ArticlePDF Available

An evaluator’s reflections and lessons learned about gang intervention strategies: an agenda for research



Purpose This paper is designed to critically review and analyze the body of research on a popular gang reduction strategy, implemented widely in the United States and a number of other countries, to: (1) assess whether researchers designed their evaluations to align with the theorized causal mechanisms that bring about reductions in violence; and (2) discuss how evidence on gang programs is generated and consumed. That review and assessment is then used to frame a research agenda for studying gang interventions. Design/methodology/approach A case study design is used to generate a multi-faceted understanding of the possible avenues for evaluation research on the law enforcement-based strategy known as the Group Violence Intervention. The paper discusses questions that remain to be answered about the strategy, such as “what type of deterrence is operating?” and if the model actually works by the threat of deterrence, and not by removing high-risk offenders and shootings from the street, what activities are needed to maintain the effect? Findings Across roughly two dozen impact evaluations of GVI, none have examined the likely cause and effect components of this multi-partner strategy in reducing the violence. Furthermore, there are many issues related to the production and generation of criminal justice evaluation research that have adversely pushed the balance of evidence on what works in gang reduction toward law enforcement programming. However, there are many strategies that researchers can use to think broadly about appropriate and holistic research and evaluation on gangs and gang programming. Practical implications The recommendations for research, if implemented, can help build a body of knowledge to move toward community-based and restorative models of gang violence reduction. Originality/value This original piece is one of the first essays to contextualize and discuss how aspects of the production of social science research on gangs may directly impact what programs and strategies are implemented on the ground.
Pre-publication copy | Caterina G. Roman
An Evaluator’s Reflections and Lessons Learned about Gang Intervention Strategies:
An Agenda for Research
Journal of Aggression, Conflict and Peace Research, 2021
ISSN: 1759-6599
Article publication date: 28 June 2021
Caterina G. Roman, PhD
Temple University, Philadelphia, PA, USA
Keywords: Focused Deterrence, Gangs, Gang Interventions, Law Enforcement, Problem-
Oriented Policing, Violence
Roman, C.G. (2021), "An evaluator’s reflections and lessons learned about gang intervention
strategies: an agenda for research", Journal of Aggression, Conflict and Peace Research, Vol.
Caterina G. Roman, PhD
Department of Criminal Justice
Temple University
5th Floor, Gladfelter Hall
1115 Polett Walk
Philadelphia, PA 19122, USA
The author would like to thank the anonymous peer reviewers, Alyssa Mendlein at Temple
University, Finn Esbensen, and John K. Roman for their review and comments on this
Pre-publication copy | Caterina G. Roman
An Evaluator’s Reflections and Lessons Learned about Gang Intervention Strategies:
An Agenda for Research
This paper is designed to critically review and analyze the body of research on a popular gang
reduction strategy, implemented widely in the United States and a number of other countries, to:
(1) assess whether researchers designed their evaluations to align with the theorized causal
mechanisms that bring about reductions in violence; and (2) discuss how evidence on gang
programs is generated and consumed. That review and assessment is then used to frame a
research agenda for studying gang interventions.
A case study design is used to generate a multi-faceted understanding of the possible avenues for
evaluation research on the law enforcement-based strategy known as the Group Violence
Intervention. The paper discusses questions that remain to be answered about the strategy, such
as “what type of deterrence is operating?” and “if the model actually works by the threat of
deterrence, and not by removing high-risk offenders and shootings from the street, what activities
are needed to maintain the effect?
Across roughly two dozen impact evaluations of GVI, none have examined the likely cause and
effect components of this multi-partner strategy in reducing the violence. Furthermore, there are
many issues related to the production and generation of criminal justice evaluation research that
have adversely pushed the balance of evidence on what works in gang reduction toward law
enforcement programming. However, there are many strategies that researchers can use to think
broadly about appropriate and holistic research and evaluation on gangs and gang programming.
The recommendations for research, if implemented, can help build a body of knowledge to move
toward community-based and restorative models of gang violence reduction.
No one would disagree that street gangs and gang violence are complex problems for
urban communities. Many scholars have brought attention to the need for careful policy and
programmatic responses that have theories of change that match the complexity behind the array
of factors that interact to engender gangs and the attendant violence. There have been some
successes in reducing gang-related violence—mostly with regard to developing interventions that
target high-risk people within the specific neighborhoods that have witnessed high levels of gun
violence (Abt and Winship, 2016; Huey et al., 2016).
One intervention that academics and government funders have deemed successful at
reducing gang violence is the Group Violence Intervention (GVI), often referred to as focused
Pre-publication copy | Caterina G. Roman
deterrence. GVI is a law enforcement-driven partnership strategy based on the principles of
deterrence and designed to reduce homicides and nonfatal shootings by street gangs. Conclusions
from two meta-analytical studies on focused deterrence models covering 15 years have led GVI
to be viewed as “evidenced-based.” Both meta-analyses showed that focused deterrence, when
aimed at gangs, has been successful at reducing aggregate levels of violence in the areas
frequented by the targeted gangs (Braga and Weisburd, 2012; Braga et al., 2018). Today, GVI has
been implemented in dozens of cities across the United States and abroad (Deuchar, 2013), and
continues to be considered a strong model for implementation amidst a time of intense scrutiny
focused on the institution of policing (Center for Policing Equity, 2020; Jaitman, 2019;
Wennmenn, 2018).
This paper examines how we arrived where we are today, with a popular gang violence
reduction strategy a key pillar of public safety for many communities, during a historical
movement to reduce the footprint of policing and systemic biases present in policing and the
criminal justice system as a whole. Does GVI—a gang violence reduction intervention that relies
on the heavy threat of punitive law enforcement action—have a place in cities today? Voices
advocating for community well-being and racial equity remind us of the importance of asking
not only “who benefits” from an intervention, but also “who is burdened?”. Even well-intended
programs and policies may have unintended consequences, particularly when they are
implemented in isolation from the larger community context (Burd-Sharps et al., 2017). Have we
done our due diligence and examined the potential costs to the community for additional or “new
and different” policing? I believe these costs are not zero, yet few studies of GVI have examined
these potential costs and burdens.
Hence, this paper uses the widespread implementation of GVI as an example to help
frame a discussion about the path toward combatting gangs without reliance on the police. More
specifically, I consider the following questions: (1) do we know enough about GVI to say
unequivocally that the model isn’t harmful or doesn’t have unintended consequences? And more
broadly, (2) are we asking the right questions when researchers set out to examine and implement
gang violence reduction models? If we are not asking the right questions, what is preventing us
from doing so? Finally, (3) what can we learn from the history undergirding how criminal justice
research is funded and directed?
Pre-publication copy | Caterina G. Roman
This reflection piece will assess these questions and propose a research agenda. Answers
to these questions can move us to think more broadly about how to conduct social science
research on gangs. The emphasis will be on building the knowledge base to increase our
understanding of the strengths and limitations of policing-based interventions while
simultaneously building knowledge on community-based restorative models of gang-violence
reduction. This essay is not a prescription for new interventions, but instead a prescription for
developing the research capacity to answer a broad range of questions that will not only help us
make progress toward reducing gang violence, but also toward understanding the path to long-
term investments in communities. Neither is this essay intended as a pointed criticism of GVI,
but a larger plea to examine our research processes in what and how we evaluate interventions.
Hence, the intended audience for this paper is the research community, and also funders seeking
to make progress in building a fair and equitable criminal justice system and increasing
community well-being.
The Group Violence Intervention
Core Model Characteristics
GVI has been part of the policing toolbox since its implementation in Boston,
Massachusetts in 1996 (see Kennedy et al., 1996). As stated above, the partnership strategy
brings together law enforcement agencies and community groups to communicate directly and
repeatedly to gangsi committing the violence that gun violence will no longer be tolerated, and if
a member of a gang commits a shooting, there will be gang-wide consequences. These
consequences, in the form of legal enforcement actions, can include, but are not limited to,
federal prosecution, the execution of outstanding warrants, requests for high bail and or longer
sentences, heightened probation and parole supervision, revocation of supervision, housing code
enforcement, and other non-traditional legal levers such as increased enforcement of outstanding
child support payments. Through direct communication and action, law enforcement partners
commit to changes in the certainty, swiftness, and severity of punishment. As Braga and
Kennedy (2020, p.9) state in a recent review and summary of focused deterrence initiatives:
“Deterrence is produced when the targeted group and other groups watching the operations
unfold understand what triggered the response and that similar actions will produce similar
enforcement operations in the future –essentially, drawing ‘cause and effect’ in the minds of
offenders and prospective offenders alike.”
Pre-publication copy | Caterina G. Roman
The intent of the strategy ultimately is to be heavy on the verbal messaging, through
formal meetings held by law enforcement leaders with members of gangs, but low on law
enforcement actions because the threat of deterrence through clear messaging should engender
reductions in violent behavior. Community leaders and organizations act as partners in delivering
a moral message throughout the community that violence will not be tolerated; the committed
messaging creates a sense of urgency, that “business-as-usual” will not continue. A social service
component complements the strategy, showing past and potential offenders that prosocial
opportunities are available as an alternative to criminal behavior. These components together
should increase perceptions of the legitimacy of policing and fairness of actions connected to the
criminal justice system (Kennedy, 2011; Kennedy and Braga, 2020).
Cities that implement GVI usually undergo a problem-solving process and analyze
violent crime data that show that a relatively large percentage of shootings (or some other violent
crime such as homicide) can be attributed to street gangs. These gangs, for instance, can be
loosely formed groups of delinquent youth who grew up together and have some kind of
peer/gang identity or are more structured street groups with and without an organizational
hierarchy. The strategy is meant to focus on the most violent groups; meetings across law
enforcement and corrections partners are held to identify gang territories and craft group
membership lists. Assessing group membership is important because group-wide consequences
following a shooting include levers pulled on all group members, regardless of involvement in
the shooting that prompted the enforcement action. But ideally, the model suggests that if threats
become credible enough, they do not need to be carried out. As the initiative unfolds over time,
group members and potential offenders (and the community at large) would see laws being
upheld in a legitimate and procedurally just manner.
