ArticlePDF Available

Intimidation and Street Gangs: Understanding the Response of Victims and Bystanders to Perceived Gang Violence

Authors:

Abstract and Figures

While research routinely examines the influence of gang membership on the quantity of violent crime involvement, less is known about the influence of gang violence on the situational characteristics of violent victimization. Felson's discussion of street gangs highlights the possible functional role gang membership plays in the commission of violent crime; what he terms othe street gang strategy.o This study examines the functionality of gang membership during violent crimes by investigating the influence of perceived gang membership on the likelihood of victim resistance, bystander intervention, and police reporting using data from the National Crime Victimization Survey. Findings offer little support for the idea that gang members intimidate victims and bystanders to the extent that their behavior during and after violence differs systematically from responses resulting from non-gang violence. Results are discussed in terms of their policy relevance and implications for future research.
Content may be subject to copyright.
PLEASE SCROLL DOWN FOR ARTICLE
This article was downloaded by:
[Michigan State University]
On:
10 August 2010
Access details:
Access Details: [subscription number 918013331]
Publisher
Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-
41 Mortimer Street, London W1T 3JH, UK
Justice Quarterly
Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t713722354
Intimidation and Street Gangs: Understanding the Response of Victims and
Bystanders to Perceived Gang Violence
Chris Melde; Callie Marie Rennison
First published on: 02 October 2009
To cite this Article Melde, Chris and Rennison, Callie Marie(2010) 'Intimidation and Street Gangs: Understanding the
Response of Victims and Bystanders to Perceived Gang Violence', Justice Quarterly, 27: 5, 619 — 666, First published on:
02 October 2009 (iFirst)
To link to this Article: DOI: 10.1080/07418820903228858
URL: http://dx.doi.org/10.1080/07418820903228858
Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf
This article may be used for research, teaching and private study purposes. Any substantial or
systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or
distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss,
actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly
or indirectly in connection with or arising out of the use of this material.
JUSTICE QUARTERLY VOLUME 27 NUMBER 5 (OCTOBER 2010)
ISSN 0741-8825 print/1745-9109 online/10/050619-48
© 2010 Academy of Criminal Justice Sciences
DOI: 10.1080/07418820903228858
Intimidation and Street Gangs:
Understanding the Response of
Victims and Bystanders to
Perceived Gang Violence
Chris Melde and Callie Marie Rennison
Taylor and FrancisRJQY_A_423059.sgm10.1080/07418820903228858Justice Quarterly0741-8825 (print)/1745-9109 (online)Original Article2009Taylor & Francis000
0000002009Dr ChrisMeldemelde@msu.edu
While research routinely examines the influence of gang membership on the
quantity of violent crime involvement, less is known about the influence of
gang violence on the situational characteristics of violent victimization.
Felson’s discussion of street gangs highlights the possible functional role gang
membership plays in the commission of violent crime; what he terms “the
street gang strategy.” This study examines the functionality of gang member-
ship during violent crimes by investigating the influence of perceived gang
membership on the likelihood of victim resistance, bystander intervention, and
police reporting using data from the National Crime Victimization Survey. Find-
ings offer little support for the idea that gang members intimidate victims and
bystanders to the extent that their behavior during and after violence differs
systematically from responses resulting from non-gang violence. Results are
discussed in terms of their policy relevance and implications for future
research.
Keywords gangs; victimization; offending; resistance; reporting; offending
Chris Melde is an assistant professor in the School of Criminal Justice at Michigan State University.
