The Contribution of Gang Membership to the Victim

Abstract

Objective:

Although a vast literature has investigated the consequences of gang membership for offending and victimization, little is known about the contribution of gang membership to the victim–offender overlap. We advance a group process theoretical model and provide an empirical extension of the victim–offender overlap to gang membership.

Method:

Using data gathered from 621 respondents in five cities, the contribution of gang membership to the victim–offender overlap is determined by examining (1) a typology of four victim–offender arrangements using multinomial logistic regression modeling and (2) the latent propensity for violent offending and victimization using multilevel item response theory modeling.

Results:

Gang members were over twice as likely as nongang members to be both victims and offenders, even after adjusting for low self-control, adherence to street codes, and routine activities. Neither contemporary theoretical perspectives on the overlap nor the reciprocal relationship between violent outcomes eliminated the association of gang membership with violent victimization and violent offending.

Conclusion:

By theoretically and empirically integrating gang membership into current knowledge on the victim–offender overlap, the results suggest that there is much to be gained for research and practice by unpacking the features of criminal and deviant networks.

Keywords

violence offending victimization overlap gang membership

The overlap between victims and offenders is on the short list of criminological “facts” (Berg 2012; Jennings, Piquero, and Reingle 2012; Lauritsen and Laub 2007). Beginning with von Hentig’s (1948) contention that perpetrators and victims are intimately linked, and Wolfgang’s (1958) findings regarding victim-precipitated homicide, research has focused increasingly on the ties between offending and victimization. Specification of the mechanisms underlying the overlap began with Wolfgang and Ferracuti’s (1967) subculture of violence thesis, but evolved with the emergence of lifestyle/routine activity (Cohen, Kluegel, and Land 1981; Hindelang, Gottfredson, and Garofalo 1978; Osgood et al. 1996) and self-control (Gottfredson and Hirschi 1990; Schreck 1999) theories. Subcultural explanations have received renewed interest due to contemporary research on neighborhoods and culture (Anderson 1999; Jacobs and Wright 2006; Kubrin and Weitzer 2003). While an impressive body of research has documented the link between offending and victimization, recent studies have shown that these findings are moderated by age and neighborhood (see Berg et al. 2012; Berg and Loeber 2011; Menard 2012), raising continued questions with regard to the etiology of the victim–offender overlap.

Notably absent from advances in research on the victim–offender overlap is an understanding of how gang membership contributes to this relationship. Given gang members’ extremely high levels of involvement in violence, this is an important topic to address. The same field research that details violent acts perpetrated by gang members also illustrates high levels of violence against gang members—28 of the 99 (Decker and Van Winkle 1996), 5 of the 38 (Levitt and Venkatesh 2001), and 3 of the 37 (Hagedorn 1991) gang members in these respective studies died violently. Studies using official data have found homicide victimization rates for gang members were 100 times that of the general population (Decker and Pyrooz 2010; see also Lattimore, Linster, and MacDonald 1997). Given the volume of research demonstrating the association between gang membership and violent offending and victimization (Esbensen et al. 2010; Krohn and Thornberry 2008), the lack of research on the effects of gang membership on their overlap is problematic. Indeed, these outcomes have been studied independently, rather than jointly, and gang membership has been conceptualized within a lifestyle/routine activity framework or limited to a control variable.

While there are good theoretical reasons for the applicability of self-control, subcultural, and lifestyles/routine activity theories to gang membership and the victim–offender overlap, we advance a theoretical model that emphasizes group process to inform the contribution of gang membership to these outcomes (e.g., Short and Strodtbeck 1965). Our contention is that the existing inventory of theories applied to the overlap cannot account for gang-related group processes. Collective identities, normative orientations, and extra-individual liabilities of gang life should influence the victim–offender overlap in ways that go unaccounted for in existing theories. Our theoretical dynamic predicts that gang membership is related to the victim–offender overlap net of existing theoretically relevant variables. The group process orientation also anticipates that gang membership will remain related to violent outcomes net of the established reciprocal link between victimization and offending (Berg et al. 2012; Ousey, Wilcox, and Fisher 2011).

This study provides a theoretical and empirical extension of the victim–offender overlap to gang membership. Accordingly, we examine two lines of questioning (1) Does gang membership increase the likelihood of the overlap between violent victimization and offending? (2) To what extent is the relationship between gang membership and violent outcomes mediated by the causal connection between offending and victimization? Multinomial logistic regression and multilevel item response theory (IRT) modeling are used to answer these questions with data gathered in five cities from 621 respondents interviewed regarding recent gang, violent offending, and victimization experiences. Violence is the focal outcome because of the generality of theory applied to the overlap, consistency with the broader literature, and the salience of violence in everyday lives. Our overall aim is to expand knowledge of the victim–offender overlap by evaluating its explanatory boundaries and assess the role of gang membership in this dynamic.

Situating Gang Membership in the Victim–Offender Overlap

Gangs are qualitatively different from other peer groups, conventional or deviant (Klein and Maxson 2006). The distinguishing features of gangs are not simply individual members or the neighborhoods where they are situated, but the group processes within the gang (Decker, Melde, and Pyrooz 2013; Esbensen and Maxson 2012; Klein and Maxson 2006; Short 1965; Short and Strodtbeck 1965; Thornberry et al. 2003). Short and Strodtbeck emphasized these mid-level dynamics as the catalyst for criminal and noncriminal activities in gangs. Individuals in gangs participate in, and are influenced by, group processes that include collective identities, status acquisition and maintenance, normative orientations toward criminal involvement, and extra-individual liabilities. Group processes elevate risks for violence, which is why individual-level studies rarely eliminate the empirical relationship between gang membership and offending and victimization despite a rich set of static and dynamic control variables.

Collective gang identities are built around the shared experiences and lore of criminal offending and victimization. Klein and Maxson (2006:205) noted “Crime and group identity are not merely fellow travelers in the gang world: they are mutual reinforcers.” Violent events in a gang take on a mythic status as participants, actions, and outcomes are recounted, embellished, and retold to add to the shared history of the gang (Ayling 2011; Decker and Van Winkle 1996; Fleisher 1998; Klein 1971, 1995). Conflict and violence are shared experiences among gang members, aiding in identity formation and the creation of normative expectations within the gang (Densley 2012; Thrasher 1927; Vigil 1988). This socialization process solidifies gang norms and values, providing members with scripts that promote and reinforce violence (Decker 1996; Melde, Taylor, and Esbensen 2009). Just as the collective gang identity helps define in-group norms and expectations, so too does it raise the risk for victimization at the hands of one’s own gang. Many gang members are susceptible to ceremonial entrance and exit processes that include violence, such as getting jumped into/out of the gang, and the potentially violent consequences of not adhering to gang rules, such as not communicating with the police or interacting with affiliates of rival gangs (Decker and Curry 2002; Pyrooz and Decker 2011).

The collective liability and norms of reciprocity within and between gangs help unpack the connections between group processes and violence. Conflict is built into the DNA of gangs: “It’s hard to find a one-gang city; gang cohesiveness thrives on gang-to-gang hostilities” (Klein 1995:30; Thrasher 1927). Gang membership entails collective liability, and as gangs replenish their ranks, new members assume ongoing conflicts. Gang violence can act as an expanding snare and entrap a wide circle of members who were not direct participants in an initial act of violence (Brunson and Miller 2009; Decker 1996; Fleisher 1998; Miller 2001; Papachristos 2009; Sanders 1994; Short and Strodtbeck 1965). One gang-related violent incident illustrates the shared risk associated with membership. If a member of gang A is shot by a member of gang B, that violent incident is generalized to all members of gang A as a result of collective gang identity. The pool of offenders from gang A expands well beyond the shooting victim. Retaliation by gang A, in turn, is generalizable to all of gang B and is not exclusive to gang B’s triggerman. All members of gang B shoulder the liability of the assault, thus expanding the pool of victims beyond the original shooter. For example, Hughes and Short (2005) describe a practice known as “wolf-packing,” where small groups of rival gang members engaged in violence, which expanded to more members, eventually escalating to large-scale gang violence.

The cycle of gang violence is driven by norms of reciprocity (Decker 1996; Decker and Pyrooz 2010; Klein 1995; Tita and Radil 2011). Violent events in the gang context have been referred to as “gifts” that necessitate reciprocation. A Chicago gang member explained:

If someone steps to you, you best roll right-the-fuck-back-up on them! Get you some revenge. If we back down from those motherfuckers [Latin Counts], they think they got the best of us. They’d think they got more heart than we do. (Papachristos 2009:117)

Left unrequited, this “gift” festers and grows, yielding even more consequential violence. Whether real or perceived, violence functions as a symbolic threat to a gang and its members (for discussions of these perceptions, see Anderson 1999; Melde et al. 2009; Stewart, Schreck, and Simons 2006). In the gang landscape, where dominance relations are constantly evaluated and each gang’s legitimacy probed by their rivals, provocations must be addressed aggressively and swiftly (Brantingham et al. 2012; Decker 1996; Katz 1988; Lauger 2012; Papachristos 2009). As a result, gang members perpetuate an “endless game of revenge” (Vigil 1988:132) that reinforces collective norms of reciprocity and roles transition from victim to offender and back again.

