What Are the Risk Factors?

Individual and Family Demographics

Traditionally, the typical gang member is male, lives in the inner city, and is a member of a racial or ethnic minority. Although these characteristics may be prevalent among gang members, it should not be assumed that all, or even the overwhelming majority of, gang members share these demographic qualities. In addition to changes in the geographical distribution of gangs (that is, the proliferation into nonurban areas) documented by Klein (1995); Curry, Ball, and Fox (1994); Curry, Ball, and Decker (1996); and the National Youth Gang Center (NYGC) surveys, research in the past 20 years has highlighted the presence of girls in gangs (Bjerregard and Smith, 1993; Chesney-Lind, 1997; Curry, 1998; and Esbensen and Winfree, 1998). Evidence also shows that gang membership is not restricted to youth from racial and ethnic minorities.

Gang behavior has been described almost exclusively as a male phenomenon. Law enforcement estimates generally indicate that more than 90 percent of gang members are male (Curry, Ball, and Fox, 1994). Early references to female gang members were usually restricted to their involvement in sexual activities or as tomboys; they were rarely included in any serious discussions about gangs. The little that was said about gang girls suggested that they were socially inept, maladjusted, and sexually promiscuous and that they suffered from low self-esteem.

Recent survey research, however, suggests that females may account for more than one-third of youth gang members (Esbensen and Winfree, 1998). In addition, a number of contemporary researchers have moved beyond the stereotypical notion that female gang members are merely auxiliary members of male gangs and have proposed gender-specific explanations of gang affiliation (Campbell, 1991; Chesney-Lind and Shelden, 1992; Fishman, 1995; Miller, 1998). Some researchers have explored the possibility that girls join gangs in search of a sense of belonging to a peer "familial" group (Giordano, 1978; Harris, 1988; Joe and Chesney-Lind, 1995). For example, in an ethnographic study of Latina gang members in male-dominated Hispanic gangs in the San Fernando Valley of California, Harris (1988) concluded that Latina gang members were lost between two worlds—Anglo and Mexican American society and culture. The complex social and cultural roles of Latinas, according to Harris, are displayed in Latina gang membership and behavior in which females found peers with whom they could relate. The females would "fight instead of flee, assault instead of articulate, and kill rather than control their aggression" (Harris, 1988:174).

Another myth about the demographics of gang youth is that they are almost exclusively members of ethnic or racial minorities. Some law enforcement estimates and studies based on law enforcement samples indicate that 85 to 90 percent of gang members are African American or Hispanic (Covey, Menard, and Franzese, 1997). However, more recent law enforcement estimates from the 1998 National Youth Gang Survey (National Youth Gang Center, in press) indicate that earlier estimates may overstate the minority representation of gang members. The survey revealed that the race or ethnicity of gang members is closely tied to the size of the community. While Caucasians constituted only 11 percent of gang members in large cities (where most gang research has taken place), they accounted for approximately 30 percent of gang members in small cities and rural counties. Lending credence to law enforcement estimates are ethnographers' depictions of gang youth, usually based on research conducted in socially disorganized communities (that is, characterized by high rates of poverty, mobility, welfare dependency, and single-parent households) in Los Angeles, New York, or other urban areas with high concentrations of minority residents. More general surveys that examine youth gangs also tend to be restricted to specific locations that do not include diverse population samples. For example, longitudinal studies in Denver and Rochester (Bjerregard and Smith, 1993; Esbensen and Huizinga, 1993; Thornberry et al., 1993), part of the OJJDP-funded Program of Research on the Causes and Correlates of Delinquency, were concentrated in high-risk neighborhoods that (by definition) included disproportionate representation of racial and ethnic minorities.

It is worthwhile to note that the early gang studies provided a rich source of information about white urban gangs. These early gangs were usually described according to nationality and/or ethnicity, not race. Researchers began to identify gang members by race in the 1950's (Spergel, 1995). This change in gang composition is closely tied to the social disorganization of urban areas and the research focus on urban youth. Covey, Menard, and Franzese (1997:240) suggested that the scarcity of non-Hispanic, white, ethnic gangs may be attributable to the smaller proportion of non-Hispanic European Americans residing in neighborhoods characterized by social disorganization.

As research expands to more representative samples of the general population, a redefinition of the racial and ethnic composition of gang members is likely. Esbensen and Lynskey (in press) report that community-level demographics are reflected in the composition of youth gangs; that is, gang members are white in primarily white communities and are African American in predominantly African American communities.