He received his PhD in criminology and criminal justice from the University of Missouri—St. Louis, in
2007. His primary research interests include program evaluation, juvenile delinquency and victim-
ization, gangs, perceptions of crime and victimization risk, and criminological theory. His recent
work has appeared in Criminology, the Journal of Research in Crime and Delinquency, and Crime
and Delinquency. Callie Marie Rennison is a lecturer at the University of Colorado—School of Public
Affairs. Previously, she was an assistant professor in the Criminology and Criminal Justice Depart-
ment at the University of Missouri—St. Louis, and a Statistician at the Department of Justice, Bureau
of Justice Statistics, in Washington, D.C. She received her PhD in political science from the Univer-
sity of Houston—University Park in 1997. Her primary research interests include victimology with a
focus on the nature, extent, predictors, and consequences of victimization. In addition, her
research has an emphasis on quantitative analyses, research methodology, and measurement. Her
research has appeared in the Journal of Quantitative Criminology, Violence and Victims, Women
and Criminal Justice, American Journal of Criminal Justice and Violence Against Women. Corre-
spondence to: Chris Melde, School of Criminal Justice, Michigan State University, 560 Baker Hall,
East Lansing, MI 48824, USA. Email: melde@msu.edu
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
620 MELDE AND RENNISON
Numerous studies have documented the disproportionate involvement of gang
members in acts of violent crime; so much that this finding is considered to be
“one of the most robust and consistent observations in criminological research”
(Thornberry, 1998, p. 147). Much less is known, however, about the influence of
gang violence on the characteristics of the victimization incident. Specifically,
does gang member involvement affect the nature of violent crime? After all, as
Short and Strodtbeck (1965) suggested, understanding gang violence necessi-
tates comprehension of the situational factors that lead to greater levels of
involvement in violence. The preponderance of evidence presented in the
media suggests that gang membership shapes the nature of violence, not just
the quantity, in that gang members are quicker to resort to disproportionately
greater degrees of violence than non-gang involved individuals (Esbensen &
Tusinski, 2007; Lane, 2002). In fact, significant legislative action has been
directed toward gang crime [e.g., sentence enhancements (Johnson, Webster,
& Connors, 1995), civil gang injunctions (Maxson, 2004)] and the fear and intim-
idation felt by local residents as a result of the presence of gangs in their
communities (Finn & Healey, 1996; Spadanuta, 2008; U.S. House of Representa-
tives, 1997a, 1997b; U.S. Senate, 1994). Only a handful of researchers have
examined the potentially unique incident-level characteristics of gang versus
non-gang crime (see e.g., Curry & Spergel, 1988; Maxson, Gordon, & Klein,
1985; Melde & Rennison, 2008; Pizzaro & McGloin, 2006; Rosenfeld, Bray, &
Egley, 1999), and while informative have focused almost exclusively on homi-
cide (for an exception see Melde & Rennison, 2008), which is a relatively rare
event even in the context of gang violence.
The current study examines the nature of gang versus non-gang, non-lethal
violence and is guided by several hypotheses drawn from Felson’s (2006) writings
on “The Street Gang Strategy.” Based on Felson’s (2006) discussion of the essential
features of street gangs, and the central role intimidation plays in the formation
and continuation of street gangs, we examine if and how the perceived gang status
of the offender affects the behavior of the victim and bystander(s) both during
and after a violent incident. Based on Felson’s (2006) assertions, we hypothesize
that victims of violent gang crime are less likely to resist their attacker(s). Second,
we hypothesize that bystanders will be less likely to intervene in gang perpetrated
violent crime. And finally, we hypothesize that victims and bystanders are less
likely to report gang involved crime to the local authorities.
To examine our hypotheses, we use data on violent crime incidents from the
National Crime Victimization Survey (NCVS). By using a nationally representative
household survey of persons over 12, we are able to evaluate Felson’s (2006)
hypotheses regarding the functionality of gang membership in a generalizable
fashion. Specifically, we seek to determine whether the perceived gang status
of the offender influences victim and bystander decision-making both during
(i.e., resistance or intervention) and after the event (i.e., reporting the inci-
dent to authorities), or whether these decisions are simply a function of other
incident-level characteristics, including victim, offender, and/or situational
factors that impact such interactions.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 621
Understanding the effect of gang membership on victim behavior is important
for a number of reasons. First, if gang members induce a lower level of crime
reporting based on resident fear of gang retaliation, official records of gang
involved crime are likely systematically under-counted, which could potentially
lead to more violence due to under-protection from the police (Baumer, 2002;
Kennedy, 1997). Second, victims of gang crime may also be denied victim assis-
tance resources that are filtered through the police department, and thus require
reporting of victimizations to the police. This may be especially problematic in
high poverty areas in large US cities where gangs are pervasive and have been
particularly problematic for many years, as members of these communities likely
already struggle to acquire these resources. Third, while evidence is mixed as to
whether forceful resistance is more or less advantageous than no resistance at
all (Tark & Kleck, 2004), the systematic control of victim or bystander behavior—
whether it be resistance or passive cooperation—during criminal incidents may
lead victims to incur a greater probability of injury. This may especially be the
case in incidents involving gang members, given the high rate of gun use by these
individuals (Bjerregaard & Lizotte, 1995; Howell, 1998; Lizotte, Tesoriero, Thorn-
berry, & Krohn, 1994; Melde & Rennison, 2008; Thornberry, Krohn, Lizotte, Smith,
& Tobin, 2003). Finally, although there has been significant legislative action tied
to the common belief that gang membership influences victim and bystander
behavior due to the ability of such groups to intimidate community residents,
empirical evidence of such an effect is wanting. The current study takes a first
step in addressing the more immediate impact gang membership has on victim
and bystander behavior during and after criminal incidents.
