Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Jan 1.
Published in final edited form as: Deviant Behav. 2021 Aug 6;43(9):1103–1119. doi: 10.1080/01639625.2021.1963644

YOUTH GANGS AND VICTIMIZATION: AN INVESTIGATION OF THE IMPACT OF GANG DYNAMICS ON EXPERIENCES OF VICTIMIZATION

John Leverso 1, Kate K O’Neill 2
PMCID: PMC9512271  NIHMSID: NIHMS1733881  PMID: 36172626

Abstract

Studies comparing gang members to similarly situated non-gang members find youth gang involvement is positively associated with experiencing simple and aggravated assault (i.e., violent victimization). This study expands on those studies by using data on gang dynamics from the Denver Youth Survey and bringing theory and concepts directly related to street gangs to the analysis of the relationship between gang membership and different forms of victimization. We focus on specific mechanisms—such as gang organization, centrality, leadership roles, identity, and tenure—that inform gang member behaviors while controlling for risky behaviors and personal histories. Findings indicate (1) gang organization is positively associated with both simple and aggravated assault victimization; (2) gang centrality is positively associated with aggravated assault victimization; and (3) being the leader of a gang is negatively associated with aggravated assault victimization. We discuss the implications of these results using a gang-informed framework.

Keywords: violence, victimization, gangs, delinquency

Youths’ violent offending and victimization can take many forms, including bullying, fighting, brandishing weapons, and in extreme cases, murder. Many of these behaviors are intrinsic parts of street gang culture and membership, and gang membership predicts involvement in youth violence (Thornberry et al. 2004). Youth violence typically involves young people victimizing other young people (David-Ferdon and Simon 2014), and while research often focuses on gang-involved youth as perpetrators (Pyrooz et al. 2016), recent scholarship suggests gang-involved youth are also uniquely susceptible to victimization (Conway-Turner et al. 2020). Intergang conflict, masculine posturing, and involvement in illicit activities are all part of the gang lifestyle, and broad criminological theories, such as routine activities theory (Cohen and Felson 1979; Hindelang et al. 1978), have been used to contextualize how gang members’ everyday lives disproportionately expose them to victimization (Spano, Freilich, and Bolland 2008; Taylor et al 2008). While studies using broad theories provide valuable insight into these correlations, they are limited in their ability to how or why within- and/or between-gang characteristics/organization facilitate violent victimization of members.

More narrowly scoped research that employs gang-informed frameworks supports assertions that gang dynamics are an important predictor of both violent offending and violent victimization. Such studies tend to use one of three theoretical models Thornberry et al. (2003) developed: selection, facilitation, and/or enhancement. The selection model posits individuals predisposed to violent behavior gravitate towards gangs because gangs allow them to engage in violent behavior. Gang-involved youth are therefore more likely to victimize and be victimized by their peers because their peers are similarly disposed toward violence. Conversely, the facilitation model argues gang-involved youth are no more violent than non-gang-involved youth, but that within-gang dynamics facilitate and encourage violence. The enhancement model combines these arguments, suggesting both individual behavioral dispositions and gang dynamics drive violence and victimization. Overall, research strongly supports the enhancement model (Delisi et al. 2009; Fox 2017; Melde, Taylor, and Esbensen 2009; Peterson, Taylor, and Esbensen 2004; Taylor et al. 2007; Wu and Pyrooz 2016), implicating gang dynamics and selection as co-occurring drivers of gang members’ violence and victimization. However, as with studies using broader criminological theories, these studies often employ comparisons of gang-involved youth to non-gang counterparts, in part due to data limitations. This, in turn, prevents full and comprehensive application of gang-informed theory and concepts and leaves links between gang dynamics and victimization of gang members undertheorized.

This paper logically extends gang-informed research on violence and moves beyond causal arguments concerning whether or not membership increases victimization by testing the differential influence of gang dynamics. We use a longitudinal sample of gang-involved youth, their offending and victimization histories, and gang dynamics to specify mechanisms through which gangs increase risk of violent victimization (simple and aggravated assault) among gang-involved youth. This data, paired with our use of gang-informed theory and concepts, allows us to account for the fact that gangs are embedded in unique contexts where features distinct to gangs have important implications for gang members’ victimization. Specifically, we analyze the roles of perceived levels of organization in the gang, gang centrality, leadership status, gang identity, and gang tenure (i.e., gang dynamics) in simple- and aggravated-assault victimization.

This research directly responds to calls for increased inclusion of gang-informed theory in gang research (Fox 2013; 2017) by using a longitudinal sample of 169 gang-involved youth (286 person years) from the Denver Youth Survey (DYS). While prior studies and datasets have addressed the influence of gang membership and gang member characteristics of violent victimization (Fox 2017; Taylor et al 2008; Wu and Pyrooz 2016), and specific gang dynamics on violent victimization (Decker, Katz, and Web 2008; Sweeten, Pyrooz, and Piquero 2013), data limitations have made it difficult to address all these factors at once. Further, samples used in the study of gang victimization are predominantly derived from school samples or jail surveys (Fox 2017; Lane, Armstrong and Fox 2019; Lane and Fox 2020; Melde et al., 2009; Peterson et al., 2004; Taylor et al., 2008; Taylor et al.,2007; Watkins and Melde 2018), which may not include individuals who do not attend school or individuals who engage in less serious offending.

In addition, DYS data on respondent behaviors and histories generally associated with violent victimization allows us to empirically assess both gang- and individual-level characteristics influence of these factors of violent victimization. Given prior research questions whether the relationship between gang membership and victimization holds when accounting for individual behavioral histories (Katz et. al 2011; Spano et al. 2008; Taylor et al. 2008) it is important to control for such characteristics, and we address these findings in our research question by asking: To what extent are gang dynamics associated with violent victimization when controlling for individual characteristics of gang members? In answering this research question, this paper illuminates mechanisms for increased victimization of gang members, and investigates them in models including risky behaviors and histories to ascertain whether these relationships hold net of these characteristics.

LITERATURE REVIEW

Victimization in the Context of the Gang

While joining a gang may decrease a member’s risk of certain types of victimization such as violence in the home or bullying at school (Rios 2011), gang members are at heightened risk of violence by both rival gang members and fellow gang members (Delisi et al. 2009; Fox 2017; Melde et al. 2009; Peterson et al. 2004; Peterson, Carson, and Fowler 2018; Wu and Pyrooz 2016). Routine activities theory (Cohen and Felson 1979; Hindelang et al. 1978) suggests the day-to-day activities of gang-involved youth places them in situations where they are likely to be victimized (Spano et al. 2008; Taylor et al. 2008). For example, street gang members regularly engage in violent disputes with members of rival gangs, and in the course of such interactions members typically perpetuate and experience violence. Street gangs are organized in a way that encourages these disputes (Decker and Van Winkle 1996; Papachristos 2009; Papachristos, Hureau, and Braga 2013), in that they hold to reciprocity norms such that attacks on individual members are tantamount to attacks on the gang as a group. Subsequently, perpetrators often attribute gang shootings to retaliation for perceived harm to their own gang (Decker and Curry 2002; Decker and Van Winkle 1996; Papachristos 2009). Such conflicts may persist over time and lead to violence among gang members who may have had little to do with the original sources of conflict.

