A Signaling Perspective on Disengagement from Gangs
James A. Densley
Metropolitan State University, USA
David C. Pyrooz
University of Colorado Boulder, USA
*Forthcoming in Justice Quarterly. This is the authors’ pre-print copy of the article. Please
download and cite the post-print copy published on the JQ website:
*James A. Densley is Associate Professor of Criminal Justice at Metropolitan State University.
His research interests include street gangs, criminal networks, violence, and social theory.
Densley is the author of How Gangs Work (Palgrave Macmillan, 2013) and co-author of
Minnesota’s Criminal Justice System (Carolina Academic Press, 2016). David C. Pyrooz is
Assistant Professor of Sociology and Faculty Associate of the Institute of Behavioral Science at
the University of Colorado Boulder. His research interests include gangs and criminal networks,
development and life course criminology, and incarceration and re-entry. He is co-author of
Confronting Gangs (OUP, 2014) and co-editor of the Handbook of Gangs (Wiley, 2015).
Correspondence to: James Densley, Metropolitan State University, School of Law Enforcement
and Criminal Justice, 9110 Brooklyn Blvd., Brooklyn Park, MN 55445, USA. E-mail:
A version of this manuscript was presented at the 2016 Stockholm Criminology Symposium, the
Eurogang XVI workshop, and the 2016 American Society of Criminology annual meeting.
Extant theoretical work on gang exit has focused more on the pushes and pulls that motivate de-
identification, rather than the mechanisms underlying the process of disengagement. How gang
members successfully communicate their unobservable “inner change” to outsiders and
eventually escape the grip of the gang is unresolved. This article reinterprets the current
inventory of knowledge on gang disengagement through the lens of signaling theory. We
position de-identification as private information that the gang member must signal to competing
stakeholders—the gang, rival gangs, crime control agents—in order to successfully progress
through the stages of disengagement established in prior theory and research. We explore how
mechanisms of signaling, screening, and credible commitments underline continuity and change
in levels of gang embeddedness and examine the role of the receivers of signals in
validating disengagement. We conclude with testable hypotheses, the generality of the
model, and associated implications for research and practice.
Keywords: Gangs; Signaling Theory; Desistance; Embeddedness; Delinquency
A longstanding myth about gangs is that gang membership is a lifelong commitment, or
one that is physically costly to cease (Howell, 2007). The truth is that gang membership is
usually a temporary status, only one to two years in duration, although much longer for some
(Melde & Esbensen, 2014; Pyrooz, 2014; Thornberry et al., 2004). Most gang members also
report passive, non-hostile departures (Carson et al., 2013; Decker & Lauritsen, 2002; Decker &
Pyrooz, 2011; Pyrooz & Decker, 2011). Disengagement from gangs is not a rare event, relatively
speaking—about 400,000 youth leave gangs each year in the United States (Pyrooz & Sweeten,
2015). Given that criminal offending rises upon entry into the gang and falls upon exit (Pyrooz et
al., 2016), an entire industry of outreach workers and interventionists, as well as leaders at all
levels of government (Densley, 2011; Tita and Papachristos, 2010), want to know how
disengagement from these purportedly “selective” criminal groups occurs.
While the literature on gang disengagement has grown in recent years (see Carson &
Vecchio, 2015), very little is known about the mechanisms of the disengagement process. In this
article, we advance a signaling perspective on gang disengagement to address this gap. We do
not report empirical data, but rather present a theoretical reinterpretation of the extant literature.
This model treats gang disengagement as a communicative exchange between privately-informed
participants who must arrive at a collective choice whether to accept disengagement or not. The
departing gang member is one of many parties in this exchange. Given that he or she may think
and say they are no longer in the gang, the question becomes: how are those thoughts and words
translated into meaningful action that is trusted by relevant parties? This is not a trivial question;
labeling theorists and gang scholars agree that the gang member’s life, liberty, and pursuit of
happiness are contingent on how others (e.g., rival gangs, police, prospective employers) react to
his or her claims.
Signaling theory has been used extensively to study information asymmetry between
parties (Spence, 2002). Applications are found in a range of disciplines, but we are most
interested in those pertaining to the study of gangs particularly and crime desistance generally
(Bushway & Apel, 2012; Densley, 2012, 2015; Maruna, 2012; Pyrooz & Densley, 2016). Our
central argument is that the “personal transformation” or “inner change” (Maruna, 2012, p. 79) at
the heart of disengagement from gangs is private information that must be communicated
strategically in order for others to develop their perceptions regarding the self-proclaimed “ex”
gang member. Communication occurs through signaling, but also the related practices of
credible commitments made by the sender and screening undertaken by receiving parties. Honest
information is crucial, as it affects the decision-making processes used by gangs and relevant
stakeholders to process this claim. A signaling perspective thus provides answers to two open
questions about gang disengagement: (1) How do “ex” gang members successfully communicate
their identity change to outsiders? (2) What makes disengagement successful, and under what
circumstances? It also provides a rich interdisciplinary framework for conceptualizing
disengagement as a process and not an event, as evidenced in empirical work on the topic.
The sections that follow begin by reviewing the existing literature on gang
disengagement. Next, we introduce signaling theory, including its applications in criminology
and its potential contribution to the study of gang disengagement. We distinguish signaling
theory from comparable theoretical perspectives, while also situating the signaling model within
the broader stages of the gang disengagement process identified in the literature. Following this,
we discuss how disengagement occurs in signaling perspective, namely, the explicit mechanisms
underlying continuity and change in levels of gang embeddedness and the role of the receivers of
signals in validating such continuity and change. We conclude with predictions stemming from
this model, identifying contingencies that are relevant for understanding the model’s generality,
and implications derived from signaling theory for gang disengagement research and practice.
Theory and Research on Disengagement from Gangs
Disengagement from gangs, defined as “the declining probability of gang membership”
(Pyrooz & Decker 2011, p. 419), differs from desistance from crime because gang membership
is a state and crime is an act. Still, extant theories of gang disengagement share a lot in common
with theories of crime desistance (Sweeten, Pyrooz, & Piquero, 2013). Maruna (2000) identifies
three broad theoretical perspectives in the general desistance literature. First, maturational reform
theories have the longest history and are based on the links between age and certain criminal
behaviors, particularly violence (Piquero, Farrington, & Blumstein, 2003; Sweeten, Piquero, &
Steinberg, 2013). Second, social control theories suggest ties to family, education, or
employment in early adulthood create a stake in conformity that explains changes in criminality
across the life course (see Laub & Sampson, 2003; LeBel et al., 2008). Finally, narrative theories
stress the significance of subjective changes in the person’s sense of self and identity, reflected in
shifting motivations, greater empathy for others, future-orientation, and consideration of deferred
gratification (Bushway & Paternoster, 2013).
Sweeten, Pyrooz, and Piquero (2013) held that gang disengagement can be understood as
the event of de-identifying as a gang member and the process of disembedding oneself from the
gang. The motives for de-identifying are well documented—essentially, deleterious gang-related
life experiences and general disillusionment push people out, while familial influences and
legitimate opportunities pull people out (Carson et al., 2013; O’Neal et al., 2016; Pyrooz &
Decker, 2011). These pushes and pulls are understood theoretically much like desistance from
crime, combining maturational reform, fatigue, and cognitive shifts in thinking (Bjorgo, 2002;
Giordano et al., 2002; Decker & Lauritsen, 2002). Bubolz and Simi (2015) present a theoretical
model of gang exit, for example, focused on the formative event of de-identification as opposed
to the summative process of disengagement. Bubolz and Simi argue youth enter into gangs with
grandiose visions of economic, familial, and physical security that gang membership inevitably
fails to satisfy. Members are sold a bill of goods, resulting in a general sense of disillusionment,
but also an incongruence in self-appraisals in identity. This, in turn, leads to negative
emotionality, social distress, and anger, which are necessary to suppress gang roles and promote
non-gang roles. Such is a compelling model, but explains why people disengage from gangs, not
how—it is focused on what Maruna et al. (2009) called “secondary desistance,” or a reform in
Our contention is the motives for leaving the gang are ubiquitous—rarely do gang
members not experience the pushes (e.g., violence, disillusionment) and pulls (e.g., parenthood,
jobs, girlfriends) that trigger de-identification. However, there is little guidance theoretically to
account for false starts and the universality of events that should lead to leaving a gang. To
elaborate, former gang members commonly identify parenthood as a trigger to put gang life
behind (e.g., Decker et al., 2014; Fleisher & Krienert, 2004; Moloney et al., 2009; 2011; O’Neal
et al., 2016), yet there is little theoretical account for why the first baby did not trigger the exit
process, but the second did. In a signaling framework, talk is cheap, which it is why theoretical
weight is not placed on whether parenthood is a motivation but how parenthood is used, which
can therefore account for why a poignant event may not produce the behavioral and identity
changes one would anticipate. Therefore, our interests rest with the process of disengagement
from gangs. That is, from the point when gang members experience the “crystallization of
discontent” (Baumeister, 1991) to the validation of “ex-gang member” status. Situating the
signaling model into this process is a point that we will return to shortly.
