Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Mar 24.
Published before final editing as: J Crime Justice. 2025 Mar 24:10.1080/0735648X.2025.2482264. doi: 10.1080/0735648X.2025.2482264

The Risks of Exposure: The Impact of Social Circles on Youth Gang Involvement and Criminal Behavior in Adulthood

John Leverso 1, David Schwager 2, Miranda Sitney 3, Kelsey Ravindren 3, Jennifer Roark 3
PMCID: PMC12370007  NIHMSID: NIHMS2070061  PMID: 40857417

Abstract

Social interactions within deviant circles during adolescence increase the likelihood of criminal behavior in the ensuing years. Similarly, exposure to gang-involved peers, parents, siblings, and cousins heightens the likelihood of gang membership. However, few studies have simultaneously considered the impact of exposure to both gang and criminal justice-involved circles. This study addresses this gap by using logistic regression and administrative data from Multnomah County, Oregon, on male adolescents who received a criminal referral between 2002 and 2017. Our findings reveal that exposure to social circles with both gang and criminal justice significantly influences gang membership, whereas exposure to non-gang-related criminal networks is more strongly associated with adult incarceration or probation. Further analysis indicates that fathers, same-generation family members, and peers play a significant role in influencing both gang membership and adult involvement with the criminal justice system, including arrest and probation. The findings illuminate the importance of examining the entire array of negative affiliations among adolescents and including a multidimensional context of youths’ deviant relationships.

Social relationships profoundly shape human behavior, serving as pathways for the transmission of norms, values, and behaviors. Within criminology, foundational theories like differential association and social learning emphasize that deviant behavior is learned through close interactions in environments where criminal or gang-related values predominate (Sutherland, 1947; Akers, 1998). Youth embedded in networks of deviant peers, family members, or romantic partners are particularly susceptible to adopting similar behaviors, as repeated exposure to these influences can normalize criminal activity (Haynie, 2002; Lonardo et al. 2008; Rowe and Farrington 1997).

Both research on gang involvement and research on criminal behavior has not addressed the possibility of cumulative effects of different types of relationships. The former typically focuses on peers only (Klein and Maxson, 2010; for exceptions see Augustyn et al., 2017; Hashimi et al., 2021; Kisser & Pyrooz 2009); the latter typically focuses on parents or peers—but not both— and typically ignores siblings (Geller et al., 2009; Mears and Siennick, 2016; Noel and Hoeben, 2022; Turney and Lanuza, 2017; Wakefield and Wilderman, 2013). This fragmented approach leaves a critical gap in understanding how overlapping social relationships collectively shape youth delinquency and gang affiliation.

At the same time existing research has not examined the simultaneous impact of exposure to both gang-involved and criminally involved individuals. Yet gang involvement represents a distinct form of deviance with unique social dynamics that warrant specific attention. While gangs overlap with broader criminal networks, features such as initiation rituals, informal rule systems, and group identities set them apart (Decker et al., 2008; Leverso and Matsueda 2019; Papachristos, Leverso and Hureau 2024; McGloin and Collins 2015). These characteristics influence not only behavior but also the intensity and durability of members’ social ties.

Our study addresses these limitations by exploring how exposure to both gang-involved and criminally involved social groups during adolescence shapes gang affiliation and adult outcomes, including incarceration and probation. By distinguishing between gang involvement and broader criminal involvement and examining both the additive and unique effects of these influences, we enhance existing theoretical frameworks—particularly learning theories. This nuanced approach disentangles the ways these distinct social networks operate, challenging the tendency of criminological theories to treat deviant influences as a monolithic category. Our findings highlight the importance of accounting for these differences to refine our understanding of deviant socialization and its long-term impacts.

Furthermore, this study provides practical insights that can guide policymakers and practitioners in crafting interventions tailored to address the complex social dynamics that contribute to delinquency and gang membership. A clearer comprehension of the roles and effects of various social relationships will support strategic targeting of interventions, potentially making them more effective. For instance, identifying significant criminal influences within a youth’s network allows for the development of early intervention programs that focus on providing at-risk youths with counseling, educational support, and positive social activities that divert their trajectories away from criminal and gang involvement.

Using administrative data from Multnomah County, Oregon, which tracks youth involved in the criminal justice system from 2002 to 2017, we examine how exposure to gang-involved and criminally involved friends and family members shapes subsequent gang involvement and adult correctional supervision, which we define as incarceration or probation for the purposes of this study. Our analysis incorporates controls across various domains (school, family, and individual) that previous research has linked to gang affiliation and criminal activity (Thornberry et al., 2003). Through this investigation, we reveal the specific roles of various deviant social relationships and their cumulative impact in shaping youth gang involvement and adult criminal behavior.

Family, Peers, and the Social Context of Criminal Behavior and Gang Membership

The impact of deviant associates on individual offending is a cornerstone of criminological research, extensively explored through Sutherland’s differential association theory (1948). According to this theory, individuals learn criminal behavior within intimate personal groups where individuals adopt attitudes and values towards crime influenced by the frequency, duration, priority, and intensity of their interactions. This theory underscores that the nature of exposure to criminal behaviors varies by source—be it parents, peers, siblings, or cousins—and each type of relationship uniquely shapes perceptions of delinquency.

Parents often act as primary role models; their involvement in crime can normalize such behaviors, seeding an early acceptance of delinquency. Peers may exert influence through mechanisms like peer pressure or group conformity, reinforcing or challenging beliefs based on group norms (Warr 2002). Siblings may integrate both peer and parental influences, playing a distinct role within the family dynamic in molding behaviors (McHale, Updegraff and Whitman 2012). In some communities cousins play a role much like that of siblings.

Parents

Research demonstrating crime can be transmitted from parent to child is extensive. For example, in their classic study, Glueck and Glueck (1950) reported parents of delinquent youth were more likely to have criminal records than parents of non-delinquent children. Since then, robust literature has found correlations between parental system involvement and child delinquency or justice system involvement (for reviews see Besemer et al., 2017; Wildeman, 2020). More recently, studies have begun to investigate the impact of parental incarceration on youth during the transition to adulthood, with similar results. For example, using data from the National Longitudinal Study of Adolescent Health and propensity score matching, Mears and Siennick (2016) found positive correlations with parental incarceration during the transition to adulthood across life domains, including criminal behavior and mental health.

The research on the direct impacts of parents’ gang status on their children’s likelihood to join gangs are mixed. Five studies find an effect. Two of these found that father gang involvement predicts son gang involvement (Augustyn et al., 2017; Kisser & Pyrooz 2009). Similarly in a study focused on intergenerational continuity in gang membership, Augustyn et al. (2017) found that a mother’s gang involvement predicted daughter gang involvement. Ethnographic research undertaken in Los Angeles has also suggested that parental gang involvement contributes to youth gang engagement (Moore, 1992; Vigil, 1988). Decker and Van Winkle (1996) is the sole dissenting study; it finds that peers and siblings are more salient factors than parents for gang involvement.

Peers

Of the three domains this research study investigates, peer relationships are the most commonly studied in relation to contact with the criminal justice system and gang involvement. During the course of adolescent development and transition into emerging adulthood, adolescents begin to transfer their dependency from their parents onto peers (Steinberg & Silverberg, 1986), leading to greater susceptibility to peer pressure. This process during adolescence has been a focus of research and is the foundation for many of the established criminological theories and youth offending research (Akers & Jensen, 2017; Miller, 2010). Social learning theory (Akers, 2017) and differential association theory (Sutherland, 1947) both suggest that crime is learned from deviant peers. In fact, Warr (2002) posited that there is no better predictor of juvenile criminal behaviors than a person’s number of delinquent friends. However, this impact is not stable over time as most adolescents both desist from crime and become more resistant to peer influence as a part of the natural maturation process. This suggests that decreased susceptibility to the influence of peers, at least in part, drives the reduction in criminal activity with age. Leverso et al. (2015) confirmed the waning of peer influence, as they found that the impact of peers on future offending was less prominent in predicting offending as the youth aged.