Additionally, gang members may be victimized by members of their own gang. Joining gangs in some cases involves violent initiation rites, such as violent beatings (Vigil 1996). In this way, gangs may be organized in favor of violence against their own members because they abide by rules or systems that violently punish members for rule-breaking behaviors. Intra-gang conflict, therefore, can readily lead to severe violence if individuals are predisposed to violence or become accustomed to violence because of these dynamics. The precise frequency of intra-gang violence is unknown, but research suggests intra-gang violence is a fairly common experience among gang members (Papachristos 2009), and one study even found gang members are more likely to be murdered by members of their own faction than by members of rival gangs (Decker and Curry 2002).

A third source of violent victimization likely stems from involvement in illicit activities, such as drug sales, where gang members may be targeted for victimization by both gang and non-gang members for instrumental reasons, such as attempts to rob them for drugs or money. Gang members’ participation in such risky activities likely leads to increases in their victimization (Spano et al. 2008; Taylor et al. 2008; Wu and Pyrooz 2016).

Gang dynamics may affect all three sources of violence—inter-gang, intra-gang, and illicit activities. That is, members’ individual relationship to their membership and/or position in the gang may make them more or less susceptible to these forms of violence. The next section describes research on such dynamics.

Theorizing Gang Dynamics

Theories specific to street gangs suggest common mechanisms through which gang dynamics increases risk of victimization: gang organization (Decker et al. 2008; Leverso and Matsueda 2019), gang identity (Hennigan and Spanovic 2012; Leverso and Matsueda 2019), and gang role centrality and leadership status (or, as more recent research terms it, gang embeddedness) (Hagedorn and Macon 1988; Klein 1997; Pyrooz, Sweeten, and Piquero 2013; Vigil 2010). Broadly, gang organization refers to the extent to which gangs have established roles for members, leadership hierarchies, and behavioral regulations. The complexity of gangs’ organizational systems varies, and members of gangs may be at higher or lower risk of victimization because of this variation. Gang identity is the extent to which individuals identify with their gang and the extent to which that identification influences their behavior. The more strongly members identify as “prototypical” gang members (Turner et al. 1987), the more likely they are to behave according to established gang norms and expectations. Depending on the nature of these expectations, behaviors motivated by gang identity may increase risk of victimization, particularly if that identity is very strong. Gang centrality refers, specifically, to how central members are to gang activities. In addition, gang centrality is related to gang leadership status inasmuch as gang members in leadership roles may be more actively involved in particular gang activities. Because gang activities are sometimes violent, both centrality and leadership status may increase victimization. While some researchers look to gang centrality and leadership to measure “gang embeddedness” as a singular concept (Pyrooz et al. 2013) we disaggregate the two in our analyses so as to unravel their individual influence on gang members’ violent victimization. Although predominantly investigated in gang research on offending and disengagement, each of these theoretical perspectives pay dividends for understanding gang member victimization.

Gang Organization.

Gang organization is a broad term, and has been used to refer to gang characteristics such as differentiated roles and coordinated activities, the presence of leadership; regular face-to-face meetings; rules, codes, and norms with sanctions for violators; and initiation rites of passage (Bouchard and Spindler 2010; Decker et. al. 2008; Pyrooz et al. 2012). Studies on gang organization often focus on its impact on delinquency and serious offending. For example, Fagan’s (1989) study of gang members in Chicago, Los Angeles, and San Diego finds gangs with higher levels of organization generally include more serious delinquents. Likewise, Bouchard and Spindler (2010) find gang organization predicts drug sales, violent offending, and general delinquency among gang-involved youth. Decker et al.’s (2008) study of gang members in a drug abuse monitoring program found that even relatively low levels of gang organization are associated with offending, while Pyrooz et al. (2012) find modest support for the relationship between gang organization and offending in a cross-national study using three samples. Given gang organization’s relationship to offending, one might imagine that proximity to delinquents and delinquency, as found in these studies, puts gang members at increased risk of victimization. Indeed, both Decker et al. (2008) and Pyrooz et al. (2012) found correlations between organization and victimization.

We conceive of two important reasons for this connection. First, the positive relationship between gang organization and violent offending supports a positive relationship between gang organization and violent victimization. The reciprocal nature of gang conflict suggests that those involved in higher levels of violent offending are also at higher risk of experiencing violent victimizations. Second, the features of gang organization could directly lead to increases in victimization. Roles and rules, for example, may promote drug sales and fighting with other gangs, which in turn increases levels of offending and risk of victimization among members. Violent initiation requirements and physical assaults as punishment are themselves forms of intra-gang victimization.

Gang Identity.

Gang identity, or the extent to which individuals identify with and are influenced by their gang, is theorized to be both a cause and consequence of how tightly bonded individuals are to their gangs. Recent research on gang identity is often developed from social identity theory (see Lauger 2020 for a review of social identity theory). Social identity theory posits individuals identify with a particular social group—such as a gang—and the strength of this identity predicts behavior regarding the group membership (Tajfel 1978). Stable or increasingly positive social identities are desirable, and therefore individuals compare the “in-groups” to which they belong to the “out-groups” to which they do not. Individuals who make favorable comparisons of their in-group to the out-group tend to hold more positive group-linked identities and are therefore more likely to continue their membership and engage in behaviors that protect their membership. Among gangs and gang members, these in- and out-group comparisons may explain why gang-involved youth are disproportionately involved in both offending and victimization: gang members fight with each other to increase status, respect, and “street cred” in relation to other gangs, and when these behaviors result in positive in-group comparisons, a stronger, more positive gang identity develops (Wood 2014). Notably, identities are not static, but are made salient by different social contexts. When social identities are activated, they experience depersonalization; an individual sees themselves as a prototypical group member (Turner et al. 1987) and behaves accordingly. Given normative behaviors of gang membership and gang life include violence and criminality, it stands to reason that these are attributes of the prototypical gang member, and that gang identity will encourage violence (Hennigang and Spanovic 2012). While the relationship between gang identity and victimization remains understudied, we argue processes of categorization, which increases normative violent and criminal behaviors of gang members, and social comparisons, increase conflict. This in turn increases both victimization and violence.

Gang Centrality and Leadership Status.

Involvement in gang activities is not equally distributed across members. Gang centrality, more recently conceptualized as embeddedness within gang activities (Pyrooz et al. 2013), is an important gang dynamic that could help us understand gang member victimization. Gang tenure, role expectations, and the extent of engagement in inter-gang conflict varies by individual, with so-called “core” members being more central to gang activities than comparatively disengaged “fringe” members (Esbensen et al. 2001; Klein 1997). Core members central to (or embedded in) the gang are involved in more offending and have longer gang tenure than do fringe members (Hagedorn and Macon 1988; Pyrooz et al. 2013; Vigil 2010). In addition more central members may gravitate to roles reserved for high-status gang members, roles which can come with increased victimization. For instance, leadership and enforcement roles are often given to gang members who have demonstrated stronger fighting ability than others. Moreover, gang leaders may be more likely to engage in violence to solidify/maintain their position in the gang hierarchy. Generally speaking, gang embeddedness is positively associated with victimization (Sweeten et al. 2013), as well as offending.