It is within the context of gangs and group processes that we derive hypotheses on the overlap between victimization and offending. This framework anticipates that the overlap should be stronger among individuals exposed to contexts where gang-related group processes transpire. Gang members should be more susceptible to these processes than nongang members. Further, the relationship between gang membership and violent outcomes should not be fully mediated by the established correlation between victimization and offending because of the group processes entwined with gang membership. Much of the attenuation that is observed should be endogenous to gang membership and the dynamics of group processes. In other words, the theoretical model predicts direct and indirect relationships between gang membership and violent outcomes.

The theories most commonly applied to the victim–offender overlap all have relevance and empirical support in the gang context—routine activity and lifestyle (e.g., Taylor et al. 2008), low self-control (e.g., Kissner and Pyrooz 2009), and subcultural (e.g., Matsuda et al. 2013) theories—but are unlikely to fully account for this relationship. Perspectives promoting risk heterogeneity (Garofalo 1987; Gottfredson and Hirschi 1990; Schreck 1999; Sparks 1982) as the common source of offending and victimization have trouble explaining away peer and alternative influences even in unidirectional studies (e.g., Battin et al. 1998; Pratt and Cullen 2000). Further, risk heterogeneity is analogous to the “selection model” of gang membership and offending, a perspective that is unsubstantiated empirically (Krohn and Thornberry 2008). Routine activity and lifestyle theories contend that the “principle of homogamy,” unstructured socializing, time use, and offending elevate victimization risk (Cohen, Kluegel, and Land 1981; Hindelang, Gottfredson, and Garofalo 1978; Jensen and Brownfield 1986; Osgood et al. 1996), but are theoretically limited in explaining the extra-individual liabilities gang membership entails. Subcultural theories emphasize legal cynicism and the code of the street as a means to earn status and resolve interpersonal disputes. Gangs cannot, however, be reduced entirely to cultures of honor, street codes, or distrust in the police; to do so discounts gang processes (Erlanger 1974; Short 1965). While some of the above group processes surely extend to co-offending networks or individuals enmeshed in the street code (Anderson 1999; Black 1983; Jacobs and Wright 2006; Kubrin and Weitzer 2003), these deviant collectives lack the criminal capital, resources, and “manpower” necessary to sustain the cycle of violence found among gangs (e.g., Black 1998:75).

Existing Research

Just as the theoretical and empirical explanations of victimization and offending have developed independently (Berg 2012; Esbensen and Huizinga 1991; Jensen and Brownfield 1986; Lauritsen and Laub 2007), so too has the understanding of the relationship between gang membership, violent offending, and violent victimization (Fox 2013; Gibson et al. 2012; Krohn and Thornberry 2008). This disjuncture has done a disservice to criminology in general and gang research in particular for understanding the linkages between these concepts.

Compared to research on gang membership and offending, studies examining the link between gang membership and victimization have been more attentive to concerns of the causal connection between offending and victimization (Hoyt, Ryan, and Cauce 1999; DeLisi et al. 2009; Katz et al. 2011; Spano, Frelich, and Bolland 2008; Taylor et al. 2007, 2008; Turanovic and Pratt 2012). Studies using matching and regression-adjustment methods typically include a measure of offending when studying the relationship between gang membership and victimization, but reveal a mixed bag of results. Further, conceptual ambiguity surrounds offending and gang membership in these studies. In some instances, gang membership is used as a proxy for risky lifestyles and routine activities (e.g., Childs, Cochran, and Gibson 2010; Hoyt, Ryan, and Cauce 1999; see Sampson and Lauritsen 1990:111-12) while others use offending (e.g., Katz et al. 2011; Rufino, Fox, and Kercher 2012; Spano, Frelich, and Bolland 2008; Taylor et al. 2007, 2008). What is unclear in this line of research is whether gang membership renders spurious the bidirectional association between components of the overlap or whether membership alters the likelihood of violence by increasing the probability of victimization and offending (see Berg 2012). Our theoretical model aligns with the latter, where we emphasize direct and indirect pathways from gang membership to victimization, offending, and their overlap.

The reciprocal relationship between victimization and offending is generally not addressed in gang research. This is problematic because victimization elevates the risk for violent offending, and offending to violent victimization, beyond risk heterogeneity and other common sources that should account for the overlap (Berg et al. 2012; Haynie and Piquero 2006; Schreck 1999; Schreck, Stewart, and Fisher 2006; Singer 1981, 1986; Turanovic and Pratt 2013). Ousey, Wilcox, and Fisher (2011) provide what is perhaps the best account to date of gang membership and the intersection of violent offending and victimization. Using four annual waves of data from students in Kentucky and a fixed effects modeling strategy to eliminate static population heterogeneity, Ousey, Wilcox, and Fisher (2011) found that gang membership was one of a group of time-varying factors that reduced the reciprocal effect sizes of offending and victimization by roughly 40 percent. Importantly, gang membership was positively related to offending and victimization—for both males and females—net of time-varying and time-stable factors. Ousey and colleagues recognized that the generalizability of their findings from mostly rural Kentucky might be limited. Further, their objectives were to address competing arguments between dynamic causal and population heterogeneity models; gang membership was simply a control variable and as a consequence its implications were not elucidated. Despite these caveats, the salience of gang membership for the victim–offender overlap was identified in these findings, but further clarification is necessary.

The Current Study

While the victim–offender overlap ranks among the most consistent facts in criminology, it has yet to be both theoretically and empirically extended to gang membership. We have presented a theoretical foundation rooted in a group process perspective that is consistent with a state-dependent framework (Berg 2012; Lauritsen and Laub 2007) because gang membership alters both the probability of violent offending and victimization, leading to our first hypothesis:

Hypothesis 1: The overlap in violent victimization and offending should be greater among gang members than nongang members, despite controls for alternative theoretical explanations;

We derive three additional hypotheses that are sensitive to the causal connection between offending and victimization:

Hypothesis 2: Gang membership will be associated with violent offending and victimization after controlling for reciprocal relationships in the outcomes;

Hypothesis 3a: Violent offending will only partially mediate the association of gang membership with violent victimization;

Hypothesis 3b: Violent victimization will only partially mediate the association of gang membership with violent offending.

Gang membership should have both direct and indirect influences on offending and victimization that our theoretical model holds is at least partially endogenous to the group processes associated with gang membership. The unfolding of these processes is not protracted; owing to its cyclical nature, gang violence is abrupt and intense and the victim–offender overlap in the gang context should be observable over brief periods.

Method

Data

This study uses data gathered from 621 respondents containing information about their experiences with gangs, violent offending, and victimization. The data were obtained from interviews conducted with youth and young adults in five U.S. cities: Cleveland, OH; Fresno, CA; Los Angeles, CA; Phoenix, AZ; and St. Louis, MO. Respondents were interviewed in settings chosen to include a large number of individuals with involvement in gangs and criminal behavior. All sites included individuals who were actively involved in gangs, as well as individuals who are not involved in gangs. In Cleveland, Los Angeles, and Phoenix, populations of individuals at high risk for criminal involvement were interviewed. Respondents in Cleveland were the clients of street outreach workers; individuals at risk for involvement in crime whose behavior was being monitored by a noncriminal justice agency. The respondents in Los Angeles and Phoenix were participants in street outreach programs that worked with gang members and former offenders seeking to change their lives. Many clients had extensive involvement in the criminal justice system. In Fresno, interviews were conducted with a jail population. In St. Louis, interviews were conducted with individuals on probation or parole. Interviews were conducted in these cities in 2011 over the course of several months. Trained members of the project staff administered surveys face-to-face in private locations at each research site (e.g., classrooms, attorney rooms).¹

Prior research demonstrates (e.g., Papachristos, Braga, and Hureau 2012; Watkins, Huebner, and Decker 2008) that data from field-based sampling strategies have many strengths for investigating research questions such as gang membership and the victim–offender overlap. First, unlike general population samples with low prevalence rates of gang involvement and offending, these data include a large number of gang members, offenders, and victims to ensure variation on our key measures. Second, because respondents were drawn from high-risk environments, this aids in holding constant various unobserved sources of between-person differences (i.e., risk heterogeneity). This is an important point (Berg 2012; Ousey, Wilcox, and Fisher 2011) and strengthens the internal validity of the findings in the current study. Finally, as we describe below, a series of steps were undertaken with the construction of the survey instrument to establish temporal ordering of our independent and dependent variables. Given the reciprocal nature of violent events associated with gangs, a cross-sectional research design is capable of capturing the victim–offender overlap, especially when coupled with a high-risk, field-based sample; an important trade-off from large, national data sources.