GVI remains popular in cities today. Proponents of GVI, including scholars of color (e.g.,
Brumfield-Young, mentioned in Byers, 2021; Brunson, 2020), note the model’s careful, data-
driven focus on the most violent gangs, clear messaging, and commensurate (rather than
aggressive or disrespectful) police actions, which in theory (and as noted earlier) should lead to
improved perceptions of police legitimacy and procedural justice. Additionally, according to the
strategy’s theory of change, the community has a voice in the model—so that public safety
moves toward co-production, not dictated by outsiders. Brunson (2020) has suggested that GVI
and related focused deterrence models have an advantage over “more divisive crime-reduction
Pre-publication copy | Caterina G. Roman
strategies” because these models do not cast a wide net over the whole of minority communities
and strategy partners can create customized social services for the individuals targeted by the
Similarly, one legal scholar (Thacher, 2016) has argued that the problem-solving
approach, with specific focus on the individuals most involved in perpetrating shootings,
combined with a careful implementation of the strategy, could help channel police discretion,
reducing (p. 536) “the inscrutable, arbitrary, and excessive use of legal authority that over-
criminalization threatens to produce.” Some international works compiling best practices to
reduce conflict and encourage peace in violent cities have leaned on focused deterrence. In a
book chapter titled “Daring the unconventional on the pathways for peace: On the ‘how’ of
sustaining peace in the city,” author Achim Wennmenn of the Centre on Conflict, Development
and Peacebuilding in Geneva, Switzerland (2018, p.185) writes that focused deterrence-related
policy options are a valid strategy for achieving peace given its theory of change related to
increases in community trust and perceptions of legitimacy.
The Group Violence Intervention in an Age of Police Reckoning
In the current age of police reckoning, it may be surprising to some that a policing model
like GVI, one that calls for extra monitoring of gangs and gang members, coupled with
potentially-wide enforcement actions on groups, has been held up by some scholars as
appropriate today. The May 2020 police killing of George Floyd brought discussions about
police violence to center stage, and created a surge in momentum to reform and reduce the role
for police across communities (Simonson, 2021; Walsh, 2021). Some scholars and community
advocates also have called for police abolition (Illing, 2020). This scrutiny of the institution of
policing has been building for years (National Academies of Sciences, Engineering and Medicine
[NASEM], 2018; Vitale, 2017), and the problems go far beyond police violence, to include racial
disparities and implicit and explicit biases, constitutional and civil rights violations, and
utilitarian perspectives that permit unconstrained police authority, where the end justifies the
means. Furthermore, police interactions can lead to civilian arrest and charges,
disproportionately sweeping people of color into a criminal justice system that is known to create
ripples of harm to individuals and communities (Travis, 2011).
Policymakers, police practitioners and the community at large had hoped that the wide-
ranging shift in the 1990s and 2000s to community and problem-solving policing by most U.S
Pre-publication copy | Caterina G. Roman
police departments (and away from reactive policing) would help improve community relations
and limit police over-surveillance and related abuses of power. But research and factual findings
from court proceedings and investigations suggest that some proactive and problem-oriented
strategies can lead to or engender violations of civil and constitutional rights (NASEM, 2018; see
also Braga, 2008 for a review of problem-oriented and proactive policing).
For instance, certain types of police surveillance tactics, and practices such as stop,
question, and frisk, have a long history of court challenges, many related to the U.S.
Constitution’s Fourteen Amendment. The Due Process Clause in the Fourteenth Amendment
limits open-ended police power with the intent to curtail racial profiling, police violence and
other potential abuses of power. In addition to the Fourteenth Amendment, the First Amendment,
which lays out rights to free association, among others, has been particularly relevant with regard
to anti-gang strategies. Specifically, in recent years, civil gang injunctions (CGI) have come
under fire and have been the subject of numerous law suits.ii CGIs, which are similar to
restraining orders but are for gang members, can result in arrest and incarceration if an individual
violates the order. In some jurisdictions, CGIs and GVI have been criticized for misidentifying
individuals as gang members and the fuzzy operationalization of the concept of “gang member.”
Both these strategies utilize chronic offender and gang lists built from police information based
not only on imperfect criminal justice data, but within a system where Black individuals tend to
be overrepresented (NASEM, 2018). Historically, neighborhoods that are targeted for CGIs or
focused deterrence-type strategies are home to people of color (Lofstrom and Raphael, 2016; and
see Scott, 2017 for a review of cities and targeted areas for focused deterrence intervention). In
addition, there may be indirect collateral consequences for individuals and communities simply
because the nature of proactive and problem-oriented policing practices fosters criminal justice
contact and involvement (NASEM, 2018).
Furthermore, in recent years studies have begun to shine more light on the factors that
cause disproportionate minority involvement in the criminal justice system. Studies have found
support for the differential selection and processing hypothesis that espouses that individuals of
color are treated differently (i.e., more unfairly and negatively) than those who are white, and
that these processes occur not only at the prosecution and sentencing levels, but also at the police
level. Study findings supporting differential selection imply that the police presence, patrolling,
and profiling that occurs in minority neighborhoods is excessive and unjust (and different from
Pre-publication copy | Caterina G. Roman
what occurs in nonminority neighborhoods) (Gase, et al., 2016; Piquero, 2008). One recent study
that analyzed court adjudications in Delaware found that the primary drivers of racial disparities
in incarceration sentences and sentence length occurred at the arrest stage (MacDonald and
Donnelly, 2019). These issues and others, and the potential harms from rights violations and
over-policing are discussed in more detail below in relation to GVI.
Past Criticisms of GVI
One particular aspect of GVI that has been denounced by some is the fact that
enforcement actions are applied against most gang members of a gang that is being targeted,
regardless of whether one was involved in the focal shooting event (Griffiths and Christian,
2015; Rowe and Sogaard, 2019). As the GVI model dictates (National Network for Safe
Communities, 2016): “Enforcement actions often pursue serious consequences of the shooter
(the individual who committed the specifically prohibited violent act) and seek to apply legal
actions, informal sanctions or uncomfortable attention to as many of that person’s group as
possible” (p.50); and “Group enforcement action can be effective even if there is no formal case
against the perpetrator for the homicide or for any crime” (p.90). Essentially, when a member of
a targeted group shoots, others can be arrested and prosecuted for minor offenses. Although in
theory, this action is not an arrest for association because the arrests are made for crimes and law
violations, albeit minor ones. One qualitative study that reviewed evidence from ethnographic
research of focused deterrence strategies operating in England, Wales and Denmark found that
law enforcement officers engaged in practices that violated First Amendment rights (i.e., freedom
of assembly) as well as due process rights (Rowe and Sogaard, 2019). The authors concluded
that law enforcement officers had too much discretion in applying levers and often applied
enforcement tactics against individuals where social services support would have been more
In Philadelphia, over-policing and unreasonable and intrusive police practices became a
major issue voiced by some stakeholders during Philadelphia Focused Deterrence, which began
with its first call-in meeting in April, 2013 (Manchester, 2018). Similar issues were highlighted
in a recent report by the Philadelphia Police Advisory Commission (2020), a civilian oversight
group that reports to the mayor. The report, developed after a series of in-depth interviews with
those involved with and touched by strategy, concluded that the strategy did not have the
necessary tools to ensure due process protections, individualized social service plans, and
Pre-publication copy | Caterina G. Roman
equitable resource provision. The report documents a number of instances where model
implementation did not align with its intended mission. Other criticisms in Philadelphia came
from the Defender Association of Philadelphia. An essay written by a Defender Association staff
member at the time (Manchester, 2018) describes a pattern of aggressive street stops where an
officer in the Philadelphia Police Department Gang Task Force stopped a gang-identified
individual 89 times over three years, with no purported cause for arrest except during the 89th
stop. The essay also suggests that there were a number of Defender Association clients caught up
in Focused Deterrence enforcement actions that resulted in the derailment of promising paths for
individuals who were no longer associated with their gang. The Defender Association, and other
stakeholders, pushed the city to specify an avenue with explicit criteria that would allow
individuals to be removed from the gang list. A news article in the online news magazine The
Appeal (Rivlin-Nadler, 2018) echoed a number of issues against Focused Deterrence as
evidenced by the title of the article: How Philadelphia’s Social Media-Driven Gang Policing Is
Stealing Years from Young People.
Beyond Philadelphia, others who have studied GVI have suggested that the complexity of
the multi-component strategy hinders implementation, and in turn, implementation problems can
result in harm to individuals and communities. It is important to note that implementation
problems are quite typical of multi-component programs, even outside of criminal justice
policies and programs. Yet, in the realm of gang prevention and intervention specifically,
implementation failure is often provided as one of the reasons why these comprehensive
initiatives tend to fail (Klein and Maxson, 2006). However, the literature chronicling potential
harms from implementation issues in police-community partnership models remains in its
Evaluators of GVI have stated that one of the potential hurdles to proper implementation
is the amount of learning that must take place among law enforcement partners to apply the
model as intended. The application of levers to specific persons within a specific group must be
carefully executed to be successful. Careful enforcement can get lost in the implementation
process as some officers may not be aware of the specifics of the initiative or and may not buy in
to a strategy that runs against business-as-usual policing practices (Griffiths and Christian, 2015).
With regard to the data collection practices to identify members of groups and relevant shooting
incidents, Corsaro and Engel (2015) reported that, in New Orleans’ implementation of GVI,
Pre-publication copy | Caterina G. Roman
officers routinely reverted back to standard definitions for coding and assessing violence, even
after multiple trainings. Another related challenge was the complicated relationships among
partners, such as having police officers work closely with community members who have been
vocal about challenging law enforcement policies and are advocating for defunding the police.
In one study that specifically sought to gain a range of stakeholders perspectives about
the systems-level processes and actions related to the execution of GVI, Buggs (2018) found
that, in Baltimore’s implementation of GVI, several stakeholders had concerns about a number of
aspects of model, some of which went beyond implementation failures. One stakeholder
indicated that there were many missed opportunities to talk about racial reconciliation and
develop shared goals. Little discussion of reconciliation and community investment took place.
Other stakeholders voiced their concern that the many moving parts of the model made it
difficult to produce and review data about who was getting arrested and why. Without a review
of enforcement actions, law enforcement are not accountable for careful adherence to the model.
Other Baltimore GVI stakeholders indicated that although social services were part of the model,
the heavy emphasis on the enforcement side created an unbalanced strategy that would
contribute to mass incarceration. It is important to note that in Baltimore, the strategy was not
successful in reducing violence in one of the two target areas where the model was implemented;
the area experienced an 8 percent increase in homicides and a 36 percent increase in shootings
during the months after the intervention.
