+1(978)310-4246 credencewriters@gmail.com
  

Instructions

My learning objective for the first assignment is three-fold; specifically, this assignment should help students develop skills associated with:

1) constructing a literature review table,

2) writing critical paragraphs in a synthesized fashion so that the commonalities of the extant research (including limitations and strengths) are plainly identified, and

3) designing a simple experiment.

The topic of concern is gaming; more specifically, the effects that playing computer/video games (independent construct) have on an individual’s level of hostility (dependent construct).

Detailed instructions:

assignment_1_experiments.pdf

Example literature synthesis paragraphs

Bichler_Norris_Dmello_Randle_2019_highlighted.pdf

Sosa Bichler Quintero_2019_highlighted.pdf

Required Reading

experiments_video_hostility_original_research_REQUIRED READING.pdf

state_hostility_index.pdf

Selected Reading

experiments_video_hostility_article1.pdf

experiments_video_hostility_article2.pdf

experiments_video_hostility_article3.pdf

experiments_video_hostility_article4.pdf

CJUS 6604 – Assignment 1
Developing an Experiment
Instructions: My learning objective for the first assignment is three-fold; specifically, this
assignment should help students develop skills associated with:
1) constructing a literature review table,
2) writing critical paragraphs in a synthesized fashion so that the commonalities of the
extant research (including limitations and strengths) are plainly identified, and
3) designing a simple experiment.
The topical focus is gaming; more specifically, the effects that playing computer games
(independent construct) have on an individual’s level of hostility (dependent construct).
Critically Reviewing the Literature
Step 1. Start by skim reading the four articles. Then, generate a small table that summarizes
the research. Your table should include details about the methods used (e.g., data source,
method of data collection, independent variables, dependent variables), analytic tool used to
test hypotheses, and key findings. Two examples appear at the end of these instructions.
Also, you may find more information in Lecture 1-Experiments and Quasi-Experiments for
examples of tables. Once your table is created (in whatever software you like) add it to a word
document so that you can write about it. Keep in mind that you may need to remove some
columns or reformat the table to fit it into word. You will have at least four rows (one for each
article)—the required reading and three articles you selected as a set from the optional
readings.
Step 2. After the table is complete, write a few paragraphs describing the studies as a group
(remember you need to synthesize), pointing out methodological consistencies and limitations,
as well as the key findings. Remember, synthesizing means that you should try to integrate
information from multiple studies rather than talk about each study in turn. Be sure to assess the
validity of this body of work. You may want to provide specific examples drawn from one of the
studies. Tables are normally used to organize literature and to write literature reviews, but they
often do not appear in published work. Learning this skill is critical to making your writing
process more efficient and improving the rigor of literature reviews.
Step 3. Now that you have reviewed some literature, you are ready to propose a study.
Remember that to be a good scientist, you need to improve upon prior work (appearing in the
table and discussed already in your response). Feel free to borrow elements from prior
elements that worked well. Be sure to cite the studies. By incorporating, or “borrowing”, good
elements you are ensuring that your work is embedded within prior work and it constitutes
incremental development of knowledge (the goal of all scientific endeavors). Provide a brief
description of the method you would use to test the effect of gaming on aggressive behavior.
You will need to specify:
1) a research question and an associated hypothesis (or two)
2) sampling strategy, including size, source, and group assignment,
3) how the gaming activity will occur,
4) how you will capture information about hostility and what you will use to accomplish
this task (you might want to use or modify the state hostility index listed as one of the
readings), and
You may opt to draw a diagram to help explain your study (see examples from Lecture 1). Be
sure to point out what you are borrowing (using citations & short statements) and what you are
improving upon from prior work (again, using citations & short statements). Your references and
citations should be typed in APA format and organized with headings/subheadings as you see
fit. There is no page requirement or limit, but if your answer is less than 2 pages you need to
think about this more carefully, and if your assignment is more than 10 pages, a good edit is
needed. You can use a font size no smaller than 11-point, line spacing no tighter than 1.5, and
margins no smaller than 1 inch. To keep your answers organized use headings. This
assignment is due on Sept. 6th at 11:59 pm (PST). EMAIL the assignment directly to me
(as an attached file in Word or PDF format).
What follows are two examples of literature review tables. You will notice that the first column
always reports the citation of the study.
Example 1. Source: Bichler, G., Norris, A., Dmello, J.R., & Randle, J. (2019). The Impact of
Civil Gang Injunctions on Networked Violence between the Bloods and the Crips. Crime and
Delinquency, 65(7): 875-915. DOI: 10.1177/0011128717739607 [online first: November 24,
2017]
Table 1. Prior Studies Evaluating the Effects of Civil Gang Injunctions
Authors
Focus of
Study
Location
Data
Method
Main Results
Followup
Blythe Street
Injunction
(1993)
Los Angeles,
CA
police data
(1991-1996)
14 injunctions
Los Angeles
County, CA
police data-part 1 crimes
(1993-1998)
increased violent
crime and drug
trafficking in
adjoined areas
5-10% decline in
violent crime
(mostly a reduction
in assaults) and no
evidence of
displacement to
adjoining areas
n/a
Grogger
(2002)
19 reporting districts
– compared target
CGI area to
surrounding areas
areas with
injunctions
compared to
matched
comparisons
LA Grand Jury
(2004)
14 injunctions
Los Angeles,
CA
areas with
injunctions
6-9% decline in
serious crime
1 year
Goulka et al. (2009)
Santa Nita
Injunction
(2006)
Santa Ana, CA
police data-part I crimes
(2003-2004)
calls for service
– crime and
disorder (20052007)
6 enjoined block
groups compared to
166 other blocks;
pre- and postinjunction and trend
analysis
sig. increase in
violent (20%) and
weapons crime
(27%) and decrease
in property crime (17%)
18 months
O’Deane &
Morreale (2011)
25 injunctions
Ventura, Los
Angeles, San
Diego & San
Bernardino
Counties
police data– part
I and II crimes
(1982-2007)
Matched pairs; 25
gang injunctions,
one year pre- and
post- injunction
part 1 crimes
significantly
decrease in first year
post-injunction;
smaller decreases
for part II1
n/a
Maxson et al.
(2005)
Verdugo Flats
Injunction
(2002)
San
Bernardino,
CA
resident surveys
4 areas (2 target & 2
controls) –
compared resident
perceptions 18
months before and 6
months postinjunction
immediate change in
primary target area:
more police
visibility, fewer
gang members
hanging around, less
intimidation and fear
6 months
Hennigan & Sloane
(2013)
3 injunctions
Los Angeles,
CA
interviews with
gang aged youth
(14-21 yrs old)
4 areas (3 target & 1
control)
mixed results: CGIs
reduce street time
but have little effect
on group cohesion,
and injunctions did
not deter crime
2 years
purposive, snowball
sampling process
found harms
including: disruption
of family
relationships and
friendships;
blockage of
opportunities;
increase in feelings
of futility and
injustice
n/a
(Year of
Publication)
Police Data
ACLU
(1997)
1 year
Surveys (primarily)
police data
(2004-2009)
Swan & Bates
(2016)
Gang-affiliated
individuals in
San Diego
County
San Diego
County, CA
22 in-depth
interviews with
people listed on
a CGI or gang
list, or policed
for being
considered gang
affiliated
Calculated percentage change as: [(pre injunction – post injunction) / pre injunction] x 100 = percent change from
baseline].
1
Example 2. Source: Sosa, V., Bichler, G. & Quintero, L., (2019). Yelping about a Good Time: Casino Popularity and Crime. Criminal Justice
Studies, 39(2): 140-164. https://doi.org/10.1080/1478601X.2019.1600820
Table 2. Selected Studies Investigating Crime at Casinos
CITATION
LOCATION
FOCUS
CRIME
TIME
METHOD
PURPOSE/ISSUE
FINDING
Albanese
(1985)
New Jersey
Atlantic City
Part I crimes
1978-1982
Correlational
study
Casinos in relation to
crime
Hakim &
New Jersey: Atlantic, 64 Communities
Buck, (1989)1 Cape May & Ocean
Counties
Chang (1996) Mississippi
Biloxi
Part I crimes
1972-1984
Pre/post
comparison
Part I crimes, other
crimes & mischief
1986-1994
Pre/post
comparison
Casinos in relation to
crime and population
at risk
Measure the impact of
casinos on local crime
Grinols,
Mustard &
Dilley (2000)
United States
3,165 counties
Part I crimes
(excluding arson)
1977-1996
Correlational
study; posttest only
Determine the
relationship between
casinos and crime
Casinos have no effect on the
serious crime; crime rose due to
other factors
Crime was higher in post casino
years and a spillover effect occurred
in surrounding areas
Decrease in overall crime rates
during the first full year of casinos
opening. Crime rates returned to the
pre-casino level during the second
year
Casinos linked to higher crime
except murder; impact increased
beginning about 3 years after the
casino opened
Gazel,
Rickman &
Thompson
(2001)
Wisconsin
all counties
Part I & other
crimes
1981-1994
Pre/post
comparison
Stitt, Nichols,
& Giacopassi
(2003)
Iowa, Missouri,
8 cities
Illinois, & Mississippi
Part I & II crimes
Pre/post
comparison
Moufakkir
(2005)
Michigan
City of Detroit & 3
counties
Part I, other crimes,
& disorderly
conduct
1987-1998
Min. 4 yrs
pre & post
casino
opening
1996-2002
Examine the
relationship between
Native American
casinos and crime
levels
Effects of new casinos
on crime and the
quality of life in the
area
Pre/during/
post
comparison
Examine crime
volume in Detroit and
its neighboring
communities
Grinols
& Mustard
(2006)
Barthe & Stitt
(2007)1
All U.S. counties
3,165 U.S.
Counties
Part I crimes
1977-1996
Pre/post
comparison
Reno, Nevada
15 casinos in
downtown Reno
Part I crimes &
other crimes
All 2003
Correlational
study
Examines how the
opening of new
casinos affect crime
Determine if casinos
and their surrounding
blocks are hot spots
that generate crime
Counties with casinos showed an
increase in crime rates; there was
also a spillover effect with counties
adjacent to casino-counties (higher
crime rates)
Some communities experienced an
increase in crime, others a
reduction, and some remained the
same: overall, crime rates do not
increase
Volume of crime did not materially
increase when the 3 casinos
opened
The effect on crime is low shortly
after a casino opens, then increases
over time
Almost 25% of Reno’s crime,
occurred within 1000 feet of the
major casinos but factoring for
population at risk, casinos do not
appear to be “hot spots” that
generate crime
1
Wheeler,
Round,
Sarre, O’Neil
(2008)
South Australia
111 local areas
Income & nonincome-generating
crimes2
2002-2003
Correlational
study
Compare electronic
gaming machine
(EGM) expenditures
to crime rates
Higher EGMs expenditures were
significantly related to incomegenerating crime rates but not nonincome-generating crime rates
Barthe & Stitt
(2009)1
Reno, Nevada
Reno, Nevada
Violent crimes,
property crimes,
and disorder crimes
Not reported. Correlational
study
Compare casino and
non-casino zones
Temporal trends in casino zones
are not very different than those
found in non-casino areas
Belanger,
Alberta, Canada
Dene & Eagle
Williams, &
River (First
Arthur
Nations
(2012)1
communities)
Pontell, Fang, Macau & Hong Kong, Macau, China
& Geis (2013) China
Casino-related
crime
2006-2010
Pre/post
comparison
Assess crime &
socioeconomic effects
on rural communities
White-collar &
economic crimes
2009-2011
Data analysis
Falls &
Thompson
(2014)1
Robbery, burglary,
larceny & motor
vehicle theft
1994-2010
Correlational
study
Examine economic
and white-collar
criminal activities
Impact of casinos on
crime rates in the host
and neighboring
counties
No reported increase in criminal
activity. Residents are aware of the
financial and social realities of
casino operations
The growing casino industry has
created a receptive environment for
various forms of corruption to thrive
The presence or size of a casino
does not increase property crime
rates in the host county or in the
nearby counties
Michigan
83 counties
Spatial aspect to the analysis of casino locations and crime.
