Criminology paper

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/320692983

PARENTHOOD AS A TURNING POINT IN THE LIFE COURSE FOR MALE AND FEMALE GANG MEMBERS: A STUDY OF ....

Article in Criminology · October 2017

DOI: 10.1111/1745-9125.12162

CITATION

1

READS

267

3 authors, including:

Some of the authors of this publication are also working on these related projects:

The LoneStar Project View project

The Overlap Between Gangs and Terror Groups View project

David C. Pyrooz

University of Colorado Boulder

68 PUBLICATIONS 1,539 CITATIONS

SEE PROFILE

Scott H. Decker

Arizona State University

187 PUBLICATIONS 5,185 CITATIONS

SEE PROFILE

All content following this page was uploaded by David C. Pyrooz on 28 October 2017.

The user has requested enhancement of the downloaded file.

  1

Parenthood as a Turning Point in the Life Course for Male and Female Gang Members:

A Study of Within-Individual Changes in Gang Membership and Criminal Behavior

DAVID C. PYROOZ* Department of Sociology

University of Colorado Boulder

JEAN MARIE MCGLOIN Department of Criminology and Criminal Justice

University of Maryland

SCOTT H. DECKER School of Criminology and Criminal Justice

Arizona State University

forthcoming in the Criminology doi: 10.1111/1745-9125.12162

This is the authors’ pre-print copy of the article Please download and cite the post-print copy published on the Criminology website:

http://onlinelibrary.wiley.com/doi/10.1111/1745-9125.12162/full

ACKNOWLEDGMENTS: This research was funded in part by the Office of Juvenile Justice and Delinquency Prevention under an award in FY 2011 (Grant # 2011-JV-FX-0004). The findings and conclusions of this paper are the sole responsibility of the authors and may not reflect the views of the OJJDP. An earlier version of this paper was presented at the 72nd meeting of the American Society of Criminology in New Orleans. The authors would like to thank Stefanie Mollborn for her comments on the paper. *Correspondence concerning this article should be addressed to David C. Pyrooz, Department of Sociology and Institute of Behavioral Science, UCB 483, University of Colorado Boulder, Boulder, CO, 80309-0483, USA. [email protected]

  2

Parenthood as a Turning Point in the Life Course for Male and Female Gang Members: A Study of Within-Individual Changes in Gang Membership and Criminal Behavior

ABSTRACT

The impact of parenthood on leaving a street gang is not well understood. This is likely because prior studies tend not to account for multiple dimensions of gang exit, possible gender differences, and/or potential selection bias. The current study uses a sample of 466 male and 163 female gang members from the National Longitudinal Study of Youth 1997 to consider the within-individual relationship between changes in parenthood and changes in claiming gang membership and offending. These data offer the opportunity to consider gender differences and birth parity (i.e., first or second child). Results from a series of fixed effects models reveal that motherhood is associated with enduring reductions in both the odds of claiming gang membership and the rate of offending, whereas fatherhood has a temporary beneficial impact on gang membership and offending only for those fathers who reside with their children. In most cases, the beneficial effect of having a child rests in becoming a parent for the first time. On the whole, our study suggests that parenthood serves as turning point for a particular group of noteworthy offenders—gang members. Keywords: gang; parenthood; disengagement; desistance; NLSY97

  3

In his ASC Sutherland address, John Laub (2006) argued that the life course framework

could serve as a valuable paradigm for criminology. The dramatic expansion of knowledge

regarding the patterns and processes of individual-level offending over the past few decades is

evidence of the validity of this assertion (Gibson and Krohn, 2013). Recent research

demonstrates that adopting a life course approach provides added value even with regard to an

already broad and deep research literature. Indeed, the impact of gang membership has been

studied for nearly a century (e.g., Klein 1995; Moore 1991; Pyrooz et al. 2016; Short and

Strodtbeck 1965; Thrasher 1927; Whyte 1943), yet recent work adopting a life course framework

undoubtedly affords a unique and refined sense of how and why gang membership facilitates

various detrimental outcomes. For example, Melde and Esbensen (2011) revisited the oft-

studied amplification effect of gang membership and demonstrated how joining a gang prompts

dramatic changes in emotional, cognitive, and social domains, which in turn increases

delinquency.

The life course framework has also pushed gang researchers to focus more attention on

the process of and reasons for exiting gang life. This research is still in its nascent stages, to be

sure, but the past few years have witnessed a dramatic rise in empirical inquiries focused on the

triggers and consequences of leaving street gangs (for reviews, see: Carson and Vecchio 2015;

Curry, Decker, and Pyrooz 2014). Despite these advances, however, broader questions remain

about the impact of life events on involvement in crime. When compared to the general criminal

desistance literature, we lack an empirical base for understanding whether certain life events

prompt gang exit. This void in knowledge is important both for understanding gang members

specifically and for a broader understanding of the impact of life course events on processes

beyond gang membership, such as delinquent involvement and transitions to adulthood.

  4

Parenthood has been identified repeatedly as a potentially salient “turning point” for gang

members, consistent with Sampson and Laub’s (2016, 327) conceptual standard, but thus far

empirical work has not provided a uniform answer regarding whether changes in parental status

promote gang exit (Fleisher and Krienert 2004; Flores 2013; Hagedorn 1994; Moloney et al.

2009, 2011; O’Neal et al. 2014; Varriale 2008). This lack of clarity likely results from several

issues. First, leaving the gang is not a simple, unidimensional concept; ideally, studies should

account for dimensions of both identity and behavior in order to offer a more complete sense of

whether parenthood truly prompts meaningful changes in gang members’ lives (see Pyrooz and

Decker, 2011: 423). Second, most inquiries have yet to simultaneously consider the impact of

parenthood for both male and female gang members (e.g., Varriale 2008), even though the

literature underscores gender differences in the meaning of both parenthood and gang life

(McGloin and DiPietro, 2013; W. H. Miller, 2011; O’Neal et al., 2014; Sanchez and Thomson,

1997). As a result, parenthood may have notably different effects on identity and behavior

depending on whether the gang member is male or female. Third, research suggests there is wide

variation in the extent to which male gang members embrace fatherhood as a social change

(Moloney et al. 2009; Roy 2006). Accounting for such variation may be important, as there is

reason to suspect that the effects of fatherhood will not be universal, but instead depend on the

degree of involvement in the child’s life.

The current study addresses these gaps using data from the National Longitudinal Survey

of Youth 1997 (NLSY97). Within these data, 629 study participants self-identified as gang

members during at least one wave and, among these individuals, 287 reported having children.

The NLSY97 offer the opportunity to compare within-individual changes in parental status with

changes in gang members’ self-identification and criminal behavior. Moreover, the data contain

enough male and female gang members to support direct gender comparisons of any observed

  5

parenthood effects. Finally, the NLSY97 also capture whether fathers reside with their child,

which we use as a proxy for involvement. In the end, this study will serve as the first test of

whether parenthood promotes leaving gang life across dimensions of identity and behavior for

both men and women.

PARENTHOOD AND LEAVING THE GANG

Ample research suggests that joining a street gang is a negative life event, setting in

motion a variety of emotional, cognitive, and social changes that support offending behavior

(though see Goldstein and Glick 1994; Sanchez-Jankowski 1991). The fact that gang

membership tends to be temporary offers little comfort in the face of estimates that more than

one million youth are currently embedded in this negative institution (Pyrooz and Sweeten,

2015). In light of the serious negative consequences that may be generated by gang membership,

scholars have become increasingly invested in understanding the mechanisms for leaving the

gang. Incorporating the broader literature on desistance offers some conceptual guidance on this

front, but one cannot simply assume that factors such as marriage and stable employment readily

translate into facilitating gang exit. After all, individuals who join gangs tend to be young,

marginalized members of society, and gang membership narrows their options for experiencing

these prosocial conventional life events even further. Still, gang members do appear to

experience one major social change with some regularity: parenthood (Krohn et al., 2011;

Thornberry, Smith, and Howard, 1997). The consequences of parenthood among gang members

remain unclear, however.

Parenthood fits Sampson and Laub’s (2016, 327) conceptual standard of a positive

turning point, in that it should “reorder short term situational inducements to crime and, over

time, redirect long-term commitments to conformity.” Subjects in Edin, Nelson, and Paranal’s

(2001) study of low skilled men, the majority of whom had criminal histories, spoke of

  6

fatherhood in ways that suggested they divided their life into periods before and after having a

child, often speaking in powerful terms about how they changed their lifestyles and their own

views of self. As one man said about the birth of his daughter:

She changed my life a lot. I was headed down the wrong path…ever since she’s been born, I slowed down a lot. You know. ‘Cause it’s like, before her, I didn’t really care too much about anything. I really just lived every day for that day. But as of now, I’m living every day for today and tomorrow (Edin et al., 2001: 23-4).

Several scholars have echoed this sentiment, arguing that having a child can operate as a “hook”

for change (Edin and Kefalas 2011; Kreager, Matsueda, and Erosheva 2010).

As the wider psychological and sociological literature discusses, parenthood is a dramatic

life event because “the role changes explicit and implicit during the transition to parenthood are

of considerable magnitude and breadth” (Feldman and Nash 1984, 63). Indeed, it can affect

numerous factors that resonate in theoretical explanations of criminal behavior (Savolainen

2009; Ziegler et al. 2017). Parenthood offers the opportunity to forge strong social bonds and

develop a stake in conformity (Hirschi 1969; Sampson and Laub 1990), it changes socializing

routines away from unstructured time with friends who may be deviant (Akers 2009; Osgood et

al. 1996; Warr 2002), and it can promote changes in identity (Giordano, Cernkovich, and

Rudolph 2002).

Despite the fact that these processes echo those set in motion by marriage, about which

there is ample research, there are comparatively few studies on parenthood and desistance from

crime (Edin, Nelson, and Paranal 2001; Monsbakken, Lyngstad, and Skardhamar 2013; Ziegler

et al. 2017). Moreover, the research that does exist is inconclusive, with studies documenting a

positive effect (e.g., Kreager, Matsueda, and Erosheva 2010), a null effect (e.g., Blokland and

Nieuwbeerta 2005), and a negative effect (e.g., Wakefield and Uggen 2004). These mixed

findings may reflect a lesson that the broader literature on parenthood has identified in recent

  7

years: parenthood is not a singular experience that produces systematic outcomes. Instead, this

life event can take on different meanings and foster different changes depending on the social

context in which it is experienced (Umberson, Pudrovska, and Reczek 2010).

For instance, parenthood among gang members is often “precocious” and unplanned

(Moloney et al., 2009; Moloney et al., 2011), which raises questions about whether it may be just

another negative disruption in their lives or if it might counterbalance the often pernicious effects

of gang membership. Moreover, there are indications that many male gang members are not

involved or committed parents (Moloney et al., 2009), which raises questions about the extent to

which fatherhood is truly a “hook for change”. Unfortunately, as with the general parenthood and

crime literature, the empirical evidence does not offer much guidance on this front.

On the one hand, Fleisher and Krienert’s (2004) study of Black female gang members in

Champaign, Illinois revealed that of those who left the gang, over half cited having a child as the

reason. Moore’s (1991) work on female gang members in Los Angeles and Bolden’s (2013)

recent research on gang members in Texas also cited motherhood and pregnancy as the primary

reason for exiting the gang. Further, in Moloney et al.’s (2009) sample of 91 current and former

male gang members in the San Francisco Bay Area, several offered “narratives of

transformation” around fatherhood, saying how it had changed their lives and shifted their

priorities. On the other hand, some quantitative studies question the robustness of these findings.

