values that adolescents communicate with one another. My examples of Antonio and Tomas in the Bell Park Group shed

light on the potential usefulness of this approach as the values of community organizers contributed to Tomas's expression

of higher educational aspirations than he had previously expressed. Yet, it remains unclear whether the values put forth by a

community organization, educational intervention, or classroom would transfer to the additional

contexts in which adolescentsinteract.

Bringing power dependence relations into the peer influence process focuses the attention of policy makers to addressing

instances where adolescents may be put into compromised positions of power that may lead to problem behavior. For

example, neighborhood violence contributed to Hector and Joe's conformity, as they were placed into the compromised

position of choosing between conforming to the group or being socially isolated. Hector and Joe's drug and alcohol

consumption may have been different if they could access alternative friends. In addition to neighborhood violence, policy

makers can also pay attention to other factors that may put adolescents into compromised positions of power, such as

romantic relationships, hazing, or pledging processes for fraternities and sororities. Policy makers may also focus on issues

of power to promote positive behavior. For example, a school could require 40 hours of service at a community organization

for students to graduate. By making the student dependent on a community organization to graduate, it is possible that the

relationship developed can lead to an increased interest in civic activities or the development of relationships with positive

role models.

ACKNOWLEDGMENTS

I am grateful to authences at the Northwestern University Ethnography Workshop, MultiDisciplinary Program in Educational

Sciences Workshop, and the 2009 Chicago Ethnography Conference for helpful feedback. I am also grateful for the

comments and suggestions of four anonymous reviewers, Tom Cook, Cecilia Ridgeway, John Diamond, Ed Telles, Nina

Johnson, Brian Sargent, Corey Fields, Lisa Troyer, and especially Mary Pattillo and Gary Alan Fine, who read countless

versions of this article. Finally, I thank the youth in Puebla whose generosity made this project possible.

References

REFERENCES

Adler, Peter A. and Patricia Adler. 1995. "Dynamics of Inclusion and Exclusion hi Pradolescent Cliques." Social Psychology

Quarterly 58:145-62.

Allen, Joseph P. and Jill Antonishak. 2008. "Adolescent Peer Influences: Beyond the Dark Side." Pp. 141-

60 inUnderstanding Peer Influence in Children and Adolescents, edited by M. Prinstein and K. A. Dodge. New York:

Guilford.

Asch, Solomon E. 1951. "Effects of Group Processes upon the Modification and Distortion of Judgment." Pp. 177-

90 in Groups, Leadership and Men, edited by H. S. Guetzkow. Oxford, UK: Carnegie Press.
Aseltine, Robert H., Jr. 1995. "A Reconsideration of Parental and Peer Influences on Adolescent Deviance." Journal of

Health and Social Behavior 36:103-21.

Berndt, Thomas J. 1979. "Developmental Changes in Conformity to Peers and Parents." Developmental Psychology 15:608-

16.

Blanton, D. H., Anne E. Stuart, and Regina J. J. M. Van den Eijnden. 2001. "An Introduction to Deviance-Regulation Theory:

The Effect of Behavioral Norms on Message Framing." Personality and Social Psychology Bulletin 27:848-58.

Blumenfeld, Phyllis C. 1992. "Classroom Learning and Motivation: Clarifying and Expanding Goal Theory." Journal of

Educational Psychology 84:272-81.

Brown, B. Bradford, Jeremy P. Bakken, Suzanne W. Ameringer, and Shelly D. Mahon. 2008. "A Comprehensive

Conceptualization of the Peer Influence Process in Adolescence." Pp. 3-44 in Understanding Peer Influence inChildren

and Adolescents, edited by M. Prinstein and K. A. Dodge. New York: Guilford.

Brown, B. Bradford, Donna R. Ciasen, and Sue A Eicher. 1986. "Perceptions of Peer Pressure, Peer Conformity

Dispositions, and Self-Reported Behavior among Adolescents." Developmental Psychology 22:521-30.

Brown, B. Bradford, Nina Mounts, Susan D Lamborn, and Laurence Steinberg. 1993. "Parenting Practices and Peer Group

Affiliation in Adolescence." CAiZd Development 64:467-82.

Brown, B. Bradford and Wendy Theobald. 1999. How Peers Matter: A Research Synthesis of Peer Influences

on Adolescent Pregnancy. Washington, DC: National Campaign to Prevent Teen Pregnancy.

Cohen, Geoffrey L. and Mitchell J. Prinstein. 2006. "Peer Contagion of Aggression and Health Risk Behavior

among Adolescent Males: An Experimental Investigation of Effects on Public Conduct and Private Attitudes." Child

Development 77:967-83.

Cohen, Jere M. 1977. "Sources of Peer Homogeneity." Sociology of Education 50:227-41.

Coleman, James S. 1966. Equality of Educational Opportunity. Washington, DC: Government Printing Office.

Cook, Karen S. and Erin Rice. 2006. "Social Exchange Theory." Pp. 53-76 in The Handbook of Social Psychology, edited by

J. Delamater. New York: Springer.

Dimitriadis, Greg. 2003. Friendship, Cliques, and Gangs: Young Black Men Coming of Age in Urban America. New York:

Teachers College Press.

Dodge, Kenneth A. and Michelle R. Sherrill. 2006. "Deviant Peer-Group Effects in Youth Mental Health Interventions." Pp.

97-121 in Deviant Peer Influences in Programs for Youth, edited by K. A. Dodge, T. A. Dishion, and J. E. Lansford. New

York: Guilford.

Dunphy, Dexter C. 1963. "The Social Structure of Urban Adolescent Peer Groups." Sociometry 3:230-46.

Eccles, Jacquelynne S., Allan Wigfield, and Ulrich Schiefele. 1998. "Motivation to Succeed." Pp. 1017-95 in The Handbook

of Child Psychology, edited by W. Damon. New York: Wiley.

Eder, Donna and Sandi enga. 2003. "Socialization in Adolescence." Pp. 157-92 in Handbook of Social Psychology, edited

by J. Delamater. New York: Kluwer.

Emerson, Richard. 1962. "Power-Dependence Relations." American Sociological Review 27:31-41.

Emerson, Richard. 1972a. "Exchange Theory, Part I: A Psychological Basis for Social Exchange." Pp. 38-57 inSociological

Theories in Progress, edited by J. Berger, M. Zelditch Jr., and B. Anderson. Boston: Houghton Mifflin.

Emerson, Richard. 1972b. "Exchange Theory, Part II: Exchange Relations and Networks." Pp. 58-87 inSociological

Theories in Progress, edited by J. Berger, M. Zelditch Jr., and B. Anderson. Boston: Houghton Mifflin.

Fine, Gary Alan. 1987. With the Boys: Little League Baseball and Preadolescent Culture. Chicago: University of Chicago

Press.

Fordham, Signithia and John U. Ogbu. 1986. "Black Students' School Success: Coping with the Burden of 'Acting White.'"

The Urban Review 18:176-206.

Frank, Kenneth A., Chandra Muller, Kathryn S. Schiller, Catherine Riegle-Crumb, Anna S. Mueller, Robert Crosnoe, and

Jennifer Pearson. 2008. "The Social Dynamics of Mathematics Coursetaking in High School." American Journal of Sociology

113:1645-696.

Gerrard, Meg, Frederick X. Gibbons, Amy E. Houlihan, Michelle L. Stock, and Elizabeth A. Pomery. 2008. "A DualProcess

Approach to Health Risk Decision Making: The Prototype Willingness Model." Developmental Review 28:29-61.

Gerrard, Meg, Frederick X. Gibbons, Monica Reis-Bergan, Linda Trudeau, Linda S. Vande Lune, and Bram Buunk. 2002.

"Inhibitory Effects of Drinker and Nondrinker Prototypes on Adolescent Alcohol Consumption." Health Psychology 21:601-

609.

Gibbons, Frederick X. and Meg Gerrard. 1997. "Health Images and their Effects on Health Behavior." Pp. 63-94 in Health,

Coping, and Well-Being: Perspectives from Social Comparison Theory, edited by B. P. Buunk and F. X. Gibbons. Mahwah,

NJ: Erlbaum.
Gibbons, Frederick X., Elizabeth A. Pomery, and Meg Gerrard. 2008. "Cognitive Social Influence: Moderation, Meditation,

Modification, and . . . the Media." Pp. 4571 in Understanding Peer Influence in Children and Adolescents, edited by M.

Prinstein and K. A. Dodge. New York: Guilford.

Giordano, P. C. 2003. "Relationships in Adolescence." Annual Review of Sociology 29:257-81.

Goffman, Erving. 1959. The Presentation of Self in Everyday Life. New York: Anchor Books.

Goffman, Erving. 1967. Interaction Ritual: Essays on Face-to-Face Behavior. Garden City, NJ: Anchor Books.

Harding, David J. 2010. Living the Drama: Community, Conflict, and Culture Among Inner-City Boys. Chicago: University of

Chicago Press.

Haynie, Dana L. 2001. "Delinquent Peers Revisited: Does Network Structure Matter?" American Journal of Sociology

106:1013-1057.

Homans, George C. 1961. Human Behavior: Its Elementary Forms. New York: Harcourt Brace.

Kandel, Denise B. 1978. "Homophily, Selection, and Socialization in Adolescent Friendships." American Journal of

Sociology 84:427-36.

Kindermann, Thomas A. 1993. "Natural Peer Groups as Contexts for Individual Development: The Case of Children's

Motivation in School." Developmental Psychology 29:970-77.

Lightfoot, Cynthia. 1997. The Culture of Adolescent Risk-Taking. New York: Guilford.

Lipsey, Mark W. 2006. "The Effects of Community-Based Group Treatment for Delinquency." Pp. 162-84 inDeviant Peer

Influences in Programs for Youth: Problems and Solutions, edited by K. A. Dodge, T. J. Dishion, and J. E. Lansford. New

York: Guilford.

MacLeod, Jay. 1995. Ain't No Makin' It: Aspirations and Attainment in a LowIncome Neighborhood. Boulder, CO: Westview.

Matsueda, Ross L. and Kathleen Anderson. 1998. "Dynamics of Delinquent Peers and Delinquent Behavior." Criminology

36:269-308.

Mounts, Nina S. and Laurence Steinberg. 1995. "An Ecological Analysis of Peer Influence on Adolescent Grade Point

Average and Drug Use." Developmental Psychology 31:915-22.

Newcomb, Michael D., G. J. Huba, and P. M. Bentler. 1986. "Determinants of Sexual and Dating Behaviors

among Adolescents." Journal of Personality and Social Psychology 50:428-38.

Prinstein, Mitchell J. 2007. "Moderators of Peer Contagion: A Longitudinal Examination of Depression Socialization

between Adolescents and Their Best Friends." Journal of Clinical Child & Adolescent Psychology 36:159-70.

Reinke, Wendy M. and Hill M. Walker. 2006. "Deviant Peer Effects in Education." Pp. 122-40 in Deviant Peer

Influences in Programs for Youth: Problems and Solutions, edited by K. A. Dodge, T. J. Dishion, and J. E. Lansford. New

York: Guilford.

Ridgeway, Cecilia L. and Henry A. Walker. 1995. "Status Structures." Pp. 281-310 in Sociological Perspectives on Social

Psychology, edited by K. S. Cook, G. A. Fine, and J. S. House. Boston: Allyn and Bacon.

Sameroff, Arnold J. and Steve C. Peck. 1998. "Contrasting Ecological Models for Adolescent Mental Health, Problem

Behavior, and Academic Achievement." Presented at the biannual meetings of the Society for Research in Adolescence,

San Diego, CA.

Savin-Williams, Ritch C. and Thomas J Berndt. 1990. "Friendship and Peer Relations." Pp. 277-307 in Ai the Threshold: The

Developing Adolescent, edited by G. R. Elliot and S. S. Feldman. Cambridge, MA: Harvard University Press.

Steinberg, Laurence. 1987. "Single Parents, Stepparents, and the Susceptibility of Adolescents to Antisocial Peer Pressure."

Child Development 58:269-75.

Stolte, John F., Gary Alan Fine, and Karen S. Cook. 2001. "Sociological Miniaturism: Seeing the Big through the

Small in Social Psychology." Annual Review of Sociology 27:387-413.

Thornberry, Terence P. and Marvin D. Krohn. 1997. 'Teers, Drug Use, and Delinquency." Pp. 218-33 in Handbook of

Antisocial Behavior, edited by D. M. Stoff, J. Breiling, and J. D. Maser. New York: Wiley.

Thye, Shane R. 2000. "A Status Value Theory of Power in Exchange Relations." America?! Sociological Review65:407-32.

Thye, Shane R., David Wilier, and Barry Markovsky. 2006. "From Status to Power: New Models at the Intersection of Two

Theories." Social Forces 84:1471-495.

Urberg, Kathryn A., Serdar M. Degirmencioglu, and Colleen Pilgrim. 1997. "Close Friend and Group Influence

on Adolescent Cigarette Smoking and Alcohol Use." Developmental Psychology 33:834-^14.

Warr, Mark. 1993. "Age, Peers, and Delinquency." Criminology 31:17-40.

Warr, Mark. 2002. Companions in Crime: The Social Aspects of Criminal Conduct. Oxford, UK: Cambridge University Press.

Willis, P. E. 1981. Learning to Labor: How Working Class Kids Get Working Class Jobs. New York: Columbia University

Press.
 Zine, Jasmin. 2001. "Muslim Youth in Canadian Schools: Education and the Politics of Religious Identity." Anthropology &

Education Quarterly 32:399-423.

Vargas, R. (2011). Being in "bad" company: Power dependence and status in adolescent susceptibility to peer influence. Social

Psychology Quarterly, 74(3), 310-332. Retrieved from http://search.proquest.com/docview/908976351?accountid=37714

COGNITIVE AND SOCIAL INFLUENCES ON GANG

INVOLVEMENT AMONG DELINQUENTS IN THREE
CHINESE CITIES
Ngan-pun NgaiAuthor Information ; Chau-kiu CheungAuthor Information ; Steven Sek-Yum NgaiAuthor Information
. Adolescence 42.166 (Summer 2007): 381-403.

1. Full text
2. Full text - PDF
3. Abstract/Details
4. References 101
Hide highlighting
Abstract
TranslateAbstract

Inasmuch as research has held the increase in youth gang activities, responsible for the escalating level of crime and

delinquency in Chinese societies, ascertaining risk or protective factors of gang involvement among Chinese youthsis

crucial. The factors include those associated with social control, social learning, and cognitive development. To investigate

these factors across different sociocultural contexts, we surveyed delinquents in three diverse Chinese cities-Hong Kong,

Guangzhou, and Shanghai. Because of the sociocultural differences among the three cities, social control by parents and

cognitive development are likely to be greater protective factors in terms of youths' gang involvement in Hong Kong, while

social control by teachers is likely to be a greater protective factor in mainland cities. Results bear out these hypotheses.

[PUBLICATION ABSTRACT]

Full text
TranslateFull text
Turn on search term navigation
Headnote

ABSTRACT

Inasmuch as research has held the increase in youth gang activities, responsible for the escalating level of crime and

delinquency in Chinese societies, ascertaining risk or protective factors of gang involvement among Chinese youthsis

crucial. The factors include those associated with social control, social learning, and cognitive development. To investigate

these factors across different sociocultural contexts, we surveyed delinquents in three diverse Chinese cities-Hong Kong,

Guangzhou, and Shanghai. Because of the sociocultural differences among the three cities, social control by parents and

cognitive development are likely to be greater protective factors in terms of youths' gang involvement in Hong Kong, while

social control by teachers is likely to be a greater protective factor in mainland cities. Results bear out these hypotheses.
Involvement in gang activities is a critical factor in the continuation and escalation of crime and delinquency among youth

(Hill et al., 1999). In China, researchers have held that gang involvement is substantially responsible for the growth of youth

crime and delinquency, noting that a great deal of crime involves youth gangs (Xi, 2001). Gang association is also

prevalent in the United States where about one-third of youth offenders in cities are associated with gangs (Yoder et al.,

2003). Thus gang involvement is an important area of concern in preventing delinquency and crime. Identifying and

understanding why young people participate in gang activities is a precondition for combating crime. Such understanding is

particularly necessary in Chinese societies, which have witnessed a dramatic growth in youth crime and delinquency (Xi,

2001; Ngai, 1994; Zhang et al., 1993). This trend is likely to persist because of the ongoing structural transformation such as

geographic mobility caused by unequal levels of economic development within China (Jin 1994; Ma et al., 1986; Zhang et

al., 1993).

Western research and theory about young people's gang involvement may not be readily applicable to Chinese societies.

Even within China, there are differences in sociocultural contexts in different parts of the country. These regional differences

suggest a need to collect data from diverse Chinese societies in investigating gang involvement. Hong Kong, Guangzhou,

and Shanghai comprise such a set of Chinese cities. The most obvious difference among these cities is in their history. The

colonial past of Hong Kong distinguishes it from the socialist history of Guangzhou and Shanghai. While Guangzhou and

Shanghai are on the mainland of China, each has its own culture, manifested in dialect, art, and drama. Guangzhou and

Hong Kong are at the southern tip of China, which is far from the Yangzi River Delta where Shanghai is located.

A gang has been denned as a group oriented and committed to anti-social, deviant, and criminal activities (Kennedy &

Baron, 1993). Typically, youths understand what a gang is and are able and willing to become involved (Yoder et al., 2003).

A gang usually recruits its members on the streets. While delinquent youths can freely form their own gangs, many Chinese

youths prefer to join gangs affiliated with the triad society (Che, 1992; Wong, 1999). The triad society has existed in China

for several hundred years, initially as an underground political association formed to revolt against rulers of the ethnic

minority. Gradually, the triad society has developed into enormous organizations that are present in most Chinese societies,

including overseas communities. The triad society, headed by adults, is keen on enlisting young members to sustain its

criminal interests. At the same time, delinquent youths are willing to join the society when they find it to be powerfully

protective. Even where they do not join the triad society formally, many delinquent youths claim to be members and name

their gangs as branches of the society. Because of the difficulty in discerning genuine triad gangs, from those that are

bogus, the police strive to dissolve all Chinese youth gangs. Rather than waiting for young gangsters to emerge, inhibiting

youth involvement is perhaps a more effective way to combat crime. The present study aspires to identify and reinforce the

inhibiting factors.

Research and Theory

Factors inhibiting delinquent youths' involvement would stem from those postulated in social control theory, social learning

theory, and cognitive development theory, which have been useful in explaining delinquency inChinese contexts (Wong,
1999; Wong, 2001; Zhang & Messner, 1996) as well as Western contexts (Benda & Corwyn, 2001; Ennett et al. 1999;

Kaplan & Liu, 1994). Social control theory refers to the informal, normative control by parents, family, school, work

organizations, helping professions, and other noncoercive and conventional institutions in preventing delinquent or criminal

activity (Warr, 1993; Rosenfield et al., 2001). It posits that involvement, attachment, and commitment to informal and

normative institutions and their activities lead youths to endorse moral, normative beliefs that are antithetical to any criminal,

delinquent, or gang persuasion (Benda & Corwyn 1997; Elliott & Menard, 1996). The essence of social control is the

individual's awareness of parental or other conventional control, attachment to social workers, teachers, or other helping

professionals, and acquisition of moral beliefs (Benda, 1999; Elliott & Menard, 1996; Hagan & McCarthy, 1997; Junger &

Marshall, 1997; Vega et al., 1993). Research findings concerning gang involvement that are consistent with social control

theory are those related to the negative effects of attachment to neighborhood, dropping out of school, unemployment, and

being in a family with legal problems (Hill et al., 1999; Klein, 1997; Yoder et al., 2003).

Social learning theory posits that delinquent behavior is a result of learning from the activities and messages of others or

from the media (Felson, 1996; Tracy & Kemef-Leonard, 1996). These activities and messages serve as either direct models

or reinforcement of behavior (Akers 1998). Friends are important models from whom youth learn delinquent behavior, and

friends' approval of that behavior reinforces the delinquency (Bao et al., 2000; Whitbeck & Hoyt, 1999). In research

on gang association, social learning theory reveals the effects of association with delinquent peers-use of subcultural

symbols and attire, and exposure to violent media (Klein, 1997; Yoder et al., 2003).

Cognitive developmental theory regards inadequate cognitive development as a cause of delinquent involvement. Cognitive

development manifests itself in one's ability to process and analyze information inorder to formulate solutions to problems

(Husband & Piatt, 1993; Okwunnabua & Duryea, 1998). That is, the ability to think about social as well as individual

problems would prepare one to resist problematic social influence and act in a socially desirable way. Very often,

delinquency results from the inability to cope with a problem. The crux of assistance in cognitive development is its support

for the individual to think about alternative, socially desirable means of solving a problem (Houston, 1998). This ability

involves empathy, perspective taking, and ethical reasoning, which are cognitive guides of one's behavior (Raine 1993). As

such, delinquency arises out of ignorance of others and society in general. This ignorance is reflected inpreoccupation with

self which can lead to impulsive pursuit of one's self-interest by delinquent means (Gibbs, 1991). Ignorance also can

result in anger and quarrels with, and even violence against others based on failure to understand and tolerate others'

actions. Cognitive development functions to regulate that impulsivity and propensity for delinquency (Bickley & Beech,

2002). Understanding or even thinking about social problems would be an indicator of cognitive development (Alter & Egan,

1997; Cheung, 1997, 1998, 1999; Weast, 1996). In research on gang involvement, factors compatible with cognitive

development theory include dropping out of school, and the age at which youth are unsupervised (Yoder et al., 2003).