For gangs, a robust set of studies has found peer influence, both through deviant peers and gang-involved peers, is associated with gang involvement. The association between deviant peers and gang involvement has been investigated using data on middle school students from the Gang Resistance Education and Training (G.R.E.A.T; Esbensen & Deschenes, 1998; Esbensen et al., 2001), data on at-risk youth from the Seattle Development Project (Hill et al., 1999), the Rochester Youth survey (Thornberry et al., 2003), and the Pittsburgh Youth Study (Lahey et al., 1999). In a review of studies on risk factors for gang membership, Klein and Maxson (2006) found research “consistently supported” the influence of peers on gang involvement. Similarly, studies have also found that having gang involved peers predicts gang membership. For example, Winfree, Backstrom, and May (1994), using data on 9th grade public school students in New Mexico, and Winfree, May, and Backstrom (1994) using data from the New Mexico Youth Authority, found a correlation between connections with gang involved peers and gang involvement. Kissner and Pyrooz (2009) reported similar findings with data from jail inmates in California.

Siblings and Cousins

The influence of siblings’ criminal behavior on youth criminal or gang involvement is less frequently examined but important, and few studies acknowledge that in many poor communities the distinction between siblings and cousins is blurred.

Rowe and Farrington (1997) highlighted that law-abiding siblings can deter criminal behaviors in their brothers or sisters, whereas those engaged in criminal activities tend to propagate such norms among their siblings. Similarly, Schwartz and Beaver (2014) found significant correlations where the repeated criminal justice involvement of one sibling substantially increased the likelihood of similar outcomes for the other siblings, with odds ratios ranging from 2.16 to 7.80. However, a comprehensive meta-analysis by Maneiro et al. (2022), which reviewed 24 studies on the influence of deviant siblings, reported a generally smaller effect size for the transmission of criminal behavior. As well, sibling attributes like birth order and gender did not significantly affect the transmission magnitude, indicating that younger siblings could influence older ones and that sisters are as likely to transmit criminal behaviors as brothers.

Similarly, both qualitative and quantitative studies have found that siblings often provide gateways into gang membership. Hill et al. (1999) found that a sibling’s deviant behavior significantly correlates with a youth’s likelihood of joining a gang. Decker and VanWinkle (1996) found that 45% of their sample of gang members had gang-involved siblings. Kissner and Pyrooz (2009) discovered that having an older sibling in a gang was a more potent predictor of one’s own gang involvement than were gang-affiliated friends or fathers. Most recently, Hashimi et al. (2021) reinforced these findings by demonstrating that sibling gang involvement is closely associated with respondent gang involvement, highlighting the critical yet understudied role siblings play in the gang-joining process.

Exposure to Gangs and Criminality: Peers, Siblings, Cousins, and Parents

The three domains of peers, parents, and siblings/cousins are critical for understanding gang involvement and adult correctional supervision. However, these influences are rarely considered collectively. Differential association theory posits that individuals learn criminal behavior through frequent and recurring associations that foster attitudes favorable to deviance (Sutherland, 1947). While much of the scholarship has focused on peers, this theory also explains how family dynamics shape the socialization of deviance. Exposure across multiple relational domains amplifies the normalization of deviant behavior by reinforcing attitudes and behaviors across different social contexts.

To fully understand these dynamics, it is essential to consider both the cumulative and individual effects of these social relationships. Exposure to deviant relationships across multiple domains likely has an additive impact on behavior, as each relationship type uniquely reinforces deviance. This layered exposure reduces competing prosocial messages and strengthens pro-deviant attitudes, creating overlapping pathways that heighten the likelihood of gang involvement and criminal behavior. Empirical evidence underscores the need to examine these cumulative effects. Lornado et al. (2008) found that exposure to deviant social relationships, including those with peers and parents, was associated with higher levels of self-reported delinquency, illustrating the additive effects of influence across domains. Thornberry et al. (2003, p. 15) observed, “delinquent behavior is expected when the individual is enmeshed in some, especially many networks, that allow or consider such behavior.” Building on this, our study seeks to disentangle how exposure across these domains influences youth gang membership and adult correctional supervision, defined as incarceration or probation.

In addition to cumulative effects, it is equally important to examine the distinct, individual impacts of specific relationship domains. Peer influence is often more predictive of immediate behaviors like gang membership because of its situational intensity and reinforcement of group norms. Parental influence, in contrast, has a stronger effect on long-term outcomes such as criminal justice involvement, reflecting their foundational role in shaping life trajectories. Siblings and cousins may influence both short- and long-term behaviors by mediating the transmission of deviant norms between family and peer groups. These differences arise from the distinct roles each domain plays in socialization, shaping behavior in both immediate and enduring ways.

Existing research supports these dynamics but often examines relationships in isolation, overlooking their cumulative and overlapping effects. For example, Kissner and Pyrooz (2009) highlighted the role of gang-involved friends in promoting gang affiliation. Parental criminality has been linked to deviance, as shown by Beijers et al. (2017), who found that parental criminality independently predicted offspring criminality. Similarly, Hashimi et al. (2021) reported that gang-involved siblings strongly impacted youth gang membership, even after controlling for other variables.

By examining exposure across these domains, this study provides a more nuanced understanding of how relational dynamics shape both youth gang involvement and adult criminal justice outcomes. It extends differential association theory by highlighting both the cumulative and domain-specific impacts of deviant socialization.

Current Study

This study assesses how exposure to deviant relationships across multiple domains—peers, parents, and siblings/cousins—shapes two critical outcomes: youth gang membership and adult correctional supervision. Guided by differential association theory, we explore how cumulative and domain-specific exposures contribute to these outcomes.

The following research questions and hypotheses guide our investigation:

Research Question 1:

Does exposure to multiple deviant relationships, including both gang-involved and criminally involved associations, increase the likelihood of youth gang membership and adult correctional supervision?

  • Hypothesis 1a: After controlling for relevant covariates, the likelihood of gang membership will be higher in youths with greater exposure to deviant social circles.

  • Hypothesis 1b: After controlling for relevant covariates, the likelihood of adult correctional supervision will be higher in youths with greater exposure to deviant social circles.

Research Question 2:

Which relationships, if deviant, have the greatest influence on our two outcomes, and does the distinction between immediate outcomes (gang involvement) and more long-term outcomes (adult correctional supervision) predict these relationships?

  • Hypothesis 2a: After controlling for relevant covariates, peer influence will exert greater influence than parent or sibling/cousin influence on gang membership.

  • Hypothesis 2b: After controlling for relevant covariates, parental influence will exert a greater influence than peer or sibling/cousin influence on adult correctional supervision.

Our approach allows us to measure the strength and specificity of the impact of youths’ exposure to different social circles on the outcomes under investigation. By highlighting the unique contributions of different social relationships and their cumulative effects, this study advances the theoretical and empirical understanding of how social environments shape deviant outcomes. These insights are important for developing targeted interventions and understanding the complex dynamics that lead youths toward delinquent behaviors and gang affiliation.

Methods

Sample

This project utilizes administrative data for self-identified male youth who received their first criminal referral (arrest), underwent a risk assessment associated with the referral, and had the referral adjudicated between 2002 and 2017. Among the 2,210 adolescents, we eliminated 20 who were not yet 18 when the dataset was assembled, as this meant their data could not include adult outcomes. An additional 382 youth were excluded because they did not name both their parents, making it impossible to determine whether they had a parent involved in gangs or the criminal justice system (369 lacked father information, 13 lacked mother information). The final analytic sample consists of 1,808 male youth. We discuss the impacts of missing data in the results section.