There are relatively few theoretical accounts of disengagement as process. Decker et al.
(2014), drawing on Ebaugh’s (1988) theory of role exit, offer a staged-based model predicated
on the notion that disengagement from gangs is not an individual effort, but rather occurs
through social interaction. The initial stage is the generation of “first doubts” about staying in the
role of gang member. At this stage, there is no commitment to leave the gang, but gang members
begin to question its efficacy and consider other options. The second stage involves actively
looking for and weighing alternative roles—known as “anticipatory socialization” (see Merton,
1957). The third stage is the “turning points” period (Carlsson, 2012), where specific experiences
(e.g., violent victimization or prison) or imperatives (family, employment) become “hooks for
change” (Giordano et al., 2002). The final stage is “post-exit validation” from within (e.g., the
individual) and without (e.g., the gang, the community, and agents of social control). Validation
is key because the teeter-tottering between current and ex-gang status is a delicate one, and the
“ties that bind” have the potential to tug individuals back into the gang (Decker et al., 2014;
Pyrooz et al., 2014). This brings us to the potential contribution of a signaling perspective.
Signaling theory tackles a fundamental problem of communication, as Gambetta (2009a,
p. 168) explains: “how can an agent, the receiver, establish whether another agent, the signaler,
is telling or otherwise conveying the truth about a state of affairs or event which the signaler
might have an interest to misrepresent”? If detained by the police or interviewing for a job, for
example, a gang member might have an interest to distance himself or herself from gang
membership. How can the police officer or prospective employer know this person is really an
“ex”? Conversely, how can the “ex” convince the police officer or business owner he is telling
the truth, whether he is telling it or not? This two-pronged question potentially arises every time
the interests between signalers and receivers diverge or collide and there is asymmetric
information—namely, the signaler is in a better position to know the truth than the receiver is.
Signs are perceivable indicators of qualities not directly observable that (1) carry
information from a signaler to a receiver and (2) influence a receiver’s behavior (Gambetta,
2009a). Signals are signs that are purposely or strategically activated or deactivated for this
cause. We may not realize a sign (e.g., a tattoo) is informing others of something about us until
others respond in a such way that makes us aware; at which point, we may choose to display the
sign intentionally, rendering it a signal (Gambetta, 2009a). The 1998 movie American History X
memorably plays with this idea in a scene where Derek Vinyard (Edward Norton) takes off his
shirt in a prison gymnasium to ensure both rival black and confederate white inmates see the
Nazi swastika emblazoned on his chest. Signals reveal some truth, therefore, but some signals
are more reliable than others in doing so. For example, the “soft” markers of gang “embodiment”
that are cheap to change in recovery, such as one’s haircut, tattoos, and clothing (Flores, 2013),
are easy-to-fake signals that, while undoubtedly important, are open to deceptive mimicry. The
“hard” markers of gang life, such as criminal records and official gang classification, substance
abuse, and verbal and nonverbal reactions, are hard-to-fake signals, thus deemed honest.
According to Cronk (2005, p. 611), signals are hard-to-fake for either one or both of two
reasons: (1) they maintain strategic costs that only honest signalers can afford; and (2) they are
direct indices of the underlying trait (i.e., inextricably connected to what they signal). Of course,
such properties are relative to the emitter of the signal. It is cheaper for the person who has a
particular quality to produce a signal of that quality, relative to someone without it. In our model,
for example, a person who has truly disengaged from the gang (or wants to disengage and is in
the process of doing so) has a lower cost for signaling than does a person who is still in the gang
(or is in two minds whether to leave).
The concept of credible commitment, that is, constraining self-interested choice to induce
beneficial actions from others (Campana & Varese, 2013), is also important for understanding
the exchange of signals. A credible commitment is any action undertaken by an agent (i.e., gang
member) to pledge himself or herself to a given course of action (i.e., disengagement) such as by
“burning one’s bridges” or denying himself or herself alternative options later on (Gambetta,
2009b). Credible commitments are signals of a sort, but different because costs are incurred ex-
post if the agent does not follow through on their commitment, whereas signaling is costly ex-
ante. In the context of crisis bargaining, Fearon (1997) influentially distinguished between
signaling and commitments as “sinking costs” and “tying hands.” Credible commitments tie
hands—they allow signal receivers to assess intentions not only in the present, but also in the
Signaling implies that the informed agent, in our case, the gang member, moves first to
reduce the information asymmetry. They signal disengagement and others receive the signal and
act accordingly. Screening, by contrast, puts the onus on those with less information, the
receivers, to screen out erroneous information and learn as much as they can about the gang
member first. Like signaling, screening is a strategy of reducing information asymmetry, but the
emphasis is on the ways the under-informed party (i.e., the stakeholders) can extract or induce
the other party (i.e., the gang member) to reveal their private information (Rothschild & Stiglitz,
1976). Screening and signaling are closely related concepts (see Spence, 1974), therefore, the
latter features in our signaling perspective.
In the tradition of Veblen’s (1899) analysis of conspicuous consumption and Akerlof’s
(1970) explanation of adverse selection, Spence’s (1974) model of job-market signaling showed
how hard-to-fake educational credentials communicated underlying work ethic. Signaling
models have since shown wide disciplinary utility, from anthropology to zoology (for a review,
see Gambetta, 2009a). In criminology specifically, signaling applications have concerned the
function of punishment beatings that deviant youth receive from the Irish Republican Army
(Hamill, 2011), the nuances of prison gang initiation rituals and inmates’ self-harm (Kaminski,
2004), offenders’ strategies to identify bona fide co-offenders (Gambetta, 2009b), and the role of
information and reputation in the mafia (Smith & Varese, 2001). Of most interest to us, signaling
has been used to explain how youth navigate entry into street gangs and offending within them
(Densley, 2012, 2015; Pyrooz & Densley, 2016) and how ex-prisoners communicate successful
rehabilitation to potential employers (Bushway & Apel, 2012; Dewitt, 2016). The current theory
builds on this work, as follows below.
Distinguishing Signaling from Comparable Perspectives
The general consensus in the criminal desistance literature is the causes and correlates of
onset are different from those of continuity and change in offending, what Uggen and Piliavin
(1998) called “asymmetrical causation.” From a signaling perspective, this is true insomuch that
the signals used to facilitate disengagement from the gang differ from those used to gain
admittance to it (see Densley, 2012, 2015). However, signaling theory implies something
different—joining and leaving the gang are symmetrical when one considers both processes in
terms of the need for one agent, the signaler, to communicate (or signal) some information to
another agent, the receiver, who must choose how to interpret the signal. For this reason,
signaling theory can potentially illuminate all stages of disengagement, from the initial first
doubts of continuity in the gang through post-exit validation (Decker et al., 2014). During the
anticipatory socialization stage, for example, gang members signal prosocial traits to exit the
gang in much the same way they signaled antisocial traits to enter it. These signals involve up-
front costs. During the turning points stage, gang members make credible commitments that
imply they will suffer future negative consequences should they renege on disengagement.
During post-exit validation, receivers screen gang members as part of a broader process of
authenticating the commitment to disengagement and the signals that accompany it. We
elaborate upon these ideas in Figure 1, which is outlined in more detail in the following section,
but caution that temporal order remains a testable hypothesis.
Figure 1 A conceptual model of gang disengagement in signaling perspective. Note. The
color gradation in the arrows represent the ramping up and tailing off of mechanisms,
where darker colors indicate greater activity. Signaling, screening, and credible
commitments are interrelated and overlap temporally.
In a response to Bushway and Apel’s (2012) signaling perspective on prisoner re-entry,
Maruna (2012) argued symbolic interactionists might question whether the unobservable quality
of “personal transformation” being signaled truly exists independent of external perceptions.
Signaling certainly downplays the “looking-glass self” (Cooley, 1922) and the self-fulfilling
nature of social interaction that lies at the heart of the labeling theory tradition (Becker, 1963;
Lemert, 1951). It is important to note, however, signaling is about external communication of
conscious intention, not internal self-concept and social identity. Further, our signaling model
positions the gang member primarily as sender rather than the receiver of signals. Indeed, the
gang member is the agent with more information. Labeling theory, alternatively, turns this
around, hence Lemert’s (1951) concept of societal reaction. While both theories draw on notions
of “credentialism” (Collins, 1979), signaling deals in reputation and indicators of future
behavior, whereas labeling concerns “stigma” and social control (Goffman, 1963). Hence why
we side more with Maruna’s (2012, p. 73) other take on desistance signaling as a both “a
revelation” and possibly “a revolution” for criminology.