Research in Chinese contexts has lent support to the explanations of social control theory, social learning theory, and

cognitive development theory for a youth's delinquency. In support of social control theory, past findings have suggested that

youths' adherence to Chinese culture, intergenerational harmony, attachment to school, and integration with society can
predict their delinquency (Wong, 1999; Wong, 2001; Zhang & Messner, 1996). Similarly, Chinese children whose parents

exert greater control over them are less likely to be involved in delinquency (Nagasawa et al., 2001; Wang et al.,

2002). In contrast, a Chinese youth associated with delinquent peers is at risk of delinquency (Ma et al., 1996, 2000).

Cognitive development has also emerged as an inhibitor of delinquency via the effects of academic achievement and

problem-solving ability (Chen et al., 1997; Wang et al., 2002).

Differences in Hong Kong, Guangzhou, and Shanghai

The above discussion leads to the formulation of a number of hypothesized predictors of youths'

expected gang involvement: parental control, attachment to teachers and social workers/counselors, moral beliefs, friends'

moral beliefs, and theorizing about social problems. These effects tend to vary among Chinese societies because of

differences in sociocultural contexts.

There are remarkable differences between the capitalist, Westernized, and individualist system of Hong Kong and the

socialist and collectivist tradition of Guangzhou and Shanghai on the Chinese mainland. The three cities' different histories

have shaped their socioeconomic development differently. On the other hand, the cities have in common that they are large,

densely populated, and leading the eco- nomic development inChina (Gu, 1999). Shanghai plays a leading role in the

development of Mainland China (Zhang, 2002). It is rapidly catching up with Hong Kong in economic development and its

rate of growth is unparallelled anywhere in the world (Zhu & Yuan, 2001). Although Shanghai is a relatively new

city in Chinese terms, it has been the industrial and commercial center of China for a thousand years. Its prosperity has

been historically attributed to the movement of wealth and talent from north of the Yantzi River to the south, due to repeated

wars and disturbances in the north (Jin & Liu, 1992). Moreover, like Hong Kong and Guangzhou, Shanghai is a port city, and

has been an important conduit for trade since the mid- 19th century. Guangzhou is the oldest city of the three, and has been

a significant socioeconomic center for about 1,200 years. Following the disintegration of China during the time of rapid

changes in dynasties, Guangzhou gradually evolved as the political and cultural stronghold in the south, which supported

the revolutionary forces against the despotic administration in the north. Thus, whereas Shanghai is famous for its industrial

and commercial success, Guangzhou is proud of its revolutionary heritage. In sharp contrast to both, Hong Kong's

development began only 150 years ago, and for most of its short history, it has been a colony of Britain. It has obviously had

the greatest exposure to Western culture among the three cities, which is what most sharply differentiates it from the two

mainland cities.

Concomitant with the Western influence in Hong Kong is its emphasis on individualism and the nuclear family, as opposed

to the two mainland cities with their collective emphasis on the extended family, clanship, and neighborhood organizations

(Chan & Lee, 1995). In an individualist society, people rely on themselves and their parents. Support from the extended

family and neighborhood organizations significantly reduces the burden on parents (Chen et al., 2000; Jessor et al., 2003).

Moreover, socialist policy on the mainland is to promote productivity through full employment, including that of mothers (Lu

et al., 2000). Mainland China therefore discourages parents from spending too much time on household tasks (Lu et al.,
2000). The role of parents is therefore less significant than in Hong Kong, which emphasizes the responsibility of the

individual and the family (Wong, 1995). While the role of parents is weaker on the mainland, that of teachers and other

professionals tend to be strong, due to its authoritarian tradition (Leung & Nann, 1995; Zheng, 1994). Despite their low

income, teachers enjoy high status in Mainland China. The role of teachers in Hong Kong has more to do with preparing

students for examinations (Moneta & Siu, 2002). The influence of teachers on students' moral development is

weaker in Hong Kong than in Mainland China.

The sociocultural differences are therefore likely to modify the hypotheses derived from social control and cognitive

development theories as follows:

1. Parental control has a stronger negative effect on youths' expected gang involvement in Hong Kong than in Guangzhou

and Shanghai.

2. Attachment to teachers has a stronger negative effect on youths' expected gang involvement inGuangzhou and Shanghai

than in Hong Kong.

3. Attachment to social workers/counselors has a stronger negative effect on youths'

expected ganginvolvement in Guangzhou and Shanghai than in Hong Kong.

4. Theorizing about social problems has a stronger negative effect on youths' expected gang involvement inHong Kong

than in Guangzhou and Shanghai.

METHOD

Data were collected by individual face-to-face interviews with 229 delinquent youths in Hong Kong between April 28 and July

15, 1999, 312 delinquent youths in Guangzhou between March 29 and June 26, 1999, and 297 delinquent

youths in Shanghai between June 24 and October 15, 1999. The delinquent youths were users of various youth service

units, which represented coverage of all services working with delinquent youths invarious cities (i.e., 24 outreaching social

work teams in Hong Kong, correctional schools for juvenile offenders, and vocational schools for young

offenders in Shanghai and Guangzhou). Social workers or counselors of the service units helped recruit delinquent youths

for the study.

The average age of the youths in the three cities was 15.8 years, and variation among the cities was not significant.

However, delinquent youths in Shanghai were disproportionately and significantly more likely to be between 12 and 14 years

of age (29.7%), whereas those in Guangzhou were less likely to be in this age range (11.8%). Of the surveyed sample, the

delinquent youths in Guangzhou and Shanghai were predominantly male, while only about two-thirds (68.9%) in Hong Kong

were male.
There was a significant difference in the number of years of formal education among delinquent youths in the three cities.

Those in Hong Kong had the highest level of education and those in Guangzhou had the lowest (9.37 vs. 6.20). A

disproportionately larger number of delinquent youths in Guangzhou were on an educational level below senior secondary.

The great majority (85.4%) in Shanghai were on a junior secondary educational level. The youths in the three cities also

significantly differed in their fathers' and mothers' education level. Parental education of delinquent youths was

highest in Shanghai and lowest in Guangzhou (see Table 1). Substantial proportions of the fathers (57.5%) and mothers

(49.6%) of the youths in Hong Kong did not have an education above the primary level. In contrast, a higher proportion

(48.5% and 54.7%) of parents inShanghai had attained a junior secondary education level.

Enlarge this image.

Table 1: Means and their differences among the three cities

There was no significant difference in student status among delinquent youths in the three cities (see Table 1). About one-

fifth (20.2%) were students. There was a significant difference in the number of jobs that the youths had held. The

delinquent youths in Hong Kong had worked at more jobs (1.46) prior to the time of the survey. Moreover, there appeared to

be a significant difference among delinquent youths in the three cities in the number of years of residence.

Those inGuangzhou tended to have been in the city for the shortest time (4.8 years, see Table 1); only 41.9% of

those inGuangzhou were natives of the city.

Measurement of moral belief, parental control, attachment to social workers/counselors, attachment to teachers, theorizing

about social problems, and friends' moral belief involved multiple indicators included in the survey questionnaire (see

Appendix). Each indicator employed a five-point rating scale to record the intensity of each factor experienced in the past six

months. The response scores ranged from 0 to 100, with 100 representing the fifth point on the response scale (the greatest

extent) and 0 (the lowest extent). The average of the indicators were the composite scores representing their respective

concepts. The composite scores had favorable reliability based on estimation of the reliability alpha coefficient (.558-.917,

see Table 2).

A measure of prior gang involvement was responses to the question: "How many times did you

participate ingang activities in the past month?" The self-report measure has proven to be adequate in tapping

youths' gang involvement (Yoder et al., 2003).

A measure of expected gang involvement was responses to the question: "How likely is it that you will

participate in gang activities in the coming month?" The responses generated scores ranging from 0 to 100, with 0 for 'Very

unlikely," 25 for "rather unlikely," 50 for "so-so," 75 for "rather likely," and 100 for 'Very likely."

Enlarge this image.

Tabic 2: Reliability alpha coefficients of composite scores

The average of all five-point rating items measured the concept of acquiescence, denned as the tendency to rate all items

uniformly high (or low) regardless of the content of the items (Zagorski, 1999). It was a method artifact which created

meaningless relationships among variables.

The analysis relied on LISREL, especially for testing the equality of regression coefficients between and across cities

(Joreskog & Sorbom, 1993). These tests examined whether increases in likelihood ratio chisquare due to equality

constraints were significant.

Analysis isolated the effects of social control, social learning, and cognitive development by controlling for a basic set of

potential confounding variables, including the youth's prior gang involvement, sex, age, education, student status, residential

stability, acquiescence, and parental education. The effects of the confounding variables would reflect alternative influences

due to gender-role socialization (Brack & Brack 1994), education (Misra, 1997; Pascarella et al., 1996), maturation (Winters

et al., 1995), and inertia or continuation of past ganginvolvement (Tracy & Kemef-Leonard, 1996). Moreover, since the effect

of age tends to be curvilinear (Haynie, 2001; Menard, 1992), the square of age is an additional control variable.

RESULTS

Youths in Hong Kong, Guangzhou, and Shanghai differed significantly on all factors examined in the study, except for age

and student status (see Table 1). Essentially, expected and prior gang involvements were highest in Hong Kong. On the

other hand, Hong Kong was the lowest on the youths' moral belief, parental control, attachment to teachers, theorizing about

social problems, and friends' moral belief. Attachment to social workers/counselors was highest in Hong Kong and

lowest in Shanghai. This finding reflects the greater severity of the problem among Hong Kong youths, which

results in higher levels of contact with and attachment to social workers/counselors. Apparently, greater attachment results

from more contact with the professionals, in line with contact theory (Morgan & Streb, 2001).

Effects on Expected Gang Involvement

Prior gang involvement and moral belief were predictors that consistently showed significant effects on

youths' gang involvement in Hong Kong, Guangzhou, and Shanghai (see Table 3). The effect (.510) of

prior gang involvement was particularly strong among youths in Guan- gzhou. This effect reflects the stability

of gang involvement. Nevertheless, the stability was not overwhelmingly high, suggesting that factors other than inertia are

required to explain expected gang involvement.

Enlarge this image.

Table 3: Standardized effects on expected gang

Supporting both social control theory and social learning theory, moral belief had the most consistent effect on youths'

expected gang involvement, among the theoretical factors. This internalized factor was more influential on the youth in Hong
Kong than on those in the mainland cities. This finding endorses the relatively higher importance of individualistic

influence inHong Kong.

The other social control factors did not consistently show negative effects on the youths' expected gang involvement in the

three cities. Whereas parental control was a significant inhibiting factor in Hong Kong ( = -.136), attachment to teachers

was a significant inhibiting factor in Guangzhou ( = -.144). In contrast, attachment to social workers/counselors tended to

be more protective in Shanghai than in Hong Kong and Guangzhou. Thus, each of the socializing agents played a different

role in preventing the youths' expected gang involvement inthe three cities. Regardless of the nature of the socializing agent,

social control from parents, teachers, or social workers/counselors had a weak negative effect on gang involvement ( = -

.050 - -.065). Moreover, the influences of the three socializing agents were not significantly different, despite the apparent

variation (see the last row in Table 3).

Friends' moral belief showed a significant negative effect on expected gang involvement in Hong Kong and Shanghai, but

not in Guangzhou. In Hong Kong and Shanghai, the negative effect of friends' moral belief was comparable to that of the

youths' own moral belief. Theorizing about social problems had a significant negative effect on expected gang involvement

only in Hong Kong. This finding lends support to the hypothesis that personal factors are more influential in the individualist

context;

Females were significantly less likely to participate in gang activities only in Hong Kong ( = - .144), after controlling for

prior gang involvement and other background characteristics. Expected gang involvement was significantly lower in older

youths in Hong Kong and Guangzhou. However, involvement significantly escalated particularly among the older

youths in Hong Kong. The relationship between age and expected gang involvement took the form of a U-shaped

curve in Hong Kong. In Hong Kong, expected ganginvolvement was also significantly higher among youths with longer

residence in the city and whose mothers were higher in education. By contrast, youths in Shanghai had significantly higher

expected ganginvolvement when their education was higher or when their fathers' education was lower. These background

characteristics displayed no significant effect on youths' expected gang involvement in Guangzhou.

Common Effects on Expected Gang Involvement and Their Variations Across Cities

Prior participation in gang activities was the strongest predictor of youths' expected gang involvement in the three cities ( =

.347, see Table 4). However, the effect varied significantly across the three cities. Thus, prior influence was the

strongest in Guangzhou and the weakest in Shanghai, suggesting differential stability ingang involvement in the three cities.

The factors of moral belief, attachment to teachers, theorizing about social problems, and friends' moral belief showed

significant common effects on youths' expected involvement. This supports the effects of social control, social learning, and

cognitive development. The effects of theorizing about social problems and friends' moral beliefs were significantly different

among the three cities. The differential effects supports hypotheses derived from the difference between individualist and

collectivist contexts. Notably, theorizing about social problems was more protective, and attachment to teachers was less
protective for youths inHong Kong, where individualism is stronger. However, other social control factors, including parental

control and attachment to social workers/counselors did not show common significant influences on

expected ganginvolvement. On average, social control from parents, social workers/counselors, and teachers was

significantly negative ( = -.058). This effect was not significantly different among the three cities.

Enlarge this image.

Table 4: Equalized standardized effects on expected gang

Sex, age, duration of residence in the city, and education made significant differences in the youths'

expected ganginvolvement. The effects of sex and age, nevertheless, significantly varied across the three cities.

Furthermore, significant variation occurred in the effect of mother education, which significantly affected youths'

expected ganginvolvement only in Hong Kong.

DISCUSSION

The expected effects of social control, social learning, and cognitive development are sustainable in this study of delinquent

youths in three Chinese cities. Notably, moral belief, friends' moral belief, attachment to teachers, and theorizing about

social problems exhibited significant common effects on the youths' expected gang involvement. Social control from parents,

teachers, or social workers/counselors, on average, was significantly negative. Furthermore, the hypotheses concerning

differentials in the effects between Hong Kong and Guangzhou and Shanghai received considerable support. In support of

Hypothesis 1, parental control was a significant inhibitive factor of expected gang involvement only in Hong Kong.

Consistent with Hypothesis 2, attachment to teachers was a significant inhibitive factor inGuangzhou but not in Hong Kong.

Supporting Hypothesis 4, theorizing about social problems was a significant inhibitive factor only in Hong Kong. Moreover,

the effects of theorizing were significantly different among the three cities. In line with Hypothesis 3, the effect of attachment

to social workers/counselors on expected gang involvement was negative in Guangzhou and Shanghai, whereas it was

positive in Hong Kong. These effects, however, were not significant. Hypothesis 3, therefore, does not receive adequate

support.

The findings largely support the generality of the protective function of social control, prosocial social learning, and cognitive

development on Chinese delinquent youths' gang involvement in terms of the common effects for the three cities. Moreover,

they support the differences in the effects due to sociocultural differences, notably between the capitalist-individualist and

socialist-collectivist contexts in Hong Kong and mainland cities. In the capitalist-individualist context of Hong Kong, individual

and parental influences are crucial, even though parents exert less control there than in mainland cities. By contrast,

teachers and social workers/counselors are more authoritative in Mainland China (Chu, 2001; Luk, 2001) and as such are

more protective in terms of youths' gang involvement in Guangzhou and Shanghai. In the socialist-collectivist context,

although control from the nuclear family is less crucial, professional and community support is more helpful. In addition, the

individual's cognitive development and moral belief are less protective against ganginvolvement in the collectivist context.
Whereas prior gang involvement had the greatest effect on expected gang involvement in Guangzhou, friends' moral belief

displayed the lowest effect there. Apparently, stability or inertia is the best explanation for gang participation and prosocial

social learning is irrelevant in Guangzhou. These findings may be attributable to the especially low proportion of girls, the

shorter duration of residence, and the lower parental education on average among delinquent youths

identified in Guangzhou. Apparently, delinquent youths inGuangzhou tend to be boys with a lower socioeconomic

background who have been residing in the city for a short time. They may not have had time to develop trusting friendships

which affect their gang involvement (White & Glick, 1999). Instead, their delinquent habit and attachment to teachers are

stronger factors affecting their gang involvement. This finding echoes the view that teachers help young migrants settle

down in the city and avoid delinquency (Goyette & Conchas, 2002; Nagasawa et al., 2001). In the absence of other effective

sources of social control, teachers are the prime agents for young migrants (Nagasawa et al., 2000). Without them, migrants

are prone to continue their habits, including clustering in ethnic enclaves and gangs (Nee & Sanders, 2001; Sampson &

Laub, 1993).

In contrast, education showed a significant positive effect, and father education had a significant negative effect on

expected gang involvement of youths in Shanghai. This may reflect the influence of parental education, which is the highest

among delinquent youth in Shanghai. The high influence of parental education underscores the importance of

education in Shanghai, which is a cultural hub in China and attracts people with higher education (Yusuf & Wu, 2002; Zhu &

Yuan, 2001). In such a city, paternal education would accomplish its greatest protective function by facilitating effective

social control over youths. However, education in Shanghai and in all three Chinese cities in general might empower the

youth to resist social control (De Haan & Schulenberg, 1997). This is evident in the prevalence of participation in social

movements among college students (Li & Song, 1992; Sherkat & Blocker, 1994). Gang involvement can thus be a side

effect of education. The concentration of intellectuals in Shanghai probably exacerbates this effect.

In Hong Kong, maternal education and duration of residence in the city were two distinctive positive predictors of youths'

expected gang involvement. The positive effect of maternal education may reflect inadequate social control, as mothers with

higher education spend more time at work and social activities (Hanlon, 1986). Inadequate maternal control may be more

conducive to youths' gang involvement in Hong Kong as there are few alternative sources of social control in an individualist

society. Alternatively, a delinquent youth with longer residence in Hong Kong is more likely to participate in gang activities

because of a climate that is conducive to gang involvement, as shown in the fact that it has the highest rate of involvement

among the cities. Simply put, a context of prevalent gang involvement would foster youths' ganginvolvement over time

(Klein, 1997).

Further Research

It would be desirable to use a longitudinal design in future research to confirm the causal effects on ganginvolvement over

time. To guarantee the generalizability of the effects, a larger sample of diverse societies is necessary. Such a sample

would shed further light on the differential effects due to societal conditions. Further research, using an enhanced design
and a wider sample is necessary, because our retrospective design cannot verify the causal effects on gang involvement

and because we use the measure of expected ganginvolvement, as opposed to observed gang involvement. Further

research is required to investigate involuntary gang involvement, given that youth are vulnerable to pressure

to join gangs (Klein, 1997; Yoder et al., 2003).

The three selected Chinese cities are not broad enough to allow a thorough investigation of the effects of different

sociocultural contexts. To do this requires a larger sample of societies. This would afford examination of how the availability

of alternative sources of social control in Mainland China affects the social control function. Such examination should

preferably control for background differences among youth in different places. The best approach would be a sampling

design that matches youths with similar characteristics. Because youths in different cities are unlikely to have exactly

matched background characteristics, counterfactual matching in analysis would be a desirable alternative (Morgan, 2001).

The logical approach is to select cases of similar background characteristics from each context for anal- ysis. This would

also control for the selection effect of the length of youths' residence in a city. To compare the effect of residence in Hong

Kong, for instance, the approach needs to compare youths in Hong Kong with their counterparts with similar characteristics

but living in another place. This approach, proposed for future research, would require a sizable database for matching.

The influence of social learning and cognitive development also requires elaboration in further research. A number of social

learning factors are missing in the present study, including peer influence and teaching, and youths' acceptance of and

learning from their peers or friends and applying what is learned (Hagan & McCarthy, 1997). The learning would particularly

reflect the value and rewards of gang involvement. Reinforcement and punishment from peers are crucial factors bolstering

social learning (Akers, 1998). Similarly, social learning also includes the influence of exposure to mass media (Huesmann et

al., 1992). Further research could investigate the effects of alternative indicators of cognitive development, including

cognitive sophistication, perspective-taking, rational thinking, moral reasoning, and problem-solving skills.

Implications

Moral beliefs, friends' moral beliefs, social control from parents, teachers, and social workers/counselors ingeneral, and

theorizing about social problems are generally protective against youths' gang involvement. Promoting these factors is

therefore helpful in cutting off the path to crime and delinquency. However, these are variations in the negative effects of

these factors due to the differences in sociocultural contexts. Whereas parental control is an effective protective

measure in a capitalist-individualist society, attachment to authoritative professionals is more significant in a socialist-

collectivist society. Bolstering social integration and cohesion in the neighborhood would be an effective way to

deter gang involvement in a collectivist society. However, the growth of high-rise residential buildings tends to diminish

neighborhood integration (Adamson, 1998) and is therefore a challenge to the prevention

of gang involvement in cities in modern China, which are undergoing urban redevelopment (Zhang, 2002). In contrast, an

individualist society can capitalize on the significance of individual and family responsibilities to resist gang influence. In a

collectivist society, cognitive development in terms of theorizing about social problems is less protective than it is in an
individual society. Hence, cognitive development plays a lesser role than expected (Xi,

2001) in preventing gang involvement in China. Moreover, education can be a risk factor for delinquent

youths' gang involvement because higher education leads to lower social control rather than better cognitive development.

