Caregiver Type and Gang Involvement: A Comparison of Female and Male Gang Members

Abstract

Gang involvement and delinquency are prominent issues frequently examined in criminal justice scholarship. Research has revealed that gang involvement increases the likelihood of delinquency/crime, and that youth participate in gangs for a variety of reasons including protection, status, and a sense of belonging. Although research has found that various social and familial factors increase the probability of gang involvement, it primarily focuses on males, and little work has compared how a youth’s primary caregiver influences the likelihood of gang involvement among both male and female gang members. The current study uses school level data to examine gang involvement and primary caregiver type among male and female youth. The results identify significant differences in gang involvement among boys and girls when examining primary caretaker. The findings have implications for theory and programming in connection with youth gang involvement for both boys and girls.

1. Introduction

Scholars continue to investigate gang involvement among America’s youth. The prior literature reveals an increased probability of criminal offending among gang members compared to non-gang members (Decker 1996; Gordon et al. 2004; Pyrooz et al. 2016; Thornberry et al. 2003). Due to the higher propensity of criminal offending among gang members, scholars continue to investigate why youths join gangs. This body of research has identified various risk factors that increase the likelihood of gang involvement, including poor academic performance, traumatic experiences, and poor family life (Deuchar et al. 2020; Miller 2001; 2008; Moore and Hagedorn 2001). While specific family characteristics have been identified to increase youths’ likelihood of joining a gang, it is unclear if family structure or primary caregiver type impacts this risk. This is problematic, as more and more youths residing in the United States are reportedly living not only with one or both of their biological parents, but with stepparents (Jacobsen and Zaatut 2022), extended family members, and even custodial grandparents (Hemez and Washington 2021). Additionally, there continues to be a much higher focus placed on examining male youths in gangs, in comparison to female gang members (Deuchar et al. 2020). Scholarship remains limited in directly comparing the characteristics of male and female youths and their risk of gang membership. Given the prevalence of both males and females in gangs, and the relationship between family life and delinquency, the current study contributes to this gap in the literature by comparing primary caregiver(s) among male and female gang-involved youths.

2. Gang Involvement and Crime

Research has frequently found that gang-involved youths are more likely to engage in delinquency (Gordon et al. 2004). In fact, the evidence shows that youths who join gangs are more likely to have engaged in crime before entering the gang than those who do not join a gang (Gordon et al. 2004). Studies on gangs and crime have consistently found that gang involvement increases the probability of participation in violence (Decker 1996; Pyrooz et al. 2016; Thornberry et al. 2003). Violence defines the gang (Decker and Van Winkle 1996), and gang members participate in violence related to both gang and non-gang issues (Moore 1978; Scott 2018). Research has also explored the various reasons for this increased likelihood of violence. Some work has found that violence is a crucial part of a gang member’s identity formation (Stretesky and Pogrebin 2007). Additionally, violence may occur due to the need to retaliate, build a reputation, friendly fighting within the gang (Collins 2009; Decker and Curry 2002), or support and protect gang members (Klein 1997). Moreover, gangs may engage in violence due to issues with race, territory, and group protection (Adamson 2000; Alonso 2004; Hagedorn 2006). Violence between rivals is frequent (Papachristos 2009; Klein 1997), and gang violence may occur due to issues related to drugs, robbery, or domestic violence (Tita and Abrahamse 2004). The strong relationship between gang involvement and crime or violence highlights the need to understand gang involvement, and, specifically, how it might vary by demographics such as gender and family structure.

3. Gender Differences and Gangs

Much of the prior literature examining gang involvement and membership has often focused on males, leaving a large gap in terms of exploring female gang members (Deuchar et al. 2020; Panfil and Peterson 2015). While female gang involvement can differ significantly from the male experience, research suggests that female gangs are typically categorized into three types: mixed-gender gangs with male and female members; female gangs that are linked with male gangs, which are referred to as “auxiliary” gangs; and independent female gangs (Deuchar et al. 2020; Miller and Brunson 2000). Of these three gang types, mixed-gender gangs are viewed as the most common, with various women resisting the label “female gang” (Deuchar et al. 2020).