To this point, the signaling model implies the gang member has the power to control
which signals to reveal. Under more traditional risk factor approaches to desistance, such as the
Risk-Need-Responsivity (R-N-R) model (Andrews & Bonta, 2006), the control agent, typically a
probation officer, determines which signs he or she will assess as having changed (Maruna,
2012). Signaling, therefore, affords gang members much needed agency and provides a model of
disengagement consistent with the common narrative of gang recovery that real change occurs
only when gang members choose to change. This, of course, also diverges from perspectives that
emphasize “desistance by default,” that is, stumbling into rather than purposefully entering into
commitments (Laub & Sampson, 2003, p. 278–79). Unlike acts of crime, states of gang
membership, we contend, require actively shedding gang identity and its attendant web of gang
embeddedness for disengagement to be realized.
Prior studies have explored the tenuous “legitimacy” of gang membership (Lauger, 2012)
and the contested nature of gang disengagement, including the ways in which gang members use
verbal communication and mainstream embodiments, often in public settings, to distance
themselves from gang masculinity and align themselves with conventional notions of manhood
(Flores, 2013). Signaling is different, however, because it focuses on the steps taken to ensure
the message is not only sent, but received. Flores (2013) argues religion drives changes in
masculinity that lead to gang exit, thus, like Decker et al. (2014), he views disengagement as
validated by securing legitimacy in the eyes of others. Signaling helps us appreciate how the eyes
of others can trust what they are seeing. While it complements existing models, we argue that
signaling theory accounts for the ways in which social interactions facilitate gang exit, thus
advancing the notion that “validation” is necessary throughout the disengagement process, not
just “post-exit” (Decker et al., 2014)—an important contribution.
Disengagement from Gangs in Signaling Perspective
Henceforth our primary outcome of interest is embeddedness in gangs, that is, individual
immersion within enduring criminal and deviant networks (Pyrooz et al., 2013). De-
identification as a gang member, we hold, occurs during the disengagement process but is neither
a necessary or a sufficient condition that must be met to observe declining or trivial levels of
embeddedness. Rather, de-identification—what longitudinal research captures when individuals
are coded as gang members in one wave and non-gang members in the successive wave—is part
and parcel of the disengagement process. Qualitative research, and the authors’ observations of
the disengagement process derived from fieldwork and survey research, reveals gang members in
their 40s, 50s, and even older for whom gang membership is an ingrained fixture of their
identity, yet are by no means active gang members, which is what Pyrooz and Decker (2011)
referred to as the veteranos or original gangsters. Instead, our model maintains that gang
disengagement will be realized through gang embeddedness, as a consequence of signals sent
and received, credible commitments made by the senders, and screening undertaken by the
receivers of signals, as outlined below.
A second, but equally important, contention of ours is that an information exchange is
central to the gang disengagement process. To borrow from Matza (1969, p. 112), being willing
to disengage from gangs is not quite the same as being able. Role ambiguity is inevitable unless
the information asymmetry between the gang member and what we term stakeholders is reduced.
Stakeholders can be thought of as the relevant parties orbiting around the gang member who
validate one’s gang status—the gang, rival gangs, family and neighborhood, and the criminal
justice system. We discuss the stakeholders’ role in the disengagement process in detail below.
Suffice it to say, gang membership is accompanied with a series of expectations and obligations
to these stakeholders. Accordingly, a gang member who, despite having committed to at least
exiting cognitively, keeps his or her cards close to the chest will remain enmeshed in the norms
and values of stakeholders who play an important role of releasing the gang member from these
expectations and obligations. But as we noted above, talk is cheap in signaling perspective, but
not talking at all is even cheaper, hence the role ambiguity.
Below we outline the essential features of disengagement from gangs in signaling
perspective, which is represented in Figure 1. We situate the signaling perspective in relation to
Decker and colleagues’ (2014) stages of gang disengagement. We are interested in the generality
of the model—not every gang member will follow all the rules outlined below, but we argue that
signaling, credible commitments, and/or screening are more likely to be present in successful
disengagement narratives. On the primary and secondary y-axes are embeddedness, our principal
outcome variable, and information asymmetry with stakeholders (i.e., receivers), which we
suspect eventually reaches near-zero with respect to disengagement, a point we revisit shortly.
Some mechanisms play a greater or lesser role at different stages, but as we outline below, it is
anticipated that the disengagement process begins with signaling, which is followed first by
screening and next by credible commitments. Even upon activation, there are aspects of these
mechanisms that are displayed earlier or later. Signaling, screening, and credible commitments
occur in the context of specific situations, often when a gang member experiences a practical
conflict between participating in gang life and something else. The examples provided below are
not exhaustive and there may well be others that fit the theory, contingent upon empirical
Owing to the fact that gangs are antisocial groups, stakeholders are looking for gang
members to display prosocial tendencies as evidence that disengagement is real. The display of
prosocial tendencies, however, can be faked. Genuine ex-gang members thus have to develop
costly or hard-to-fake signals of prosocial intentions and propensities to avoid being perpetually
under suspicion from police, probation officers, and others. Hard-to-fake signals are reasonably
honest; they supply others with information that can be used to make social decisions (Cronk,
2005), including information about one’s affiliation with a gang.
Grief and remorse. Owing to the positive relationship between gang membership and
violent offending, violence has been viewed as the elixir of gang life and an honest signal of
utility to the gang that pulls youth in (Densley, 2013). The violence associated with gang
membership, particularly vicarious and direct victimization, however, can also generate feelings
of disillusionment and fatigue that push youth out (Decker & Van Winkle, 1996). Gang members
may be traumatized by the violence they have seen or done (Kerig et al., 2013; Moore, 1991).
They may even regret it. Gang members also grieve loved ones lost to violence. The question for
gang youth is how to communicate they have reached their upper limit on violence or that
violence has gone too far without admonishing the group and bringing that violence back onto
Emotions are often hard to fake and difficult to conceal, thus they are reliable indicators
of a person’s underlying intentions and propensities. Winegard et al. (2014, p. 175) observe, “It
is difficult for people to portray emotional reactions that they are not experiencing, which is one
reason actors are lauded for their abilities.” It is difficult to cry when one is not sad (Frank,
1988), for instance, or to Duchenne smile without true enjoyment (Mehu et al., 2007). Typically,
people experience and express sadness in isolation, thus it is not informative. Public expressions
of sadness, however, become signals that are informative and, we argue, facilitate gang exit.
Hence why gang “recovery” programs, like other 12-step programs (see Carr, 2010), are set up to
facilitate “public talk” (Flores, 2013) that differentiates between honest and dishonest
proclamations of grief and contrition.
Grief is a hard-to-fake signal because its expression is costly and its display conveys
information about the underlying pro-social proclivities of the griever (Winegard et al., 2014).
Ruminating about the dead and other expressions of grief interrupt a person’s ability to perform
daily tasks, but simple professions and consciously controlled displays of grief are easy-to-fake,
thus not so reliable as the more protracted symptoms of grief, such as withdrawal, disinterest,
sadness, and dysfunction. In the hyper-masculine world of the gang (Messerschmidt, 1993),
crying itself is costly, whereas the stylized and scripted “pouring out liquor” for the dead homies
is not (Densley, 2013, p. 165). Prolonged grief limits participation in the gang’s social activities,
increases susceptibility to illness, visits to health care professionals, and even suicidal ideation
(Winegard et al., 2014). Combined, the above are costly displays that diminish the individual’s
fitness for gang membership.
Physical fitness for gang membership. Gang violence leaves physical scars that
accompany the emotional ones. Gambetta (2009b) observes, scars from incised and puncture
wounds or ballistic trauma indicate that someone has been through many fights and has survived.
These are hard-to-fake signals of authenticity and fighting ability that gangs covet (Densley,
2012; Ralph, 2014). Once “battle scars” become debilitating, however, they serve a new purpose
to signal that disengagement with the gang is occurring. A wheelchair, catheter, or prosthesis, for
example, can be activated to signal toughness, “resilience” (Ralph 2014, p. 141), or, importantly,
disengagement, ultimately symbolizing that one’s best days as a gang member are behind them.
A permeant disability is a sign since it does not change at the discretion of the sender—indeed,
one has no choice but to display it—but it becomes a signal of disengagement should the gang
member choose to use it in such a way. Every scar has a story. Depending on how that story is
told, the receiver may begin to question whether the sender still has the physical capacity to
perform the role of gang member and, from an intervention standpoint, whether the perceived
benefits of gang membership are worth the costs (Ralph, 2014). In a similar vein, deliberate self-
harm (DSH) among gang members, such as wrist-cutting, could be another form of signaling, in
the sense it serves as a strategic “cry for help” (see Hagan et al., 2008). Suicidal thoughts
increase upon entry into a gang (Watkins & Melde, 2016). DSH can signal a lot of things
(Gambetta, 2009b) and low-level DSH may be purely manipulative, but life-threatening suicide
attempts, while perhaps rare among gang members, would be genuinely informative.