Income-generating crimes: robbery and extortion, burglary, break and enter, fraud, forgery, false pretenses, and larceny. Non-income-generating
crimes: offences against the person, damage (property and environmental), offences against good order, driving, and other offences.
2
End of Assignment Instructions
739607
research-article2017
CADXXX10.1177/0011128717739607Crime & DelinquencyBichler et al.
Article
The Impact of Civil
Gang Injunctions on
Networked Violence
Between the Bloods
and the Crips
Crime & Delinquency
2019, Vol. 65(7) 875­–915
© The Author(s) 2017
Article reuse guidelines:
sagepub.com/journals-permissions
https://doi.org/10.1177/0011128717739607
DOI: 10.1177/0011128717739607
journals.sagepub.com/home/cad
Gisela Bichler1, Alexis Norris1, Jared R. Dmello2,
and Jasmin Randle1
Abstract
Comparing the centrality of gangs and changing structure in attack behavior,
this study examines the effects of civil gang injunctions (CGIs) on violence
involving 23 gangs (seven Bloods and 16 Crips) operating in Southern
California. We mapped violence networks by linking defendants and victims
named in 272 court cases prosecuted in the City of Los Angeles (19972015), involving at least one conviction for a violent crime and a defendant
tried as an adult. The results show that a small number of gangs are centrally
located in a dynamic web of non-reciprocated conflict that exhibited
complex hierarchical structures. These results raise four implications for
combating gang violence.
Keywords
social network analysis, civil gang injunctions, gangs, violence, Bloods and
Crips
1California State University, San Bernardino, CA, USA
2University of Massachusetts, Lowell, MA, USA
Corresponding Author:
Gisela Bichler, California State University, San Bernardino, 5500 University Parkway,
San Bernardino, CA 92407, USA.
Email: gbichler@csusb.edu
876
Crime & Delinquency 65(7)
Introduction
Civil gang injunctions (CGIs) impose significant behavioral restrictions on
individuals, that is, setting curfews, prohibiting free movement, and restricting social activity. Yet, there are few evaluations of the effect that CGIs have
on sanctioned groups and the community they aim to protect. Notably, where
CGIs are evaluated, there is a tendency to examine community-level impacts,
rather than how the injunctions influence individual-level behavior—with the
noted exceptions of Swan and Bates (2016) and Hennigan and Sloane (2013).
This trend is problematic because emerging from recent street gang research
is convincing evidence of the interconnected nature of violence and the
importance of social connections in accounting for patterns of conflict.
Studies using social network analysis (SNA) show that the risk for involvement in violence spreads in a contagious fashion (e.g., Green, Horel, &
Papachristos, 2017; Papachristos, Hureau, & Braga, 2013; Papachristos,
Wildeman, & Roberto, 2015).1 When a gang member becomes embroiled in
a dispute, causing injury or perceived harm to reputation or status, the individual (or group) will react in some fashion. It follows that any perceived
attack or harm will have repercussions beyond the targeted individuals as
ripple effects extend across a social neighborhood. The imposition of a CGI
is, without doubt, a clear public admonition of a group’s behavior. As such, it
should trigger a shift in violent behavior, in either the selection of targets,
direction of attack, or frequency of victimization.
By understanding how violence spreads through social networks, there is
a better chance of developing focused-deterrence strategies that minimize
displaced aggression, reduce gang conflict, and ultimately, improve public
safety. The social structure of gang violence has been investigated within a
single gang (e.g., McCuish, Bouchard, & Corrado, 2015), within an identifiable neighborhood or region (e.g., Tita & Radil, 2011; Randle & Bichler,
2017), and across cities, that is, Boston (Papachristos et al., 2013), Chicago
(Papachristos, 2009), Montreal (Descormiers & Morselli, 2011), and Newark
(McGloin, 2005). To the best of our knowledge, this study is the first to investigate the effects of CGIs on the structure of group-on-group conflict. The
present study investigates the structure of inter-gang violence involving 23
gangs based in the City of Los Angeles, California, as revealed by prosecutions over a 19-year study period (January 1, 1997, to December 31, 2015).
Our primary aim is to assess the effects of CGIs by exposing shifts in the web
of violence at the local level, comparing pre- and post-injunction violence
networks for specific gangs, as well as the aggregate effect across the study
period. In doing so, this study investigates how the imposition of CGIs alters
the tendency of gangs to direct serious violence at non-gang-involved
Bichler et al.
877
individuals and engage in new conflict, attacking gangs they did not victimize in the period before the injunction.
To provide context for this study, we provide a brief discussion of focused
deterrence, describe CGIs enacted in Los Angeles, review prior efforts to
evaluate the impact of CGIs, and report on the historic rivalry between the
Bloods and the Crips. Next, we outline current thinking about how gang violence transmits through social networks. After introducing the research aims,
we detail the research method, describe the networks, and present our results.
Two sets of results are reported: First, we identify the groups involved in the
most violence—attacking behavior pre- and post-injunction, as well as victimization for both observation periods—and then, we conduct a triadic census to investigate the nature of structural change between observation periods.
The article concludes with a discussion of the implications that a social network approach has for combating gang violence with CGIs.
Study Context
Focused Deterrence
Problem-oriented policing literature suggests that law-enforcement interventions would be most effective when focused on the people, places, and
the context driving the problem (Braga et al., 1999; Sherman & Weisburd,
1995). Given this, focused-deterrent strategies aim to direct attention
toward individuals driving the problems in targeted areas. Originated to
reduce gang and group involved violence in Boston (e.g., Braga, Pierce,
McDevitt, Bond, & Cronin, 2008; Braga et al., 1999; Kennedy, 1997),
focused-deterrent strategies have been used to combat gun violence
(Corsaro & McGarrell, 2009; Papachristos, Meares, & Fagans, 2007) and
crime associated with drug markets (Corsaro, Brunson, & McGarrell, 2010;
Hipple, Corsaro, & McGarrell, 2010).
As a problem-oriented approach to resolving crime problems, focuseddeterrence tactics use an iterative strategy to identify and analyze crime problems, address those problems, and assess and adjust the response. Initiatives
typically involve (a) creating a multi-agency working group of law-enforcement agencies; (b) identifying offenders, groups, and behavior patterns driving the crime problem; (c) developing a clear deterrent message to offenders
and groups of offenders that employ a wide range of sanctions (pulling levers)
to persuade them to stop their behavior; (d) focusing social services and community resources on targeted offenders and groups to complement lawenforcement efforts; and (e) directly, clearly, and repeatedly communicating
to offenders why they are receiving this special attention (Braga et al., 2008;
878
Crime & Delinquency 65(7)
Kennedy, 1997). Multi-agency working groups coordinate efforts to create
consequences for targeted individuals: By giving a direct, explicit, and consistent message threatening clear sanctions, focused deterrence is meant to
deter and control serious violence in communities by increasing the perceived
cost of engaging in those behaviors. Designed to disrupt behavior that facilitates gang conflict and community intimidation, CGIs constitute a focuseddeterrent strategy aimed at reducing gang-involved violence.
CGIs
First used in the City of Los Angeles in 1987, CGIs establish “safety zones”
wherein members of the target gang(s) are prohibited from engaging in a
predetermined list of activities, ranging from illegal behavior (e.g., drug selling) to social nuisances (e.g., hanging out in public while flashing gang signs
to intimidate area residents) and general restrictions (e.g., curfews). The
focus is on behavior that occurs in public. Similar to other nuisance abatement litigation, violators face civil sanctions, such as financial penalties, for
failing to abide by the behavioral orders.
The legal frameworks used in California to establish that gang behavior
constitutes a public nuisance are Sections 3479 and 3480 of the California
Civil Code. To impose a sanction against named individuals or a group of
people, the behavior in question must affect a definable group (e.g., community or neighborhood) and be considered:
. . . injurious to health or is indecent or offensive . . . , or an obstruction of the
free use of property, so as to interfere with the comfortable enjoyment of life or
property, or unlawfully obstructs the free passage or use, in the customary
manner, of any navigable lake, river, bay, stream, canal, or basin, or any public
park, square, street, or highway. (Cal. Civ. Code, § 3479)
Several restrictions are common to all CGIs enacted in Los Angeles. Of
central importance is the restriction against driving, standing, sitting, walking, gathering, or appearing anywhere in public view or any place accessible
to the public, with any known member of the enjoined gang.2 In addition,
most injunctions prohibit gang-involved individuals from engaging in acts of
intimidation including threatening behavior or obstructing someone’s free
passage or movement, possessing graffiti or vandalism tools, making graffiti
or damaging property, or violating municipal curfews. Moreover, these orders
routinely prohibit enjoined gang members from possessing any firearm,
ammunition, or illegal weapon, or being around someone in possession of
such weapons. Prohibitions also restrict gang members from selling,
Bichler et al.
879
possessing, or using any controlled substance or related paraphernalia without a prescription.
The courts can also impose gang specific restrictions within the safety
zones. These prohibitions are usually associated with the activities of the
group in question. Examples include,
••
••
••
••
••
No reckless driving or obstructing traffic;
No lookouts or loitering;
No trespass, that is, Gilbert Lindsay Park or Alameda Swap Meet;
No identity theft; and
No recruiting children.
Given the information needed to establish that individuals or a group
imposes a significant nuisance to a community, prosecutors work closely
with law-enforcement agencies integrating gang intelligence with crime
reports, testimonials, and evidence from prior criminal cases. This information identifies specific individuals or groups to sanction, and defines the
scope of the injunction, in regard to establishing the physical parameters of
safety zones (mapping the area) and developing a list of prohibited behaviors.
As law enforcement is responsible for enforcing the terms of the injunction,
it is critical that representatives of the agency are involved in setting the
behavioral expectations that they will have to enforce.
Currently, there are 46 injunctions targeting about 72 gangs in the City of
Los Angeles. Admittedly, several injunctions name cliques as well as individual gangs, making it difficult to assess exactly how many different gangs
(as opposed to cliques) are involved. For instance, 18th Street gang is
enjoined, as are three cliques: 18th Street (Hollywood), 18th Street (Pico
Union), and 18th Street (Wilshire). An additional complication is that subgroups and cliques do not always follow a naming convention that indicates
the gang association, blurring the line between cliques and gangs. Moreover,
gangs are known by several names (e.g., Barrio Van Nuys, Van Nuys, and
BVN) and similar names with different spellings (e.g., 83 Street Hoovers also
are referred to as 8-Trey Hoovers and Eight Tray Crips). Notably, half of the
injunctions target gangs generally and half of the CGIs name specific individuals (807 people) and their gangs. The Superior Court of California
imposed most of these injunctions before 2010 (see Table 1).
The restrictions imposed by CGIs are indefinite. For this reason, about
35% of injunctions include a clause permitting individuals, demonstrating no
gang activity for a period of 3 years (no felonies, no associating with gang
members, and no new gang tattoos) and evidence of gainful employment or
school attendance for at least 18 months, to petition to have the injunction
880
Crime & Delinquency 65(7)
Table 1. Enactment and Characteristics of Los Angeles CGIs.