Theoretical Background
Theories of victim decision-making propose that victim behavior is directed by
two overarching goals during criminal incidents. The first is to lessen the imme-
diate pain and suffering that is likely to occur, and the second is to reduce the
prospect of future victimization (Baumer, 2002; Gottfredson & Gottfredson,
1988; Greenberg & Ruback, 1992; Ruback, Greenberg, & Westcott, 1984). To this
end, victims have a number of options from which to choose, including: direct
forceful resistance, non-forceful resistance (Baumer, Horney, Felson, &
Lauritsen, 2003), notifying the police (Baumer, 2002; Baumer, Felson, &
Messner, 2003), or doing nothing at all. Extant research uncovered a number of
victim, offender, and situational characteristics that influence such victim
behavior (see e.g., Baumer, 2002; Baumer, Horney, et al., 2003; Baumer &
Lauritsen, in press; Felson, Baumer, & Messner, 2000; Felson, Messner, & Hoskin,
1999). Missing from this literature, however, is an examination of the influence
of the perceived gang membership status of the offender on victim and
bystander behavior. After all, gang members are unlike many offenders in that
they purposefully make their presence known in many communities, which has
the potential to impact victim and bystander behavior in a systematic manner.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
622 MELDE AND RENNISON
Gang members represent a relatively unique portion of the offending popula-
tion. As opposed to a more covert (i.e., cryptic) method of conduct, they often
use very open signs of affiliation, including easily identifiable and sometimes
nationally known names (i.e., Blood, Crip, MS-13), colors, clothing, tattoos, and
graffiti to announce their presence. According to Felson (2006), the tendency
for gangs and their members to use common names, symbols, and colors is no
accident. On the contrary, gangs do so for very pragmatic reasons. Taking on
the name of a well-known gang offers immediate recognition and respect. As
Katz (1988) argued, gangs set themselves apart through their capacity to instill
fear in others around them. By taking on the name and corresponding symbols of
commonly recognized gangs, individual members can use the often violent
reputation of these outside groups without actually having to build such status
for themselves. Members of such gangs, thus, are afforded a level of respect
and deference that cannot be obtained easily either individually or as a group.
Much like Katz (1988), Felson (2006) argues that the ability to intimidate, or
create dread, is the essential feature of a street gang. Felson’s (2006, p. 316)
definition of a gang highlights this notion by claiming that a gang is “a very local
group of youths who intimidate others with overt displays of affiliation.” That
is, gangs are only functional, or perhaps even a gang at all, when they provide
members with the ability to intimidate others. It is through intimidation that
gang members perceive they can both protect themselves from potential attack
and commit crimes with little fear of reprisal.
Felson (2006) describes this process of gang behavior and symbolism as a
specific case of mimicry. Mimicry, in general, is a process whereby “an organism
(the mimic) simulates signal properties of a second living organism (the model)
which are perceived as signals of interest by a third living organism (the opera-
tor), such that the mimic gains in fitness as a result of the operator identifying
it as an example of the model” (Vane-Wright, 1980, p. 4).