When examining gender differences and gang involvement, research reveals that female gang members commit the same type of offenses as males, but to a lesser extent (Deuchar et al. 2020; Esbensen and Winfree 1998). Official data also reveals that females are reportedly less involved in violent crimes in comparison to male gang members (Deuchar et al. 2020). According to Gover et al. (2009) and others (Miller and Decker 2001), female gang members are often limited in participating in the distribution of drugs or crimes involving firearms, which can reduce their risk of serious injury or death. However, research suggests that gang involvement can increase the risk of engaging in more serious crimes for female members (Deuchar et al. 2020). In comparison to male gang members, female gang members are known to display higher rates of delinquency (Deuchar et al. 2020; Peterson et al. 2001). This may be a product of females falling victim to peer influence (Deuchar et al. 2020) or the threat of abuse or victimization (Harding 2014; Havard et al. 2021).

Previous research suggests that many females who become gang-involved have experienced a number of difficult family obstacles while growing up. Female gang members reported experiencing drug addiction within the home, domestic violence, absent fathers, and maternal neglect due to rejection or addiction (Deuchar et al. 2020). For these reasons, gang membership can be seen as a safety net for females who have experienced traumatic past experiences (Moore and Hagedorn 2001), especially for impoverished, minority females who reside within male-dominated, high-crime neighborhoods (Miller 2001, 2008).

4. Gangs, Delinquency, and Family Characteristics

Much of the extant literature investigating juvenile delinquency and gang membership has focused on risk factors associated with one’s family and home life. Research suggests that family structure and other characteristics can impact a juvenile’s propensity to engage in delinquent behavior (Scott et al. 2022) or join a gang (Lenzi et al. 2015; Wood et al. 1997). Households in which there is poor parental involvement, a lack of supervision, and inconsistent discipline have been shown to increase the likelihood of youth gang membership (Lenzi et al. 2015; Wood et al. 1997) and participation in delinquent or criminal behavior (Jacobsen and Zaatut 2022; Scott et al. 2022; Williams et al. 1997). Kakar (1998) argued that caregivers who struggled to manage the behavior of youths exhibited problem behaviors themselves, allowed conflicts and fighting to occur within the family, and supported attitudes favorable towards criminal behavior, fostered the development of delinquency in juveniles. Previous research suggests youths raised by parents who also engage in criminal behavior may be at an increased risk of exhibiting this same conduct (Kakar 1998; Nijhof et al. 2009). According to Nijhof et al. (2009), youths with both parents who reported a criminal history committed significantly more offenses than youths with one criminal parent or no criminal parents.

Parental rejection or a lack of warmth or affection between caregivers and youths has also been associated with the development of juvenile delinquency and gang membership (Kakar 1998). Parental abuse, neglect, or maltreatment can impact a youth’s risk of delinquency, as well, and can increase their risk of violence later in life (Manzoni and Schwarzenegger 2019; Mersky and Reynolds 2007; Watts 2017). In their longitudinal analysis of minority youths from low-income areas of Chicago, Mersky and Reynolds (2007) found that youths who experienced maltreatment or physical abuse before age 12 were significantly more likely to have a juvenile petition or adult conviction for violent offending. Across 26 different countries, adolescents who were assaulted by their parents were more than twice as likely to use violence against others (Manzoni and Schwarzenegger 2019). According to Connolly and Jackson (2019), factors that contribute to the association between gang membership and adolescents are found within households and neighborhoods that play a role in shaping the mental and physical health of youths over a period of time. Most of this research has focused on males and has not directly compared how these factors influence female gang involvement, however.

5. Family Structure, Delinquency, and Gang Membership

Based on these risk and protective factors, prior research suggests that the household and family structure can impact a juvenile’s risk of delinquent behavior or gang membership. If consistent supervision and discipline are necessary to reduce the risk of engaging in delinquent behavior, then certain family types may be perceived as more equipped to provide such characteristics (Gottfredson and Hirschi 1990; Jacobsen and Zaatut 2022). In comparison to other family structures, scholars have long argued that juveniles who reside in two-parent households are at a reduced risk of engaging in delinquent behavior because they are likely to experience a higher level of supervision and monitoring than juveniles who reside in single-parent households, with a biological parent and stepparent, or an extended family member such as a custodial grandparent, aunt/uncle, or older sibling (also known as kinship care) (Gottfredson and Hirschi 1990; Jacobsen and Zaatut 2022; Scott et al. 2022). Two-parent households, or intact families, are argued to be in the best position to care for and support the needs of developing juveniles, oftentimes with higher incomes and more resources than other family types (Brannigan et al. 2002; Jacobsen and Zaatut 2022; Mack et al. 2007; Wong 2017).