Dependent Variables

Our dependent variables are measures of violence that include simple assault, aggravated assault, and robbery. The wording of these forms of violence was as follows: “hit someone with your fists” or “hit by someone with fists” (simple assault), “shot at or attacked someone with a weapon” or “been attacked by someone with a weapon” (aggravated assault), and “used a weapon or force to get money or things from people” or “had a weapon or force used against you to get money or things” (robbery). Respondents self-reported whether they were involved with or experienced each of these three forms of violence in the six months prior to the interview. This ensures that violence is proximal to time-varying factors, as we detail later. All items were dummy coded, with those responding “yes” and “no” coded 1 and 0, respectively.²

We model the violent victim–offender overlap using two distinct measurement approaches, categories of violence and propensity for violence. Our first approach uses a categorical measure to decompose violence into four categories, including respondents who were (1) nonvictim, nonoffender, (2) victim only, (3) offender only, and (4) victim and offender. A categorical measure affords us the ability to assess the probability of assignment conditional on a vector of covariates (e.g., Muftic and Hunt 2013). Gang research illustrates discrepant victimization findings when shifting from prevalence to variety or frequency measures of violent outcomes (Ozer and Engel 2012; Taylor et al. 2007, 2008) because gang members experience multiple events or types. Our second approach, therefore, employs a measure of violence using IRT modeling for the respective offending and victimization outcomes. This allows us to capture latent or underlying propensity for violence (for excellent discussions, see Osgood and Schreck 2007; Sweeten 2012). As a result, we examine three outcomes: categories of violence, propensity for violent offending, and propensity for violent victimization.

Independent Variables

Gang Membership

The key independent variable in this study is gang membership. Individuals who are not gang members will not be exposed and susceptible to group processes in the same manner as active members. Self-report methodology was used to determine gang membership, an approach that is “quite robust and results in classification of a group of youth who are substantively different” (Esbensen and Carson 2012:471; Esbensen et al. 2001; Thornberry et al. 2003; Webb, Katz, and Decker 2006). Respondents were asked “Are you currently in a gang?” Those responding “yes” and “no” were recorded as 1 and 0, respectively.

To establish the empirical relationship between gang membership and the dependent variables, it is important that violent offending and victimization occurred during periods of gang membership. To ensure that gang membership overlapped with the time period of our dependent variables (i.e., six months), a follow-up question was asked about the onset of gang membership (year, month) and evaluated in relation to the interview date. With one exception, all current gang members reported joining their gang at least six months before the interview, which means that any recorded violence occurred during active periods of gang membership.³

Theoretical Constructs

Measures of several individual-level criminological theories relevant to the victim–offender overlap were included in this study, including self-control, code of the street, and routine activity. By including measures derived from these theories, our analysis contributes to the nascent body of research on competing explanations of the overlap. To the best of our knowledge, no study has assessed these three theories simultaneously.

Low self-control is examined to tap individual differences in risk heterogeneity (Schreck 1999; Schreck, Stewart, and Fisher 2006). The construct is comprised of four items from the Brief Self-Control scale (Tangney, Baumeister, and Boone 2004), which has been used effectively in research on the overlap (Holtfreter et al. 2010). Respondents were asked to indicate how well statements such as “I do certain things that are bad for me, if they are fun” described them on a scale of 0 (not at all) to 4 (very much). Higher scores correspond with poorer levels of self-control. The items were averaged and the construct exhibited acceptable levels of internal consistency (Cronbach’s α = .71; average interitem r = .37).

Code of the street is included as part of contemporary examinations of subcultural theory (Anderson 1999), and consists of six items drawn from Stewart and Simons (2010). Respondents were asked to report how well they agreed with statements such as “If someone uses violence against you, it is important that you use violence against him or her to get even” on a scale from 0 (strongly disagree) to 3 (strongly agree). Higher scores reflect stronger adherence to the street code. The items were averaged and the psychometric properties of the construct exceed conventional standards (Cronbach’s α = .81; average interitem r = .42).

Routine activities draws from Osgood and colleagues’ (1996; Osgood and Anderson 2004) individual-level emphasis on time use patterns and unstructured socializing. Two items were combined to capture routine activities by asking respondents how many hours they spent (outside of work or school) hanging out with their friends in (1) public places like malls, parks, corners, or stores and (2) private place like houses, apartment complexes, or backyards (interitem r = .50). Five possible responses categories ranged from 0 (zero hours) to 4 (20 or more hours). Responses were mean adjusted (e.g., 1–5 hours = 3 hours) to approximate a meaningful distribution, then summed for public and private routine activities, resulting in a range of 0 to 40 hours hanging out with friends.

Violent offending and violent victimization are examined in the reciprocal portion of the analysis. Consistent with the theoretical model presented previously, offending is expected to have a causal connection with victimization, and vice versa, to demonstrate an indirect relationship between gang membership and the overlap. A variety score is used for both variables, which is the sum of item types (ranging from 0 to 3), because it closely approximates a latent IRT measure (r > .93; see Sweeten 2012).

Control Variables

Several control variables are also included in our analysis, including age (in years), male (=1, “female” = 0), first generation (=1, “foreign born” = 0) and second generation (=1, “parents foreign born” = 0), race Black (=1, “other/White” = 0), ethnicity Hispanic (=1, “other/White” = 0), and parental education (in years). The typical respondent is in their twenties (M = 27), male (84 percent), Black or Hispanic (41 percent and 47 percent), and their parents were born in the United States (63 percent; see Table 1). Given our purposive sampling strategy, these are characteristics we would expect to find of respondents drawn from inner cities with involvement in or at high risk for involvement in the criminal justice system.

Table 1.

Descriptive Statistics by Gang Membership Status for the Study Variables.

	Full Sample	Gang Member	Nongang
N/%	621/100%	185/30%	436/70%	Significant Difference
	M or % (SD)	M or % (SD)	M or % (SD)	Significant Difference
Categories of violence
Nonoffender, nonvictim	59	46	66	*
Victim only	9	9	8	ns
Offender only	12	14	11	ns
Overlap: offender and victim	20	31	15	*
Violence propensity
Victimization variety	0.42 (0.76)	0.63 (0.90)	0.33 (0.68)	*
Offending variety	0.43 (0.74)	0.71 (0.95)	0.31 (0.58)	*
Demographic control variables
Age	26.70 (9.20)	25.95 (8.37)	27.10 (9.60)	ns
Male	84	83	84	ns
Black	41	38	42	ns
Hispanic	47	56	43	*
First generation	8	8	8	ns
Second generation	29	35	27	*
Parental education	10.48 (3.26)	10.48 (3.37)	10.47 (3.21)	ns
Theoretical constructs
Low self-control (0–4)	1.78 (1.00)	2.06 (0.99)	1.65 (0.99)	*
Code of the street (0–3)	1.52 (0.71)	1.79 (0.67)	1.41 (0.70)	*
Routine activities (0–40)	14.80 (11.9)	17.01 (11.8)	13.91 (11.8)	*

Note: N = 621. SD = standard deviation.

Statistical differences between gang and nongang members were determined using independent sample t-tests and chi-square tests.

*p < .05, ns p > .05.

Analytic Strategy

We begin our analysis by first reporting the descriptive statistics of the study variables. Next, we compare differences in the outcome and explanatory variables for gang-involved and nongang respondents. This second step provides an understanding of differences in involvement and experiences with violence and whether our theoretical model has merit. To test our first hypothesis, multinomial logistic regression is used to relate gang membership and theoretical constructs to our categorical measure of the violent victim–offender overlap (Long 1997). Here, victims and offenders (i.e., the overlap) serve as the reference category, allowing us to compare the remaining three categories of violence to our key outcome of interest. This approach extends our descriptive understanding of differences in the overlap of gang membership by statistically accounting for extraneous demographic and theoretical influences.

The final stage of our analysis uses multilevel IRT modeling to relate gang membership and other explanatory covariates to the propensity for violent offending and victimization (Osgood, McMorris, and Potenza 2002; Raudenbush, Johnson, and Sampson 2003). We begin by examining whether gang membership is related to each respective outcome. To test our hypotheses, violent offending is added to the violent victimization model, and vice versa (e.g., Berg et al. 2012). This approach yields insight into the contribution of gang membership to offending and victimization while accounting for the connection between the respective outcome (i.e., the direct and indirect hypotheses), and the extent that self-control, code of the street, and routine activities explain the link between violent offending and victimization.

All analyses were carried out in Stata 12.0 (StataCorp, College Station, TX). The multinomial logistic regression models were conducted using the mlogit program. The multilevel IRT models were conducted using the gllamm (generalized linear latent and mixed models) suite and a two-level model was estimated, where offending and victimization items (level 1), respectively, are nested within people (level 2). Results from the level-2 structural models are reported. All analyses adjusted for clustering by site and we report robust standard errors. Collinearity diagnostics revealed variance inflation factor scores and condition indices below established thresholds. All theoretical constructs were standardized in the multivariate models. This four-part strategy expands knowledge on the relationship between gang membership, victimization, and offending.

Results

Descriptive and Bivariate Statistics

Table 1 displays the descriptive statistics for the study variables. Respondents reported a high degree of involvement in violence. Indeed, in the six months prior to the interview, 41 percent of the sample either was a victim (9 percent), an offender (12 percent), or both (20 percent). Despite such high rates of violence, the majority of respondents were not victims or offenders. Among those involved in violence, half were either victims or offenders while the other half comprised the overlap. The violence variety scores illustrate variability across respondents, where a modest portion reported two or more forms of violent offending (8 percent) and victimization (11 percent) during the study period.