Renouncement and recovery. Gang “renouncement” programs in prison, such as the Gang
Renouncement and Disassociation program in Texas (Burman, 2012), or “recovery” programs
outside of prison, such as those described by Flores (2013), force gang members to perform
costly behaviors, such as submitting a written statement confirming their desire to renounce their
membership to a particular gang, and submit to laborious rituals and programming designed to
create emotional, physical, and social separation from the gang. Graduation from or successful
completion of such initiatives signals disengagement in the way other ceremonial rights-of-
passage work. There is a fundamental difference, however, between gang members who
voluntarily participate in renouncement and recovery compared to those who do not (Bushway &
Reuter, 1997). While some have questioned the sincerity of prison conversion (Thomas &
Zaitzow, 2006), comparison to the costly rituals and requirements associated with confirmation
in the Church is useful (Sosis, 2003). Not everyone who attends Church decides to enroll in or is
confirmed by the Church. Likewise, not every gang member who goes to prison completes the
renouncement and dissociation process, hence it becomes cost discriminating.
Education and training. Bushway and Apel (2012) maintain that in the absence of perfect
information about potential employees, employers can reliably look to signals like education and
skills training completion as indicators of the otherwise unobservable quality of commitment to
legitimate work. Bushway and Apel found that offenders who volunteer to participate in a
rigorous and challenging employment program, and especially those who successfully complete
such a program, signal that they are fundamentally different from other offenders who do not
choose to participate in programming. Voluntary completion of post-prison employment
programs served as a hard-to-fake “desistance signal” used to assess offenders (Maruna, 2012).
We suspect the same rules apply for gang members seeking to make a change. Moreover,
education and training credentials signal a reversal of previous “incompetence” in non-gang
endeavors, which Gambetta (2009b, p. 42) argued, initially allowed offenders to signal to other
potential co-offenders how their legitimate options outside of gang work were severely limited.
In other words, education and training beyond that which is compulsory signal disengagement to
both potential employers and former gang colleagues.
As work with ex-offenders demonstrates, it can take several years or more to truly
ascertain whether an individual with a criminal record has definitively desisted from crime
(Kurlychek, Brame, & Bushway, 2006). Beyond signaling his or her intent to disengage,
therefore, a gang member must also commit to the actions he or she has signaled they will take.
For these commitments to be believed, they must be credible. Gang members achieve such
credibility through consistency of action over time, and the kind of costly signaling that exhibits
their will to disengage. Subscribers to the transtheoretical model, the dominant model of health
behavior change, call this “action” and “maintenance” (Prochaska, Norcross, & DiClemente,
1994). There is a reason probation or parole conditions in the criminal justice system include
rehabilitative terms, such as the attending therapy, submitting to random drug testing, avoiding
places and/or people, requiring that the individual not commit another crime, and when possible,
the maintaining of gainful employment and/or education. Credible commitments cut off gang-
related options either by “burning bridges” (see Gambetta, 2009b) or by building bridges to
viable gang alternatives.
Burning bridges. Michael Corleone’s (Al Pacino) famous line from The Godfather: Part
III— “Just when I thought I was out…they pull me back in”—captures how many gang members
feel when they re-engage with the gang against their will. This is what former gang members
describe as the “teeter tottering” in the disengagement process (Decker et al., 2014). Declaring
oneself an “ex” gang member is not the same as having no gang ties. Even for those who leave
the gang, ties to gang life remain in the form of family, friends, and neighborhood affiliations
(Pyrooz et al., 2014). Not to mention ties to the gang's online community, which can be far more
difficult to delete (Storrod & Densley, 2017). Less willingness to reach consensus, more
destructive comments, and fewer sacrifices for the group (Wiselquist et al., 1999) can signal
relative independence, but they do not sever the “ties that bind” in such a way that the gang
member is committed to life outside the group. When a gang member burns bridges, however,
there is no going back (Gambetta, 2009b).
Perhaps the best example of burning bridges pertains to the prison system’s debriefing
procedure, by which gang members renounce their gang membership and divulge gang-related
information in order to earn release from segregation back into the prison’s general population or
some other privilege (see Burman, 2012). In an extra-legal world where “snitches get stitches”
and the consequences for informing on other gang members can be catastrophic (Whitman &
Davis, 2007), debriefing is a commitment that potentially incurs costs into perpetuity. Indeed,
collateral violence is one of the reasons why debriefing is so controversial. Another example of
burning bridges would be moving to a new city, neighborhood, or school, thus “knifing-off” all
ties to the gang (Decker & Pyrooz, 2011). The up-front costs associated with moving and starting
over signal something, but it is the deferred costs of separation from family, friends, and
community that make this an ongoing, and credible, commitment to disengagement.
Burning bridges is extremely costly. It is probably reserved for those gang members
looking to make an immediate exit from gangs—akin to “cold turkey” drug cessation (i.e.,
without tapering of lowered doses)—or gang members who were once more embedded in gangs,
challenged to escape the “grip of the group” (van Gemert & Fleisher, 2005). Few gang members
will burn bridges to the point of no return. Indeed, there are a handful of urban ethnographers
who, themselves, were gang members at one point in their lives then went on to study their own
peer group or other gang members (e.g., Duran, 2013; Rios, 2011). But this constitutes the
exception and not the norm, and even then, the bridge between the gang and the ex-gang
member, while not fully incinerated, resembles the rickety bridge with fallen boards over
troubled waters you see in a Hollywood adventure.
Building bridges. “Nothing stops a bullet like a job,” was once the slogan of Homeboy
Industries, a celebrated gang intervention project (Boyle, 2011), and it aptly captures the theory
that “turning points,” such as getting a job, redirect criminal paths toward more prosocial
outcomes by means of informal social control (Laub & Sampson, 2003). In crude terms, a job
gives a gang member a reason to get up in the morning and a reason not to “do gang” the night
before. The challenge of course is that active gang members also have jobs. We argue, therefore,
that a job per se is not the mechanism for disengagement, but rather the mechanism is the
credible commitment to disengagement that certain jobs convey to others. Jobs that are, at the
very least, stable and full-time, but also specifically orientated toward helping oneself and others,
live out the promise of disengagement.
While there are no estimates of the size of the gang industry, programs like Homeboy
Industries and ex-gang members in intervention programs (Cheng, 2017; Densley, 2011; Lopez-
Aguado, 2013; Kennedy, 2011; Klein, 2011) show that street gang outreach work has
experienced a resurgence since Klein (1971) notably questioned its utility. Whether intervention
“works” is moot in signaling perspective because work in gang intervention commits ex-gang
members to disengagement, which alone is informative for two reasons. First, “By voluntarily
forgoing delinquent opportunities … [gang members] signal to potential spouses and employers
that it is safe to build them into their future plans in an orderly and effective manner” (Massoglia
& Uggen, 2010, p. 552). Second, by “sharing one’s experiences, strength, and hope”, mentoring
others, modelling recovery, and, for some, “making a career of helping others who are not as far
along in the recovery and/or reintegration process”, gang members “give something back”
(LeBel et al., 2015, p. 110–11) in the mode of other costly altruistic acts (see Bowles & Gintis,
2003). Gang members achieve “earned redemption” via the positive contributions through which
they might make good to their communities (Bazemore, 1999).
Many gang interventionists, street workers, or staff members working in prisoner re-entry
programs are former gang members or incarcerated persons (e.g., Kennedy, 2011). Their
criminal credential explains why—and not just because the gang industry is willing to hire
individuals with significant prior criminal records (Densley, 2011). The criminal credential is a
social status that functions as a major source of stratification (Pager, 2007); it is a “social stigma”
that disqualifies offenders from full social acceptance, such as obtaining meaningful work,
earning a high school diploma or post- secondary degree, or building a strong, participatory civic
life (Goffman, 1963, p. 9). In the strengths-based role of the “wounded healer” (Maruna, 2001)
or “professional ex-” (Brown, 1991), however, the criminal credential that once regulated access
to the gang and opportunity within it, is repurposed. Here the criminal credential lends credibility
and is what unites the “recovering” and “recovered” gang member in their struggle, the past
group member with the present group member. It is an example, therefore, of a signal that lives
very much in the eye of the beholder.
To some extent, credible commitments highlight ways in which a gang member’s agency
is limited by outside actors. In many cases, for example, gang members are ordered to cease
contact with their gangs via conditions of probation, parole, or civil gang injunctions (e.g.,
Hennigan & Sloan, 2013). Likewise, employment or moving to a new neighborhood or city may
be more a function of legitimate opportunities and parental decisions than individual agency.
Nevertheless, gang members’ still have a choice—namely, what signal to send and how to send
it. Some will comply with legal mandates, using supervision and conditions of probation as a
cover for exiting the gang with little resistance from other gang members. Others will resist
them, wearing civil or criminal sanctions as a badge of honor. Receivers also adjust their
expectations of signalers according to context, imposing different costs upon them. Gainful
employment in a booming economy or community where all gang members have jobs provides
little information of value about disengagement. Even when preoccupied with work or physically
cut-off from the old gang, moreover, gang members can still identify with gangs and stay
involved (particularly now that gangs have a virtual presence—see Patton, Eschmann, & Butler,
2013; Moule, Pyrooz, & Decker, 2014). De-identification is a prerequisite for disengagement,
otherwise there is no underlying, unobserved, state to be signaled.