Prior research examining the impact of family structure on a youth’s likelihood to engage in delinquent or antisocial behavior has remained relatively consistent. Juveniles who reside with both biological parents are less likely to exhibit problematic behavior than youths in other household types (Apel and Kaukinen 2008; Brown and Rinelli 2010; Jacobsen and Zaatut 2022; Juby and Farrington 2001). Youths who are raised in single-parent households are more likely to use drugs and alcohol and engage in a number of delinquent behaviors, including property and violent crime (Brown and Rinelli 2010; Demuth and Brown 2004; Fagan and Wright 2012; Griffin et al. 2000; Kroese et al. 2021; Rebellon 2002; Schroeder et al. 2010). Several scholars note, however, that this relationship between family structure and juvenile delinquency is still dependent on how involved caregivers are, the level of supervision provided, and the overall relationship between the caregivers and youths (Demuth and Brown 2004; Goulette et al. 2016; Jacobsen and Zaatut 2022; Mack et al. 2007).

While the previous research is pretty clear in identifying two-parent households as superior to other family types in protecting juveniles from participating in delinquent behavior, the number of youths residing with two parents has been decreasing (Hemez and Washington 2021). Juveniles not only reside with one or both of their biological parents, but with stepparents (Jacobsen and Zaatut 2022), extended family members, and even custodial grandparents (Hemez and Washington 2021). In 2020, for example, it was estimated that roughly 18 million juveniles reportedly lived with only one of their biological parents (Hemez and Washington 2021). During this same year, roughly 3 million children were reported to be residing with a relative. Neither of the child’s biological parents were present at the time. Nearly 55 percent of these youths had a grandparent living with them (Hemez and Washington 2021). Whether through formal or informal placement, kinship care and custodial grandparent-headed households can take shape for several reasons, including parental death, incarceration, abandonment, unemployment, mental illness, drug abuse, or teenage pregnancy (Edwards 2003; Edwards and Mumford 2005; Goulette et al. 2016; Kelley et al. 2011). While these living arrangements have increased over time (Hemez and Washington 2021), scholars continue to examine the challenges facing both caregivers and juveniles who reside in kinship care and how juveniles’ behavior is impacted.

The extant literature surrounding custodial grandparent-headed households and kinship care remains mixed on whether these living arrangements are beneficial to the youths who are placed in them (Lee and Villagrana 2015; Sands et al. 2009; Smith and Palmieri 2007). Lee and Villagrana (2015) and others (Dubowitz et al. 1994; Smith and Palmieri 2007) argue that juveniles who are removed from one living arrangement and placed in the care of another family member might experience trauma, and, as a result, may exhibit aggressive or attention-seeking behaviors (Dubowitz et al. 1994; Sands et al. 2009). This trauma could lead youths to develop psychological issues (Lee and Villagrana 2015; Smith and Palmieri 2007). Conversely, Sands and colleagues (2009) concluded that juveniles who resided in grandparent-headed households exhibited appreciation for the care provided by their grandparents. While several of the juveniles experienced abuse and neglect before residing with their grandparents, interviewed juveniles reportedly felt safe and loved (Sands et al. 2009). Grandparents were described as providing a protective environment for their grandchildren that allowed them to excel at school and with their peers (Sands et al. 2009). This may be a result of kinship care providing consistency and less disruption for youths, allowing youths to develop healthier well-being (Winokur et al. 2014). Youths raised in kinship care who reported having a warm and consistent relationship with their caregivers were found to be protected from engaging in substance abuse or delinquent behavior (Goulette et al. 2016; Jacobsen and Zaatut 2022; Johnson-Garner and Meyers 2003). Finally, scholars argue that residing in kinship care can create resiliency in youth, allowing them to thrive in the care of their relatives (Goulette et al. 2016; Sands et al. 2009; Washington et al. 2013, 2014).