Partitioning the respondents by gang membership status reveals initial support for our hypotheses. Gang members are twice as likely to fall into the overlapping category of victims of violence and violent offenders. An important distinction is that unlike nongang respondents (34 percent), involvement in violence is the modal category for gang members (54 percent). Further, gang members have violence variety scores that are about twice as great as the respondents who are not in gangs. These findings provide a benchmark for understanding differential participation in violence by gang membership, and are consistent with prior research (Melde and Esbensen 2013; Pyrooz and Decker 2013). Table 1 also reveals that, despite the purposive sampling strategy that absorbs a great deal of unobserved heterogeneity, other important differences that are central to the theoretical underpinnings of the overlap can be observed. Respondents in gangs have lower levels of self-control, stronger adherence to street codes, and riskier routine activities than nongang respondents. As these differences could render spurious the bivariate observations, we move to the multivariate analysis.

Multinomial Logit Models Predicting Violent Overlap Categories

Table 2 reports the results of the multivariate multinomial regression model, where the victim–offender overlap category serves as the base outcome. Here, we test our first hypothesis that existing theoretical perspectives will not fully account for the association of gang membership with the victim–offender overlap. The results are consistent with this hypothesis: Gang membership elevates the probability of the overlap between violent offending and victimization. The relative log odds of being in the violence overlap category, as opposed to the remaining violence categories, was statistically greater if respondents were gang involved. Converting the log odds to an odds ratio indicates that gang members are 48 percent less likely to be victims only [100 × (1 − e ^(−0.65))], 39 percent less likely to be offenders only [100 × (1 − e ^(−0.49))], and 64 percent less likely to be nonvictims/nonoffenders [100 × (1 − e ^(−1.02))] than they are victims and offenders when compared to nongang respondents. The observed results for gang membership are consistently statistically significant due to differences in the victim-offender overlap: there is a full 12 percentage point adjusted marginal difference between gang and nongang members. A supplementary logistic regression analysis (1 = overlap, 0 = all other categories) revealed an odds ratio of 2.36 for gang members (b = 0.86, SE = 0.21).

Table 2.

Multinomial Logistic Regression Model of the Violent Victim–Offender Overlap.

	Victim Only		Offender Only		Nonvictim, Nonoffender
	Base Outcome: Victim and Offender (Overlap)
	Coef.	(SE)	Coef.	(SE)	Coef.	(SE)
Gang membership	−0.65	(.23)*	−0.49	(.19)*	−1.02	(.30)*
Age^a	0.77	(.13)*	0.41	(.20)*	1.07	(.26)*
Male	1.21	(.51)*	0.07	(.32)	0.17	(.24)
Black	−0.44	(.23)#	−0.22	(.50)	0.04	(.57)
Hispanic	−0.59	(.46)*	0.06	(.56)	0.16	(.50)
First generation	0.91	(.69)	0.86	(.44)#	1.15	(.56)*
Second generation	0.04	(.51)	0.04	(.49)	0.04	(.58)
Parental education	0.08	(.04)*	0.18	(.06)*	0.16	(.04)*
Low self-control^a	−0.12	(.23)	0.22	(.23)	−0.06	(.18)
Code of the street^a	0.07	(.10)	−0.39	(.18)*	−0.40	(.21)#
Routine activities^a	−0.36	(.16)*	−0.34	(.24)	−0.50	(.08)*
Constant	−1.69	(.66)*	−1.92	(.73)*	−0.29	(.57)

Note: N = 621; Coef. = coefficient. Standard errors (SEs) are robust and adjust for clustering by site.

^aStandardized.

*p < .05, ^# p < .10; McFadden’s pseudo R ² = .14.

Among the theoretical constructs, we find little support for low self-control theory to distinguish across any of the categories. Code of the street and routine activity theories find marginal statistical support, where a standard deviation increase in these constructs decreased the relative log odds of being a victim only (routine activities), an offender only (code of the street), and a nonparticipant (both theories) compared to being a victim and an offender.⁴ The multinomial logistic regression results, however, are shielded from reciprocal or bidirectional causal connections in violence. We next turn to a modeling strategy that can better approximate the contribution of gang membership to these outcomes when accounting for the connection between offending and victimization (Berg 2012; Lauritsen, Sampson, and Laub 1991).

Multilevel Logistic IRT Models of Propensity for Violent Victimization and Violent Offending

We examine the remaining hypotheses—direct and indirect—on the relationship between gang membership, violent victimization, and violent offending. If the theory guiding these hypotheses is correct, gang membership should (a) have a direct relationship with violent outcomes, (b) be only partially confounded by low self-control, street codes, and routine activities, and (c) operate indirectly through violence outcomes, resulting in partial mediation.⁵ The latter expectation is central to our theoretical model, which contends that an empirical relationship between gang membership and the violent outcomes will persist because of group processes assumed with involvement in gangs. The answers to these hypotheses are found in the multivariate two-level logistic IRT models.

Table 3 displays the results where violent victimization is regressed on gang membership, violent offending, and the remaining covariates in three stages. Model 1 indicates that gang membership increases the log odds of violent victimization by 0.91 (p < .05), a 27 percent reduction from a gang-only model (not reported: b = 1.25, p < .05). Model 2 reveals a strong relationship between violent offending and violent victimization, where a standardized change of 0.58 is observed. Nearly two-thirds of the variance is explained in this model, based on the initial variance from an unconditional model that only adjusts for site and item seriousness. Model 3 tests whether membership contributes to violent victimization net of involvement in offending. The direct hypothesis is confirmed: A log odds increase of 0.27 (p < .05) in violent victimization is associated with gang membership. Violent offending explains away an additional 70 percent of the relationship between gang membership and violent victimization. This is consistent with prior empirical work that finds much of the effect of gang membership on victimization operates through offending (e.g., Katz et al. 2011). Violent offending, however, does not account entirely for the association between gang membership and violent victimization, which is consistent with the group process framework.

Table 3.

Logistic Item Response Theory (IRT) Models of Propensity for Violent Victimization.

	Violent Victimization
	Model 1		Model 2		Model 3
	Coef.	(SE)	Coef.	(SE)	Coef.	(SE)
Gang membership	0.91	(.23)*			0.27	(.10)*
Violent offending^a			0.58	(.05)*	0.57	(.05)*
Age^a	−0.72	(.28)*	−0.27	(.16)#	−0.29	(.18)
Male	0.11	(.13)	0.13	(.22)	0.13	(.24)
Black	−0.53	(.57)	−0.55	(.42)	−0.59	(.43)
Hispanic	−0.66	(.52)	−0.62	(.45)	−0.68	(.43)
First generation	−1.20	(.60)*	−0.90	(.47)#	−0.91	(.46)*
Second generation	−0.25	(.52)	−0.24	(.26)	−0.27	(.26)
Parental education^a	−0.59	(.09)*	−0.42	(.10)*	−0.43	(.09)*
Low self-control^a	−0.08	(.18)	−0.17	(.17)	−0.17	(.17)
Code of the street^a	0.43	(.15)*	0.39	(.10)*	0.36	(.11)*
Routine activities^a	0.54	(.12)*	0.31	(.12)*	0.31	(.12)*
Constant	−1.89		−1.56		−1.59
Log likelihood	−604		−552		−550
Variance explained	37%		63%		65%

Note: N = 621; Coef. = coefficient. Standard errors (SEs) are robust and adjust for clustering by site.

^aStandardized.

*p < .05, ^# p < .10.

Table 4 presents the findings where violent offending is regressed on gang membership, violent victimization, and the remaining covariates across three stages. Models 1 and 2 indicate that gang membership and violent victimization correspond with independent changes of 1.51 and 0.91, respectively, in the log odds of the propensity to offend violently. Violent victimization, however, explains a much greater portion of the variance than gang membership. The theoretical constructs only account for about 20 percent of the relationship between gang membership and violent offending. Turning to model 3, the results indicate that gang membership and violent victimization have an additive relationship with violent offending. The attenuation of the gang membership coefficient for violent offending was more modest than in the violent victimization model (30 percent vs. 70 percent), but gang membership results in a 1.03 (p < .05) increase in the log odds of violent offending.⁶

Table 4.

Logistic Item Response Theory (IRT) Models of Propensity for Violent Offending.

	Violent Offending
	Model 1		Model 2		Model 3
	Coef.	(SE)	Coef.	(SE)	Coef.	(SE)
Gang membership	1.51	(.31)*			1.03	(.15)*
Violent victimization^a			0.91	(.10)*	0.85	(.10)*
Age^a	−1.24	(.23)*	−0.83	(.14)*	−0.85	(.14)*
Male	−0.20	(.58)	−0.13	(.67)	−0.12	(.61)
Black	−0.02	(.43)	0.32	(.24)	0.15	(.25)
Hispanic	−0.04	(.29)	0.47	(.27)#	0.22	(.22)
First generation	−0.71	(.86)	0.01	(.81)	−0.04	(.77)
Second generation	0.00	(.65)	0.16	(.40)	0.08	(.39)
Parental education^a	−0.35	(.03)*	0.04	(.02)*	−0.02	(.03)
Low self-control^a	0.17	(.20)	0.29	(.20)	0.25	(.18)
Code of the street^a	0.20	(.23)	0.08	(.19)	−0.03	(.21)
Routine activities^a	0.62	(.14)*	0.28	(.18)	0.28	(.17)
Constant	5.17		−5.25		−5.27
Log likelihood	−534		−487		−480
Variance explained	37%		56%		61%

Note: N = 621; Coef. = coefficient. Standard errors (SEs) are robust and adjust for clustering by site.