Roman and colleagues (2020), the evaluators for Philadelphia Focused Deterrence,
studied some of the actions of GVI related to a number of law enforcement levers, but limitations
in the availability of data hindered their ability to link lever-related actions to outcomes. In their
assessment of arrests, probation sanctions and prosecution, they found that, overall, those levers
appeared to be put to use carefully, without resulting in unjustified or excessive arrests or
probation sanctioning. However, they did find that probationers who were members of groups
subject to GVI enforcement actions were issued arrest warrants 300 times the average for the
Anti-Violence Unit probationers (the category of probationers used for a business-as-usual
comparison). Although the data did not permit them to ascertain whether the warrant was
generated because of enhanced attention via the strategy, it should be noted that these warrants
result in detainers. A detainer, often called a “hold,” will result in incarceration even if an
individual can afford bail or would be released otherwise. The probationer must remain jailed
Pre-publication copy | Caterina G. Roman
until they are scheduled to appear or (if a new charge) the case is resolved, which can take
months. These individuals are jailed even under the presumption of innocence. Detainers have
been mostly overlooked as a policy problem, but in recent years, more attention has been brought
to bear on the issue as research has begun to show that even brief stays in jail can amount to
significant personal, social, and fiscal costs (Sawyer et al., 2020.) A related issue that remains to
be studied here is whether due process rights, and related rights, are being violated because the
heightened surveillance on probationers, which resulted in additional criminal justice system
contacts, may not have been warranted if these individuals were unfairly on the gang list.
With regard to proactive policing (under which GVI falls), a salient argument also has
been made that, regardless of constitutional and civil rights issues, there is a duty of responsible
administration by policing agencies, where careful attention to proactive policing techniques in
minority neighborhood must ensue, coupled with the use of discretion that must be accountable
(Sabel and Simon, 2015). Sabol and Simon, both Columbia University legal scholars, argue that
although racially disproportionate harm may not come about through specific intent, it may come
about due to indifference or ignorance. They further state (p.43): “since problem-oriented
policing has rarely been rigorously and comprehensively implemented, it is difficult to separate
limitations that arise from inadequate implementation from those that are inherent in the model.”
The arguments of Sabol and Simon have been echoed by the NASEM Committee on Proactive
Policing in their recent scientific review and resulting report (2018) on proactive policing.
Outside of GVI, a few studies have examined downstream harms from over-surveillance.
Studies show that police practices that involve surveillance and arrest can alienate marginalized
populations and create distrust of law enforcement, regardless of whether the intent is to increase
perceptions of fairness (La Vigne et al., 2014; Tyler et al., 2014; Young and Petersilia, 2016).
Research also suggests that, in turn, low levels of legitimacy may have a causal effect on
violence. Kane (2005) found that incidents of police misconduct and over-policing (i.e.,
indicators of low legitimacy) were linked with increases in violence crime. Other research shows
that when examining the spectrum of gang program evaluations, interventions using law
enforcement officers as the provider of the intervention were more likely to witness increases in
problem behavior by members of the treatment group, compared to members of the non-
intervention group (Rubensen et al., 2020a; Rubensen et al., 2020b).
Limitations of the Current Knowledge Base on GVI
Pre-publication copy | Caterina G. Roman
Discussing these issues and addressing the possible consequences of the GVI model are
important because the overwhelming majority of evaluations of GVI have not examined
outcomes beyond aggregate reductions in violence, nor other relevant changes in norms,
perceptions and beliefs held by those in the targeted groups (NASEM, 2018). Regarding the
former, for an initiative that expects groups to change their behavior, the outcomes measured in
GVI have been predominantly at the community level (e.g., reductions in homicides or shootings
or other relevant violent crimes aggregated across the target areas), not those incidents attributed
to the target gangs.iv GVI espouses changes to individual-level and group-level attitudes, norms
or behaviors; these changes take place in order for community-level change to occur. In a chapter
in Police Innovation: Contrasting Perspectives, David Kennedy (2006), the chief architect of the
strategy, states: “Pulling levers or focused deterrence strategies deploy enforcement, services, the
moral voices of communities, and deliberate communication in order to create a powerful
deterrent to particular behavior by particular offenders” (p.156). Kennedy goes on to state that
special enforcement operations are designed to substantially influence the context of group
behavior to sanction groups whose members commit serious violence. Kennedy is suggesting
that it is general deterrence that is operating, not specific deterrence. When discussing the
successes of the strategy in the same chapter, he further states: “...gang members, violent
offenders, shooters—are substantially changing their behavior without being taken off the street”
(p.162). As such, the suggestion is that the threat of punishment is at work (general deterrence)
as opposed to specific deterrence, which signifies that individuals will refrain from crime after
receiving punishment (Gibbs, 1975). But, to date, no evaluator has explicitly tested how
deterrence is operating; no evaluator has examined their GVI outcomes in tandem with
measurement of the incarceration and incapacitation effects of the strategy (NASEM, 2018). And
we know anecdotally from examples of levers and action that groups in GVI that don’t take the
“don’t shoot” message to heart can be met with arrest and incarceration, sometimes very long
sentences (Kennedy, 2011).
The small handful of studies that have addressed group-level change in violent behavior
(Braga et al., 2014; Papachristos and Kirk, 2015; Roman et al., 2019) show mixed findings. No
studies have examined the extent of removal of individuals from the street and whether this
incapacitation likely played a role in reductions in violence. A few studies have examined some
individual-level outcomes related to reduced offending for gang-focused GVI, but those studies
Pre-publication copy | Caterina G. Roman
simply examined whether or not the individuals were arrested after receiving the GVI message
(or assessed the amount of time until arrest) (see McGarrell, 2020 for a review). Furthermore,
most of these studies found no evidence of reduced re-offending at the individual level
(McGarrell, 2020).v This should lead one to ask a few germane questions: if those individuals did
not change their offending patterns, did anyone change their behavior? Would these individuals
who were arrested have been arrested in absence of GVI? If these individuals were not shooters,
were these individuals deserving of arrest and incarceration? Were the sentences commensurate
with the crime? If the answer is “no” to any of the latter three questions, then what were the
negative effects from these individuals becoming entwined in the criminal justice system? Other
questions to ask include: if the model actually works by the simple threat of deterrence, and not
by removing high-risk offenders and shootings from the street, must there be constant police
presence and surveillance to maintain the effect? One recent study of Kansas City GVI set out to
address this question and found that although the strategy, implemented with fidelity, was able to
significantly reduce homicides and gun-involved aggravated assaults in the first year, the effect
wore off after 12 months (Fox and Novak, 2018). Homicides returned to their pre-
implementation levels in the second year and by the third year (with the program ongoing with
no change in implementation), gun-involved aggravated assaults increased to levels higher than
the pre-implementation period.
In addition to the stark dearth of evaluation studies that consider the questions listed
above, there exists a large void with regard to examining the effect of the strategy on norms and
attitudes. We know little about whether the public and potential targets of GVI believe that police
actually hold the capacity to effect change across participants and the general public with regard
to perceptions and attitudes about police and legitimacy (NASEM, 2018). Brunson, mentioned
earlier as a proponent of GVI, states (2015, p.508): “However, the empirical evidence [on GVI]
is limited about whether offenders, members of their criminal networks, family, and communities
view the police as having the moral, not just legal, authority to promote and enforce the
deterrence message.” He adds a footnote that indicates a few past studies have administered
questionnaires to clients about actions related to sharing the message and attitudes after receiving
the message but that “most of these findings have not been published” (2015, p.508). One recent
study of GVI in Chicago that did survey group members about perceptions did not find any
significant differences between baseline and follow-up surveys (either one or two years apart) on
Pre-publication copy | Caterina G. Roman
measures related to likelihood of sanctions, concerns about violence and precautionary
behaviors, and acceptance of legal authority (Fontaine et al., 2017).
Looking across the evaluations and the outcomes measured, it remains surprising that few
evaluations have attempted to break down the “black box” of focused deterrence. The review
above indicates that across the roughly two dozen impact evaluations of GVI, no one has
examined the likely cause and effect components of this multi-partner strategy in reducing the
violence. When a few studies did delve further than an examination of only aggregate,
community-level outcomes, results mostly show null or negative findings.
So, why aren’t evaluators assessing how the levers play out across the groups and the
individuals within the groups? Why haven’t more evaluators included surveys or other measures
that capture perceptions and values after implementation of the strategy? Without deeper
examination of these aspects, what we are left with is only the assumption that individual-level
behavior change of violent gang members is occurring, based on evidence related to reductions
in aggregate rates of violence across the area targeted (i.e., the area in which the intervention
gangs are located). How can we say GVI leads to community-level change and then ignore the
full range of possible changes to the individuals and communities who become part of the
intervention, including the potentially adverse changes? To answer these questions, I turn now to
possible contextual forces that have culminated in a “what works” criminal justice literature that
appears to favor law enforcement initiatives over community-based solutions for violence and
From Research to Practice
How policy-relevant criminal justice research gets disseminated to government and
community stakeholders with regard to evidence-based practices and programs is fraught with a
number of issues that result in a clouded picture of what really works, what could work, and
what is worth the investment (Prendergast, 2011). Scholars have posited that, with regard to
violent crime and gang-related strategies, these issues have led to a body of evidence in criminal
justice that appears to favor policing initiatives over non-policing initiatives (Curry, 2010; John
Jay College Research Advisory Group on Preventing and Reducing Community Violence, 2020;
Lipsey et al., 2010). Researchers have offered a number of reasons for this, but here I focus on
three reasons: (1) data from police agencies on police outcomes of arrests and criminal activity
tend to be more readily available than outcome data in other substantive domains and further
Pre-publication copy | Caterina G. Roman
down the criminal justice system processing stream (i.e., prosecution, sentencing and
incarceration data); (2) there are simply more policing-based gang programs than non-policing
gang programs, more dedicated funding streams for evaluation of policing programs, and
policing programs are often “easier” or less resource intensive to evaluate than other community-
based programs; and (3) publication biases limit a true understanding of what works and why.
Availability and Validity of Data
Evaluation research on gang programs, including all the evaluations of GVI, has
historically measured reductions in violent crime using law enforcement data (Ridgeway, 2018).
Police data are often available through open data portals and do not require a great deal of time
or resources to analyze, making police administrative data relatively easy to access and analyze.
These data often have standardized definitions for crime types and reporting practices, making
the data generally reliable with regard to comparisons across jurisdictions. Across the different
types of law enforcement and criminal justice system processing data, it is much more difficult
for researchers to obtain and analyze prosecution data (Olsen et al., 2018) and incarceration data
(Binswanger et al., 2019) than police data. These issues may partially explain why so few
evaluations of GVI use court or corrections data. Also, the over-reliance on police administrative
data to understand crime, justice, and program-related outcomes while forsaking outcome data
representing other aspects of criminal justice processing likely has created a body of literature on
what works that is skewed toward policing programs and interventions. Furthermore, the easy
availability of police record data also hinders the likelihood that evaluators of policing programs
will seek criminal justice outcome data that go beyond arrest and incidents, or other non-crime
outcomes for that matter (Eckblom and Pease, 1995), which limits the ability to assess harms
from involvement in police-based interventions.