Similarly, higher education in the mother in an individualist society can weaken social control over youths' ganginvolvement.

Hence, education is not necessarily a protective factor.

References

REFERENCES

Adamson, C. (1998). Tribute, turf, honor, and the American street gang: Patterns of continuity and change since 1820.

Theoretical Criminology, 2(1), 57-84.

Akers, R. L. (1998). Social learning and social structure: A general theory of crime and deviance. Boston, MA: Northwestern

University Press.

Akers, R. L., & Gang, L. (1999). Age, social learning, and social bonding in adolescent substance use. Deviant Behavior, 19,

1-25.

Alter, C, & Egan, M. (1997). Logic modeling: A tool for teaching critical thinking in social work practice. Journal of Social

Work Education, 33(1), 85-102.

Bao, W., Whitbeck, L. B., & Hoyt, D. R. (2000). Abuse, support and depression among homeless and runaway adolescents.

Journal of Health & Social Behavior, 41, 408-420.

Bellair, P. E., & Roscigno, V. J. (2000). Local labor-market opportunity and adolescent delinquency. Social Forces, 78(4),

1509-1538.

Benda, B. B. (1999). Testing the problem syndrome among young males in boot camp: Use of theoretical elaboration with

reciprocal relationships. Social Work, 23(1), 28-41.

Benda, B. B., & Corwyn, R. F. (2001). Are the effects of religion on crime mediated, moderated, and misrepresented by

inappropriate measures? Journal of Social Service Research, 27(3), 57-86.

Bickley, J. A, & Beech, A. R. (2002). An investigation of the Ward and Hudson pathways model of the sexual offense

process with child abusers. Journal of Interpersonal Violence, 17(4), 371-393.

Brack, C. J., & Brack, G. (1994). Dimensions underlying problem behaviors, emotions, and related psychosocial

factors in early and middle adolescents. Journal of Early Adolescence, 14(3), 344-359.

Cao, Z., & Wang, S. (1993). Summary of studies of causes of crime in China. Qinhuangdao, China: China Political Legal

University.
Ceci, S. J. (1996). On intelligence: A bioecological treatise on intellectual development. Cambridge, MA: Harvard University

Press.

Chan, H., & Lee, R. P. L. (1995). Hong Kong families: At the crossroads of modernism and tradionalism. Journal of

Comparative Family Studies, 24(1) 83-99.

Che, W. (1992). Problems of juvenile delinquency in Hong Kong. Hong Kong: China Book.

Chen, X., Rubin, K H., & Li, D. (1997). Relations between academic achievement and social adjustment: Evidence from

Chinese children. Developmental Psychology, 33(3), 518-525.

Chen, X., Liu, M., Li, B., Cen, G., Chen, J., & Wang, L. (2000). Maternal authoritative and authoritarian attitudes and mother-

child interactions and relationships in urban China. International Journal of Behavioral Development, 24(1), 119-126.

Cheung, C. (1997). World understanding and well-being in a marital context. Journal of Genetic Psychology, 158(1), 41-58.

Cheung, C. (1998). Sophistication in theorizing about social problems as a condition for the good life. Genetic, Social,

General Psychology Monographs, 124(3), 353-374.

Cheung, C. (1999). Variation in structuralist and individualist explanations among classes in Hong Kong. Sociological

Spectrum, 19(1), 93-118.

Chu, C. (2001). The changing concept of Zhong (loyalty): Emerging new Chinese political culture. In S. Hua (Ed.), Chinese

political culture, 1989-2000 (pp. 42-69). Armonk, NY: Sharpe.

Crutchfield, R. D., & Pitchford, S. R. (1997). Work and crime: The effects of labor stratification. Social Forces, 76(1), 93-118.

De Haan, L. G., & Schulenberg, J. (1997). The covariation of religion and politics during the transition to young adulthood:

Challenging global identity assumptions. Journal of Adolescence, 20, 537-552.

Eisenberg, N., Carlo, N., Murphy, B., & van Court, P. (1995). Prosocial development in late adolescence: A longitudinal

study. Child Development, 66, 1179-1190.

Elliott, D. S., & Menard, S. (1996). Delinquent friends and delinquent behavior: Temporal and developmental patterns. In D.

Hawkins (Ed.), Delinquency and crime: Current theories (pp. 28-67). Cambridge: Cambridge University Press.

Elliott, D. S., Huzinga, D., & Menard, S. (1989). Multiple problem youth: Delinquency, substance use, and mental health

problems. New York: Springer- Verlag.

Ennett, S. T., Bailey, S. L., & Federman, E. B. (1999). Social network characteristics associated with risky behaviors among

runaway and homeless youth. Journal of Health & Social Behavior, 40(1), 63-78.
Felson, R. B. (1996). Mass media effects on violent behavior. Annual Review of Sociology, 22, 103-128.

Gibbs, J. C. (1991). Sociomoral developmental delay and cognitive distortion: Implications for the treatment of antisocial

youth. In W. M. Kurtines & J. L. Gewirtz, Handbook of moral behavior and development, Vol. 3: Application (pp. 95-110).

Hillsdale, NJ: Lawrence Erlbaum.

Gottlieb, N. H., Galavotti, C, McCuan, R. A., & McAlister, A. L. (1990). Specification of a social-cognitive model predicting

smoking cessation in a Mexican-American population: A prospective study. Cognitive Therapy and Research, 14(6), 529-

542.

Goyette, K. A., & Conchas, G. Q. (2002). Family and non-family roots of social capital among Vietnamese and Mexican

American children. Research in Sociology of Education, 13, 41-71.

Gu, Z. (1999). Chinese city geography. Beijing, China: Commercial.

Hagan, J., & McCarthy, B. (1997). Mean streets: Youth crime and homelessness. Cambridge, UK: Cambridge University

Press.

Hagan, J., Merkens, H., & Boehnke, K. (1995). Delinquency and disdain: Social capital and the control of right-wing

extremism among East and West Berlin youth. American Journal of Sociology, 100(4), 1028-1052.

Hagan, J. (1988). Structural criminology. Cambridge: Polity.

Halpern, D. (1996). Changing data, changing minds: What the data on cognitive sex differences tell us and what we hear.

Learning and Individual Differences, 8(1), 73-82.

Hanlon, M. D. (1986). Age and commitment to work: A literature review and multivariate analysis. Research on Aging, 8(2),

289-316.

Haynie, D. L. (2001). Development peers revisisted: Does network structure matter? American Journal of Sociology, 106(4),

1013-1057.

Heimer, K, & Matsueda, R. L. (1994). Role-taking, role commitment, and delinquency: A theory of differential social control.

American Sociological Review, 59, 365-390.

Hernstein, R. J., & Murray, C. (1994). The bell curve: Intelligence and class structure in American Ufe. New York: Free

Press.

Hill, K. G., Howell, J. C, Hawkins, J. D., & Battin-Pearson, S. R. (1999). Childhood risk factors

for adolescentgang membership: Results from the Seattle Social Development Project. Journal of Research in Crime &

Delinquency, 36(3), 300-322.

Houston, J. (1998). Making sense with offenders: Personal constructs, therapy and change. Chichester, TJK: Wiley.

Huesmann, L. R., Eron, L. D., Berkowitz, L., & Chaffee, S. (1992). The effects of television violence on aggression: A reply

to a skeptic. In P. Suedfeld & P. E. Tetlock, Psychology and social policy (pp. 191-200). New York: Hemisphere.

Husband, S. D., & Piatt, J. J. (1993). The cognitive skills component in substance abuse treatment incorrectional settings: A

brief review. Journal of Drug Issues, 23(1), 31-42.

Jang, S. J. (1999). Age-varying effects of family, school and peers on delinquency: A multilevel modeling test of interactional

theory. Criminology, 37(3), 643-685.

Jessor, R., Turbin, M., Costa, F. M., Zhang, Q. D. H. C, & Wang, C. (2003). Adolescent problem behavior in China and the

United States: A cross-national study of psychosocial protection factors. Journal of Research inAdolescence, 13(3), 329-

360.

Jin, G., & Liu, Q. (1992). Prosperity and crisis: On the ultra stable structure of Chinese feudal society. Hong Kong: Chinese

University Press.

Jin, Z. (1994). The People's Republic of China. In K Hurrelmann (Ed.), International Handbook of Adolescence. Westport,

CT: Greenwood.

Jones, R. (1987). Like distant relatives: Adolescents' perceptions of social work and social workers. Aldershot: Gower.

Joreskog, K G., & Sorbom, D. (1993). LISREL 8 User's Reference Guide. Chicago, IL: Scientific Software.

Junger, M., & Marshall, I. H. (1997). The interethnic generalizability of social control theory: Empirical test. Journal of

Research in Crime and Delinquency, 34(1), 79-112.

Junger, M., & Polder, W. (1993). Religiosity, religious climate, and delinquency among ethnic groups in the Netherlands.

British Journal of Crimonology, 33, 416-435.

Kaplan, H. B., & Liu, X. (1994). A longitudinal analysis of mediating variables in the drug use-dropping out relationship.

Criminology, 32(3), 415-439.

Kennedy, L. W., & Baron, S. W. (1993). Routine activities and a subculture of violence: A study of violence on the street.

Journal of Research in Crime and Delinquency, 30(1), 88-112.

Klein, M. W. (1997). The American street gang: Its nature, prevalence, and control. New York: Oxford University Press.

Leung, J. C. B., & Nann, R. C. (1995). Authority and benevolence: Social welfare in China. Hong Kong: Chinese University

Press.
Li, J., & Song, D. (1992). Political consciousness and political behavior of contemporary China's youth. Beijing, China:

China's People Public Order University.

Lu, Z. Z., Maume, D. J., & Bellas, M. L. (2000). Chinese husbands' participation in household labor. Journal of Comparative

Family Studies, 31(2), 191-215.

Luk, P., & Fong, Y. Y. (2001). Comparing contexts for developing personal and social education in Hong Kong. Comparative

Education, 37(1), 65-87.

Ma, H. K, Shek, D. T. L., Cheung, P. C, Lam, C. O. B. (2000). Parental, peer, and teacher influences on the social behavior

of Hong Kong Chinese adolescents. Journal of Genetic Psychology, 161(1), 65-78.

Ma, H., Shek, D. T. L., Cheung, P., & Lee, R. Y. P. (1996). The relations of prosocial and antisocial behavior to personality

and peer relationships of Hong Kong Chinese adolescents. Journal of Genetic Psychology, 157(3), 255-266.

Ma, J., Huai, Y., Luo, F., Ran, J., & Zhong, Y. (1986). Introduction to China's juvenile delinquency. Beijing, China: Beijing

Yanshan.

Menard, S. (1992). Demographic and theoretical variables in the age-period-cohort analysis of illegal behavior. Journal of

Research in Crime & Delinquency, 29(2), 178-199.

Miller, J., & Brunson, R. K. (2000). Gender dynamics in youth gangs: A comparison of males' and females' accounts. Justice

Quarterly, 17(3), 419-448.

Misra, J. (1997). Teaching stratification: Stimulating interest and critical thinking through research projects. Teaching

Sociology, 25(4), 278-291.

Miyazawa, S. (1993). The enigma of Japan as a testing ground for cross-cultural criminalogical studies. Annata

Internationales de Criminologie, 32, 81-102.

Moneta, G. B., & Siu, C. M. Y. (2002). Trait intrinsic and extrinsic motivations, academic performance, and creativity in Hong

Kong college students. Journal of College Student Development, 43(5), 665-683.

Morgan, S. L. (2001). Counterfactuals, causal effect heterogeneity, and the Catholic school effect on learning. Sociology of

Education, 74, 341-374.

Morgan, W., & Streb, M. (2001). Building citizenship: How student voice in service-learning develops civic values. Social

Science Quarterly, 82(1), 154-169.

Nagasawa, R., Qian, Z., & Wong, P. (2001). Theory of segmented assimilation and the adoption of marijuana use and

delinquent behavior by Asian-Pacific youth. Sociological Quarterly, 42(3), 351-370.

Nagasawa, R., Qian, Z., & Wong, P. (2000). Social control theory as a theory of conformity: The case of Asia/Pacific drug

and alcohol nonuse. Sociological Perspectives, 43(4), 581-603.

Nee, V., & Sanders, J. (2001). Trust in ethnic ties: Social capital and immigrants. In K. S. Cook (Ed.), Trust insociety (pp.

374-392. New York: Russell Sage foundation.

Ngai, N. P. (1994). Youth deviance in China. Pp. 18.1-18.21 in M. Brosseau & Chi-kin (Eds.), China review 1994. Hong

Kong: Chinese University Press.

Okwunnabua, J. O., & Duryea, E. J. (1998). A synthesis of etiology and prevention of drug abuse in youth: Application and

critique of the epidemiologic model. American Journal of Health Studies, 14(1), 31-41.

Pascarella, E., Edison, M., Nora, A., & Hagedorn, L. S. (1996). Effects of teacher organization/preparation and teacher

skill/clarity on general cognitive skills in college. Journal of College Student Development, 37(1), 7-19.

Raine, A. (1993). The psychopathology of crime: Criminal behavior as a clinical disorder. San Diego, CA: Academic Press.

Rosenfeld, R., Messner, S. F., & Baumer, E. P. (2001). Social capital and homicide. Social Forces, 80(1), 283-309.

Sampson, R. J., & Groves, W. B. (1989). Community structure and crime, testing social-disorganization theory. American

Journal of Sociology, 94, 774-802.

Sampson, R., & Laub, J. H. (1993). Crime in the making: Pathways and turning points through life. Cambridge, MA: Harvard

University Press.

Sherkrat, D. E., & Blocker, T. J. (1994). The political development of sixties' activists: Identifying the influence of class,

gender, and socialization on protest participation. Social Forces, 72(3), 821-842.

Tracy, P. E., & Kemef-Leonard, K. (1996). Continuity and discontinuity in criminal careers. New York: Pleunum.

Vander Ven, T. M., Cullen, F. T., Carrozza, M. ?., & Wright, J. P. (2001). Home alone: The impact of maternal employment

on delinquency. Social Problems, 48(2), 236-257.

Vega, W. A, Gil, A. G., Warheit, G. J., Zimmerman, R. S., & Apospori, E. (1993). Acculturation and delinquent behavior

among Cuban American adolescents: Toward an empirical model. American Journal of Community Psychology, 21(1), 113-

125.

Wang, G. T., Qiao, H., Wei, S., & Zhang, J. (2002). Adolescent social bond, self-control and deviant behavior inChina.

International Journal of Contemporary Sociology, 39(1), 52-68.

Warr, M. (1993). Parents, peers, and delinquency. Social Forces, 72(1), 247-264.
Weast, D. (1996). Alternative teaching strategies: The case for critical thinking. Teaching Sociology, 24(2), 189-194.

Whitbeck, L. B., & Hoyt, D. R. (1999). Nowhere to grow: Homeless and runaway adolescents and their families. New York:

Aldine de Gruyter.

White, M. J., & Glick, J. E. (1999). The impact of immigration on residential segregation. In F. D. Bean & S. Bell Rose (Eds.),

Immigration and opportunity: Race, ethnicity, and employment in the United States (pp. 345-372. New York: Russell Sage

Foundation.

Winters, K. C, Stinchfield, R. D., & Kim, L. G. (1995). Monitoring adolescent gambling in Minnesota. Journal of Gambling

Studies, 11(2), 165-183.

Wong, D. S. W. (2001). Pathways to delinquency in Hong Kong and Guangzhou (South China). International Journal of

Adolescence & Youth, 10(1/2), 91-115.

Wong, S. K. (1999). Delinquency of Chinese-Canadian youth: A test of opportunity, control and intergeneration conflict

theories. Youth & Society, 29(1), 112-133.

Wong, T. W. P. (1995). Economic culture and distributive justice. In S. Lau, M. Lee, P. Wan, & S. Wong (Eds.), Indicators of

social development: Hong Kong 1993 (pp. 367-398). Hong Kong: Hong Kong Institute of Asian-Pacific Studies.

Xi, J. (2001). Research reports on China's youth. Beijing, China: China Youth.

Yoder, K. A., Whitbeck, L. B., & Hoyt, O. R. (2003). Gang involvement and membership among homeless and runaway

youth. Youth & Society, 34(4), 441-467.

Yusuf, S., & Wu, W. (2002). Pathways to a world city: Shanghai rising in an era of globalization. Urban Studies, 39(7), 1213-

1240.

Zagorski, K (1999). Egalitarianism, perception of conflicts, and support for transformation in Poland. In S. Svallfors & P.

Taylor-Gooby (Eds.), The end of the Welfare State? Responses to state retenchment (pp. 190-217. London: Routledge.

Zhang, L., & Messner, S. F. (1996). School attachment and official delinquency status in the People's Republic of China.

Sociological Forum, 11(2), 285-303.

Zhang, L., Zhang, L., & Lei, X. (1993). China's juvenile delinquency and justice. Beijing, China: World Knowledge.

Zhang, T. (2002). Urban development and a socialist pro-growth coalition in Shanghai. Urban Affairs Review, 37(4), 475-

499.

Zheng, Y. (1994). Development and democracy: Are they compatible in China. Political Science Quarterly, 109(2), 235-259.
Zhu, J., & Yuan, Z. (2001). Shanghai employment report. Shanghai, China: Shanghai People.

AuthorAffiliation

Ngan-pun Ngai, Ph.D., Professor, Chau-kiu Cheung, Ph.D., Research Assistant Professor, and Steven Sek-Yum Ngai,

Ph.D., Associate Professor, Department of Social Work, The Chinese University of Hong Kong.

The authors thank the Social Science and Educational Panel of the Faculty of Social Science at the Chinese University of

Hong Kong for a grant in support of this research.

Please send requests for reprints to Chau-kiu Cheung, Department of applied Social Studies, City University of Hong Kong,

Tat Chee Avenue, Kowloon, Hong Kong, China. E-mail: cheungck@cuhk.edu.hk

Appendix

Appendix: Items of Measurement

Parental control (Hagan, 1988): How much did your parent do?

Encouragement to study and work hard; knowing the whereabouts of youths; talking with youths; joint activities; advice on

school or work; prohibition against watching certain television programs.

Attachment to social workers I counselors (Jones, 1987): How much did you receive / have / find / do?

Good impression of the counselor; helpfulness of the counselor; acceptance of the counselor's opinion; liking to disclose

personal business to the counselor; seeking help from the counselor.

Attachment to teachers (Miyazawa, 1993): How much did you receive I have I find I do?

Teachers' care; good impression of teachers; acceptance of teachers' guidance; helpfulness of teachers; usefulness of

teacher's help; seeking help from teachers.

Moral belief, disapproval of offenses (Elliot et al., 1989): How much did you disapprove of?

Flirting with women; taking illicit substances; fraud; theft; bullying; vandalism; assault; participation in gangactivity.

Friends' moral belief or disapproval of offenses (Vega et al., 1993): How much did your friends disapprove of?

Rape; taking heroin, opium, or crack; robbery; bullying; assault; vandalism; participation in gang activity.

Theorizing about social problems: How much did you do?

 Finding causes of problems among friends; thinking about causes of problems in the family; finding causes of

problems in school or work; having personal views on social problems; care about causes of social problems; liking to

analyze political problems.

Copyright Libra Publishers Incorporated Summer 2007

Ngan-pun Ngai, Chau-kiu Cheung, & Steven Sek-Yum Ngai. (2007). COGNITIVE AND SOCIAL INFLUENCES ON GANG INVOLVEMENT

AMONG DELINQUENTS IN THREE CHINESE CITIES. Adolescence, 42(166), 381-403. Retrieved from

http://search.proquest.com/docview/195942127?accountid=37714

Where the streets cross the classroom: A study of Latino students'

perspectives on cultural identity in city schools and neighborhood
gangs
Katz, Susan RobertaAuthor Information . Bilingual Research Journal 20.3/4 (Summer 1996): 603-631.

1. Full text
2. Full text - PDF
3. Abstract/Details
Hide highlighting
Abstract
TranslateAbstract

This study is an ethnography of eight middle school students, all immigrants from Central America and Mexico with siblings

or close friends active in gangs.

Full text
TranslateFull text
Turn on search term navigation

Headnote

Susan Roberta Katz(1)

Headnote

Abstract

Headnote

This study is an ethnography of eight middle school students, all immigrants from Central America and Mexico with siblings

or close friends active in gangs. The investigation attempts to solve the puzzle of why these teenagers from marginalized

ethnic communities were increasingly involved in neighborhood gangs and how schooling potentially contributed to that

involvement. Though the students were seeking a positive cultural identity as Latinos, they felt thwarted at school by

negative stereotypes. Suggestions are made that assist educators in maximizing opportunities for youth to achieve a

positive sense of identity and minimize their chances of engaging in self-- destructive behavior.
They put us in ESL classes, and then they forget about us. They call us trouble-makers. They don't care if we succeed.