^aStandardized.

*p < .05, ^# p < .10.

When the results of Tables 3 and 4 are considered together, these findings confirm the final three hypotheses of the study. The direct association of gang membership with violence outcomes is not entirely confounded or mediated by any alternative explanation. Further, the relationship between gang membership and the victim–offender overlap operates indirectly through violent offending and violent victimization. To be sure, the group processes associated with gangs are expected to increase involvement in violent offending and risks for violent victimization. Yet, the mechanisms that link violence outcomes—offending to victimization, victimization to offending—do not fully mediate the effect of membership in a gang, nor do other theoretical constructs such as low self-control, street codes, or routine activities, which leads to our discussion of the implications of these findings.

Discussion

The strong link between offending and victimization has been the subject of intense theoretical and empirical scrutiny for nearly 70 years. The prominent explanations for this relationship include the constitutional characteristics of individuals, how and where people spend their time, diversity of cultural norms and values, and the neighborhoods and contexts where individuals reside. Mid-level dynamics, in particular gang participation and group processes, have largely been ignored in examinations of the victim–offender overlap despite the empirical standing of theories emphasizing socialization and group influence (Pratt et al. 2010; Warr 2002). To that end, the current study extended a theoretical model that emphasized group processes to understand how gang membership contributes to the victim–offender overlap vis-à-vis existing theoretical paradigms, including self-control, routine activity, and code of the street. We use Figure 1 to illustrate three main points from our findings that merit further discussion.

Figure 1.

The linkages between gang membership and violence. Note: Solid lines represent direct pathways; dashed lines represent indirect pathways.

First, the results provide empirical support for extending knowledge of the victim–offender overlap to the context of gangs. To date, the interrelationship between offending, victimization, and gang membership has been neither well understood nor researched. This observation is curious, as two decades of studies have revealed consistently that gang membership is linked to higher levels of offending and victimization (Deckeret al. 2013; Fox 2013; Krohn and Thornberry 2008). The multinomial logistic regression results indicate that respondents who reported membership in a gang were more likely than nongang members to engage in violent offending and experience violent victimization than they were to be only victims or offenders or nonvictims and nonoffenders. In other words, gang members are not distinctly offenders or victims; instead, gang membership is a common source of both forms of violence, as reflected in path A in Figure 1.

Nothing is inherently criminal about gang membership, meaning path A is merely an indirect path to violence. But the past two decades of individual-level research on gangs rarely finds that gang membership is statistically and substantively unrelated to offending or victimization, especially when concentrating on violence. What makes gang membership compelling in this research is that controlling for numerous mediators and confounders—observed and unobserved sources of selection bias—does not eliminate this relationship. That is, path A should be mediated, but it is not. Some have unpacked this “black box” by showing how movement into and out of gangs corresponds with changes in social bonds, unstructured socializing, street code adherence, peer networks, temperance, and victimization (Matsuda et al. 2013; Melde and Esbensen 2011; Sweeten, Pyrooz, and Piquero 2013). But even then, these studies find the relationship between gang membership and offending is only partially mediated, which is why theoretically and empirically integrating this literature within the criminological reality of the victim–offender overlap is particularly timely, leading to our next point.

Second, the mechanisms linking gang membership to the victim–offender overlap are only partially found in the current inventory of theories used to explain the victim–offender overlap. Our results showed that controlling for constructs—those operating in path A—derived from risk heterogeneity, routine activity, and subcultural theories did not account fully for the relationship between gang membership and victimization or offending. When we turned to bidirectional models as represented in path B, using offending to explain variation in violent victimization (i.e., consistent with risky lifestyles) and victimization to explain variation in violent offending (i.e., consistent with legal cynicism, general strain),⁷ gang membership remained a stable, albeit more modest, correlate of the outcomes. Our contention was that contemporary theories of the victim-offender overlap are limited when studying violence within the context of gangs and gang members. Instead, we held that the remaining connection between gang membership and violence and violence is found in path C—group processes.

The mid-level dynamics of group process suggest that gangs are greater than the sum of their parts. Gang membership widens the pool of victims and offenders through collective identities, normative orientations, status acquisition and maintenance, extra-individual liabilities, and reciprocity in conflict. One former Phoenix gang member recalled the following incident to us:

We were kind of coming down the street and [he] turned around and he’s like, “You guys there’s a car following us.” All of us like reached under the seat to grab our stuff … you know our stuff, like our guns. We reach under and we’re like loading it, but by the time that we could like pull back and load it, uh, shots fired in the window, in the back window. And like the window just shattered like that so, um … I mean she fell over you know, and we all like kinda ducked and we pull into a Circle K and um … And then I looked over at [her] and she’s still laying there and I’m like, “[Victim name] get up. Get up.” … we all rushed to the hospital. We ended up calling [Person name], to tell him, and he got the rest of the gang together and went to go shoot the other guy. He went … they did go shoot the other guy. They went and killed the other guy, but they ended up shooting his brother too, and his brother’s paralyzed now.

As this young woman described, gang violence is neither isolated nor static but connected and dynamic. Papachristos (2009; Papachristos, Hureau, and Braga 2013) detailed the institutionalized conflict networks of gangs in Boston and Chicago that create an “organizational memory” in the gang landscape to prompt future violence. Therefore, to fully understand the empirical relationship between gang membership and the victim–offender overlap, it is necessary to account for the group processes and dynamics in the gang. We anticipate that these processes are not necessarily limited to street gangs and that there is much to be gained by examining this relationship in a social network framework. The resources, characteristics, contexts, and composition of the social networks that youth and young adults are embedded in should offer important insights into victim and offender roles (McGloin and Kirk 2010; Papachristos 2010; Pyrooz, Sweeten, and Piquero 2013; Young and Rees 2013).

Third, the finding that gang membership elevates risk for the overlap in offending and victimization illustrates one of the main challenges associated with responding to gang violence. Figure 1 shows not only how group processes lead to violent victimization and offending but also how incidents of violence enhance group processes (Klein 1971; Papachristos, Hureau, and Braga 2013). As the quote above illustrates, violence begets violence. Our results support interventions that target violent offenders and victims and add to a literature demonstrating that individuals move easily between these roles (Berg 2012; Jennings, Piquero, and Reingle 2012). The collective liabilities of gangs may also be used by law enforcement as a means of inhibiting violent behavior. Indeed, “lever-pulling” strategies that draw law enforcement attention to all members of a gang for only a single member’s behavior have been shown to be somewhat effective in reducing violence (e.g., Kennedy, Piehl, and Braga 1996; Papachristos 2011). Such interventions draw on the collective liability of the gang.

Criminologists understandably focus discussions of policy and crime control on offending. Equally important to consider are strategies that reduce victimization. Indeed, because of the close nexus between victimization and offending, criminal justice policy may be ripe for victimization reduction strategies. The results from the Ceasefire Chicago evaluation (Skogan et al. 2009), specifically the role of “violence interrupters,” offer a promising strategy to interrupt high rates of victimization among gang members by intervening to prevent retaliation. Reducing victimization in the context of gangs may pay additional dividends because it diminishes the motivations for retaliation and may thereby also reduce group solidarity, a process often enhanced by violent gang activity. Recent research by Whitehill and colleagues (2013) provides insights in the mediation practices of these individuals, but not the efficacy of these practices. Having interrupters who focus on victims—particularly gang-involved victims—should become a high priority given the causal connection between victimization and offending. Further, conflict interventions should be sufficiently broad as to include as many gang members as possible. Such activities should be carefully planned so as not to enhance group solidarity and identity by focusing attention on the gang as a group. That is to say, such interventions must be cognizant of the larger group processes that operate in the immediate gang networks and how they contribute to emerging and long-standing gang conflicts. Communicating a message to other gang members about refraining from retaliation may also be appropriate (Tita, Riley, and Greenwood 2003; Kennedy, Piehl, and Braga 1996). A more victim-centered form of lever pulling is described by Kennedy (2011). Working group members visited gunshot wound victims, their families, and members of the victim’s gang, in the hospital. While granting sympathy to the victim, fellow gang members were warned that retaliation would be met with a strong response from law enforcement.