Although outcome data on arrest and incidents is often readily available to researchers, it
is much less likely that specific information on gangs and gang-related crime is available to
researchers (Klein and Maxson, 2006). Many police departments do not keep consistent and
reliable data on gangs and gang crime over time, seriously limiting the ability for evaluators to
carefully assess the implementation and application of GVI and any group-level outcomes that
are pertinent to the model’s theory of change. Even if data are valid and available to researchers,
this information is often considered “intelligence,” a category of information that is usually not
shareable with researchers because of confidentiality laws. The limitations of gang-related data
Pre-publication copy | Caterina G. Roman
and measures severely hinder the opportunity for evaluators to describe the context of gangs
across the target area for GVI or assess changes in group processes that may be due to the
intervention. Although GVI research teams in Boston and Chicago have used social network
analytical techniques to assess characteristics of gangs related to alliances and conflicts with
other gangs, there has not been published work that assesses changes to within-gang structures
after implementation of GVI (see for example, McGloin, 2005).
Opportunities for and Ease of Evaluation
Research that reviews existing gang policies and programs and complementary
evaluations highlights the dearth of evaluations for community-based gang intervention
programs and prevention programs, as compared to law enforcement programs (Butts and
Roman, 2010; Klein and Maxson, 2006; Wong et al., 2016). These scholars have suggested
evaluation opportunities for community-based models are limited because it is much easier to get
police involved in gang control than neighborhoods and communities (Klein and Maxson, 2006,
p.238), and hence, more policing programs equate to more opportunities to evaluate policing
programs. In addition, policing programs historically have been tied to strong, sustainable
government funding sources that include funding for evaluation that is not similarly available for
community-based or large, multi-partner programs. For instance, the federal government has
promoted GVI across a number of large programs and grant solicitations, and these solicitations
often require or urge grantees to work directly with a research partner.
Under the United States Department of Justice, the programs and grant solicitations
include, but are not limited to, Project Safe Neighborhoods (PSN) via funding to United States
Attorneys’ Offices, the Byrne Criminal Justice Innovation (BCJI) program, and the Community-
Based Violence Prevention (CBVP) program. Each of these platforms has a long-term funding
stream. The PSN platform, which highlights focused deterrence as an evidence-based model, was
established in 2001 and renewed in 2017. PSN funds law enforcement-led crime reduction
efforts in all 94 U.S. judicial districts. In addition, these programs and portfolios often also have
resources to support technical assistance for research and evaluation. Many programs that are not
operating at a level to support evaluation methods or would need a complex, multi-component
research design—typical of community-based gang prevention and intervention programs—are
at a deficit because the lack of funding for evaluation also affects whether a program makes it
onto the synthesized lists of evidence-based or “proven” programs (Hohl et al., 2019; Roman et
Pre-publication copy | Caterina G. Roman
al., 2018). This can lead to a vicious cycle—no evaluation evidence reduces the ability to secure
new funding because the program is not considered evidence-based or there is no solid outcome
data to show a program’s initial successes to a potential funder. Related to this, in general, social
science research funding is scarce, and many funders often have to choose between evaluation
research and direct service provision. Services will take priority (and they should). Particularly in
times of tight budgets and across the philanthropic sector in general, it is rare a research
component gets funded unless research is a stated priority of the funding organization (Freedman
et al., 2013; Crowley and Pearl, 2020).
Publication Bias
Publication bias and selective outcome-reporting bias represent major threats to the
validity of systematic reviews and reduce our ability to produce valid conclusions based on a
body of evidence (Mlinarić et al., 2017). Research has shown that studies published tend to be
for those programs that showed positive results (i.e., publication bias). Selective outcome
reporting occurs when researchers analyze numerous outcomes, such as reductions in nonfatal
shootings, homicides, robberies, but only report on the outcome that showed statistically
significant reductions post-implementation of the intervention. These biases occur not just in the
criminal justice field but across academic disciplines (Fan, 2012; Pratt, 2010; Rothstein, 2008).
These issues are amplified in meta-analytical studies and complicated by research indicating that
weaker research methods correlate with more positive study results. Littell (2008) examined
whether biases existed in research review summary articles in the field of child and family
prevention services by analyzing the methods of extant meta-analyses and found that non-
significant results were often ignored and positive results were over-emphasized. A study that set
out to examine how research design in criminal justice evaluations might be related to outcomes
found that the weaker the research design (i.e. less internal validity), the more likely a study is to
report a result in favor of treatment (Weisburd et al., 2001). The study also found weaker designs
to be associated with a decreased likelihood of reporting a harmful effect of treatment. There are
implications here specific to the evaluation of community-based gang interventions because most
evaluations do not utilize experimental methods, which, if could be used, would help limit some
of these
When all of these “evidence-to-policy and action” issues discussed above are taken into
context with regard to the body of evidence on gang prevention and intervention, the result likely
Pre-publication copy | Caterina G. Roman
skews what policymakers see and understand. This is exemplified by the following account, as
reported by the John Jay College Research Advisory Group on Preventing and Reducing
Community Violence (2020, p.1):
if one searches for research on violence prevention it is immediately apparent that
evaluations of policing interventions are more common than non-policing strategies. A
search of community-based violence interventions on using the
filters for strong evidence (“effective”), topic (“crime and crime prevention”), and setting
(“high crime neighborhood”) yielded only 17 programs. Of those 17 programs, 14
involved the police as either the lead agency or a key partner, and at least 5 of the 14 were
based on the “focused deterrence” law enforcement strategy. Changing the filter for
evidence rating to “promising” yielded 34 programs, again with the overwhelming
majority being police programs and those related to focused deterrence.
Toward a Community Investment-Focused Research Agenda on Gangs
Using GVI as a case study and taking into consideration the issues articulated in this
paper, there are a number of recommendations that come to mind to re-orient gang research and
evaluation. Orienting gang research with an understanding of past limitations can help build a
strong body of evidence on programs and models that utilize the strengths of communities and
support (or has implications for) community investment. As the evidence builds, there will be
greater likelihood that successful efforts can be taken to scale.
For gang reduction interventions that have a law enforcement component, include research
questions assessing aspects of fairness in criminal justice system processes and the footprint
of police surveillance and mass incarceration. With strategies like GVI (or any program),
evaluators should ask, and build compatible research methods to answer, the following
questions: Will this strategy directly lead to reductions in arrests and/or incarceration and
the promotion of a more efficient, effective and fair criminal justice system? Will the law
enforcement levers/actions generate more investigative stops, criminal summonses,
probation detainers, high bail, and misdemeanor arrests—levers that are associated with
racial discrimination, oppression, and inequity? Does the strategy have components that will
make equitable investments in neighborhoods and help strengthen the resiliency of
neighborhoods? Discrimination may not necessarily be conscious or intentional, but it is built
Pre-publication copy | Caterina G. Roman
into the institutional practices that buoy many of the levers of GVI and the criminal justice
system as a whole. As the NASEM Committee on Proactive Policing carefully articulates in
their consensus study report’s chapter on racial bias and disparities in proactive policing
(2018, p.265): “while some [criminal justice] policies might plausibly be considered to be
grounded in “race neutral” reasons, it is critical to understand that these policies can have
important negative economic, social, and health impacts on non-White people in society.”
Relatedly, utilize qualitative methods in gang strategy evaluations to take a deep dive into
exploring causal processes and aspects of criminal justice system functioning that may be
difficult to measure quantitatively or with administrative data. Questions to examine include:
How did law enforcement use their discretionary powers across the intervention? Or, Were
legal values upheld in the policing model? Qualitative methods can also help advance what
we know about the hypothesized intervening psychological constructs that accompany
behavior change, such as inclination to comply with authorities, willingness to pursue group
goals, perceived procedural justice and legitimacy.
Carefully map out a program/strategy’s theory of change and assess the measures needed to
capture the various components. A strong evaluation design will at least align with the
outcomes articulated by the strategy. For GVI, one could ask: What kind of deterrence is
operating? What factors are associated with gang members’ adherence to the “don’t shoot”
message? Take the time to develop a thoughtful evaluation approach and acquire the funding
needed to measure the range of outcomes. This is easier said than done, but it is possible to
obtain funding that allows for resource-intensive research components, such as in-person
surveys, long-term longitudinal data collection, or mixed methods designs incorporating in-
depth and longitudinal participant interviews. Even if original funding sources dry up,
consider obtaining additional funding to unpack new hypotheses that might have arisen
during the evaluation. Connect with colleagues or seek advice from those who have tackled
the more resource-intensive evaluation designs. These partnerships will provide important
learning opportunities and build scholarship in this limited area.
Engage the community in the many phases of gang research, including the development of
research questions and measurement of outcomes. Listen carefully to what is important to
community members, stakeholders, and gang members, and how your research will benefit
communities. A researcher’s description of benefits or “impact” may be very different from
those living in gang-filled neighborhoods and those involved in violent behavior. Relatedly,
Pre-publication copy | Caterina G. Roman
pay attention to scholars who are advocating change in how social science research is
conducted. For instance, Bell (2017; 2019) advocates for radical change (in both method and
theory), stressing that social scientists have not captured the nuanced realities Black
Americans face on a daily basis and how socio-structural issues are linked to crime and
violence across communities (Bell, 2019, p.710). Her recent work offers suggestions and
examples of research methods and frameworks that use community assets and co-produce
knowledge. Another great source of information for thinking about and developing research
designs for community-centered research include the Robert Wood Johnson Foundation’s
research and funding portfolio on Building a Culture of Health.vii
Engage in basic research that allows for a deeper understanding of the group processes
related to gang maintenance and violence. As Ransford and Slutkin suggest (2017, p.19):
“Crucially, the process through which violence spreads must itself be understood –
particularly the importance of social approval and norms, both very powerful forces that are
neurologically driven. The reversal of violence outbreaks requires working on these
processes through the health and other sectors and involving credible health workers with
access, trust and skills.” For instance, research that seeks to uncover how these norms are
developed and transmitted will have huge implications for prevention and intervention. So
too will evaluation research that carefully measures changes in these processes due to
intervention efforts.
Pay attention to definitional issues around the terms “gang” and “gang member” and develop
an understanding of locally-relevant definitions and how they differ by type of stakeholder.