Three key screening strategies— information gathering, vouching, and costly
extraction—assist those with the least information, the receivers, in overcoming their
informational handicap regarding gang disengagement. Screening occurs when (1) signaling fails
(i.e., the signal is not received) or is ambiguous or (2) where the sender lacks credibility in the
eye of the receiver. We suspect, for example, it is often the perceptions of “gang members” in the
aggregate (i.e., that they lack honesty, are unreliable, are potentially dangerous) that give rise to
the use of screening. Further, screening likely comes after signaling in the disengagement
process because it is the initial signals of disengagement, incongruous with prior or expected
gang behaviors, that prompt the receivers to probe for more information. If stakeholders are
relying on others, including other gang members, to vouch for the fact that another person is not
a gang member, for example, then signaling is already in process, because the disengaging gang
member already had to signal to the gang his or her intentions.
Information gathering and vouching. If gang membership is common knowledge in gang
circles and the best way to identify a gang member is simply to “ask another gang member,” as
Densley (2015, p. 238) argues, then it holds the best way to confirm whether someone is out of
the gang would be to ask the gang. Embedded gang members are best placed to judge whether
someone else is in or not. They can vouch for the ex-gang member, or leverage community ties
within “selective environments” (Gambetta 2009b, p. 9) to gather the all-important “word on the
street.” Others can too, if they are deemed credible or authoritative sources. Probation officers
and employers, for example, can externally validate the ex-gang member’s claims, lending a bit
of their own credibility to those who lack it, at least until the gang member can accumulate
enough credibility to stand on their own (Maruna, 2012). Journalists can do the same, to the
extent that some gang members use media profiles and public interviews to broadcast
disengagement. In Densley’s (2013, p. 140) ethnography, for example, one gang member
explains, “…some of the TV interviews…You’ve just got to have other people to verify…It’s
like a public way of saying this is not my game no more.” In both cases, the cost lies in the fact
that a third party is willing to stake his or her reputation on the validity of the gang member’s
professed disengagement. This is extremely hard-to-fake.
Costly extraction. A signaling perspective also accounts for those instances of hostile
departure, when someone leaves the gang by means of an event, such as being “beaten out” or
having to commit a crime (Decker & Pyrooz, 2011). Hostile departures, first identified in the
qualitative literature (Decker & Lauritsen, 2002; Vigil, 1988), are non-normative methods for
exiting gangs, as only roughly 20% of juvenile arrestees (Pyrooz & Decker, 2011) and school-
based youth (Carson et al., 2013) reported such an occurrence. What is important to our model is
that these events are initiated by the gang, thus putting a price on exit that is too expensive for
those who are undecided about disengagement. If it benefits the gang to have a member leave,
moreover, this can be a way of expediting disengagement with or without their consent. Hostile
departures ceremoniously put distance between the now ex-gang member and the gang. They are
costly, given they are physical, and hard-to-fake, not least because they are public and witnessed
The Role of the Receivers
Thus far, we have grouped stakeholders together, but of course different stakeholders
have different expectations as the receivers of signals and, in turn, the senders of signals will
likely need to adapt their signaling strategy to conform with these expectations. Some
stakeholders, namely those who work in offender rehabilitation, for example, are naturally more
open to disengagement signals owing to their “belief in redeemability” (Reich, 2017). People
who subscribe to the myth of “once a gang member, always a gang member”, by contrast, will be
far less receptive to disengagement signals, regardless of their construction. Family members
might be willing to accept less in the way of a credible commitment than, say, an employer,
given the more intimate knowledge the former likely has of gang life and the exit process. In
simple terms, one’s friends are more likely to accept “the new normal,” whereas one’s enemies
will not, a point that is supported in research on gang disengagement (Decker et al., 2014). A
gang member’s capacity to successfully juggle various signal receivers, in turn, is likely in itself
an important variable in the success or failure of the disengagement process.
Researchers have discovered that effectiveness of the signaling process very much
depends upon the characteristics of the receiver (for a review, see Connelly et al., 2011). If
receivers are not actively looking for the signals, not paying attention, for example, the process
will not work (Connelly et al., 2011). This is important considering that some policy imperatives
(e.g., “zero tolerance”) actively promote not looking for gang disengagement signals. Different
receivers may also attend to different signals or interpret the same signal differently depending
upon who sends it, based in part upon their own explicit and implicit biases (Glaser, 2014). In
any event, receivers may interpret signals differently than what signalers have intended.
Signaling occurs under conditions of information asymmetry, thus what constitutes a
signal is contingent upon the amount and type of information the receiver already possesses. We
suspect control agents in law enforcement or corrections, who benefit from firsthand knowledge
of the nature and extent of one’s criminal record, will respond more cautiously to a gang
member’s disengagement signals than, say, the average community member; not least because
the criminal justice professional has little to gain and much to lose from misreading the signals.
Moreover, when a signal is interpreted by others in a particular way, an individual who is unsure
about how to interpret the signal may look to imitation as a means of decision making, creating a
“bandwagon effect,” where signals are interpreted in a manner that may or may not be accurate
(Connelly et al., 2011, p. 55).
Much like how prison systems such as California apply uneven weights to the source
items (e.g., gang tattoos, photographs, debriefing reports) used to classify someone as a gang
member (see Pyrooz, 2016, p. 163), receivers may also apply weights to signals in accordance
with preconceived notions about their importance. With experience, for example, we suspect
people learn how best to receive signals; hence why some gang interventionists are more
resistant to manipulation and/or are adept at identifying individuals primed for disengagement.
Logically, those gang members who send signals aligning themselves with the perceived
characteristics of the prototypical “ex-gang member,” thus tapping into this prior knowledge and
experience, should be more effective in reducing the initial information asymmetry among
stakeholders because these new “formers” gain from the pre-existing reputations of similar
formers. For this reason, signal precedence (i.e., the order in which signals are received) can
affect receiver interpretation (Connelly et al., 2011). Without past precedent, moreover, gang
members may be forced into trial and error in an effort to gauge which signals solicit the best
Beyond a given stakeholder’s ability or motivation to receive signals, external
characteristics also affect if and how signals are received. The signaling environment, external
referents, even other signalers can create distortion that affects the mere observability of the
signals (Connelly et al., 2011). If there is a lot of media attention about “dangerous gang
members” at the time of the signaling exchange, for example, disengagement signals could be
drowned out by the noise.
Signalers and receivers need to “talk the same language” lest the signal not reach its
destination. Upon receiving signals, the receiver selects an action to take. They can send
feedback to signalers to confirm that their signals have been interpreted in the desired way or
they can impose costs on dishonest signalers to maintain signal honesty. Criminal justice
professionals utilize formal legal sanctions to this effect, whereas community members
(including other gang members) use informal or extra-legal sanctions (Densley, 2013). Signal
costs are intrinsic to signal production, but if the receiver is in a position to identify and punish
deception should it occur within the time frame of his or her decision, then the receiver can take
appropriate steps to verify the signal, including screening the signaler.
Many theoretical hypotheses are generated from a signaling perspective on gang
disengagement as outlined. It is expected that if signals correlate with the unobservable qualities
of disengagement and are communicated (1) reliably in terms of cost or convention (i.e., hard-to-
fake), (2) consistently in terms of agreement between signals from one source, and (3) frequently
in terms of number of times the message is transmitted (Connelly et al., 2011), then they will be
received and interpreted to correspond with a variety of empirically detectable changes in
associations, attitudes, behaviors, and identity, as follows.
First, our primary contention is that the mechanisms of signaling, commitments, and
screening produce reductions in levels of gang embeddedness. Importantly, while any one of the
mechanisms should reduce levels of embeddedness, when they are combined and point in the
same direction, we would anticipate effects on embeddedness that are larger in magnitude. To be
sure, we place a premium on gang embeddedness over gang de-identification; the latter, we
contend, is part and parcel of the disengagement process and we see less value in specifying its
occurrence and timing in this process for the precise reasons we have described above. The “old
heads” and “original gangsters” who continue to claim gang membership but are not at all
embedded in gang activity highlight the subtle but significant difference between de-
identification and disengagement proper.
Second, and related, are features of the exit process that should relate to the mechanisms
of the model. It has been our argument from the beginning, based both on the tenets of signaling
and the extant literature, that gang disengagement is a process rather than an event. Consistency
or agreement between signals throughout this process is important because in real life signals are
rarely separating in the sense of a perfect Bayesian equilibrium (Cho & Kreps, 1987). Instead,
receivers search for clusters of semi-sorting signals that, if pointing in the same direction,
together come close to fully discriminating (see Gambetta & Hamill, 2005). When signals of
disengagement point in the same direction, we predict disengagement should manifest quicker,
resembling a “knifing-off” rather than a long, drawn-out exit process. When signals are
frequently displayed, moreover, the outcome will be fewer ambivalent role exits and false starts
that others (e.g., Decker et al., 2014) have observed.