For example, using national longitudinal data, Varriale (2008) found no evidence that pregnancy

occurred in the prior or same interview period as when subjects de-identified as gang members.

One possible reason for the lack of agreement in the empirical research may rest in the

fact that the outcome measures are not always consistent. Some studies focus on changes in

gang member identity, some on offending, some on other risky behaviors like substance use, and

some on an amalgam of antisocial outcomes that typically define gang life, such as hanging out

  8

with other gang members. Of course, this makes comparisons difficult, but it also highlights an

important reality about the process of gang exit. As Decker, Pyrooz and Moule (2014) note,

leaving gang life is an intermittent, gradual, and complex process. Part of the reason it is

“complex” is because it involves changes across several domains, including friendship, routine

activities, value commitments and identity—much like parenthood does. This reflects Decker et

al.’s (2014) observation that there are at least two broad dimensions of identity change in leaving

the gang: (1) shifts in roles and (2) shifts in commitment to activities related to a particular

identity. When individuals join gangs, they alter their personal identity (i.e., adopt the label of

gang member) as well as amplify their illegal behaviors in a way that is consistent with this new

identity. Thus, leaving the gang is not simply about de-identification from the gang member

label, but also about disembedding oneself from its consequent connections and snares (see, e.g.,

Pyrooz, Sweeten, and Piquero 2013). Ideally then, studies should account for both gang

membership identity and behavior patterns when studying gang exit.

CHANGES IN GANG MEMBERSHIP AND OFFENDING

Changing one’s identity as a gang member is at the very core of leaving a gang. Several

theories from the broader “desistance from crime” literature emphasize the crucial role of

identity change (e.g., Giordano, Cernkovich, and Rudolph 2002; Maruna 2001). For example, a

good marriage or a quality job provide an opportunity to step away from a deviant view of self

and embrace a positive identity that is reinforced by conventional acts and social relationships

(Laub and Sampson 2003). Typically, measuring shifts in an offender’s identity is challenging,

with qualitative work offering the best insight into internalized labels and perceptions (e.g.,

Giordano, Cernkovich, and Rudolph 2002). When it comes to exiting the gang, however,

assessing at least one dimension of identity change is relatively more straightforward. Upon

joining a street gang, individuals adopt a new label as part of their social identity. When

  9

departing the gang, they leave the label of “gang member” behind. This is a complicated process

with many twists and turns, but self-reported gang status nonetheless can offer an explicit

indicator of whether individuals are taking strides to exit gang life through de-identification. For

youth on a disruptive path marred by commitment to gang life and values, having a child may

prompt a shift in one’s identity and role commitments away from “gang member” and towards

that of parent instead.

Of course, claiming gang membership is not simply a social label without consequence.

Decades of research under numerous methodological specifications demonstrate that gang

membership amplifies offending behavior, including serious and violent crime (for a meta-

analysis of 179 studies, see Pyrooz et al. 2016). Although there is evidence that joining a gang

can have detrimental effects that long outlast the duration of the actual membership, the

amplification effect in offending is most pronounced during the time when individuals are active

in gang life. Therefore, if a life event truly prompts gang exit, there should be marked declines

in criminal activity as part of this transition (Pyrooz and Decker, 2011). This argument reflects

the smorgasbord of illegal activities available to gang members who are “on the set”—present

and available for drug dealing, auto theft, assault and the like. Clearly, the requirements of

parenting could cut into such time. The question is whether such presumed changes in gang

membership and offending behavior should emerge for both male and female gang members.

Motherhood, Gang Membership, and Offending

Motherhood is often portrayed as reflecting the very core of female identity. Edin and

Kafalas (2011) found that economically disadvantaged single mothers viewed motherhood as an

essential part of their identity, providing direction and meaning to their lives. Kreager et al.

(2010) echoed this argument, saying that having a child is viewed as the primary life event

marking the transition to adulthood for women in disadvantaged communities, especially in light

  10

of the fact that marriage and employment are often seen as less attainable. Indeed, they found

that the transition to motherhood reduced offending and substance use among 500 women from

disadvantaged neighborhoods in Denver. Such work suggests that motherhood may offer a

valued new role, one that is largely inconsistent with being a gang member.

Fellow gang members may even encourage the idea that the identities of mother and gang

member are inconsistent. There is evidence that female gang members are often viewed as “soft”

and vulnerable (J. Miller, 2001; W. H. Miller, 2011; Thrasher, 1927); if a female gang member is

pregnant or taking care of an infant, then her capacity to back up fellow members or respond to

threats of violence are significantly hampered (see also Fleisher and Krienert, 2004). In other

words, motherhood may diminish the instrumental value girls offer the gang, which can help

promote an identity transformation away from the “gang member” label. Further, research

suggests that female gang members generally are not as well-embedded in gangs as are male

gang members (Decker, Pyrooz, and Moule, 2014; J. Miller and Decker, 2001; Peterson, Miller,

and Esbensen, 2001). In particular, females report that the gang is less important to them and that

they have less contact with the gang than males do. Breaking the bonds of membership in the

face of so deep a family obligation as having a child may be less difficult for females than it is

for males. Perhaps it is not surprising then that some studies have suggested pregnancy and

motherhood can be the primary motivation for gang de-identification among females. For

example, in Fleisher and Krienert’s (2004) field study of Black female gang members, 64 percent

of women were no longer active gang members said their primary reason for leaving the gang

was pregnancy. Therefore, we hypothesize that within-individual changes in motherhood will be

associated with a reduction in the odds of claiming gang membership.

In thinking about the social changes that accompany parenthood, there is also reason to

suspect that female gang members who become mothers will reduce their offending behavior. In

  11

most cases of parenthood, women serve as the primary caregiver, which fundamentally shifts

routine activities (Sanchez and Thomson, 1997; Thompson and Walker, 1989). In their

qualitative study of 65 women in street gangs who became mothers, Moloney et al. (2011: 1)

concluded: “For many of the young women, motherhood entails a retreat from the street and a

renewed emphasis on time spent in the home”. The constant care an infant requires may mean

that opportunities for criminal activity markedly decline for female gang members, even if the

commitment to deviant values does not. For example, Fleisher and Krienert (2004: 619) found

that even active gang members reported changes in their gang activities upon becoming mothers.

Accordingly, we hypothesize that within-individual changes in motherhood will also be

associated with a reduction in the rate of offending.

Fatherhood, Gang Membership, and Offending

Being part of a gang offers one way for men to embrace marginalized masculinity scripts

to achieve status (Hagedorn 1998; Rios 2011). Fatherhood, alternatively, can provide gang

members with the opportunity to seize conventional masculinity that focuses on being a strong

provider and protector (Flores 2013; Marsiglio and Pleck 2005)—roles that may be inconsistent

with the often present-oriented nature of gang life (Anderson 1990; Fleisher 1998; Katz 1988). In

their interviews with 91 current and former gang members, Moloney et al. (2009) observed that

many of their subjects said becoming a father saved their lives and fundamentally shifted the

way they viewed themselves. As one subject said:

I had never thought about a future before. My future was never important to me just besides day-to-day. It didn’t matter. Now I had to start thinkin’ about, ‘Oh my God, I got a kid. I gotta start thinkin’ about a future’ (p. 314).

Fatherhood can be an opportunity to “start over,” offering men a second chance to live in a more

conventional and productive way (Roy and Lucas 2006).

  12

Yet, young men often resist the constraints on autonomy that can come with fatherhood

(Marsiglio and Cohan 1997). For example, male gang members may view a caretaking role as

inconsistent with what it means to be a man and instead view it as reflective of femininity.

Among such men, fatherhood may spark a renewed commitment to the modes of masculine

identity captured in gang life, particularly when “conventional” masculinity is often frowned

upon by fellow gang members (Flores, 2013).

When turning to a consideration of the relationship between fatherhood and offending,

many of the same theoretical processes that are likely at work with motherhood may also operate

here. Fatherhood can produce incentives, as well as social and financial obligations, that shift

daily routines. Upon having a child, male gang members may find themselves seeking legitimate

employment and helping to care for the child, which can reduce unstructured time and

socializing with friends. Still, after analyzing data from the National Survey of Families and

Households, Sanchez and Thomson (1997, 747) argue that “parenthood crystallizes a gendered

division of labor, largely by reshaping wives’, not husbands’ routines” (see also Thompson and

Walker 1989). Perhaps it is not surprising then that when using data from the Toledo Adolescent

Relationships Study, Giordano et al. (2011) found that within-individual changes in parental

status were related to (reduced) offending for female subjects but not for male subjects (see also

Thornberry, Smith, and Howard 1997). Other work even indicates that fatherhood—especially

young fatherhood—may amplify offending (Unruh, Bullis, and Yovanoff 2003). In the end,

however, to our knowledge no quantitative research has directly examined the relationship

between changes in parental status and offending among male gang members, leaving this an

open question.

Thus, the potential influence of parenthood for male gang members seems less

straightforward. We believe paternal involvement may clarify that influence. After all, the

  13

biological fact of being a parent for males does not necessarily align with the social experience

of being a parent like it does for females. For example, Moloney et al. (2009) note that only one-

third of the 91 fathers who were current or former gang members lived with their children and 15

percent had no contact with their child. In other words, one should not assume that gang

members who are fathers experience the social changes that are thought to be coincident with it.

The Moderating Role of Paternal Involvement

There are numerous studies of the impact of paternal involvement, but the vast majority

of them focus on child outcomes (e.g., Furstenberg, Morgan, and Allison 1987; King 1994). The

inquiries focused on fathers suggest that participation in their children’s lives improves the

father-child relationship and may also increase the father’s psychological well-being (Russell

and Radin 1983). In thinking about the role of father involvement, however, it is important to

recognize that it can be, and has been, operationalized in numerous ways (e.g., financial support,

time spent with the child). Lamb (2000) argued that paternal involvement should be thought of as

extending to three dimensions: (1) engagement, which refers to one-on-one interactions, (2)

accessibility, which refers to a father’s availability even when not directly engaged in a shared

activity, and (3) responsibility, which refers to ownership over decisions related to the child’s

care. Researchers often opt for relatively simple proxies, however, which are perhaps not ideal,

but can still capture variations in paternal involvement. Perhaps the most dominant of these is

the residential status of the father, which Nelson (2004) argues shapes the degree to which

fathers are involved in their children’s lives. As one example, Gee et al. (2007) found that

father-child cohabitation predicted the degree to which fathers were involved in playing with

their child, offering affection to the child, and participating in activities such as feeding the child

and helping to put him/her to bed.

  14

From a theoretical point of view, it is reasonable to suspect that fathers who live with

their child would be more likely to have shifts in their identity and behavior than would non-

residential fathers. Indeed, residential status may reflect a decision to be an involved father,

and/or it could lead to greater involvement. On the one hand, male gang members who embrace

this new role as a parent may actively decide to live with their child while also shifting their

identity and behavioral routines. Conversely, gang members who have no interest in taking on

the role of father are probably unlikely to live with their child or to change their identity/reduce

their offending behavior. At the same time, living with one’s child can have an appreciable

impact on both identity and behavior. Residential fathers have a daily reminder of this social

change, as they most routinely interact in some manner with their child. The consistent presence

of the person who defines a new social role for them may urge a shift in identity towards the

embrace of a vision of self that diverges from one of gang member. Moreover, living with one’s

child also results in greater expectations for help in childcare activities, which would cut into

time that could otherwise be spent on risky activities, such as hanging out with (deviant) friends.