Characteristic
Injunctions (N = 46)
Year enacted
2000-2005
2006-2010
2011-2015
Example characteristics
Prohibited for hanging out near schools
No alcohol
Curfew
Renunciation or opt-out
%
58.7
34.8
6.5
13.0
87.0
73.9
34.8
Note. CGI = civil gang injunction.
against them lifted. In addition, recent California legislation established statutory provisions for individuals to contest their inclusion in gang databases
(AB-2298 Criminal Gangs).3 This legislation expands the possibility for
removing names from all gang databases, extending beyond just those under
a current injunction. As the process for removing oneself from a gang injunction is very complex, posing undue legal challenges (Crawford, 2009), the
impact of new legislation on the ability of individuals to remove themselves
from the lens of injunction enforcement remains unclear.
Effectiveness of CGIs
Prior studies investigating the effectiveness of CGIs typically use police data
(crime reports or calls for service), comparing areas with injunctions with
matched control areas (e.g., Grogger, 2002) or surrounding areas without
injunctions (e.g., American Civil Liberties Union [ACLU] of Southern
California, 1997; Goulka et al., 2009). Results of these past studies have been
mixed. Some studies reported in Table 2 found that injunctions were associated with a reduction in violent or serious crime (Grogger, 2002; Los Angeles
County Civil Grand Jury, 2004). For example, Grogger (2002) found that
compared with matched neighborhoods without injunctions, neighborhoods
with injunctions experienced a 5% to 10% decline in violent crime during the
year following enactment with no evidence of displacement to adjoining
areas. Other studies found an increase in violent crimes in enjoined areas
(ACLU, 1997; Goulka et al., 2009). While prior research finds mixed results,
881
14 injunctions
Blythe Street
Injunction
(1993)
Focus of study
O’Deane and Morreale 25 injunctions
(2011)
Los Angeles County 14 injunctions
Civil Grand Jury
(2004)
Goulka et al. (2009) Santa Nita
Injunction
(2006)
Grogger (2002)
Police data
ACLU (1997)
Authors (year of
publication)
Table 2. Prior Studies of CGIs.
Data
Police data—Part I crimes
(1993-1998)
Calls for service—Crime
and disorder (20052007)
Ventura, Los Angeles, Police data—Part I and II
crimes (1982-2007)
San Diego and San
Bernardino Counties
Santa Ana, California
Los Angeles, California Police data—Part I crimes
(2003-2004)
Los Angeles County,
California
Los Angeles, California Police data (1991-1996)
Location
Main results
Follow-up
Significant increase in violent
6 enjoined block groups
(20%) and weapons crime
compared with 166
(27%) and decrease in
other blocks; pre- and
property crime (−17%)
post-injunction and
trend analysis
Part I crimes significantly
Matched pairs; 25 gang
decrease in first year postinjunctions, one year
injunction; smaller decreases
pre- and post-injunction
for Part IIa
(continued)
NA
18 months
NA
Increased violent crime and
19 reporting districts—
drug trafficking in adjoined
Compared target CGI
areas
area with surrounding
areas
1 year
5%-10% decline in violent
Areas with injunctions
crime (mostly a reduction
compared with matched
in assaults) and no evidence
comparisons
of displacement to adjoining
areas
Areas with injunctions
6%-9% decline in serious crime 1 year
Method
882
Focus of study
San Bernardino,
California
Location
Resident surveys
Data
Method
Main results
Follow-up
aCalculated percentage change as [(pre-injunction – post-injunction) / pre-injunction] × 100 = percent change from baseline].
Note. CGI = civil gang injunction; ALCU = American Civil Liberties Union of Southern California.
6 months
Immediate change in primary
4 areas (2 target and 2
target area: more police
controls)—Compared
visibility, fewer gang
resident perceptions 18
members hanging around,
months before and 6
less intimidation and fear
months post-injunction
2 years
Mixed results: CGIs reduce
Hennigan and Sloane 3 injunctions
Los Angeles, California Interviews with gang-aged 4 areas (3 target and 1
control)
street time but have little
(2013)
youth (14-21 years old)
effect on group cohesion,
Police data (2004-2009)
and injunctions did not deter
crime
NA
Purposive, snowball
Found harms including
22 in-depth interviews
Swan and Bates (2016) Gang-affiliated San Diego County,
sampling process
disruption of family
California
with people listed on
individuals
relationships and friendships;
a CGI or gang list,
in San Diego
blockage of opportunities;
or policed for being
County
increase in feelings of futility
considered gang affiliated
and injustice
Surveys (primarily)
Maxson et al. (2005) Verdugo Flats
Injunction
(2002)
Authors (year of
publication)
Table 2. (continued)
Bichler et al.
883
a consensus exists that the effects are short lived (e.g., Maxson, Hennigan, &
Sloane, 2005; O’Deane & Morreale, 2011).
Of interest to the current study are the findings of Hennigan and Sloane
(2013) and Swan and Bates (2016). Using interviews with gang-aged males
(14-21 years old), Hennigan and Sloane (2013) examined gang-involved
youths’ perceived risk of being caught and punished for criminal activity, the
effect of gang injunctions on gang cohesion, and gang identification. Even
though no differences emerged between study groups regarding the expectation of being caught and punished for criminal and violent activity, Hennigan
and Sloane (2013) found that gang-involved youth in CGI areas were less
likely to identify with their gang (weaker social identity) and reported spending less street time together (lower street cohesion) than gang-involved youth
in the control area with no injunctions.
Swan and Bates (2016) examined the perceptions of gang suppression
measures by interviewing those listed on CGIs in San Diego County, focusing on the hidden harms these measures have on enjoined individuals. First,
they found that respondents felt that gang suppression measures reduced their
opportunities to find housing, extend their education, and find jobs—prospects that would enhance successful integration into the community. Second,
injunctions blocked them from pursing conventional relationships for many
years: Individuals were unable to associate with family members in enforcement areas, and constant police harassment and heavy surveillance, prohibited socializing with pro-social friends who were not gang involved.4 Third,
respondents said gang activities did not stop with the impositions of CGIs;
rather, gang activities shifted to neighborhoods without gangs or to rival gang
territory, which exacerbated existing conflicts—a finding consistent with
emerging research on networked gang violence.
Networked Gang Violence
In a series of articles, Papachristos and colleagues (e.g., Green et al., 2017;
Papachristos, 2009, 2013; Papachristos et al., 2013) put forth a convincing
argument about the contagious nature of violence and the need for deterrent
strategies to consider the networked nature of gang violence. Violence, like
most phenomena (i.e., information, disease, and emotions), transmit through
a social network, moving from person-to-person like an infection. Described
as “hyperdyadic contagion” by Christakis and Fowler (2009), individuals
respond to what they learn or experience, and in turn, this reaction facilitates
additional ripple effects back toward the origin and forward, toward new
people. Influence typically extends three steps from the origin, that is, to the
friend-of-a-friend-of-a-friend. The structure of the local social neighborhood
884
Crime & Delinquency 65(7)
is critical to understanding how contagions progress: The local social neighborhood surrounding each person influences how information reaches them
and whom their reactions affect. Social neighborhoods include overlapping
family, friend, and acquaintance networks. Each type of relation (e.g., kinship
vs. friendship) forms a unique network, and because people can have multiple relations with the same person, such as working with a cousin, multiple
ties could exist linking pairs of people. This multiplexity serves to create
overlapping networks through which social phenomena—that is, ideas, perceptions, and risk—could be transmitted.
Papachristos (2013) argued that because of this network effect, deterrent
strategies must consider the social structure within which individual gang
members are embedded, as well as the group’s cohesion, and linkage to other
groups. This argument is consistent with recent scholarship that shows that
the behavior of gang members is constrained (and enabled) by the social
network within which the individual and their group are enmeshed (e.g.,
Descormiers & Morselli, 2011; McCuish et al., 2015; Papachristos, 2009;
Papachristos et al., 2013; Papachristos et al., 2015).
Gang members do not kill because they are poor, black, or young or live in a
socially disadvantaged neighborhood. They kill because they live in a structured
set of social relations in which violence works its way through a series of
connected individuals. (Papachristos, 2009, p. 75)
Social networks place “adversaries in positions where each must attempt to
defend, maintain, or repair their reputation” (Papachristos, 2009, p. 76).
Building from a growing body of work examining the structure of gang violence (see, for example, Descormiers & Morselli, 2011; McCuish et al.,
2015), Papachristos (2009) argued that patterns of networked violence
emerge from the aggregation of individual-level disputes associated with historic rivalries, retaliation for preceived harms, and efforts to retain or elevate
social status. As factions jostle for social position, conflict can emerge
between groups that are otherwise thought to be in allegiance as perceived
disturbances trigger new waves of violence (Decker & Curry, 2002;
Descormiers & Morselli, 2011). Simply stated, violence spreads through
social networks as individuals react to the behavior of others (Papachristos,
2009). Moreover, social connectivity may better account for conflict patterns
than spatial proximity or contiguity: Although related, Tita and Radil (2011)
found that the importance of social connections, or linkages, “extends well
beyond simple spatial contiguity,” increasing the salience of network-oriented research on gang communities and conflict patterns (p. 538). It is
expected that although gang violence is likely to involve a spatial aspect (i.e.,
Bichler et al.
885
conflicts spanning city borders to encompass new groups), the affiliations
and relationships of individuals within the gang communities will have a
stronger impact on how the characteristics of violence evolves over time.
Previous studies found evidence of three types of reaction, each generating identifiable social network structures. Perceived harm can trigger an act
of retaliatory violence as groups struggle for dominance. In Chicago, for
example, reciprocal attacks accounted for 37% of gang-related homicides
(Papachristos, 2009), and in a subsequent study, reciprocity was found to be
a significant predictor of fatal and nonfatal gunshot injuries in Boston and
Chicago (Papachristos et al., 2013). Violence could also spread to those not
directly involved. For instance, there may be a knock-on or domino effect,
where the victimized group, responds by attacking another, less dominant
group—a social pecking order emerges (Papachristos, 2009; Randle &
Bichler, 2017). In addition, groups may exhibit high levels of aggression by
attacking many other groups or a group may come to be repeatedly victimized (Descormiers & Morselli, 2011; Papachristos et al., 2013).
Current Research Focus and Hypotheses
The focus of this study is on the Bloods and Crips, two of the most infamous,
predominantly African American gang coalitions rooted in Los Angeles with
a historic reputation for networked hostility. Raymond Washington and
“Tookie” formed the Crips in the 1960s to protect themselves from criminal
activity in their neighborhood. Initially confined to high schools within the
Los Angeles area, in a little over a decade, membership grew and subsets
emerged in cities throughout California. Some of this growth can be attributed to established gangs adopting the Crips name (e.g., Main Street Crips,
Rollin 20 Crips), despite being independent gangs with leadership of their
own. In the 1970s, Sylvester Scott and Vincent Owens formed the Compton
Prius, in response to the growing number of Crips subsets. After a conflict
with the Compton Crips in which the Compton Prius were badly beaten, several sets of the Prius gang and a number of other gangs in the area, that were
threated or attacked by the Crips, joined together to form the Bloods. In other
words, the Bloods formed as clusters of networked individuals banded
together in a defensive reaction to being attacked by Crips gang members.