The cross-sectional nature of our data naturally introduces some constraints to the interpretation of our findings. However, the measures of our key variables ensure that our outcomes occurred during periods of gang membership. There is no doubt that a nexus exists between victimization and offending, yet it is not easy—and potentially not useful—to determine which came first, particularly in the arena of gang violence where a group process model anticipates cyclical, rapid incidents of violence. Even longitudinal data provide challenges to establishing time order to the extent that waves are spaced over long intervals (e.g., one year). Examining gang membership and the overlap longitudinally requires shorter time periods between waves and life history calendars to elaborate on this relationship (e.g., Griffin and Armstrong 2003; Horney, Osgood, and Marshall 1995; McGloin et al. 2007), as well as detailed inquiry into the events that precipitated violence, particularly around the points of transitioning into and out of gangs.

Results from the current study provide a foundation for future research. While this study did not directly observe gang-related group processes, the linkages in Figure 1 can be used to investigate the features of gangs and other criminal networks as well as to provide important insight into the victim–offender overlap. Further, as recent research has demonstrated, neighborhood context also structures this overlap (Berg and Loeber 2011; Berg et al. 2012). We encourage future research to explore whether contextual measures will “chip away” at the influences of gang membership on violence, but we do not anticipate that neighborhood factors would alter the results of the current study given that respondents were drawn from similar risky contexts. Instead, we view measures of group processes as key directions forward to understand the overlap in violence, but we encourage expanding the inventory of offending and victimization types beyond violence. Such work will need to be conducted with targeted samples of offenders and those at high risk for involvement in crime as victims and offenders. Recent research has applied a variety of theoretical perspectives to the overlap across diverse research contexts (Berg 2012; Jennings, Piquero, and Reingle 2012; Lauritsen and Laub 2007; Menard 2012). This study provided a theoretical and empirical extension of the victim–offender overlap to gang membership, broadening the theoretical assets available to understand this relationship and expanding the generality of the overlap.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research and/or authorship of this article: This project was supported in part with funding from Google Ideas. We are grateful for their support. The content of this article, however, is solely the responsibility of the authors and does not necessarily represent the official views of Google.

Notes

References

Anderson

Elijah

. 1999. Code of the Street: Decency, Violence, and the Moral Life of the Inner City. New York: W. W. Norton.

Ayling

Julie

. 2011. “Gang Change and Evolutionary Theory.” Crime, Law, and Social Change 56:1–26.

Battin

Sara R.

Hill

Karl G.

Abbott

Robert D.

Catalano

Richard F.

David Hawkins

. 1998. “The Contribution of Gang Membership to Delinquency Beyond Delinquent Friends.” Criminology 36:93–116.

Berg

Mark T.

2012. “The Overlap of Violent Offending and Violent Victimization: Assessing the Evidence and Explanations.” Pp. 17–38 in Violent Offenders: Research, Theory, and Practice, edited by DeLisi

Matt

. Burlington, MA: Jones and Bartlett.

Berg

Mark T.

Loeber

Rolf

. 2011. “Examining the Neighborhood Context of the Violent Offending-Victimization Relationship: A Prospective Investigation.” Journal of Quantitative Criminology 27:427–51.

Berg

Mark T.

Stewart

Eric A.

Schreck

Christopher J.

Simons

Ronald L.

. 2012. “The Victim-Offender Overlap in Context: Examining the Role of Neighborhood Street Culture.” Criminology 50:359–90.

Black

Donald

. 1983. “Crime as Social Control.” American Sociological Review 48:34–45.

Black

Donald

. 1998. The Social Structure of Right and Wrong. San Diego, CA: Academic Press.

Brantingham

P. Jeffrey

Tita

George E.

Short

Martin B.

Reid

Shannon E.

. 2012. “The Ecology of Gang Territorial Boundaries.” Criminology 50:851–85.

10.

Brunson

Rod K.

Miller

Jodi

. 2009. “Schools, Neighborhoods, and Adolescent Conflicts: A Situational Examination of Reciprocal Dynamics.” Justice Quarterly 26:183–210.

11.

Childs

Kristina K.

Cochran

John K.

Gibson

Chris L.

. 2010. “Self-control, Gang Membership, and Victimization: An Integrated Approach.” Journal of Crime and Justice 32:35–54.

12.

Cohen

Lawrence E.

Kluegel

James R.

Land

Kenneth C.

1981. “Social Inequality and Predatory Criminal Victimization: An Exposition and Test of a Formal Theory.” American Sociological Review 46:505–24.

13.

Decker

Scott H.

1996. “Collective and Normative Features of Gang Violence.” Justice Quarterly 13:243–64.

14.

Decker

Scott H.

David Curry

. 2002. “Gangs, Gang Homicides, and Gang Loyalty: Organized Crime or Disorganized Criminals.” Journal of Criminal Justice 30:343–52.

15.

Decker

Scott H.

Melde

Chris

Pyrooz

David C.

. 2013. “What Do We Know about Gangs and Gang Members and Where Do We Go from Here?” Justice Quarterly 30:369–402.

16.

Decker

Scott H.

Pyrooz

David C.

. 2010. “Gang Violence Worldwide: Context, Culture, and Country.” Pp. 128–55 in Small Arms Survey 2010: Gangs, Groups, and Guns. NewYork: Cambridge University Press.

17.

Decker

Scott H.

Winkle

Barrick Van

. 1996. Life in the Gang. New York: Cambridge University Press.

18.

DeLisi

Matt

Barnes

J. C.

Beaver

Kevin M.

Gibson

Chris L.

. 2009. “Delinquent Gangs and Adolescent Victimization Revisited: A Propensity Score Matching Approach.” Criminal Justice and Behavior 36:808–23.

19.

Densley

James A.

2012. “Street Gang Recruitment.” Social Problems 59:301–21.

20.

Erlanger

Howard S.

1974. “The Empirical Status of the Sub-cultures of Violence Thesis.” Social Problems 22:280–92.

21.

Esbensen

Finn-Aage

Carson

Dena C.

. 2012. “Who are the Gangsters? An Examination of the Age, Race/Ethnicity, Sex, and Immigration Status of Self-reported Gang Members in a Seven City Study of American Youth.” Journal of Contemporary Criminal Justice 28:465–81.

22.

Esbensen

Finn-Aage

Huizinga

David

. 1991. “Juvenile Victimization and Delinquency.” Youth & Society 23:202–28.

23.

Esbensen

Finn-Aage

Maxson

Cheryl L.

. 2012. “The Intersection of Gang Definition and Group Process: Concluding Observations.” Pp. 303–315 in Youth Gangs in International Perspective, edited by Esbensen

Finn-Aage

Maxson

Cheryl L.

. New York, NY: Springer-Verlag.

24.

Esbensen

Finn-Aage

Peterson

Dana

Taylor

Terrance J.

Freng

Adrienne

. 2010. Youth Violence: Sex and Race Differences in Offending, Victimization, and Gang Membership. Philadelphia, PA: Temple University Press.

25.

Esbensen

Finn-Aage

Thomas Winfree

Jr He

Taylor

Terrance J.

. 2001. “Youth Gangs and Definitional Issues: When is a Gang a Gang, and Why Does It Matter?” Crime and Delinquency 47:105–30.

26.

Fleisher

Mark S.

1998. Dead End Kids: Gang Girls and the Boys They Know. Madison: University of Wisconsin Press.

27.

Fox

Kate

. 2013. “New Developments and Implications for Understanding the Victimization of Gang Members.” Violence and Victims. doi:10.1891/0886-6708.VV-D-12-00100

28.

Garofalo

James

. 1987. “Reassessing the Lifestyle Model of Criminal Victimization.” Pp. 23–42 in Positive Criminology, edited by Gottfredson

Michael R.

Hirschi

Travis

. Thousand Oaks, CA: Sage.

29.

Gibson

Chris L.

Swatt

Matt L.

Mitchell Miller

Jennings

Wesley G.

Gover

Angela R.

. 2012. “The Causal Relationship between Gang Joining and Violent Victimization: A Critical Review and Directions for Future Research.” Journal of Criminal Justice 40:490–501. doi:2012.08.001

30.

Gottfredson

Michael R.

Hirschi

Travis

. 1990. A General Theory of Crime. Stanford, CA: Stanford University Press.

31.

Griffin

Marie L.

Armstrong

Gaylene S.

. 2003. “The Effect of Local Life Circumstances on Female Probationers’ Offending.” Justice Quarterly 20:213–39.

32.

Hagedorn

John M.

1991. “Gangs, Neighborhoods, and Public Policy.” Social Problems 38:529–42.

33.

Haynie

Dana L.

Piquero

Alex

. 2006. “Pubertal Development and Victimization Risk: Differences across Gender.” Journal of Research in Crime and Delinquency 43:3–35.

34.

Hindelang

Michael J.

Gottfredson

Michael R.

Garofalo

James

. 1978. Victims of Personal Crime: An Empirical Foundation for a Theory of Personal Victimization. Cambridge, MA: Ballinger.

35.

Holtfreter

Kristy

Reisig

Michael D.

Piquero

Nicole L.

Piquero

Alex

. 2010. “Low Self-control and Fraud: Offending, Victimization, and Their Overlap.” Criminal Justice and Behavior 37:188–203.

36.

Horney

Julie

Wayne Osgood

Marshall

Ineke H.

. 1995. “Criminal Careers in the Short-term: Intra-individual Variability in Crime and Its Relation to Local Life Circumstances.” American Sociological Review 60:655–73.