Presumably, non-residential fathers would be less likely to experience such changes.

Although the literature on fatherhood and desistance tends not to consider involvement,

there are two important exceptions using national data. First, Landers, Mitchell, and Coates

(2015) investigated the impact of teenage fatherhood among delinquents and found that

fatherhood reduced offending among those adolescents who lived with their children, but

increased offending for non-residential fathers. Second, Ziegler et al. (2017) also found that

offenders who were residential parents were more likely to desist than were non-parents, but that

non-residential parents did not experience the same benefit. In light of such work and theoretical

expectations, our hypothesis is that we will observe within-individual reductions in the

  15

likelihood of gang membership and offending when gang members live with their child, but that

there will be no such changes for gang members who do not live with their child.

ADDITIONAL NUANCES TO THE PARENTHOOD EFFECT

When studying the potential impact of parenthood on gang exit, it is important to

consider four issues, all of which affect the specifics of the empirical analysis. The first is a

general one that applies to all turning point inquiries, whereas the latter three are specific to a

focus on parenthood. First, turning points are not randomly assigned, which means it is essential

to account for alternative explanations of observed desistance. Unfortunately, extant research

has largely failed to account for important selection processes that may confound the effect of

parental status (see Kreager et al., 2010). Shifting the focus to whether intra-individual changes

in parental status are related to changes in various dimensions of gang exit, rather than

comparing parents to non-parents, helps to reduce these concerns and offers a more rigorous test

of whether having a child produces changes that span identity and behavior. In this way,

individuals serve as their own counterfactual and one is able to focus on within-individual

changes, which is not only consistent with the arguments put forth in this paper, but also with life

course approaches in general.

Second, parenthood is thought to coincide with another major life event that research

consistently relates with desistance from crime: marriage. Indeed, the “civilizing” effect of

marriage is often discussed in terms that suggest a family unit, including children, promotes

change (Popenoe 1996). Still, the percentage of U.S. births to unmarried women has drastically

increased over the past four decades (Ventura 2009). Even though rates of cohabitation have

increased during this same time, it is clear that one should not presume parenthood is inherently

confounded with a stable romantic partnership, especially among certain sub-groups of the

population. This may be reflected in Edin and Kafalas’ (2011) qualitative work, in which women

  16

from disadvantaged communities saw motherhood as the primary marker of adulthood while

marriage was often viewed as unattainable. Even if marriage and parenthood are not always

intertwined, accounting for any potential confound with marriage (or cohabitation) is essential if

one hopes to make inferences about parenthood in particular (Farrington and West 1995). After

all, both parenthood and marriage invoke similar theoretical mechanisms for why there may be

changes in identity and behavioral routines, including social controls, peer processes, time use

patterns, and symbolic interactionism. Moreover, if they do co-occur, there are also indications

that marriage is the “real” turning point, not having children. Even though the Glueck men

identified fatherhood as a factor that motivated change in their lives (Laub and Sampson 2003),

the quantitative analysis did not identify it as a turning point once it accounted for marriage

(Sampson and Laub 1993). Moreover, using data from the National Youth Survey, Warr (1998)

found that individuals who were unmarried with children had no change in their time spent with

friends or exposure to deviant peers, and that married persons with children were not

significantly different on these domains from married persons without children. This led him to

conclude that marriage, not having children, altered socialization routines and promoted

desistance. Yet, gang members offer a different social context in which to test the independent

impact of parenthood, a context that may offer a different salience to this life transition (Kreager,

Matsueda, and Erosheva 2010).

Third, the research on the impact of parenthood has usually focused on the transition to

parenthood, that is, the birth of the first child. In this way, it has equated the effects of becoming

a parent with having a child. To be sure, shifting from being childless to being a parent is a

profound transition—this social distinction is experienced and defined in stark terms, and

arguably more so than having a second child (LaRossa and Sinha 2006). Still, having another

child results in significant caretaking responsibilities, as one has to supervise, support, and care

  17

for two children. Having two children also sends a clear message that one’s role in life has

dramatically changed from those childfree days. As a result, the theoretical mechanisms that

supposedly connect parenthood to gang exit may be invoked even more upon the arrival of a

second child (Stewart 1990). At the same time, however, second children arrive to more

experienced parents (Whiteman, McHale, and Crouter 2003), which may dampen the starkness

of this experience. In the end, distinguishing whether the hypothesized mechanisms that may

promote gang exit are tied to the initial transition to parenthood and/or to having a(ny) child rests

in considering the birth of more than only the first child. No study has considered the impact of

becoming a parent for the second time on offender outcomes; doing so would lend theoretical

insight on the manner by which parenthood is (or is not) connected to gang exit.

Fourth, it is important to recognize that the experience of parenthood may not be

consistent over time, as the demands of parenthood change (Ziegler et al. 2017). The initial

“shock” of having a child is often a more challenging experience than anticipated (Barnes and

Balber 2007). Newborns require round-the-clock care, which substantially changes parental

routines. These sudden, encompassing caretaking demands may also reinforce one’s sense that

their social position in life has changed and that their gang member identity may not be

compatible with this new role. The question is whether such anticipated changes endure over

time, or are instead tied to the shock of the immediate experience. On the one hand, as time

passes, one arguably has time to invest in the social bond to the child, to acclimate to

conventional expectations, and to embrace the parental identity. In this way, it is reminiscent of

arguments that the marriage effect endures because of greater social investments (Laub, Nagin,

and Sampson 1998). On the other hand, the need for constant supervision and care generally

diminish as the child ages, which may offer more freedom to engage in routines less oriented

around family life (e.g., socializing with friends). Interestingly, like the overall effect of

  18

parenthood on gang exit, this too may be gendered, as research suggests that father involvement

tends to wane as the child ages (Cutrona et al. 1998; Fagot et al. 1998). Even though mothers are

primary caretakers, the immediate parental demands may be so encompassing as to require help,

which the father may fulfill. As those demands wane, however, mothers may continue to focus

on caretaking while fathers find that their load lessens. This is speculative, of course, but it

certainly underscores the importance of considering the timing of the parenthood effect across

gender to improve understanding of the gang exit process.

METHODS

DATA

Data from the National Longitudinal Survey of Youth 1997 (NLSY97) were used to

determine whether changes in parental status corresponded with disengagement from gangs and

desistance from crime. The NLSY97, sponsored by the Bureau of Labor Statistics and

coordinated by the National Opinion Research Center, is representative of youth born between

1980 and 1984 and living in the United States at the time of the study. A stratified multi-stage

probability sampling design was used to generate a sample of 8,984 respondents. The first stage

relied on standard area-probability sampling methods using over 90,000 housing units. The

second stage relied on the selection of a target sample of age-eligible youth. The NLSY97 data

consist of a “cross-sectional” sample (N=6,748) designed to achieve representativeness through

self-weighting and a “supplemental” sample (N=2,236) of black and Hispanic youth.

Respondents in the NLSY97 were first surveyed in 1997, when they were between ages 12 and

18 years. There have been 15 additional rounds of data collection through 2013. With high rates

of retention throughout the study (>80%), the data are appropriate to conduct the type of analyses

needed to address our research questions. The NLSY97 data are well-established in

criminological research generally (Siennick 2011; Sweeten, Bushway, and Paternoster 2009;

  19

Uggen and Thompson 2003) and gang research particularly (Bellair and McNulty 2009;

Bjerregaard 2010; Pyrooz and Sweeten 2015).

Given our theoretical interests, we restricted the data to a subgroup of respondents with a

history of gang membership. Measures of gang membership were included in waves 1 to 9, and

676 respondents self-reported membership in a gang at least once during that period. We further

restricted our sample by orienting the data around the first observed report of gang membership.

Similar to Sweeten and colleagues (2013), we track a sample of active gang members and

examine changes in their identity and behavior over time. A final restriction was made to focus

only on respondents without any children prior to the first observed report of gang membership

because it was impossible to disentangle the effects of parenthood from that of gang

membership. After these cuts, our final analytic sample consists of 629 respondents born

between 1980 and 1984 without children prior to the onset of gang membership and is used to

examine the effects of parenthood prior to age 26.1

These data provide several opportunities to bolster confidence in making inferences with

respect to testing our hypotheses and to expand beyond prior inquiries of parenthood as a

potential turning point. First, the NLSY97 contain as many as nine panels of data, which permits

fixed effects modeling and thus controls for time-invariant selection effects. Second, the data

also contain time-varying information on other (positive) social events that may coincide with

changes in parenthood, most notably marriage/cohabitation, which allows us to estimate the

independent effects of parenthood. Third, the data also allow us to model both identity and

behavioral dimensions of gang exit that may change with parenthood. Fourth, because the

                                                                                                                          1 The 47 respondents who were parents prior to the onset of gang membership were statistically indistinguishable from the 629 respondents who were not across the following factors: offending (at gang onset and throughout the study period), illegal earnings (at gang onset and throughout the study period), gender, age cohort, Hispanic, Asian, Native American, and generational status. The excluded respondents differed in terms of gang membership (35.7% vs. 27.4%), age at first observed gang membership (19.6 vs. 16.4 years), the prevalence of births across the study period, the total number of children (2.4% vs. 0.7%), and percent black (55.3% vs. 37.4%).

  20

NLSY97 has information on birth parity, we can extend past prior work by considering the

effects of having a first and a second child. Finally, our modeling approach will distinguish

between immediate and delayed effects of parenthood.

MEASURES

Dependent Variables

Gang membership is a dichotomous indicator of whether a respondent self-reported

membership in a gang across the recall period of a given study wave. Respondents were asked if

they “ever belonged to a gang in the last 12 months” (baseline interview) or “since the date of

last interview” (waves 2 to 9). Those who responded “yes” to these questions were coded 1 and

those who responded “no” were coded 0, which is consistent with a long line of research that

measures gang membership using mutually exclusive binary attributes. Self-nomination is a

reliable and valid method to distinguish respondents who are currently, formerly, or have never

been in a gang (Curry 2000; Decker et al. 2014; Esbensen et al. 2001; Webb, Katz, and Decker

2006). Table 1 reveals that respondents’ self-reported membership in a gang 27.4 percent of the

completed interviews between waves 1 and 9, with an intra-class correlation of 0.075.

Offending is a summated category scale composed of self-reports to six measures of

offending behavior, including the intentional destruction of property, theft exceeding $50, other

property-related crimes (e.g., selling/receiving stolen property), selling illegal drugs, carrying a

hand gun, and attacking or seriously assaulting someone. Respondents were asked the number of

times they committed the offenses in the last year (baseline interview) or since the date of the

last interview. Each of the six items were recoded to four response categories (0=never, 1=one

time, 2=two-three times, 3=four-nine times, and 4=10 or more times),2 then summed to create a

                                                                                                                          2 The scaling and recall period for gun carrying focused on the number of days a gun was carried in the last 30 days. The response options were: 0 (never), 1 (one day), 2 (two-three days), 3 (4-29 days), and 4 (30 days).

  21

scale ranging from 0 to 24. This scale is advantageous because it ensures that our outcome is not

overly dominated by less serious forms of offending, has sufficient variation, and is found to be a

valid scale that closely approximates criminal propensity (Sweeten, 2012: 554). Beginning in

wave 8, these measures were restricted to a subgroup of respondents, which is why our analyses

of offending are limited to waves 1 to 7. Respondents scored an average of 3.2 on the offending

scale across the study period (Table 1), with an intra-class correlation of 0.256.