The gangs give us something we don't get in school. We feel like we belong -Ricardo Garcia, 15, a member of

a gang in Richmond, California.

San Francisco Examiner, July 10, 1994

Front page stories in the Sunday newspaper told of two major weekend events impacting the lives of youth inthe Northern

California city of Bahia. One hopeful article described a peace-making conference of mostly African-American

former gang members inspired by the year-long truce in Los Angeles between the Crips and the Bloods. The second article

depicted a drive-by shooting in the heart of Bahia's barrio, which wounded four innocent victims. The Latino community was

not included in the gang truce. Instead each weekend seemed to announce a new retaliation between the Surefos and the

Nortedos - the Latino parallels of Crips and Bloods.

When I was teaching middle school in 1992-1993, students would regularly come into my classroom first thing every

Monday morning to give their personal accounts of the events in the newspapers. On the Monday after this particular

weekend, Julio told me his older brother was in lockup because of the drive-by. The police had come to his house early

Sunday morning to inform his family that Elvin was in jail but would not reveal the nature of the crime. Twenty-four hours

later, the police returned to report that Elvin, a Norteno, was being held for attempted murder. He had shot five Surenos in a

drive-by shooting; two were dead.

In the beginning of that school year, Elvin was 15 years old and a student in my seventh grade ESL (English-as-a-Second

Language) class. He would speak to me about his gang activity with the Nortenos and confessed he had done drive-bys.

When I asked him how he felt about his actions, he simply responded, "I don't know. I never found out if I killed anyone or

not." He believed his actions were justified since he was retaliating for his homeboys.

Thinking I could be trusted, he showed me his gun at the end of one day in late September. Wanting to honor his trust, I let

him leave school, but not without misgivings. If I reported him, he would get expelled and be out in the streets forever more.

If I did not report him, I would be delivering the message that I condoned the possession of guns at school. Furthermore, as

a friend later warned, I could get fired. After hours of internal struggle, I finally reported his action to the vice-principal, and

Elvin was expelled the next day. Expelled to nowhere, just as I suspected. He hasn't been to school since.

Many months after his expulsion, I saw Elvin at the Cinco de Mayo parade, walking in a single file with his homeboys. When

I called out to him, "Hey, how are you doing?," he just stared right through me. That same afternoon, Suredos fired at

Nortenos while driving cars along the route of the parade, and a week later, Elvin shot five Surenos in revenge. Soon

afterwards he was convicted with a long jail sentence.

Elvin's brother, Julio, had taken a different route. He preferred to participate in "tagging crews," groups that paint graffiti on

walls. Julio cut classes, ran around the in halls, and stole from stores. But he didn't kill.
In a public television program on the LA uprisings, writer John Edgar Wideman wisely noted, "If people just took the time to

talk to these kids, they'd see that they are operating within a very logical framework." I wanted to know: What was the logical

framework of these kids, my own students?

Purpose of this Study

This last question became the focus of my doctoral dissertation: to help solve the puzzle of why teenagers from marginalized

ethnic communities were increasingly involved in neighborhood gangs and how schooling potentially contributed to that

involvement. During ten years of teaching middle school in Bahia from 1984-1994, I noticed that each year more of my

students were becoming turned off to school and attracted to gangmembership. Gangs have existed in these communities

since the 1920s, but in the 1990s children were joining gangs at increasingly younger ages and using weapons of greater

danger. What were these youth seeking that drove them to take such high risks? What forces were operating within school

that pushed them out of school?

In 1992 I was fortunate to have the same group of students assigned to my ESL class for the third consecutive year and,

consequently I got to know them very well. Four of these students - all Central American stopped coming to their assigned

classes in the spring of their eighth grade year. Instead they would show up at odd times and seek me out during my free

period. They started telling me stories - detailed, intimate narratives of their lives away from school. One boy was kicked out

of both parents' homes, fending for himself and his girlfriend. He ended up living with gang members for food and protection.

Each time we had a conversation, I found myself wishing I had a tape recorder with me. I felt that these stories needed to be

heard by other teachers and educators to gain a deeper understanding of why these kids were so alienated from school, a

place which seemed trivial and boring to them in comparison to their neighborhood alive with high drama.

Given my concern, I decided to focus my doctoral dissertation for the Graduate School of Education at UC Berkeley on

issues related to my Latino students' perspectives of schools and gangs. The following year a number of students in my

seventh grade ESL class were coincidentally the younger brothers or sisters of the students I had taught for three

consecutive years. I conducted a year-long ethnographic case study of eight of these students; seven were first- or second-

generation immigrants from Central America and one was from Mexico. Through interviews and observations, I examined

how these students were searching for a positive cultural identity as Latinos at school, but felt thwarted in this search by

negative, stereotypical perceptions by both peers and teachers. Consequently, they chose to look to their communities for

the source of their cultural identity and interacted with neighborhood gangs.

My purpose was to suggest ways educators could maximize opportunities for youth to achieve a positive sense of identity

and possibly minimize the chances for youth to engage in self-destructive behavior. I recognize, however, that schools exist

within the confines of society, and that fundamental change in the lives of our youth can take place only with a deep

reorganization of social structures and values.

Theoretical Framework
Adolescence is a unique stage of life, floating between childhood and adulthood, a time in which human beings

particularly in the West are struggling to determine their place in the sun, often for the first time. Youth of color in a

heterogeneous society, such as the US, face daunting challenges in that they must figure out their place within the

mainstream as well as within their own racial or ethnic group. Their search for identity involves negotiating among three

aspects of identity: social, cultural and ethnic (Ferdman, 1990, p. 192).

In a multiethnic society, all three aspects are tightly interwoven. When the images of success valued by the dominant culture

are not accessible to members of other cultures, these members struggle to discover a positive identity for themselves within

the dominant culture. As a result, they must search within their own ethnic groups or peer subgroups for a positive identity.

These groups may, in turn, be organized in opposition to the images of the dominant culture precisely for the purpose of

creating a positive ethnic identity for its members.

The complexity of an ethnic identity results from the fact that any definition of ethnicity within a heterogeneous society

contains both internal and external meanings. An internal definition of ethnicity includes how people within an ethnic group

perceive themselves as well as other members of the group. For example, how one Nicaraguan family describes another

Nicaraguan family involves an internal definition of ethnicity. In contrast, an external definition includes how people outside

the ethnic group perceive those within the ethnic group (Adams, 1989). How the US media depict Haitian immigrants

involves external definitions of ethnicity. In a society riddled with racial and ethnic discrimination, these two definitions often

come into conflict.

Anthropologists are beginning to critique the common practice of basing the definition of ethnicity exclusively on "cultural

features." As Adams (1989) writes:

Of course, cultural features are characteristic of ethnic groups; but to use them to define an ethnic group is

unworkable in the long run. The reason for this apparent contradiction is that the central issue at stake inidentifying

ethnicities is not a configuration of cultural features, but the reproduction of the internal identity. Incontrast, external

definitions of an ethnicity can only be expressed in terms of external features. The outsider has no alternative but to give an

external, or categorical, definition; he or she cannot ever really "know" the nature of the internal identity that may exist in the

minds of the members (p. 6).

Those outside a particular ethnic group recognize members of that group by their shared "cultural features" (e.g. style of

dress, food, dance). Consequently, outsiders are much more likely to form ethnic stereotypes, resulting from an external

view of group membership. Those within the group have a much more sophisticated and subtle understanding of what

involves membership; they grasp the complexity and mixed feelings about those same cultural features. Internal definitions

center on self-identity, adaptive and mutable according to historical context (Adams, p. 12).

Latino Identity: Conflicts between Internal and External Definitions The identification "Latino" means something very

different in Latin America than it does in the United States. In Latin America, one's identity is foremost national: one is
Salvadoran, Guatemalan, or Mexican. Each national identity embodies a distinct history, geography and culture. Even within

Central America, the difference between Guatemala with 80% indigenous people and Costa Rica with only 5% is striking.

The term "Latino" is too broad to be truly significant.

Migration to the United States changes the ways Latin Americans define themselves and the ways others define them. A girl

growing up in El Salvador may have a strong sense of her national history and land. When she emigrates to the United

States, she is suddenly called "Latina" or "Hispanic" by school officials, government workers, classmates. For the first

time in her life, she becomes a "minority," a term meaning "less than," the victim of racial and ethnic discrimination.

Arriaza (1995) describes how migrating from Guatemala to the US as an adult affected his sense of identity:

But the host culture labeled me "Latino" and pushed me into a larger group of the "brown" population. I resisted this

massifying of my Guatemalan culture, but deep in my mind I already knew that certainly, and after a while, I would end up

appropriating the ascribed values - whatever they were - of being "Latino." I only hoped that I would have the time and the

room to define such values on my terms, and not those designated by the host culture. Although the historical me is

Guatemalan, from my perspective I now am first Guatemalan and then Latino. However, in the school where I teach the

African-American kids define me as Mexican, and my colleagues as "Latin." In some instances I am in the position Indians

are today in Guatemala... I found myself, for the first time ever, on the side of the target culture, in terms of my ethnicity. Not

even in Guatemala I had felt more Guatemalan than I felt in the States for a while. The first years living here I found a sorely

needed refuge in a growing pride for my Spanish and Maya ancestry (pp. 20-21).

In this way, Latino ethnic identity in the US is not just shaped by members of the ethnic group itself; it contains both internal

and external definitions. External definitions are not value-free, but develop within a social and historical context. When

external definitions are created within a heterogeneous society, where all ethnic groups are not equal, those definitions may

be used to categorize people along a hierarchical structure. In this way, external definitions may result in negative

stereotypes.

What happens when those stereotypes, created for historical and political reasons, become commonplace conceptions of

what it means to be "Latino" or "Hispanic"? After the 1992 L.A. uprisings, my students composed their own versions of a

poem entitled, "Just Because I'm Black." A Salvadoran girl wrote the following poem:

Just because I'm Hispanic

I don't eat burritos

I don't eat tacos

I'm not from Mexico.

Just because I'm Hispanic

I don't have an accent

I'm not a gangster

I don't hate other races.

Just because I'm Hispanic

I'm not better than you

You are not better than me

We're not stupid.

Just because I'm Hispanic

I'm not a prostitute

I'm not going to be a prostitute

I'm just a normal person.

(Weinberg, 1992)

The same negative images repeated themselves over and over again in my Latino students' poems. Latinos were

supposedly stupid, gang members, thieves, prostitutes. These were the perceptions which the students believed others had

of them, undoubtedly very painful perceptions to accept. The students were faced with several choices: 1) They could

internalize these negative images of their ethnicity as part of their cultural identity. 2) They could reject these negative

images and build upon the positive associations created within their own ethnic group. 3) They could resist the stereotypes

by turning them on their heads and slapping society's faces with them. They could say, "Yeah, I steal. Wanna see how much

more money I make than you?" Or, "Yeah, I get bad grades. Wanna see how much more I know about life than you?" In this

way, the students created an alternative cultural identity.

In reconstructing a cultural identity at the group level, language is possibly the most important symbolic element involved

(Ferdman, 1990). Language includes not only the ways in which people within a culture speak and write, but also the signs

which they use to communicate (Williams, 1977). In this sense, graffiti represents much more than an act of defacing

property; it is an act of secret communication among group members. When youth "tag" (spray paint their code names on

walls), they are using names for themselves that only group members can understand, creating a sign that means

belonging. This study incorporated this sense of language in identifying how elements of gang culture were manifested

within the classroom.

Research Questions
Given this theoretical context, I identified key spheres involved in the students' construction of their cultural identity: peers at

school, teachers at school, and groups in their neighborhood. I explored the students' own perspectives on peer groups at

school by interviewing them about their friendship choices. I asked them questions about what role ethnicity and language

played in both their selection of friends and in becoming a privileged group at school. I interviewed teachers about their

perceptions of the students and compared these with the students' perceptions of the teachers. I also asked the students

about their views of gangs in their neighborhood and about their associations with these gangs. Finally, I examined what

symbolic elements, such as language and literacy, shape their sense of cultural identity and to what extent those elements

relate to being Latino.

Methodology

School

This study took place in Coolidge Middle School in Bahia School District. Bahia, located in Northern California, is

representative of the current trend in urban education. Of its 64,000 students (1990-91), 75% were from homes in which

languages other than English were spoken and 28% were Limited English or Non-English Proficient (NEP/LEP). Within

Bahia was Coolidge Middle School, with 1400 students, located in a middle-class Asian and European-American

neighborhood. Due to a court order to desegregate in 1983, Coolidge received about 300 students who were bussed in from

both the Latino district and one of Bahia's African-American neighborhoods. Both of these targeted areas were

approximately 30 minutes away by bus.

As evident in Table 1, the Gifted and Talented Education (GATE) Program was predominantly Asian and European-

American; the Special Education Program was disproportionately African-American. Over one-third of the Latino students at

Coolidge were enrolled in either ESL or Special Ed. These figures show why Coolidge was characterized by

educators in Bahia as "apartheid."

Moreover, the ESL Program was housed for ten years until 1992 in portable bungalows, minutes from the main campus.

Teachers in the halls frequently told the ESL students, "Speak English! You're in America!" and their native English-speaking

peers often teased them for their foreign accents or customs. However, when students tried to "exit" the ESL program and

enter the mainstream, they encountered roadblocks. The district's Bilingual Office discouraged moving the students unless

they met the difficult criteria for exiting, based upon standardized test scores and length of time in the US The school

administration also claimed the "regular" classes had no space. Thus, many ESL students remained in the program for all

their years at Coolidge, even if they were born in the US and spoke English fluently.

Community

Seven of the eight participants lived in Las Palmas, the Latino section of Bahia. (The eighth focal student lived in one of the

African-American neighborhoods.) Las Palmas was distinct from other barrios in California inthat this section was

predominantly Central American, not largely Mexican or Chicano (Cordoba, 1989). One-third of Latino youth (ages 12-
17) in this area lived below the poverty level, as compared to 20% of Latino children in the city overall (1990 census). From

1992 to 1993, 24 gang-related shootings were reported city-wide (Ferris, 1992).

Enlarge this image.

Table 1

In 1993, approximately seven Latino gangs (called clicas) operated within Las Palmas, many named after the

streets in which they functioned. Each clica had from twenty to sixty central members, with many more peripheral members

called "taggers" (those who did graffiti) or "wannabes" (those who "want to be" in a gang and are awaiting acceptance). Of a

total of If,700 youth in the barrio from ages ten to seventeen, about 200-500 were affiliated with gangs. Most members were

male, but female gang members existed as well (Ross, 1993).

Most of the gangs in the barrio were associated with either the Nortenos or the Surenos, two arch rival statewide networks.

Norteros and Surenos distinguished themselves by nationality, culture and language. The Nortenos consisted mostly of

Central Americans, primarily spoke English, and blended Latino and African-American cultures. They "claimed" [used] the

color of red to show their association with the Bloods. Incontrast, the Surenos included Mexicans exclusively, spoke Spanish

only, and maintained elements of traditional Mexican culture. Surenos were linked to the Crips, "claiming" the color blue.

Participants

Students. Four male and four female Latino students served as focal subjects. All were enrolled in the intermediate-level

ESL seventh grade class throughout 1992-1993. They were selected on the basis of several criteria: 1) The four girls and

the four boys had naturally formed same-sex friendship groups within the class. 2) All lived in neighborhoods with

high gang activity. 3) All were considered "at risk," as defined by grades of mostly Ds and Fs.

The focal students were affected by neighborhood gangs through personal connections. All of the boys and girls had either

siblings, relatives, or close friends in gangs. They were bussed into Coolidge from their community where gangs were

prevalent. They had scored below 26% on CTBS and were designated as "Limited English Proficient," with Spanish as their

native language. Seven families immigrated from Central America (Nicaragua, El Salvador and Guatemala); one came from

Mexico. Four students spent all their school years in the US.

Data collection

Data were collected to investigate students' perspectives of their place within the school, their relationships with other ethnic

groups, and their position within their community. To tap student perspectives, I collected varied kinds of data. The process

of data collection took seven months and occurred in several stages. During the first stage of five months, I documented oral

and written responses to curriculum projects related to gangmembership and ethnic identity. I conducted daily observation

and weekly audiotaping of class discussions about these issues. I kept field notes daily by writing records of classroom
events at the end of the school day or at home in the evening. In addition, I collected and xeroxed student journals every two

weeks. In these journals, students wrote down their reflections about activities at school and in their lives.

Enlarge this image.

Table 2

During the second stage of data collection of two months, I conducted audiotaped interviews of students and faculty. I

interviewed the eight focal students and three faculty members who taught the students. The teachers included a European-

American woman and a Filipino woman in the ESL department as well as a Mexican-American man in physical education.

Data analysis

I conducted the analysis of my data in several stages. After developing individual files for each student, I color-coded each

piece of data according to my original study questions and created portraits of each student. I then organized the data into

themes, including Issues of Identity: Social World of School Lives with Teachers, Issues of Identity: Social World of School

Lives with Peers, and Issues of Identity: Social Sphere of Neighborhood Gangs.

Findings

All the focal students chose other Latinos as their closest friends. They unanimously believed that Latino students

experienced the greatest discrimination from teachers and peers: Latinos were the ones blamed for wrongdoings, the ones

associated with gang violence, the ones condemned to school failure, and the ones invisibly categorized as "Mexicans."

In general, the teachers perceived the focal students as "discipline problems." Some felt relief when the students were

absent; others habitually sent them out of the classroom. The adults did not have a hopeful view toward the future of these

students, predicting they would become gang members or drop out of school. While the students agreed they were trouble-

makers at school, they had confidence in their ability to survive in life with or without academic success. The students all

stated that they did not intend to become "gang-bangers," viewing gangs as too dangerous. Even though they had close

association with gang members, they claimed that they did not feel pressure to join. On the other hand, they felt pressure

from their parents to not join, given the dangers faced by their older siblings.

The focal students perceived Asian-American students as the most privileged group at school. They felt that Asian students

were the ones consistently chosen for leadership positions. The symbolic elements of this group were: Asian-American

ethnicity, high grades, polite behavior in class, use of Standard English without slang in speech, use of school

penmanship in writing, neat dress and hairstyles.

In contrast, the symbolic elements of the focal students' friendship group were: Latino ethnicity, low grades,

getting in trouble in class, street slang in speech, stylized handwriting, oversized T-shirts and sweatshirts, sagging pants,

yarn necklaces with crosses, long hair for girls with top section piled up high into a rubber band, short hair for boys with
some parts of head shaved. Their roots were not from the culture of school but from the culture of the streets, as

represented in "gangsta rap." These symbolic elements delivered the message: "We are not like you. We don't want to be

like you."

The students' cultural identity related to being Latino in that they saw being "Latino" as a marker of a minority group which

was oppressed in this society. They viewed themselves as distinct from other groups, especially Asians. They wore woven

crosses around their necks, icons of their Roman Catholic religion. Otherwise, the symbolic elements of their cultural identity

drew more from North American turban life than from Latin American civilization. They had little connection to their

homelands in Central America nor much interest indeveloping that connection. They did, however, resent when others

assumed they were from a nation other than their own. The students were expected to assimilate at school.

The students who either were born in the US (Julio, Alicia and Veronica), or who arrived before the age of five (Hector), had

the most difficulties at school in terms of academics and behavior. This finding confirms other studies on immigration and

academic performance in the US. Research in bilingual education has shown that older learners acquire a second language

more rapidly and fully than younger learners, due to their greater proficiency in their native language and intellectual

maturity. Age on arrival to US positively correlates with mastery of English and other subjects taught in English (Cummins,

1981). Perhaps struggles with the language contributed to Julio, Alicia and Hector's academic problems, but this explanation

does not answer why their performance declined over the years.

Assimilation can result in a loss of ethnic identity and the creation of a new cultural identity to take its place. As Darder

(1992) writes, "To be a good US citizen, students of color have been expected to assimilate to 'American' standards and

values and in so doing, discard the values of their primary culture, breaking free of all bonds to a cultural or ethnic identity"

(p. 2). My focal students were supposed to do the same. Most did "break free" of their ties to their ethnic identity in many

ways: they had lost their Spanish, they knew little about their national histories, and they did not listen or dance to Latin

music. On the other hand, mainstream US culture did not exactly welcome them with open arms. Consequently, they

developed a new cultural identity which drew its source from urban American pop culture of resistance, or "gangsta rap."

In their interviews, the focal students revealed that they felt discriminated against by both teachers and other students at

Coolidge Middle School. The students believed that they were perceived through the confines of stereotypes and were not

accepted as individuals. Consequently, they turned to each other for support and formed their own tight-knit and exclusive

peer groups, which provided them with an alternative cultural identity.

Alicia angrily commented in her interview, "They [teachers] think that most of the Latin people are in gangs, and they try to

put them down." Ironically, none of the focal students expressed a desire to join a gang, stating that they

thought gangs were a negative force. They all spoke about the destructive nature of gang violence and mourned the loss of

friends or acquaintances killed as a result of that violence. However, all did want to affiliate with some kind of "posse" - some

kind of group which would provide friendship, support and cultural identity.
Focus on Julio Cortez

To show how the findings of this study came alive in the day-to-day experiences of the focal students, I will provide an in-

depth look at one of those students, Julio Cortez. I chose Julio because he was the most outspoken and the one most

grappling with gang membership, largely due to the involvement of his older brother.