Gang tenure (i.e., length of time in the gang) is likely associated with the gang dynamics highlighted in this paper. Given gang membership constitutes a high-risk state, lengthier gang tenure may increase risk of victimization. For example, individuals with longer tenure may be more active members than are newer gang members, and may be more likely to hold leadership positions. Given these likely associations, we control for gang tenure alongside our other gang dynamics, and discuss its influence accordingly. Taken together research suggests gang organization, centrality, leadership status, identity, and tenure, directly relate to violent offending, and likely influence victimization as well. Thus, we hypothesize:

H: Gang dynamics are associated with both simple- and aggravated-assault victimization, net of risky behavior and histories.

Because some research questions the unique contribution of gang membership to victimization, after accounting for individual offending histories (Gibson et al. 2009) or controlling for offending (Katz et al. 2011; Spano et al. 2008; Taylor et. al 2008), we test our hypothesis with risky behavioral histories to see if the hypothesized relationship holds net of these characteristics.

CURRENT STUDY

This study advances the scholarship of gang victimization in three important ways. First, while a majority of research has found that gang members are more likely than other youth to be victims of violence, potential mechanisms remain understudied. Thus, by investigating associations among gang member victimization and gang dynamics this study theoretically and practically contributes to existing literature. Second, our sample of high-risk youth complements other samples of adjudicated youth and school-enrolled children. Substantively, school-based samples may not include individuals who do not attend school and carceral samples may include only more serious offenders. Our high-risk sample will complement knowledge on gang victimization of individuals in other contexts, as many gang members may be truants and/or dropouts who have avoided adjudication (Pyrooz 2014).

Finally, our models include controls for past and present violent offending and current weapon carrying to assess the effects of gang dynamics net of individual behaviors. In cases where gang dynamics, particularly gang organization, were examined in previous research pertaining to victimization, the designs were bivariate (Decker et al. 2009) or cross sectional (Pyrooz et al. 2012), and lacked controls for individual level histories. We improve on these designs with multivariate models, which include controls for both past and present offending histories. In doing so, our study acknowledges selection-based arguments in order to isolate the impact of gang dynamics.

MATERIAL AND METHODS

The DYS is a longitudinal study of delinquency and drug use in high-risk neighborhoods in Denver, Colorado, from 1987 to 1998. The period in which these data were collected has implications for interpreting our results. The Bloods and Crips formed in Denver in the 1980s, resulting in a splintering of gangs, increases in gang conflict, and a proliferation of gangs and gang activities (Durán 2013). Given this context, the results of this study are most relevant to areas and time periods where gangs are undergoing proliferation and increased activity.

DYS data were gathered annually, creating ten waves. The first includes a total of 1,526 completed interviews. A child survey was used for respondents aged 7-11. After age 11 all respondents were administered a youth survey until they were aged approximately 18. Starting with Wave 6, all respondents aged 18 and over were given the adult survey. We use data from the youth and adult surveys because the child survey lacks questions about gang membership. We restrict the study to respondents who self-report being members of a street gang and report their gang is involved in fights with other gangs or participates in illegal activities (regardless of whether the respondent reported being involved with fights or illegal activities personally). This restricted measurement of gang membership is consistent with those used in other studies using the DYS (Esbensen and Huizinga 1993; Leverso and Matsueda 2019), and is necessary to ensure respondents are referring to a street gang and not a peer group.3 In total, 226 individuals representing 404 person years report gang membership during the survey period. However, because variables about gang organization that are central to our investigation were introduced in Wave 3, we further restrict our analysis to respondents reporting gang membership in Waves 3 or later, resulting in a sample of 201 persons and 360 person years. After list-wise deletion of missing key variables we have a final analytic sample of 286 person years representing 169 individuals.

Dependent Variables

Simple- and Aggravated-Assault Victimization.

The DYS contains three measures of violent victimization: how many times in the past year a respondent was strong-armed by someone with a weapon, hit by someone trying to hurt them, or attacked by someone with a weapon trying to seriously hurt or kill them. The strong-arm robbery variable was dropped as an outcome because it lacked sufficient variation to be a dependent variable (very few respondents reported this type of victimization). Of the two remaining variables, we treated being hit as a measure of simple-assault victimization and being attacked with a weapon as a measure of aggravated-assault victimization. While this contrasts with previous research where all three measures are combined to make a summed frequency scale (Childs, Cochran, and Gibson 2009; Gibson et al. 2009; Peterson et al. 2004; Taylor et al. 2007), there is good reason to use individual survey items to represent simple- and aggravated-assault victimization.

To start, prior research on gang member victimization cautions against using items without internal consistency (Fox 2013). The Cronbach’s alpha for the three-item scale is 0.14, indicating robbery, simple-assault, and aggravated-assault victimization cannot be reliably combined to measure violent victimization as a single construct. Further analysis of simple- and aggravated-assault victimization suggests they are not strongly related either. Specifically a correlation test found a week correlation (α=0.07), and prior literature asserts differences between these types of violent victimization in regards to severity (Hart and Rennison 2003), frequency (Lauritsen and Rezey 2013), and correlates (Lauritsen and Heimer 2008). Specific to gang research, scholars have warned about using theoretically different constructs of victimization in one measure (Ozer and Engel 2012). We therefore measure simple-assault victimization as a self-reported count of how many times the respondent was hit by someone trying to hurt them in the past year, and aggravated-assault victimization as a count of how many times the respondent was attacked by someone with a weapon trying to seriously hurt or kill them in the past year.

Independent Variables: Gang Dynamics

Perceived Gang Organization.

Following Leverso and Matsueda’s (2019) work on gangs using the DYS data, we measure gang organization based on subjective perceptions of organizational features. While a “true” and objective state of gang organization exists, the development of any gang organization measure necessarily relies on individual perception and reports thereof. Given perceptions of gang membership are as likely, if not more likely, to motivate individual behaviors as is the true state of gang organization (see Matsueda 2006; Thomas and Thomas 1928) we posit this is a reasonable approach given data limitations. Indeed, Pyrooz and Decker (2019) find support for using self-reported perceptions to research group level characteristics.

Gang organization is a summed scale of time-varying dichotomous variables. For each year a respondent answered “yes” to being a member of a street gang they were also asked to respond to a series of questions about the organization of that gang. Respondents were given eight measures of gang organization: “Tell me if the following describes your gang”: (1) “there are initiation rites,” (2) “the gang has established leaders,” (3) “the gang has regular meetings,” (4) “the gang has specific rules or codes,” (5) “gang members have a specific role,” (6) “there are specific roles for girls,” (7) “there are specific roles for each age group” and (8) “the gang has colors and symbols.” High scores indicate greater organizational structure in the gang. Exploratory and confirmatory factor analyses indicated a one-factor solution and the Cronbach’s alpha is 0.68. Similar questions and scales have been successfully used in previous studies (Bouchard and Spindler 2010; Decker 1996; Decker et al. 2007).