In 2005, the last year the LAPD (Los Angeles Police Department) updated
their public statistics on the population of specific gangs, the Crips had a
reported 113 gangs with 10,792 members while the Bloods had a reported 45
gangs with 4,415 members.5 The number of identifiable groups grows to 265
if cliques are considered. Cliques are acknowledged subgroups of gangs that
have established identities and turf. Generally, there are more Crips gangs
886
Crime & Delinquency 65(7)
than Bloods: The ratio is about 3 to 1 (Descormiers & Morselli, 2011, citing
Delaney, 2006; Landre, Miller, & Porter, 1997). Notably, of the 72 enjoined
gangs based in Los Angeles, 23 have been, or are currently, affiliated with the
Bloods (30%) or the Crips (70%). For a list of the 23 gangs, please see the
appendix.
Scholars suggest that the combined effects of evolution, expansion, and
clique formation is likely to increase inter-group violence due to greater competition over resources and reputation (e.g., Howell, 2012, p. 18), lead to the
formation of alliances that foster positive associations among gangs (e.g.,
Descormiers & Morselli, 2011), and facilitate an observable increase in intraconsortium fighting (e.g., Decker & Curry, 2002). Despite the historic and
widely acknowledged rivalry among the Bloods and the Crips, only two prior
studies focus on mapping the structure of conflict among these groups.
Through interviews with 20 youth gang members, Descormiers and Morselli
(2011) mapped inter-coalition conflicts and intracoalition alliances among
Bloods and Crips active in Montreal, Canada. In addition to finding network
structures consistent with a Bloods and Crips rivalry (76.7% of conflicts),
these authors also identified exceptions—9.4% of the conflicts involved
intracoalitional issues. For instance, a Crips gang would attack another Cripsaffiliated group if they were not adequately representing Crips. Using 284
court cases to map the structure of serious violence occurring in Los Angeles
(January 1, 2002, to December 31, 2010), Randle and Bichler (2017) discovered that of the 147 different Bloods and Crips gangs involved, most gangon-gang violence (62%) involved intracoalition conflict, wherein Crips
attacked Crips and Bloods attacked Bloods. These authors found that only
about 38% of 205 violent victimizations reflected the classic Bloods and
Crips rivalry. Moreover, much of the gang-on-gang violence (42%) involved
a directed line, where one group victimized another, who victimized a third
group. The remaining victimization patterns (58%) exhibited two-star formations, wherein a group attacked or was attacked by multiple groups.
Drawing upon the literature reviewed above in concert with prior research
documenting the contagious nature of gang violence, our general working
hypothesis is that, on average, substantive structural change will occur among
enjoined gangs between the pre- and post-injunction periods. From this general framework, we examine three predictions about the structure of gang
violence.
1.
As enjoined groups react to the injunction and adjust routine behaviors, possibly venturing into the territories of other groups, serious
violence post-injunction will involve conflict with new groups or
exhibit an increase in violence against non-gang individuals. One
Bichler et al.
2.
3.
887
consequence would be an increase in two-star structures wherein, a
group attacks or is victimized by several other groups.
Local hierarchies will intensify post-injunction. Enjoined gangs will
become more aggressive toward others as groups vie for dominance
over their counterparts, observed as an increase in the knock-on or
domino pattern, suggesting a shift in the inter-gang social pecking
order. The structural manifestation would be an increase in directed
lines, wherein, Gang A attacks Gang B, which in turn, attacks Gang C.
Retaliation (reciprocated attacks) will increase. CGIs will tarnish the
street reputation of the enjoined gang, and other groups previously
dominated by the gang, will be emboldened, leading to greater levels
of retaliatory violence.
Alternatively, CGIs will have no effect. Gang violence will continue
unabated, with no apparent structural change between the pre- and postinjunction period. This means that, on average, there will be no substantive
change in network structure for the enjoined gangs.
Method
Identifying Incidents of Serious Gang Violence
Twenty-three enjoined gangs affiliated with Bloods and Crips constitute our
seeds. In social network terms, a seed is a starting point in the generation of
a social network. By using the gang’s name as a search term, we identified
cases wherein at least one defendant or victim was a known gang member at
the time of the incident. With California’s enhanced penalties for gang-related
crime and the extensive intelligence network designed to support the system,
information about gang affiliation is commonly reported in case details.
Limiting the scope of this investigation, we applied a set of eligibility criteria
to narrow our focus to serious violence, occurring within a specified timeframe, and in a predetermined study area. To be included in the study, the
case must involve
1.
2.
3.
4.
At least one charge/conviction for serious violence, that is, robbery,
assault with a deadly weapon, attempted homicide, or homicide;
At least one defendant tried as an adult;
At least one defendant or victim known to be a member of a seed gang
based in the City of Los Angeles at the time of the incident; and
A crime occurring between January 1, 1997, and December 31, 2015.
888
Crime & Delinquency 65(7)
To find these cases, we searched LexisNexis for the name of each group:
LexisNexis is an electronic library that provides information about federal
and state cases.
Our search process identified 284 criminal convictions that satisfied all
eligibility criteria. Of these cases, 272 were coded.6 In addition to recording
all named individuals—defendants, accomplices, victims, witnesses, and the
judge—we captured information about their gang or group affiliation, as
well as other relevant demographic and case details, such as city of residence, gender, ethnicity, age, city where the attack took place, type of crime,
and number of people involved. Non-gang-affiliated individuals were coded
“non-gang community,” “LAPD” for officers from the Los Angeles Police
Department, or “LACSD” for deputies from the Los Angeles County
Sheriff’s Department: This designation was used primarily for the victims’
“group” status.7
Constructing Networks of Gang Violence
As illustrated in Figure 1, the process used to build the master, directed and
valued network involves multiple steps. In the figure, squares represent
individuals and circles indicate their gang affiliation. The lines connecting
individuals (or gangs) illustrate the direction of the attack as the arrowheads
point to the victimized party. Each step in the process is listed down the left
side of the image. Network generation begins at the dyad level, that is, with
pairs of individuals, in this case, each offender and victim pairing. This means
that the existence of multiple codefendants and victims results in many
offender–victim pairings. The 272 cases identified for this study resulted in
1,002 victimization dyads. A few cases did not report the date of the incident,
and thus, there were 984 usable victimization dyads. Seven people were victims and offenders (due to involvement in multiple cases).
Notably, 45% of the cases named a lone offender.8 Co-offending typically
involved one other person (32% of cases) or two other people (16%). Only
6% of cases involved three co-offenders and 2% involved four people. These
cases generated 375 co-offending activities, of which 87% involved members
of the same gang or individuals (friends or relatives) associated with the gang
and 13% involved collaborations among different gangs that appeared to be
within the same coalition (Bloods or Crips). Only one partnership occurred in
both study periods (Rollin 20s and Black P Stones). The 16 pre-CGI cooffending partnerships involved 19 gangs, whereas, among 20 gangs cooffending post-CGI, there were 32 instances of co-offending.
Most of the victimizations involved gun violence (93%) that occurred in a
public area—10% of victimizations occurred in open spaces like a park or
Bichler et al.
889
Figure 1. Process used to convert individual victimizations to a master network
of gang violence.
Note. Circles represent gang groups and squares depict individuals. Bold, underlined text
denotes seed gangs. Line thickness indicates the number of victimizations and arrowheads
indicate the direction of the attack.
field, 47% of incidents occurred in a public street or parking lot, and 28% happened in or just outside of a business, such as a convenience store or fast food
restaurant. About 15% of victimizations occurred just outside a residence.
While a number of different violent crimes were included in the case eligibility criteria, most incidents involved murder (62%) or attempted murder (15%).
Robberies that did not lead to death constituted 16% of the victimizations and
assault, carjacking, and rape comprised the remaining victimizations.
Aggregating individual-level victimizations to the gang-level, to reflect
group-on-group activity produces valued links between groups. For example,
in the second case illustrated in Figure 1, two members of Gang “G2” victimized a member of the Gang “G3” resulting in a thicker line, which would be
890
Crime & Delinquency 65(7)
valued with a score of two. Aggregation resulted in 114 groups, of which 111
were uniquely identified gangs. As noted previously, the three non-gang
groups were the community, the LAPD, and the LACSD. On average, each
gang (excluding non-gang groups) initiated 8.7 victimizations with scores
ranging from 0 to 112.
We mapped two sets of networks from the group-to-group links. (a) CGI
master networks aggregate victimizations during the pre-CGI observation
and the post-CGI observation periods. As our search protocol captures the
activity of non-enjoined gangs, the resulting violence network permits investigation of how each target gang (subject of a CGI) fits within a context, a
community-level social neighborhood if you will, of violent relations. (b)
Next, we construct egocentric networks to examine each of the seed gangs
(23 enjoined gangs). Each egocentric network includes the ego—the focal
gang—and all others, referred to as alters, with whom the ego is in conflict
with as an aggressor or a victim, as well as the conflicts among alters. Table
3 describes these networks: Egocentric networks for the groups involved in
the most violence (at least 10 attacks or victimizations) are reported. Network
mapping and analysis used UCINET 6 (Borgatti, Everett, & Freeman, 2002).
Analytic Strategy
At the gang-level, CGIs should materially alter the local web of violence
within which groups are embedded. To untangle this web, we compare the
social structure of violence pre-injunction and post-injunction for each gang.
As gangs are enjoined at different times, the number of attacking behaviors
and victimization are averaged by the number of years, pre- and post-CGI
within the study period. As described below, attacking behavior was captured
with outdegree centrality and victimization measured by indegree centrality.
These statistics are described in greater detail shortly. Using a case study
approach, we compared the local networks (ego networks) of the most violent gangs to see if there are similarities among groups, as well as change in
their level of violence.
We investigate patterns of relations among sets of three groups with a triad
census. A triad census is an inventory of the different structures observed in
the network. Repeating a triad census for egocentric networks pre- and postCGI reveals how the networks change. Some of the key structures of interest
are illustrated in Table 4 under the subheading, local hierarchy.
Complex patterns of violence exist if three groups are in conflict. For
instance, a group may be targeted by two gangs who also attack each other:
This transitive structure (i←j, i←h, and j↔ h) indicates that the focal gang,
designated by i, is embedded in a more complex pattern of violence than if it
891
Bichler et al.
Table 3. Description of Networks.
Network
Nodesa
All incidents
109
CGI master networks
Pre-CGI
68
Post-CGI
74
Egocentric networks
Bloods-affiliated gangs
   Black P Stones
9
  Bounty Hunter
14
Bloods
  Swan Bloods
6
Crips-affiliated gangs
  48th Street
2
Gangster Crips
  Geer Street
3
  Grape Street
14
Crips
  Main Street
5
  Rollin 40s
9
  Rollin 60s
7
  Venice Shoreline
7
Crips
Ties/
Size of main
victimization
component Density
Components (groups)
pairsa
(%)b
178/984
3
105
8.0
89/387
108/599
3
4
61
68
8.1
10.6
12/72
8/182