While much of the prior literature has remained relatively positive on the development of kinship care and the impact it may have on juveniles, scholars have highlighted a number of issues facing youths who reside within these households (Goulette et al. 2016). Scholars argue that youths who are raised in kinship care may have a more difficult time adjusting at home or school if they are exposed to inconsistent discipline and parenting practices while in kinship care (Dolbin-MacNab and Keiley 2006; Goulette et al. 2016; Hayslip and Kaminski 2005; Kelley et al. 2011; Rodgers-Farmer 1999; Shelton and Harold 2008; Smith and Richardson 2008; Smith et al. 2008). Smith et al. (2008, 2019) reported youths who were being raised by a custodial grandparent have trouble interacting with other juveniles their age and were more likely to develop both internalizing and externalizing problems over time. Other research suggests that emotional and behavioral problems exhibited by juveniles raised in grandparent-headed households can grow increasingly intense and more frequent as they age (Goulette et al. 2016; Neely-Barnes et al. 2010) and may increase their risk of problematic outcomes in the future.

As a result, it is unclear whether juveniles who are raised by a custodial grandparent are at an increased risk of engaging in delinquent behavior or joining a gang. While scholars have begun to explore a number of outcomes associated with the criminal justice and juvenile justice systems (Foster et al. 2004; Goulette et al. 2016; Ryan et al. 2010), this research remains relatively limited (Yorgason and Gustafson 2014). Using a nationally representative sample of youths, Goulette et al. (2016) found that youths raised by a custodial grandparent were stopped by the police more often than youths raised by other caregiver types. A youth’s self-reported delinquent behavior from adolescence to young adulthood was not found to be influenced by caregiver type, however (Goulette et al. 2016). In their analysis of 37 youths living with a custodial grandparent, Campbell and colleagues (2006) concluded these individuals were more likely to re-offend than juveniles raised by a biological parent. Youths who were living with a grandparent reported several risk factors, including a pattern of problematic behavior earlier in life and a lack of supervision at home (Campbell et al. 2006). In Los Angeles, California, it was concluded that, in comparison to female youths, Black and White males who were raised by a relative were more likely to be arrested for committing a criminal offense (Ryan et al. 2010).

6. Current Study

The current study aims to build and extend the reviewed prior literature in multiple ways. Caregiver types are compared to one another using a large sample of youths from the state of Arizona. Oftentimes, research involving juveniles who reside with custodial grandparents uses small samples (Campbell et al. 2006) or combines kinship care and grandparent-headed households within the analysis (Ryan et al. 2010). The current analysis isolates the influence of grandparents, both biological parents, single parents, and mixed families to investigate whether caregiver types impact the risk of gang membership differently for male and female juveniles. While there is a large gap in the literature surrounding primary caregiver type and both male and female gang members, the current study contributes to the field by examining a non-traditional but growing group of juveniles and their risk of gang membership.

7. Research Question

What is the relationship between gang involvement and primary caregivers among male and female youths?

Hypothesis 1.

The presence of both mom and dad reduces the likelihood of gang membership for both male and female youths.

Hypothesis 2.

The presence of a single parent (either mom or dad) increases the likelihood of gang membership for both male and female youths.

Hypothesis 3.

The presence of a grandparent as a primary caregiver increases the likelihood of gang membership for both male and female youths.

Hypothesis 4.

The presence of an alternative or other family structure as a primary caregiver increases the likelihood of gang membership for both male and female youths.

8. Data and Methods

The data used for this study come from the 2018 Arizona Youth Survey (AYS). This survey is given to a sample of 8th, 10th, and 12th grade youths throughout all 15 counties in the state every two years, by the Arizona Criminal Justice Commission in conjunction with Arizona State University’s School of Criminology and Criminal Justice. The Arizona Youth Survey examines gang involvement and substance abuse, among other behaviors of youths. The survey is derived from the nationally recognized Risk and Protective Factor model incorporated into the Communities that Care Survey (Hawkins et al. 1992). The current study only analyzes the 2018 survey data, which had an overall total of 48,991 participants.1 Selected survey questions related to gang involvement and caregiver status varied depending on the survey year; 2018 included the most variables and cases for our analyses. This data collection effort covered youth demographics, family structure, perceived mental health support, academics, substance use, and gang involvement. The data for this study were requested from the Arizona Criminal Justice Commission. Upon the provision of a plan for how the data was going to be analyzed, the Commission approved and provided the data. Once the data were received, the primary investigators began cleaning the data and developing the necessary codebook. Due to the categorical nature of the dependent and independent variables, chi-square analyses are conducted, as well as multivariable logistic regression analyses. The analyses were conducted through the use of statistical analysis software, STATA 17.