This knowledge will help stakeholders better identify gang youth and those at high risk for
gang involvement who might benefit from innovative programs. We know from the work of
Mark Lipsey (2009) that many program failures have been associated with an inability to
enroll high-risk youth into programs. It is not unusual to find that well-placed and well-
intentioned programs have failed to reach the youth who need help the most (Melde et al.,
2011). Research in this area undoubtedly will also lead to more successes with a youth
population who has not yet engaged in violent behavior or those who have only just begun to
engage in illegal behavior. Evidence of successes on the prevention side of gang programs is
sorely lacking (Klein and Maxson, 2006).
Explicitly seek out evaluation opportunities to study violence and gang reduction programs
and strategies that aim to mitigate racial inequities and bolster social resilience, prosocial
Pre-publication copy | Caterina G. Roman
networks, and legitimacy. As Klein and Maxson (2006, p.88) reiterate: “Evaluated programs
yield the sorts of data we need to build more effective programs.”
Publish findings about failed programs, not just the successful ones (Berman and Fox, 2010).
Many journals are interested in case studies that highlight lessons learned regarding
implementation and program maintenance. In addition, narrative accounts that detail
successful and failed evaluation designs can be particularly useful for strengthening the
evaluation methods and practices in the criminal justice field. More broadly, a nuanced
understanding of failure can turn into innovation such as the program development work in
which distinguished gang experts Irving Spergel and Malcolm Klein became involved. They
both carefully studied the failures of early gang programs (related to both program and
evaluation design failures) and worked with community and government agencies to develop
new and innovative programs and evaluation strategies that could overcome those failures
(Curry, 2010). Their successes became models for the country. Writing about and learning
from challenges and failures can help ensure research integrity and buoy the likelihood that
research will have implications for policy and practice. Essentially, engaging in and
publishing information from systematic process/implementation evaluation studies of gang
interventions would improve the field’s ability to innovate and build evidence.viii
Last, and related to many of the recommendations above, as Burd-Sharps and colleagues
write (2010, p.61): “measure the change you wish to see in the world.” Regardless of the
challenges of any research design, by applying what they term a “well-being framework”—
one that is based in an understanding of how people experience well-being— research will be
more likely to produce findings that have implications for promoting well-being and equity.
For instance, with regard to GVI, most scholars would agree that procedural justice and
police legitimacy are important to communities and have benefits beyond crime reduction,
such as improved clearance rates (Tyler and Fagan, 2008). In addition, GVI espouses the
attainment and spread of anti-violence norms. Imagine such a place where anti-violence
norms and values are widespread and guide all behavior. These constructs that promote well-
being, when part of a theory of change, need to be measured.
In conclusion, it is essential to think broadly about how scholarship dedicated to
generating evidence, in the aggregate, is produced, consumed, and used to effect change—in this
case—for reductions in gang violence. Researchers must push for valid, careful, community-
engaged, comprehensive research designs, and funding that is commensurate. The field of gang
Pre-publication copy | Caterina G. Roman
violence reduction needs sound evaluation studies that can elucidate the causal aspects behind
the critical components of these programs—findings that show that the theory of change is truly
a theory of change. And, as stated in the introduction, these evaluations should always be asking
who benefits and who is burdened?
Pre-publication copy | Caterina G. Roman
Abt, T. and Winship, C. (2016), “What works in reducing community violence: a meta-review
and field study for the Northern Triangle”, available at:
Community-Violence-Final-Report.pdf (accessed 10 December 2020).
Bell, M.C. (2017), “Police reform and the dismantling of legal estrangement”, Yale Law Journal,
Vol. 126 No. 7, pp.2054-2150.
Bell, M.C. (2019), “Safety, friendship, and dreams”, Harvard Civil Rights - Civil Liberties Law
Review, Vol. 54 No. 2, pp.703-739.
Bell, M.C. (2020), “Black security and the conundrum of policing”, Just Security.
Berman, G. and Fox, A. (2010), Trial and Error in Criminal Justice Reform: Learning from
Failure, Urban Institute Press, Washington, DC.
Binswanger, I. A. et al. (2019), “Principles to Guide National Data Collection on the Health of
Persons in the Criminal Justice System”, Public Health Reports, Vol. 134(1_suppl), pp.
Braga, A. A. (2008). Problem-Oriented Policing and Crime Prevention (2nd ed.). Boulder, CO:
Lynne Rienner.
Braga, A.A., and Kennedy, D.M. (2020), A framework for addressing violence and serious crime:
Focused deterrence, legitimacy, and prevention. Cambridge, UK: Cambridge University
Braga, A.A. and Weisburd, D.L. (2012), “The effects of focused deterrence strategies on crime: a
systematic review and meta-analysis of the empirical evidence”, Journal of Research in
Crime and Delinquency, Vol. 49 No. 3, pp.323-358.
Braga, A.A., Weisburd, D.L. and Turchan, B. (2018), “Focused deterrence strategies and crime
control: an updated systematic review and meta-analysis of the empirical evidence”,
Criminology & Public Policy, Vol. 17 No. 1, pp.205-250.
Braga, A.A. and Winship, C. (2006), “Partnerships, accountability, and innovation: clarifying
Boston’s experience with pulling levers”, Weisburd, D. and Braga, A. A. (Ed.s), Police
Innovation: Contrasting Perspectives, Cambridge University Press, Cambridge, England,
Pre-publication copy | Caterina G. Roman
Brantingham, P. J., Sundback, N., Yuan, B., & Chan, K. (2017), “GRYD intervention incident
response & gang crime: 2017 Evaluation Report”, Los Angeles Mayor’s Office of Gang
Reduction and Youth Development, Los Angeles, CA.
Brantingham, P. J., Tita, G., and Herz, D. (2021), “The impact of the city of Los Angeles
Mayor’s Office of Gang Reduction and Youth Development (GRYD) comprehensive
strategy on crime in the city of Los Angeles”, Justice Evaluation Journal, DOI:
Brunson, R.K. (2020), “Protests focus on over-policing. But under-policing is also deadly”, The
Washington Post, 12 June, available at:
crime/2020/06/12/b5d1fd26-ac0c-11ea-9063-e69bd6520940_story.html (accessed 12
June 2020).
Brunson, R.K. (2015), “Focused deterrence and improved police–community relations:
unpacking the proverbial black box”, Criminology & Public Policy, Vol. 14 No. 3, pp.
Buggs, S.A.L. (2018), Evaluating efforts to reduce gun violence in Baltimore: Drug law
enforcement, Cure Violence, and Focused Deterrence (Dissertation), Johns Hopkins
University, available at:
DISSERTATION-2018.pdf?sequence=1&isAllowed=y (accessed 8 January 2021).
Burd-Sharps, S., Lee, B. Y, Matsuoka, K., and Riley, C. (2017). “Measure the change you want
to see in the world.” Plough, A. (Ed.) Knowledge to Action: Accelerating Progress in
Health, Well-being, and Equity, Oxford University Press, pp.61-82.
Butts, J.A., and Roman, C.G. (2010), “A community youth development approach to gang
control programs”, Chaskin, R., (Ed.) Youth Gangs and Community Intervention,
Columbia University Press, pp.175-205.
Butts, J.A., Roman, C.G., Bostwick, L. and Porter, J.R. (2015), “Cure Violence: a public health
model to reduce gun violence”, Annual Review of Public Health, Vol. 36 No. 1, pp.39-53.
Byers, C. 2021, KSDK News. Will St. Louis Mayor-elect Tishaura Jones' crime deterrence plan
work? The I-Team asks the experts (April 7). Available:
Pre-publication copy | Caterina G. Roman
experts/63-91a55f83-8e8b-409d-80ec-8b9c39059a87 (Accessed April 8, 2021)
Cahill, M., Jannetta, J. Tiry, E., Lowry, S., Becker-Cohen, M., Paddock, E., and Serakos, M.
(2015), “Evaluation of the Los Angeles Gang Reduction and Youth Development
Program. Year 4 evaluation report”, Available at:
development-program-year-4-evaluation-report/view/full_report (accessed March 30, 2021).
Center for Policing Equity (2020), “Center for Policing Equity releases critical steps for
exploring how public safety resources are allocated”, available at:
resources-allocation (accessed 5 January 2021).
Corsaro, N. and Engel, R.S. (2015), “Most challenging of contexts: assessing the impact of
focused deterrence on serious violence in New Orleans”, Criminology & Public Policy,
Vol. 14 No. 3, pp.471-505.
Crowley, M. and Pearl, B. (2020), “Reimagining federal grants for public safety and criminal
justice reform”, Center for American Progress, Washington, D.C., Available at:
justice-reform/ (Accessed 15 December 2020).
Curry, G.D. (2010), “From knowledge and response and back again: theory and evaluation in
responding to gangs”, Chaskin, R., (Ed.) Youth Gangs and Community Intervention,
Columbia University Press, pp.109-126.
Deuchar, R. (2013), Policing Youth Violence: Transatlantic Connections, IOE Press, London,
Dunworth, T., Hayeslip, D., Lowry, S., Kim, K., Kotonias, C., & Pacifici, L. (2013), “Evaluation
of the Los Angeles Gang Reduction and Youth Development Program. Year 3 final
report”, Washington, D.C.: The Urban Institute, Harder Company. Available at:
Report.pdf (accessed March 30, 2021)
Pre-publication copy | Caterina G. Roman
Engel, R.S., Tillyer, M.S. and Corsaro, N. (2013), “Reducing gang violence using focused
deterrence: evaluating the Cincinnati Initiative to Reduce Violence (CIRV)”, Justice
Quarterly, Vol. 30 No. 3, pp.403-439.
Fan, M. D. (2012), "Panopticism for Police: Structural Reform Bargaining and Police Regulation
by Data-Driven Surveillance", Wash. L. Rev. 87, 93.
Fontaine, J. Jannetta, J., Papachristos, A.V., and Leitson, D. (2017), “Put the guns down:
outcomes and impacts of the Chicago Violence Reduction Strategy”, Urban Institute
Report, Washington, DC. Available:
inalized_0.pdf (Accessed 2 December 2019).
Fox, A.M. and Novak, K.J. (2018), “Collaborating to reduce violence: the impact of focused
deterrence in Kansas City”, Police Quarterly, Vol. 21 No. 3, pp.283-308.
Freedman, A. M., Kuester, S. A., and Jernigan, J. (2013), "Peer reviewed: evaluating public
health resources: what happens when funding disappears?" Preventing chronic disease,
Vol. 10. Pp.E190.
Gase, L. N., Glenn, B. A., Gomez, L. M., Kuo, T., Inkelas, M., & Ponce, N. A. (2016),
Understanding Racial and Ethnic Disparities in Arrest: The Role of Individual, Home,
School, and Community Characteristics. Race and social problems, Vol. 8, Issue 4, pp.