16.7
4.4
3/30


23.3
0/2


0
3/6
9/182


50.0
5.0
4/20
9/72
8/42
1/42


20.0
12.5
19.1
2.4
Note. CGI = civil gang injunction.
aFor egocentric networks, size counts the number of alters (groups connected to the ego),
ties reflect the number of relations among alters excluding the ego, and pairs counts the
number of victimizations (valued ties) among alters.
bWhen used with an egocentric network, density is calculated by dividing the number of ties
among alters by the number of pairs (among alters), and then, this figure is multiplied by 100
to generate a percent.
were simply attacked by two gangs. Referred to as transitive triads or transitivity, summing the number of these complex attack patterns provides a way
to identify the level of complex embedding. Our measure of complex embedding includes seven types of transitive structures where j and h represent
other gangs in conflict with the focal gang denoted by i: (1) i→j←h, i →h; (2)
i ←j←h, i →h; (3) i ←j→h, i ↔h; (4) i →j←h, i ↔h; (5) i →j→h, i ↔h; (6)
i →j↔h, i ↔h; and (7) i ↔j↔h, i ↔h. These structures are indicative of a
deeper level of submersion or entanglement with other gangs. Examining the
892
Crime & Delinquency 65(7)
Table 4. Indicators of Structural Change in Networked Violence.
Structural variable
Visual representation
Retaliation (reciprocity)
Local hierarchy
Chain-like knock-on
or domino effects
(directed or transitive ties)
Intensified activity
More attacks
(Outward two-star
formation)
More victimizations
(Inward two-star
formation)
Note. Dashed lines exhibit the nature of the change expected between pre- and postinjunction periods.
effects of CGIs with a range of structural metrics provides a comprehensive
framework through which to document the impact that these civil instruments have on prosecuted gang violence.
Results
Groups Most Involved in Violence
Attacking behavior. Attacking behavior varied among the 23 focal gangs.
Only seven of the focal groups were involved in at least 10 attacks, with at
least five incidents occurring before the injunction and five incidents postCGI. We set this threshold to ensure that the gangs examined were sufficiently active during the observation periods for a change in behavior to be
observable.9 Table 5 describes the aggressive behavior of seven gangs in
comparison with the non-gang community. In addition to reporting on the
outdegree centrality scores, the table includes the average number of attacks
by year. We use raw outdegree centrality scores calculated for the directed,
valued networks, as our interest is the amount of victimization caused by
each gang. Outdegree centrality is a network metric that captures the number
of victimizations originating from each gang: Raw scores are counts and
rather than normalizing by the size of the networks, we opted to average
scores by time. Dividing the annualized value for the post-CGI period by the
893
Bichler et al.
Table 5. Focal Gangs Initiating at Least 10 Attacks With at Least Five Attacks Preand Post-Injunction.
Attacks (outdegree
centrality)a,b
Focal gang
Bloods-affiliated gangs
Black P Stones—CGI
9/21/2006
Bounty Hunter Bloods—
CGI 8/26/2003
Crips-affiliated gangs
Grape Street Crips—CGI
5/25/2005
Venice Shoreline Crips—
CGI 10/18/2000
Rollin 60s—CGI
11/24/2003
48th Street Gangster
Crips—CGI 4/7/2005
Geer Street—CGI
9/22/2006
Non-gang communitya
Average
(per year)
Total
Pre
Post
Pre
Post
Change
112
26
86
2.9
9.6
3.3
69
14
55
2.3
4.6
2.0
94
37
57
4.6
7.1
1.5
31
9
22
3.0
1.5
0.5
26
7
19
1.2
1.6
1.4
14
6
8
0.8
1.0
1.3
14
6
8
0.7
0.9
1.3
23
13
10
—
—
(0.8)
Note. CGI = civil gang injunction.
aPre- and Post-CGI is determined by the focal gang associated with incident.
bNormed values are available upon request.
pre-CGI period provides an indication of the change. As network size can
affect this analysis, an examination of density (a percent metric) and normed
betweenness (standardized metric) was included in the egocentric analysis
described in a subsequent section.
Of the two highly active Bloods gangs, the Black P Stones exhibit the
most aggression overall with an outdegree centrality score of 112, a value
almost twice that of the Bounty Hunter Bloods. Comparing pre- and postCGI activities, we find that the violence initiated by the Black P Stones was
on average 3 times higher during the post-CGI observation period. While the
Bounty Hunter Bloods also exhibit an increase in violence post-CGI, the
increase was less dramatic.
Five Crips-affiliated gangs meet the threshold level of aggression. The
most violent by far was the Grape Street Crips. Enjoined on May 25, 2005,
this group initiated an average of 4.6 victimizations per year before being
894
Crime & Delinquency 65(7)
subject to behavioral restrictions. Their use of violence increased post-CGI to
an average of seven victimizations per year. This pattern of increased violence repeats for three other gangs. Only the Venice Shoreline Crips were
observed to reduce their use of violence.
Table 6 shows evidence that egocentric networks changed, increasing in
size and density, and trending toward greater betweenness post-CGI.10 Size
denotes the number of alters (other groups) that the focal gang is involved
with. For outward egocentric networks, size indicates the number of groups
attacked by the focal gang (referred to as the ego). Table 6 reports that size
was consistent or increased for most of the groups post-CGI, suggesting that
the increased attack behavior reported in Table 5 was distributed among more
groups.
Density calculates the percent of alters that are connected while ignoring
all connectivity to the ego. In other words, the density of the egonetwork
indicates how much the groups in conflict with the focal gang attack each
other. For example, if Gang A is in conflict with four other gangs, but those
four other gangs are also battling each other, then Gang A is deeply enmeshed
in a dense web of violence. Even if Gang A refrains from violence, networked
violence will continue. Alternatively, if the other four gangs are not battling
each other, then Gang A is not constrained by an intense web of conflict. In
this scenario, if Gang A refrains from violence, the violence network surrounding them might dissolve. Increasing density suggests a constricting of
the local social neighborhood in that the ego is enmeshed in a more integrated
world of violence than it had been previously. For example, none of the
groups that Grape Street targeted were observed to attack each other during
the pre-CGI period; however, density increased to 16.7% post-CGI. This
increase in alters engaging in conflict post-CGI suggests that the Grape Street
Crips became more enmeshed in a web of violence following the imposition
of the CGI. As reported in Table 6, density increased for most of the focal
gangs in the post-CGI period.
The column named, nEgoBe, reports normed ego betweenness. This metric is normed to account for differences in network size. As used in a directed
egocentric network, betweenness measures the extent to which each ego sits
between pairs of others in a directed chain: In this domino or knock-on effect,
a group attacks the ego group, who then attacks a different group. For
instance, while the size and density of the Bounty Hunter Bloods egocentric
network increased, we found a more substantive change in their position of
betweenness. Pre-CGI, the Bounty Hunter Bloods attacked one group, then
post-CGI, they attacked four groups, who themselves were involved in conflict, and 41.7% of this change put the Bounty Hunters in the middle of a
knock-on or domino chain of violence. Notably, both Bloods gangs were
895
Bichler et al.
Table 6. Outward Egocentric Descriptive Statistics for Pre- and Post-CGI.
Focal gang (year CGI
imposed)
Bloods-affiliated gangs
Black P Stones (2006)
Bounty Hunter
Bloods (2003)
Crips-affiliated gangs
Grape Street Crips
(2005)
Venice Shoreline
Crips (2000)
Rollin 60s (2003)
48th Street Gangster
Crips (2005)
Geer Street (2006)
Pre-CGI
Size Density(%)
Post-CGI
nEgoBe
Size
Density(%) nEgoBe
3
1
0.00
—
0.00
—
3
4
0.00
8.33
33.33
41.67
4
0.00
0.00
7
16.67
40.48
2
0.00
50.00
4
8.33
0.00
3
1
16.67
—
16.67
—
3
1
50.00
—
8.33
—
3
16.67
0.00
2
50.00
0.00
Note. CGI = civil gang injunction.
situated within more chains of violence post-CGI; whereas, no clear pattern
emerged for the Crips gangs examined—Some groups were centrally positioned in more chains of violence post-CGI and others were positioned within
fewer chains.
Victimization. Table 7 reports victimization levels for all gangs suffering at
least 10 victimizations during the observation period. Five gangs, identified
with an asterisk, also appear in Table 5 as being among the most aggressive.
Of the Bloods gangs, the Bounty Hunter Bloods suffered the most victimizations, and the level almost doubled following their enjoinment. The Black P
Stones, though targeted less in the pre-injunction period, experienced 3.5
times more victimizations post-enjoinment. While the Crips gangs suffered
fewer victimizations, two groups—the Rollin 40s and the Grape Street
Crips—suffered five and 6.5 times more attacks, respectively, during the
post-enjoinment observation. Other Crips gangs experienced a substantial
reduction in victimization. Most striking, and contrary to the expressed aim
of injunctions, the non-gang community experienced more overall violence
after enjoinment. As the community was not subject to an injunction, we
coded pre–post-injunction victimization based on the status of the attacking
gang. Our results show that 1.8 times more victimizations were committed by
enjoined gangs against the non-gang community post-injunction. Yearly
averages were not calculated.
896
Crime & Delinquency 65(7)
Table 7. Focal Gangs Experiencing at Least 10 Victimizations.
Victimizations
(indegree
centrality)a,b
Focal gang
Bloods-affiliated gangs
*Bounty Hunter Bloods–
CGI 8/26/2003
*Black P Stones–CGI
9/21/2006
Swan Bloods–CGI
12/15/2009
Crips-affiliated gangs
Rollin 40s–CGI 4/10/2000
*Grape Street Crips–CGI
5/25/2005
*Rollin 60s–CGI
11/24/2003
*Venice Shoreline
Crips–CGI 10/18/2000
Main Street Crips–CGI
12/15/2009
Non-gang community
Average
(per year)
Total
Pre
Post
Pre
Post
Change
50
11
39
1.8
3.3
1.8
27
6
21
0.7
2.3
3.5
14
14
0
1.2
0.0
0.0
26
15
1
2
25
13
0.3
0.3
1.7
1.6
5.0
6.5
15
10
5
1.7
0.4
0.3
13
11
2
3.7
0.1
0.0
12
12
0
1.0
0.0
0.0
630
227
401
—
—
(1.8)
Note. CGI = civil gang injunction.
aPre- and Post-injunction designations were determined by the focal gang associated with the
incident.
bNormed values are available upon request.
Examining the size, density, and normed ego betweenness for victimization, Table 8 shows that groups experiencing an increase in victimization
were being attacked by more gangs. We observed that egocentric networks
for three gangs—the Bounty Hunter Bloods, Black P Stones, and Rollin
40s—exhibited greater density, suggesting that post-CGI conflict increased
among alters, deepening the web of violence around these focal gangs.
Betweenness statistics confirm that the five gangs experiencing more violence were situated between more groups in victimization chains (domino or
knock-on effects) following the imposition of injunctions.
The Jaccard coefficient of similarity, reported in Table 9, compares two
networks and measures the proportion of the networks that are the same
between observation periods: Jaccard is used to indicate whether the patterns
of relations observed in two networks are consistent. Generally, scores below
897
Bichler et al.
Table 8. Inward Egocentric Descriptive Statistics for Pre- and Post-CGI.
Focal gangs (year CGI
imposed)
Bloods-affiliated gangs
*Bounty Hunter
Bloods (2003)
*Black P Stones
(2006)
Swan Bloods (2009)
Crips-affiliated gangs
*Grape Street Crips
(2005)
Main Street Crips
(2009)
Rollin 40s (2000)
*Rollin 60s (2003)
*Venice Shoreline
Crips (2000)
Pre-CGI
Size
Post-CGI
Density (%) nEgoBe
Size
Density (%) nEgoBe
2
0.00
0.00
5
5.00
15.00
3
0.00
0.00
9
8.33
22.22
3
0.00
0.00
0
0.00
0.00
1
—
—
5
0.00
40.00
2
0.00
0.00
0
0.00
0.00
1
1
3
—
—
0.00
—
—
33.33
6
3
1
10.00