9. Dependent Variable

The dependent variable in this analysis is gang involvement measured in two ways. There is extensive debate in the literature on how to define gang membership (Bjerregaard 2002; Petersen 2000; Roman et al. 2016), but research consistently argues that self-report data is the most accurate way to measure gang membership (Esbensen et al. 2001). The current study will utilize self-reports to examine current gang membership or if the juvenile has ever been a member of a gang. The dependent variable, gang membership, is measured in two ways, including current gang membership, and ever gang membership. For both variables, the youths were asked if they have ever been involved in a gang and if they are currently involved in a gang. Two variables were then created with the youths who responded “yes” to either of these questions. These variables include current gang membership, where 0 = non-gang membership, and 1 = current gang membership, and ever gang membership, where 0 = never gang membership, and 1 = ever gang membership.

10. Independent Variables

The independent variables for this study are primary caregiver type and gender. Youths were originally asked to identify their primary female and male caregivers. They were provided several options including mother/father, stepparent, grandparent, foster parent, other adult, and none. These responses were used to create dichotomous variables (yes/no) capturing if the youth’s primary caregiver(s) were both parents, single parent, grandparent(s), or other primary caregiver. Gender, in this survey, is defined as male or female, which is measured with a question asking the youth if they are male or female. This is a dichotomous variable, where 0 = Male and 1 = Female. Interaction variables with gender and the different primary caregiver types are created and included in the multivariate analysis to examine the relationship more effectively with gang involvement.

11. Control Variables

A number of control variables are included in the current analyses. A youth’s race (White, Hispanic, Black, or other) is included as a control variable, due to the unique relationship between race and crime (Gabbidon and Greene 2018; Kovera 2019).

As described in

Table 1. Control Variable Descriptives for Gang Boys and Girls.

Race%	Current Gang		Ever Gang
	Male (n = 832)	Female (n = 348)	Male (n = 1514)	Female (n = 711)
White	24.7	15	24.6	15.2
Latino	38.5	38.8	40.9	42.9
Black	17.9	12.6	15.3	11.4
Other	18.9	33.6	19.3	30.5
Mean Age in Years	15.49	14.95	15.62	15.08
Lunch%	54.1	59.4	53.9	61
Have Sibling(s)%	71.6	68.3	71	71.2
Caretaker Completed High School%	85.5	83.5	84.9	84.3

, while racial group composition among males and females is similar across both Latino and Black youths, there are noticeable differences when looking at White and other racial groups2. The analysis also included participation in the free lunch program as a control, due to the relationship between crime and socioeconomic status (Agnew et al. 2008). The data in Table 1 identify that gang girls receive free lunches more often than gang boys. Furthermore, research has consistently revealed a relationship between family dynamics and both delinquency (Kierkus and Hewitt 2009) and gang involvement (Whitlock 2004). Therefore, whether or not the youths have siblings, and whether or not their primary caregiver completed high school, are included as control variables in the analyses. The descriptive data also show similar frequencies for both current gang involvement and ever gang involvement.

12. Results

The bivariate results directly compare male and female gang-involved youths and reveal a significant difference in having both parents as primary caregivers and current gang involvement for male and female youths.

According to

Table 2. Gang Boy and Girl Comparison.

Primary Caregiver	Current Gang Boy (n = 832)	Current Gang Girl (n = 348)	Ever Gang Boy (n = 1514)	Ever Gang Girl (n = 711)
Single parent	15.90%	18.40%	16.50%	18.60%
Both parents	*** 61.5%	50.00%	*** 61%	48.30%
Grandparent(s)	** 4%	8.60%	*** 4.1%	8.60%
Mixed	18.70%	23.00%	18.40%	24.50%

*** p < 0.001; ** p < 0.01.

, a significantly greater proportion of current gang boys have both parents as primary caregivers compared to gang girls. There is also a significant difference between having grandparent(s) as primary caregiver(s) and current gang involvement. Specifically, there is a greater proportion of current gang girls who have grandparent(s) as primary caregivers than current gang boys.