37.

Hoyt

Dan R.

Ryan

Kimberly D.

Cauce

Ana Marie

. 1999. “Personal Victimization in a High-risk Environment: Homeless and Runaway Adolescents.” Journal of Research in Crime and Delinquency 36:371–92.

38.

Hughes

Lorine A.

Short

James F.

, Jr. 2005. “Disputes Involving Youth Street Gang Members: Micro-social Contexts.” Criminology 43:43–76.

39.

Jacobs

Bruce A.

Wright

Richard

. 2006. Street Justice: Retaliation in the Criminal Underworld. New York: Cambridge University Press.

40.

Jennings

Wesley G.

Piquero

Alex R.

Reingle

Jennifer M.

. 2012. “On the Overlap between Victimization and Offending: A Review of the Literature.” Aggression and Violent Behavior 17:16–26.

41.

Jensen

Gary F.

Brownfield

David

. 1986. “Gender, Lifestyles, and Victimization: Beyond Routine Activity.” Violence and Victims 1:143–65.

42.

Katz

Jack

. 1988. The Seductions of Crime. New York: Basic Books.

43.

Katz

Charles M.

Webb

Vincent J.

Fox

Kathleen A.

Shaffer

Jennifer N.

. 2011. “Understanding the Relationship between Violent Victimization and Gang Membership.” Journal of Criminal Justice 39:48–59.

44.

Kennedy

David M.

2011. Don’t Shoot: One Man, a Street Fellowship, and the End of Violence in Inner-city America. New York: Bloomsbury Publishing.

45.

Kennedy

David M.

Piehl

Anne M.

Braga

Anthony M.

. 1996. “Youth Violence in Boston: Gun Markets, Serious Youth Offenders, and a Use-reduction Strategy.” Law and Contemporary Problems 59:147–96.

46.

Kissner

Jason

Pyrooz

David C.

. 2009. “Self-control, Differential Association, and Gang Membership: A Theoretical and Empirical Extension of the Literature.” Journal of Criminal Justice 37:478–87.

47.

Klein

Malcolm W.

1971. Street Gangs and Street Workers. Englewood Cliffs, NJ: Prentice Hall.

48.

Klein

Malcolm W.

1995. The American Street Gang: Its Nature, Prevalence, and Control. New York: Oxford University Press.

49.

Klein

Malcolm W.

Maxson

Cheryl L.

. 2006. Street Gang Patterns and Policies. New York: Oxford University Press.

50.

Krohn

Marvin D.

Thornberry

Terence P.

. 2008. “Longitudinal Perspectives on Adolescent Street Gangs.” Pp. 128–60 in The Long View of Crime: A Synthesis of Longitudinal Research, edited by Liberman

Akiva M.

. Washington, DC: National Institute of Justice.

51.

Kubrin

Charis E.

Weitzer

Ronald

. 2003. “Retaliatory Homicide: Concentrated Disadvantage and Neighborhood Culture.” Social Problems 50:157–80.

52.

Lattimore

Pamela K.

Linster

Richard L.

MacDonald

John M.

. 1997. “Risk of Death among Serious Young Offenders.” Journal of Research in Crime and Delinquency 34:187–209.

53.

Lauger

Timothy R.

2012. Real Gangstas: Legitimacy, Reputation, and Violence in an Intergang Environment. New Brunswick, NJ: Rutgers University Press.

54.

Lauritsen

Janet L.

Laub

John H.

. 2007. “Understanding the Link between Victimization and Offending: New Reflections on an Old Idea.” Crime Prevention Studies 22:55–75.

55.

Lauritsen

Janet L.

Sampson

Robert J.

Laub

John H.

. 1991. “The Link between Offending and Victimization among Adolescents.” Criminology 29:265–91.

56.

Levitt

Stephen D.

Venkatesh

Sudhir A.

. 2001. “

Growing up in the Projects: The Economic Lives of a Cohort of Men Who Came of Age in Chicago Public Housing.”

American Economic Review 91:79–84.

57.

Long

J. Scott

. 1997. Regression Models for Categorical and Limited Dependent Variables. Thousand Oaks, CA: Sage.

58.

Matsuda

Kristy N.

Melde

Chris

Taylor

Terrance J.

Freng

Adrienne

Esbensen

Finn-Aage

. 2013. “Gang Membership and Adherence to the ‘Code of the Street.’” Justice Quarterly 30:440–68.

59.

McGloin

Jean M.

Kirk

David

. 2010. “Social Network Analysis.” Pp. 209–24 in Handbook of Quantitative Criminology, edited by Piquero

Alex R.

Weisburd

David

. New York: Springer.

60.

McGloin

Jean M.

Sullivan

Christopher J.

Piquero

Alex R.

Pratt

Travis C.

. 2007. “Local Life Circumstances and Offending Specialization/Versatility: Comparing Opportunity and Propensity Models.” Journal of Research in Crime and Delinquency 44:321–46.

61.

Melde

Chris

Esbensen

Finn-Aage

. 2011. “Gang membership as a turning point in the life course.” Criminology 49:513–552.

62.

Melde

Chris

Esbensen

Finn-Aage

. 2013. “Gangs and Violence: Disentangling the Impact of Gang Membership on the Level and Nature of Offending.” Journal of Quantitative Criminology 29:143–66.

63.

Melde

Chris

Taylor

Terrance J.

Esbensen

Finn-Aage

. 2009. “‘I Got Your Back’: An Examination of the Protective Function of Gang Membership in Adolescence.” Criminology 47:565–94.

64.

Menard

Scott

. 2012. “Age, Criminal Victimization, and Offending: Changing Relationships from Adolescence to Middle Adulthood.” Victims and Offenders 7:227–54.

65.

Miller

Jody

. 2001. One of the Guys: Girls, Gangs, and Gender. New York: Oxford University Press.

66.

Muftic

Lisa R.

Hunt

Donald E.

. 2013. “Victim Precipitation: Further Understanding the Linkage between Victimization and Offending in Homicide.” Homicide Studies 17:239–54.

67.

Osgood

D. Wayne

Anderson

Amy L.

. 2004. “Unstructured Socializing and Rates of Delinquency.” Criminology 42:519–50.

68.

Osgood

D. Wayne

McMorris

Barbara J.

Potenza

Maria T.

. 2002. “Analyzing Multiple-item Measures of Crime and Deviance I: Item Response Theory Scaling.” Journal of Quantitative Criminology 18:267–96.

69.

Osgood

D. Wayne

Schreck

Christopher J.

. 2007. “A New Methods for Studying the Extent, Stability, and Predictors of Individual Specialization in Violence.” Criminology 45:273–312.

70.

Osgood

D. Wayne.

Wilson

Janet K.

O’Malley

Patrick M.

Bachman

Jerald G.

Johnston

Lloyd D.

. 1996. “Routine Activities and Individual Deviant Behavior.” American Sociological Review 61:635–55.

71.

Ousey

Graham C.

Wilcox

Pamela

Fisher

Bonnie S.

. 2011. “Something Old, Something New: Revisiting Competing Hypotheses of the Victimization-Offending Relationship among Adolescents.” Journal of Quantitative Criminology 27:53–84.

72.

Ozer

M. Murat

Engel

Robin S.

. 2012. “Revisiting the Use of propensity Score Matching to Understand the Relationship between Gang Membership and Violent Victimization: A Cautionary Note.” Justice Quarterly 29:105–24.

73.

Papachristos

Andrew V.

2009. “Murder by Structure: Dominance Relations and the Social Structure of Gang Homicide.” American Journal of Sociology 115:74–128.

74.

Papachristos

Andrew V.

2010. “The Coming of a Networked Criminology?” Pp. 101–40 in Measuring Crime and Criminality: Advances in Criminological Theory, edited by MacDonald

John

. New Brunswick, NJ: Transaction.

75.

Papachristos

Andrew V.

2011. “Too Big to Fail.” Criminology & Public Policy 10:1053–61.

76.

Papachristos

Andrew V.

Braga

Anthony A.

Hureau

David M.

. 2012. “Social Networks and the Risk of Gunshot Injury.” Journal of Urban Health 89:992–1003.

77.

Papachristos

Andrew V.

Hureau

David M.

Braga

Anthony A.

. 2013. “The Corner and the Crew: The Influence of Geography and Social Networks on Gang Violence.” American Sociological Review 78:417–47.

78.

Pratt

Travis C.

Cullen

Francis T.

. 2000. “The Empirical Status of Gottfredson and Hirschi’s General Theory of Crime: A Meta-analysis.” Criminology 38:931–64.

79.

Pratt

Travis C.

Cullen

Francis T.

Sellers

Christine S.

Thomas Winfree

Madensen

Tamara

Daigle

Leah

Fearn

Noelle E.

Gau

Jacinta M.

. 2010. “The Empirical Status of Social Learning Theory: A Meta-analysis.” Justice Quarterly 27:765–802.

80.

Pyrooz

David C.

Sweeten

Gary

Piquero

Alex R.

. 2013. “

Continuity and Change in Gang Membership and Gang Embeddedness.”