Procedure

Data was collected across four administrative databases: Oregon State Corrections Information System (CIS), Multnomah County Criminal Records Information Management and Exchange System (CRIMES), Oregon County Juvenile Justice Information System (JJIS) which includes the Juvenile Crime Prevention (JCP) Risk Assessment Tool given to all youth at intake, and Federal Probation and Pretrial Services Automated Case Tracking System (PACTS). Table 1 lists all variables that were created from each data source.

Table 1:

Summary of Data Sources and Variables

Data Source Description Variables Created
CIS (Corrections Information System) Contains records for Oregon state adult prison, probation, and parole. Adult criminal outcomes (prison or probation).
CRIMES (Criminal Records Information Management and Exchange System) Managed by the Multnomah County district attorney, tracks criminal investigations. Gang membership status of youth, their family and peers. Adult arrests of family and peers.
JJIS (Juvenile Justice Information System) Statewide tool capturing information about youth in juvenile justice systems. Youth, their family, and peers’ juvenile justice involvement. All control variables were accessed via the Juvenile Crime Prevention Tool within the JJIS.
PACTS (Probation and Pretrial Services Automated Case Tracking System) Manages risk and needs assessments, case management for adults under federal supervision. Federal probation information.
Open in a new tab

All data were collected in 2019 and 2021. In 2019 information on youth with a criminal referral who would be over the age of 18 in 2021 was pooled together from several of the databases. Then in 2021 adult incarceration and probation information was appended to the dataset. The end result was a pooled, person-level dataset.

Outcome variables

The outcomes for this study, detailed in Table 2, include gang membership and adult correctional supervision. Gang membership was determined using data from the district attorney’s office, which contains images that may display gang identifiers or explicit mentions in case notes of gang-related activities, such as “this youth is part of a gang.” This variable is coded as 1 for identified gang members and 0 otherwise, with 389 youths (22% of the sample) identified as gang members.

Table 2:

Variable Descriptions and Descriptive Statistics (N = 1,808)

Variable Name Description Mean (%) SD
Outcomes
Documented gang activity Youth documented as a gang member 22.00
Escalation into Adult System Youth arrested as adult 47.00
Level of Exposure
Gang Exposure Index (0–3) Exposure to gang and criminal involved family and peers. 0.37 0.67
Exposure to criminal legal circles (0–4) Exposure to criminal involved family and peers. 1.38 1.04
Type of Exposure
Father’s criminal justice involvement Father had contact with criminal justice system 19.00
Father criminal justice & gang involved Father documented as gang member 9.00
Mother’s criminal justice involvement Mother had documented contact with the criminal justice system 15.00
Same generation criminal justice involved Sibling or cousin had contact with the criminal justice system 27.00
Same generation criminal justice & gang involved Sibling or cousin documented as a gang member 8.00
Peer criminal justice involved Peers had contact with the criminal justice system 41.00
Peer criminal justice & gang involved Peers were documented as a gang member 19.00
School
Significant school attachment Youth had significant attachments, beliefs, commitment, and/or involvement in school 48.00
Total school disruptions Number of school disruptions recorded by system personnel 0.50 0.80
Family
Poor family supervision Family is unaware where the youth goes and with whom 44.00
Dependency referral Youth had been involved in a child dependency case 23.00
Individual Characteristics
Race/Ethnicity
Youth identifies as White 42.00
Youth identifies as African American 34.00
Youth identifies as Hispanic 18.00
Youth identifies as Native American or Other 6.00
Impulsive behavior Youth has exhibited a pattern of impulsivity combined with aggressive behavior 29.00
Age of first criminal referral Age at first criminal referral 14.45 1.63
Total juvenile criminal referrals Number of criminal referrals before turning 18 5.05 4.30
Total days spent in detention Number of days a youth spent in detention within the juvenile legal system 44.93 67.26
Weapons crime charge ever Youth had a weapon charge before turning 18 years old 20.00
Years Since 18th Birthday Number of years since a youth’s 18th birthday 10.77 3.72
Open in a new tab

The second outcome, adult correctional supervision, is a dichotomous variable measuring whether a youth was subsequently sentenced to Oregon state prison, state probation, or federal probation (1=yes, 0 otherwise).1 Data was sourced from CIS, and nearly 50% of our sample (n = 851) had adult criminal justice contact.

Family influences

We utilized data from Oregon’s JJIS and CRIMES databases to examine the influence of known associates—parents, siblings/cousins, and peers—on youth outcomes. Known associates were identified, and their criminal or gang involvement was assessed through database queries. To operationalize this, we focused on criminal convictions as indicators of criminal justice system contact. Because our dataset is based on administrative criminal justice records, it inherently captures only those individuals who have had documented interactions with the criminal justice system. As a result, this dataset does not include gang-involved individuals who have never interacted with the criminal justice system. We therefore created a three-tiered categorical variable for each known associate: “0” for criminal justice or gang involvement, “1” for criminal justice contact but no gang involvement, and “2” for criminal justice contact and gang involvement. For analysis, we recoded these categorical variables into dichotomous variables for each family unit. Criminal justice contact was recoded as no (=0) and yes (=1). Gang involvement, originally coded as “2,” was recoded as no (=0) and yes (=1) for consistency in the dichotomous framework.

Family influence variables included father’s, mother’s, and sibling/cousin criminal justice contact (adult or juvenile convictions) and gang membership (fathers and siblings/cousins only). Due to the low proportion of women in gangs in Portland (approximately 2%), mothers’ gang membership was excluded.

Approximately 20% of the sample (n=334) had fathers with at least one conviction, and 164 had fathers involved in gangs. Maternal criminal justice contact was present for 275 youth. Missing data regarding fathers were addressed by excluding youth with unidentifiable fathers due to missing birthdates. For youth with absent or unknown fathers (n=372), fathers were coded as uninvolved in both the criminal justice system and gangs, as absent fathers cannot exert direct influence.2

Siblings and cousins were grouped due to insights from preliminary fieldwork, where gang members described cousins as playing roles similar to siblings within their households. In our sample, 27% (n=488) had siblings/cousins with criminal justice contact, and 8% (n=142) had siblings/cousins involved in gangs.

Peer influences

In this study, a peer is defined as a friend or neighbor who has a significant relationship with the peer whose age is within 6 years of the youth’s. In our research site, neighbors tend to have close ties, with frequent group activities and strong connections among children of similar age living nearby. Generally, a criminal justice system worker working with the youth and their family determined whether the association with the neighbor was significant. The counselor used a dropdown list of possible relationships where only one could be chosen, and “neighbor” was often used to indicate “a friend who lives next door.”

Much as with familial influences we used dichotomous variables (no=0) to measure peer criminal justice contact and peer gang involvement. Our sample included 746 individuals (41%) who had a connection with a non-kin peer who had a criminal conviction as an adult or juvenile. It included 355 individuals (19%) with an association to a non-kin peer who was a gang member. Our coding decisions were robust to sensitivity analyses based on the number of deviant peer associations.3

Deviant exposure index

To understand how the level of exposure within gang circles might affect outcomes, an additive exposure index was created by combining peer gang involvement (1/0), same-generation kin gang and criminal justice involvement (1/0), and father gang and criminal justice involvement (1/0). The resulting variable ranges from 0 (no exposure) to 3 (high exposure), with a mean of 0.37 and a standard deviation of 0.67.

Similarly, for exposure within a criminal circle, an additive exposure index was created by combining peer criminal legal involvement (1/0), same-generation criminal legal involvement (1/0), father criminal legal involvement (1/0), and mother criminal legal involvement (1/0). This variable ranges from 0 (no criminal legal exposure) to 4 (high criminal legal exposure), with a mean of 1.38 and a standard deviation of 1.04.