Julio was born in Bahia, one year after his family arrived from Guatemala. He was the middle son of three and turned

14 in late summer of 1993. With his short stature, Julio could have passed for years younger. He normally was

dressed in loose pants, sagging on his hips, and oversized T-shirts or sweatshirts. In warm weather, he wore big shorts

which fell below his knees. His huge tennis-shoes and heavy socks gave the impression that his feet were the largest part of

his body. Circling around his neck was a string of black yarn, ending with a cross. His brown hair was stylishly cut, nearly

totally shaved in the back but long in the front.

As I previously described, Julio's older brother, Elvin, was a leader of a major Norteno gang in Las Palmas. Despite his own

devoted membership, Elvin refused to let Julio participate in gang activities. Julio and Elvin's father had recently died in the

summer of 1992, but their parents had already divorced. The two boys were living with their mother and her second

husband, a Chicano who owned a salsa nightclub. Their mother was a teenager when she gave birth to her sons, and she

told me she was quite overwhelmed about how to help them. Her husband believed Elvin was a lost cause, but he

expressed faith that Julio could turn around with strict discipline.

Issues of Identity: Social Sphere of School

School history. Julio attended a bilingual Spanish program in elementary school. He was bused to this school since it was

located in a Chinese neighborhood far away from the barrio. He did very well in upper primary level, scoring 96th

percentile in math computation in third grade. While his first grade teacher expressed concern over his absenteeism, his

report card until fifth grade showed almost all Bs or Cs, with some As.

The bilingual resource teacher at that school completed the testing and paper work to redesignate him as Fluent English

Proficient, since he met the state criteria of attending school in the United States for over seven years and receiving grades

of C or above. The school principal and his mother approved the redesignation inwriting, which meant he would be

placed in "regular" classes permanently. However, the bilingual department of Bahia School District sent back the forms with

a rejection, claiming his writing sample was not passable. Julio was categorized as Limited English Proficient and

placed in ESL classes, where he stayed for his subsequent years of schooling.

Once at Coolidge Middle School in the ESL program, Julio's academic record deteriorated. His CTBS scores sunk to the first

percentile in all areas in sixth grade, rising slightly to 13% in Reading and 14% in Math inseventh grade. He received nearly

straight Fs in both sixth and seventh grades. Two weeks before the end of seventh grade, the head counselor informed him

that he would not be promoted, even if he attended summer school. She suggested he repeat the grade at another school.

He cut nearly all of his classes, except for language arts and math. While he did go to those particular classes, he almost
never did assignments. When he did attend his science and social studies classes, his teachers kicked him out for

misbehavior.

Ironically, even when he was suspended from school several times during the year, he would still go to his language arts

and math classes. He would beg me, his language arts teacher, to allow him to stay in the class even when he could

legitimately leave for home. Ultimately, however, Julio flunked the year for his poor attendance. On one hand, Julio seemed

attached to school in that he wanted to stay even when suspended; on the other hand, he was resistant, as evidenced in his

cutting. His counselor explained this contradiction by describing him as a stubborn and spoiled kid, trying to set his own

terms. My own sense was that he was desperately attempting to gain some control over his life within a school system which

had rendered him powerless.

Julio's academic performance was in sharp contrast to perceptions of his natural quickness. In conversation, Julio seemed

to some adults and many students as extremely alert. His sense of humor was well-honed, razor sharp. He was the most

talented "capper" (master of sarcastic put-downs) in his seventh grade class. His peers commented that he could have a

successful career as a stand-up comic. His favorite activity in my class was to get up on the tables and perform raps, which

he made up spontaneously. He noticed and remarked on almost every interaction among adults and students, showing keen

observational skills. Julio was able to correctly complete a vocabulary quiz - without ever studying - faster than anyone in the

class by considering context clues. He had been known to give other students the answers to spelling tests but, then,

refusing to do the test himself. He prided himself on never taking a book home. In short, Julio seemed to be trying to control

his life through resistance of school norms.

Ethnic identity. Although Julio did not express much interest in his homeland of Guatemala, he showed a strong desire to

identify himself as Latino. He organized a group at school called "Respect Latin People" (RLP), and he chose Latinos as his

closest friends. Being born here, he never experienced himself as a Guatemalteco, but as a Latino in a broader sense.

Moreover, Julio spoke of the necessity of Latino-Black unity. Interestingly, he allied himself with the street gang (Nortenos)

which was not exclusively Latino or Spanish-speaking, incontrast to its rival (Surefos).

Friendships. In the class Julio's closest friends were Eduardo, Ruben, and Hector - all Latinos. At times, he expressed

disdain for Hector because he took acid and acted stoned at school. Julio did smoke marijuana, but he was too cool to

reveal it in his actions or appearance. His closest friend among the three was Eduardo, with whom he occasionally argued

and fought. His group, RLP, included Latinos of both sexes. But gender rivalry prevailed over ethnic solidarity when RLP fell

apart as the seventh grade boys and girls started feuding in the middle of the school year. Instead of writing love notes to

each other, as they did in the beginning of the year, the boys and girls refused to speak to each other, except for insults. As

Julio stated: "We're going to bust those ho's." ("We're going to get those girls in trouble.") His analysis was that the friction

between the seventh grade Latino girls and boys was due to envy:

Julio: They think I like Claudia [eighth grade girl], right?

Hector: They get jealous.

Julio: And, like, they get jealous if I hug her. They get jealous and Alicia don't talk to me. And I say, "Fuck it!" They say,

"She's our friend and we'll let you talk to her." And I don't care.

Ruben:You guys will still talk to them.

Julio: I won't talk to them. Yo mama.

Even though the war between the sexes was raging, Julio and the other Latino students were still focused almost exclusively

on each other. Their frame of reference was Latinos. They might not have been talking about each other with words of

sweetness and light, but it was each other that they were talking about. Inother words, they were acting like what they were -

adolescents. Julio said it best: They were going through puberty.

Ethnic relations. Ethnic identity was achieved partially through creating distinction with others. For Julio, the line of

demarcation was between the Chinese students and the Latinos. The labels of "Chinese" and "nerds" were

synonymous in his mind. When I asked the boys, "What groups do you see as having power in school?," Hector immediately

said, "The nerds." Julio clarified: "The Chinos." A bit later, in response to the same question, Ruben answered, "Chinese."

Julio affirmed, "Fuck yeah."

Julio spoke in a condemning way about students who were academically successful ("nerds"). Yet his resentment was

mixed with keen awareness of their privilege and his perceived inaccessibility to that privilege. His attitude was clear in these

two excerpts from my field notes:

On Thursday, Julio came into the library. He saw a GATE [Gifted and Talented] class come in and said, "When I see those

kids, something goes on inside of me that makes me want to fight." I asked what it was, and he said it was that they were

"wannabe's." I asked what it was about them that made him feel that - their hair, clothes, what? He just said it was what they

were like. He said we better hold him down because he wanted to fight them. (11/16/92)

I was showing students the University of California Academic Talent Development Program yearbook from last summer to

encourage them to apply for this year. Julio said he wanted to go, but when he looked at the yearbook, he threw it down: "No

Latinos. They're all white and Chinese." He was absolutely correct. (1/21/93)

Julio's attitude toward African-American students was different. He befriended African-American males, especially those he

knew were involved in "tagging" or illegal activities. In discussing the L.A. uprisings inMay 1992, he called Latino-Black unity,

"sick" (meaning, "cool"). His way of talking was clearly influenced by his contact with African-Americans, employing features

of what linguists call "Black English Vernacular."

Cultural Identity:
Creating an Alternative with Language

So, if you want to really hurt me, talk badly about my language. Ethnic identity is twin skin to linguistic identity - I am my

language. (Anzaldua, 1990, p. 206)

Oral Performance Style

The oral performance style of Julio as well as the other students exhibited a use of street slang, laced with humor, and little

evidence of Spanish. Mastery of street slang for these native Spanish speakers was, in many ways, like learning a third

language.

Use of street slang. Julio's speech was a synthesis of Chicano and African-American street dialects, very urban American.

His talk was influenced by the jargon of "gangsta rap" (the music of contemporary inner city rap artists), as evidenced in the

frequent use of these terms such as "yo mama," "hella," "nigga" for "boys," and "ho's" or "bitches" for "girls."

Of all the students, Julio's oral performance style displayed the greatest use of African-American street slang. He

successfully mimicked the oratorical style of Malcolm X in presenting an oral rendition of a Greek myth to the class.

Both in and out of class, he often talked in rap-like rhymes, both spontaneous and prepared. When he performed his raps,

he gestured animatedly, to the delight of his classmates. Use of humor. Julio showed his enjoyment in playing with word

meaning and tone. Julio's oral performance style with peers was based on sarcastic humor, as in this interview

example in which he was competing for the biggest laugh:

SW: How come in elementary school you got really good grades?

Julio: `Cause I was sick [cool].

SW: And what happened?

Julio: I went through puberty.

Eduardo: He got horny.

Julio also took delight in the use of imitation:

SW: When you were in elementary school, were you in a bilingual class that was in two languages?

Julio: Frankly, my dear, I have to think about this [imitating Rhett Butler in Gone with the Wind. What did you say?

Use of Spanish. Interestingly, Julio used more Spanish in the classroom than other male students even though he was born

here and the others had come more recently. Publicly, Julio seemed to use Spanish mostly for insults (like calling a teacher

"vieja" - old) or curse words ("chinga tu madre"). Since Julio used Spanish words in his day-to-day vocabulary, he would

have had difficulty relating to non-Spanish speakers. On the other hand, much of his lexicon was street slang, so he would
have had more barriers with students who did not speak that language. He looked down on Surenos (who spoke only

Spanish). In other words, Julio would have more difficulty communicating with a Spanish speaker newly arrived to US who

had no mastery of street talk than with speaker of street slang who had no knowledge of Spanish.

Written Language

Use of nicknames. Vigil (1988) writes that nicknames are "the adolescent's way to capture and caricature peer quirks, and

by the naming make them accepted and normal; it's as if poking fun publicly works to destigmatize a funny or odd personal

quality" (p. 113). In this vein, all the focal boys and girls had original nicknames, which they used almost

exclusively in written but not oral communication. Julio's nickname was "Mini," making ftu of his shortness. Julio always used

his nickname for his signature, as did the other students. For Julio and the others, their nicknames became their "tags" -

their labels, their signs to the world - displayed on their personal stationery or on building walls.

Use of stylization. Julio's style of handwriting was as critical to him as his speech, if not more so. When Julio wrote, he used

graffiti lettering. In fact, he claimed he no longer could write otherwise. (However, I had a sample of his handwriting in sixth

grade which was school-style penmanship). He also used stylized misspellings, as in "krushed" for "crushed" (a term

referring to doing lots of graffiti) and "kry" for ("cry"). Figure I is an example of his "tag," which he illustrated during a class

period:

Handwriting was the single-most important entry into tagging crew membership. As a result, students who wished

to join tagging crews practiced their writing constantly to master the style. Julio once asked me if I could get him a lettering

book which had Old English-style letters in it. I took him to the library, and we made xeroxes of these pages. He copied this

alphabet for days. This book became the most popular item in the class. Other students, including the Latinas, approached

me for copies of the Old English alphabet. Julio was trying to come up with an original graffiti style which combined cholo

(style of Chicano street gangs) and Old English writing.

Language and literacy provided ways in which students were trying to take control over their lives by communicating in their

own style. Graffiti and tag names represented forms of critical literacy, reshaping written language for one's own purposes.

To understand the meanings embedded in graffiti, one must be a member of a particular literacy community. Not everyone

can "read" graffiti; you must be taught by an insider who can decipher the codes - the nicknames, the acronyms, the use of

Roman numerals.

Enlarge this image.

Figure 1

In these ways, Julio and the other students brought the streets into the classroom. They expressed much alienation from

school as an academic world. They did not want to read any books that were linked to school inany way, even if a

former gang member wrote them. To gain a place for themselves in school, the students created an alternative cultural

identity. This identity was based on oral and written language forms from the streets, the world in which they felt included.
Cultural Identity and Gang Membership

What happened when the students left school at the end of the day? It was not difficult to understand why youth in Las

Palmas would want to join a gang. They had the natural desire and need to belong to a group. An ex-gang member turned

poet, Luis Rodriguez (1993), wrote in Always Running that teens in a gang want just what children in an established

organization like the YMCA or Little League want: respect, sense of belonging, and protection.

Need to Belong/Nothing to Belong to

Schools in California once provided after-school activities, such as clubs or sports teams, which offered opportunities to

belong to a group. However, budget cuts, beginning in the 1980s and growing ever since, have severely restricted those

activities. In addition, busing in Bahia has made participation in the few remaining activities nearly impossible. For example,

at Coolidge, the students from Las Palmas got a free bus ride home school immediately after the school bell rang. If they

stayed later, they were forced to take public transportation, which was both expensive, time-consuming, and potentially

dangerous. Bus stops near their own neighborhoods were not safe places to be in the dark. These students

were, in essence, discouraged from belonging to official school groups.

A further difficulty was that many "legitimate" community groups, like the Boy or Girl Scouts, fraternities or sororities, lacked

credibility because they either did not exist in barrios like Las Palmas, or else had a history of racist practice. Consequently,

the only neighborhood group accessible for urban youth to join was one distinguished by its illegal or antisocial activities -

a gang.

Julio's Brother, Elvin

Julio's brother, Elvin, was a prominent leader of the largest Norteno neighborhood street gang until he was sent to a juvenile

detention program. Years earlier, Elvin had graduated from elementary school with a good school record. His CTBS

scores in 1991 were high enough that his school counselor recommended him insixth grade for a program at Coolidge to

prepare African-American and Latino students for honors classes. This counselor said that Elvin turned down the

opportunity. When I asked Elvin, he said he had never heard of this recommendation; in fact, he acted surprised to discover

that he had been considered for the honors program. Julio, present when I mentioned this, smirked: "My brother? Honors?"

Although Elvin left Coolidge in the fall, his former classmates frequently recounted and discussed his gangactivities for the

rest of the school year. In fact, students greeted each other many Monday mornings with, "Did you hear what happened to

Elvin this weekend?" Stories of the bullet wounds in his leg from a drive-by shooting or his arrests or his stays at the local

youth detention center circulated in the classroom. His ganginvolvement clearly had an impact -upon not only Julio, but the

other students who knew him well.

People in the mainstream often speak of gang members as "immoral" or "lacking values." However, Elvin clearly was

operating within his own (as well as his gang's) system of ethics. He was single-mindedly committed to protecting his

own gang. He was willing to put his life on the line to avenge an injury or death of one of his "homeys."

Elvin 's Influence on Julio

Julio gave the external impression of being a "gang-banger" (gang member) with his style of dress and talk. Although he had

previously joined a gang, he was no longer a member at the time of the study. He said he did not feel under pressure to join:

"They come up to you and say: `Do you want to get in?' `No.' `Well, fuck you, then'." He "knew a lot of gang-bangers,"

including some who had been killed. "I had hella (many) friends that died," he proclaimed. He said that if he were sent to

Juvenile Hall, he wouldn't worry about being attacked because he knew most of the inmates.

Julio's friendship and neighborhood networks were influenced by those of his brother Elvin, yet distinct from them. While he

did engage in activities, such as "tagging" (graffiti writing) and robbery, which were illegal, he did not

participate in gang violence which involved assault. In fact, Elvin told me that he prohibited Julio from joining his gang, taking

the role of the protective older brother, or the paternalistic "man of the house." He didn't want his mother to suffer from

having two sons involved in such risky behavior.

Julio spent much of his time with his tagging partner, Rafael, who was a high school student. He had been arrested with

Rafael for breaking into the high school and some corner stores in the barrio. Below is his description in his journal (spelling

left as original, punctuation added) from March 1993:

Dear Julio

On Friday night I left my house and stad [stayed] up all night Saturday. Broke in Las Palmas High and police came and then

we got cout [caught] at gun point. Then I was sent home Then I went to Rafy's hous [house] and I got cout at some store we

broke into. Narcs had us on gun point. Then we went to 250 Burbank [juvenile hall]. Got fingure print tuke [finger prints

taken]. Went to Juvie. Then I ate went to my room and went to school and saw hella peaple [people] and this time I ain't

playing. was dieing to get out and I got out back on the street. Went straight home and sed [said] I will try to do

good in skew] [school] - so it's cool to kry but alone and cry. Later well I rote [wrote] this becouse I wasn't crying but I felt like

because they were gana [going to] give me two mons [months] with no visitors.

Well By.

While Julio distinguished himself from Elvin, he frequently recounted his brother's activities. He treated Elvin as a war hero,

defending the territory. On October 16, 1992, Julio came in early in the morning to tell me a secret. He said that Elvin had

been shot in the leg and was in an intensive care unit in the local hospital.
Then on May 11, 1993, Julio came into my classroom to tell me that the police (nicknamed "5-0") came to his family's house

at 4 am, Sunday morning, to inform them that Elvin was locked up, but they couldn't reveal any more information. The next

day, Julio said the charge was attempted murder. On May 13, Julio cut school. I told him on May 14 that the attendance

office called his mother, but he said his mother wasn't home since she had gone to see Elvin in jail.

Inevitably, Elvin had a huge impact on Julio. He was simultaneously worshipful and terrified of his brother. He was excited

by the degree of danger in his brother's life, and yet personally witnessed the destructive consequences. His mother and

stepfather spoke of Julio as "bad," but not "as bad" as Elvin. This characterization seemed to describe his general role in the

minds of adults in his community and in school.

His overall assessment of gangs was that they were not positive. Ironically, in a minimum standards writing test, Julio wrote

about the dangers of gangs and drugs. The prompt was to select a major problem in the US today and to propose solutions.

This was Julio's example:

Gangs and Drugs

To help the city you have to stop graffiti, drugs and no sale on weapons stop all gang members cause it will stop the people

dying cause of drugs and gang members and it will stop people not to be afraid to come out there house. So scared they

have to be packed with a nine millimeter or a twelve gage [gauge] that is worse. It's better to enjoy life the way it is not the

way the gang and drugs make it. It's better to enjoy what talent you got not wot [what] you want.

At the end of the schoolyear, Julio failed-seventh grade at Coolidge because of absenteeism and was referred to an

alternative public school for students involved in gang activity. In the following year, he came to visit, informing me he never

attended the school and was flunking out. He told me that he spent his days on the streets. Two months later, I saw him on

a Las Palmas street corner, gesturing dramatically while talking to friends circled around him. All I noticed was the color of

his oversized T-shirt: red.

Conclusion

Recently in the fall of 1996, I encountered Julio at the same corner. Now age 16, he had just returned to Bahia after being

sent to a juvenile detention camp in Arizona on charges of attempted murder. Spontaneously he gave me an update on all

the focal students. Julio himself is now regularly going to classes at his alternative school and is aspiring for admission to the

local university. He wants to be a writer, particularly of filmscripts so he can capture the Bahia gang scene. His brother Elvin

is in a federal penitentiary for murder. Of the four boys, only Eduardo is not gang-banging. Veronica has been in and out of

a gang, and Karla is pregnant and has been in and out of school. Julio's news created a picture in stark contrast to three

years earlier.

At the time of the study in 1992-1993, none of the students was a gang members even though gang activities were

pervasive in their lives. All of them were 12, 13 or 14 years old - the critical age for 1990s youth to joingangs. They were
seriously weighing the pros and cons of gang membership, but they seemed to land on the negative side, mostly from

having seen the destructive consequences within their families and friends. While writing about the dangers of gangs, the

boys particularly also romanticized the adventures of gang members like Elvin.

On the other hand, all the students dressed in the "gangsta" style of sagging, baggy pants and oversized sweat shirts, and

they talked the "gangsta" talk. They were indeed influenced by their older siblings or immediate family members. In this

sense, all acted like "wannabes." Perhaps intimately knowing gang members and personally witnessing their arrests,

wounds and stays in emergency rooms or jails discouraged them from joining at that time. In 1994, I wrote:

However, the combination of their school experiences, ranging from marginalization to flunking, along with neighborhood

peer pressure might push them to join in later years. Clearly, they were at a very vulnerable moment of their lives in this

regard, a moment which could determine their futures. (Weinberg, p. 138).

Those words proved to be prophetic.

Luis Rodriguez (1993) clearly speaks for my students:

I've talked to enough gang members and low-level dope dealers to know that they would quit today if they had a productive,

livable-wage job. You'll find people who don't care about who they hurt, but nobody I know wants to sell death to their

children, their neighbors and friends.

If there was a viable alternative, they would stop. If we all had a choice, I'm convinced nobody would choose la vida loca, the

"insane nation" - to "gang bang." But it's going to take collective action and a plan (p. 251).

As educators, we need to listen seriously to the words of Rodriguez and youth like my focal students. They would not

choose a life of violence if they had the choice. It is time we all work together to build other options, and soon, before we

lose an entire generation.

Footnote

1 The dissertation, written under the name Susan Katz Weinberg and entitled the same as the article, was granted in 1994

by the University of California, Berkeley. The Dissertation Committee was chaired by Dr. Anne Haas Dyson with Drs.