Independent Variables

Our theoretical interests rest in how changes in parenthood status among a sample of

active gang members relate to changes in gang membership and offending. There are two

distinguishing features of the measurement strategy we employ to test the effects of becoming a

biological parent on these outcomes. First, we account for birth parity, that is, distinguishing the

influences of first-time parenthood from second-time parenthood. Table 1 indicates that over half

of the gang members did not have children during the study period, while 26 percent had only

one child. Another 14 percent of the sample reportedly had their first and second child during the

study period.3 It is important to be able to distinguish the unique effects of these births to

determine if one, or both, are driving changes in behavior and identity. Second, we distinguish

between the immediate effect of becoming a parent, what we term a birth shock, from the later

effect of being a parent, what we term a post-birth lag, for first and second births.

Indicators of parenthood are drawn from the NLSY97 created variables for the birth of

biological children, which was in continuous month format and then recorded within a wave

recall period (and within 12 months of the baseline interview) based on the month of an

interview. If an interview and a birth shared the same month, we assume the birth fell within the                                                                                                                           3 While 5 and 9 percent of males and females become a parent for a third time during the study period, the follow up period to observed lagged effects was limited due to right censoring. There were 29 persons and only 54 person- waves for the post-birth lag period, which contributed to large standard errors and greater uncertainty in detecting trends.

  22

recall period. Every person-wave observation was coded according to the parental status of the

subject at that period: (1) not a parent during this wave (i.e., pre-birth period); (2) became a

parent for the first time in this wave (i.e., birth shock #1); (3) is a parent of one child during this

wave, but it is not the year of first child’s birth (i.e., post-birth lag #1); (4) became a parent for

the second time in this wave (i.e., birth shock #2); and, (5) is a parent of two children during this

wave, but it is not the year of the second child’s birth (i.e., post-birth lag #2). In order to

consider the relationship between within-individual change in these parental status categories

and within-individual changes in gang membership and offending, we converted the latter four

parental status categories into dummy variables: birth shock #1, post-birth lag #1, birth shock #2,

and post-birth lag #2. For example, birth shock #1 is coded 1 the wave in which the first child

was born and 0 for all other person-waves. Likewise, post-birth lag #1 is coded 0 until the wave

after the child’s birth, when it switches to being coded 1, then returns to 0 the wave the second

child is born if there is indeed an additional birth. An identical coded scheme is used for all birth

shock #2 and post-birth lag #2. Importantly, none of the parenthood status variables could be

coded 1 simultaneously. Because pre-parenthood person-waves are the excluded category, the

effects of these parenthood statuses are interpreted in reference to when the subjects were not

parents. This approach allows us to parse out the unique effects of the distinct parenthood

statuses, rather than dividing the lives of gang members into aggregate periods of before and

after parenthood using one variable. We provide an illustration of our dummy coding schema in

the Supplemental Text.

We report the results of two additional forms of coding and modeling strategies in the

Supplemental Text: (1) we estimated the pooled effects of parenthood, in which all periods after

the birth of a child are given a value of 1 (i.e., we did not distinguish between the immediate and

delayed effects) until an additional birth, using fixed effects modeling, and (2) we estimated the

  23

additive effects of the parenthood sequences using a discontinuous or piecewise growth curve

modeling strategy (Singer and Willett 2003). On the whole, the takeaway from these approaches

is not substantively different from what we report below, but offers an alternative way of

thinking about the shocks and lags associated with changes in parental status.

Some scholars may wonder if mechanisms set in motion by pregnancy, rather than the

birth of a child, are responsible for the possible for producing disengagement from gangs and

desistance from crime. Unfortunately, the research design of the NLSY97 presents

methodological challenges to disentangle pregnancy from parenthood. Indeed, the 287 first births

observed during the study period occurred, on average, 10 months after the prior wave’s

interview, which means that both pregnancy and birth occurred within a single recall period. In

fact, only about 23 percent of females entered into their third trimester of pregnancy at the time

of the prior wave’s interview. Should pregnancy, and not parenthood, function as the turning

point for gang members, our results would be conservative. However, supplemental findings

suggest that this is not this case.4

Moderator Variable

The NLSY97 data contain information at each wave about the relationship between the

respondents and the persons with whom they live. We use this variable to examine if fatherhood

had differential effects on behavior and identity based on whether the child lived with her or his

father. There were 90 possible relationship types. Respondents were coded “1” if they lived with

their biological daughter or son and “0” otherwise, which was used to construct indicators of

residential and non-residential fatherhood birth shocks and post-birth lags. We concentrate only

                                                                                                                          4 Supplemental analysis revealed that a time-varying measure of pregnancy (females) and impregnation (males) increased the likelihood of gang membership for females (b=1.69, p<.05) but not males (b=0.11, p=.65), and was unrelated to offending for both genders (F: b=0.06, p=0.78; M: b=0.13, p=.20). It is worth noting that there is considerable missing data on pregnancy/impregnation, as the NLSY97 restricts these items to respondents who exceed age 14 and self-report to having had sex.

  24

on males because motherhood almost always involves living with the child. Indeed, a child

resided with the mother 92 percent of the time on the birth wave and 89 percent of the remaining

waves, whereas a child resided with the father only 50 percent of the time on the birth wave and

41 percent of the remaining waves.

Control Variables

We include a series of time-varying covariates to control for spuriousness. First, in order

to estimate an effect of changes in parental status that is independent from the effects of changes

in romantic partnership status, we control for cohabitation, that is, living with a spouse or

romantic partner (=1, not living with spouse or romantic partner=0), based on household member

roster data.5 Given that other positive life events may also occur at the same time as parenthood,

we also account for whether the subject is in school, (=1, not attending school=0), and legally

employed (=1, not legally employed=0) at each wave. Linear and quadratic time trends,

measured as waves from first observed membership in a gang (e.g., 0=wave 1, 1=wave 2, …),

and months between waves to account for the exposure period, were also included as control

variables.

-- Table 1 --

ANALYTIC STRATEGY

We track a sample of 629 active gang members over as many as nine years to examine if

changes in parental status correspond with changes in gang membership and offending. Given

that our theoretical interests focus on intra-individual change, and that we have longitudinal data

containing interviews nested within people, we fit a series of fixed effects models in Stata 14.0.

                                                                                                                          5 There is evidence that cohabitation and marriage exhibit differential impacts on desistance from crime, with marriage exerting stronger effects (Forrest 2014; Monsbakken, Lyngstad, and Skardhamar 2013). Marriage was very rare of among gang members, which led us to combine cohabitating with a romantic partner and spouse. None of the females reported a single instance of marriage during active periods of gang membership, while there were only 10 person-pooled instances of marriage among males. When distinguishing cohabitation and marriage as time-varying covariates, the sign and significance of the effects of parenthood on gang membership and offending did not change.

  25

Each individual serves as her or his own counterfactual, where we focus on whether the effects

of birth shocks and lags differ from pre-birth periods, net of linear and quadratic time trends as

well as observable time-varying covariates. This identification strategy further guards our

findings from unobservable time-stable individual characteristics (Allison 2009). Logistic and

negative binomial regression is used because gang membership and offending are binary and

count variables, respectively.6 Naturally, respondents with no variation in parental status do not

contribute to the birth shock and lag results, nor do respondents contribute to the within-

individual results without variation in the outcomes.7

We proceed by further describing our sample, with a special emphasis on the differences

and similarities among male and female gang members. Next, we report the results of our fixed

effects models, again comparing the differential effects of parenthood by gender. Finally, in light

of the variation in whether the biological child was in the respondent’s residence, we consider

whether there are differential effects of parenthood among male gang members based on

residential status, which is our proxy for father involvement. Given our theoretical emphasis on

parenthood functioning as a turning point for gang members, our hypotheses are directional,

which leads us to interpret p values below .10 as statistically significant.

RESULTS

Changes in parental status were observed regularly among the sample of male and female

gang members. There were nearly 4,100 person-waves included in our data—2,950 for males,

1,131 for females—and a parental status change occurred in slightly over 11 percent of them.                                                                                                                           6 We assume the effect of rejoining a gang is the inverse of leaving, although our sample—starting with active gang members—ensures that our results overwhelmingly reflect the effects of parenthood on leaving. There are nearly six times the number of instances of leaving (N=690) than rejoining (N=121) a gang. There were no differences in parenthood, gang membership, or offending between multi-wave intermittent or continuous gang members.   7 We lose cases for 37 respondents who were gang members at all waves and 49 respondents who reported no offending at all waves. None of the 629 respondents lacked variation on both outcomes. A mixed effects modeling strategy with time-stable control variables nonetheless produced results that were statistically and substantively similar to the fixed effects results, and we elected to report the latter, which, as an anonymous reviewer noted, was more consistent with our theoretical interests.

  26

Recall that we track our respondents beginning with their first observed report of gang

membership and that none of them had children prior to the onset of gang membership.

Therefore, we are capturing 460 instances of where a male or female with a history of gang

membership became a parent. Females were more likely than males to become parents

throughout the study period; both the prevalence and frequency of births were statistically

greater for female than male gang members. The majority of male gang members did not have

children (58 percent), whereas the majority of female gang members did have children (56

percent). Given that males constituted the bulk of gang members (74 percent) in our data, the

prevalence of births among female gang members only underscores the appropriateness of

gender-partitioned analyses.

Male and female gang members also differed across our outcomes. Males spent more

person-waves in gangs than females—a difference of eight percentage points. Males also

engaged in more offending behavior than females—a difference of about one-quarter standard

deviation. These results are devoid of an explicit linkage to parenthood. Indeed, the results could

mean that male gang members may have more to gain from parenthood than females or that

parenthood did not influence male gang members in positive ways like it did for females.

Although both hypotheses are plausible, as we have outlined above in our theoretical framework,

we turn to our fixed effects models which are better capable of identifying whether parenthood

has beneficial influences on our outcomes and whether those effects are immediate and enduring.

PARENTHOOD EFFECTS BY GENDER

Table 2 reports the results from the fixed effects models that examine the immediate and

lagged effects of parenthood on gang membership. The results from the full sample indicate that

becoming a first-time parent maintained an effect on gang membership that was negative and

statistically significant at the one-tailed level (b=-0.68, p=0.08). This effect was fleeting,

  27

however, as respondents were just as likely to be gang members during the post-birth lag period

as they were during the pre-birth period. This suggests that the effects of parenthood are

beneficial, but temporary, and confined to becoming a parent for the first time.

Disaggregating the results by gender reveals a different story, however. Upon becoming a

first-time parent, females witnessed an immediate 93 percent (100*[1-e-2.65]) reduction in their

likelihood of gang membership that was enduring—the shock and lag effects are virtually

identical. The second child, however, did not correspond to similar reductions for females,

although the coefficients for the first and second birth shocks and lags, respectively, were

statistically indistinguishable (Wald test statistic: shock 𝜒! = 0.53, p=.47; lag 𝜒! = 0.37, p=.54).