Domain-based control variables

We include a number of theoretically relevant control variables that are based on domains associated with gang and criminal justice system involvement. These domains, identified as risk factors in established literature, encompass school, family, and individual characteristics (Farrington, 1998, 2000). The variables were accessed via the JJIS and derived from the JCP Risk Assessment Tool. Criminal justice system workers—including probation officers, counselors, and social workers—administer the tool to youth in the state of Oregon upon their referral to a juvenile department, conducting interviews with the youth, family, and other sources and reviewing official school, court, and other records. To ensure consistency and accuracy in our analysis, we utilize data from the first assessment administered to each youth, addressing the possibility that some may enter the system multiple times.

School and Family Control Variables

Using the results of the JCP assessments, we include variables related to school and family domains as both are significantly linked to both gang membership and delinquency (Cesar, Walker, & Fernandez, 2024; Esbensen et al., 2009; Sampson & Laub, 1993; Thornberry et al., 2003). The variable significant school attachment is dichotomous, with 1 indicating that the criminal justice system worker assessed the youth as having a commitment to school, an attachment to school, and beliefs that motivated them to succeed academically; 48% (874) of youths were assessed as having significant school attachment. Total school disruptions is a count of incidents such as expulsions, withdrawals, or dropouts. There were an average of 0.5 disruptions per youth (standard deviation = 0.80) across 1,808 cases.

A 1 for poor family supervision indicates that the interviewer assessed the youth as lacking adequate supervision, evidenced by a family’s lack of awareness of the youth’s whereabouts, companions, and activities; 44% (793) of the sample had experienced poor family supervision. Finally, a 1 for child dependency indicates the youth had been involved in a dependency case, typically initiated by the court when a child is removed from the home due to neglect or abuse; this applied to 23% (407) of youths in the sample.

Individual-level control variables

Previous history with the criminal justice system, demographic variables such as race and age (Esbensen & Carson 2012; Pyrooz & Sweeten 2015), and impulsiveness (Pyrooz, Melde, Coffman, & Meldrum, 2021; Pratt & Cullen, 2000) are associated with gang membership and criminal justice involvement. Youths’ race and ethnicity data were collected from the state juvenile database, which categorizes individuals as Asian, Black, White, and Native American. From these categories, we derived four race/ethnicity variables: Non-Hispanic White (N = 756), Hispanic (N = 328), African American (N = 614), and Other (N = 110). Each is treated as a dichotomous variable where “1” represents presence in the category and “0” represents absence. Impulsive behavior is measured as a dichotomous variable where 29% (529 youths) exhibited patterns of impulsivity combined with aggressive behavior. Age at first criminal referral is a continuous variable with an average age of 14.45 years (SD = 1.63). Total juvenile criminal referrals counts the number of criminal referrals received before turning 18, with an average of 5.05 (SD = 4.30). Total days spent in detention records the number of days youths spent in detention, averaging 44.93 days (SD = 67.26). The assessment of weapons crime charge ever shows that 20% (368 youths) had a weapons charge before turning 18. Lastly, years since 18th birthday is a continuous variable capturing the number of years since a youth’s 18th birthday at the time of data gathering, with an average of 10.77 years (SD = 3.72).

Analytic strategy

We began by conducting a missing data/case analysis to compare individuals included in the study to those excluded. This step allowed us to identify patterns or inconsistencies in data exclusion and assess the degree to which our analytical sample represents the general population from the administrative database. While our findings are not intended to be generalizable beyond this specific context, this analysis enhances our understanding of the sample’s characteristics and ensures the robustness of our conclusions within this context.

Following this assessment, we compute bivariate relationships for our two primary outcomes—gang membership and adult correctional supervision —across all family variables. These bivariate analyses provide an unadjusted view of the associations, serving as a foundation for interpreting the adjusted relationships presented in the multivariable models.

We then employ four multivariable logistic regression models (see Table 4) to investigate the influence of familial, peer, and parental involvement in criminal and gang activities on these key outcomes. The first pair of multivariable logistic regression models assesses the overall effect of the Criminal Exposure Index and the Gang Exposure Index—combining influences from family, peers, and parents—while controlling for theoretically relevant risk factors. These models establish a baseline understanding of how exposure to deviant circles correlates with gang involvement and progression into the adult system.

Table 4.

Multivariable logistic regression models predicting documented gang activity and adult correctional supervision N = 1808

A. Documented Gang Activity B. Adult Correctional Supervision C. Documented Gang Activity D. Adult Correctional Supervision
FOCAL VARIABLES OR SE OR SE OR SE OR SE
Exposure Index
Eposure to gang circles 2*** 0.09 1 0.11
Expousre to criminal legal circles 1.19* 0.11 1.28*** 0.07
Type of Exposure
Father criminal justice involved 1.57* 0.2 1.24 0.15
Father criminal justice & gang involved 2.04** 0.23 1.47† 0.22
Mother criminal justice involved 1.27 0.2 1.21 0.16
Same generation criminal justice system 1.02 0.16 1.22 0.13
Same generation criminal justice & gang involved 2.12*** 0.23 1.52† 0.23
Peers criminal justice involved 1.15 0.18 1.38* 0.13
Peers criminal justice & gang involved 2.79*** 0.20 1.11 0.18
CONTROL VARIABLES
Documented Gang Invovlement 2.04*** 0.16 2.06*** 0.16
Years Since 18th Birthday 1.2*** 0.02 1.2*** 0.02
School
Significant school attachment 0.96 0.15 0.99 0.12 0.96 0.15 0.99 0.12
Total school disruptions 1.05 0.09 1.1 0.07 1.04 0.09 1.1 0.08
Family
Poor family supervision 1.06 0.15 1.04 0.12 1.06 0.15 1.04 0.12
Dependency referral ever (Yes) 0.52*** 0.18 0.99 0.14 0.52*** 0.19 1 0.14
Individual Characteristics
Race/ethnicity (White)
African American 6.24*** 0.22 0.83 0.15 6.55*** 0.23 0.83 0.15
Hispanic 6.61*** 0.24 0.95 0.16 6.96*** 0.24 0.95 0.16
Other race 3.51*** 0.35 0.55* 0.25 3.52*** 0.35 0.55* 0.25
Impulsive behavior 1.18 0.16 0.94 0.13 1.19 0.16 0.93 0.13
Age of first criminal referral 1.1 0.05 1.11** 0.04 1.1 0.05 1.12** 0.04
Total juvenile criminal referrals 1.22 0.15 2.79*** 0.12 1.2 0.16 2.82*** 0.13
Weapons crime charge ever (Yes) 4*** 0.16 0.98 0.15 4*** 0.16 0.98 0.15
Total days spent in detention 1.3*** 0.05 1.24*** 0.04 1.29*** 0.05 1.25*** 0.04
Intercept 0.00228*** 0.91 0.00188*** 0.70 0.00229*** 0.92 0.00174*** 0.70
χ2 600.1*** 517.1*** 605.4*** 520.20***
Open in a new tab

Notes: OR = Odds Ratio; SE=Standard Error for ln(OR) .

†

p < .10,

*

p≤.05

**

p≤.01

***

p≤.001

In the subsequent pair of models, we explore which specific types of relationships— parents, siblings/cousins, and peers—have the greatest impact on the likelihood of being system-identified as a gang member and escalating into the adult system. These models include the same control variables but treat sibling/cousin, peer, and parental influences as separate predictors, allowing us to pinpoint the relative impact of each type of deviant network exposure. This structured approach aims to elucidate both the level and the specific types of deviant network exposure that most significantly affect gang involvement and subsequent contact with the criminal justice system in adulthood.

Results

Excluded observations and bivariate associations

We examined the distribution of missing data on the key variables in our analysis by comparing the distribution of all variables in the study for included observations (n = 1,808) against excluded observations (n = 382). Those excluded are significantly less likely to be documented as gang members (15% vs. 22% respectively), less likely to have adult correctional supervision (34% to 44% respectively) and less likely to have friends or family documented as gang members or adult correctional supervision. They were also less likely to have criminal referrals, spent less time in detention, and were older when they got their first referral. We found no substantial differences between included and excluded observations on school attachment, impulsiveness, race/ethnicity, or having a weapons referral. Taken together, these results indicate that those included in the analytical sample could be assumed to be at higher risk of gang and criminal involvement than those who had missing data, but there were no other differences.