Gang Identity.

We follow previous research (Leverso and Matsueda 2019) and measure gang identity using six items measured from 1 to 5: “being in a gang makes me feel important,” “being in a gang makes me feel respected,” “being a member of the gang makes me feel like I’m a useful person to have around,” “being a member of the gang makes me feel like I really belong somewhere,” “I really enjoy being a member of the gang,” and “how important to you is the gang and their activities?” Exploratory and confirmatory factor analyses indicated a one-factor solution and Cronbach’s alpha was 0.88 for the scale. The scale score is the mean of all items divided by six so the scale remains bounded between 1 and 5. Higher scores indicate a stronger gang identity.

Gang Centrality.

We measure gang centrality with two variables. First, following previous research, we use a five-concentric-ring response scale (Esbensen et al. 2001; Melde, Diem, and Drake 2012). Respondents were told, “The target scale represents the activities that go on in your gang. How far out from the center of things are you?” This item was reverse-coded and higher scores indicated greater role centrality. Second, we include a separate variable asking respondents if they were the leader of the gang as there is evidence leadership and embeddedness go hand-in-hand (Pyrooz et al. 2013). Finally, we include a variable for the length of time an individual was a member of a gang (i.e. gang tenure) because gang membership is a high-risk state, which suggests greater time spent in the gang may be associated with increases in victimization. This variable is a summed score of how many waves an individual reports being in a gang. For example, an individual reporting gang membership in three waves would be scored a three.

Independent Variables: Risky Behaviors and Histories

We control for violent offending using a dichotomous measure of whether or not a respondent reports committing simple or aggravated assault within the past year and we include a lagged (t-1) violent offending variable to control for whether or not respondents were involved in violent offending in the prior year. In addition, we separately control for whether or not a respondent has carried a hidden weapon in the past year because Spano et al. (2008) has found that carrying a weapon renders gang membership a non-significant predictor of victimization. We also include behavioral measures of frequency of drinking beer or hard liquor in the last year. Finally, we include a dichotomous, lagged victimization variable to control for whether or not respondents had experienced any violent victimization in the prior year (see Table 1).

Table 1.

Variable Descriptions and Descriptive Statistics, Denver Youth Survey Waves 3-10

Variable Name Description Mean
(%)		 S.D. N
Victimization
 Simple Assault Self-reported count of victimization indicating how many times the respondent was hit by someone trying to hurt them in the past year. 2.19 11.07 169
 Aggravated Assault Self-reported count of victimization indicating how many times the respondent was attacked by someone with a weapon trying to seriously hurt or kill them in the past year. 0.87 4.29 169
 Prior Victimization 1 = Respondent reported experiencing a simple or aggravated assault in the previous year. 31 52
0 = Respondent did not report experiencing a simple or aggravated assault in the previous year. 69 117
Gang Dynamics
 Gang Organization Gang Organization Summed Scale 0-8. 4.50 2.04 169
 Gang Centrality If the Target Scale represents the activities that go on in your gang, how far out you are from the center, 1 = The outer edge, 5 = The center. 3.33 1.17 169
 Leadership 1 = Respondent reported being a gang leader. 9 15
0 = Respondent reported not being a gang leader. 91 154
 Gang Identity Gang identity: Six item scale. 2.92 0.88 169
 Gang Tenure
  1 Year Respondent reported gang involvement in one time point. 45 75
  2 Year Respondent reported gang involvement in two time points. 28 47
  3 Year Respondent reported gang involvement in three time points. 15 25
  4 Year Respondent reported gang involvement in four time points. 9 15
  5 Year Respondent reported gang involvement in five time points. 2 4
  6 Year Respondent reported gang involvement in six time points. 1 3
Risky Behaviors and Histories
 Violent Offending 1 = Respondent reported involvement involved in simple or aggravated assaults in the past year. 46 78
0 = Respondent reported not being involved in simple or aggravated assaults in the past year. 54 91
 Carrying a Hidden Weapon 1 = Respondent reported a carrying a hidden weapon. 52 88
0 = Respondent reported a carrying a hidden weapon. 48 81
 Beer Drinking Frequency Beer drinking frequency scale: in the last year, how often did you drink beer? 60.34 103.32 169
 Hard Liquor Drinking Frequency Hard liquor drinking frequency scale: in the last year, how often did you drink hard liquor? 18.51 54.20 169
Background and Contextual Variables
 Age Self-reported age 16.7 2.38 169
 Male 1 = Male 76 129
 Female 1 = Female 24 40
 Hispanic 1= Hispanic 54 91
 Non-Hispanic Black 1 = Black 34 57
 White/Other 1 = White or other race 12 21
 Income  Reported household income in 10,000s. 13,533 10,095 169
Open in a new tab

Independent Variables Background and Contextual Variables

A detailed description of background and contextual variables appears in Table 1. We include demographic measures of race, ethnicity, gender, socioeconomic status as measured by self-reported income, and age.

Analytic Strategy

To investigate the effects of gang dynamics on gang members’ violent victimization, we present our sample as a pooled time series. Our outcome measures of victimization are counts, and, therefore, we assume the counts follow a Poisson distribution. In estimating Poisson regressions, we found overdispersion—the variance exceeds the mean due to a large number of zeros. The negative binomial corrects for overdispersion, providing consistent and asymptotically efficient estimates of parameters and consistent estimates of standard errors. We use robust standard errors to correct for clustering on individuals and control for previous crime and victimization at t-1. Our analysis then proceeds in two steps. We first present descriptive statistics for our full sample at the person level taken from their first year of gang membership (N= 169). Next, we present person-year negative binomial regression models for both simple- and aggravated-assault victimization based on gang identity, gang organization, gang centrality, and risky behaviors and histories and controls (N= 286). Specifically, we test the hypothesis that gang dynamics are associated with both simple- and aggravated-assault victimization, net of risky behavior and histories.

RESULTS

Descriptive Statistics and Excluded Observations

Descriptive statistics for our sample of gang members are presented in Table 1. Consistent with prior research, DYS gang members report a relatively brief period of membership: Almost half of respondents report only one year of gang tenure, 28% report two years of gang tenure, and 27% report three to six years of gang tenure. The composition of the sample reflects the ethnic and racial composition of the neighborhoods in which they reside: 54% of respondents are Hispanic and more than one-third are Black.4 The mean age respondents reported joining a gang is 16.7 (SD = 2.38).

We examined the distribution of missing data on the key variables in our analysis by comparing the distribution of 24 variables for included observations (n = 169) against excluded observations (n = 57). Pertaining to the majority of variables in our models we find no substantial differences between included and excluded observations. However, findings regarding gang tenure, violent offending, and carrying a weapon did indicate some statistically significant differences between included and excluded observations – though the practical implications of these differences are mostly incidental.5 As expected given having only one data point increases the likelihood of list-wise deletion, respondents missing from analysis are more likely to report only one year of gang tenure. In addition, the 57 respondents excluded from analyses were more likely to report carrying a weapon (71% vs. 52%) and more likely to report involvement in violent offending (64% vs. 46%). While these differences do not substantively alter our interpretation of our results concerning carrying a weapon, violent offending, and violent victimization, it is worth noting more serious offenders may be underrepresented in our analyses.