Males did not incur any benefits from parenthood in terms of lowering the likelihood of gang

membership, whether captured as having a first or second child. Thus far the results suggest that

parenthood is beneficial in knifing off identities closely linked to proximal and long-term

criminal and non-criminal consequences, but only for females and not males. Equality of

coefficient tests (Clogg, Petkova, and Haritou 1995) reveal no statistically significant differences

between males and females in the effect of birth shock #1 on gang membership (and offending),

but we urge caution in placing too much confidence in the test.8

-- Table 2 here –

Table 3 contains the results of three fixed effects models relating offending to parenthood

in the full sample and then partitioned by gender. As with the model on gang membership, there

is a statistically significant aggregate effect associated with the shock of becoming a parent for

the first time on offending. But, there is also an immediate effect of becoming a parent for the                                                                                                                           8 This conclusion could be misleading for two reasons. First, by disaggregating parenthood into shocks (birth #1: N*T=313) and lags (birth #1: N*T=503), we have reduced power to detect statistically significant differences. As we report in the Supplemental Text, pooling periods of parenthood increases power (birth #1 N*T=816) and revealed statistically significant gender differences in the coefficients for our key parenthood status measures. Second, there is disagreement about the appropriateness of using equality of coefficients tests in generalized linear models (Allison 1999; Long 2009; Mood 2010; Williams 2009), which is what we use in this study.

  28

second time. Like gang membership, the effects of parental status changes on offending appear

to be fleeting; neither of the post-birth lag effects are distinguishable from the pre-birth period.

Disaggregating the results by gender reveals that parenthood maintains beneficial effects

for females but null effects for males. The expected rate of offending falls by 47 percent

(100*[1-e-0.63]) among females with the birth of a first child, yet remains statistically similar to

pre-parenthood periods of offending among males (b=-.08, p=0.56). Further, the immediate

effects of first-time parenthood on offending for females appear to be enduring, as there is a

negative and statistically significant post-birth #1 lag effect. Further, the birth #2 shock is also

negative and statistically significant, which suggests universal benefits of parenthood in reducing

offending for females across birth parity. That said, although the unstandardized coefficient for

the post-birth #2 lag is as large as both birth #1 coefficients, it does not differ statistically from

zero.9

Table 3 also illustrates that, much like the parenthood effects on gang membership, there

were no statistically significant differences in the offending model between the shock and lag

coefficients for the first and second births for females (Wald test statistic: shock 𝜒! = 0.93,

p=.33; lag 𝜒! = 0.10, p=.75). These results offer a second measure of confidence that

parenthood is functioning as a turning point for female but not for male gang members.

While our framework did not argue that the impact of motherhood on offending (only)

depends on changes in gang membership, but it would be instructive to know if our observed

motherhood effect(s) on offending is mediated in any way through gang membership. The

addition of gang membership as a time-varying mediating variable did not alter the substantive

                                                                                                                          9 It is important to remind readers that among females who became parents, 49 percent had only one child and there were 123 person-periods to identify the post-birth #1 lag effect. There were fewer than half the person-periods (N*T=52) to identify the post-birth #2 lag effect for the remaining women who had more than one child.  

  29

or statistical significance of the offending effect among females associated with birth shock #1

(∆b=-13%, p=.04), post-birth lag #1 (∆b=+1%, p=.095), or birth shock #2 (∆b=+11%, p=.02)

-- Table 3 --

THE MODERATING EFFEECT OF FATHER INVOLVEMENT

If turning points require fundamental shifts in social bonds, informal social controls,

routines, and self-concept, fatherhood in and of itself may not universally function as a “hook for

change.” Some fathers may maintain an active presence in the life of their child, while others do

not (Mollborn and Lovegrove, 2011), which might correspond to differences in how fatherhood

influences the dimensions of desistance (see Landers, Mitchell, and Coates, 2015). We therefore

further evaluate the potential for parenthood to function as a turning point for residential and

non-residential fathers.

The results presented in Table 4 largely confirm our suspicion that the impact of

fatherhood on male gang members may depend on their involvement. The results show that if

fatherhood has the potential to promote positive change among gang members, it is most likely

to occur among first-time fathers who reside with their child.10 Net of the time trend and time-

varying control variables (note that we had to exclude cohabitation as a covariate because of its

statistical overlap with residential fatherhood), we observed that gang members who become

fathers and lived with their child experienced immediate reductions in the likelihood of gang

membership (72 percent: 100*[1-e-1.26]) and the rate of offending (40 percent: 100*[1-e-0.50]). We

did not observe similar benefits of parenthood for criminal behavior or identity among male gang

members who did not reside with their child.                                                                                                                           10 We also considered whether marriage/cohabitation moderated the relationship. We did not find any moderating effects of cohabitation on gang membership, but did for offending (b=-.63, p=.04). However, cohabitation and residential fatherhood are highly correlated (r=0.75, p<.01). Only 10 fathers cohabitated but do not live with their child (compared to 68 who do) and 13 fathers who do not cohabitate and live with their child (compared to 96 who do not). The 96 fathers who neither cohabitate nor live with their child were the highest risk group, while the remaining groups were similar—we urge caution when interpreting off-diagonals with small cell sizes.

  30

The positive effect of fatherhood for male gang members who live with their children is

consistent in sign and significance and on par in magnitude with the effects of motherhood on

gang membership and offending. Unlike the results for female gang members, which revealed

that the parenthood effect was enduring, the benefits of fatherhood for residential fathers appear

to be fleeting. Indeed, none of the post-birth lag variables maintained effects that were

distinguishable statistically from the pre-birth periods in gang membership and offending. The

fatherhood effect was also tied to the transition to parenthood (i.e. the first child) rather than to

having a second child. As with the motherhood effect on offending, we investigated whether the

fatherhood effect on offending (for residential fathers) was mediated by changes in gang

membership. There is some evidence that this effect may be operating through the attendant

shifts that come with de-identifying as a gang member (∆b=-20%, p=.04).

Figure 1 provides a graphical summary of our key results. This graph only focuses on the

“shock” of first birth and its effects on the likelihood of gang membership and expected rate of

offending (reporting one-tailed confidence intervals), but it nonetheless nicely captures the

overarching story from the multiple models. Clearly, becoming a parent has a much more

dramatic impact on females than on males, particularly for gang membership. For instance, there

is a 93 percent reduction in the likelihood of gang membership for female gang membership, but

a statistically null 28 percent reduction among male gang members. Similarly, the expected rate

of offending drops by 47 percent when a female gang member becomes a parent, but a

statistically null 8 percent reduction among male gang members. The graph also illustrates the

importance of differentiating between male gang members who live with their children and those

who do not. The predicted difference in the effects of parenthood on gang membership and

offending for males compared to females shrinks considerably for residential fathers, while it

remains statistically null for non-residential fathers.  

  31

-- Figure 1 --

DISCUSSION

In light of its impact on behavioral, economic, psychological, and social outcomes, it is

possible to frame parenthood as one of the most dramatic transitions a person can experience.

Because this powerful change can promote connections to conventional roles and behavior and

can impose changes in routine activities, it has naturally been considered as a possible turning

point in the lives of offenders. Recent work offers persuasive evidence that having a child can

promote desistance among certain offenders (e.g., Kreager, Matsueda, and Erosheva, 2010). For

years, scholars have raised the possibility that parenthood serves as an important turning point

for a particular group of active and problematic offenders: gang members. Unfortunately, extant

empirical work does not offer clear or consistent evidence on this front, due in no small part to

the conceptual and methodological complexities that such a convincing analysis imposes, which

has limited our understanding about the process of leaving the gang.

Researchers have spent decades identifying the micro- and macro-level factors that

promote gang membership, resulting in a solid knowledge base regarding selection into street

gangs, including the motivations, risk factors, processes, and consequences (Decker, Melde, and

Pyrooz 2013; Densley 2015; Esbensen et al. 2010; Howell and Griffiths 2015; Klein and Maxson

2006). In comparison, our knowledge regarding the other side of membership—exiting gang

life—is notably underdeveloped. Recent years certainly have seen growth in empirical attention

to this issue (e.g., Carson and Vecchio, 2015; Pyrooz, Decker, and Webb, 2014; Sweeten,

Pyrooz, and Piquero, 2013), but it nonetheless has not yet established whether certain life events

promote transitioning out of gang life. Compared to established turning points identified in the

life course literature—marriage, stable employment, and military service (Laub and Sampson,

2001)—parenthood is among the most probable life events encountered by and available to gang

  32

members, particularly to the youth who constitute the bulk of gang members. There is some

support for a parenthood effect in qualitative work, but the underdeveloped status of the

quantitative research is notable.

The current study investigated whether within-individual changes in parental status were

related both to changes in gang membership status and offending across male and female gang

members in the NLSY97 dataset. In doing so, we believe it offers a rigorous assessment of

whether parenthood promotes gang exit. After all, because of the role of gang identity, elevated

levels of criminal activity, and susceptibility to group process, gang members represent a “high

bar” for measuring the effects of parenthood on criminal trajectories.

Our findings suggested that parenthood is an important factor in the lives of gang

members, but that its impact varies in important ways. Most notably, there are clear gender

differences in the parenthood effect. Becoming a mother for the first time is related to a large

drop in two dimensions of gang life—the likelihood of claiming gang membership and the rate

of offending. Importantly, these effects are lasting ones, suggesting that motherhood promotes

changes that “stick.” Interestingly, having a second child promotes an immediate reduction in

offending, but this change is temporary. What might this mean? The transition from being

childless to being a mother is a profound one—it results in sudden and encompassing changes, as

one must shift focus towards the constant demands of raising a child. This dramatic change may

fundamentally shift how a female gang member sees herself; if she embraces the conventional

identity of “mother” and sees this a marker of adulthood, then that should be a lasting change

that is not affected by how many children one has. However, having more than one child

undoubtedly increases the amount of time spent on caretaking activities, at least initially when a

mother must balance the needs of a newborn with her other child’s demands. In this way, the

opportunities for offending may be especially limited, even if having prior parental experience

  33

could allow women to adapt more quickly. Though speculative, this may explain why the

beneficial effect of a second child is temporary and only affects offending.

The NLSY97 data confirm that women are often primary caregivers, given that almost all

female gang members lived with their children. In contrast, only about half of the male gang

members lived in the same residence as their child(ren). Given that living with one’s child may

be an indicator of both parental involvement and responsibility, we investigated whether it

conditioned the effect of parenthood on male gang members. The results suggested that there

were stark differences: male gang members who became fathers and lived with their child

demonstrated significant reductions in their likelihood of claiming gang membership and

offending rate. In contrast, we observed no beneficial effect of fatherhood for male gang

members who did not live with their child. Clearly, the answer to whether fatherhood promotes

leaving the gang hinges on the level of parental involvement. In other words, fatherhood appears

to have little impact on its own; actively being a parent is the hook for change. Thus, we should

be careful to not treat fatherhood as a monolithic experience.

It is also important to note that this beneficial effect for residential fathers only emerged

for transitioning to being a father for the first time (i.e., there was no effect of becoming a father

for the second time) and that, unlike for motherhood, this benefit was not enduring (i.e., the

lagged effects were not statistically significant). Both of these findings may speak to the fact

that fathers, even those involved in the lives of their children, rarely take lead responsibility in

caring for their children. When a first child is born, the initial parenting demands may be so

numerous, encompassing and constant that they spill over and affect a father living in the same

house. For a man who has no experience with fatherhood, this immediate change upon the

child’s birth may be so jarring and dramatic that it ushers in both identity and behavioral change.

  34

As the child ages, however, and the constant demands wane somewhat, fatherhood may simply

not be as disruptive to daily routines and outside social connections.

These findings expand our understanding of why and how individuals exit gang life, but

they also clearly also raise questions that should guide future work. First, we raised several

potential mechanisms both in framing our inquiry and in interpreting the findings. There is sound

life course theory underlying our suggestions, but the fact remains that we were unable to

determine with precision the mechanisms responsible for the observed relationships. Clearly, an

important next step for criminological research on this topic would be to isolate where shifts in

routines, personal roles, and other factors are linked in theorized ways to parental status changes.