0.00
6.67
33.33
Note. CGI = civil gang injunction.
.2 suggest the networks are not the same and values above .6 are indicative of
a high level of similarity (or stability); when scores fall within this range, the
network exhibits substantive evolution (Snijders, Van De Bunt, & Steglich,
2010). The proportions are simple to interpret: A value of .2 indicates that
20% of the links are the same in both networks.
Our analysis revealed that only one gang, Geer Street Crips, with a score
of .67 (or 67% similarity)—calculated from two continued conflicts divided
by three observed violent relations summed between both time periods—
could be classed as being embroiled in the same network with some change.
Venice Shoreline Crips exhibits a moderate level of similarity that is suggestive of continuing conflict, but all other groups exhibit such high levels of
change that they appear to be enmeshed in new networks of conflict during
the post-CGI period.
Dissecting the observed change, we find that for five of the seven gangs
listed in Table 9, most of the change observed involves the start of new conflicts, meaning that between pre- and post-injunction observations, gangs
were more likely to become involved in new conflicts, rather than reciprocate
or terminate a conflict. Thus, comparing pre- and post-CGI egocentric networks, this study finds some evidence to suggest that material change
898
Note. CGI = civil gang injunction.
Bloods-affiliated gangs
Black P Stones (2006)
Bounty Hunter Bloods (2003)
Crips-affiliated gangs
Grape Street Crips (2005)
Venice Shoreline Crips (2000)
Rollin 60s (2003)
48th Street Gangster Crips (2005)
Geer Street (2006)
Focal gangs (year CGI imposed)
9
14
14
7
7
2
3
.07
.29
.14
.00
.67
Total violent
relations
.22
.07
Jaccard
coefficient
1
2
1
0
2
2
1
9
3
4
1
0
4
10
New
conflicts
0
0
0
0
0
1
0
Reciprocated
attacks
Nature of conflict with groups
(comparing pre- to post-injunction)
Continued
conflicts
Table 9. Description of Change in Conflict Between Ego and Other Groups.
4
2
2
1
1
2
3
Terminated
conflicts
Bichler et al.
899
occurred in the dyadic structure of conflict following the imposition of
injunctions—Network violence is highly dynamic at this level of analysis.
Triadic Structural Change
Examining structural change through a triad census, we find mixed results
(see Table 10). For 60% of the gangs evaluated, reciprocal violence was null,
indicating no change; however, the remaining four gangs saw an increase in
reciprocated violence. The intensification of conflict post-injunction, however, tended to take the form of an in-star formation; 60% of gangs experienced an increase in attacks from pairs of other gangs. Notably, this means
that for the remaining four gangs, out-star formations increased, indicating an
intensification of attack behavior—about 30% of gangs developed out-star
activity post-injunction when this structure was not observed prior to CGI
imposition.
Directed and complex patterns (transitivity) exhibited the most dramatic
changes. For instance, the Black P Stones and the Bounty Hunter Bloods
became more deeply enmeshed in a complex pattern of violent conflict following CGI imposition, as did the Grape Street Crips, Rollin 40s, and Rollin
60s. Thus, comparing pre- and post-CGI triadic structure, this study finds
some evidence to suggest material change occurred in the triadic structure of
conflict following the imposition of injunctions.
Discussion
Key Findings
The results indicate material difference in attack behavior post-injunction. Of
the seven gangs that met the minimum threshold of inclusion, only the Venice
Shoreline Crips experienced a reduction in average attacks after the injunction went into effect. This finding is in line with prior work that found
increased violent crime post-injunction (see ACLU, 1997; Goulka et al.,
2009). Bloods gangs also experienced a change of a larger magnitude than
their Crips counterparts, possibly indicative of coalition differences, such as
differences in the number of cliques or focus of criminal behavior. Density
increases post-injunction, also suggesting greater integration of conflict postinjunction, the exception being the Geer Street Crips (a finding reinforced
through the Jaccard coefficient for relative similarity between the pre- and
post-injunction networks). On average, our sample indicates that networked
violence appeared to increase after the injunction was instituted; this includes
violence against non-gang individuals.
900
0/0
0/0
Main Street (2009)
Venice Shoreline Crips (2000)
0/0
0/5
2/0
0/4
1/0
2/0
1/0
0/0
0/16
0/2
0/0
0/3
6/0
3/23
1/9
(In-star)
Complex
3/0
2/4
0/0
0/13
0/2
0/0
3/15
7/0
2/14
4/3
2/1
3/1
1/1
0/14
2/12
0/0
1/23
3/0
1/23
2/15
(Directed) (Transitive)
Chain-like
No change
No change
No change
Increase
No change
No change
Increase
No change
Increase
Increase
Reciprocal
No change
Increase
Decrease
Increase
Decrease
No change
Increase
Decrease
Increase
Decrease
(Out-star)
Extreme
increase
Decrease
Decrease
(Directed)
Chain-like
Decrease
Decrease
Decrease
Decrease
No change No change
Increase
Extreme
increase
No change No change
Increase
Increase
Increase
No change
Extreme
increase
Extreme
increase
Decrease
(In-star)
Intensified conflict
Nature of change
No change
Extreme
increase
No change
Increase
Extreme
increase
Decrease
Decrease
Extreme
increase
Extreme
increase
Decrease
(Transitive)
Complex
Note. Increase refers to up to 100% increase, extreme increase indicates greater than 400% increase, decrease indicates up to 100% decrease. CGI = civil gang injunction.
0/0
0/3
0/0
0/0
6/8
0/0
0/19
Geer Street (2006)
Rollin 40s (2000)
Rollin 60s (2003)
2/0
0/0
Swan Bloods (2009)
Crips-affiliated gangs
48th Street Gangster Crips (2005)
Grape Street Crips (2005)
0/1
3/2
(Out-star)
0/6
0/1
Reciprocal
Bounty Hunter Bloods (2003)
Bloods-affiliated gangs
Black P Stones (2006)
Focal gangs (year CGI imposed)
Intensified conflict
Count of structures (pre/post)
Table 10. Pre- and Post-CGI Triad Census of the Most Aggressive and Victimized Gangs.
Bichler et al.
901
Hennigan and Sloane (2013) found little evidence of change in reported
group cohesion post-injunction (despite spending less time together in the
streets), and we found that local hierarchies intensified post-injunction:
Considering these results together, it is plausible that the enjoined gangs continued to function as a group and they had to intensify their fight for dominance within their social neighborhood because five of the groups committing
the most attacks post-injunction were also the groups most victimized during
the same period. These groups are Black P Stones (Bloods), Bounty Hunter
Bloods, Grape Street Crips, Venice Shoreline Crips, and Rollin 60s (Crips).
Although these gangs were operating within a larger sphere of gang violence
in the greater Los Angeles area, few gangs were engaged in reciprocal violence. Instead, the current study found evidence of victimization involving a
knock-on or domino chain of violence—Groups were situated between others
in such a fashion that a pecking order might exist. Simple structures were
embedded within more complicated triadic structures that suggest that the
imposition of CGIs is associated with a deepening of the inter-gang conflict.
While our data do not permit us to investigate why these structures developed,
we can speculate from anecdotal information. The imposition of an injunction
involves a public statement, direct notification, and an increase in visible
police presence and this may be why previous studies that found CGIs to be
effective often found those effects to be short lived (e.g., Maxson et al., 2005;
O’Deane & Morreale, 2011). Pre-injunction conflicts terminate due to public
scrutiny and intensive police activity, thereby producing a temporary lull in
gang activity. The targeted gang retreats (to some extent) from public view, in
their neighborhood at least, only to reappear elsewhere. Whether the targeted
gang is perceived as weak, responds with increased aggression, or splits into
factions, their reaction to the CGI triggers a shift in the local scene that generates conflicts with new groups. Post-CGI, groups become even more enmeshed
in communities of violence, albeit a different “community” than before enjoinment. Although somewhat speculative, this argument concurs with qualitative
research by Swan and Bates (2016) and is consistent with studies of networked
gang violence (e.g., Papachristos, 2009; Papachristos et al., 2013).
Overall, our findings suggest that once a gang is subject to a CGI, other
gangs may be empowered to attack in a bid to increase their own reputations.
In this sense, the injunction can be viewed as a mechanism for reducing the
informal street-level power differential between two gangs, a notion that is
accepted within the theoretical literature (see Quinney, 1970, 1977). Swan
and Bates’s (2016) finding of increased futility and injustice feelings among
gang-affiliated individuals post-injunction is in line with this conflict theoretical implication. In seeking to expand the evaluation literature on CGIs,
the results of the current study suggest that injunctions are associated with a
902
Crime & Delinquency 65(7)
noticeable change in networked violence, though not in the intended direction—Networked violence between gangs actually increased post-injunction.
Several implications follow from these findings.
Implications
CGIs impose significant, long-lasting restrictions on individuals, and yet,
there are few scientific evaluations of the effect that CGIs have on the individuals and groups involved. Factoring for the significant costs associated
with implementing and enforcing injunctions,11 scholars are beginning to
question whether the limits placed on personal mobility, ability to congregate, and freedom of expression are justified; perhaps, there are other focuseddeterrence strategies that are more effective and cheaper. Alternatively, there
may be a way to modify CGIs to strengthen their impact without adding to
the social and financial costs they pose. Building on emerging research about
the interconnected nature of gang-related violence, these results stand to
advance crime prevention policy in four ways.
Implication 1: Expand data sources to include modern forms of inter-gang
communication.
CGIs aim to restrict social interactions among gang members, with the intent
of reducing violent conflict. Gang members, like the rest of us, live to a large
degree in hyperspace—the social world generated by the interconnection
between online social activity (e.g., Facebook and other social media) and
physical activity. In rethinking routine activities theory, P. L. Brantingham
and Brantingham (2015) argued in favor of expanding the conceptualization
of routine activity leading to offender convergence, as well as victim and
offender interaction, to include online behavior and how this activity intersects with the physical world (hyperspace). Communication is less and less a
matter of direct face-to-face contact and our explanations of crime, and more
importantly, our responses, must adjust accordingly.
It follows that prohibiting “hanging out” in physical space may not curtail
violence stemming from online interactions and inflammatory statements.
With the current stipulations imposed by injunctions, gangs may continue to
terrorize communities through virtual interactions without violating the terms
of their injunction. Searching each focal gang name, we found that 78% had
multiple YouTube videos that could be considered recruiting tools that also
serve to intimidate the community and inflame rivalries. Some videos had
more than 2 million views. Searching all 72 enjoined gangs, we discovered
that 73% had similar videos. Undoubtedly, this migration out of physical
Bichler et al.
903
space may alter the nature of gang violence. Consequently, any effort to map
the structure of inter-gang conflict should include information obtained by
mining law-enforcement records, as well as investigation of various social