Experiences of Victimization

Table 2 presents models predicting self-reported simple- and aggravated-assault victimization. When discussing results from negative binomial models, we convert coefficients into percent changes in-text, using the following formula:

(expβ-1)×100=percentchangeinvictimizationcounts

Table 2.

The Effects of Gang Membership on Reported Victimization (Negative Binomial Models)

Simple Assault
Victimization		 Aggravated Assault
Victimization
Observations 286 286
Individual Respondents 169 169
 Intercept 0.36*** (1.19) −7.62*** (1.74)
Gang Dynamics
 Gang Organization 0.20** (0.07) 0.27* (0.11)
 Gang Centrality 0.11 (0.11) 0.71*** (0.15)
 Leadership 0.18 (0.40) −0.96† (0.53)
 Gang Identity −0.20
(0.15)		 −0.08
(0.16)
 Gang Tenure −0.04
(0.11)		 0.08 (0.13)
Risky Behaviors and Histories
 Prior Victimization 0.01 (0.31) 0.21 (0.32)
 Violent Offending 1.32*** (0.29) 0.89* (0.35)
 Violent Offendingt-1 −0.46 (0.29) 0.16 (0.30)
 Carrying a Hidden Weapon 1.12*** (0.26) 1.34** (0.38)
 Beer Drinking 0.04** (0.01) 0.04*** (0.17)
 Hard Liquor Drinking −0.04 (0.03) −0.04 (0.03)
Background and Contextual Variables
 Age −0.17** (0.06) 0.09 (0.07)
 Male 0.92* (0.37) 0.00 (0.45)
 Black −0.01 (0.32) −0.43 (0.40)
 White/Other Race −0.28 (0.37) 0.19 (0.43)
 Income in $10,000s −0.18 (0.21) 0.13 (0.27)
Fit Statistics
 Wald Chi-Square (16) 81.30 104.44
 Log-Likelihood −340.7762 −231.1857
 BIC 783.3602 564.178
Open in a new tab
†

p < .10

*

p≤0.05

**

p≤0.01

***

p≤0.001.

Robust standard errors in parentheses.

Given the frequency of beer and hard liquor consumption we divided the values by 10 so the interpretation is in terms of 10's of drinks.

Coefficients converted to percent changes in-text using the following formula: (expβ-1)×100=percent change in victimization counts

Overall, results support our hypothesis that gang dynamics are associated with simple- and aggravated-assault victimization. However, some gang dynamics are clearly more important than others with regards to predicting violent victimization, and the salience of gang dynamics is not uniform across the two types of victimization.

Simple Assault

Of our focal variables, only gang organization is associated with simple-assault victimization: a one-unit increase in perceived gang organization is associated with a 22% increase in simple-assault victimization counts among gang members. These results indicate that as perceptions of gang organization increase, the number of times per year individual gang members are victims of simple assault do as well. Such results suggest gangs are organized in favor of violence not only towards outsider threats, but towards their own members as well. Gang identity and centrality were not significantly associated with simple-assault victimization, a difference we explore further in our Discussion section.

Among individual behaviors and histories, violent offending, carrying a hidden weapon, and beer drinking are positively and significantly associated with simple-assault victimization counts. Gang members who report having engaged in one or more violent offenses in the past year see a 274% increase in victimization counts, though prior histories of violent offending are not a significant predictor of simple-assault victimization. Carrying a hidden weapon also exerts a sizable effect on simple-assault victimization; it is associated with a 206% increase in victimization counts. Drinking beer 10 times a year is associated with a 4% increase in reports of victimization. Of the background and contextual variables, age is associated with a decrease in simple-assault victimization, while being male is associated with a 151% increase in victimization. These results imply perceived gang organization and individual histories are important correlates of gang members’ simple-assault victimization, and reveal that not all gang dynamics are associated with simple-assault victimization.

Aggravated Assault

Our results for aggravated assault-victimization also partially support our hypothesis. Differentiating it from our results for simple-assault victimization is the fact that both perceived gang organization and gang centrality are significant predictors of aggravated-assault victimization. The positive association between perceived gang organization and aggravated-assault victimization is slightly larger in magnitude (31% increase with each additional feature of gang membership) than is the positive association between perceived gang organization and simple-assault victimization (22%). Further, a one-unit increase in gang centrality is associated with a 103% increase in aggravated-assault victimization. This is interpreted to mean gang members who are (or believe themselves to be) heavily involved in gang activities experience aggravated assaults in considerably greater numbers than do gang members who are further removed from gang activities. On the other hand respondents who report being a leader in their gang experience approximately 62% fewer incidents of aggravated-assault victimization than do non-leaders (p < 0.10).

As with simple-assault victimization, individual behaviors and histories are associated with aggravated-assault victimization. Having carried a hidden weapon in the past year is associated with an 282% increase in counts of aggravated-assault victimization, having been involved in one or more violent offenses is associated with a 144% increase in counts of aggravated-assault, and drinking beer 10 times a year is again associated with a 4% increase in reports of aggravated-assault victimization. No demographic variables are associated with experiences of aggravated-assault victimization.

DISCUSSION

Previous research finds gang membership is associated with increases in victimization (Delisi et al. 2009; Fox 2017; Melde et al. 2009; Peterson et al. 2004; Taylor et al. 2007; Wu and Pyrooz 2016). However, a scarcity of gang-only samples means researchers generally compare gang members to similarly situated non-gang members using broader criminological theory. As a result, most research on how gang membership influences victimization offers a partial account of how gang dynamics influence youths’ victimization trajectories. By focusing on specific mechanisms shown to inform gang member behavior (Decker et al. 2008; Hennigan and Spanovic 2012; Leverso and Matsueda 2019; Hagedorn and Macon 1988; Klien 1997; Vigil 2010; Pyrooz et al. 2013), we are able to empirically support our gang-informed framework and add nuance to the gang/victimization relationship. In addition, our inclusion of both gang dynamics and individual behaviors and histories known to influence victimization (Katz et al. 2011; Spano et al. 2008; Taylor et al. 2008) allows us to test the importance of gang dynamics net of these factors.

We find varying levels of support for our hypothesis, gang dynamics are associated with both simple- and aggravated-assault victimization, net of risky behavior and histories. While perceived gang organization is positively and significantly associated with both simple- and aggravated-assault victimization, gang centrality and leadership are associated only with aggravated assault, and gang identity does not appear to be associated with violent victimization. Consistent with Ozer and Engel (2012), our findings underscore the importance of disaggregating types of violent victimization in research on its correlates. One reason is that simple assault is considerably more prevalent than the more severely injurious victimization experience of aggravated assault (see Morgan and Truman 2019): DYS gang member respondents reported having experienced simple assault at more than double the amount of aggravated assault. Moreover, there are substantive differences in the experiences of our measures of simple and aggravated assault, being hit and being hit with a weapon by someone trying to cause death or serious injury (aggravated assault). For example, the former could describe an impulsive schoolyard fist fight among youths, whereas the latter could imply some premeditation (obtaining a weapon) and considerably more malicious intent. In line with this, our findings suggest that while these two types of violent victimization likely share some explanatory mechanisms, there are also important differences in how and why these events come to pass. While perceived gang organization is a crucial factor in both simple- and aggravated-assault victimization, gang centrality and leadership emerge as important predictors of aggravated-, but not simple-, assault victimization.