Guadalupe Valdes, Sarah W. Freedman, and Margarita Melville as members.

References

References

References

Adams, R. N. (1989). Internal and external ethnicities: With special reference to Central America. Texas Papers on Latin

America Pre-publication working papers of the Institute of Latin American Studies. Austin: University of Texas. Anzaldua, G.

(1990). How to tame a wild tongue. Out there: Marginalization and contemporary cultures. Cambridge, MA: MIT Press.
References

Arriaza, G. (1995). Geography of identity: Issues of cultural identity among first generation Mesoamerican migrants.

Unpublished manuscript, University of California, Berkeley. Cordoba, C. B. (1989). The Las Palmas District: The ethnic

diversity of the Latin American enclave in Bahia, California. Journal of La Raza Studies, 2(1), 21-32. Cummins, J. (1981).

The role of primary language development in promoting educational success for language minority children. In Schooling

and language minority students: A theoretical framework. Los Angeles: California State Department of Education,

Evaluation, Dissemination and Assessment Center.

References

Darder, A. (1992). Bicultural identity and the development of voice. Unpublished paper submitted for CABE Conference,

Winter 1993. Ferdman, B. M. (1990). Literacy and cultural identity. Harvard Educational Review, 60(2), 181-204. Ferris, S.

(1992). What pulls youngsters into Las Palmas gangs? Bahia Examiner, September 30, 1992. Melville, M. B. (1988).

Hispanics: Race, class, or ethnicity. The Journal of Ethnic Studies, 16(1), 67-84. Nawa, F. (Fall, 1993). Loyalty, respect,

revenge in Las Palmas. Yo!, p. 2. O'Connor, J. D. (1993). Violent night inLas Palmas District. Bahia Examiner, May 3, 1993,

A-1 to A-14.

References

Rodriguez, L. J. (1993). Always running: La vida loca: Gang days in L.A. Willimantic, CT: Curbstone Press. Ross, J. (1993).

G.E.T. vs. gangs: Life is getting cheap for Las Palmas youth. Las Palmas Life Bilingue, 19-20. Vigil, J. D. (1988).

Barrio gangs: Street life and identity in Southern California. Austin: University of Texas Press. Weinberg, S. K. (1994).

Where the streets cross the classroom: A study of Latino students' perspectives on cultural identity in city schools and

neighborhood gangs. Unpublished doctoral dissertation, UC Berkeley. Weinberg, S. K. (Ed.) (1992). Don't stereotype me: A

book of poems. California: Coolidge Middle School. Williams, R. (1977). Marxism and literature. New York: Oxford University

Press.

AuthorAffiliation

University of San Francisco

Copyright National Association for Bilingual Education Summer 1996

Katz, S. R. (1996). Where the streets cross the classroom: A study of latino students' perspectives on cultural identity in city

schools and neighborhood gangs. Bilingual Research Journal, 20(3), 603-631. Retrieved from

http://search.proquest.com/docview/222042624?accountid=37714

OUT 5

Matusitz, J., & Repass, M. (2009). Gangs in nigeria: An updated examination. Crime, Law and Social Change, 52(5), 495-511.

doi:http://dx.doi.org/10.1007/s10611-009-9208-y
 Pride and prejudice in high school gang members
Wang, Alvin YAuthor Information . Adolescence 29.114 (Summer 1994): 279-91.

1. Full text
2. Full text - PDF
3. Abstract/Details
Hide highlighting
Abstract
TranslateAbstract

This study compared gang versus nongang high school students along measures of self-esteem, racial attitudes, and their

self-professed role models. A total of 78 Caucasian (65 nongang and 13 gang members) and 77 African-American students

(41 nongang and 36 gang members) participated. Results indicated that gang members had significantly lower levels of self-

esteem compared to their nongang peers. All students, regardless of ethnicity, manifested negative racial stereotyping

toward racial outgroups; gang members were not more racially prejudiced compared to other students. The role model data

revealed that overall, gangmembers could name fewer role models than did their nongang peers. African-American students

who were not gang members were much more likely to mention a parent or teacher as a role model. A regression analysis

indicated that the absence of parental or teacher roles models was the best predictor of gangmembership. These results are

conceptualized within the framework provided by social identity theory.

Full Text
TranslateFull text
Turn on search term navigation
This study evaluated self-esteem and racial attitudes of high school gang members compared to their uninvolved (nongang)

peer. While previous research has documented linkages between ethnicity and self-esteem in adolescents, none have

examined this issue within the context of gang membership (Richman, Clark, & Brown, 1985; Tashakkori & Thompson,

1991). For instance, the question of whether gang members' racial attitudes differ from their uninvolved peers has not been

explored despite the observation that gangsgenerally tend to form along ethnically distinct boundaries (Johnstone, 1983).

There is also a great deal ofresearch on the self-esteem of adolescents and juvenile delinquents, but little is known

concerning the self-esteem of adolescents who are members of gangs (Arthur, 1989; Kaplan, 1975).

What is known is that levels of self-esteem in African-American youth tend to be comparable or higher than their Caucasian

peers (Richman et al., 1985; Tashakkori & Thompson, 1991). This finding invalidates an earlier notion that African-

Americans have a debased sense of self, formerly termed "black self-hatred" (Porter & Washington, 1979; Simmons, 1978);

Other research has shown that adolescents (and adults) manifest favorable stereotypes toward their own racial group at the

expense of outgroups (Clark, 1985; Wooten & Brown, 1990). The ethnocentric bias displayed by both Caucasian and

African-Americans has caused some researchers to conclude that individuals identify with groups in ways that enhance their

self-esteem (White & Parham, 1990; Porter & Washington, 1979).

A similar conclusion was reached by Kaplan (1975) in his review of the literature on the self-identity ofdelinquent youth. In

developing the theory of self-esteem motivation, Kaplan (1975, p. 73) states that "persons characterized by negative self-
attitudes are motivated to adopt deviant response patterns...that are associated with the enhancement of self-attitudes."

Thus, gang membership and its accompanying behavior (e.g., acting-out and fighting) has the effect of improving one's self-

attitudes, particularly when positive experiences and social support are not available elsewhere (e.g., family and schools).

Rosenberg & Rosenberg (1978) examined this claim using a regression analysis of a nationwide sample of tenth-grade

boys. Their finding that self-esteem was a powerful causal factor of delinquency was interpreted as support for Kaplan's

(1975) theory ofthe self-esteem motive. Nevertheless, in none of the aforementioned studies were any of the delinquent

youth identified as being gang members. Consequently, the relationship of self-esteem to gang membership remains

unanswered. In fact, the only evidence that gang members might possess different (i.e., lower) levels of self-esteem relative

to their nongang peers is anecdotal rather than empirical in nature (Arthur, 1989).

Therefore, the present study was designed to explore the linkages between ethnicity and self-esteem in high school

student gang members compared to their nongang peers. The conceptual basis for this comparison was the framework

offered by social identity theory (Tajfel, 1974, 1982; Tajfel & Turner, 1979). According to this theory, the social identity that is

formed as a function of group membership is comprised of three components: motivational, cognitive, and sociocultural

(Dovidio & Gaertner, 1986; Hamilton & Trolier, 1986). The motivational component is based on the premise that individuals

attempt to form positive evaluations ofthemselves. Thus, people are motivated to identify with groups in ways that enhance

rather than detract from their personal self-esteem. The cognitive component involves the process of social categorization

responsible for the ingroup and outgroup stereotyping that may lead to intergroup discrimination, conflict, and prejudice

(Tajfel, 1974, 1982). The soiocultural component is based on the process of social learning which, in turn, hinges on the

types of role models available. These role models can exert a powerful influence on the extent to which prejudiced attitudes

and behaviors will be exhibited (Dovidio & Gaertner, 1986).

Given this conceptual framework, three components of social identity were evaluated in the present study: measures of self-

esteem, ethnic ingroup versus outgroup stereotyping, and the number and type of self-professed role models. The students

sampled were male adolescents attending the 11th and 12th grades; gang membership was determined on the

basis of referrals made by teachers and counselors. The two gangs studied were ethnically distinct: Caucasian or African-

American; their nongang peers were students attending the same classes.

The design and materials thus described allowed for the testing of three hypotheses. First, it was hypothesized

that gang members would have lower levels of self-esteem. This finding would be compatible with the earlier research on

delinquent youth (Kaplan, 1975; Rosenberg & Rosenberg, 1978). The second hypothesis was that gang members would

display greater ethnocentrism. Prior research within the context of social identity theory suggests that intergroup conflict

(e.g., gang warfare) may contribute to negative outgroup stereotyping (Tajfel, 1974, 1982; Vaughn, 1988). Similarly, in a

study of Chicago gang recruitment, Johnstone (1983) reported that perceived racial tension was greater among African-

American gang membersthan in their nongang peers. Because gangs tend to form along ethnic lines, it was expected

that gangmembers would display extreme negative attitudes toward their ethnic outgroup compared to the attitudes oftheir

uninvolved peers. The third hypothesis concerned the role models who were mentioned by both gangmembers and
uninvolved youth when asked to "name three people you most admire." One relatively straightforward prediction from social

exchange theory is that adolescents' self-professed role models would tend to be of the same ethnicity. Theoretically,

however, it was not clear whether the number or type (e.g., parent, teacher, sports figure) of role models would change as a

function of gang membership.

One final issue was addressed in the present study. Because several measures were collected from gangmembers and their

nongang peers, it was possible to determine which factors might best predict gangmembership (Rosenberg & Rosenberg,

1978). Specifically, by treating gang membership as a dependent variable in a regression solution, the present study would

be able to assess the degree to which various factors (e.g., self-esteem, racial attitudes, ethnicity, number and type of role

models) contribute to membership in a gang.

METHOD

SUBJECTS

All of the subjects (N = 155) were 11th-and 12th-grade students attending two public high schools located in the suburbs

surrounding Orlando, Florida. Both schools enroll comparable numbers of African-American students (35% and 38%).

The gang members were identified by teachers and counselors in Beeline High School (a fictitious name was supplied here

at the request of school administrators) a students who were referred to a new school program ("Project Redirect") designed

to foster social skills, coping strategies, and adaptive problem solving in teenagers. The criteria used for identifying these

teenagers were that they either belonged to a gang and/or evidenced a prior history of school demerits based on interracial

conflict (i.e., in-school fighting and acting-out). The gangs in Beeline High School were formed along racial lines (either

Caucasian or African-American).

Because Beeline school administrators were willing to allow testing only of Project Redirect students, it was necessary to

obtain "nongang" student data from another comparable school. Therefore, five classes representing a broad

range of academic achievement levels were obtained from Sanford High School. These classes included one remedial class

and another four classes randomly selected from the regular school curriculum. For purposes of gang versus nongang

comparisons, the data from the five nongang classes were combined, thereby yielding a fairly large, representative

sample of 11th and 12th graders who were not members of gangs. At the time of data collection there was no

evidence of any gang activity at Sanford High School according to teachers and the school principal.

MATERIALS AND PROCEDURE

The questionnaires were administered in classes of about 25 students each. Students were given a pamphlet containing

four sheets of paper stapled in the following order: cover sheet, self-esteem scale, Racial Attitude Adjective Checklist

(RAAC) for Caucasians, and RAAC for African-Americans. The cover requested information concerning the student's age,

gender, and race. In order to obtain the names of each student's self-professed role models, an additional question asked

students to "please name three people you most admire." Students completed the questionnaires on an individual basis and
were instructed not to discuss their responses with anyone else during this time. A total of about 50 minutes was required to

collect the data from each class.

The self-esteem scale (Group Dynamics and Strategy Training, Inc.) was chosen for use by Beeline School administrators

as a means of evaluating students referred to Project Redirect. This self-report scale was comprised of 25 statements

describing both positive (e.g., "I normally anticipate new endeavors with a positive expectancy and confidence") and

negative feelings of self-esteem (e.g., "I am very concerned about what others think of me"). The thirteen even-numbered

statements were positive, and the twelve odd-numbered statements expressed negative feelings of self-esteem. Students

responded to each of the statements using a 4-point, Likert-type scale to indicate how well each statement described them

(from 0 = not true to 3 = true). Three scores were computed from this scale: positive self-esteem, negative self-esteem, and

overall self-esteem (positive minus negative score). A pilot study involving 38 high school students indicated that the positive

and negative self-esteem statements had average Chronbach's alphas of .56 and .73, respectively. The test-retest reliability

for overall self-esteem scores after two months was r(36) = .73.

The RAAC was designed as a "covert measure of racial attitudes toward one's own and other racial groups" (Wooten &

Brown, 1990, p. 3). The adjectives on the RAAC were derived from the Adjective Checklist (ACL) which was originally

developed as a projective test of personality. Based on pilot research (Wooten & Brown, 1990) a total of 333 RAAC items

was selected based on their positive or negative weights along an ethnic stereotyping scale. Consequently, 239 negative

items (e.g., lazy, vicious) and 94 positive items (e.g., intelligent, ambitious) comprised the RAAC. Each RAAC yields a

positive and negative score as well as an overall score (positive minus negative score). Wooten and Brown (1990) reported

a Cronbach's alpha of .79 for the overall RAAC scores of 127 college students.

RESULTS

A total of 155 male high school students (51 gang members and 117 nongang students) was tested. However, due to the

nature of the RAAC, students who were not Caucasian or African-American were omitted from the analyses reported below.

Also, because there were only six female gang members, the data from female students were excluded from all analyses.

The resulting subsample sizes were: 78 Caucasians (65 nongang and 13 gang members) and 77 African-American students

(41 nongang and 36 gang members). However, dfs may vary as a function of incomplete data. Unless reported otherwise, a

significance level of p = .01 was applied to all statistical tests.

SELF-ESTEEM

Table 1 shows the mean positive, negative, and overall (positive minus negative) self-esteem scores' as a function of group

membership.(Table I omitted) A 2 (Race) x 2 (gang vs. nongang) Analysis of Variance (ANOVA) indicated

that gang members had significantly lower levels of overall self-esteem than did nongang students, F(1, 119) = 20.44, MS, =

.41. No other relationships were statistically significant. The lack of a main effect for Race indicates that African-American

students did not have lower self-esteem scores compared to their Caucasian peers. This is in line with other studies (Porter
& Washington, 1979; Simmons, 1978) that also found comparable levels of self-esteem for these two ethnic groups. It

should also be pointed out that this finding runs counter to pre-1960s reports that African-Americans possess a debased

self-concept which some researchers have termed "black self-hatred" (Dreger & Miller, 1960).

A separate 2 x 2 x 2 ANOVA was applied to the positive and negative self-esteem scores of students. The three factors

analyzed were Race, Condition (gang vs. nongang), and Score (positive vs. negative). In this analysis a main effect was

obtained for Score indicating that, overall, students' positive self-esteem scores were higher than negative self-esteem

scores, F(1, 119) = 8.59 MS sub e = .20. A subsequent analysis revealed that among the Caucasian students,

nongang members (M = 1.88) had higher positive self-esteem than did gangmembers (M = 1.21), Fisher's Least Significant

Difference (LSD) = 62. However, no difference was found in the positive self-esteem for African-American nongang

versus gang members. There was also a tendency for negative self-esteem scores to be higher in gang members. This was

the case for both Caucasian and African-American students, Fisher's LSD = .43, p < .05.

RACIAL ATTITUDES

Table 2 shows the mean RAAC scores (positive, negative, and overall) as a function of Race of Subject, Race of Target,

and Condition (gang vs. nongang). (Table 2 omitted) A 2 x 2 x 2 ANOVA on overall RAAC scores revealed a highly

significant interaction between Race of Subject and Race of Target, F(1, 151) = 29.18, MS sub e = 1017.63.

Inspection of Figure 1 suggests that all students exhibited greater negative stereotyping toward their outgroup than to their

own racial group. (Figure 1 omitted) A Fisher's LSD = 19.95, p < .05 indicated that Caucasian students have significantly

more negative stereotyping of African-Americans compared to their own racial group. A corresponding pattern was found for

African-American students: they displayed significantly greater negative stereotyping toward Caucasians compared to their

own racial group. This cross-over interaction replicates the results of other research using the RAAC (Wooten & Brown,

1990). No other Fs reached statistical significance.

ROLE MODELS

Each student was asked to name the three people they most admire. The total number of times and a breakdown by

profession are shown in Table 3. (Table 3 omitted) A 2 x 2 ANOVA was performed on the total number of models named

with Race of Subject and Condition (gang vs. nongang) as the two factors. Only a marginally significant effect of Condition

was obtained indicating that gang members (M = 1.9) could name fewer role models compared to their nongang peers (M =

2.4), F(1, 123) = 3.44, p < .07.

Separate 2 (Race of Subject) x 2 (Condition) ANOVAS were performed on the three categories of role models shown in

Table 3. The classification of role models into categories was rather straightforward (e.g., Michael Jordan as a sports figure,

or "my mom" as a parent). These analyses revealed that, as predicted, Caucasian students were more likely to name

Caucasian role models than were their African-American peers; the same ethnocentrism was evident for the African-

American students, F(1, 123) = 35.45, MS sub e = .87. However, African-American students were more likely to claim a
sports figure as a role model compared to Caucasian students, F(1, 123) = 5.50, MS sub e = .86. Finally, when the

number of parental or teacher role models were analyzed, a marginally significant two-way interaction was obtained, F(1,

123) = 3.52, MS sub e = .47, p < .07. Inspection of Table 3 indicates that this was primarily due to the fact that African-

American students who were not gang members were much more likely to mention parents or teachers as role models

compared to all other groups of students.

REGRESSION ANALYSES

Separate multiple regression analyses were applied to the data of Caucasian and African-American students. Both analyses

used a backward-stepping solution with Condition (gang vs. nongang) treated as the dependent variable; Race and all of the

test scores served as the independent variables (e.g., positive, negative and overall self-esteem, Caucasian and African-

American RAAC scores, the number of type of role Models). The results for both analyses were similar in that parental role

models emerged as the strongest predictor of gangmembership for both Caucasian, partial r = .62, F(1, 58) = 8.25, multiple r

sup 2 , 29, as well as African-American students, partial r = .60, F(1, 33) = 17.76, multiple r sup 2 = 36. Specifically, the

absence of a parent or teacher role model was highly predictive of gang involvement. For Caucasian students, the

number of Caucasian role models named was also a valid predictor of gang membership, partial r = .54, F(2, 56) = 23.42.

This finding was primarily due to the fact that the parents of Caucasian students (and therefore parental role models) were

also Caucasian--a relationship that would, of course, not apply to the African-American students. No other factors survived

further stepping in either regression analysis. Hence, whether students were able to name their parents and teachers as

people whom they most admire appears to be a strong predictor of ganginvolvement. Interestingly, none of the

measures of self-esteem were predictive of gang membership.

DISCUSSION

This study tested three hypotheses concerning the self-esteem and racial attitudes of gang members versus their nongang

peers. These hypotheses were based on the assumption that social identity consisted of three components: motivational

(self-esteem), cognitive (racial stereotyping), and sociocultural (the number and type of self-professed role models). In

general, modest support was obtained for social identity theory (Tajfel, 1974, 1982) and Kaplan's (1975) theory of self-

esteem motivation. The following is a summary of the study's main findings as they relate to each hypothesis.

SELF-ESTEEM

As predicted, gang members of both ethnicities possessed lower levels of overall self-esteem compared to their nongang

peers. In part, this was due to the fact that the negative component of self-esteem was greater for both Caucasian and

African-American gang members compared to their uninvolved peers. Thus, support was found for Kaplan's (1975)

theory of self-esteem motivation which states that negative self-attitudes may prompt individuals to engage in deviant

patterns of behavior (e.g., gang activity). In addition, the present study did not find any evidence for the existence of "black

self-hatred." While no overall differences were found as a function of ethnicity, it is worth noting that for nongang students,
the overall self-esteem score ofAfrican-Americans (M = .76) as higher than that of Caucasians (M = .64). This is in line with

other recent investigations which report that African-American students possess comparable (or even higher) levels ofself-

esteem relative to their Caucasian peers (Porter & Washington, 1979; Simmons, 1978).

RACIAL ATTITUDES

Contrary to expectations, no support was found for the second hypothesis that gang members would display higher

levels of racial prejudice than would nongang peers. When RAAC scores were analyzed, neither the ANOVAs nor

regression solutions produced any significant relationship between racial attitudes and gangmembership. Indeed, strong

ethnocentrism was exhibited by all students regardless of their ethnicity (Clark, 1985; Wooten & Brown, 1990),

but gang members were not more prejudiced than their nongang peers. This is somewhat surprising given the fact

that gang membership usually follows ethnic boundaries and that gangmembers tend to perceive more racial tension in their

neighborhoods than do their uninvolved peers (Johnstone, 1983).

How might these disparate findings be reconciled? Recall that within the context of social identity theory, intergroup hostility

is believed to enhance both group identity and outgroup bias (Tajfel, 1974, 1982; Vaughn, 1988). It then seemed probable

that gang members would be more racially biased than their uninvolved peers, especially if there is a history of intergang

conflict. However, in retrospect, it may be that for the gangmembers studied here, the "salient" ingroups and outgroups were

defined largely in terms of gangmembership rather than ethnicity. If this is the case, then racial tensions would be

discounted by gangmembers as the primary cause for intergang conflict. Under these circumstances, intergang hostility

would not lead to more biased racial attitudes by gang members. Future research is needed to address this interesting issue

regarding the relative salience of various groups (e.g., ethnic, gang, family) as they pertain to gang members.