media channels. Interactions—whether electronic or face-to-face, in real time
or asynchronous—transmit information that affects perceptions, beliefs, attitudes, and behavior. Thus, efforts to map interactions that fuel conflict should
extend to electronic media.
Implication 2: Agents of the criminal justice system need to be specific
about gang and clique involvement and the local social neighborhood
within which individuals and groups are embedded.
Repeated network-oriented efforts to track clique and group membership are
vital to understanding conflicts. Sierra-Arevalo and Papachristos (2015)
argued that regular and routine efforts to conduct gang audits, that map the
alliances, rivalries, and conflicts of gangs, as well as the current claimed turf,
are critical to supporting gang-focused gang suppression and deterrence
efforts. Group-on-group conflict is dynamic. Viewing interactions in a network format, as well as the spatial or geographic context of inter-gang activity, will significantly enhance suppression efforts. As illustrated by Tita and
Randil (2011), groups do not necessarily attack those sharing turf or those in
adjacent areas, but rather, they may also victimize groups that are socially
proximate. For example, rival gang members may attend the same high
school but they live and claim turf in neighborhoods located miles apart. By
demonstrating a method for identifying the structural effects of CGIs on networked violence, this study draws attention to an analytic process that can be
used in support of the development of future injunctions. A network approach
can help refine the scope of an injunction and improve target selection so as
to enhance the impact of these sanctions and minimize the likelihood of triggering new conflict.
With this said, it is important to keep in mind that deconstructing the
aggregate and long-term impacts of injunctions is a difficult process. Several
factors complicate efforts to map the social structure of gang violence. First,
entity resolution is challenging. Many gangs have multiple names and membership boundaries are dynamic, that is, large gangs tend to form factions/
cliques with their own set of names and some individuals have multiple affiliations over time. These intersecting issues make it difficult to assign the correct group membership to individuals involved in conflict.12 Also, many
gangs claim the same territory. Overlapping territory adds confusion to the
identification of gang members. Thus, despite efforts to identify gang-associated individuals for focused deterrence or prosecution, clear and current
904
Crime & Delinquency 65(7)
information about group affiliation can be illusive. Given these challenges,
and in light of the dynamic nature of gang involvement, future efforts may
benefit from using an SNA approach to identify group involvement using
information about interactions, instead of formal gang membership.
Groups can be defined by the structure of interactions. Rather than imposing labels about group membership, it might be more advantageous to classify the probability of membership by the primary social group (e.g.,
co-arrested for burglary, social media friends, and observed hanging out
together) and the homogeneity of their behavioral characteristics (e.g., attends
the same school, lives in the same apartment complex, similar age, and similar record of performance in school). For instance, mapping interactions that
are identified in field interrogation reports in conjunction with other investigative information (e.g., surveillance, wiretaps, and social media interactions), makes it possible to map a person’s network. If a large proportion of
the personal network, maybe 90% of contacts, exhibit gang behavior, then
there is a higher probability that she or he is more enmeshed in a gang-oriented lifestyle than someone who has less than 10% of gang-affiliated associates. By applying clique statistics, it is possible to identify cohesive subsets
of people with appropriate confidence intervals. Readers should note that the
term clique is used in SNA to refer to a set of analytic tools used to identify
groups based on local patterns of connectivity. Moving beyond gang labels,
clique analysis may improve efforts to identify criminal groups if group
cohesiveness is integrated with an assessment of the homogeneity of behavioral characteristics.
Implication 3: Gang violence is dynamic and not contained by neighborhood, community, or city boundaries—a regional approach is required.
Prior research found that the effects of CGIs are short lived (e.g., Maxson
et al., 2005; O’Deane & Morreale, 2011). This finding is supported by the
present study: Individual behavior aggregated to the gang-level exhibits substantial change in the structure of conflict. As prior studies were conducted at
the community-level, this consistency is not unexpected. One explanation for
this phenomena can be drawn from a major tenet of social network theory—
relations among people form and dissolve as individuals interact and respond
to the perceived reactions of others (see Christakis & Fowler, 2009;
Wasserman & Faust, 1994). At the individual level, small changes may not be
immediately apparent, when aggregated; however, the change becomes substantial. While the transient nature of gang involvement combined with the
outcome of the violent event (e.g., death or imprisonment) may partially
Bichler et al.
905
account for our findings, residential migration is also a factor in Southern
California.
While the focal point of this study was the City of Los Angeles, the sampling strategy cast a wide net, and inadvertently captured some of the compound effects of CGIs throughout a five county region—Los Angeles,
Orange, Ventura, Riverside, and San Bernardino. One of the contributing factors is the ongoing pattern of residential migration throughout the region
caused by economic and housing crises, employment, and familial relations.
As a result, Los Angeles–based gangs and inter-gang conflict extends beyond
the boundaries of the City of Los Angeles. Though not investigated directly,
it is plausible that residential instability in part accounts for some of the triadic complexity—seeking to avoid being seen in their neighborhoods with
others, gang members are sent to live with relatives in other cities. Displacing
gang members may inadvertently spread conflict. Due to the complexity of
conflict, there is a need to coordinate across precincts, we advise that suppression efforts focus on the groups involved in the most violence (victimization and aggression) because the structure of group-on-group conflict is more
apt to be chain-like—a domino effect suggestive of a social pecking order.
Uncovering how gang violence ripples through a regional social network
helps to identify where interagency cooperation could maximize the benefits
of anti-gang efforts (Randle & Bichler, 2017). Prosecutors, law enforcement,
probation officers, and those involved in anti-gang programming will be able
to use our results to facilitate an open dialogue about CGIs and the possible
alternatives, to weaken the web of violence and improve community safety
across the metropolitan region.
Implication 4: Typically, focused-deterrence initiatives offer, or try to
offer, some kind of social services for the targeted group, and thus,
while CGIs may be thought of as a focused-deterrent strategy, they may
not be so.
This study found that violence increased post-injunction. A plausible explanation comes from the work of Swan and Bates (2016). Gang members have gang
ties extending across familial relations and gang members have pro-social relations with individuals living in the safe zones. Because all public socializing is
under scrutiny, when enjoined individuals interact with non-gang involved
individuals, the non-gang associates could be labeled as gang affiliates. As
such, this by-product of injunction status may compromise a reformed gang
member’s ability to form pro-social networks of support. Moreover, Swan and
Bates (2016) found that enjoined gang members were often blocked from
social services. Taken together, it is plausible that as currently implemented in
906
Crime & Delinquency 65(7)
Los Angeles, CGIs satisfy most of the characteristics of a focused-deterrent
tactic (Braga et al., 2008; Kennedy, 1997); with one critical omission, there are
no mechanisms to provide targeted social services that will support the formation of a pro-social network of support. This may suggest that this missing
element is a quintessential component of focused deterrence.
Limitations and Future Research
Five notable limitations restrict the generalizability of our findings. Foremost,
as noted above, there was little consistency in how gang experts described
gangs and gang affiliations during their testimony. This situation made it
harder to identify which gang and clique the offenders and victims belonged
to. Thus, future research should begin with an effort to map existing gang and
clique structures, as well as inter-gang associations and conflict from a range
of sources, before attempting to extract information from case notes. While
such an effort should aid entity resolution, it is imperative that greater effort
be made on the part of gang experts to identify group involvement, perhaps
focusing more on the current social group as opposed to formal gang membership. Greater effort to describe current gang associations during investigations and trials improves research, but it is also necessary for CGIs to be
effective because prosecutors and law enforcement must have a clear understanding of the extent of gang involvement in the local social neighborhood,
as well as the nature of competitive and collaborative relations among gang
members, to enforce court orders.
Second, we relied on LexisNexis to identify relevant cases. The problem
with using this search engine is that only appeal cases are cataloged for
California. Restricting our analysis to appeal cases may produce networks
representing murder cases, as these cases tend to receive lengthy prison sentences and generate appeals; however, it is unlikely that these findings represent less than lethal robbery and assault. For this reason, future research in
California should restrict case eligibility to murder and attempted murder
until case archives include the initial trials. Alternatively, other sources
should be explored.
The third notable limitation is that we do not control for the number of gang
members or size of the safety zone. As of 2007, there were an estimated 450
gangs in the city of Los Angeles with about 45,000 gang members. Some of
the gangs examined have in excess of a thousand members and claim territory
in different parts of Los Angeles.13 Larger gangs with more cliques, spread
over a greater area, are likely to generate more conflict, compared with smaller
geographically restricted groups. Moreover, more conflict is likely when there
are multiple injunctions covering gangs with overlapping territories. For
Bichler et al.
907
example, the Fremont Passage safety zone covers Swans Bloods, F-13, 7 Trey
Hustlers, and Main Street Crips. Other injunctions show safety zones overlapping for the Rolling 40s, Hustler Crips, and 46 Neighborhood Crips. Thus,
future research should control for these factors when assessing which groups
are more aggressive. Alternatively, as noted above, group membership can be
defined by social activity using clique identification metrics available through
SNA, rather than imposed by official labels.
The fourth limitation pertains to the interaction between high levels of
aggression and the number of gang members. Together, these factors increase
the likelihood of becoming enmeshed in a more complex web of conflict. If a
gang is well connected, in this case meaning that they are more aggressive,
there is an increased possibility of additional interactions (Hao, Park, & Pei,
2017). Moreover, larger gangs face a greater opportunity for new violence.
Because the current study sought to evaluate pre- and post-injunction violence for gangs with sufficient activity and that larger groups are more likely