The more reliable, frequent, and consistent the signals of disengagement, the more
reliable, frequent, and consistent the outcomes of disengagement because the senders and
receivers are approaching information symmetry. If fellow gang members better understand the
intentions of the exiting gang member, for example, there should be fewer instances of hostile
departures or departures that are met with resistance from the gang; although this will be tied to
the circumstances of departure (e.g., Pyrooz & Decker 2011). Retirement differs from
resignation or dismissal in legitimate work, and we suspect this is also true in “gang work.” The
above, of course, remain empirical questions—many have observed, but have not explained,
variability in the exit process (e.g., Decker & Lauritsen, 2002; Vigil, 1988).
Third, the theoretical predictions stemming from a signaling perspective also extend to
outcomes of non-gang-related significance. It is expected that signaling, commitments, and
screening be related inversely to offending and victimization. These mechanisms should
introduce distance between the gang and the self. Accordingly, former gang members will be less
susceptible to gang-related group processes that embed gang members in networks of conflict.
For example, a former gang member who has made his post-gang life intentions apparent will
should be less likely to be called on to retaliate against rival gangs in the event of violence.
Similarly, we would anticipate attitudinal shifts, as former gang members will also have fewer
obligations to maintain “front stage” appearances (Goffman, 1959) that contradict the
disengagement signal, including notions of gang masculinity (Flores, 2013), adherence to the
code of the street (Matsuda et al., 2013), and legal cynicism (Papachristos et al., 2012).
Model Generality and Contingencies
Curry (2000, p. 1254) questioned whether there was a “single unified gang problem” in
the United States or tale of two gang problems. The first problem, found (mostly) in survey
research, involved gang members who were demographically younger, female and white, and
more delinquent and marginally embedded in informal-diffuse gangs. The second problem,
found (mostly) in ethnographic research and law enforcement depictions of gangs, involved gang
members who were demographically older, male and minority, and criminal, justice-system
involved, and deeply embedded in instrumental-rational gangs. It is our contention that a
signaling model of gang disengagement can account for the varieties of gangs and gang
members, although it is important that we recognize the generality of the model from the outset.
This leads us to argue that the contingencies we describe are questions of external, not internal,
validity, all of which are worthy of empirical investigation.
First, individual immersion within gangs is fluid and not all gang members start the
disengagement process from the same point of embeddedness. Research supports that barriers to
disengagement are likely greater for more embedded gang members (i.e., those with criminal
convictions, prison time, or long lists of enemies) than less embedded gang members (Pyrooz,
Sweeten, & Piquero, 2013). As a result, we anticipate that embedded gang members will be held
to a standard of disengagement beyond a reasonable doubt, compared to less embedded gang
members, who will be adjudicated based on the preponderance of evidence. In other words,
embedded gang members will need more costly, consistent, and frequent signals and
commitments than their less embedded counterparts en route to disengagement in order to
convince a naturally skeptical audience. Stanley Tookie Williams (2004), one of the early leaders
in the Los Angeles Crips, authored several anti-violence books signaling his “redemption” and
was even nominated for the Nobel Peace Prize, which more than meets the hard-to-fake
threshold. Williams was held to a high burden of proof owing to his long criminal history and
varied offending pattern. In denying him clemency, California Governor Arnold Schwarzenegger
(2005) wrote, “Is Williams’ redemption complete and sincere, or is it just a hollow promise?”
Second, not all gangs are created equal (Thrasher, 1927), which recognizes that the norms
associated with disengagement in one gang may differ from the next. It is plausible that the
signals needed to disengage from large, “traditional” or vertical gangs that are accustomed to
member turnover, therefore, may differ from the signals needed to exit small, “compressed” or
“specialty” gangs for whom every member counts (Klein & Maxson, 2006). Likewise, leaving
highly “evolved” gangs (Ayling, 2011; Densley, 2014), or gangs with especially high levels of
criminality or organizational structure, may incur different costs, commensurate with the
collective reputation of the group. Some gangs may have their own internal disengagement
procedures that bypass the need for members to burn bridges or utilize some of signals discussed.
The Almighty Latin King and Queen Nation’s manifesto, for example, encourages its members
to obtain legitimate employment, go to school, earn a degree, and even provide scholarships to
members in need. Desistance signaling thus still occurs, but very much within an existing
structure at the level of the gang, which may or may not make it easier for members to disengage
and convince others that impending departure is, in fact, legitimate.
Third, signals may also vary across gang and gang member demographics, such as
gender, race/ethnicity, and age. Signals that successfully convey disengagement from black
gangs, for example, may not work for Latino gangs, and vice versa. And if gender and age are
correlated with levels of gang embeddedness, women and children who never persisted in gang
activity may have both less ability to show signals of disengagement and less need to show (or
cultivate activities that show) commitment to disengagement. There is no doubt some of the
examples above (e.g., renouncement and recovery, burning bridges) pertain more to older gang
members than younger.
We have advanced a theoretical account of the gang disengagement process that is rooted
in a signaling perspective, emphasizing the mechanisms that close the gap in information
asymmetry between the gang member’s professed “inner change” and others’ healthy skepticism.
Disengagement from gangs in signaling perspective relies not on “variables” or individual risk
factors, but rather how such life events and states are used by the gang member according to
“specific situational characteristics and processes of interaction” (Hughes, 2006, p. 39). This
model prioritizes processes over events and methods over motivations, which we view as central
to understanding how leaving the gang is communicated and disengagement from gangs is
realized. Although our emphasis is on gangs, we suspect the ideas presented here apply generally
to disengagement from many forms of extra-legal groups beyond the scope of this paper but ripe
for investigation (e.g., terrorists, far right, religious extremists, organized crime, outlaw
motorcycle groups). We see this paper as a necessary first step.
Signaling has been criticized on the grounds that it explains too much, and cannot be
falsified (e.g., McAdams, 2001). However, the idea that any action can be a signal does not
render signaling theory malleable or non-falsifiable, only complex (Posner, 2002). And
simplicity is not the only criterion for a good theory in the face of competing explanations. Both
human capital theory and signaling theory predict that productivity and wages increase with
education, for example, and many economists doubted that it would be possible to discriminate
between these alternative explanations by empirical data, but experimental or other
methodological designs have made it possible to test the central propositions of the theory
(Weiss, 1995). After all, signaling theory relies on the rational actor model, which has been
subject to decades of careful empirical study, including in the area of criminal desistance (see
Paternoster et al., 2015). Signaling is not likely to explain all of gang disengagement. It is simply
a hypothesis, for which there is some suggestive evidence. The question is whether it explains
enough of the disengagement process that gang scholars and interventionists alike should
consider problems of asymmetric information in the future.
There are increasing efforts to link research on disengagement from gangs to the
practices of the gang intervention industry (see Roman, Decker, & Pyrooz, 2017). Equally, there
are large implications for policy and programming that stem from a signaling perspective, chief
among them procedures that facilitate the mechanisms that lessen knowledge disparities between
signalers and receivers. For example, stakeholders must provide opportunities for gang members
to signal their disengagement, as opposed to the typical risk management orientation of reading
outward signs of gang desistance or persistence (Maruna, 2012). Stakeholders must be given a
platform to “vouch” for ex-gang members, much like in the problem analysis stage of the highly-
acclaimed Group Violence Intervention strategy (National Network for Safe Communities,
2013). If disengagement lives in the eye of the beholder, gang interventionists would also require
education and training on how to receive and respond to signals sent. Consistent with notions of
“reintegrative shaming” (Braithwaite, 1989) and the transtheoretical model of health behavior
change (Prochaska, Norcross, & DiClemente, 1994), rituals that support reintegration and
prevent relapse, sending signals of acknowledgement and encouragement back to the gang
member, could make or break his or her disengagement efforts.
Akerlof G. (1970). The market for ‘lemons’: quality uncertainty and the market mechanism. The
Quarterly Journal of Economics, 84, 488–500.
Andrews, D. & Bonta, J. (2006). The psychology of criminal conduct, Fourth Edition. Newark,
Ayling, J. (2011). Gang change and evolutionary theory. Crime, Law and Social Change, 56, 1–
Baumeister, R. (1991). Meanings of life. New York, NY: Guilford Press.
Bazemore, G. (1999). After shaming, whither reintegration: Restorative justice and relational
rehabilitation. In G. Bazemore & L. Walgrave (eds)., Restorative juvenile justice (pp.
155–194). Monsey, NY: Criminal Justice Press.
Becker, H. (1963). Outsiders. New York: Free Press.
Bjorgo, T. (2002). Exit neo-nazism: Reducing recruitment and promoting disengagement from
racist groups, NUPI Working Paper 627. Oslo: Norwegian Institute of International
Bowles, S. & Gintis, H. (2003). Origins of human cooperation. In P. Hammerstein (ed.), Genetic
and cultural evolution of cooperation. Cambridge, MA: MIT Press.
Boyle, G. (2011). Tattoos on the heart. New York: Free Press.
Braithwaite, J. (1989). Crime, shame and reintegration. Cambridge, UK: Cambridge University
Brown, J D. (1991). The professional ex-: An alternative for exiting the deviant career. The
Sociological Quarterly, 32, 219-230.