1
In this case, gang
1. Although Felson (2006) believes that gangs represent a specific case of Müllerian mimicry, our
understanding of human learning and behavior leads us to believe that this is not an appropriate
example of such a dynamic. Müllerian mimicry is the process through which separate species
develop similar warning signals to deter would-be attackers, which results in mutual benefit for
both species (Kapan, 2001). For instance, separate butterfly species grow to mimic the color pattern
(i.e., warning coloration) of distasteful species to reduce the threat of attack from common preda-
tors (Kapan, 2001). Importantly, Müllerian mimicry does not imply that the mimic is any less—or
more—dangerous than the species it comes to look like (the model). In the case of our butterfly
example, both species have no real defensive capabilities beyond their ability to fly, it just so
happens that one species does not taste very good to their common predator(s). Lastly, Müllerian
mimicry must be beneficial for both species. Because butterfly predators use Pavlovian learning
(i.e., trial and error), both of these relatively defenseless species benefit because it increases the
pool of potential prey in a given area per the number of predators—the rate of predation through
trial and error is held constant—which reduces the likelihood members of each separate species will
be eaten during the trial phase, and thus both groups receive benefit (Speed, Alderson, Hardman, &
Ruxton, 2000); it is simply a case of safety in numbers. Humans have the ability to share knowledge
through human interaction and various forms of media. Thus, not every human has to have toiled
with a gang member to learn of their dangerousness, and thus avoid doing so in the future, as the
theory of Müllerian mimicry assumes (see Ihalainen et al., 2007 for a review). Thus, we have chosen
to keep our discussion of mimicry at a more general level.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 623
members (the mimics) have learned to use easily recognizable warning signals
developed by more famous gangs (the models) to create fear on the part of
community members and rival gangs (the operators). According to Felson (2006,
p. 312), this practice serves four purposes, both defensive and offensive:
(1) To scare off enemies
(2) To get your personal victims to comply with your wishes
(3) To discourage bystanders from interfering with your offenses
(4) To discourage witnesses from speaking against you later
Apart from the defensive (i.e., protective) role of gang membership, which has
been found to be more perceptual than real (Melde, Taylor, & Esbensen, 2009),
affiliation with such groups is hypothesized to provide an advantage in predatory
acts of violence. While it has been consistently documented that gang member-
ship is associated with an increased quantity of violent crime involvement,
suggesting gang membership plays a substantive role in violent offending, less is
known about the nature of this violence, and the potentially unique role gang
membership plays at the incident level. According to Felson’s (2006) proposi-
tions regarding the essential features of gang membership, being recognized as
a gang member provides these individuals an offensive advantage in criminal
incidents. More specifically, victims and bystanders will be less likely to resist,
intervene, or report the incident to authorities when they believe the offender
is a gang member.
Psychological research details how the process of mimicry can influence
common perceptions of gang members as disproportionately violent in nature.
As Mitchell (1996) describes in his discussion of the process of deception, what
makes our response to gangs:
both predictable and ubiquitous is our reliance upon pragmatic presumptions to
organize our understanding of and responses to the world. Such pragmatic
presumptions derive from our knowledge of regularity … With knowledge of
regularities, one can make predictions about someone from myriad signs they
exhibit, such as clothing and hairstyle … A regularity may also be expected
which is not based on usualness, much as a stereotype is expected but not
statistically normal, and this expectation of regularity is as easily used for
deception as is statistical normality …. (pp. 824-825)
The media’s focus on the most sensational and violent gang crimes likely
creates and later reinforces the common assumption (i.e., stereotype) that
gang members are disproportionately violent, and thus crimes committed by
gang members are more likely to turn violent than if committed by non-gang
members (Esbensen & Tusinski, 2007). Respondents in Lane’s (2002, pp. 457-
458) study of public perceptions of gang violence alluded to the seeming random
and severe nature of gang violence by stating:
I think they’re dangerous because they have very little value for human life
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
624 MELDE AND RENNISON
They have to kill somebody—to be in a gang and to make them powerful … They
kill for nothing.
In the end, the common belief that gang members are unduly violent may influ-
ence victim and bystander behavior in criminal incidents in hopes that they can
lessen the possibility of immediate injury and later retaliation by the gang. In
fact, lawmakers have assumed that such a process is at work in communities
across the USA.
The fear producing aspect of street gangs has received a great deal of atten-
tion from local and national policymakers (Finn & Healey, 1996; Spadanuta,
2008; U.S. House of Representatives, 1997a, 1997b; U.S. Senate, 1994). Concern
over the level of fear produced by street gangs has been recognized as a major
problem in the arrest and later conviction of gang members, as it is believed
that victims and witnesses are less likely to cooperate with authorities for fear
of gang retaliation. For instance, testimony at a 1997 U.S. House Subcommittee
on Crime, which focused on “Gang-Related Witness Intimidation and Retalia-
tion,” noted that “Fear and intimidation are the foundation of gang dominance
in our communities …” (U.S. House of Representatives, 1997b, p. 5; as cited in
Lane & Meeker, 2003). Since that time, concern over witness intimidation has
led to significant legislative action across the USA. By 2006, 19 states had
passed witness intimidation laws as a direct result of concerns related to the
threat posed by gangs (Spadanuta, 2008).