The finding that perceived gang organization is broadly associated with increases in violent victimization is unsurprising given prior literature indicates gangs are generally organized in ways that increase both within- and between-gang violence (Decker and Curry 2002; Decker and Van Winkle 1996; Papachristos 2009). Initiation-type rituals (Vigil 1996), corporal punishment for violations of gang code (Padilla 1992) and impulsive physical scuffles amongst mostly young and male gang members may contribute to gang members’ experiences with simple-assault victimization. In addition, there is a feedback loop between violent offending and victimization (Pyrooz, Moule, and Decker, 2014). Because gang organization is associated with greater levels of violent offending, it follows that members of gangs with greater levels of organization may be victimized at higher rates than gangs with less organization. Indeed, we find positive correlations between violent offending and violent victimization (see Table 2), and perceived gang organization and violent offending (supplementary analyses not shown).

While perceived gang organization is associated with both simple- and aggravated-assault victimization, gang centrality and leadership are only associated with the less common but more severe-in-its-consequences aggravated-assault victimization. We suggest this finding is related to what these two gang dynamics have in common: individual agency. Where gang organization and perceptions thereof may be more closely tied to structural characteristics of the gang, centrality and leadership are more dependent on individual gang members’ characteristics, desires, and abilities. Gang organization in favor of violence generally places gang members at higher risk of offending and victimization than similarly situated youth, but there is also variation among gang members regarding risk-taking and offending behaviors. Our findings suggest the gang-specific consequences of this behavioral diversity are expressed through gang centrality and leadership status. Specifically, we argue that as centrality to gang activities increases, risk of exposure to serious victimization does too. Simply put, gang activities are high risk, so it follows that the more gang activities one participates in the higher one’s risk of victimization.

This assertion is supported by our finding that gang centrality is associated with increases in aggravated-assault victimization (see also Sweeten et al. 2013). Further, more central gang members may be more likely to engage in retaliatory violence, which often involves acts of violence above and beyond simple assault (Papachristos et al. 2013). In support of this statement, supplemental analyses indicate more “central” DYS gang members report more involvement in violent offending. It therefore appears that while gang organization uniformly influences risk of victimization among gang members, regardless of their within-gang status, central members of the gang experience contexts for risk above and beyond those of other members, and those contexts place them at greater risk of experiencing severe violent victimization. However, while gang dynamics indicative of being a more active member increase risk of aggravated-assault victimization, leadership status appears to exert the opposite effect.

The finding that being a leader is not related to simple assault and is negatively associated with aggravated-assault victimization is surprising. While our null finding for simple assault supports the assertion that individual-level gang dynamics are not especially predictive of simple-assault victimization, the tentative finding that leadership roles insulate members from aggravated-assault victimization appears to counter our narrative that “core” gang members are exposed to serious victimization risks more than peripheral members. However, we argue this finding may be because gang members’ perceptions of leaders as among the toughest people in a gang deters would-be assailants. In addition, leadership status may in fact be tied to capacity for violence, and it is possible gang leaders are more likely to forcibly end conflicts before they escalate past what we identify here as simple assault, and have greater control over decisions to fight (see Short and Strodtbeck 1965). Leadership roles, therefore, may afford members some protection from aggravated-assault victimization, though leaders experience simple assault victimization at the same rates as other members.

The finding that gang identity is not significantly associated with violent victimization is somewhat surprising, and we believe it has two likely explanations. First, as with other null findings regarding simple-assault victimization, it is likely that gang organization is the most salient predictor of this victimization type and that its influence eclipses the influence of gang dynamics at the individual level. Second, with regards to aggravated-assault victimization, it may be possible to identify strongly as a gang member but to be relatively uninvolved in high-risk gang activities. While gang centrality and identity likely overlap in important ways, they are not strongly correlated (α=0.26). Thus, it is possible that even among members who strongly identify with their gang, there is enough variation in taste for risk among these members that some are substantially more likely than others to steer clear of high-conflict, high-risk contexts. This finding therefore suggests a defining characteristic among deeply embedded gang members is their willingness to expose themselves to risky contexts, and that this exposure is an important predictor of aggravated-assault victimization among gang members. Regardless, given the relative novelty of this finding we encourage future researchers to explore this relationship using other datasets and analytical methods.

Of individual behaviors and histories, violent offending, carrying a weapon, and drinking beer were significantly associated with both types of violent victimization. Regarding simple assault, and taken together with our null findings regarding individual-level gang dynamics, we interpret results to mean the influence of individual-level characteristics that exist independent of respondents’ gang membership eclipses the influence of individual-level gang dynamics. Taking our results regarding aggravated assault together with the finding that gang centrality and leadership are important correlates of aggravated-assault victimization suggests both gang-dependent and non-gang-dependent individual-level characteristics influence aggravated-assault victimization. The finding that carrying a weapon exerts strong effects on aggravated-assault victimization aligns with prior research on the importance of individual behaviors, but contradicts prior research that indicates gang dynamics are spurious if one accounts for carrying a weapon (Spano et al. 2008). Further, our analyses do not support the notion that prior victimization and/or offending are key factors in predicting gang members’ violent victimization. Thus, it appears prior histories of violent victimization and offending among gang members are less important than are their current gang contexts.

The above findings improve our knowledge of gang member victimization in important ways. DYS data on youths in high-risk communities facilitates a new vantage point in understanding the relationship between gangs and victimization in that it allows us to compare gang members to other gang members in an emergent gang city. In addition, this paper demonstrates the advantages of using gang-informed theory and concepts to better understand how different gang dynamics influence different types of violent victimization among gang-involved youth. The findings that the influence of gang dynamics varies based on both the type of dynamic and the type of victimization being analyzed should motivate future researchers to continue to disaggregate gang dynamics and victimization to further unpack this complex relationship.

This study has some limitations, and these limitations demonstrate future research is necessary to further support the findings and arguments presented in this paper. To start, while our sample affords us a rare opportunity to empirically compare gang members to other gang members, it is somewhat small and yields less statistical power than do larger samples (see Long and Fresses 2006). Thus our estimates of effect sizes may be conservative. Limitations related to our sample size may be especially relevant with regards to our argument that leadership roles insulate gang members from aggravated-assault victimization, given the leadership coefficient is only significant at p ≤ 0.10. Therefore, we present this finding as suggestive and encourage future researchers to critically evaluate this relationship. Finally, our measure of prior violent victimization necessarily includes victimization both within and prior to gang membership. While we explored restricting this variable to one state or the other, we ultimately determined it would have reduced the sample size in a way that would preclude multivariate models. For this reason, we suggest future researchers work to disaggregate pre- and post-gang-membership victimization in analyses wherever possible.