Family characteristics of gang members, such as family structure and parental education and income, also have been revised, because the traditional stereotype of gang members as urban, minority males from single-parent families is too restrictive. In fact, gang youth are found in intact two-parent, single-parent, and recombined families. In addition, gang youth are not limited to homes in which parents have low educational achievement or low incomes. Klein (1995:75-76) summarizes gang characteristics as follows (emphasis added):

[I]t is not sufficient to say that gang members come from lower-income areas, from minority populations, or from homes more often characterized by absent parents or reconstituted families. It is not sufficient because most youths from such areas, such groups, and such families do not join gangs.

Although it would be erroneous to conclude that demographic characteristics alone can explain gang affiliation, individual factors are nevertheless clearly associated with gang membership; that is, minority youth residing in single-parent households are at greater risk for joining gangs than are white youth from two-parent households.

Personal Attributes

Some researchers (for example, Yablonsky, 1962) have found that, compared with nongang youth, gang members are more socially inept, have lower self-esteem, and, in general, have sociopathic characteristics. Moffitt (1993) stated that youth gang members are likely to be "life-course persistent offenders." To what extent are such depictions accurate? Are gang youth substantially different from nongang youth? Recent surveys in which gang and nongang youth's attitudes were compared found few consistent differences.³ This lack of consistent findings, however, may reflect differences in survey methods and question content. Comparisons between gang and nongang youth have been reported from Rochester (Bjerregard and Smith, 1993), Seattle (Hill et al., 1999), and San Diego (Maxson, Whitlock, and Klein, 1998). These authors used different questions and different sampling methods and reported slightly different findings. In the Seattle study, Hill and colleagues (1999) found that gang youth held more antisocial beliefs, while Maxson, Whitlock, and Klein (1998), among others, found that gang members had more delinquent self-concepts (based on statements such as the following: "I'm the kind of person who gets into fights a lot, is a bad kid, gets into trouble, and does things against the law."), had greater tendencies to resolve conflicts by threats, and had experienced more critical stressful events. On a more generic level, both the Seattle and San Diego studies found significant differences between gang and nongang youth within multiple contexts; that is, individual, school, peer, family, and community characteristics.

Extending this comparative approach, Esbensen, Huizinga, and Weiher (1993) examined gang youth, serious youthful offenders who were not gang members, and nondelinquent youth. Their findings indicated that the nondelinquent youth were different from the delinquent and gang youth—nondelinquent youth reported lower levels of commitment to delinquent peers, lower levels of social isolation, lower tolerance for deviance, and higher levels of commitment to positive peers. In a partial replication of the study by Esbensen, Huizinga, and Weiher (1993), Deschenes and Esbensen (1997) found a continuum extending from nondelinquent to minor delinquent to serious delinquent to gang member. Based on delinquency scores, they categorized eighth grade students into one of these four classifications. On every measure tested, gang members were significantly different from each of the other groups but were clearly the most distinct from nondelinquents (generally, at least one standard deviation above the mean). Gang members were more impulsive, engaged in more risk-seeking behavior, were less committed to school, and reported less communication with, and lower levels of attachment to, their parents. Nongang youth were more committed to prosocial peers and less committed to delinquent peers.

Using a somewhat different approach, Esbensen et al. (in press) examined differences among gang members. They classified gang members on a continuum, beginning with a broad definition of gang members and gradually restricting the definition to include only those youth who claimed to be core members of a delinquent gang that had a certain level of organizational structure. They found significant attitudinal and behavioral differences between core gang members and those more broadly classified as gang members. They did not find any differences in regard to demographic factors.

In another report from the Seattle study, Battin-Pearson and colleagues (1997) compared nongang youth, transient gang youth (members for 1 year or less), and stable gang youth (members for 2 or more years). Both the transient and stable gang members differed significantly from the nongang youth on a variety of attitudinal and behavioral measures. However, few distinctions between the transient and stable gang members were found. The measures on which differences occurred tended to represent individual- and peer-level measures (for example, personal attitudes and delinquency of friends).

Research shows that the notion of youth joining gangs for life is a myth. While some members make the gang a lifelong endeavor, findings from three longitudinal studies indicate that one-half to two-thirds are members for 1 year or less (Battin-Pearson et al., 1997; Esbensen and Huizinga, 1993; Thornberry, 1998).