Given these significant findings, multivariable logistic regression models are conducted with current gang involvement and ever gang involvement as dependent variables, and with primary caregiver type and gender interacted as independent variables. These analyses allow the level of spuriousness in the findings to be reduced with the inclusion of various control variables, including age, race, siblings, parental education, and free/reduced lunch. Additionally, three models are conducted with the variable of both biological parents as primary caregivers serving as the reference group, due to it being the largest category (

Table 3. Comparing Boys and Girls.

	Model 1 Current Gang/Non-Gang	Model 2 Current Gang/Non-Gang	Model 3 Ever Gang/Never Gang	Model 4 Ever Gang/Never Gang
	n = 35,213	n = 35,213	n = 36,629	n = 36,629
Independent/Control Variables	Odds Ratio (Std. Err.)	Odds Ratio (Std. Err.)	Odds Ratio (Std. Err.)	Odds Ratio (Std. Err.)
Gender	0.341 *** (0.028)	0.321 *** (0.032)	0.393 *** (0.023)	0.365 *** (0.026)
Single parent	1.42 ** (0.154)	1.42 ** (0.179)	1.55 *** (0.123)	1.53 *** (0.144)
Grandparent(s)	2.3 *** (0.379)	1.69 * (0.365)	2.62 *** (0.312)	1.86 *** (0.298)
Mixed	3.3 *** (0.506)	3.21 *** (0.582)	3.03 *** (0.363)	2.95 *** (0.429)
Gender × singleparent	---	0.995 (0.23)	---	1.06 (0.172)
Gender × grandparent	---	2.37 ** (0.77)	---	2.38 *** (0.554)
Gender × mixed	---	1.1 (0.348)	---	1.09 (0.263)
Age	0.929 ** (0.02)	0.929 ** (0.02)	0.972 † (0.016)	0.972 † (0.016)
Other	2.02 *** (0.29)	2.02 *** (0.29)	2.22 *** (0.228)	2.22 *** (0.228)
Hispanic	1.59 *** (0.136)	1.59 *** (0.136)	1.68 *** (0.107)	1.68 *** (0.107)
Black	3.67 *** (0.347)	3.68 *** (0.348)	2.96 *** (0.223)	2.96 *** (0.222)
Reduced or Free Lunch	1.05 (0.084)	1.05 (0.083)	1.12 * (0.066)	1.12 † (0.066)
Have Sibling(s)	0.864 (0.078)	0.86 † (0.077)	0.834 ** (0.055)	0.829 ** (0.055)
Parent High School	0.815 (0.091)	0.813 † (0.091	0.808 ** (0.065)	0.806 ** (0.065)

*** p < 0.001; ** p < 0.01; * p < 0.05; ^† p < 0.10.

).

The results in model 1 reveal various significant findings when comparing current gang and non-gang youth. The results identify a significant and positive association when comparing a single parent and both biological parents as primary caregiver(s) with current gang involvement. Additionally, the association between grandparents and current gang involvement is statistically significant and positive when compared to both biological parents as primary caregivers. Furthermore, youths from mixed families are significantly more likely to be current gang members than kids from families with both parents as primary caregivers. Additionally, gang-involved youths are more likely to report having a single parent or mixed family as primary caregiver(s) compared to both biological parents than non-gang-involved youth. Furthermore, there is a significant and positive association between identifying as Hispanic, Black, or Other racial group compared to White and reporting current gang involvement.

Model 2 includes primary caregiver variables interacted with gender. When grandparent is interacted with gender the results reveal a significant and positive association with current gang involvement. This highlights the significant role of gender in association with grandparent compared to both biological parents as primary caregiver type and current gang involvement. When gender is interacted with single parent and mixed the association lacks significance. Other significant findings from Model 2 include youth that report being male are at an increased probability of reporting being currently gang-involved compared to youth that report being female. Additionally, gang-involved youth are more likely to report having a single parent or mixed family as a primary caregiver compared to non-gang-involved youth. Furthermore, there is a significant and positive association between identifying as Hispanic, Black, or Other racial group compared to White and reporting current gang involvement.