Regardless of its limitations, this study demonstrates the value of using gang-informed theory to analyze the relationship between gang dynamics and violent victimization in gang contexts. This research moves past the causal argument that gang membership leads to higher rates of victimization and suggests gang dynamics – particularly gang organization, gang centrality, and leadership – act as mechanisms to influence simple- and aggravated-assault victimization among gang members. Further, our results are robust to the inclusion of mechanisms for victimization suggested in prior literature, such as drinking, offending, and prior victimization. Taken together, these results clearly illustrate the advantages of investigating gang dynamics using gang-informed theory by showing this approach can help researchers understand an important dimension of youths’ violent victimization.

Footnotes

3

While research demonstrates self-report alone is a valid measure of gang membership (Esbensen et al. 2001) the more restricted definition used here is necessary to rule out youths in non-criminal peer groups. For example, one participant reported being a member of a gang, but further analysis into the name of this gang revealed the respondent was referring to a church choir.

4

Unfortunately, the DYS did not disaggregate between race and ethnicity during data collection. Therefore, “Hispanic” is grouped along with racial identities despite the fact that it is an ethnicity. Thus, Black, White, and Asian Hispanics were forced to select either their race or their ethnicity during data collection.

5

Due to nontrivial levels of missingness, additional analyses were conducted using multiple imputation techniques. Missing values for our covariates were imputed using Stata 15’s chained command and imputing the data 10 times. Our imputation models adjusted for all independent variables with missingness. However, missing values for outcome variables were not imputed to prevent the introduction of bias. Sample size in these models increased to 343. Results, available on request, were consistent with the reported models.