ROLE MODELS

Evidence was obtained for social identity theory's prediction that adolescents were more likely to name role models from

their own ethnic group compared to their ethnic outgroup. Also, there was a tendency for gangmembers to name fewer role

models than did their nongang peers; among the African-American students, nongang youth were three times more likely to

mention a parent or teacher as a role model compared to gangmembers. However, the most striking aspect of the role

model data emerged from the multiple regression analysis. According to this analysis, the absence of either a parent or

teacher role model was the beat predictor of gang membership. The predictive validity of parental role models corroborates

other research suggesting that intrafamilial socialization and parental support are among the most significant

determinants of gang membership, particularly for ethnic minority youth (Adler, Ovando, & Hocevar, 1984; Johnstone, 1983).

Finally, it should be reiterated that the present study involved adolescent males who were either of Caucasian or African-

American descent. Additional research is needed to determine whether similar processes of social identity are responsible

for gang membership among female adolescents and youth of other ethnicities (e.g., Asian and Hispanic). Future research

should also consider potential differences between gang members who are still attending school compared to school drop-
outs. Possibly, the predictive validity of self-esteem for gang membership will vary as a function of academic achievement,

particularly if a distinction is made between general self-esteem versus specific (e.g., academic) self-esteem (Richman et

al., 1985).

REFERENCES

Adler, P., Ovando, C., & Hocevar, D. (1984). Familiar correlates of gang membership: An exploratory study ofMexican-

American youth. Hispanic Journal of Behavioral Sciences, 6, 65-76.

Arthur, R. F. (1989). How to help gangs win the self-esteem battle. School Administrator, 46, 18-20.

Clark, M. L. (1985). Social stereotypes and self-concept in black and white college students. Journal of Social Psychology,

125, 753-760.

Dovidio, J. F., & Gaertner, S. L. (1986). Prejudice, discrimination, and racism: Historical trends and contemporary

approaches. In J. F. Dovidio, BE & L. Gaertner (Eds.), Prejudice, discrimination, and racism. Orlando, FL: Academic Press.

Dreger, R. M., L Miller, K. S. (1960). Comparative psychological studies of Negroes and whites in the United States.

Psychological Bulletin, 57, 361-402.

Hamilton, D. L., & Trolier, T. K. (1986). Stereotypes and stereotyping: An overview of the cognitive approach. In J. F.

Dovidio, & S. L. Gaertner (Eds.), Prejudice, discrimination and racism. Orlando, FL: Academic Press.

Johnstone, J. W. C. (1983). Recruitment to a youth gang. Youth and Society, 14, 281-300.

Kaplan, H. B. (1975). Self-attitudes and deviant behavior. Pacific Palisades, CA: Goodyear.

Porter, J. R., & Washington, R. E. (1979). Black identity and self-esteem: A review of studies on black self-concept 1968-

1978. Annual Review of Sociology, 5, 53-74.

Richman, C. L., Clark, M. L., & Brown, K. P. (1985). General and specific self-esteem in late adolescent students: Race by

gender by SES effects. Adolescence, 20, 555-566.

Rosenberg, F. R., & Rosenberg, M. (1978). Self-esteem and delinquency. Journal of Youth and Adolescence. 7, 272-291.

Simmons, R. (1978). Blacks and high self-esteem: A puzzle. Social Psychology, 41, 54-57.

Tajfel, H. (1974). Social identity and intergroup behavior. Social Science Information, 13, 65-93.

Tajfel, H. (1982). Social psychology of intergroup relations. Annual Review of Psychology, 33, 1-39.
Tajfel, H., & Turner, J. C. (1979). An integrative theory of social conflict. In W. Austin, & S. Worchel (Eds.), The social

psychology of intergroup relations. Monterey, CA: Brooks/Cole.

Tashakkori, A., & Thompson, V. D. (1991). Race differences in self-perception and locus of control during adolescence and

early adulthood: Methodological implications. Genetic, Social, and General Psychology Monographs, 117, 135-152.

Vaughn, G. M. (1988). The psychology of intergroup discrimination. New Zealand Journal of Psychology, 17, 1-14.

White, J. L., & Parham, T. A. (1990). The psychology of blacks: An African-American perspective (2nd ed.). Englewood

Cliffs, NJ: Prentice-Hall.

Wooten, W., & Brown, K. (1990). The assessment of racial attitudes among African-Americans, Asian-Americans,

Caucasian-Americans and Hispanic-Americans using the racial attitude adjective checklist (RAAC). Paper presented at the

annual meeting of the Southeastern Psychological Association, New Orleans, LA.

Portions of this data were presented at the April 1992 meeting of the Southwestern Psychological Association, Austin,

Texas. The author acknowledges the assistance of Tammy Chamberlain in gathering the data and the principal and staff at

Sanford High School.

Reprint requests to Alvin Wang, Ph.D. Department of Psychology, University of Central Florida 32816.

Word count: 3949

Wang, A. Y. (1994). Pride and prejudice in high school gang members. Adolescence, 29(114), 279-91. Retrieved from
http://search.proquest.com/docview/195934781?accountid=37714

An exploratory study of gang-affiliated young men's perceptions

and experiences of sexuality and gender relations
Fleury, velyneAuthor Information ; Fernet, MylneAuthor Information . The Canadian Journal of Human Sexuality
21.1 (2012): 1-15.

1. Full text
2. Full text - PDF
3. Abstract/Details
4. References 49
 Hide highlighting
Abstract
TranslateAbstract

The literature on gangs frequently associates them with gender inequality and violence against women. Some recent studies

that have addressed the sexual practices of gang-affiliated young men have painted a portrait of their romantic and sexual

experiences that is incomplete because violence and similar issues are emphasized to the detriment of overall affective and

sexual experience. The perspectives of such young men is largely absent. The present study used a qualitative

analysis of interviews with gang-affiliated young men to explore their perceptions and experiences of gang life in

terms of sexual exploitation, violence towards young women, gender relations, and romantic and sexual relationships. Our

goal was to uncover the underlying affective, sexuality-related, and identityrelated aspects of their relationships with women.

Our findings revealed some paradoxical differences between their self-reported experiences and their identified perceptions

and aspirations for a better life. [PUBLICATION ABSTRACT]

Full Text
TranslateFull text
Turn on search term navigation

Headnote

Abstract: The literature on gangs frequently associates them with gender inequality and violence against women. Some

recent studies that have addressed the sexual practices of gang-affiliated young men have painted a portrait of their

romantic and sexual experiences that is incomplete because violence and similar issues are emphasized to the

detriment of overall affective and sexual experience. The perspectives of such young men is largely absent. The present

study used a qualitative analysis of interviews with gang-affiliated young men to explore their perceptions and

experiences of gang life in terms of sexual exploitation, violence towards young women, gender relations, and romantic and

sexual relationships. Our goal was to uncover the underlying affective, sexuality-related, and identityrelated aspects of their

relationships with women. Our findings revealed some paradoxical differences between their self-reported experiences and

their identified perceptions and aspirations for a better life.

Acknowledgements: This study received financial support from the Facult des Sciences Humaines of the Universit du

Qubec Montral and the Centre de Recherche Interdisciplinaire sur la Violence Familiale et la Violence Faite aux

Femmes (CRI-VIFF). For generously sharing their stories, warm thanks goes to Will, Tempo, Franois, Nellyville, Pop,

Mexique, Alex, Pantera, Loco, and Max.

Introduction

Few studies have adequately investigated the etiology of sexism, sexual exploitation, or aggression in gangs. The available

data come from studies on young women who associate with gangs (Chesney- Lind, Shelden, & Joe, 1996; Dorais &

Corriveau, 2006; Fournier, Cousineau, & Hamel, 2004) or studies that do not specifically address gang-related sexuality

(Danyko, Arlia, & Martinez, 2002; Decker & Van Winkle, 1996; Hamel, Fredette, Blais, Bertt, & Cousineau, 1998; Perreault

& Bibeau, 2003; Sanders, 1994). To our knowledge, Totten (2000) is the only author who has considered the
viewpoint of gang-affiliated young men on violence in romantic relationships. Most studies depict a fragmented portrait of the

sexual and romantic experiences of gang-affiliated young men where violence is at the forefront to the detriment of a

broader overall picture of their experience. The literature reviewed below contextualizes the perceptions and

experiences of gang-affiliated young men in regard to their sexuality and gender relations.

Background

Many studies have described the backgrounds and social environments (suburbs, family, school) experienced by gang-

affiliated youth (Covey, Menard, & Franzese, 1992; Goldstein, 1991; Hamel, Fredette, Blais, Bertt, & Cousineau, 1998;

Maxson, Whitlock, & Klein, 1998; Wood, Furlong, Scozzari, & Sosna, 1997). Collectively, such studies can help us to

understand the processes of gang affiliation and disaffiliation experienced by young men.

It should be noted that there is no established consensus on the definition of gangs, gang members, or gang activities.

Therefore, it is difficult to assess the extent of criminal gang activity and related issues such as delinquency, violence, and

sexual exploitation. Moreover, gang members have different images and develop different types of relationships within the

various groups to which they belong. Hebert, Hamel, and Savoie (1997) defined the term "gang" as "a community of persons

(adolescents, young adults or adults) who share a common identity, who regularly interact in small or large groups, and take

part in criminal and/or violent activities" (p. 41).

Family environment and violence against women

The family environment predictors of gang affiliation are complex and cannot be easily generalized. In many cases,

deunification (e.g., immigration, separation, foster family) and parents' personal problems (e.g., alcoholism, mental health)

have deprived many gang-affiliated young men of their parents' presence (Danyko et al., 2002; Hamel et al., 1998; Maxson,

Whitlock & Klein, 1998; Patton, 1998). Many gangaffiliated young men come from single-parent families and have been

brought up by women (Danyko et al., 2002; Hamel et al., 1998; Patton, 1998). Although they may speak about their mothers

with respect, this respect does not extend to other women (Patton, 1998). The father's violence toward his wife or children

has impacted the family experience ofmany gang-affiliated young men (Danyko et al., 2002; Hamel et al., 1998; Patton,

1998). All the gangaffiliated young men met by Totten (2000) who were violent against women mentioned that their parents

adhered rigidly to traditional family and gender roles.

In many other cases, parents' personal behaviour is not considered a root cause of gang affiliation, or at least not the main

one. Many studies have pointed to multiple predictors of gang affiliation such as economic marginalization, poverty, peer

influences, school, individual characteristics, experience of victimization, and community conditions (Hill, Hawkins, & Battin-

Pearson,1999; Thornberry, Krohn, Lizotte, Smith, & Tobin, 2003; Thornberry, Lizotte, Krohn, Smith, & Porter, 2003).

Gender relations and sexuality in gangs

Stereotyped sex roles, including domination over and violence towards women, are learned in families, and they appear to

crystallize in gangs (Totten, 2000). As is the case with violence and delinquency, sexual conquests and insensitivity to

young women also result in greater gang status and help to maintain a masculine image for gang-affiliated young men

(Dorais & Corriveau, 2006). Nevertheless, a reputation for masculinity is never permanent, and must be continuously

reaffirmed (Miller & White, 2003).

Whether or not they are exploited, women become a central priority for gang-affiliated young men (Girard & Ttreault, 2005;

Palmer & Tilley, 1995; Perreault & Bibeau, 2003). The prestige and status associated with gangaffiliation appears to be

attractive to some young women. Dorais and Corriveau (2006) note that, regardless ofthe qualities and appearance of gang-

affiliated young men, gang membership increases tenfold the possibility of having sexual relationships with young women.

This attraction persists even when the young women know that the members they are with have sexually abused other

young women (Palmer & Tilley, 1995). Some young women approach gang-affiliated young men and offer themselves as

sexual partners (Dorais & Corriveau, 2006; Palmer & Tilley, 1995; Perreault & Bibeau, 2003).

Relationships between young women and gangaffiliated young men are characterized in the literature as involving

seduction, manipulation, and violence (Miller & White, 2003; Perreault & Bibeau, 2003). The most virile and violent young

men are perceived to be the most respected. Generally, young women are placed in an inferior position (Chesney-Lind et

al., 1996; Miller, 1998). Unless these young women are protected by a gangmember (e.g., brother, boyfriend), they may

become regular victims of psychological, physical, and sexual violence perpetrated by the members of their

own gang (Miller, 1998). However, it seems that the role of the girls is changing. Some girls now occupy central positions

and create their own niche of criminal activities (Chesney-Lind et al.). Individually, not all young men are involved to the

same extent in violence towards young women. Some young men initiate violence, whereas others take part in it mainly to

avoid being rejected. Those who premeditate their acts are more conscious, detached, and centered on the advantages,

and less sensitive to the well-being of the young women. Others seek the acceptance of peers through their first violent act.

Subsequent acts would bring secondary benefits and advantages (Perreault & Bibeau, 2003).

Conflicted views, double standards, sexual prowess, and insensitivity

Gang-affiliated young men are characterized as having a negative view of women (Chesney-Lind & al., 1996; Miller, 1998;

Patton, 1998; Perreault & Bibeau, 2003; Sanders, 1994; Totten, 2000) and embracing sexual stereotypes (Girard &

Ttreault, 2006). The sexual exploitation, sexual abuse, and violence experienced by a substantial number of young women

who associate with gangs is thought to stem from this perception ofwomen (Fournier et al., 2004; Patton, 1998; Sanders,

1994). As is the case in other social groups, a majority ofgang-affiliated young men report negative impressions of young

women who engage in one-night stands (Dorais & Corriveau, 2006; Miller & White, 2003; Perreault & Bibeau, 2003).

Because relations with multiple partners and hypersexuality bolster the masculine image and serve as a model for

masculine behaviour (Dorais & Corriveau, 2006; Miller & White, 2003), there is peer pressure for gang members to be

sexually active (Totten, 2000).

At-risk sexual behaviours

At-risk sexual behaviours are associated with gangaffiliation (Voisin et al., 2004). Often occurring at parties, sexual relations

are usually preceded by the consumption of psychotropic substances (Brooks, Lee, Stover, & Barkley, 2009; Crosby,

Salazar, & DiClemente, 2004; Silverman et al., 2006; Voisin et al., 2004). In addition, young women are frequently

intoxicated, voluntarily or not, when they engage in group sexual relations or aggressive and exploitative relationships

(Fournier et al., 2004; Sanders, 1994). Compared to other troubled youth, gang-associated young men have more sexual

partners and more frequent sexual behaviours (Crosby et al., 2004; Palmer & Tilley, 1995; Raj et al., 2007; Voisin et al.,

2004). They are less likely to use condoms, and consequently more likely to be involved in a pregnancy or to contract a

sexually transmitted infection (STI) (Voisin et al.).

Sexual aggression and gangbangs

Sexual aggression and collective rape (known as gangbangs) are crimes that are seldom reported by young

male gang members (Perreault & Bibeau, 2003; Sanders, 1994). When gang-affiliated young men address the subject, they

do not speak openly, preferring instead to report anecdotes or the experiences of their peers. Sanders (2004) noted that

perceptions of sexual aggression among such young men do not correspond to public or clinical perceptions. In

general, gang-affiliated young men consider that young women take part in a game where they know the rules. Furthermore,

they emphasize that "good" girls are either not invited to their parties, do not attend, or leave before the sexual "games"

start. Because many young women who are associated with gangs share these perceptions, the extent of sexual aggression

is likely to be underestimated (Fournier et al., 2004).

Sexual exploitation

Pimping contributes to the identity and status of gangaffiliated young men (Dorais & Corriveau, 2006). Gang-affiliated young

men who devote themselves to pimping generally target vulnerable young women (Dorais & Corriveau, 2006; Fournier et al.,

2004; Perreault & Bibeau, 2003) and Perreault and Bibeau (2003) stress that sexual exploitation requires more than threats

to maintain young women in that situation. Pimps learn skills from their peers such as seduction and "love-bombing" (e.g.,

overwhelming a young woman with attention and friendliness) (Dorais & Corriveau; Perreault & Bibeau).

Violence in romantic relationships

According to Totten (2000), young men who are violent towards women display some confusion about the components of a

healthy egalitarian relationship and that those who are considered marginal (e.g., gang-affiliated) are generally more violent

than those who are not gang-affiliated. Fournier et al. (2004) found that women who reported being victims of violence in

their romantic relationships identified jealousy, possessiveness, control, and obedience as factors that tainted their romantic

relationships with gang-affiliated young men. In some cases, such young men may consider young women as their property

(Dorais & Corriveau, 2006; Girard & Ttreault, 2006).

The present study

The literature on gang-affiliated young men tends to emphasize violence and related factors but does not capture the

perspectives of the young men who live and seek acceptance in this environment. The present qualitative study sought to

explore gang-affiliated young men's perceptions and experiences of sexual exploitation and violence against women, gender

relationships, sexuality and love.

Methods

Participants

Participants were recruited from and interviewed at the Centre jeunesse de Montral - Institut universitaire (n = 3) or

community resource centres (n = 7). Given that the study population was considered marginal, difficult to reach, and

generally wary about institutions and authority, the collaboration provided by trusted social and community workers was

essential. We therefore used the expert sorting method, which meant asking people who were familiar with the milieu to

recruit the participants (Angers, 1996). The previously cited definition of"gang" (Hbert et al., 1997) was adopted for the

present study but gang membership was determined by the perceptions of the social and community workers we consulted

and by the young men who agreed to participate.

Conceptual framework

Grounded theory (Glaser & Strauss, 1967) was utilized for this study. It aims primarily to understand the meaning of an

event by collecting data, determining key points, grouping them into concepts and categories, and from these, generating

conceptualizations as patterns emerge (Paill, 1994). The process does not begin with a pre-established conceptual design.

Instead, the concepts emerge from the data collected in the field. Supported by inductive logic, new concepts, assumptions

and theories were generated from empirical observations and grounded within them (Dorais, 1993).

Data collection and analysis

Five topics were covered by the interviews: (1) gang experiences; (2) family models; (3) masculinity; (4) relationships

between males and females and gang-related sexuality; and (5) suggestions for intervention. Interviews lasted from 45 to

100 minutes and were conducted in French. The interview framework was validated by a criminologist with expertise

in gangs.

Data were recorded on audiotape and transcribed verbatim. Data were analyzed according to grounded theory procedures

(Strauss & Corbin, 1997; 1998). Analyses were run using ATLAS/ti v.5 (PC software). As proposed by Laperrire (1997), an

initial open and exhaustive coding of all the interviews was carried out. Thus, data were deconstructed, extracted, and

broken down into meaning units (Deslauriers, 1991; L'cuyer, 1987). Categories were then developed inductively from the

meaning units. This categorization was accompanied by theoretical and analytical reflection. All meaning units and
conceptual categories were reorganized and the unifying links were identified. These links, grounded in the empirical data,

were continuously remodelled throughout the analysis process.

Inter-judge reliability was determined by having another researcher approve the process and review the analysis

(Deslauriers & Krisit, 1997) in order to ensure the application of appropriate procedures (Deslauriers, 1991) and the

relevance of the suggested conclusions (Laperrire, 1997).

Ethical considerations

This study received the approval of the ethics committee of the Dpartement de sexologie at the Universit du Qubec

Montral and was conducted with the support and collaboration of the Centre jeunesse de Montral - Institut universitaire.

Participation in this study was voluntary. At the time of the interview, written consent was obtained. If individual names or

precise locations related to illegal or criminal acts were mentioned, the researcher was required to immediately end the

interview and inform the authorities. Participants were also informed of this. Responses were treated anonymously and

confidentially. All participants' names are fictitious and were chosen by the participants. Each participant received a

compensation of $20.

Results

Sample characteristics

Ten young men (18-25 years old; mean age 20 years), affiliated or formerly affiliated with criminal or violent gangs, took part

in individual semi-structured interviews. For the eight participants who specified it, the average duration of their affiliation

with a gang was three years. At the time of the interview, six participants maintained regular social contacts

with gang members but were not involved in violent or criminal activities. Three participants were continuously affiliated, and

only one participant had definitively broken ties with a gang (Table 1).

Eight participants had received ongoing services after a youth protection evaluation. Whether related to gangaffiliation or for

other reasons, seven participants had committed offences and had been referred under the Youth Criminal Justice Act

(YCJA). The mean age at the time of first placement was 15.5 years. At the time ofthe interviews, one participant was

staying at a youth centre and two others were on probation: one under the YCJA and the other after serving time in prison

for assault (Table 2).

One participant had been found guilty of sexual assault and reported taking part in group sexual activities. Another had been

recently accused of sexual assault, but his romantic partner had subsequently withdrawn her complaint. One participant had

served a prison sentence for assault on his former romantic partner. Another had sexually exploited young women by

pimping (Table 3).

Six participants were in a relationship at the time of the interview. Only one lived with his partner and was taking care of a

child. Two were fathers but had separated from the mothers (Table 4).