Perhaps most notably, research should account for the role of peer processes, including time

spent with friends and whether such friends are fellow gang members and/or involved in

offending. In a seminal piece on the mechanisms of desistance, Warr (1998) demonstrated in the

National Youth Survey that marriage promoted changes in peer socialization, including a notable

decline in time spent with friends and exposure to criminally-involved friends. Unfortunately,

the NLSY97 did not collect peer information in a systematic enough manner to estimate within-

individual changes in our models. Investigating whether peer risk factors change in meaningful

ways that mirror the effects we observe here—in terms of gender, timing, and birth parity—

would go a long way in specifying the mechanisms that connect parenthood to gang exit.

Indeed, delving more deeply into the day-to-day social consequences of parenthood among gang

members more generally would not only be theoretically insightful, but it would also offer

critical guidance to service providers on the ground who are implementing the programs to

facilitate such positive changes.

Second, the parameters of the NLSY97 data ensured that we studied parenthood at a

young age—teenage years through emerging adulthood. On the one hand, this is fitting, as gang

  35

membership tends to wane dramatically after young adulthood. Therefore, if one seeks to

understand gang exit, this is an appropriate timeframe. On the other hand, research suggests that

delayed parenthood has known benefits for parents (e.g., Mirowsky and Ross, 2002). In this way,

though sample size of gang members would certainly decline, if future research could extend

their focus beyond the mid-twenties, it may reveal even stronger protective effects.

Third, our data were advantaged by having information on whether male gang members

resided with their children, allowing us to uncover the moderating effect of father involvement.

While paternal residence is an often-employed proxy for involvement, the fact remains that it

does not offer detail about the many dimensions of paternal involvement (Lamb 2000; see also

Nelson 2004). If future work were able to capture factors such as financial support, participation

in childcare, time spent with the child, and responsibility in making parental decisions, this could

further elucidate our understanding of the process whereby some male gang members manage

both their parental responsibilities and their exit from gang life.

Fourth, our investigation focused on the potential benefits of parenthood for gang

members, but as the research on intergenerational offending reminds us (Thornberry 2009),

researchers would be remiss if at some point they did not also consider the children born into

these contexts. Individuals who join street gangs typically have enduring risk factors that

promoted their initial membership and likely accrue several additional deleterious consequences

that go along with membership itself (Thornberry et al., 2003). As a result, they may not have

particularly strong parenting skills or support networks, which may unfortunately disadvantage

the child (Augustyn, Thornberry, and Krohn, 2014). In other words, the stage may be set for

intergenerational transmission of gang membership and its associated behaviors in part because

of persistent exposure of children to risk factors for gang membership (Howell and Egley Jr.

2005).

  36

In conclusion, parenthood is an important turning point that life course researchers turn

to when contrasting time-varying event and state dependencies against time-stable population

heterogeneity (Nagin and Paternoster 2000). Our study represents an important and needed

extension on this front, as it suggests that parenthood does indeed matter even for a group of

offenders engaged in substantial and serious forms of criminal activity. Most importantly, this

research has taken strides to settle divergent findings by considering the role of identity and

behavior changes for both male and female gang members. Even so, we see this as an initial

rather than final step to understanding how and why parenthood matters to a population that is

are central to criminological theory and research and criminal justice policy and practice.

  37

REFERENCES

Akers, Ronald L. 2009. Social Learning and Social Structure: A General Theory of Crime and Deviance. Piscataway, NJ: Transaction Publishers.

Allison, Paul D. 1999. “Comparing Logit and Probit Coefficients across Groups.” Sociological Methods & Research 28 (2): 186–208.

———. 2009. Fixed Effects Regression Models. Vol. 160. Quantitative Applications in the Social Sciences. Thousand Oaks, CA: Sage.

Anderson, Elijah. 1990. Streetwise: Race, Class, and Change in an Urban Community. Chicago: University of Chicago Press.

Augustyn, Megan Bears, Terence P. Thornberry, and Marvin D. Krohn. 2014. “Gang Membership and Pathways to Maladaptive Parenting.” Journal of Research on Adolescence 24 (2): 252–67.

Barnes, Diana Lynn, and Leigh G. Balber. 2007. The Journey to Parenthood: Myths, Reality and What Really Matters. Abingdon, UK: Radcliffe Publishing.

Bellair, Paul E., and Thomas L. McNulty. 2009. “Gang Membership, Drug Selling, and Violence in Neighborhood Context.” Justice Quarterly 26 (4): 644–69. doi:10.1080/07418820802593394.

Bjerregaard, Beth. 2010. “Gang Membership and Drug Involvement: Untangling the Complex Relationship.” Crime & Delinquency 56 (1): 3–34. doi:10.1177/0011128707307217.

Blokland, Arjan AJ, and Paul Nieuwbeerta. 2005. “The Effects of Life Circumstances on Longitudinal Trajectories of Offending.” Criminology 43 (4): 1203–1240.

Bolden, Christian. 2013. “Tales from the Hood: An Emic Perspective on Gang Joining and Gang Desistance.” Criminal Justice Review, 0734016813509267.

Carson, Dena C., and J. Michael Vecchio. 2015. “Leaving the Gang: A Review and Thoughts on Future Reearch.” In The Handbook of Gangs, edited by Scott H Decker and David C. Pyrooz, 257–75. Chichester, West Sussex: Wiley-Blackwell.

Clogg, Clifford C., Eva Petkova, and Adamantios Haritou. 1995. “Statistical Methods for Comparing Regression Coefficients between Models.” American Journal of Sociology, 1261–1293.

Curry, G. David. 2000. “Self-Reported Gang Involvement and Officially Recorded Delinquency.” Criminology 38 (4): 1253–74. doi:10.1111/j.1745-9125.2000.tb01422.x.

Curry, G. David, Scott H. Decker, and David C. Pyrooz. 2014. Confronting Gangs: Crime and Community. Third edition. New York: Oxford University Press.

Cutrona, Carolyn E., Robert M. Hessling, Pamela L. Bacon, and Daniel W. Russell. 1998. “Predictors and Correlates of Continuing Involvement with the Baby’s Father among Adolescent Mothers.” Journal of Family Psychology 12 (3): 369.

Decker, Scott H., Chris Melde, and David C. Pyrooz. 2013. “What Do We Know about Gangs and Gang Members and Where Do We Go from Here?” Justice Quarterly 30 (3): 369– 402. doi:10.1080/07418825.2012.732101.

Decker, Scott H., David C. Pyrooz, and Richard K. Jr Moule. 2014. “Disengagement from Gangs as Role Transitions.” Journal of Research on Adolescence 24 (2): 268–283.

Decker, Scott H., David C. Pyrooz, Gary Sweeten, and Richard K. Jr Moule. 2014. “Validating Self-Nomination in Gang Research: Assessing Differences in Gang Embeddedness across Non-, Current, and Former Gang Members.” Journal of Quantitative Criminology 30 (4): 577–98. doi:10.1007/s10940-014-9215-8.

  38

Densley, James A. 2015. “Joining the Gang: A Process of Supply and Demand.” In The Handbook of Gangs, edited by Scott H. Decker and David C. Pyrooz, 235–56. Chichester, West Sussex: Wiley-Blackwell.

Edin, Kathryn, and Maria Kefalas. 2011. Promises I Can Keep: Why Poor Women Put Motherhood before Marriage. Berkeley: University of California Press.

Edin, Kathryn, Timothy Jon Nelson, and Rechelle Paranal. 2001. Fatherhood and Incarceration as Potential Turning Points in the Criminal Careers of Unskilled Men. Evanston, IL: Institute for Policy Research, Northwestern University.

Esbensen, Finn-Aage, Dana Peterson, Terrance J. Taylor, and Adrienne Freng. 2010. Youth Violence: Sex and Race Differences in Offending, Victimization, and Gang Membership. Philadelphia: Temple University Press.

Esbensen, Finn-Aage, L. Thomas Winfree, Ni He, and Terrance J. Taylor. 2001. “Youth Gangs and Definitional Issues: When Is a Gang a Gang, and Why Does It Matter?” Crime & Delinquency 47 (1): 105–30. doi:10.1177/0011128701047001005.

Fagot, Beverly I., Katherine C. Pears, Deborah M. Capaldi, Lynn Crosby, and Craig S. Leve. 1998. “Becoming an Adolescent Father: Precursors and Parenting.” Developmental Psychology 34 (6): 1209.

Farrington, David P., and Donald J. West. 1995. “Effects of Marriage, Separation, and Children on Offending by Adult Males.” In Current Perspectives on Aging and the Life Cycle, edited by Zena Smith Blau, 4:249–281. Greenwich, CT: JAI Press.

Feldman, S. Shirley, and Sharon Churnin Nash. 1984. “The Transition from Expectancy to Parenthood: Impact of the Firstborn Child on Men and Women.” Sex Roles 11 (1–2): 61– 78.

Fleisher, Mark S. 1998. Dead End Kids: Gang Girls and the Boys They Know. Univ of Wisconsin Press.

Fleisher, Mark S., and Jessie L. Krienert. 2004. “Life-Course Events, Social Networks, and the Emergence of Violence among Female Gang Members.” Journal of Community Psychology 32 (5): 607–622.

Flores, Edward. 2013. God’s Gangs: Barrio Ministry, Masculinity, and Gang Recovery. New York: NYU Press.

Forrest, Walter. 2014. “Cohabitation, Relationship Quality, and Desistance from Crime.” Journal of Marriage and Family 76 (3): 539–56. doi:10.1111/jomf.12105.

Furstenberg, Frank F. Jr, S. Philip Morgan, and Paul D. Allison. 1987. “Paternal Participation and Children’s Well-Being after Marital Dissolution.” American Sociological Review, 695–701.

Gee, Christina B., Christopher M. McNerney, Michael J. Reiter, and Suzanne C. Leaman. 2007. “Adolescent and Young Adult Mothers’ Relationship Quality during the Transition to Parenthood: Associations with Father Involvement in Fragile Families.” Journal of Youth and Adolescence 36 (2): 213–224.

Gibson, Chris L., and Marvin D. Krohn. 2013. The Handbook of Life Course Criminology. Springer New York.

Gilman, Amanda B., Karl G. Hill, and J. David Hawkins. 2014. “Long-Term Consequences of Adolescent Gang Membership for Adult Functioning.” American Journal of Public Health 104 (5): 938–945.

Giordano, Peggy C., Stephen A. Cernkovich, and Jennifer L. Rudolph. 2002. “Gender, Crime, and Desistance: Toward a Theory of Cognitive Transformation.” American Journal of Sociology 107 (4): 990–1064.

  39

Giordano, Peggy C., Patrick M. Seffrin, Wendy D. Manning, and Monica A. Longmore. 2011. “Parenthood and Crime: The Role of Wantedness, Relationships with Partners, and SES.” Journal of Criminal Justice 39 (5): 405–416.

Goldstein, Arnold P., and Barry Glick. 1994. The Prosocial Gang: Implementing Aggression Replacement Training. Thousand Oaks, CA: Sage.

Hagedorn, John M. 1994. “Homeboys, Dope Fiends, Legits, and New Jacks.” Criminology 32 (2): 197–219.

———. 1998. People and Folks: Gangs, Crime and the Underclass in a Rustbelt City. 2nd ed. Lake View Press.

Hirschi, Travis. 1969. Causes of Delinquency. Berkeley: University of California Press. Howell, James C., and Arlen Egley Jr. 2005. “Moving Risk Factors into Developmental Theories

of Gang Membership.” Youth Violence and Juvenile Justice 3 (4): 334–54. doi:10.1177/1541204005278679.