REFERENCES

  1. Bouchard Martin, and Spindler Andrea. 2010. “Groups, gangs, and delinquency: Does organization matter?” Journal of Criminal Justice 38(5): 921–933. [Google Scholar]
  2. Childs Kristina K., Cochran John K., and Gibson Christopher L.. 2009. “Self-control, gang membership, and victimization: An integrated approach.” Journal of Crime and Justice 32(1): 35–60. [Google Scholar]
  3. Cohen Lawrence E., and Felson Marcus. 1979. “Social change and crime rate trends: A routine activity approach.” American Sociological Review 588–608. [Google Scholar]
  4. Conway-Turner Jameela, Visconti Kari, and Winsler Adam. 2020. “The Role of Gang Involvement as a Protective Factor in the Association Between Peer Victimization and Negative Emotionality.” Youth and Society 52(3): 469–489. [Google Scholar]
  5. David-Ferdon Corrinne, and Simon Thomas R.. 2014. “Preventing Youth Violence: Opportunities for Action.” Centers for Disease Control and Prevention. [Google Scholar]
  6. Decker Scott H., Katz Charles M., and Webb Vincent J.. 2008. “Understanding the black box of gang organization: Implications for involvement in violent crime, drug sales, and violent victimization.” Crime and delinquency 54(1): 153–172. [Google Scholar]
  7. Decker Scott H., and Van Winkle Barrik. 1996. Life in the gang: Family, friends, and violence. Cambridge University Press. [Google Scholar]
  8. Decker Scott H., and Curry David G.. 2002. “Gangs, gang homicides, and gang loyalty: Organized crimes or disorganized criminals.” Journal of Criminal Justice 30(4): 343–352. [Google Scholar]
  9. DeLisi Matt, Barnes JC, Beaver Kevin M., and Gibson Chris L.. 2009. “Delinquent gangs and adolescent victimization revisited: A propensity score matching approach.” Criminal Justice and Behavior 36(8): 808–823. [Google Scholar]
  10. Durán Robert. 2013. Gang life in two cities: An insiders journey. Columbia University Press. [Google Scholar]
  11. Esbensen Finn-Aage and Huizinga David. 1993. “Gangs, drugs, and delinquency in a survey of urban youth.” Criminology 31(4): 565–589. [Google Scholar]
  12. Esbensen Finn-Aage, Winfree Thomas L. Jr, He Ni, and Taylor Terrence J.. 2001. “Youth gangs and definitional issues: When is a gang a gang, and why does it matter?” Crime and delinquency 47(1): 105–130. [Google Scholar]
  13. Fagan Jeffery. 1989. “The social organization of drug use and drug dealing among urban gangs.” Criminology 27(4): 633–670. [Google Scholar]
  14. Fox Kathleen A. 2013. “New developments and implications for understanding the victimization of gang members.” Violence and victims 28(6): 1015–1040. [DOI] [PubMed] [Google Scholar]
  15. Fox Kathleen A. 2017. “Gangs, gender, and violent victimization.” Victims and Offenders 12(1): 43–70. [Google Scholar]
  16. Gibson Chris L., Miller Mitchell J., Jennings Wesley G., Swatt Marc, and Gover Angela. 2009. “Using propensity score matching to understand the relationship between gang membership and violent victimization: A research note.” Justice Quarterly 26(4): 625–643. [Google Scholar]
  17. Hagedorn John M., and Macon Perry. 1988. People and Folks. Gangs, Crime and the Underclass in a Rustbelt City. Lake View Press, PO Box 578279, Chicago, IL 60657: (paperback: ISBN-0-941702-21-9; clothbound: ISBN-0-941702-20-0). [Google Scholar]
  18. Hart Timothy C. and Rennison Callie M.. 2003. Reporting crime to the police, 1992-2000. US Department of Justice, Office of Justice Programs; Washington, DC. [Google Scholar]
  19. Hindelang Michael J., Gottfredson Michael R., and Garofalo Jane. 1978. Victims of personal crime: An empirical foundation for a theory of personal victimization. Cambridge, MA: Ballinger. [Google Scholar]
  20. Hennigan Karen, and Spanovic Marija. 2012. “Gang dynamics through the lens of social identity theory.” Pp. 127–149 in Youth gangs in international perspective. Springer, New York, NY. [Google Scholar]
  21. Katz Charles M., Webb Vincent J., Fox Kate, and Shaffer Jennifer N.. 2011. “Understanding the relationship between violent victimization and gang membership.” Journal of Criminal Justice 39(1): 48–59. [Google Scholar]
  22. Klein Malcom W. 1997. The American street gang: Its nature, prevalence, and control. Oxford University Press. [Google Scholar]
  23. Lane Jodi, Armstrong Gaylene S., and Fox Kathleen A.. 2019. “Fear of victimization among incarcerated youths: Examining the effects of institutional “neighborhood” characteristics and gang membership.” Youth & Society 51(3): 417–439. [Google Scholar]
  24. Lane Jodi, and Fox Kathleen A.. 2020. “Race and ethnic differences in fear of property, personal, and gang victimization among people involved in crime: testing the effects of perceived neighborhood characteristics.” Victims & Offenders 15(4): 395–417. [Google Scholar]
  25. Lauger Timothy R. "Gangs, identity, and cultural performance. 2020." Sociology Compass 14 (4). [Google Scholar]
  26. Lauritsen Janet L. and Heimer Karen. 2008. The Gender Gap in Violent Victimization, 1973–2004. Journal of Quantitative Criminology 24(2): 125–147. [Google Scholar]
  27. Lauritsen Janet L. and Rezey Maribeth L.. 2013. Measuring the prevalence of crime with the national crime victimization survey. U.S. Department of Justice Office of Justice Programs Bureau of Justice Statistics. Washington, DC. [Google Scholar]
  28. Leverso John, and Matsueda Ross L.. 2019. “Gang organization and gang identity: An investigation of enduring gang membership.” Journal of quantitative criminology 35(4): 797–829. [Google Scholar]
  29. Long J. Scott, and Freese Jeremy. 2006. Regression models for categorical dependent variables using Stata. Vol. 7. Stata press.. [Google Scholar]
  30. Matsueda RL (2006) Criminological implications of the thought of George Herbert Mead. In: Deflem M (ed) Sociological theory and criminological research: views from Europe and the United States. Elsevier, Oxford, pp 77–108 [Google Scholar]
  31. Melde Chris, Taylor Terrence J., and Esbensen Finn-Aage. 2009. “‘I got your back:’ An examination of the protective function of gang membership in adolescence.” Criminology 47(2): 565–594. [Google Scholar]
  32. Melde Chris, Diem Chelsea, and Drake Gregory. 2012. “Identifying correlates of stable gang membership.” Journal of Contemporary Criminal Justice 28(4): 482–498. [Google Scholar]
  33. Morgan RE, and Truman JL. “Criminal victimization 2020. National Crime Victimization Survey 2019, Bureau of Justice Statistics. [Google Scholar]
  34. Ozer M. Murat, and 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(1): 105–124. [Google Scholar]
  35. Padilla FM (1992) The gang as an American enterprise. Rutgers University Press, Rutgers. [Google Scholar]
  36. Papachristos Andrew V. 2009. “Murder by structure: Dominance relations and the social structure of gang homicide.” American journal of sociology 115(1): 74–128. [DOI] [PubMed] [Google Scholar]
  37. Papachristos Andrew V., Hureau David M., and Braga Anthony A.. 2013. “The corner and the crew: The influence of geography and social networks on gang violence.” American sociological review 78(3): 417–447. [Google Scholar]
  38. Peterson Dana, Taylor Terrence J., and Esbensen Finn-Aage. 2004. “Gang membership and violent victimization.” Justice Quarterly 21(4): 793–815. [Google Scholar]
  39. Peterson Dana, Carson Dena C., and Fowler Eric. 2018. “What’s sex (composition) got to do with it? The importance of sex composition of gangs for female and male members’ offending and victimization.” Justice Quarterly 35(6): 941–976. [Google Scholar]
  40. Pyrooz David C. 2014. “From colors and guns to caps and gowns? The effects of gang membership on educational attainment”. Journal of research in crime and delinquency 51(1): 56–87. [Google Scholar]
  41. Pyrooz Davod C., Fox Andrew M., Katz Charles M., and Decker Scott H.. 2012. “Gang organization, offending, and victimization: A cross-national analysis.” Pp. 85–105 in Youth gangs in international perspective. Springer, New York, NY. [Google Scholar]
  42. Pyrooz David C., Sweeten Gary, and Piquero Alex R.. 2013. “Continuity and change in gang membership and gang embeddedness.” Journal of Research in Crime and Delinquency 50(2): 239–271. [Google Scholar]
  43. Pyrooz David C., Moule Richard K. Jr, and Decker Scott H.. 2014. “The contribution of gang membership to the victim–offender overlap.” Journal of Research in Crime and Delinquency 51(3): 315–348. [Google Scholar]
  44. Pyrooz David C., Turanovic Jillian J., Decker Scott H., and Wu Jun. 2016. “Taking stock of the relationship between gang membership and offending: A meta-analysis.” Criminal Justice and Behavior 43(3): 365–397. [Google Scholar]
  45. Pyrooz David C., and Decker Scott H.. 2019. Competing for control: Gangs and the social order of prisons. Cambridge University Press. [Google Scholar]
  46. Rios Victor M. (2011). Punished: Policing the lives of Black and Latino boys. NYU Press. [Google Scholar]
  47. Short James F., and Strodtbeck Fred L.. 1965. Group process and gang delinquency. Chicago: University of Chicago Press. [Google Scholar]
  48. Spano Richard, Freilich Joshua D., and 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(2): 381–410. [Google Scholar]
  49. Sweeten Gary, Pyrooz David C., and Piquero Alex R.. 2013. “Disengaging from gangs and desistance from crime.” Justice Quarterly 30(3): 469–500. [Google Scholar]
  50. Tajfel Henri E. 1978. Differentiation between social groups: Studies in the social psychology of intergroup relations. Academic Press. [Google Scholar]
  51. Taylor Terrence J., Peterson Dana, Esbensen Finn-Aage, and Freng Adrienne. 2007. “Gang membership as a risk factor for adolescent violent victimization.” Journal of Research in Crime and Delinquency 44(4): 351–380. [Google Scholar]
  52. Taylor Terrence J., Freng Adrienne, Esbensen Finn-Aage, and Peterson Dana. 2008. “Youth gang membership and serious violent victimization: The importance of lifestyles and routine activities.” Journal of interpersonal violence 23(10): 1441–1464. [DOI] [PubMed] [Google Scholar]
  53. Thomas William Isaac, and Thomas Dorothy Swaine. 1928. “The methodology of behavior study.” The child in America: Behavior problems and programs. 553–576. [Google Scholar]
  54. Thornberry Terrence P., Huizinga David, and Loeber Rolf. 2004. “The causes and correlates studies: Findings and policy implications.” Juv. Just 9: 3. [Google Scholar]
  55. Thornberry Terrence P., Krohn Marvin D., Lizotte Allen J., Tobin Kimberly, and Smith Carolyn A.. 2003. Gangs and delinquency in developmental perspective. Cambridge University Press. [Google Scholar]
  56. Turner John C., Hogg Michael A., Oakes Penelope J., Reicher Stephan D., and Wetherell Margaret S.. 1987. Rediscovering the social group: A self-categorization theory. Basil Blackwell. [Google Scholar]
  57. Vigil James D. 1996. “Street baptism: Chicano gang initiation.” Human Organization 55(2): 149–153. [Google Scholar]
  58. Vigil James D. 2010. Barrio gangs: Street life and identity in Southern California. University of Texas Press. [Google Scholar]
  59. Watkins Adam M., and Melde Chris. 2018. “Gangs, gender, and involvement in crime, victimization, and exposure to violence.” Journal of criminal justice 57:11–25. [Google Scholar]
  60. Wood Jane L. 2014. “Understanding gang membership: The significance of group processes.” Group Processes and Intergroup Relations 17(6): 710–729. [Google Scholar]
  61. Wu Jun, and Pyrooz David C.. 2016. “Uncovering the pathways between gang membership and violent victimization.” Journal of quantitative criminology 32(4): 531–559. [Google Scholar]

RESOURCES