There are also multiple significant findings revealed in model 3, which include ever gang membership as the dependent variable. The relationship between ever gang involvement with single parent, grandparent, and mixed compared to both biological parents as primary caregiver(s) are significant and positive. Other significant findings from Model 3 include youth that report being male are at an increased probability of reporting being ever gang-involved compared to youth that report being female. Furthermore, there is a significant and positive association between identifying as Hispanic, Black, or Other compared to White and reporting being ever gang-involved.

Model 4 includes primary caregiver variables interacting with gender, and ever gang membership as the dependent variable. There are multiple significant findings revealed in model 4. The relationship between ever gang involvement with single parent, grandparent, and/or mixed family compared to both biological parents as primary caregiver(s) is significant and positive, but this changes when the primary caregiver type is interacted with gender. When gender is interacted with grandparent as the primary caregiver, the results reveal a significant and positive association with ever gang involvement when compared to youths having both biological parents as primary caregivers. When gender is interacted with single parent and/or mixed, the association lacks significance. Other significant findings from Model 4 include youths who report being male are at an increased probability of reporting being ever gang-involved compared to youths who report being female. Furthermore, there is a significant and positive association between identifying as Hispanic, Black, or Other compared to White and reporting being ever gang-involved.

13. Discussion

Scholars and practitioners continue to examine risk factors associated with a youth’s likelihood of becoming involved with a gang. Prior research suggests that gang membership can contribute significantly to a youth’s likelihood of engaging in delinquent or violent behavior and can put the youth at an increased risk of injury or death (Decker et al. 2013). While much of this research has continued to focus on male youths, there is growing concern that females can become gang-involved as well. Although prior research suggests some similarities between the likelihood of gang membership for male and female youths (Thornberry et al. 2003), the current study attempts to contribute to the limited literature surrounding female youths and the reasons they become involved in a gang. According to Peterson and Panfil (2017), “gender should be considered, as it interacts with other statuses and shapes experiences” (p. 338).

Overall, the current study reveals significant findings between a youth’s primary caregiver type and gender with the likelihood of gang membership. When interacting gender with having a grandparent as a primary caregiver compared to both biological parents, the analyses reveal a significant and positive association with current gang involvement. These findings held true for youths who had ever been involved in a gang or who were currently gang-involved. These noted differences between gang involvement and gender with primary caregiver type somewhat support the work of Klein and Maxson (2006), who argued that familial structure and attachment may not be significant predictors of gang involvement for male youths. Male youths may arguably become gang-involved for different reasons than female youths, including status, survival, and excitement.

Our findings related to grandparent-headed households, gender, and gang involvement are consistent with work identifying that a higher likelihood of gang involvement among girls occurs when families are highly dysfunctional (Archer and Grascia 2006; Deuchar et al. 2020; Valdez 2007). Many female gang members experience drug addiction, domestic violence, absent fathers, and maternal neglect while growing up (Deuchar et al. 2020). While male youths become gang members for protection and excitement, female youths may see joining a gang as an opportunity to acquire new friends or family (Maxson and Whitlock 2002), or as a safe haven for those who have experienced traumatic past experiences (Moore and Hagedorn 2001). It is possible that when a primary caregiver is non-traditional, such as a grandparent, girls are more likely to seek out gangs for more familial support. This is consistent with past research highlighting the greater emotional need that gangs might fill for girls as compared to boys (Esbensen 2010; Miller 2001). While scholars argue that family characteristics may not be related to predicting gang membership (Klein and Maxson 2006), the current study suggests that more focus should be placed on gender when examining familial characteristics of gang members. For those youths who reside with a custodial grandparent, examining the reasons for a change in care, as well as the parenting and discipline styles of these non-traditional families, would further inform the literature on grandparent-headed households (Goulette et al. 2016).

In consideration of gang prevention and intervention programs, the results suggest that semi-specialized gendered programming could be helpful in successfully reducing the likelihood of gang involvement and improving the effectiveness of gang intervention. This is inconsistent with past research, which suggests that, due to few significant risk factors for gang-joining among boys and girls, there is no need to develop gender-specific programming (Klein and Maxson 2006). However, new family-focused gang programs could be developed to specifically aid in reducing the likelihood of gang-joining and assist with desistance among gang-involved girls and boys. For example, developing a gender-focused gang program utilizing a Structural Community Family Therapy Model could be quite beneficial (McNeil et al. 2013). This type of model specifically focuses on dynamics in the home related to family structure, income, and parenting. By taking gender into consideration, gang prevention and/or intervention could potentially be much more effective (Petersen and Howell 2013). Understanding that taking gender into consideration, lacking both parents, or even a single parent, potentially impacts youths’ likelihood of gang involvement is essential for effective approaches to treatment. Conversely, the added knowledge that interacting gender and gang involvement reveals a significant association with specific primary caregiver types suggests other familial factors may have more of an influence on gang involvement, such as parenting skills, siblings, or other environmental factors.