Journal of Research in Crime and Delinquency 50:239–71.

81.

Pyrooz

David C.

Decker

Scott H.

. 2011. “Motives and Methods for Leaving the Gang: Understanding the Process of Gang Desistance.” Journal of Criminal Justice 39:417–25.

82.

Pyrooz

David C.

Decker

Scott H.

. 2013. “Delinquent behavior, violence, and gang involvement in China.” Journal of Quantitative Criminology 29:251–272.

83.

Raudenbush

Stephen W.

Johnson

Christopher

Sampson

Robert J.

. 2003. “A Multivariate, Multilevel Rasch Model with Application to Self-reported Criminal Behavior.” Sociological Methodology 33:169–211.

84.

Rufino

Katrina A.

Fox

Kathleen A.

Kercher

Glen A.

. 2012. “Gang Membership and Crime Victimization among Prison Inmates.” American Journal of Criminal Justice 37:321–37.

85.

Sampson

Robert J.

Lauritsen

Janet L.

. 1990. “Deviant Lifestyles, Proximity to Crime, and the Offender-Victim Link in Personal Violence.” Journal of Research in Crime and Delinquency 27:110–39.

86.

Sanders

William B.

1994. Gangbangs and Drive-bys: Grounded Culture and Juvenile Gang Violence. New York: Aldine.

87.

Schreck

Christopher J.

1999. “Criminal Victimization and Low Self-control: An Extension and Test of a General Theory of Crime.” Justice Quarterly 16:633–54.

88.

Schreck

Christopher J.

Stewart

Eric A.

Fisher

Bonnie S.

. 2006. “Self-control, Victimization, and Their Influence on Risky Lifestyles: A Longitudinal Analysis Using Panel Data.” Journal of Quantitative Criminology 22:319–40.

89.

Schreck

Christopher J.

Stewart

Eric A.

Wayne Osgood

. 2008. “A Reappraisal of the Overlap of Violent Offenders and Victims.” Criminology 46:871–906.

90.

Short

James F.

1965. “Social Structure and Group Processes in Explanations of Gang Delinquency.” Pp. 155–89 in Problems of Youth: Transition to Adulthood in a Changing World, edited by Sherif

Muzafer

Sherif

Carolyn Wood

. Chicago, IL: Aldine.

91.

Short

James F.

Strodtbeck

Fred L.

. 1965. Group Process and Gang Delinquency. Chicago, IL: University of Chicago Press.

92.

Singer

Simon I.

1981. “Homogeneous Victim-Offender Populations: A Review and Some Research Implications.” Journal of Criminal Law and Criminology 72:779–88.

93.

Singer

Simon I.

1986. “Victims of Serious Violence and Their Criminal Behavior: Subcultural Theory and Beyond.” Violence and Victims 1:61–70.

94.

Skogan

Wesley

Hartnett

Susan

Bump

Natalie

DuBois

Jill

. 2009. Evaluation of Chicago Ceasefire. Accessed October 8, 2013. http://www.skogan.org/files/Evaluation_of_CeaseFire-Chicago_Main_Report.03-2009.pdf.

95.

Spano

Richard

Freilich

Joshua D.

Bolland

John

. 2008. “Gang Membership, Gun Carrying, and Employment: Applying Routine Activities Theory to Explain Violent Victimization among Inner City, Minority Youth Living in Extreme Poverty.” Justice Quarterly 25:381–410.

96.

Sparks

Richard F.

1982. Research on Victims of Crime: Accomplishments, Issues, and New Directions. Washington, DC: Government Printing Office.

97.

Stewart

Eric A.

Schreck

Christopher

Simons

Ronald L.

. 2006. “‘I Ain’t Gonna Let No One Disrespect Me’: Does the Code of the Street Reduce or Increase Violent Victimization among African American Adolescents?” Journal of Research in Crime and Delinquency 43:427–58.

98.

Stewart

Eric A.

Simons

Ronald L.

. 2010. “Race, Code of the Street, and Violent Delinquency: A Multilevel Investigation of Neighborhood Street Culture and Individual Norms of Violence.” Criminology 48:569–605.

99.

Sweeten

Gary

. 2012. “Scaling Criminal Offending.” Journal of Quantitative Criminology 28:533–57.

100.

Sweeten

Gary

Pyrooz

David C.

Piquero

Alex R.

. 2013. “Disengaging from Gangs and Desistance from Crime.” Justice Quarterly 30:469–500.

101.

Tangney

June P.

Baumeister

Roy F.

Boone

Angie Luzio

. 2004. “High Self-control Predicts Good Adjustment, Less Pathology, Better Grades, and Interpersonal Success.” Journal of Personality 72:271–322.

102.

Taylor

Terrance J.

Freng

Adrienne

Esbensen

Finn-Aage

Peterson

Dana

. 2008. “Youth Gang Membership and Serious Violent Victimization: The Importance of Lifestyle/Routine Activities.” Journal of Interpersonal Violence 10:1–24.

103.

Taylor

Terrance J.

Peterson

Dana

Esbensen

Finn-Aage

Freng

Adrienne

. 2007. “Gang Membership as a Risk Factor for Adolescent Violent Victimization.” Journal of Research in Crime and Delinquency 44:351–80.

104.

Thornberry

Terence P.

Krohn

Marvin D.

Lizotte

Alan J.

Smith

Carolyn A.

Tobin

Kimberly

. 2003. Gangs and Delinquency in Developmental Perspective. Cambridge, England: Cambridge University Press.

105.

Thrasher

Frederic M.

1927. The Gang. Chicago: University of Chicago Press.

106.

Tita

George E.

Radil

Steven M.

. 2011. “Spatializing the Social Networks of Gangs to Explore Patterns of Violence.” Journal of Quantitative Criminology 27:521–45.

107.

Tita

George

Riley

Jack

Greenwood

Peter

. 2003. “From Boston to Boyle Heights: The Process and Prospects of a ‘Pulling Levers’ Strategy in a Los Angeles Barrio.” Pp. 102–30 in Policing Gangs and Youth Violence, edited by Decker

Scott H.

. Belmont, CA: Wadsworth.

108.

Turanovic

Jillian J.

Pratt

Travis C.

. 2012. “‘Can’t Stop, Won’t Stop’: Self-control, Risky Lifestyles, and Repeat Victimization.” Journal of Quantitative Criminology. doi:10.1007 /s10940-012-9188-4

109.

Turanovic

Jillian J.

Pratt

Travis C.

. 2013. “The Consequences of Maladaptive Coping: Integrating General Strain and Self-control Theories to Specify a Causal Pathway between Victimization and Offending.” Journal of Quantitative Criminology 29:321–45.

110.

Vigil

James D.

1988. Barrio Gangs: Street Life and Identity in Southern California. Austin: University of Texas Press.

111.

von Hentig

Hans

. 1948. The Criminal and His Victim: Studies in the Sociobiology of Crime. New York: Schocken Books.

112.

Warr

Mark

. 2002. Companion in Crime: The Social Aspects of Criminal Conduct. New York: Cambridge University Press.

113.

Watkins

Adam M.

Huebner

Beth M.

Decker

Scott H.

. 2008. “Patterns of Gun Acquisition, Carrying, and Use among Juvenile and Adult Arrestees: Evidence from a High-crime City.” Justice Quarterly 25:674–700.

114.

Webb

Vincent J.

Katz

Charles M.

Decker

Scott H.

. 2006. “Assessing the Validity of Self-reports by Gang Members: Results from the Arrester Drug Abuse Monitoring Program.” Crime and Delinquency 52:232–52.

115.

Whitehill

Jennifer M.

Webster

Daniel W.

Frattaroli

Shannon

Parker

Elizabeth M.

. 2013. “Interrupting Violence: How the CeaseFire Program Prevents Imminent Gun Violence through Conflict Mediation.” Journal of Urban Health. doi:10.1007/s11524-013-9796-9.

116.

Wolfgang

Marvin E.

1958. Patterns in Criminal Homicide. Philadelphia: University of Pennsylvania Press.

117.

Wolfgang

Marvin E.

Ferracuti

Franco

. 1967. The Subculture of Violence: Towards an Integrated Theory in Criminology. London, England: Tavistock.

118.

Young

Jacob T. N.

Rees

Carter

. 2013. “Social Networks in Delinquency and Adolescence: Implications for Life-course Criminology.” Pp. 159–82 in Handbook of Life-course Criminology, edited by Gibson

Chris L.

Krohn

Marvin D.

. New York: Springer-Verlag.

The Contribution of Gang Membership to the Victim–Offender Overlap

Abstract

Objective:

Method:

Results:

Conclusion:

Keywords

Situating Gang Membership in the Victim–Offender Overlap

Existing Research

The Current Study

Method

Data

Dependent Variables

Independent Variables

Gang Membership

Theoretical Constructs

Control Variables

Analytic Strategy

Results

Descriptive and Bivariate Statistics

Multinomial Logit Models Predicting Violent Overlap Categories

Multilevel Logistic IRT Models of Propensity for Violent Victimization and Violent Offending

Discussion

Footnotes

Declaration of Conflicting Interests

Funding

Notes

References