Howell, James C., and Elizabeth Griffiths. 2015. Gangs in America’s Communities. 2nd ed. Thousand Oaks, CA: Sage.

Katz, Jack. 1988. Seductions of Crime. New York: Basic Books. King, Valarie. 1994. “Nonresident Father Involvement and Child Well-Being: Can Dads Make a

Difference?” Journal of Family Issues 15 (1): 78–96. Klein, Malcolm W. 1995. The American Street Gang: Its Nature, Prevalence and Control. New

York, NY: Oxford University Press. Klein, Malcolm W., and Cheryl L. Maxson. 2006. Street Gang Patterns and Policies. New York:

Oxford University Press. Kreager, Derek A., Ross L. Matsueda, and Elena A. Erosheva. 2010. “Motherhood and Criminal

Desistance in Disadvantaged Neighborhoods.” Criminology 48 (1): 221–258. Krohn, Marvin D., Jeffrey T. Ward, Terence P. Thornberry, Alan J. Lizotte, and Rebekah Chu.

2011. “The Cascading Effects of Adolescent Gang Involvement across the Life Course.” Criminology 49 (4): 991–1028.

Lamb, Michael E. 2000. “The History of Research on Father Involvement: An Overview.” Marriage & Family Review 29 (2–3): 23–42.

Landers, Monica D., Ojmarrh Mitchell, and Erica E. Coates. 2015. “Teenage Fatherhood as a Potential Turning Point in the Lives of Delinquent Youth.” Journal of Child and Family Studies 24 (6): 1685–1696.

LaRossa, Ralph, and Cynthia B. Sinha. 2006. “Constructing the Transition to Parenthood.” Sociological Inquiry 76 (4): 433–457.

Laub, John H. 2006. “Edwin H. Sutherland and the Michael-Adler Report: Searching for the Soul of Criminology Seventy Years Later.” Criminology 44 (2): 235–58. doi:10.1046/j.1365-2281.2001.0382a.x.

Laub, John H., Daniel S. Nagin, and Robert J. Sampson. 1998. “Trajectories of Change in Criminal Offending: Good Marriages and the Desistance Process.” American Sociological Review, 225–238.

Laub, John H., and Robert J. Sampson. 2001. “Understanding Desistance from Crime.” Crime and Justice, 1–69.

———. 2003. Shared Beginnings, Divergent Lives: Delinquent Boys to Age 70. Cambridge, MA: Harvard University Press. http://psycnet.apa.org/psycinfo/2003-88395-000.

Lizotte, Alan J., Matthew D. Phillips, Marvin D. Krohn, Terence P. Thornberry, Shawn D. Bushway, and Nicole M. Schmidt. 2015. “Like Parent like Child? The Role of Delayed Childrearing in Breaking the Link between Parent’s Offending and Their Children’s

  40

Antisocial Behavior.” Justice Quarterly 32 (3): 410–44. doi:10.1080/07418825.2013.771206.

Long, J. Scott. 2009. “Group Comparisons in Logit and Probit Using Predicted Probabilities.” Department of Sociology, University of Indiana. Unpublished Manuscript. http://www.indiana.edu/~jslsoc/files_research/groupdif/groupwithprobabilities/groups- with-prob-2009-06-25.pdf.

Marsiglio, William, and Mark Cohan. 1997. “Young Fathers and Child Development.” In The Role of the Father in Child Development, edited by Michael E. Lamb, 3rd ed., 227–44. Hoboken, NJ: John Wiley & Sons.

Marsiglio, William, and Joseph H. Pleck. 2005. “Fatherhood and Masculinities.” The Handbook of Studies on Men and Masculinities, 249–269.

Maruna, Shadd. 2001. Making Good: How Ex-Convicts Reform and Rebuild Their Lives. Washington, DC: American Psychological Association.

McGloin, Jean Marie, and Stephanie DiPietro. 2013. “Girls, Friends and Delinquency.” In The Oxford Handbook of Criminological Theory, edited by Francis T. Cullen and Pamela Wilcox, 294–312. New York: Oxford University Press.

Melde, Chris, and Finn-Aage Esbensen. 2011. “Gang Membership as a Turning Point in the Life Course.” Criminology 49 (2): 513–52.

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

Miller, Jody, and Scott H. Decker. 2001. “Young Women and Gang Violence: Gender, Street Offending, and Violent Victimization in Gangs.” Justice Quarterly 18 (1): 115–40. doi:10.1080/07418820100094841.

Miller, Walter H. 2011. City Gangs. Phoenix, AZ: Arizona State University, School of Criminology and Criminal Justice. http://gangresearch.asu.edu/walter_miller_library/walter-b.-miller-book.

Mirowsky, John, and Catherine E. Ross. 2002. “Depression, Parenthood, and Age at First Birth.” Social Science & Medicine 54 (8): 1281–1298.

Mollborn, Stefanie, and Peter J. Lovegrove. 2011. “How Teenage Fathers Matter for Children: Evidence from the ECLS-B.” Journal of Family Issues 32 (1): 3–30.

Moloney, Molly, Geoffrey P. Hunt, Karen Joe-Laidler, and Kathleen MacKenzie. 2011. “Young Mother (in the) Hood: Gang Girls’ Negotiation of New Identities.” Journal of Youth Studies 14 (1): 1–19. doi:10.1080/13676261.2010.506531.

Moloney, Molly, Kathleen MacKenzie, Geoffrey Hunt, and Karen Joe-Laidler. 2009. “The Path and Promise of Fatherhood for Gang Members.” British Journal of Criminology 49 (3): 305–25.

Monsbakken, Christian Weisa eth, Torkild Hovde Lyngstad, and Torbjørn Skardhamar. 2013. “Crime and the Transition to Parenthood the Role of Sex and Relationship Context.” British Journal of Criminology 53 (1): 129–148.

Mood, Carina. 2010. “Logistic Regression: Why We Cannot Do What We Think We Can Do, and What We Can Do about It.” European Sociological Review 26 (1): 67–82.

Moore, Joan. 1991. Going down to the Barrio: Homeboys and Homegirls in Change. Philadelphia: Temple University Press.

Nagin, Daniel, and Raymond Paternoster. 2000. “Population Heterogeneity and State Dependence: State of the Evidence and Directions for Future Research.” Journal of Quantitative Criminology 16 (2): 117–44. doi:10.1023/A:1007502804941.

Nelson, Timothy J. 2004. “Low-Income Fathers.” Annual Review of Sociology 30: 427–451.

  41

O’Neal, Eryn Nicole, Scott H. Decker, Richard K. Moule, and David C. Pyrooz. 2014. “Girls, Gangs, and Getting out: Gender Differences and Similarities in Leaving the Gang.” Youth Violence and Juvenile Justice, September, 1541204014551426. doi:10.1177/1541204014551426.

Osgood, D. Wayne, Janet K. Wilson, Patrick M. O’malley, Jerald G. Bachman, and Lloyd D. Johnston. 1996. “Routine Activities and Individual Deviant Behavior.” American Sociological Review, 635–655.

Peterson, Dana, Jody Miller, and Finn-Aage Esbensen. 2001. “The Impact of Sex Composition on Gangs and Gang Member Delinquency.” Criminology 39 (2): 411–40. doi:10.1111/j.1745-9125.2001.tb00928.x.

Popenoe, David. 1996. Life without Father: Compelling New Evidence That Fatherhood and Marriage Are Indispensable for the Good of Children and Society. Simon and Schuster.

Pyrooz, David C. 2014. “‘From Your First Cigarette to Your Last Dyin’ Day’: The Patterning of Gang Membership in the Life-Course.” Journal of Quantitative Criminology 30 (2): 349– 72. doi:10.1007/s10940-013-9206-1.

Pyrooz, David C., and Scott H. Decker. 2011. “Motives and Methods for Leaving the Gang: Understanding the Process of Gang Desistance.” Journal of Criminal Justice 39 (5): 417– 25. doi:10.1016/j.jcrimjus.2011.07.001.

Pyrooz, David C., Scott H. Decker, and Vincent J. Webb. 2014. “The Ties That Bind: Desistance from Gangs.” Crime & Delinquency 60 (4): 491–516. doi:10.1177/0011128710372191.

Pyrooz, David C., and Gary Sweeten. 2015. “Gang Membership Between Ages 5 and 17 Years in the United States.” Journal of Adolescent Health 56 (4): 414–19. doi:10.1016/j.jadohealth.2014.11.018.

Pyrooz, David C., Gary Sweeten, and Alex R. Piquero. 2013. “Continuity and Change in Gang Membership and Gang Embeddedness.” Journal of Research in Crime and Delinquency 50 (2): 239–71. doi:10.1177/0022427811434830.

Pyrooz, David C, Jillian J. Turanovic, Scott H. Decker, and Jun Wu. 2016. “Taking Stock of the Relationship between Gang Membership and Offending: A Meta-Analysis.” Criminal Justice & Behavior 43 (3): 365–97.

Rios, Victor M. 2011. Punished: Policing the Lives of Black and Latino Boys. New York: NYU Press.

Roy, Kevin M. 2006. “Father Stories: A Life Course Examination of Paternal Identity among Low-Income African American Men.” Journal of Family Issues 27 (1): 31–54. doi:10.1177/0192513X05275432.

Roy, Kevin M., and Kristen Lucas. 2006. “Generativity as Second Chance: Low-Income Fathers and Transformation of the Difficult Past.” Research in Human Development 3 (2–3): 139–159.

Russell, Graeme, and Norma Radin. 1983. “Increased Paternal Participation: The Father’s Perspective.” In Fatherhood and Family Policy, edited by Michael E. Lamb and Abraham Sagi, 139–165. Hillsdale, NJ: Lawrence Erlbaum Associates.

Sampson, Robert J., and John H. Laub. 1990. “Crime and Deviance over the Life Course: The Salience of Adult Social Bonds.” American Sociological Review, 609–627.

———. 1993. Crime in the Making: Pathways and Turning Points through Life. Cambridge, MA: Harvard University Press.

———. 2016. “Turning Points and the Future of Life-Course Criminology: Reflections on the 1986 Criminal Careers Report.” Journal of Research in Crime and Delinquency 53 (3): 321–335.

  42

Sanchez, Laura, and Elizabeth Thomson. 1997. “Becoming Mothers and Fathers: Parenthood, Gender, and the Division of Labor.” Gender & Society 11 (6): 747–772.

Sanchez-Jankowski, Martin. 1991. Islands in the Street: Gangs and American Urban Society. Berkeley: University of California Press.

Savolainen, Jukka. 2009. “Work, Family and Criminal Desistance Adult Social Bonds in a Nordic Welfare State.” British Journal of Criminology 49 (3): 285–304.

Short, James F., and Fred L. Strodtbeck. 1965. Group Process and Gang Delinquency. Chicago: University of Chicago Press Chicago.

Siennick, Sonja E. 2011. “Tough Love? Crime and Parental Assistance in Young Adulthood.” Criminology 49 (1): 163–195.

Singer, Judith D., and John B. Willett. 2003. Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. Oxford University Press.

Stewart Jr, Robert B. 1990. The Second Child: Family Transition and Adjustment. Newbury Park, CA: Sage.

Sweeten, Gary. 2012. “Scaling Criminal Offending.” Journal of Quantitative Criminology 28 (3): 533–557.