Griffiths, E. and Christian, J. (2015), “Considering focused deterrence in the age of Ferguson,
Baltimore, North Charleston, and beyond”, Criminology & Public Policy, Vol. 14 No. 3,
Harcourt, B.E. (2009), Illusion of Order: The False Promise of Broken Windows Policing,
Harvard University Press, Cambridge, MA.
Hennigan, K. and Spanovic, M. (2012), “Gang dynamics through the lens of social identity
theory”, Esbensen, F.A. and Maxson, C.L. (Ed.s), Youth Gangs in International
Perspective, Springer, New York, NY, pp.127-149.
Hohl, B. C., Kondo, M.C., Kajeepeta, S. MacDonald, J. M. Theall, K., Zimmerman, M.A., and
Branas, C. C. (2019), “Creating safe and healthy neighborhoods with place-based
violence interventions”, Health Affairs, Vol. 38 No. 10, pp.1687-1694.
Pre-publication copy | Caterina G. Roman
Huey, S.J.J., Lewine, G. and Rubenson, M. (2016), “A brief review and meta-analysis of gang
intervention trials in North America”, Maxson, C.L. and Esbensen, F.A. (Ed.s), Gang
Transitions and Transformations in an International Context, Springer, New York, NY,
Hyde, L.W., Shaw, D.S. and Moilanen, K.L. (2010), “Developmental precursors of moral
disengagement and the role of moral disengagement in the development of antisocial
behavior”, Journal of Abnormal Child Psychology, Vol. 38 No. 2, pp.197-209. DOI:
Illing, S. (2020), The “abolish the police” movement, explained by 7 scholars and activists. (June 12):
Jaitman, L. (2019), “Frontiers in the economics of crime: lessons for Latin America and the
Caribbean”, Lat Am Econ Rev, Vol. 28 No. 1, pp.1-36. DOI: 10.1186/s40503-019-0081-5
John Jay College Research Advisory Group on Preventing and Reducing Community Violence
(2020), “Reducing violence without police: a review of research evidence”, New York,
NY: Research and Evaluation Center, John Jay College of Criminal Justice, City
University of New York. Available at:
(Accessed 14 December 2020).
Kane, R.J. (2005), “Compromised police legitimacy as a predictor of violent crime in structurally
disadvantaged communities”, Criminology, Vol. 43 No. 2, pp.469-498.
Kennedy, D.M. (2006), “Policing and the lessons of Focused Deterrence”, D.L. Weisburd and
A.A. Braga (Ed.s), Police Innovation: Contrasting Perspectives, Cambridge University
Press, New York, NY, pp. 205-226.
Kennedy, D.M. (2011), Don’t shoot: One man, a street fellowship, and the end of violence in
inner-city America, New York, NY: Bloomsbury.
Kennedy, D.M., Piehl, A.M., and Braga, A.A. (1996), “Youth violence in Boston: Gun markets,
serious youth offenders, and a use-reduction strategy”, Law and Contemporary
Problems, Vol. 59, pp. 147-96.
Klein, M.W. and Maxson, C.L. (2006), Street Gang Patterns and Policies, Oxford University
Press, New York, NY.
Pre-publication copy | Caterina G. Roman
La Vigne, N.G., Lachman, P., Rao, S. and Matthews, A. (2014), “Stop and frisk: balancing crime
control with community relations”, available at:
Balancing-Crime-Control-with-Community-Relations.PDF (accessed 7 November 2020).
Lipsey, M.W. (2009), “The primary factors that characterize effective interventions with juvenile
offenders: a meta-analytic overview”, Victims and Offenders: An International Journal of
Evidence-based Research, Policy, and Practice, Vol. 4 No. 2, pp.124-147.
Lipsey, M. W., Howell, J. C., Kelly, M. R., Chapman, G., & Carver, D. (2010), “Improving the
effectiveness of juvenile justice programs: A new perspective on evidence-based
practice”, Washington, DC: Georgetown University, Center for Juvenile Justice Reform.
Littell, J. H. (2008). "Evidence-based or biased? The quality of published reviews of evidence-
based practices", Children and Youth Services Review, Vol. 30 No. 11, pp.1299-1317.
Lofstrom, M., and Raphael, S. (2016), Crime, the criminal justice system, and socioeconomic
inequality. Journal of Economic Perspectives, Vol. 30, No. 2, pp.103–126.
MacDonald, J. and Donnelly, E. (2019),“Evaluating the role of race in sentencing: an entropy
weighting analysis.” Justice Quarterly, Vol. 36, No. 4, pp.656- 681.
Manchester, M. (2018). “Focused Deterrence: an alternative justice system in South
justice-system-in-south-philadelphia/ (Accessed 09 January 2021).
McGarrell, E., (2020), “Focused deterrence violence prevention at community and individual
levels”, Marquette Law Review, Vol. 103, No. 3, pp.963.
McGloin J.M. (2005), “Policy and intervention considerations of network analysis of street
gangs”, Criminology and Public Policy, Vol. 4, pp.607–635.
Melde, C. Gavazzi, S., McGarrell, E.F., and Bynum, T. (2011), “On the efficacy of targeted gang
interventions: can we identify those most at risk?”, Youth Violence and Juvenile Justice,
Vol. 9 No. 4, pp. 279–294.
Mlinarić, A., Horvat, M., and Smolcic, V.S. (2017), “Dealing with the positive publication bias:
Why you should really publish your negative results”, Biochemia medica, Vol. 27 No. 3,
Pre-publication copy | Caterina G. Roman
National Academies of Sciences, Engineering, and Medicine [NASEM]. (2018), Proactive
policing: Effects on crime and communities, Washington, DC: The National Academies
National Network for Safe Communities. (2016), “Group Violence Intervention: an
implementation guide”, Washington, DC: US Department of Justice, Community
Oriented Police Services. Available at:
content/uploads/2017/10/GVI_Guide_2016.pdf (accessed 14 January 2021).
Olsen, Robin, Courtney, L., Warnberg, C., and Samuels, J. (2018), "Collecting and Using Data
for Prosecutorial Decisionmaking", Findings from 2018 National Survey of State
Prosecutors’ Offices, The Urban Institute.
Papachristos, A.V. and Kirk, D.S. (2015), “Changing the street dynamic: evaluating Chicago’s
Group Violence Reduction Strategy”, Criminology and Public Policy, Vol. 14 No. 3,
Piquero, A.R. (2008), Disproportionate minority contact. The Future of Children, Vol. 18, No. 2,
pp. 59-79.
Police Advisory Commission (2020), “Data-driven efforts to combat gun violence: a discussion
on civil liberty protections”, available at: (accessed 6
November 2020)
Pratt, Travis C. (2010), "Meta‐analysis in criminal justice and criminology: what it is, when it's
useful, and what to watch out for", Journal of Criminal Justice Education Vol 21, no. 2,
Ransford, C. and Slutkin, G. (2017), “Seeing and treating violence as a health issue”, Brookman,
F., Maguire, E.R. and Maguire M. (Eds.), The Handbook of Homicide, Wiley-Blackwell,
Chichester, England, pp.601-625.
Prendergast, Michael L. “Issues in defining and applying evidence-based practices criteria for
treatment of criminal-justice involved clients”, Journal of psychoactive drugs Vol. Suppl
7 (2011): 10-8.
Pre-publication copy | Caterina G. Roman
Rivlin-Nadler, M. (2018), “How Philadelphia’s social media-driven gang policing is stealing
years from young people”, The Appeal, 19 January, available at:
years-from-young-people-fa6a8dacead9/ (accessed 1 February 2018).
Roman, C.G., Forney, M., Hyatt, J.M., Klein, H.J. and Link, N.W. (2020), “Law enforcement
activities of Philadelphia’s Group Violence Intervention: an examination of arrest, case
processing, and probation levers”, Police Quarterly, Vol. 23 No. 2, pp.232-261.
Roman, C.G., Link, N.W., Hyatt, J.M., Bhati, A. and Forney, M. (2019), “Assessing the gang-
level and community-level effects of the Philadelphia Focused deterrence strategy”,
Journal of Experimental Criminology, Vol. 15 No. 4, pp.499-527.
Roman, C. G.. Klein, H. J., and Wolff, K. T. (2018). "Quasi-experimental designs for
community-level public health violence reduction interventions: a case study in the
challenges of selecting the counterfactual", Journal of Experimental Criminology, Vol. 14
No. 2, pp.155-185.
Rothstein, H.R. (2008), “Publication bias as a threat to the validity of meta-analytic results.”
Journal of Experimental Criminology, Vol. 4, pp.61–81.
Rowe, M. and Sogaard, T.F. (2020), “‘Playing the man, not the ball’: targeting organised
criminals, intelligence and the problems with pulling levers”, Policing and Society, Vol.
30 No. 2, pp.120-135.
Rubenson, M., Galbraith, K. and Huey, S.J. (2020a), “Understanding adverse effects in gang-
focused interventions: a critical review”, Melde, C. and Weerman, F. (Eds.), Gangs in the
Era of Internet and Social Media, Springer International, pp.271-290.
Rubenson, M. P., Galbraith, K., Shin, O., Beam, C. R., & Huey, S. J. J. (2020b), “When helping
hurts? Toward a nuanced interpretation of adverse effects in gang-focused interventions”,
Clinical Psychology: Science and Practice. Advance online publication.
Sabel, C.F. and Simon, W.H. (2014), “Due process of administration: the problem of police
accountability [2507280]”, working paper, Stanford Law School, Stanford University,
Stanford, 1 October.
Sawyer, W., Jones, A. and Troilo, M. (2020), “Technical violations, immigration detainers, and
other bad reasons to keep people in jail: how ICE and probation and parole detainers (or
“holds”) contribute to unnecessary jailing”, Prison Policy Initiative, 18 March, available
Pre-publication copy | Caterina G. Roman
at: (accessed 10 November
Scott, M.S. (2017), Focused deterrence of high-risk individuals. Strategies for policing
innovation: Problem-oriented guides for police response guide series Number 13. CNA
Corporation and Arizona Board of Regents.
Short, J. and Hughes, L. (2010), Promoting research integrity in community-based intervention
research, Chaskin, R., (Ed.) Youth Gangs and Community Intervention. Columbia
University Press, pp.127-151.
Simonson, J. (2021), Police Reform through a Power Lens. The Yale Law Journal, Vol. 130 No.
4. Available:
(Accessed March 30, 2021).