Next, we computed bivariate relationships between the exposure variables and the two outcomes to determine if there were any significant associations before adding controls (see Table 3). Exposure to criminal and gang behaviors within social circles significantly influences the odds of individual gang involvement. General exposure to criminal justice involvement increases the odds of gang involvement by 293% (OR = 3.93, p < .001), while exposure to gang involvement, inherently linked to criminal justice contact, also increases the odds (OR = 1.85, p < .001). Having a father with criminal justice contact raises the odds of gang involvement by 31% (OR = 1.31, p < .10), and if the father is both criminal justice and gang-involved, the odds increase by over five times (OR = 5.03, p < .001). A mother’s involvement in the criminal justice system increases the odds of gang involvement by 52% (OR = 1.52, p < .01). For same-generation siblings/cousins, criminal justice contact increases the odds of gang involvement by 74% (OR = 1.74, p < .01), while gang involvement among these family members, which also includes criminal justice contact, nearly triples the odds (OR = 6.79, p < .001). A peer’s criminal justice contact increases the odds of gang involvement by 44% (OR = 1.44, p < .05), while peer gang involvement, which inherently includes criminal justice contact, dramatically increases the odds—by more than eight times (OR = 8.26, p < .001).

Table 3.

Bivariate logistic regression models predicting gang membership and adult correctional supervision N = 1808

Independent variable Documented Gang Activity Adult Correctional Supervision
Exposure Index OR SE OR SE
Exposure to gang circles 1.85 *** 0.06 1.66 *** 0.05
Exposure to criminal legal circles 3.93 *** 0.09 1.86 *** 0.08
Type of Exposure
Father criminal justice system involved 1.31 † 0.15 1.54 ** 0.12
Father criminal justice & gang involved 5.03 *** 0.17 2.41 *** 0.17
Mother criminal justice system involved 1.52 ** 0.15 1.45 ** 0.13
Same generation criminal justice system 1.74 ** 0.08 1.88 ** 0.11
Same generation criminal justice & gang involved 6.79 *** 0.19 3.04 *** 0.19
Peers criminal justice system involved 1.44 * 0.15 1.91 * 0.11
Peers criminal justice & gang involved 8.26 *** 0.16 3.27 *** 0.14
Open in a new tab

Notes: OR = Odds Ratio; SE=Standard Error for ln(OR);

†

p < .10

*

p≤.05

**

p≤.01

***

p≤.001

Exposure to criminal and gang behaviors significantly influences adult correctional supervision. Each additional level of exposure increases the odds of adult correctional supervision by 86% (OR = 1.86, p < .001), while each additional level of gang exposure—closely tied to criminal justice contact—is associated with a 66% increase in the odds (OR = 1.66, p < .001). Similarly, having a father with criminal justice contact increases the odds by 54% (OR = 1.54, p < .01). However, if the father is a gang member and involved in the criminal justice system, the odds of the youth being arrested as an adult more than double (OR = 2.41, p < .001). For mothers involved in the criminal justice system, the odds of adult correctional supervision increase by 45% (OR = 1.45, p < .01). Criminal justice contact of a sibling or cousin nearly doubles the odds of adult correctional supervision (OR = 1.88, p < .01), and having siblings or cousins who are both gang-involved and have had criminal justice contact triples the odds (OR = 3.04, p < .001). Finally having associations with peers with criminal justice contact increases the odds of adult correctional supervision by 91% (OR = 1.91, p < .05); if a peer is also gang involved, odds are more than triple (OR = 3.27, p < .001). Taken together, the bivariate analyses show strong associations between family variables and the outcomes of gang membership and escalation to the adult criminal system.

Level of Exposure

Table 4 presents multivariate logistic regression models of gang membership and adult correctional supervision. We hypothesized that the higher the degree of gang and criminal legal exposure, the greater the likelihood of gang membership and adult correctional supervision. After controlling for domain-specific variables related to school, family, and individual characteristics, we find that gang and criminal justice exposure impact gang involvement (Model A) and that criminal justice exposure influences further contact with adult correctional supervision (Model B). For gang involvement, knowing at least one system-identified gang-involved individual in a given category (peer, sibling/cousin, or father) doubles the odds of being identified as a gang member (OR = 2.00, p < .001). However, gang exposure does not have a significant impact on the odds of escalating into the adult system. On the other hand, each additional connection to someone with a criminal record increases the odds of adult incarceration or probation by 28% (OR = 1.28, p < .001), and gang membership increases the odds by 19% (OR = 1.19, p < .05). These results largely support Hypotheses 1a and 1b.

Domain-specific control variables revealed associations mostly between the family and individual domains on the outcomes. In the school and family domains, school disruptions, school supervision, and family supervision were not significantly related to either outcome. However, the results concerning dependencies (court intervention in the family) contradicted the findings of established literature. Research indicates that early involvement with the system increases the likelihood of a youth’s gang involvement (Kerig & Mendez, 2022; Kubik et al., 2019). Our data revealed that, when holding all other variables constant, those who experienced a dependency were 48% less likely (OR = 0.52, p > .000) to be documented as a gang member than those who did not. Given this result we double-checked the coding of the variable and it appears accurate. Further research is clearly needed to determine if that early contact is a proxy for a successful intervention or a proxy for an omitted variable.

Turning to individual-level characteristics we see that compared to identifying as White, identifying as African American increased the odds of being identified as a gang member by 524% (OR = 6.24, p > .001). Being Hispanic increased the odds by 561% (OR = 6.61, p > .001) and identifying as other race increased the odds by 251% (OR = 3.51, p > .001). However being identified as other race decreased the odds of escalation to the adult system (OR = 0.55, p > .05) and no significant differences were found with African Americans and Hispanics for escalation to the adult system. Days spent in detention were all positively and significantly associated with both gang affiliation and escalation to the adult system, weapons charges were only significantly associated with gang affiliation, and age at first referral was only associated with escalation to the adult system (OR = 1.11, p > .05). Impulsiveness was not significantly associated with gang affiliation or with escalation to the adult system. Finally, exposure time was significantly related to adult contact with the justice system. Unsurprisingly more exposure time increased the probability of later adult contact with the justice system.

Exposure Type

The data reveal that the exposure to criminal justice and gang members in family and peer systems significantly affects youths’ outcomes. Controlling for all variables in the model, we found that fathers with both gang and criminal justice involvement (OR = 2.04, p < .001), same-generation kin with both criminal justice and gang involvement (OR = 2.12, p < .001), and non-kin peers with both criminal justice contact and gang involvement (OR = 2.79, p < .001) were all positively associated with a youth being identified as gang-involved. Among these factors, peers exerted the strongest influence, followed by same-generation kin and then fathers. These results support our hypothesis (2a) that peers have the strongest impact due to the direct and daily interactions they are more likely to have with the youth. Additionally, fathers involved only in the criminal legal system were also positively associated with youth gang involvement, though with a smaller effect size (OR = 1.57, p < .001).

In terms of adult correctional supervision, the data indicate that peers involved with the criminal justice system significantly increased the likelihood of a youth’s escalation into the adult system (OR = 1.38, p < .001). Additionally, both father’s gang and criminal justice involvement (OR = 1.47, p < .01) and the involvement of same-generation family members (OR = 1.52, p <.01) showed marginal significance. While these findings do not fully support Hypothesis 2b, they provide partial support, as the father’s involvement in the criminal justice system showed a slightly weaker effect compared to that of same-generation family members.