Bubolz, B. F., & Simi, P. (2015). Disillusionment and change: A cognitive-emotional theory of
gang exit. Deviant Behavior, 36, 330–345.
Burman, M. (2012). Resocializing and repairing homies within the Texas prison system
[Unpublished Doctoral Dissertation]. Austin, TX: University of Texas at Austin.
Bushway, S. & Apel, R. (2012). A signaling perspective on employment-based reentry
programming. Criminology & Public Policy, 11, 21–50.
Bushway, S. & Paternoster, R. (2013). Desistance from crime: a review and ideas for moving
forward. In C. L. Gibson & M. D. Krohn (Eds.), Handbook of life-course criminology
(pp. 213–231). New York: Springer.
Bushway, S. & Reuter V. (1997). Labor markets and crime risk factors. In L. Sherman (Ed.),
Preventing crime: What works, what doesn’t, what’s promising (pp 198-240).
Washington DC: National Institute of Justice.
Campana, P. & Varese, F. (2013). Cooperation in criminal organizations: Kinship and violence
as credible commitments. Rationality and Society, 25, 263–289.
Carlsson, C. (2012). Using ‘turning points’ to understand processes of change in offenders.
British Journal of Criminology 52, 1–16.
Carr, S. (2010). Scripting addiction. Princeton, NJ: Princeton University Press.
Carson, D., Peterson, D., & Esbensen, F-A. (2013). Youth gang desistance. Criminal Justice
Review, 38, 510–534.
Carson, D. & Vecchio, M. (2015). Leaving the gang: A review and thoughts on future research.
In S. Decker & D. Pyrooz (eds.), The Handbook of gangs. Hoboken, NJ: Wiley.
Cheng, T. (2017). Violence prevention and targeting the elusive gang member. Law & Society
Review, 51, 42–69.
Cho, I-K, & Kreps, D. (1987). Signaling games and stable equilibria. The Quarterly Journal of
Economics, 102, 179–222.
Collins, R. (1979). The credential society. New York: Academic Press.
Connelly, B., Certo, S., Ireland, R., & Ruetzel, C. (2011). Signaling theory: a review and
assessment. Journal of Management, 37, 39–67.
Cooley, C. (1922). Human nature and the social order, revised edition. New York: Scribners.
Cronk, L. (2005). The application of animal signaling theory to human phenomena: Some
thoughts and clarifications. Social Science Information, 44, 603–620.
Curry, G. D. (2000). Self-reported gang involvement and officially recorded delinquency.
Criminology, 38, 1253–1274.
Decker, S. & Lauritsen, J. (2002). Leaving the gang. In C. R. Huff (ed.), Gangs in America Third
Edition (pp. 51–67). Thousand Oaks, CA: Sage.
Decker, S. H & Pyrooz, D. C. (2011). Timing is everything: Gangs, gang violence and the life
course. In M. DeLisi & K. M. Beaver (Eds.), Criminological theory: A life-course
approach (pp. 149-164). Boston, MA: Jones and Bartlett.
Decker, S.H., Pyrooz, D.C., & Moule, R.M., Jr. (2014). Gang disengagement as role transitions.
Journal of Research on Adolescence, 24, 268–283.
Decker, S. & van Winkle, B. (1996). Life in the gang. New York: Cambridge University Press.
Densley, J. (2011). Ganging up on gangs: Why the gangs intervention industry needs an
intervention. The British Journal of Forensic Practice, 13, 12–23.
Densley, J. (2012). Street gang recruitment: Signaling, screening and selection. Social Problems,
Densley, J. (2013). How gangs work. New York: Palgrave Macmillan.
Densley, J. (2014). It’s gang life, but not as we know it: The evolution of gang business. Crime
& Delinquency, 60, 517–546.
Densley, J. (2015). Joining the gang: A process of supply and demand. In S. Decker & D. Pyrooz
(eds.), The Handbook of gangs. Hoboken, NJ: Wiley.
DeWitt, S. (2016). Signaling desistance: Identifying long-term desisters using short-term signals
[Unpublished Doctoral Dissertation]. New Brunswick, NJ: Rutgers University.
Durán, R. (2013). Gang life in two cities: An insider’s journey. New York: Columbia University
Ebaugh, H. (1988). Becoming an ex. Chicago, IL: University of Chicago Press.
Fearon, J.D. (1997). Signaling foreign policy interests: Tying hands versus sinking costs.
Journal of Conflict Resolution, 41, 68–90.
Fleisher, M. & Krienert, J. (2004). Life-course events, social networks, and the emergence of
violence among female gang members. Journal of Community Psychology, 32, 607–622.
Flores, E. O. (2013). God’s gangs. New York: New York University Press.
Flores, E. O. (2016). ‘Grow your hair out’: Chicano gang masculinity and embodiment in
recovery. Social Problems, 63, 590-604.
Frank, R. (1988). Passions within reason. New York, NY: Norton.
Gambetta, D. (2009a). Signaling. In P. Hedstrom and P. Bearman (Eds.), The Oxford handbook
of analytical sociology (pp. 168–192). Oxford: OUP.
Gambetta, D. (2009b). Codes of the underworld. Princeton, NJ: Princeton University Press.
Gambetta, D. & Hamill, H. (2005). Streetwise. New York: Russell Sage Foundation.
Giordano, P., Cernkovich, S., & Rudolph, J. (2002). Gender, crime, and desistance: Toward a
theory of cognitive transformation. American Journal of Sociology, 107, 990-1064.
Glaser, J. (2014). Suspect race: Causes and consequences of racial profiling. New York: OUP.
Goffman, E. (1963). Stigma. Englewood Cliffs, NY: Prentice-Hall.
Goffman, E. (1959). The presentation of self in everyday life. New York: Anchor.
Hagen, E., Watson, P., & Hammerstein, P. (2008). Gestures of despair and hope: A view on
deliberate self-harm from economics and evolutionary biology. Biological Theory, 3,
Hamill, H. (2011). The hoods. Princeton, NJ: Princeton University Press.
Hennigan, K. & Sloane, D. (2013) Improving civil gang injunctions. Criminology & Public
Policy, 12, 7–41.
Howell, J. C. (2007). Menacing or mimicking? Realities of youth gangs. Juvenile and Family
Court Journal, 58, 39–50.
Hughes, L. (2006). Studying youth gangs: The importance of context. In J. Short & L. Hughes
(Eds.), Studying youth gangs (pp. 37–46). Lanham, MD: AltaMira Press.
Kaminski, M. (2004). Games prisoners play. Princeton: Princeton University Press.
Kennedy, D. (2011). Whither streetwork? The place of outreach workers in community violence
prevention. Criminology & Public Policy, 10, 1045–1051.
Kerig, P., Wainryb, C., Sinayobye Twali, M., & Chaplo, S. (2013). America’s child soldiers:
toward a research agenda for studying gang-involved youth in the United States. Journal
of Aggression, Maltreatment & Trauma, 22, 773–95.
Klein, M. (1971). Street gangs and street workers. Englewood Cliffs, NJ: Prentice-Hall.
Klein, M. (2011). Comprehensive gang and violence reduction programs: Reinventing the square
wheel. Criminology and Public Policy, 10, 1037–1044.
Klein, M. & Maxson, C. (2006). Street gang patterns and policies. New York: OUP.
Kurlychek, M., Brame, R., & Bushway, S. (2006). Scarlet letters and recidivism: Does an old
criminal record predict future offending? Criminology & Public Policy, 483–504.
Laub, J. & Sampson, R. (2003). Shared beginnings, divergent lives. Cambridge, MA: Harvard
Lauger, T. (2012). Real gangstas. New Brunswick, NJ: Rutgers University Press.
LeBel, T., Burnett, R., Maruna, S., & Bushway, S. (2008). The ‘chicken and egg’ of subjective
social factors in desistance from crime. European Journal of Criminology, 5, 131–159.
Lebel, T., Richie, M., & Maruna, S. (2015). Helping Others as A Response to Reconcile a
Criminal Past: The Role of the Wounded Healer in Prisoner Reentry Programs. Criminal
Justice and Behavior, 42, 108-120.
Lemert, E. (1951). Social pathology. New York: Mcgraw-Hill.
Lopez-Aguado, P. (2013). Working between two worlds: Gang intervention and street liminality.
Ethnography, 14, 186–206.
Moloney, M., MacKenzie, K., Hunt, G., & Joe-Laidler, K. (2009). The path and promise of
fatherhood for gang members. British Journal of Criminology, 49, 305– 325.
Moloney, M., Hunt, G., Joe-Laidler, K., & MacKenzie, K. (2011). Young mother (in the) hood:
Gang girls’ negotiation of new identities. Journal of Youth Studies, 14, 1–19.
Maruna, S. (2000). Desistance from crime and offender rehabilitation: A tale of two research
literatures. Offender Programs Report, 4, 1–13.
Maruna, S. (2001). Making good. Washington DC: APA Books.