Conceptual analysis

The results obtained in the analysis were grouped into three conceptual categories: (1) Absent, stereotyped, or violent

family; (2) Maintaining a masculine image through violence, criminality, emotional insensitivity, and sexual prowess; and (3)

the need for support and sexual education to learn how to express their masculinity non-violently and to engage in healthy,

egalitarian sexual relationships. The three conceptual categories, with their associated themes and sub-themes are

summarized in Table 5. The documentation for these categories, themes and subthemes is presented below.

Category 1: Absent, stereotypical, or violent family models

The young men generally came from a traditional, stereotypical family (7/10) in which the man was the provider and the

mother took care of the home. Almost all of them have witnessed (8/10) or have been a victim (5/10) of family violence.

Immigration, separation, placement in a youth centre and parents' problems (i.e., alcoholism) had deprived many young men

(7/10) from their father's model or other males' models. Violence or the absence of a male model impacted the familial

experience. This category had two themes.

Defined by opposition to a violent father

Although they admitted having suffered from their father's absence or from violence, half the participants wished to resemble

their father. Four others distanced themselves from the paternal model, although in practice, they perpetuated their father's

domineering or violent behaviors towards women.

When I was a kid, my father-often under the influence of alcohol-was looking at my agenda and if I was late in anything he

would smash my face .... One day ... he got angry and he started smash my face like I was a dog. ... Pah! Pah! Pah! He

kicked me with his fists and his feet. Man, I stood up-Paf! ... We got into a fight. I was bleeding ... my father too. I went to my

room, I locked myself in. My mother came in to talk to me. I said, no, mom, sorry but him, he doesn't care. ... I'm his son, not

a dog. Why is he talking like that? I'm sick of it! ... Another day, there was a birthday at home, and then everybody left. My

father wanted to go but my mother didn't want to let him leave. My father got angry, he said, "OK, you're my wife, not my

mother." He started to hit her in front of everybody, including his own friends. I saw that nobody was doing anything. ... I

called the police, the police came in [laughing].... My male model? I don't know. (Mexique, 22) My dad beat my mum. Me, in

my relationship, I don't want us to be like my dad and my mom ... because I like the person who's with me and if I love her, I

won't hurt her, I won't beat her, or I won't cheat on her .... (Pantera, 21)

Family messages about sexuality and gender relations: absent or incomplete

Only two participants reported that they had received information about sexuality: the first from his father about STI/HIV and

the second from his mother about respect toward women.

My mum never spoke to me about that. She just said, "You never hit a girl, that's it, you don't even touch them with a flower

petal, you can't touch them." She never told me how to treat them. So for me, I have a pretty aggressive nature. When

there's something that I don't like, I will say it and not in the best way. (Alex, 22)

I've never talked about sexuality with my dad .... With his father, we do "be careful." On some aspects, I learned by myself.

My mother, it's only, "Be careful, protect yourself." (Franois, 18)

Category 2: Maintaining a masculine image in a gang through violence, criminality, emotional insensitivity, and sexual

prowess

All the young men reported that in a gang, violence towards young women was accepted and even encouraged. The results

also suggest that violent behaviour towards women varied across gangs and across gang members. Whereas some gang-

affiliated young men were highly sexually active and engaged in sexual exploitation, others had difficulty approaching young

women. Although the experiences differed, the perceptions and norms related to sexual roles were common to all

participants. This category has eight themes most of which have two or more sub-themes.

The macho male sub-culture and conflicting views of women

All participants reported that male supremacy prevailed in their gang. Yet they also held conflicting views ofwomen. On the

one hand, they idealized, dominated, and desexualized the women they cared for (mother, sisters, girlfriend). On the other

hand, they devalued young women, used them for their own pleasure, and sometimes sexually exploited or assaulted "easy"

girls.

I don't like it [when my girl] comes to the park with me. Because girls that hang out in that park, they aren't seen as good

girls.... They're whores and fuck, you go with them to just chill, not to have something serious. (Mexique, 22)

Peer influence

Seven participants mentioned the influence of peers on their consumption habits as well as their delinquency behaviours

and behaviours towards women. Participants expressed their admiration for peer models. Like them, they wished to obtain

status, money, valuable objects, and women quickly and without too much effort. Peers sometimes explicitly encouraged

violence, delinquency, and the sexual exploitation of women.

Ever since I was a kid, I always thought about money and that I wanted money. ... My cousins, they always had pretty girls

.... I started imitating them. They talked all sweet to girls, all gentleman-like. Just until they wanted to pimp them, then they

started doing their shit. That's how I learned to talk to girls; you have to tell them what they want to hear .... It's just that the

guys are too manipulative. It's brainwashing what they're doing [laughs]. There's nothing you can do against it. (Max, 19)
Opportunities for and obstacles to sexual encounters

Most participants (8/10) reported that gang affiliation provided opportunities for sexual gratification.

Opportunities for uncommitted sexual encounters In the gang, sexual relations were distinguished from committed romantic

relationships. Often occurring at a party, sex was usually preceded by the consumption ofpsychotropic drugs.

In general, it's true, there are some who could exchange girlfriends. There are some who have sex lives, it's like two, three,

four girls .... In reality, it's rare that it doesn't work. Sexual relations: girls like it, sometimes tripping out with guys. (Pop, 24)

One guy who drinks alcohol, who's drunk and the girl is also drunk, it's more just for fun. It's more to feel close. It's just for

one night. (Franois, 18)

They trip out. More guys, more girls. And they exchange girls. It's like if you're getting a glass of water for someone and he

just keeps passing it on: "Yo, I'm done with her. Yo, you go, man." That's it! And yeah. No respect, and they don't care, it's

what's they see, that's what they're gonna do. (Will, 19) Obstacles to sexual encounters

One participant mentioned that certain gangs were unable to attract young women and that he was not very sexually active

when he belonged to a gang.

There are girls that say, "I'm gonna chill." ... What they're looking for is to have fun, sleep with a guy, after they sleep with

another guy. Like, I've never really experienced [that]. When you're in a gang, you don't have much of a sexuality, you aren't

really active. (Tempo, 19)

Risk taking and STI/HIV

More than half of the participants (6/10) reported unprotected sex during gang-affiliated sexual encounters.

The people that do gangbangs, they don't use protection. Well, in any case, that's what the girls told me. To me they

seemed dirty. It's for that reason that I never wanted to try it, because I could catch AIDS from another guy. ... The girls,

they're stupid .... Seems to me that a girl who goes out on the weekend, if she knows she's going to hook up, she would take

some condoms. (Max, 19)

Proving one's virility through insensitivity and violence

In gang life, violence is central and a principal way to express masculinity. A majority of the participants (8/10) reported the

existence of implicit rules that required insensitivity toward young women. Most participants (7/10) reported that to prove

their virility they needed to be sexually active.

Being insensitive
Within the context of the gang, all participants felt that they had to demonstrate toughness toward young women and to hide

their sensitivity.

We have to prove that we're tough with the girls. ... If you're weak, you don't have to show that you're weak. Otherwise the

others, they'll treat you like you're the lowest.(Pantera, 21)

More than half of the participants (6/10) described the insensitivity of their relationships with young women. They indicated

that they had to remain distant, insensitive and refrain from demonstrating empathy toward young women.

In a gang, there's no love towards a girl [laughs] ... When you're in a gang, you learn that you shouldn't love a girl. You can't

say, "Okay, I'm gonna go out with that girl, she's gonna be my girlfriend forever." ... When you're in a gang, you become

cold-hearted. (Tempo, 19)

Being sexually active so as not to lose face

Most of the participants (7/10) reported that if a young man refused an opportunity for sex his virility might be questioned.

Guys rarely, really rarely [refuse having sex]. They could say, "Yo, a girl I don't even know if she's infected with something.

Yo, ten guys on one girl ... I'm not down with that." He moves. After [the others say to him], "Yo chicken (you're a little

scardy), pussy hole" stufflike that. ... Since he was treated like a pussy, for sure it's a straight up battle. It's a serious

disrespect. Like, he could shoot the other guy just 'cause he was treated like a pussy hole. Like taunts or stufflike that. (Will,

19)

Actively participating in or being a silent witness to violence toward young women

All participants reported that violence toward young women was accepted within their gang.

Not opposing violence even when in disagreement The majority (7/10) perceived themselves as silent witnesses even

though some felt uncomfortable with the behaviour or disagreed with it but none openly opposed it out of fear of peer

rejection or violence.

Me, personally, [sexually assaulting a girl] isn't my style: "Wow! Calm down you guys." But for sure (if you said that) you're

gonna get told offor start a fight. (Alex, 22)

Minimizing their behaviours

At first, almost all participants denied that they were or had been a member of a gang. They also denied any behaviour

associated with gangs (i.e. drug use, violence and crime). But many young men reported engaging in psychological or

physical violence (6/10) or sexual exploitation (1/10). Regardless of whether or not they recognized their responsibility, all

minimized their actions or projected the blame onto their peers or the young women themselves.
I have a lot of problems with the police. For fights, aggressions, assault on a cop, breaking of bail conditions ... a lot of shit.

Now I got [a problem] for conjugal violence, because of my wife. It was just a slap. But a slap, for them, is already enough.

For me it's nothing. It was a way for me to calm her down because she had given me a slap too. (Mexique, 22)

I have too much respect for me and the girls, you know .... I find disgusting to make dance a girl that hasn't any resource,

and then take the cash .... I'm still seeing this girl [the girl he pimped] and my opinion remains the same. It's a girl that's very

limited with life. She only tries to find boys who would love her, to find masculine affection, but it doesn't work. Boys all know

she's stupid and everything. Christ, she put my name on her arm, her leg, she almost attempted suicide. I made her lost her

mind. So, she probably has big affective problem. Any guy could have done so [pimp her]. (Max, 19)

Within gang culture, sexual exploitation was considered to be a legitimate way to make money. Participants reported that

young men who sexually exploited women showed no empathy toward them.

There's no remorse in that kind of business. You have no feelings, you can't have feelings. Me, I surely had too many. That's

why I stopped .... If I'm gonna scam, I'll sell drugs. I'll make myself more cash. Well, not more cash because the girls, it's

about $500 to $1000 a night. It's just that it's shitty. You just make the girl fall in love with you and then you send her out to

go dance .... It's like a sale. (Max, 19)

Having a confused view of consent

According to a number of participants, general, legal, and societal norms related to consent are not respected in

the gang culture.

The majority are consenting but they're drunk. It's happened where one couldn't remember what had happened. But there

were never any complaints or anything .... Since they were all drunk, I think that they just sort of let it happen. (Alex, 22)

They think only of their pleasure. The girls have no choice. They can't say, "Oh, no!" and just close their legs. It's like a

collective rape, that's exactly what it is. [Except the guys] see it more like a game. Cat and mouse. It's fucked up. (Max, 19)

Romantic relationships and paternity

Like many young men in other social groups, those affiliated with gangs want to find love and some of them also want to

have a family.

Maintaining a romantic relationship

Some of the participants spoke about the challenge of maintaining relationships within gang culture.

A guy who's in love with a girl, there's gonna be jealousy. When you send a girl out to dance, for sure you're not sending her

to a club where it's $10 a dance. You're gonna look for a club where the girl can hook up with the client for $80. A guy who's
in love with a girl will be jealous, he's not gonna want to send her out dancing. So it's better if you get yourself a girl that you

won't fall in love with. (Max, 19)

Peers sometimes create conflicts in order to break a couple up because in their view romantic relationships tend to limit the

young men's investment in the gang.

Loverboy. He's always with his girlfriend. Sometimes you'll say, "Yo, are you coming with us?" Answer: "Oh, no, I'm gonna

check up on my girl." It's like he's saying, "Oh no, you're a loverboy. ... You're just here to screw around." They go to the

movies, they do activities together. ... Then the others [those who don't have girlfriends] have to go collect money. "Okay, I'll

check up on you later, I'll call you, ciao!" (Will, 19)

Being in love with a young woman who is not associated with a gang

Without admitting it to their peers, the majority of the participants dreamed of being in love (5/10) or were in love (4/10) with

a young women who was not associated with a gang which could become a source ofconflict.

With girls, I prefer that it's calmer. ... She has the right to have fun, that's for sure. It's a girl, it's human. But, it's more to stay

at her place. It's more for staying at home. ... She's going to love me like I am: hard with my feelings. (Pantera, 21)

They gonna laugh at him, saying: "Yo, why you don't make her dance [pimping], you know?" Something like that. Sometime

they don't even show respect for your girlfriend. Like: "Yo, give her to me and you'll see, I'm going to make her dance."

Sometime they have no respect and you get mad. (Will, 19)

Having difficulty attracting young women who are not associated with a gang

Some of the participants (6/10) reported that the "good" women were not interested in gang-affiliated young men because

they didn't like the associated images of drug use and violence.

When you're in a gang, there are girls that don't like it. Like if you're in a gang, you're a player. You're not just seeing her. ...

When you're not in a gang, she's going to have more trust in you. (Nellyville, 18)

Distancing or disaffiliation

Most participants (8/10) said that committed romantic relationships and paternity had motivated them to distance themselves

or disaffiliate from a gang.

So if she's sincerely interested in you and you are too, it seems to me that you're going to have to make a change. It's what

happened to me. .... It's because of her that I stopped the first time. And that I started going to church and became Christian.

Slowly, but surely, I succeeded in staying two years sober. (Alex, 22)
All my friends, the most of them, they all stopped because of it [paternity]. Maybe the same thing will happen to me. I want

my baby. (Mexique, 22)

Now, I'm more adult. Now, I don't react. I have my family already, I have a son. So I think of him more than ofmyself. I don't

want him to be like I was. I don't want him to react, whether he's on the streets, whether he's smoking, drinking. I prefer that

he goes to school and he plays sports. It's for that reason that I don't want to know anything about it. (Pantera, 21)

Violence in romantic relationships

Several participants reported that as much violence occurred in romantic relationships with gangaffiliated young women

(4/10) as in relationships with young women who were not gang-affiliated (3/10).

Violence in gang-related romantic relationships

Four participants talked about the violence in their romantic relationships. Violence generally occurred in public (at a party or

on the street) and in the presence of gang members. Three of these four participants mentioned that real or anticipated

infidelity could lead to violence by the young men.

Sometimes, everyone spoke a lot about her [ex-girlfriend]. ... Sometimes I drank too much and I was too drunk. Then a

bunch of stuffthat people were telling me so I'd jump on her. Sometimes I would push her, stufflike that. (Loco, 19)

Violence in romantic relationships with young women who are not associated with gangs

Three participants explained that violence toward young women was often motivated by jealousy or infidelity. One of them

stated that negative comments by peers could also give a motive for violence.

Me, yeah, I can go out. My girl, she has to stay home. You have to be tough because on the streets you learn so much

stuffso that after, you don't really trust other people. It hardens your heart. It crosses my mind, you gotta behave like a

severe man. (Pantera, 21)

Category 3: The need for support and sexuality education

When questioned about possible ways to improve their gender and sexual relations, half the young men brought up

disaffiliation from gangs.

Discussions with significant adults

Half of the participants believed that adults could play a major role to help young men become aware of the negative

consequences of gang affiliation, violence, and consumption habits. They felt it was important for adults to speak to them

about their future. One specified that adults should inspire trust, stimulate motivation, and be there for young men when
problems arise. Another proposed talking about family relationships with gang-affiliated young men because, according to

him, these relationships were the cause of the problem as well as part of the solution.

Also talk to them about how it happens in a family relationship. Because they often have parents that beat them all the time.

When you have violent parents, you're better offbeing with a gang than your parents. ... And if possible, talk to their parents,

to pay them, to give them more attention, more time for them. Because they often don't have the time. (Tempo, 19)

Peer influence

One participant mentioned the importance of getting credible, trustworthy people involved, such as family members or young

men who had formerly belonged to a gang.

Well, it's gotta be guys who have experienced it that they go see. Like, let's say I would go see a big gang guy, a leader

would say, "What's he talking about?" Like, if it's a guy that talks like him, it's a guy that's really changed, he's gonna know

that it's real. ... He needs to see an example. It needs to be someone close so that he takes it seriously. (Franois, 18)

Receiving sexuality education

Two participants proposed interventions aimed at reducing violence in romantic relationships as well as sexual aggression.

One proposed dealing with violence by increasing young men's awareness about the future consequences (e.g., infections,

imprisonment). Some proposed that young women could be encouraged to denounce violence.

With the guys, they cannot be violent. Violence does nothing. ... Giving women more respect. To talk to them, what they're

thinking of doing later. If they're thinking about having a family, that they don't do that. (Tempo, 19)

Two young men raised the need for information about STI/HIV. They suggested that significant adults could provide support,

not only by educating them, but also by helping them become aware of the potential consequences of their sexual

behaviours.

You have to talk to them about protection. I don't know if it's a money problem. ... [Social or community workers] should go

talk about it and say, "Guys, you really need to protect yourself because viruses propagate really quickly." Protecting

themselves, they have to talk about it. Sometimes we don't talk about it with them and we want them to learn on their own.

They're not going to learn on their own. You have to talk to them about it. (Tempo, 19)

Discussion

The literature reviewed for the present study documented the association of gang affiliation with sexual stereo-typing,

insensitivity, sexual exploitation, and violence against women. We anticipated that gang-affiliated young men's perceptions,

experiences and issues related to sexuality and gender relations extended beyond violent behaviours. Our interviews with

young men who had been or were currently associated with gangs provided insight into those issues and expanded our
perspective. The participants found it difficult to talk about their personal experiences, particularly when discussing sexuality

and gender relations. Several admitted that they were addressing these subjects for the first time with an adult. Generally,

they dissociated themselves from all violent acts towards young women, even if they had participated in

domineering gangbehaviours or been accused of violence associated either with a gang or in a love or sexual relationship.

They revealed general confusion about the concepts of violence and consent and acknowledged the

influence ofgang expectations on their public actions and the way they portrayed themselves.

Most participants followed implicit gang rules that fostered insensitivity, dominance, or violence towards women. To prove

their virility, they engaged in multiple sexual encounters or group sexual aggressions, sometimes against their will. When

they had sexual relations, the young men described their sexuality as deprived of intimacy (gangbangs, relations without

attachment, sexual exploitation). Objects of desire were segregated: love without desire for "good" girls, and desire without

love for "bad" girls. The erotic scenarios ofgang-affiliated young men, like those of other youth, were influenced by their

environment. The gangstandards and values described by the participants included: male domination, emotional

insensitivity, sexual double standards, use of woman as objects of pleasure, and use of women for exchange or exploitation

(e.g., pimping). Some criminal gangs appeared to provide relatively frequent opportunities to take part in sexual activities

involving domination, lack of consent, or violence.

In contrast to these overt and expected attitudes and behaviours, the majority of our participants dreamed ofbeing with a

young woman who was not associated with a gang, and who held traditional values. Yet their criminal, violent, and sexual

activities prevented this type of young woman from being interested in them.

Our findings illustrate the gap between the sexual experiences of gang members and the committed romantic relationships

that the majority of the young men hoped for. An intimate relationship requires a certain amount of sensitivity and emotional

proximity (self-expression, listening, and empathy). These attributes are not valued by gangs or, in some cases, by

the gang members' families. Several of the young men had difficulty developing these relational skills. Generally, committed

romantic relationships appeared to be incompatible with affiliation to a criminal gang. It is therefore possible that

several of these young men will be unable to commit to a relationship until they reach adulthood or decide to leave the gang.

Study limitations

This investigation was designed as an exploratory, qualitative study but, in retrospect, triangulation methods (Laperrire,

1997) would have broadened and diversified our perspective on the topic. The lack of empirical saturation, the small sample

size, and the homogeneity of the participants (e.g., recruitment areas, ethnic origin, age) suggest that generalization of the

findings should be done with caution (Pires, 1997). While the social and community workers we consulted were helpful, they

may have selected young men with whom they had a good relationship. As a result, young men who had not been met by

social or community workers (and might therefore have been more disposed to criminality and gang affiliation), may have

been underrepresented. When they were addressing sexuality and violence towards women, the young men in our sample
were often evasive and tended to express more ambiguous and contradictory statements. It is possible that these topics

were consciously or unconsciously censored because of social desirability.

We used systematic theoretical sampling to maximize transferability of the results. However, the problems cited above

related to recruitment, the lack of empirical saturation due to the limited number of subjects, and sample homogeneity made

it impossible to paint a complete portrait of the issues studied. Strategically selected variables allowed some

diversification of the cases covered. Moreover, the detailed description ofthe methods could help other researchers establish

the extent to which the characteristics of the studied population are similar to those of other populations (Deslauriers &

Krisit 1997). Data on the antecedents ofviolence and violence towards women should have been gathered, and should

receive particular attention in future studies. Given the importance of the identity-related, emotional, and sexual aspects

addressed in this study, a deeper understanding could have been gained had we examined a broader sample.

Fleury, ., & Fernet, M. (2012). An exploratory study of gang-affiliated young men's perceptions and experiences of sexuality and

gender relations. The Canadian Journal of Human Sexuality, 21(1), 1-15. Retrieved from

http://search.proquest.com/docview/1032532443?accountid=37714

OUT 6

Galatzer-Levy, R. (2002). Created in others' eyes. Adolescent Psychiatry, 26, 43. Retrieved from

http://search.proquest.com/docview/206081173?accountid=37714