Thacher, D. (2016). "Channeling police discretion: The hidden potential of focused deterrence,"
University of Chicago Legal Forum, Vol. 2016, Article 13, pp. 533-577.
Travis, J. (2011), “Summoning the superheroes: harnessing science and passion to create a more
effective and humane response to crime.” Keynote address on the occasion of the twenty-
fifth anniversary of The Sentencing Project. (October 11, 2011). Washington, DC.
Tyler, T.R., and Fagan, J. (2008), Legitimacy and Cooperation: Why do people help the police
fight crime in their communities? Ohio State Journal of Criminal Law, Vol. 6, No.
Tyler, T.R., Fagan, J. and Geller, A. (2014), “Street stops and police legitimacy: teachable
moments in young urban men’s legal socialization”, Journal of Empirical Legal Studies,
Vol. 11 No. 4, pp.751-785.
Vitale, A. (2017), The end of policing. London: Verso Books.
Walsh, C. (2021), Solving racial disparities in policing. The Harvard Gazette (February 23).
policing/ (Accessed March 30, 2021).
Webster, D.W., Whitehill J.M., Vernick, J.S. and Curriero, F.C. (2013), “Effects of Baltimore’s
Safe Streets Program on gun violence: a replication of Chicago’s CeaseFire program”,
Journal of Urban Health, Vol. 90 No. 1, pp.27-40.
Pre-publication copy | Caterina G. Roman
Weisburd, D., Lum, C.M. and Petrosino, A. (2001), "Does research design affect study outcomes
in criminal justice?", The Annals of the American Academy of Political and Social
Science Vol. 578 No. 1, pp.50-70.
Wellford, C.F., Pepper, J.V. and Petrie, C.V. (2005), Firearms and Violence: A Critical Review,
The National Academies Press, Washington, DC.
Wennmann, A. (2018), “Daring the unconventional on the pathways for peace: on the ‘how’ of
sustaining peace in the city”, Wennmann, A. and Jütersonke, O. (Ed.s), Urban Safety and
Peacebuilding: New Perspectives on Sustaining Peace in the City, Taylor & Francis
Group, Milton Park, England, pp.178-194.
Whitehill, J.M., Webster, D.W., Frattaroli S. and Parker, E.M. (2014), “Interrupting violence:
how the CeaseFire program prevents imminent gun violence through conflict mediation”,
Journal of Urban Health, Vol. 91 No. 1, pp.84-95.
Whitehill, J.M., Webster, D.W. and Vernick, J.S. (2013), “Street conflict mediation to prevent
youth violence: conflict characteristics and outcomes”, Injury Prevention, Vol. 19 No. 3,
Wong, J. S., Bouchard, J., Gravel, J., et al. (2016), “Can at-risk youth be diverted from crime?: a
meta-analysis of restorative diversion programs”, Criminal Justice and Behavior, Vol. 43
No. 10, pp.1310-1329.
Wood, J. L., Alleyne, E. and Beresford, H. (2016), “Deterring gangs: criminal justice approaches
and psychological perspectives”, Bornstein, B.H. and Miller, M.K. (Ed.s.), Advances in
Psychology and Law: Vol. 2, Springer International, pp.305-336. DOI: 10.1007/978-3-
Young, K. M. and Petersilia, J. (2016), “Keeping track: surveillance, control, and the expansion
of the carceral state”, Harvard Law Review, Vol. 129 No. 7, pp.1318-1360.
i The term “gang” is used interchangeably with “group.”
ii Gang injunctions are used to limit gatherings of gangs in specific area. The laws typically specify that certain
individuals cannot occupy or use a specified area in a manner that constitutes a public nuisance. Gang injunctions have also
been criticized for violations of the 1st, 4th, 5th, and 9th Amendments.
iii It also is worth noting that large, multi-component gang reduction programs that balance an array of prevention
and intervention services can be successful, as evidenced by recent studies evaluating the efforts of Los Angeles’ Gang
Reduction and Youth Development (GRYD) strategy (Brantingham et al., 2017; Brantingham et al., 2021). Brantingham and
colleagues’ most recent evaluation measured the combined effects of community engagement, prevention, intervention and
violence interruption and found significant reductions in violent crime. Furthermore, independent implementation
assessments of GRYD found the strategy to be well-designed and, overall, carefully implemented (Cahill et al., 2015;
Dunworth, Hayeslip et al., 2013).
iv In Table 2 of their systematic review and meta-analysis of focused deterrence strategies (including those targeted
to gangs), Braga, Weisburd and Turchan (2018) list out the focal outcome measures for the 24 evaluations included in their
review. The strongest evidence of success came from those studies where focused deterrence was aimed at gangs, as
opposed to individuals or drug markets.
v One study found GVI-focused individuals who attended a call-in notification meeting had longer times to arrest
than a matched comparison group (Circo, et al., 2020).
vi I recognize the inherent difficulty of utilizing experimental methods in community-based evaluations that assess
strategies across large geographic areas. A detailed discussion of these challenges can be found in Roman, Klein, and Wolff,
vii See
viii Although many criminal justice evaluators engage in process analyses alongside impact evaluation, a look at
published journal articles yields few process evaluations on gang interventions. Research that includes detailed information
on program process and implementation often are left to unpublished reports or technical reports that are not widely
... The academic spotlight has become fixated on gangs due to what is often referred to as the "gang problem", i.e., the observed relationship between gangs (and gang membership) and a range of negative social, economic, and health related outcomessuch as crime, victimisation, homelessness, low educational and employment achievement, and mental health difficulties (Chu et al., 2012;Pyrooz, 2014;Pyrooz et al., 2016;. Despite the severity of harms constitutive of the gang problem, there is still much ground to be covered in the development of effective, reliable, ethical, and evidence-based responses (Bjerregaard, 2015;Boxer, 2019;Decker, 2016;Klein, 2001;Mallion & Wood, 2020b;McDaniel & Sayegh, 2020;Pyrooz & Decker, 2019;Roman, 2021;Thornberry et al., 2018). ...
... While the core aims of approaches such as focused-deterrence are to understand and address the causes, triggers, and facilitators of crime, evaluations of focused deterrence programmes have been labelled as "black box" evaluations with little to no examination of the specific programmatic elements thought to be responsible for changes in offence-related outcomes Roman, 2021). In discussing focused deterrence approaches, Roman (2021) states: …across the roughly two dozen impact evaluations of GVI, no one has examined the likely cause and effect components of this multi-partner strategy in reducing violence. ...
... While the core aims of approaches such as focused-deterrence are to understand and address the causes, triggers, and facilitators of crime, evaluations of focused deterrence programmes have been labelled as "black box" evaluations with little to no examination of the specific programmatic elements thought to be responsible for changes in offence-related outcomes Roman, 2021). In discussing focused deterrence approaches, Roman (2021) states: …across the roughly two dozen impact evaluations of GVI, no one has examined the likely cause and effect components of this multi-partner strategy in reducing violence. ...
Full-text available
Gangs are associated with a range of social, criminal, and economic harms. Yet, after almost a century of dedicated research, the development of effective and ethical responses to such harms has proven difficult. Recent attempts to address this have seen the establishment of the Eurogang Program; an international group of gang researchers and practitioners coordinated around a consensus definition of gangs. Since its recent inception, the Eurogang Program has quickly become the dominant framework of research and practice. While much is being staked on the success of the Eurogang Program, the suitability of such a programme for progressing gang research is yet to be thoroughly examined. In this thesis I therefore conduct a meta-theoretical examination of the state of gang research and particularly the Eurogang Program and its associated practices. By examining the frameworks underpinning gang research and drawing upon insights from the philosophy of science, I characterise the Eurogang approach as an attempt to coordinate gang research through means of unification (i.e., through the privileging of particular research perspectives and strategies to achieve coordination through consensus). I draw attention to some major limitations of these unificatory attempts and emphasise how the consensus Eurogang definition does not appropriately set up researchers to be able to develop the various kinds of conceptual and theoretical understandings of gangs required to improve gang policy and practice. Instead, I make the case for a framework known as epistemic pluralism, in which researchers do not pursue consensus but rather cultivate multiple systems of knowing to serve a variety of different research purposes. After establishing the benefit of epistemic pluralism, I examine how such a framework may be applied to the gang field. This involves specific discussion of the various aims of gang researchers and the roles that conceptual strategies (i.e., definitional, classificatory, and explanatory approaches) play in providing the pragmatic and epistemic (i.e., knowledge-related) insights required to meet them. These discussions offer a novel perspective on the roles of conceptual strategies in the process of knowledge production and justification. Having established the general kinds of strategies required for different research purposes, I then consider some specific examples of conceptual strategies that are relevant to meeting the various needs of gang researchers. This takes the form of the Conceptual Framework for Gang Research (CFGR). This novel approach offers greater opportunities for more meaningful kinds of research coordination and maximises the likelihood of establishing the conceptual and theoretical understandings of gangs required to improve gang policy and practice. The value of this thesis as a case study for pursuing epistemic pluralism in the sciences is also discussed.
... Furthermore, given current challenges involving police legitimacy, community-police relations, and procedural justice, empirical evaluations should assess the perceptions and responses of the individuals subject to these interventions (Griffiths & Christian, 2015). To that extent, evaluations should consider whether GVI may increase over-surveillance for some individuals (Roman, 2021;Webster, 2022). ...
... Strategies such as GVI seek to improve the public's perception of procedural justice, especially in neighborhoods in which governmental entities have historically imposed overly aggressive and disparate criminal justice sanctions (Braga, 2012;Kennedy, 2019;Braga, Weisburd, & Turchan, 2018;Braga & Kennedy, 2020;Webster, 2022;Roman, 2021). By tailoring intensive crime-prevention efforts at specific individuals and groups involved in problem activities, GVI inherently differs from law enforcement strategies that indiscriminately surveille entire communities. ...
... While it is true that Glasgow had great success with focused deterrence and other violence reduction measures imported from the US (Deuchar, 2013;Williams et al., 2014), London did not (Densley & Jones, 2016). I am generally an advocate for focused deterrence, but the UK must think about the risks of different responses to gangs in the absence of rigorous evidence (especially evidence gathered locally), which often are higher (in terms of negative externalities for communities) with police-led initiatives than they are for community-led programmes, all else being equal (Roman, 2021). ...
In this chapter, ‘US and UK Gangs: Research, Policy and Practice’, James Densley takes stock of what we now know about UK gangs in comparison to their US counterparts. He therefore examines points of convergence and divergence, and highlights what Britain can learn from the US experience of gangs and vice versa in order to set a research agenda for the next decade. He notes that although there is reluctance among UK criminologists to embrace gang research, UK gang studies should be viewed as foundational to UK criminology. They are important, he argues, for understanding the lives of children and young people because gangs are integral social groups for those involved. Moreover, gang membership changes people’s lives and there is a pressing need to respond to its consequences.