Maruna, S. (2012). Elements of successful desistance signaling. Criminology & Public Policy,
Maruna, S., LeBel, T., Naples, M., & Mitchell, N. (2009). Looking-glass identity transformation:
Pygmalion and golem in the rehabilitation process. In B. Veysey, J. Christian & D.
Martinez (eds.), How offenders transform their lives. Cullompton, UK: Willan.
Massoglia, M. & Uggen, C. (2010). Settling down and aging out: Toward an interactionist theory
of desistance and the transition to adulthood. American Journal of Sociology, 116, 543–
Matsuda, K., Melde, C., Taylor, T., Freng, A., & Esbensen, F-A. (2013). Gang membership and
adherence to the ‘code of the street’. Justice Quarterly, 30, 440–68.
Matza, D. (1969). Becoming deviant. Englewood Cliffs, NJ: Prentice-Hall.
McAdams, R. (2001). Signaling discount rates: Law, norms, and economic methodology. Yale
Law Journal, 110, 625–689.
Mehu, M., Little, A., & Dunbar, R. (2007). Duchenne smiles and the perception of generosity
and sociability in faces. Journal of Evolutionary Psychology, 5, 183-(196.
Melde, C. & Esbensen, F-A. (2014). The relative impact of gang status transitions: Identifying
the mechanisms of change in delinquency. Journal of Research in Crime and
Delinquency, 51, 349–376.
Merton, R. (1957). Continuities in the theory of reference groups and social structure. In R.
Merton (ed)., Social theory and social structure. New York, NY: Free Press.
Messerschmidt, J. (1993). Masculinities and crime. Lanham, MD: Rowman & Littlefield.
Moore, J. (1991). Going down to the barrio. Philadelphia, PA: Temple University Press.
Moule, R., Pyrooz, D., & Decker, S. (2014). Internet adoption and online behaviour among
American street gangs: Integrating gangs and organizational theory. British Journal of
Criminology, 54, 1186–1206.
National Network for Safe Communities. (2013). Group violence intervention: An
implementation guide. Washington, DC: U.S. Department of Justice, COPS.
O’Neal, E. N., Decker, S., Moule, R., & Pyrooz, D. (2016). Girls, gangs, and getting out: Gender
differences and similarities in leaving the gang. Youth Violence and Juvenile Justice, 14,
Papachristos, A., Meares, T., & Fagan, J. (2012). Why do criminals obey the law? The influence
of legitimacy and social networks on active gun offenders. Journal of Criminal Law and
Criminology, 102, 397–440.
Paternoster, R., Bachman R., Bushway S., Kerrison E., & O’Connell D. (2015). Human agency
and explanations of criminal desistance: Arguments for a rational choice theory. Journal
of Developmental and Life-Course Criminology, 1, 209–235.
Patton, D., Eschmann, R., & Butler, D. (2013). Internet banging: New trends in social media,
gang violence, masculinity and hip hop. Computers in Human Behavior, 29, 54–59.
Piquero, A., Farrington, D., & Blumstein, A. (2003). The criminal career paradigm. In M. Tonry
(ed.), Crime and justice: A review of research Vol. 30. Chicago, IL: University of
Pager, D. (2007). Marked. Chicago, IL: University of Chicago Press.
Posner, E. (2002). The signaling model of social norms: Further thoughts. University of
Richmond Law Review, 36, 465–480.
Prochaska, J., Norcross, J., & DiClemente, C. (1994). Changing for good. New York: W.
Pyrooz, D. (2014). ‘From your first cigarette to your last dyin’ day’: The patterning of gang
membership in the life-course. Journal of Quantitative Criminology, 30, 349–372.
Pyrooz, D. (2016). Gang affiliation and restrictive housing in U.S. Prisons. In M. Garcia (Ed.),
Restrictive housing in the U.S.: Issues, challenges, and future directions (pp. 117–64).
Washington, DC: U.S. Department of Justice, National Institute of Justice.
Pyrooz, D. & Densley, J. (2016). Selection into street gangs: signaling theory, gang membership,
and criminal offending. Journal of Research in Crime and Delinquency, 53, 447–481.
Pyrooz, D. & Decker, S. (2011). Motives and methods for leaving the gang: understanding the
process of gang desistance. Journal of Criminal Justice, 39, 417–425.
Pyrooz, D. & Sweeten, G. (2015). Gang membership between ages 5 and 17 years in the United
States. Journal of Adolescent Health, 56, 414–419.
Pyrooz, D., Decker, S., & Webb, V. (2014). The ties that bind: Desistance from gangs. Crime
and Delinquency, 60, 491–516.
Pyrooz, D., Sweeten, G., & Piquero, A. (2013). Continuity and change in gang membership and
gang embeddedness. Journal of Research in Crime and Delinquency, 50, 239–271.
Pyrooz, D., Turanovic, J., Decker, S., & Wu, J. (2016). Taking stock of the relationship between
gang membership and offending: a meta-analysis. Criminal Justice and Behavior, 43,
Ralph, L. (2014). Renegade dreams. Chicago, IL: University of Chicago Press.
Reich, S. (2017). An exception to the rule: Belief in redeemability, desistance signals, and the
employer’s decision to hire a job applicant with a criminal record. Journal of Offender
Rehabilitation, 56, 110–136.
Rios, V. (2011). Punished: Policing the lives of Black and Latino boys. New York: NYU Press.
Roman, C. G., Decker, S. H., and Pyrooz, D. C. (2017). Leveraging the pushes and pulls of gang
disengagement to improve gang intervention: findings from three multi-site studies and a
review of relevant gang programs. Journal of Crime and Justice. Online First. DOI:
Rothschild, M. & Stiglitz, J. (1976). Equilibrium in competitive insurance markets: an essay on
the economics of imperfect information. The Quarterly Journal of Economics, 90, 629–
Schwarzenegger, A. (2005). Statement of decision: Request for clemency by Stanley Williams.
Sosis, R. (2003). Why aren’t we all Hutterites? Costly signaling theory and religious behavior.
Human Nature, 14, 91–127.
Smith, A. & Varese, F. (2001). Payment, protection and punishment: The role of information and
reputation in the Mafia. Rationality and Society, 13, 349–393.
Spence, M. (1974). Market signaling. Cambridge, MA: Harvard University Press.
Spence, M. (2002). Signaling in retrospect and the informational structure of markets. American
Economic Review, 92, 434–459.
Storrod, M. & Densley, J. (2017). ‘Going viral’ and ‘Going country’: The expressive and
instrumental activities of street gangs on social media. Journal of Youth Studies, 20, 677–
Sweeten, G., Piquero, A., & Steinberg, L. (2013). Age and the explanation of crime, revisited.
Journal of Youth and Adolescence, 42, 921–938.
Sweeten, G., Pyrooz, D., & Piquero, A. (2013). Disengaging from gangs and desistance from
crime. Justice Quarterly, 30, 469–500.
Thomas, J. & Zaitzow, B. (2006). Conning or conversion? The role of religion in prison coping.
The Prison Journal, 86, 242–259.
Thornberry, T., Huizinga, D., & Loeber, R. (2004). The Causes and correlates studies: Findings
and policy implications. Juvenile Justice, 9, 3–19.
Thrasher, F. (1927). The gang. Chicago, IL: University of Chicago Press.
Tita, G. E., & Papachristos, A. V. (2010). The evolution of gang policy: Balancing intervention
and suppression. In R. J. Chaskin (Ed.), Youth gangs and community intervention:
Research, practice, and evidence (pp. 24–48). New York: Columbia University Press.
Uggen, C. & Piliavin, I. (1998). Asymmetrical causation and criminal desistance. Journal of
Criminal Law and Criminology, 88, 1399–1422.
van Gemert, F. & Fleisher, M. (2005). In the grip of the group. In S. Decker & F. Weerman
(eds.), European street gangs and troublesome youth groups (pp. 11–29). Lanham, MD:
Veblen, T. (1899). The theory of the leisure class. New York: Macmillan.
Vigil, J. D. (1988). Barrio gangs. Austin, TX: University of Texas Press.
Watkins, A. & Melde, C. (2016). Bad medicine: The relationship between gang membership,
depression, self-esteem, and suicidal behavior. Criminal Justice and Behavior, 43, 1107–
Weiss, A. (1995). Human capital vs. signaling explanations of wages. Journal of Economic
Perspectives, 9, 133–154.
Whitman, J. & Davis, R. (2007). Snitches get stitches. Washington, D.C.: National Center for
Victims of Crime.
Wieselquist, J., Rusbult, C., Foster, C., & Agnew, C. (1999). Commitment, pro-relationship
behavior, and trust in close relationships. Journal of Personality and Social Psychology,
Williams, S. T. (2004). Blue rage, black redemption: A memoir. Pleasant Hill, CA: Damamli.
Winegard, B., Reynolds, T., Baumeister, R., Winegard, B., Maner, J. (2014). Grief functions as
an honest indicator of commitment. Personality and Social Psychology Review, 18, 168-