Discussion

Differential association suggests that associations with deviant others impact deviant behavior. Previous research has, for the most part, supported this supposition, finding deviant parents, siblings, and peers influence the behaviors of youth (Maneiro, et al., 2022; Miller, 2010; Rowe & Farrington, 1997; Wildeman, 2020). A less investigated parallel track of research has found gang-involved parents, siblings/cousins, and peers influence youth gang involvement (Augustyn et al., 2017; Hashimi et al., 2021; Kissner & Pyrooz, 2009). The present study brings these parallel lines of research into conversation with one another and develops and tests four hypotheses related to the impact of deviant influences on youth deviant behavior.

Our first set of hypotheses focused on the impact of exposure to gang-involved and criminal justice-involved family and friends. Based on the premise that cumulative and varied exposure to deviant behaviors across multiple relational domains intensifies the normalization and learning of criminal behaviors, we hypothesized that youth with higher levels of exposure to deviant social circles would have significantly greater odds of gang membership (1A) and adult correctional supervision (1b). Our hypotheses were mostly supported. Our exposure index showed that associations with criminal justice-involved and criminal justice-and-gang-involved social circles were related to a youth joining a gang and being under correctional supervision as an adult. However, only criminal justice exposure impacted being under correctional supervision as an adult. This finding suggests that while both types of exposure involve interactions with the criminal justice system, the influence of purely criminal justice-involved peers and family members may have a more direct and lasting impact on future criminal justice contact.

Our second set of hypotheses focused on the individual impacts of parents, siblings/cousins, and peers on gang membership and later contact with the criminal justice system. Hypothesis 2a posited that peer influence would be more predictive of gang membership due to the immediate behaviors and direct, daily interactions that reinforce group norms. This hypothesis was supported, as gang-involved peers with criminal justice contact exerted the strongest effects on gang membership.

Hypothesis 2b related to later contact with the criminal justice system, where we hypothesized that fathers would have the largest impact, given their foundational role in shaping behavior over time. While fathers did exert a strong influence, siblings and first-generation family members had a slightly stronger effect. This suggests that, although fathers are important for shaping early behavior, the influence of siblings and first-generation family members becomes more significant during certain life stages—particularly adolescence and early adulthood, when peer-like relationships with siblings and cousins are more salient. These findings emphasize the need to consider the broader familial context, beyond parental influence, when examining long-term behavioral outcomes.

This study’s findings support several important conclusions. In the context of gang research, when accounting for exposure to both criminal justice-involved networks and those involved in both criminal justice and gangs within the same model, the data reveal a stronger association between exposure to gang circles and subsequent gang involvement. This suggests that family and friends involved in gangs exert a greater influence on youth gang membership than those solely involved in the criminal justice system, indicating that gangs and their unique group processes (McGloin & Collins, 2015) have a greater impact on outcomes than criminal justice contact alone. These findings contribute to a growing body of research demonstrating how gang-involved family members influence gang involvement (Augustyn et al., 2017; Hashimi et al., 2021; Kissner & Pyrooz, 2009).

Additionally, this study builds on the concept of embeddedness in gang literature (Pyrooz, Sweeten, & Piquero, 2013; Sweeten, Pyrooz, & Piquero, 2012) by examining the role of exposure as a dimension of embeddedness. While embeddedness traditionally focuses on the depth of immersion within deviant networks, our study emphasizes how the breadth of relationships—including less intense or enduring connections—shapes behavior. This approach captures how deviant family members, peers, or acquaintances may influence individuals even when these relationships are peripheral to their primary social network. By incorporating both the breadth and depth of deviant relationships, this integrated perspective offers a more complete understanding of how social environments shape individuals.

For studies of criminal behavior, this research provides evidence that exposure to criminal justice-involved circles, rather than those involved in both criminal justice and gangs, increases the likelihood of adult contact with the criminal justice system. This extends research on the transition to adulthood (Mears & Siennick, 2016) by incorporating a more comprehensive set of family variables, rather than focusing on a single point of contact (Noel & Hoeben, 2022; Turney & Lanuza, 2017), and it demonstrates the importance of exposure in this context (Lornado et al., 2008). These findings also underscore the importance of distinguishing between gang-related and criminal justice-related exposures in theory and practice, as these influences are not homogenous and should be treated separately in theoretical models. Additionally, consistent with Warr (2002), the influence of peers remains one of the strongest predictors of deviant behavior, highlighting the critical role peer relationships play in understanding both gang involvement and later-life contact with the criminal justice system (McGloin & Thomas, 2019).

Finally, this research has important implications for policy and evidence-based intervention programming. Much of the existing gang intervention efforts have focused on educating middle school youth about the consequences of gang involvement (Esbensen et al., 2012, 2013). However, these findings suggest that interventions should be expanded to address both family and peer systems. Family-based interventions, particularly those targeting the entire family unit, may be especially effective in promoting gang disengagement. Additionally, a clearer understanding of the distinct roles and effects of various social relationships enables more strategically targeted interventions. For example, by identifying significant criminal influences within a youth’s network, early intervention programs can be tailored to offer counseling, educational support, and positive social activities to steer at-risk youth away from criminal and gang involvement. Ultimately, interventions should focus on a targeted approach to youth social circles, rather than relying solely on individual therapeutic methods.

Several limitations should be considered when interpreting our results. One key limitation is that our administrative data do not distinguish between gang membership and mere criminal justice contact, making it difficult to isolate gang-specific effects from those attributable to criminal justice involvement. This means we cannot rule out the possibility that observed effects stem from criminal justice contact rather than gang involvement. Additionally, our data lack temporal detail, which is particularly important when analyzing the father variable. Prior research (Leverso & Hess, 2020; Pyrooz & Decker, 2011) reveals variation in connections with gangs among men who, like most fathers of youth old enough to be in our data, have effectively aged out of active involvement. We cannot explore how such variations might influence likelihood of gang involvement among the sons of (former) gang members. Moreover, our data do not provide insight into the level of contact between fathers and youth. While we are confident that fathers marked as involved had some level of contact with the focal youth—whether direct or indirect—the lack of specificity remains a limitation. Future research should aim to collect more nuanced data to better understand the intergenerational transmission of gang involvement

Another limitation is the reliance on administrative records to identify gang-involved individuals. Unlike self-reported data, administrative classification depends on system personnel’s documentation, which could lead to the misidentification of non-gang members as gang affiliates. However, previous studies (Curry, 2000, 2015; Decker & Pyrooz, 2010; Pyrooz et al., 2020) suggest that administrative and self-reported data often overlap significantly in both street and institutional contexts, somewhat mitigating this concern.

The nature of our data collection also presents a limitation, as we had access to sentencing outcomes (prison or probation) for the sample participants, while the data available for their networks predominantly included court outcomes such as convictions. Although this discrepancy exists, the close legal and procedural relationship between convictions and sentencing reduces its impact on our findings.

A further limitation concerns the sequence of events. While we assume that progression to the adult criminal system follows adolescent involvement with the variables and controls, this cannot be definitively confirmed. Although we are confident that a father’s gang history precedes the youth’s, the sequence may vary for other variables. To address this, we conducted sensitivity analyses, restricting peers and first-generation family members to those who had prior gang involvement and were older than the youth. These results aligned with our main findings, but our models should still be interpreted as demonstrating associations rather than causality.

Lastly, the study is limited by the exclusion of female gang involvement, which is more prevalent in other cities, and our findings may not be generalizable to mothers’ influence or associations’ influence on young women. Future research should address this.

Despite these limitations, administrative data offer significant advantages that make them a valuable resource for research. By encompassing a broad range of information on adolescents, their peers, and family members, these data provide unique insights into the social contexts that shape youth behavior. The findings demonstrate that exposure to criminal justice involvement within one’s social network significantly increases the likelihood of both gang membership and adult correctional supervision, with the impact of knowing a system-identified gang-involved individual doubling the odds of gang involvement. Additionally, the influence of criminal justice exposure on adult correctional supervision remains significant even after accounting for gang membership, highlighting the enduring effects of early social environments on long-term outcomes. Thus, moving forward, research should consider the breadth and complexity of social circles, as the cumulative impact of these relationships may play a critical role in shaping pathways into and through the criminal justice system.