Clearly, prosecution and law enforcement officials are concerned with the
potential impact gangs have on their ability to arrest and later convict gang-
involved individuals for the crimes they commit. While public policy and limited
research is clearly supportive of Felson’s (2006) arguments regarding the ability
of gangs to intimidate individuals in a manner conducive to the commission of
violent crime, systematic empirical research on the impact gang members have
on victim and bystander behavior both during and after violent incidents is lacking.
Current Study
The current study seeks to determine whether or not gang membership influ-
ences victim and bystander decision-making by using Felson’s (2006) discussion
of “the street gang strategy” as a guiding theoretical framework. Specifically,
we investigate the relative likelihood of victim resistance and/or bystander
involvement in violent criminal encounters in which the perpetrator is
perceived to be a gang member, as well as the relative likelihood of the victim
or bystander(s) reporting the event to the police after such incidents. If Felson’s
(2006) assessments regarding gangs are accurate, we would expect to find that
victims are less likely to resist, and bystanders are less likely to intervene, in
situations where the perpetrator is perceived to be a gang member. Also, we
would expect that victims and bystanders would be less likely to report the
incident to the local authorities, for fear of gang retaliation. We also examine
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 625
whether the effect of gang membership on victim and bystander behavior is
mediated by other situational characteristics associated with victim actions. In
all, we seek to discern whether or not gang membership provides advantages in
the commission of violent crime, net of other factors known to influence victim
resistance and police reporting (e.g., injury, weapon presence, age).
Data, Measures and Analytic Strategy
Data
Felson’s (2006) description of the essential features of street gangs, including
the primary role intimidation plays in facilitating criminal activities in these
social circles, requires a better understanding of incident-level characteristics
of gang and non-gang crime. Unfortunately, data commonly used to examine
gang violence (e.g., police data, self-report data, and ethnographic research)
often do not include much if any information on incident-level characteristics.
Victimization data, however, offer a unique perspective on the nature of gang
crime. As Block (1981, p. 747) described, “Criminal justice system records are
highly appropriate to study the effect of victim characteristics or victim-
offender dynamics on police, prosecution, and court decisions. They are not
appropriate to study the background of the criminal events, the relationship of
victim and offender, or the dynamic of victim-offender interaction. To study
these, knowledge of the victim’s perspective and environment is most appropri-
ate.” After all, “human behavior is primarily driven by perception and not by
facts or by what is understood as facts by risk analysts and scientists” (Renn,
2005, p. 31). Because of this reality, data on the actual gang membership status
of the offender is relatively unimportant when studying victim behavior, as it is
the victim’s perception of the offender’s status that drives their decision-
making and behavior.
Fortunately data are available that offer information on the victim’s percep-
tion of the offender’s gang status as well as numerous situational characteristics
of criminal incidents: the National Crime Victimization Survey (NCVS). These
data represent the only nationally representative data that include detailed
information on the victims, offenders, and situational features of non-fatal
violence in the USA, and with this information, a partial test of Felson’s
assertions regarding the functionality of gang membership is possible.
Specifically, we use 1992 to 2005 National Crime Victimization Survey (NCVS)
data to test our hypotheses. These data are publicly available through the
National Archive of Criminal Justice Data (NACJD) and are collected using a
stratified, multistage cluster (XE “NCVS Sampling: Stratified, multi-stage
cluster”) design (Hubble, 1995; Rennison & Rand, 2007). All persons age 12 or
older are interviewed every six months for a three-year period in sampled
households. On average, from 1992 to 2005, approximately 99,800 households
and 184,100 persons have been interviewed annually for the survey. Response
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
626 MELDE AND RENNISON
rates during this period range from 91% to 96% for households and from 84% to
94% for individuals, which is well above general recommendations for consent
rates needed to ensure low sample bias (Babbie, 1973; Lueptow, Mueller,
Hammes, & Master, 1977; Sewell & Hauser, 1975). Findings from the NCVS are
representative of the non-institutionalized population of persons age 12 or older
in the USA (Garofalo, 1990; Rennison & Rand, 2007).
The NCVS offers several advantages for the examination of gang violence.