Although risk factors for gang-joining are generally similar for both boys and girls, the current study is consistent with past research revealing increases in the likelihood of gang involvement when taking gender into consideration on specific indicators (Petersen and Howell 2013). Therefore, semi-specialized gender programming may help prevent and reduce gang involvement more effectively. Although not specific to gang youths, the Strengthening Families Program could be customized to focus on the specific needs of gang-involved boys and girls. The program’s goals are to decrease the use of drugs and participation in delinquent behaviors (Kumpfer and Alvarado 1998). This program helps with improving and enhancing healthy parenting and youth skills, which could be potentially beneficial for gang-involved youths if customized to the needs of their specific family structure to improve effectiveness. It is also important to acknowledge that youths whose primary caregiver(s) are grandparent(s) are more likely to have traumatic experiences related to parental loss, abuse, or drug use (Goulette et al. 2016; Kelley et al. 2011). These are factors that may contribute to a youth’s higher likelihood of being gang-involved and should be taken into consideration and applied in any gender-specific programming.

14. Limitations and Future Directions

While the current study significantly contributes to the body of literature on gang-involved youth, custodial grandparent-headed households and their influence on youth behavior, and the relationship between familial structure and delinquency, the study is not without limitations. First, the results of the current study are based on only one year of the Arizona Youth Survey (AYS). For this reason, the cause cannot be determined between primary caregiver type and the likelihood of gang involvement for male and female youths. Additionally, the time order of the independent and dependent variables is unclear. Future research examining whether primary caregiver type influences gang involvement in youths should focus on longitudinal data in which a youth’s caregiver can be identified prior to the juvenile’s involvement with a gang. Despite this, however, the current study revealed significant differences in primary caregivers among the gang-involved when interacting gang involvement with gender. This highlights a need for future research in this area. Second, by analyzing data from one year of the AYS, the generalizability of the findings is limited. Additional research in other jurisdictions will need to be conducted to more clearly understand how being raised by a custodial grandparent compared to other primary caregiver types influences a youth’s risk of becoming involved in a gang.

The current analyses are also not exhaustive. It is likely that there are other relevant factors to consider when exploring factors influencing family structure and gang involvement among male and female youths. Future research should explore additional variables related to households with different primary caregiver types, including parenting styles, discipline strategies, and the relationship between caregivers and youths. Future research should also further explore the relationship between parents/guardianship and female gang members. In-depth interviews could help shed light on female gang members’ experiences, and specifically what about their family life pushes or pulls them towards gang involvement. This would help to better inform whether specialized programming or treatment is needed, to what extent, and how to more effectively tailor such programming. Additional studies comparing more specific familial struggles, as well as other gang factors such as time in gang, age at joining, and gang structure among boy and girl gang members, would also be beneficial. Finally, because the AYS is a survey administered in schools, youths who do not attend school, or who were absent on the day of data collection, are not included in the current analyses. Examining primary caregiver types and gang involvement may be particularly important to investigate for these youths. Other data sources should be considered for future investigations.

15. Conclusions

The current study attempts to contribute to the ever-growing body of research surrounding gang-involved youth, family, and, specifically, custodial grandparent-headed households and their impact on youths’ behavior. While much of the prior research on gang involvement has focused on males, the current study investigates how primary caregiver types can influence the likelihood of gang involvement when interacting with gender. When gender and both current gang membership and ever gang membership were interacted, the results revealed a significant and positive association with being in the care of a grandparent. The practical implications related to this finding suggest that semi-specialized gendered prevention and intervention programs may be worthwhile in addressing the reasons behind why these youths become gang-involved. While the current study contradicts the work of others (Klein and Maxson 2006), future research is needed to investigate the various types of primary caregiver households more closely, as well as other familial structures, and their impact on gang involvement and, specifically, youth behavior.