to become enmeshed in conflict, these results are most applicable to the largest, more violent organizations in the Los Angeles area. Future research
should control for organizational size of the gang to ensure this characteristic
does not affect the results.
Finally, we do not control for the spatial proximity of gang turf. P. J.
Brantingham, Tita, Short, and Reid (2012) presented an argument for the
emergence of gang territories by competition, noting specifically, the relative
lack of overlap between territories. P. J. Brantingham and colleagues (2012)
suggested that rapid competition between gangs early on in turf wars will
result in changes to the territory, suggesting that gang injunctions covering
areas with multiple feuding gangs might curb this competitive nature before
a gang can build a stronghold in the area. Furthering this discussion, Smith,
Bertozzi, Brantingham, Tita, and Valasik (2012) found that for a rivalry to be
expected in a given area, there should be sufficient overlap of the densities of
two gang networks. Moreover, studies have found that while spatiality is
important, these elements could be non-contiguous (Huddleston, Fox, &
Brown, 2012; Tita & Randil, 2011); rivalries do not necessarily spill over into
neighboring territories. Thus, future research should account for the spatial
dispersion of conflicts, controlling for characteristics of both contiguous and
non-contiguous inter-gang conflicts.
Conclusion
CGIs impose a nuisance abatement order on a group involved in behavior
that is injurious to health, is considered indecent or offensive, and poses a
significant interference on the free use of public space and enjoyment of life
908
Crime & Delinquency 65(7)
and property. Standard prohibitions include refraining from drug use and
sales, possession of firearms, involvement in graffiti and vandalism, associating with known gang members, or intimidating people in the neighborhood.
Though used in many jurisdictions, the City of Los Angeles currently has the
most CGIs in the United States, with 46 permanent injunctions placed against
72 gangs. Using SNA, this study is the first to test the dynamic effects of
CGIs on the social structure of serious violence. Investigating inter-group
violence pre- and post-injunction, these results document the dynamic and
stable properties of street gang conflict which may help to improve focuseddeterrence efforts.
Of critical importance, CGIs are intended to reduce gang-related harms to
the community. Gang members, however, could perceive these sanctions as
admonishments or attacks by the community, and as such, court orders may
trigger significant change in the nature of gang violence, that is, CGIs may
increase the likelihood that gangs target citizens from the neighborhood. This
study found that attacks on members of the community continued despite the
presence of an injunction. In addition, for six out of seven very active focal
groups, violence initiated post-CGI was greater than before sanction.
Examining the nature of these attacks reveals the dynamic nature of conflict.
Taken together, these results suggest that while CGIs affect gang behavior, a
more comprehensive set of interdiction efforts may be needed to reduce gang
conflict and resulting violence.
Improving the potential for CGIs to reduce gang violence requires several
considerations. (a) A greater range of information, including harvesting from
social media channels, is needed to better identify inter-gang communication
and conflict, as well as identify gang and clique affiliation. (b) Data integration and analysis with network analytics will improve our efforts to track the
evolution of gang conflict and develop a package of intervention strategies
tailored to the current social landscape. (c) Mapping how violence ripples
through a regional social network may promote interagency cooperation that
could maximize the benefits of anti-gang efforts. (d) Until CGIs include
mechanisms to provide targeted social services that will support the formation of a pro-social network of support, they may not constitute a fully developed focus deterrent strategy.
Appendix
Enjoined Bloods- and Crips-Affiliated Gangs
As of 2007, there were an estimated 450 gangs in the city of Los Angeles with
about 45,000 gang members. In Los Angeles County, there was estimated to
Bichler et al.
909
be about 1,300 gangs with 150,000 gang members. Membership for each
group or sub-clique is not available (City of Los Angeles, 2007).
Bloods Coalition. Established: 1972
Estimated number of groups: 30 in the City of Los Angeles and 65 in Los
Angeles County
Bloods-affiliated gangs in the study:
••
••
••
••
••
••
••
Bounty Hunter Bloods (CGI August 26, 2003)
Black P Stones (CGI September 21, 2006)
All for Crime (CGI January 23, 2009)
Pueblo Bishops (CGI January 23, 2009)
Blood Stone Villains (CGI January 23, 2009)
East Side Pain (CGI June 11, 2009)
Swan Blood (CGI December 15, 2009)
Crips Coalition. Established: 1969
Estimated number of groups: 100 in the city of Los Angeles and 200 in the
County of Los Angeles
Crips-affiliated gang in the study:
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
••
42nd Street Gangster Crips (CGI April 7, 2005)
43rd Street Gangster Crips (CGI April 7, 2005)
46 Neighborhood Crips (CGI April 10, 2000)
46 Top Dollar Hustler Crips (CGI April 10, 2000)
48th Street Gangster Crips (clique of Rollin 40s) (CGI April 7, 2005)
7-Trey Hustlers/Gangster Crips (CGI December 15, 2009)
Geer Street (CGI September 21, 2006)
Grape Street (CGI May 25, 2005)
Harbor City (CGI February 1, 2000)
Main Street (CGI December 15, 2009)
Rollin 40s (CGI April 10, 2000)
Rollin 60s (November 24, 2003)
School Yard Crips (September 22, 2006)
Venice Shoreline Crips (October 18, 2000)
Playboys (CGI September 21, 2006)
Hoover Criminals Gang (CGI September 13, 2002)a
aThough not currently a Crips gang, the Hoover Criminals Gang was for a
time affiliated with the Crips, and for that reason, this group was included in
the study.
910
Crime & Delinquency 65(7)
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research,
authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Notes
1. Social network analysis (SNA) is an interdisciplinary, multidisciplinary field of
study dedicated to understanding how social relations, and the larger context
these interactions create, effect perceptions, beliefs, decisions, and behavior
(Wasserman & Faust, 1994). A deep and rapidly expanding body of research illustrates that SNA theory, methods, and analytic tools can be applied toward understanding crime (e.g., Morselli, 2009, 2014) and criminality (e.g., Carrington,
2011; Papachristos, 2011), and that crime prevention and suppression efforts,
including anti-gang strategies, could benefit from adopting a network approach
(see Bichler & Malm, 2015, in press).
2. The term enjoined denotes that a gang or person is subject to a civil injunction.
3. The Governor of California approved this legislation on September 28, 2016,
and its impact on court proceedings is still to be determined. Contested designations may change the number of gang enhancement penalties implemented, with
implications for sentence length. It is also plausible that networked violence will
change as well, with certain individuals being able to remove themselves from
the lens of police scrutiny. Because the current study only includes cases occurring prior to enactment of this legislation, AB-2298 does not affect the validity
of reliability of our findings.
4. There were no official expiration dates for civil gang injunctions (CGIs) in San
Diego County; however, respondents in the study noted that police officers
would tell them that as long as they were on good behavior, their CGI would
expire in 5 years. This lead to confusion when they were sanctioned for violations of the CGI years after they thought the CGI had expired.
5. Beyond the City of Los Angeles, there are more than 1,100 gangs in 115 cities
in the United States with “Bloods” or “Crips” in their name (Howell, 2012, p.
13). Underscoring the magnitude of their growth, Valdez (2007) asserted that all
black street gangs on the West coast affiliate themselves with the Bloods or the
Crips.
6. Two relevant injunctions, filed in 2013, are too recent to permit a complete follow-up period, and thus, associated cases were not included in the present study.
The gangs, however, were included in the study when conflict involved one of
the other seed gangs.
7. As aptly noted by a reviewer, gangs do not have equal membership. Larger gangs
are likely to be involved in more conflict. To address this issue, a case study
Bichler et al.
8.
9.
10.
11.
911
approach was used here, and as will be explained shortly, several of the key metrics used to examine network structure were standardized measures. With this
said, it is important to acknowledge that gangs may not be the best unit of analysis for this type of research. An individual’s social group is more dynamic, centered at the local level (an egocentric network). While the egocentric network is
embedded within the larger community, an individual’s specific connections and
reach best predict behavioral trends and, by extension, criminality (Pattiselanno,
Dijkstra, Steglich, Vollebergh, & Veenstra, 2015). Thus, clique affiliation would
have been more informative; however, testimony provided by gang experts was
inconsistent regarding clique identification. As such, official data may not accurately depict the behavioral context of an individual’s accession to gang involvement (Malm, Bichler, & Nash, 2011). We return to this issue in the “Discussion”
section.
It is possible that some accomplices were not identified or described during the
trial, or that individuals were named, but the information was not deemed relevant for inclusion in the summary of the facts reported in an appeal.
One of the reviewers suggested doing a sensitivity analysis to investigate the
effects of these criteria. While we concur with the reviewer, a proper sensitivity
analysis was not feasible with this sample. Reducing the threshold or not specifying involvement in at least five attacks before and following an injunction doubles
the number of gangs available for the analysis; however, doing so also introduces
a major threat to the analysis. It was not possible to determine if empty cells
indicated that behavior changed or if the group simply rebranded itself (changed
its name or split into factions). In other situations, groups appeared in the data
only after the imposition of the injunction leaving us to speculate whether the
increased surveillance by law enforcement improved detection of violence that
previously would have gone unsolved. Thus, reducing the thresholds would not
help us to identify whether violence changed with the imposition of the CGI; in
fact, it would cloud our ability to detect change. Alternatively, we could raise the
threshold; however, raising the threshold decreases the sample size.
These metrics are described briefly here. For more information about these metrics and other analytics available for egocentric networks, please see Borgatti,
Everett, and Johnson (2013) or Wasserman and Faust (1994).
While few assessments of the financial burden posed by CGIs exist, it is reasonable to suggest that this anti-crime policy does incur considerable costs as
implementation requires effort over and above conventional criminal justice system activity. For instance, Grogger (2005) estimated that the 14 injunctions he
studied in Los Angeles (1993-1998) cost upward of 1.4 to 2.1 million dollars to
prosecute, and possibly an equal amount for enforcement; however, given the
possible reduction in crime, this may be warranted if metrics include the social
and economic costs of crime prevented. Goulka et al. (2009) documented that the
enforcement and prosecution costs associated with one injunction in Santa Ana,
Califor…
Purchase answer to see full
attachment

  
error: Content is protected !!