First, the data include a measure of the perceived gang membership status of
the offender. Second, the data include theoretically relevant situational char-
acteristics of victimization (e.g., weapons used, injuries sustained). Third,
because the NCVS is a large, nationally representative, dataset, disaggregation
by multiple characteristics is feasible. Unlike official records, NCVS data can be
disaggregated by the gang member’s gender, race, and age simultaneously. This
is a particularly important advantage as there is a “crucial” need for offending
research using disaggregated data (Heimer, 2000, p. 433). And finally, NCVS
data offer information on violence regardless of whether the police were noti-
fied, thus avoiding some of the limitations of data associated with official
records.
NCVS data have some limitations pertinent to the present purposes. First, our
ability to model bystander intervention as a function of the gang membership
status of the offender is impaired by the inability to determine how the
bystander perceived the offender’s gang status. In the subsequent models, gang
membership of the offender is based on the victim’s perception. Because there
is no measure in the NCVS regarding the bystander’s perception of the
offender’s gang status, we use the victim’s perception as a proxy measure.
Therefore, all models using bystander intervention as the dependent variable
should be interpreted with this caveat in mind. A final limitation is that we are
not able to determine the precise time when the victim formed their conclusion
as to the gang membership status of the offender. It could be that this determi-
nation was completed well after the reported incident, and thus would have had
no effect on their behavior. While not perfect, the NCVS offers an important
opportunity to investigate Felson’s theory on the effects of gang membership on
the situational dynamics of violent crime.
The present analyses were conducted on a sample describing 29,511 threat-
ened, attempted, and completed violent victimizations between 1992 and
2005.
2
This total includes 2,216 victimizations committed by an offender
perceived to be a gang member.
Measures
Three hypotheses derived from Felson’s “street gang strategy” are tested. The
first hypothesis is that victim resistance is less likely when a violent crime is
2. Unweighted sample size.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 627
committed by a gang versus a non-gang member. Second, bystander interven-
tion is less likely when violent victimization is perpetrated by a gang versus a
non-gang member. And finally, police reporting of a violent crime is signifi-
cantly less likely when the offender is perceived to be a gang versus a non-gang
member. The measures used to test these hypotheses are briefly described
below. For greater detail regarding these measures, please see the Appendix.
Dependent variables
Three dependent variables are used in tests of Felson’s hypotheses: “victim
resistance,” “bystander intervention,” and “police reporting.” Victim resistance
is measured using three dichotomous variables distinguishing among forceful
resistance, non-forceful resistance, and no resistance by the victim (see
Baumer, Horney, et al., 2003; Block & Skogan, 1986). Bystander intervention is
measured using a dichotomous variable that indicates whether a bystander
intervened or failed to intervene during a crime incident.
3
Analyses using
bystander intervention as a dependent variable are restricted only to incidents
in which the victim reported a bystander was present. And finally, police
reporting is measured using a dichotomous variable distinguishing victimizations
reported from those not reported to the police.
Independent variables
The primary independent variable in this research is whether the offender was
perceived to be a gang member. In the NCVS, respondents reporting a violent
crime (rape, sexual assault, robbery, aggravated assault, and simple assault)
are asked: “Was the offender a member of a street gang, or don’t you know?”
Therefore, “gang status” is measured as a dichotomous variable where offend-
ers perceived as gang members are coded as “1” and those who were not are
coded as “0.”
4
While this variable does not allow us to identify the specific
signals used by the victim (i.e., operator) to identify the offender (i.e., mimic)
as a gang member, the crucial point in the process is that the operator identi-
fies the mimic as an example of the model. As Ihalainen, Lindstrom, & Mappes
3. Analyses using bystander intervention as a dependent variable are restricted to victimizations in
which a bystander was present. While this variable does not allow us to determine if the bystander
believed the perpetrator was a gang member, we include this variable in our study because it repre-
sents a key argument Felson makes in his theory of gang crime. Findings based on this measure of
bystander intervention, however, are merely speculative, and should be interpreted with caution.
4. Those who reported that they were unsure if the offender was a gang member or not were coded
zero in analyses reported in the current paper. We coded these incidents in this manner because it is
unlikely that victim or bystander behavior would be altered by the gang status of the offender
unless the victim was more certain of this possibility. We conducted supplementary analyses which
excluded those cases where the victim reported “don’t know” to the gang question, with substan-
tively identical results.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
628 MELDE AND RENNISON
(2007) highlight, mimicry does not have to be perfect to communicate the
potential for danger to the operator (i.e., victim), because it is to their advan-
tage to be cautious, and thus even imperfect mimicry is often accepted. If
Felson’s (2006) assertions regarding the street gang strategy are valid, and gang
membership is a form of mimicry, the offender (i.e., mimic) will benefit by the
operator (i.e., victim) identifying them as a gang member, no matter how this
identification takes place.