Peer Group, School, and Community Factors

One consistent finding from research on gangs, as is the case for research on delinquency in general, is the overarching influence of peers on adolescent behavior (Battin-Pearson et al., 1997; Menard and Elliott, 1994; Warr and Stafford, 1991). In their comparison of stable and transient gang youth, Battin-Pearson and colleagues reported that the strongest predictors of sustained gang affiliation were a high level of interaction with antisocial peers and a low level of interaction with prosocial peers. Researchers have examined the influence of peers through a variety of measures, including exposure to delinquent peers, attachment to delinquent peers, and commitment to delinquent peers. Regardless of how this peer affiliation is measured, the results are the same: Association with delinquent peers is one of the strongest predictors (that is, risk factors) of gang membership.

Gang researchers examine school factors less frequently than other factors. However, they have found that these issues are consistently associated with the risk of joining gangs. Research indicates that gang youth are less committed to school than nongang youth (Bjerregard and Smith, 1993; Esbensen and Deschenes, 1998; Hill et al., 1999; Maxson, Whitlock, and Klein, 1998). Some gender differences have been reported in regard to this issue. In OJJDP's Rochester study, expectations for educational attainment were predictive of gang membership for girls but not for boys. In a similar vein, Esbensen and Deschenes (1998) found that commitment to school was lower among gang girls than nongang girls. No such differences were found for boys. Studies that examine juveniles' cultures and ethnic backgrounds also attest to the role of school factors in explaining gang membership (Campbell, 1991; Fleisher, 1998).

The community is the domain examined most frequently in regard to both the emergence of gangs and the factors associated with joining gangs. Numerous studies indicate that poverty, unemployment, the absence of meaningful jobs, and social disorganization contribute to the presence of gangs (Curry and Thomas, 1992; Fagan, 1990; Hagedorn, 1988, 1991; Huff, 1990; Vigil, 1988). There is little debate that gangs are more prominent in urban areas and that they are more likely to emerge in economically distressed neighborhoods. However, as previously stated, surveys conducted by NYGC during the 1990's identified the proliferation of youth gangs in rural and suburban communities. Except for law enforcement identification of this phenomenon, few systematic studies have explored these rural and suburban youth gangs. Winfree, Vigil-Backstrom, and Mays (1994) studied youth gang members in Las Cruces, NM, and Esbensen and Lynskey (in press) looked at gang youth in rural areas and small cities that were included in an 11-site study. Although neither of these reports addressed environmental characteristics, they did indicate a substantial level of violence by these gang members.

The traditional image of American youth gangs is characterized by urban social disorganization and economic marginalization; the housing projects or barrios of Chicago, Los Angeles, and New York are viewed as the stereotypical homes of youth gang members. The publication of Wilson's (1987) account of the underclass—those members of society who are truly disadvantaged and affected by changes in social and economic conditions—has renewed interest in the social disorganization perspective advanced by Thrasher (1927) and Shaw and McKay (1942). Los Angeles barrio gangs, according to Vigil (1988) and Moore (1991), are a product of economic restructuring and street socialization. Vigil (1988:9) refers to the multiple marginality (that is, the combined disadvantages of low socioeconomic status, street socialization, and segregation) of both male and female gang members who live in these socially disorganized areas. In addition to the pressures of marginal economics, these gang members experience the added burden of having marginal ethnic and personal identities. They look for identity and stability in the gang and adopt the cholo subculture—customs that are associated with an attachment to and identification with gangs—that includes alcohol and drug use, conflict, and violence. According to Moore (1991:137-138):

Gangs as youth groups develop among the socially marginal adolescents for whom school and family do not work. Agencies of street socialization take on increased importance under changing economic circumstances, and have an increased impact on younger kids.

Social structural conditions, which have resulted in a lack of education and employment and in lives of poverty without opportunities (Short, 1996), are compounded for females, who experience the additional burden of sexual discrimination and traditional role expectations (Fishman, 1995; Swart, 1995). Social structural conditions alone, however, cannot account for the presence of gangs. Fagan (1990:207) comments that "inner-city youths in this study live in areas where social controls have weakened and opportunities for success in legitimate activities are limited. Nevertheless, participation in gangs is selective, and most youths avoid gang life." Therefore, addressing structural factors is not the only plausible strategy for gang prevention or intervention.