Acknowledgement:

We thank Dr. Jerald Herting for his insightful comments.

Funding:

Research reported in this manuscript was funded by the National Institute of Justice W.E.B Dubois grant 2017-MU-MU-0055. Partial support for this research came from a Eunice Kennedy Shriver National Institute of Child Health and Human Development research infrastructure grant, P2C HD042828, to the Center for Studies in Demography & Ecology at the University of Washington.

Biographies

John Leverso is an assistant professor at the University of Cincinnati’s School of Criminal Justice. His research focuses on urban street gangs, later-in-life outcomes for justice-involved youth, and health criminology.

David Schwager was a Senior Data Analyst with Multnomah County’s Department of Community Justice, and has recently transferred to the county’s Health Department. He earned a B.S. and M.S. in statistics from Cornell University and Portland State University, respectively. He worked for market research and electron microscope companies before taking his skills into the public sector, where he can support and improve knowledge and decision making in government.

Kelsey Ravindren is a Senior Data Analyst with Multnomah County’s Department of Community Justice. She earned her Master’s Degree in Criminal Justice from the University of Cincinnati. Kelsey’s research focuses on juvenile justice, best practices in community supervision, and correctional rehabilitation. She has collaborated on numerous research projects aimed at improving probation practices with both adults and juveniles including Effective Practices in Community Supervision (EPICS), Subject-Matter Expert Training (SMART) Supervision, and Second Chance for Women

Miranda Sitney is a research scientist at the Multnomah County Department of Health. Her research focuses on juvenile male sexual offending, father and son relationships, and the harmful impacts of long detention facilities for youth.

Jennifer Roark is an independent researcher specializing in the criminal justice system. Formerly the Research and Planning Manager for the Department of Community Justice, her work focuses on examining key issues in criminal justice policy, reform, and system impacts.

Footnotes

Conflict of Interest: The authors declare that they have no conflict of interest.

1

We considered differentiating by level of correctional sentencing (i.e., prison vs. probation) as indicators of varying levels of criminal involvement, and performed sensitivity tests with such differentiation. These tests indicated that both the direction and magnitude of odds ratios were remarkably consistent with those reported in our main findings. Given this consistency, we retained the combined indicator for correctional sentencing in our analyses to maintain a focused set of results.

2

To verify the robustness of our decision to code missing fathers as gang uninvolved, we conducted a sensitivity test by modeling the father’s gang and criminal justice involvement as a three-category variable: Not Gang Father, Gang Father, and Absent Father. The results were consistent with our main results. Therefore, we retain the coding of missing fathers as gang uninvolved in our analyses.

3

Our coding strategy raises questions about whether the number of peers or siblings/cousins influences our outcomes. For example, we treat someone with 1 gang-involved friend the same as someone with 10 gang-involved friends, though the level of influence might well differ. However, findings from a sensitivity analysis treating the number of gang-involved or criminally involved peers and same-generation family members as counts rather than as a dichotomous variable (one or more vs. none) were highly consistent with those reported in our original models.