Situational control variables
In the total violence model, each of the four violent crime types is used as
controls: rape/sexual assault, robbery, aggravated assault, and simple assault.
Each category of crime includes threatened, attempted, and completed violence.
Injuries to the victim are accounted for using three dichotomous variables: no
injury (no injury), minor injury (minor), and serious injury (serious). Weapons are
controlled for with a series of dichotomous variables: no weapon, firearm, knife
or other cutting instrument (knife), and other weapon (other). Third party pres-
ence (bystander) is measured as a dichotomous variable (1 = bystander present
and 0 = no bystander present) as is the number of offenders (offenders) (0 = one
offender and 1 = two or more offenders).
Offender characteristic controls
Offender’s gender is measured using three dichotomous variables: male,
female, group/unknown. To account for the offender’s race, four dichotomous
variables are used: white, black, other, group/unknown.
5
Our models control
for offender’s age using these categories: less than 18, 19 to 29, 30 plus, and
group/unknown. And finally, four dichotomous variables are used to distinguish
among four victim and offender relationships: stranger, family member, friend,
and intimate.
Victim characteristic controls
Victim’s gender is coded zero for males and one for females (female). Victim’s
age (age) in years is a continuous measure that ranges from 12 to 90. The race
and Hispanic origin of the victim (race) is accounted for in the analyses using
four categories: non-Hispanic white (white), non-Hispanic black (black), non-
Hispanic other (other), and Hispanic of any race (Hispanic). The annual
5. Offender characteristics include “group/unknown” categories. While it would be preferable to
have these categories separated in the analysis, the low sample size in each of these categories
prohibitied this.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 629
household income of a victim (income) is measured using 14 income categories
of unequal width in the NCVS. As has been done in previous work using this
measure (see e.g., Baumer, 2002), this variable is treated as a continuous
measure. Marital status is accounted for using five dichotomous variables:
married, never married, widowed, divorced, and separated. A binary control
variable for employment status (employed) is included where 1 = employed and
0 = unemployed. Whether the victim lives in an owned or rented home (own) is
addressed in the models with a dichotomous variable in which 0 = rented and 1
= owned. And finally, three dichotomous variables are used to account for the
type of area where the victim lives: urban, suburban, and rural.
Analytic Strategy
The first dependent variable measures a victim’s resistance during a violent
victimization using three categories (force, non-force, and no resistance). Given
the polytomous nature of this variable, we use multinomial logistic regression to
estimate the influence of offender’s gang status (and the controls) on victim
resistance (Hosmer & Lemeshow, 2000). In the second and third models, the
dependent variables are dichotomous in nature (i.e., bystander intervention
and police reporting) and binary logistic regression models are used (Hosmer &
Lemeshow, 2000). Standard statistical packages do not take into account the
complex sample design of the NCVS so all regression models were estimated
using Stata’s survey regression procedure which accounts for complex sample
designs (StataCorp, 2005). Model fit statistics reported are based on parallel
regressions which do not account for sample design effects because such statis-
tics are not available using the survey regression procedures (Hosmer &
Lemeshow, 2000). All analyses utilize appropriate weights available on the file.
Our analyses proceed in the following fashion. We first estimate bivariate
models of victim resistance, bystander intervention, and police reporting for
total violence and each of the four violent crimes separately. We next estimate
the same series of models using relevant controls in a multivariate setting to
determine if any influences of perceived gang status observed in the bivariate
models are due to victim, offender, or incident-level factors associated with
gang status of the offender.
Results
Table 1 presents percentage distributions for the variables included in the anal-
yses. The percentages are consistent with previous findings based on NCVS data
(see e.g., Baumer, Horney, et al., 2003; Catalano, 2006; Hart & Rennison, 2003;
Melde & Rennison, 2008; Planty, 2002; Rennison, 2001; Skogan, 1977). Findings
show that 29% of victims used forceful resistance, 33% used non-forceful resis-
tance, and 39% offered no resistance during non-fatal violence occurring from
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
630 MELDE AND RENNISON
Table 1 Percentage distributions for variables included in analysis of the effects of gang violence on victim resistance, bystander intervention
and police reporting, 1992-2005 NCVS, n = 29,511
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 631
1992 to 2005. An examination of bystander intervention indicates that when
violence is committed with a bystander present, 52% of the victimizations
involved an intervening bystander, and 48% did not. And 45% of violence was
reported to the police.