... Furthermore, she highlights many mechanismsincluding practices of federal funding agencies and researchersthat have led us to a situation where the field's policy recommendation becomes almost inevitably tilted toward law enforcement-led strategies. Roman (2021) offers several insightful recommendations for the field to move forward. ...
Full-text available
In this editorial, we want to highlight areas where we, as gang scholars, need to be better. We believe that many of us, unwittingly or not, have been part of the problem. As a community of scholars, we need to recognize the power we have in guiding perceptions of a phenomena and reactions to it. We argue that most of our past failures have been to inadequately use this power or even failing to recognize that we have it. Through our silence and tacit endorsement of the status quo, we have allowed the criminal justice system and its actor to take ownership of the concept of the gang and let them – along with the media – proliferate ideas and myths we know are wrong. We watched – and sometimes assisted – as they used the mythical idea of the gang to justify policies and practices that have contributed to uphold White supremacy and destroyed trust in institutions among communities of color.
Spurred by the success of public health violence interventions, and accelerated by policy pressure to reduce violence without exacerbating overpolicing and mass incarceration, streetwork programs—those that provide anti‐violence services by neighborhood‐based workers who perform their work beyond the walls of parochial institutions—have positioned themselves as the most important non–law‐enforcement violence prevention option available to urban policy makers. Yet despite their importance, the state of the field seems difficult to interpret for academics and practitioners alike. In this article, we make several contributions that bring forth new findings and deliver new perspectives on streetwork as a violence reduction strategy. First, we offer an extended analytic review of the streetwork evaluation literature that connects the study of contemporary public health violence interventions to a preceding tradition of criminologically inspired streetwork studies. Second, we present the results of an impact evaluation of StreetSafe Boston (SSB)—a multiyear streetwork intervention that served 20 Boston gangs. We find that the SSB intervention had no detectable effect on violence among the gangs that it served. We conclude by offering a framework for understanding a field at multiple crossroads: past and present, proclaimed successes and failures, help and harm.
Full-text available
The City of Los Angeles Mayor’s Office of Gang Reduction and Youth Development (GRYD) program was conceived as a comprehensive response to gang violence. Unlike most comprehensive approaches, suppression was excluded from the primary model. Program services including community engagement, gang prevention and intervention services, and street-based violence interruption, were formally launched in late 2011. Strict geographic eligibility criteria mean that GRYD services were available in some Los Angeles communities and not others. Using the geographic structure of GRYD, we use a place-based difference-in-differences model to estimate the effect of GRYD services on both violent and property crime. The analyses suggest a reduction in violent crime of around 18% in areas exposed to GRYD Comprehensive Strategy services, including aggravated assault and robbery. Similar declines are not observed in property crimes including burglary and car theft. Comparison with evaluations of placed-based gang injunctions demonstrate that GRYD is able to achieve nearly one-half of the reductions in crime without a suppression focus.
Full-text available
The number of jurisdictions implementing focused deterrence strategies targeted at gangs continues unabated. Although recent research suggests positive impacts of the strategy on reductions in gun violence, little is known about the particular mechanisms operating behind the strategy. This article provides a descriptive analysis of the law enforcement activities or levers undertaken after enforcement operations in Philadelphia as a part of the focused deterrence strategy. The article quantifies the execution of levers related to arrest, case processing, and probation sanctioning during enforcement activities after shootings. The results show that Philadelphia achieved success in implementing the enforcement levers as intended, and there was little evidence that arrest practices were overly aggressive. The authors suggest that future evaluations seek to carefully document the wide array of levers used in concert with an assessment of community understanding of and reactions to the strategy as well as an examination of reactions of group members targeted.
Full-text available
Crime and violence generate many distortions in the allocation of private and public resources and engender economic and social costs that hinder development. In Latin America and the Caribbean, which is the most violent region on earth, the costs of crime represent at least 3.5% of the regional gross domestic product, twice as much as in developed countries. Despite the magnitude of the security problem, the region is lagging in the production of rigorous research on crime and the application of evidence-based policies to fight and deter crime. This paper uses the crime economics framework to shed light on the main drivers of crime and proposes avenues for future research and action in the region to reduce crime and its social and economic costs.
Full-text available
Objectives Violence reduction initiatives based on focused deterrence strategies have gained attention in recent years due to their empirical support. The evaluations have generally assessed the impact of this intervention on trends in gun violence at the aggregate level, but not at the gang level. The current study evaluates both the community- and gang-level impacts of the Philadelphia Focused Deterrence strategy. Methods The intervention was assessed using a quasi-experimental design that measured trends in shootings over a twelve-year period, including two years after the implementation of the initiative. Propensity scoring and matching techniques were used to match neighborhoods and gangs, and a number of regression models were run to assess impact. ResultsAlthough a statistically significant reduction in total shootings across the treated neighborhoods was observed when compared to matched neighborhoods, the findings at the gang level were mixed. Models comparing shootings around gang territories showed significant reductions when compared to shootings around the territories of matched gangs, but pre-post-only models of treated gangs using the more rigorous measure of gang-involved shootings did not show evidence of impact. Conclusions The findings suggest that focused deterrence may provide a mechanism for general deterrence among a broad pool of potential offenders. Specifically, violent gangs, even when targeted, may not be affected similarly for a variety of reasons. To better understand who is receiving the deterrence message and responding to it, future evaluations of focused deterrence strategies, when assessing impact, should include measures of the dosage of the message and other components relative to individuals and their groups.
Full-text available
This research examines the impact of focused deterrence on homicide and gun violence in Kansas City, MO. In 2014, a coalition of police, prosecutors, city officials, researchers, and others implemented Kansas City No Violence Alliance, a focused deterrence violence reduction strategy. Using street-level intelligence and analysis, groups involved with violence were identified and notified of the consequences for future violent incidents. Leveraging existing social services, members opting for nonviolence were offered assistance. This study evaluates the impact on violence over 3 years of implementation. Using 2009–2016 police incident data on homicide (including group member involved homicide) and gun-involved aggravated assault, time series models were estimated to determine the effects of focused deterrence during 2014–2016. Analysis indicated that focused deterrence implementation resulted in an immediate reduction in homicides and gun-involved aggravated assaults. This effect began to diminish around the 12-month postintervention point. During the third year, overall and group member involved homicide numbers returned to preimplementation levels, and gun-involved aggravated assaults exceeded those levels. After achieving significant first-year reductions, despite robust implementation and fidelity, violence returned to preimplementation levels by the third year. Limitations to the focused deterrence model and the need for continuous evaluation and innovation are discussed.
Programs that aim to reduce gang involvement and violence can unintentionally produce adverse outcomes. A recent systematic review identified 41 controlled evaluations of gang-focused interventions, eight of which produced statistically significant adverse effects (i.e., effects that favored control groups). Understanding what caused these effects and whether they indicate that programs actually harmed participants requires careful investigation. In this critical review, we provide an overview of how gang-focused interventions can yield adverse effects and specify when effects are more likely to indicate harmful programs. We then critically review four program evaluations with adverse effects that exemplify different implementation problems (i.e., implementation failure), faulty theories about behavior change (i.e., theory failure), and evaluation methods that did not adequately measure outcomes (i.e., measurement failure). We offer hypotheses about what may have caused these adverse effects and conclude by recommending ways to maximize confidence that measured outcomes reflect real intervention effects, rather than artifacts of research design, using both time-tested methods and new technologies.
Violence is a leading cause of death and disability in the United States and abroad, with far-reaching consequences for individuals and communities. Interventions that address environmental and social contexts have the potential for greater populationwide effects, yet research has been slow to identify and rigorously evaluate these types of interventions to reduce violence. Several urban communities across the US are conducting experimental and quasi-experimental community-based research to examine the effect of place-based interventions on violence. Using examples from Philadelphia, Pennsylvania; Flint, Michigan; Youngstown, Ohio; and New Orleans, Louisiana, we describe how place-based interventions that remediate vacant land and abandoned buildings work to reduce violence. These examples support the potential for place-based interventions to create far-reaching and sustainable improvements in the health and safety of communities that experience significant disadvantage. These interventions warrant the attention of community stakeholders, funders, and policy makers.
Focused on deterrence popular model to address community-level violence, however little research has examined the individual-level effect of deterrent messaging on subsequent offending. To answer this question, we utilize data on 254 gang- and group-involved probationers and parolees who attended offender “call-in” meetings as part of the Detroit Ceasefire. We employ inverse-probability weighting to construct a counterfactual comparison group from a sample of gang-involved young adults who were not subject to the Ceasefire call-in. We then use a Cox regression to estimate time to re-arrest. We find that individuals who were delivered a deterrent message at a call-in meeting had a longer time to re-arrest compared to a weighted comparison group for up to 3 years following the meeting.
In efforts to combat organised crime, police forces have adopted variations on the pulling levers approach to individuals and groups identified with gun crime, drug supply and other serious offences. Once identified, those individuals and their networks are targeted for interventions from criminal justice agencies and their partners. When levers are pulled, criminals find their lives made intolerable by the attentions of multiple agencies. Identifying the right people for this sort of attention and the quality and currency of police intelligence are, then, key concerns for such strategies. But the choice of levers, and their implications also raise some difficult questions. The law is explicitly applied differently to those associated with organised crime than to anyone else. This paper reviews evidence from two ethnographic studies, one of police officers in three police forces in England and Wales, the other from a third-party policing arrangement in the Danish night-time economy. We seek to understand the ways in which levers are understood and used, raising questions about the efficacy of pulling levers.
Research Summary Focused deterrence strategies are increasingly being applied to prevent and control gang and group‐involved violence, overt drug markets, and individual repeat offenders. Our updated examination of the effects of focused deterrence strategies on crime followed the systematic review protocols and conventions of the Campbell Collaboration. Twenty‐four quasi‐experimental evaluations were identified in this systematic review. The results of our meta‐analysis demonstrate that focused deterrence strategies are associated with an overall statistically significant, moderate crime reduction effect. Nevertheless, program effect sizes varied by program type and were smaller for evaluations with more rigorous research designs. Policy Implications The available empirical evidence suggests these strategies generate noteworthy crime reduction impacts and should be part of a broader portfolio of crime reduction strategies available to policy makers and practitioners. Investments still need to be made, however, to strengthen the overall rigor of program evaluations and improve our understanding of key program activities associated with observed crime reduction impacts.