Sweeten, Gary, Shawn D. Bushway, and Raymond Paternoster. 2009. “Does Dropping out of School Mean Dropping into Delinquency?” Criminology 47 (1): 47–91.

Sweeten, Gary, David C. Pyrooz, and Alex R. Piquero. 2013. “Disengaging from Gangs and Desistance from Crime.” Justice Quarterly 30 (3): 469–500.

Thompson, Linda, and Alexis J. Walker. 1989. “Gender in Families: Women and Men in Marriage, Work, and Parenthood.” Journal of Marriage and the Family, 845–871.

Thornberry, Terence P. 2009. “The Apple Doesn’t Fall far from the Tree (or Does It?): Intergenerational Patterns of Antisocial Behavior—the American Society of Criminology 2008 Sutherland Address.” Criminology 47 (2): 297–325.

Thornberry, Terence P., Marvin D. Krohn, Alan J. Lizotte, Carolyn A. Smith, and Kimberly Tobin. 2003. Gangs and Delinquency in Developmental Perspective. New York: Cambridge University Press.

Thornberry, Terence P., Carolyn A. Smith, and Gregory J. Howard. 1997. “Risk Factors for Teenage Fatherhood.” Journal of Marriage and the Family, 505–522.

Thrasher, Frederic M. 1927. The Gang: A Study of 1,313 Gangs in Chicago. Chicago: University of Chicago Press.

Uggen, Christopher, and Melissa Thompson. 2003. “The Socioeconomic Determinants of Ill-­‐ gotten Gains: Within-­‐person Changes in Drug Use and Illegal Earnings.” American Journal of Sociology 109 (1): 146–85. doi:10.1086/ajs.2003.109.issue-1.

Umberson, Debra, Tetyana Pudrovska, and Corinne Reczek. 2010. “Parenthood, Childlessness, and Well-Being: A Life Course Perspective.” Journal of Marriage and Family 72 (3): 612–29. doi:10.1111/j.1741-3737.2010.00721.x.

Unruh, Deanne, Michael Bullis, and Paul Yovanoff. 2003. “Community Reintegration Outcomes for Formerly Incarcerated Adolescent Fathers and Nonfathers.” Journal of Emotional and Behavioral Disorders 11 (3): 144–56. doi:10.1177/10634266030110030201.

Varriale, Jennifer A. 2008. “Female Gang Members and Desistance: Pregnancy as a Possible Exit Strategy.” Journal of Gang Research 15 (4): 35–64.

Ventura, Stephanie J. 2009. Changing Patterns of Nonmarital Childbearing in the United States. NCHS Data Brief, No 18. Hyattsville, MD: National Center for Health Statistics. https://www.cdc.gov/nchs/products/databriefs/db18.htm.

  43

Wakefield, Sara, and Chris Uggen. 2004. “Having a Kid Changes Everything? The Effect of Parenthood on Subsequent Crime.” presented at the Annual Meeting of the American Sociological Association, San Francisco, CA.

Warr, Mark. 1998. “Life-Course Transitions and Desistance from Crime.” Criminology 36 (2): 183–216.

———. 2002. Companions in Crime: The Social Aspects of Criminal Conduct. Cambridge: Cambridge University Press.

Webb, Vincent J., Charles M. Katz, and Scott H. Decker. 2006. “Assessing the Validity of Self- Reports by Gang Members: Results from the Arrestee Drug Abuse Monitoring Program.” Crime & Delinquency 52 (2): 232–252.

Whiteman, Shawn D., Susan M. McHale, and Ann C. Crouter. 2003. “What Parents Learn from Experience: The First Child as a First Draft?” Journal of Marriage and Family 65 (3): 608–621.

Whyte, William F. 1943. Street Corner Society. Chicago: University of Chicago Press. Williams, Richard. 2009. “Using Heterogeneous Choice Models to Compare Logit and Probit

Coefficients across Groups.” Sociological Methods & Research 37 (4): 531–559. Ziegler, Jessica A., Danielle C. Kuhl, Raymond R. Swisher, and Jorge M. Chavez. 2017.

“Parenthood Residency Status and Criminal Desistance across Neighborhood Contexts.” Deviant Behavior 38 (1): 17–33.

  44

Table 1. Descriptive Statistics for the Study Variables by Gender Full Sample Females Males

N= 629 163 466 Sig. Mean (SD) Mean (SD) Mean (SD) Diff. Outcome Variables Gang membership a 27.4% 21.6% 29.6% * Offending b 3.18 (5.08) 2.29 (3.99) 3.52 (5.41) * Parenthood Variables Births (time varying) a 11.3% 13.7% 10.3% * Number of Births c 0.73 (0.98) 0.95 (1.05) 0.65 (0.94) * 0 54.4% 44.2% 57.9% 1 25.8% 27.6% 25.1% 2 13.8% 19.0% 12.0% 3 or more 6.0% 9.2% 5.0% Time-Varying Control Variables Cohabitation a 21.7% 26.0% 20.0% * In school a 33.0% 38.3% 31.1% * Employed a 59.6% 56.7% 60.8% * Months between waves a 13.58 (5.69) 13.31 (4.94) 13.68 (5.95)

NOTES: All descriptive statistics reported for time-varying covariates are post-initial self-reported gang membership during the study period. Statistically significant differences between males and females were determined using independent group t-tests and Pearson’s chi-squared tests.

ABBREVIATIONS: SD = standard deviation; Sig. Diff. = statistically significant difference * p < .05 (two-tailed test) a person-waves 1 to 9, N=4,081 valid person-waves (2,950 males, 1,131 females); b person-waves 1 to 7, N=3,027 valid person-waves (2,187 males, 840 females); c persons N=629 (466 males, 163 females)

45  

Table 2. Logistic Regression Fixed Effects Models Relating Gang Membership to Birth Shocks and Lags by Gender, Waves 1-9

DV: Gang Membership

Full Sample Females Males b (SE) z b (SE) z b (SE) z

Parenthood covariates Birth #1 shock —0.680 (.389) 1.75 # —2.648 (1.420) 1.86 # —0.327 (.413) 0.79 Post-birth #1 lag —0.037 (.423) 0.09 —2.774 (1.438) 1.93 # 0.416 (.451) 1.06 Birth #2 shock 0.263 (.538) 0.49 —1.598 (1.379) 1.16 0.634 (.599) 0.92 Post-birth #2 lag 0.171 (.683) 0.25 —1.775 (0.549) 1.15 0.585 (.770) 0.76 Time-varying controls Linear time trend —2.338 (.124) 18.8 * —2.969 (.334) 8.90 * —2.224 (.139) 15.9 * Quadratic time trend 0.217 (.015) 14.3 * 0.306 (.040) 7.61 * 0.200 (.017) 11.8 * Cohabitation —0.452 (.297) 1.52 —0.217 (.643) 0.34 —0.549 (.340) 1.61 In school 0.243 (.236) 1.03 0.099 (.593) 0.17 0.308 (.263) 1.17 Employed —0.032 (.183) 0.18 —0.982 (.448) 2.19 * 0.177 (.204) 0.87 Exposure 0.026 (.015) 1.73 # —0.043 (.046) 0.93 0.034 (.015) 2.18 * Model 𝜒! 1644* 507* 1140* Persons 556 149 407 Person-Waves 3,414 964 2,450 NOTE: The intra-individual reference group for the parenthood status variables is person-periods in which the

subject had no children (i.e., was not a parent). ABBREVIATIONS: DV = dependent variable; b = unstandardized coefficient; (SE) = standard error; z = z test

statistic (absolute value) # p < .10; * p <.05 (two-tailed test)

46  

Table 3. Negative Binomial Regression Fixed Effects Models Relating Offending to Birth Shocks and Lags by Gender, Waves 1-7

DV: Offending

Full Sample Females Males b (SE) z b (SE) z b (SE) z

Parenthood covariates Birth #1 shock —0.232 (.124) 1.87 # —0.628 (.272) 2.31 * —0.082 (.140) 0.58 Post-birth #1 lag —0.111 (.133) 0.83 —0.449 (.272) 1.65 # 0.027 (.154) 0.18 Birth #2 shock —0.396 (.223) 1.78 # —1.229 (.599) 2.05 * —0.167 (.242) 0.69 Post-birth #2 lag —0.016 (.226) 0.07 —0.631 (.542) 1.16 0.157 (.252) 0.63 Time-varying controls Linear time trend —0.620 (.046) 13.6 * —0.700 (.093) 7.54 * —0.602 (.053) 11.4 * Quadratic time trend 0.065 (.008) 7.79 * 0.072 (.017) 4.18 * 0.064 (.010) 6.71 * Cohabitation —0.159 (.107) 1.49 —0.403 (.213) 1.89 # —0.065 (.125) 0.52 In school 0.052 (.074) 0.70 —0.078 (.157) 0.49 0.072 (.084) 0.86 Employed 0.161 (.062) 2.58 * 0.141 (.126) 1.12 0.166 (.072) 2.31 * Exposure 0.015 (.005) 2.92 * 0.005 (.012) 0.37 0.016 (.005) 2.98 * Model 𝜒! 609* 229* 387* Persons 508 137 371 Person-Waves 2,524 701 1,823 NOTE: The intra-individual reference group for the parenthood status variables is person-periods in which the

subject had no children (i.e., was not a parent). ABBREVIATIONS: DV = dependent variable; b = unstandardized coefficient; (SE) = standard error; z = z test

statistic (absolute value) # p < .10; * p <.05 (two-tailed test)

47  

Table 4. Logistic and Negative Binomial Regression Fixed Effects Models Relating Gang Membership and Offending to Birth Shocks and Lags by Residential and Non-Residential Parenthood, Males Only DV: Gang Membership DV: Offending b (se) z b (se) z Residential father Birth #1 shock —1.262 (.634) 1.99 * —0.503 (.260) 1.93 # Post-birth #1 lag 0.020 (.656) 0.03 —0.078 (.232) 0.34 Birth #2 shock 0.134 (.712) 0.19 —0.394 (.328) 1.20 Post-birth #2 lag —0.608 (.938) 0.65 —0.066 (.325) 0.20 Non-residential father Birth #1 shock —0.125 (.441) 0.28 0.110 (.154) 0.71 Post-birth #1 lag 0.250 (.471) 0.53 0.080 (.182) 0.44 Birth #2 shock 0.484 (.780) 0.62 0.068 (.329) 0.21 Post-birth #2 lag 0.859 (.841) 1.02 0.320 (.341) 0.94 Model 𝜒! 1202** 405** Persons 427 379 Person-Waves 2,668 1,932 NOTE: The intra-individual reference group for the parenthood status variables is person-periods in which the

subject had no children (i.e., was not a parent).All models include controls for linear and quadratic time trends, time-varying covariates (except cohabitation), and months between waves.

ABBREVIATIONS: DV = dependent variable; b = unstandardized coefficient; (SE) = standard error; z = z test statistic (absolute value)

# p < .10; * p < .05 (two-tailed test)

48  

Figure 1. Percent Change in the Odds of Gang Membership and Expected Rate of Offending Linked to the Shock of Becoming a Parent for Females and Males, First Birth Only

Note: Within-individual change reported for the first births only. End caps on the lines represent upper and lower 90% confidence intervals.

-100% -90% -80% -70% -60% -50% -40% -30% -20% -10%

0% 10% 20% 30% 40% 50%

% C

ha ng

e in

th e

O dd

s or

E xp

ec te

d R

at e

Gang Membership Offending

Mothers Fathers Residential Fathers Non-Residential Fathers

View publication statsView publication stats