References

  1. Akers RL, & Jensen GF (2017). The empirical status of social learning theory of crime and deviance: The past, present, and future. Taking stock, 37–76. [Google Scholar]
  2. Akers R (2017). Social learning and social structure: A general theory of crime and deviance. Routledge. [Google Scholar]
  3. Augustyn MB, Ward JT, & Krohn MD (2017). Exploring intergenerational continuity in gang membership. Journal of crime and justice, 40(3), 252–274. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Besemer S, Ahmad SI, Hinshaw SP, & Farrington DP (2017). A systematic review and meta-analysis of the intergenerational transmission of criminal behavior. Aggression and violent behavior, 37, 161–178. [Google Scholar]
  5. Beijers J, Bijleveld C, van de Weijer S, & Liefbroer A (2017). “All in the family?” The relationship between sibling offending and offending risk. Journal of Developmental and Life-Course Criminology, 3(1), 1–14. [Google Scholar]
  6. Cesar GT, Walker DA, & Fernandez T (2024). On gangs and family: Primary, secondary, and surrogate family. In The handbook of gangs. [Google Scholar]
  7. Curry GD (2000). Self-reported gang involvement and officially recorded delinquency. Criminology, 38(4), 1253–1274. [Google Scholar]
  8. Decker SH, Pyrooz DC, Densley JA 2022. On Gangs. Philadelphia: Temple Univ. Press [Google Scholar]
  9. Decker SH, & Van Winkle B (2020). Life in the gang: Family, friends, and violence. In Crime, Inequality and the State (pp. 338–352). Routledge. [Google Scholar]
  10. Decker SH, Katz CM, & Webb VJ (2008). Understanding the black box of gang organization: Implications for involvement in violent crime, drug sales, and violent victimization. Crime & delinquency, 54(1), 153–172. [Google Scholar]
  11. Esbensen FA, & Carson DC (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(4), 465–481. [Google Scholar]
  12. Esbensen FA, & Deschenes EP (1998). A multisite examination of youth gang membership: Does gender matter?. Criminology, 36(4), 799–828. [Google Scholar]
  13. Esbensen FA, Winfree LT Jr, He N, & Taylor TJ (2001). Youth gangs and definitional issues: When is a gang a gang, and why does it matter?. Crime & delinquency, 47(1), 105–130. [Google Scholar]
  14. Esbensen F, Osgood D, Peterson D, Taylor T, & Carson D (2013). Short- and long-term outcome results from a multisite evaluation of the G.R.E.A.T. Program. Criminology and Public Policy, 12(3), 375–411. DOI: 10.1111/1745-9133.12048 [DOI] [Google Scholar]
  15. Esbensen F, Peterson D, Taylor T, & Osgood D (2012). Results from a multi-site evaluation of the GREAT program. Justice Quarterly, 29(1), 125–151. DOI: 10.1080/07418825.2011.585995 [DOI] [Google Scholar]
  16. Farrington DP (1998). Predictors, causes, and correlates of male youth violence. Crime and justice, 24, 421–475. [Google Scholar]
  17. Farrington DP (2000). Explaining and preventing crime: The globalization of knowledge—The American Society of Criminology 1999 presidential address. Criminology, 38(1), 1–24. [Google Scholar]
  18. Geller A, Garfinkel I, Cooper CE, & Mincy RB (2009). Parental incarceration and child well-being: implications for urban families. Social science quarterly, 90(5), 1186–1202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Gilman AB, Hill KG, & Hawkins JD (2014). Long-term consequences of adolescent gang membership for adult functioning. American journal of public health, 104(5), 938–945. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Glueck S, & Glueck E (1950). Unraveling juvenile delinquency. Juv. Ct. Judges J, 2, 32. [Google Scholar]
  21. Huey S, Lewine G, & Rubenson M (2016). A brief review and meta-analysis of gang intervention trials in North America. In Maxson C & Esbensen F (Eds.), Gang transitions and transformations in an international context (217–233). Springer. [Google Scholar]
  22. Jankowski MS (1991). Islands in the street: Gangs and American urban society (Vol. 159). Berkeley: University of California Press. [Google Scholar]
  23. Kerig PK, & Mendez L (2022). The Role of Trauma in the Developmental Trajectories of Gang-Involved Youth. Psychology of Gang Involvement, 51–69. [Google Scholar]
  24. Kissner J, & Pyrooz DC (2009). Self-control, differential association, and gang membership: A theoretical and empirical extension of the literature. Journal of Criminal Justice, 37(5), 478–487. [Google Scholar]
  25. Klein MW, & Maxson CL (2010). Street gang patterns and policies. Studies in Crime and Public Policy. [Google Scholar]
  26. Krohn MD, Ward JT, Thornberry TP, Lizotte AJ, & Chu R (2011). The cascading effects of adolescent gang involvement across the life course. Criminology, 49(4), 991–1028. [Google Scholar]
  27. Kubik J, Docherty M, & Boxer P (2019). The impact of childhood maltreatment on adolescent gang involvement. Child abuse & neglect, 96, 104096. [DOI] [PubMed] [Google Scholar]
  28. Hashimi S, Wakefield S, & Apel R (2021). Sibling transmission of gang involvement. Journal of research in crime and delinquency, 58(5), 507–544. [Google Scholar]
  29. Hill KG, Howell JC, Hawkins JD, & Battin-Pearson SR (1999). Childhood risk factors for adolescent gang membership: Results from the Seattle Social Development Project. Journal of research in Crime and Delinquency, 36(3), 300–322. [Google Scholar]
  30. Lahey BB, Gordon RA, Loeber R, Stouthamer-Loeber M, & Farrington DP (1999). Boys who join gangs: A prospective study of predictors of first gang entry. Journal of Abnormal Child Psychology, 27, 261–276. [DOI] [PubMed] [Google Scholar]
  31. Leverso J, Bielby W, & Hoelter LF (2015). Back on the streets: Maturation and risk factors for recidivism among serious juvenile offenders. Journal of Adolescence, 41, 67–75. [DOI] [PubMed] [Google Scholar]
  32. Leverso J, & Matsueda RL (2019). Gang organization and gang identity: An investigation of enduring gang membership. Journal of quantitative criminology, 35, 797–829. [Google Scholar]
  33. Leverso J, & Hess C (2021). From the hood to the home: masculinity maturation of Chicago street gang members. Sociological Perspectives, 64(6), 1206–1223. [Google Scholar]
  34. Lonardo RA, Giordano PC, Longmore MA, & Manning WD (2009). Parents, friends, and romantic partners: Enmeshment in deviant networks and adolescent delinquency involvement. Journal of youth and adolescence, 38, 367–383. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Maneiro L, Ziti Y, Van Geel M, Gómez-Fraguela XA, & Vedder P (2022). The role of deviant siblings in delinquency: a meta-analysis of longitudinal studies. Aggression and violent behavior, 101780. [Google Scholar]
  36. McGloin JM, & Collins ME (2015). Micro-level processes of the gang. The handbook of gangs, 276–293. [Google Scholar]
  37. McGloin JM, & Thomas KJ (2019). Peer influence and delinquency. Annual Review of Criminology, 2(1), 241–264. [Google Scholar]
  38. McHale SM, Updegraff KA, & Whiteman SD (2012). Sibling relationships and influences in childhood and adolescence. Journal of marriage and family, 74(5), 913–930. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Mears DP, & Siennick SE (2016). Young adult outcomes and the life-course penalties of parental incarceration. Journal of Research in Crime and Delinquency, 53(1), 3–35. [Google Scholar]
  40. Melde C, & Esbensen FA (2011). Gang membership as a turning point in the life course. Criminology, 49(2), 513–552. [Google Scholar]
  41. Miller HV (2010). If your friends jumped off of a bridge, would you do it too? Delinquent peers and susceptibility to peer influence. Justice Quarterly, 27(4), 473–491. [Google Scholar]
  42. Moore JW, Garcia R, Moore JW, & Garcia C (1978). Homeboys: Gangs, drugs, and prison in the barrios of Los Angeles (p. x). Philadelphia: Temple University Press. [Google Scholar]
  43. Moore J (2010). Going down to the barrio: Homeboys and homegirls in change. Temple University Press. [Google Scholar]
  44. Noel ME, & Hoeben EM (2022). “The Glass is Half Full”: Narratives from Young Adults on Parental Incarceration and Emerging Adulthood. Journal of child and family studies, 31(11), 3125–3139. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Pratt TC, & Cullen FT (2000). The empirical status of Gottfredson and Hirschi’s general theory of crime: A meta-analysis. Criminology, 38(3), 931–964. [Google Scholar]
  46. Pyrooz DC, Turanovic JJ, Decker SH, & Wu J (2016). Taking stock of the relationship between gang membership and offending: A meta-analysis. Criminal Justice and Behavior, 43(3), 365–397. [Google Scholar]
  47. Pyrooz DC, & Decker SH (2011). Motives and methods for leaving the gang: Understanding the process of gang desistance. Journal of Criminal Justice, 39(5), 417–425. [Google Scholar]
  48. Pyrooz DC, Melde C, Coffman DL, & Meldrum RC (2021). Selection, stability, and spuriousness: Testing Gottfredson and Hirschi’s propositions to reinterpret street gangs in self-control perspective. Criminology, 59(2), 224–253. [Google Scholar]
  49. Pyrooz DC, & Sweeten G (2015). Gang membership between ages 5 and 17 years in the United States. Journal of Adolescent Health, 56(4), 414–419. [DOI] [PubMed] [Google Scholar]
  50. Pyrooz DC, Decker SH, & Owens E (2020). Do prison administrative and survey data sources tell the same story? A multitrait, multimethod examination with application to gangs. Crime & Delinquency, 66(5), 627–662. [Google Scholar]
  51. Pyrooz DC, & Decker SH (2011). Motives and methods for leaving the gang: Understanding the process of gang desistance. Journal of Criminal Justice, 39(5), 417–425. [Google Scholar]
  52. Rowe DC, & Farrington DP (1997). The familial transmission of criminal convictions. Criminology, 35(1), 177–202. [Google Scholar]
  53. Sampson RJ, & Laub JH (1993). Crime in the making: Pathways and turning points through life. Crime & Delinquency, 39(3), 396–396. [Google Scholar]
  54. Schwartz JA, & Beaver KM (2014). Exploring whether genetic differences between siblings explain sibling differences in criminal justice outcomes. Comprehensive psychiatry, 55(1), 93–103 [DOI] [PubMed] [Google Scholar]
  55. Shaw CR, & McKay HD (1969). Juvenile delinquency and urban areas (rev. ed.). Chicago: University of Chicago Press. [Google Scholar]
  56. Smetana JG, Campione-Barr N, & Metzger A (2006). Adolescent development in interpersonal and societal contexts. Annu. Rev. Psychol, 57, 255–284. [DOI] [PubMed] [Google Scholar]
  57. Steinberg L, & Silverberg SB (1986). The vicissitudes of autonomy in early adolescence. Child development, 841–851. [DOI] [PubMed] [Google Scholar]
  58. Sutherland Edwin H. 1947. Principles of Criminology, 4th ed. Philadelphia, PA: J. B. Lippincott. [Google Scholar]
  59. Thornberry TP, Krohn MD, Lizotte AJ, Smith CA, & Tobin K Gangs and Delinquency in Developmental Perspective. [Google Scholar]
  60. Turney K, & Lanuza YR (2017). Parental incarceration and the transition to adulthood. Journal of Marriage and Family, 79(5), 1314–1330. [Google Scholar]
  61. Vigil JD (1988). Barrio gangs: Street life and identity in Southern California. University of Texas Press. [Google Scholar]
  62. Wakefield S, & Wildeman C (2013). Children of the prison boom: Mass incarceration and the future of American inequality. Oxford University Press. [Google Scholar]
  63. Warr M (2002). Companions in crime: The social aspects of criminal conduct. Cambridge University Press. [Google Scholar]
  64. Wildeman C (2020). The intergenerational transmission of criminal justice contact. Annual Review of Criminology, 3, 217–244. [Google Scholar]
  65. Winfree LT Jr, Bäckström TV, & Mays GL (1994). Social learning theory, self-reported delinquency, and youth gangs: A new twist on a general theory of crime and delinquency. Youth & Society, 26(2), 147–177. [Google Scholar]
  66. Winfree LT Jr, Mays GL, & Vigil-Bäckström T (1994). Youth gangs and incarcerated delinquents: Exploring the ties between gang membership, delinquency, and social learning theory. Justice Quarterly, 11(2), 229–256. [Google Scholar]

RESOURCES