Findings in Table 1 also demonstrate the demographic characteristics of
victims, offenders, and incidents, which are consistent with previous findings.
For example, most violence comes in the form of simple assault (65%) while
rape/sexual assault accounts for the smallest percentage of non-fatal violence
(4%). In most victimizations, no injuries were sustained (74%), no weapons were
used (67%), offenders were primarily male (80%), white (58%),
6
and strangers
(47%). Most victims were male (56%), white (72%), and never married (57%). We
now turn to the primary focus of the paper which is the influence of the
offender’s gang status on victim resistance, bystander intervention, and police
reporting.
Bivariate Results
Table 2 presents bivariate effects of the offender’s perceived gang status on
victim resistance (Panel A), bystander intervention (Panel B), and police report-
ing (Panel C) for total violence as well as the four disaggregated types of violent
crime. Our first hypothesis derived from Felson (2006) is that by using intimida-
tion via mimicry, violence committed by gang members will involve significantly
less victim resistance than violence committed by a non-gang offender.
Contrary to this hypothesis, findings in Panel A indicate that victims are no less
likely to resist when the offender is thought to be a gang member.
Our second hypothesis suggests that gang member violence will involve signif-
icantly less bystander intervention than victimization committed by non-gang
offenders. Bivariate results of this test are presented in Panel B and contrary to
expectations, among victimizations involving a bystander, when the offender is
perceived to be a gang member bystanders are more likely to intervene in over-
all violence and during simple assaults. Further, perceived gang membership of
the offender did not influence bystander intervention for rape/sexual assault,
robbery, or aggravated assault.
Panel C offers findings for a bivariate test of the influence of perceived gang
status on police reporting. These findings offer mixed support for our hypothe-
sis. For total violence, when the offender is believed to be a gang member,
there is a greater likelihood of police reporting. However, an investigation of
robbery demonstrates that when the offender is thought to be a gang member,
the police are less likely to be contacted. Finally, perceived gang membership
did not influence police reporting for rape/sexual assault, simple assault, or
aggravated assault.
6. While most offenders were white, black offenders (25%) were disproportionately represented
among offenders relative to their representation in the US general population.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
632 MELDE AND RENNISON
Table 2 Bivariate association between gang status of offender and victim resistance, bystander intervention and police reporting by type of
violence
Panel A victim resistance—multinomial regression
Total violence Rape/sexual assault Robbery Aggravated assault Simple assault
Forceful
resistance
Non-forceful
resistance
Forceful
resistance
Non-forceful
resistance
Forceful
resistance
Non-forceful
resistance
Forceful
resistance
Non-forceful
resistance
Forceful
resistance
Non-forceful
resistance
Gang membership 0.08
(0.06)
0.03
(0.07)
0.18
(0.51)
0.14
(0.48)
0.22
(0.15)
0.28
0.19
0.03
(0.10)
0.18
(0.11)
0.08
(0.09)
0.07
(0.09)
Intercept 0.30*
(0.02)
0.17
(0.02)
0.06
(0.08)
0.05
(0.09)
0.35*
(0.05)
0.05*
(0.05)
0.14
(0.04)
0.06*
(0.04)
0.35*
(0.02)
0.19*
(0.02)
Model chi-square 9.43*
n = 29,511
0.21
n = 1,208
10.36*
n = 3,156
3.59
n = 5,989
2.50
n = 19,158
Panel B bystander intervention—binomial regression Panel C police reporting—binomial regression
Total
violence
Rape/sexual
assault Robbery
Aggravated
assault
Simple
assault
Total
violence
Rape/sexual
assault Robbery
Aggravated
assault
Simple
assault
Gang membership 0.20*
(0.06)
0.35
(0.52)
0.18
(0.15)
0.11
(0.11)
0.25*
(0.08)
0.14*
(0.06)
0.23
(0.34)
0.55*
(0.12)
0.05
(0.09)
0.15
(0.07)
Intercept 0.01
(0.02)
0.05
(0.12)
0.21*
(0.06)
0.11*
(0.04)
0.04
(0.02)
0.23*
(0.02)
0.65*
(0.07)
0.36*
(0.05)
0.26*
(0.03)
0.44*
(0.02)
Model chi-square 19.71*
n = 19,668
0.067
n = 352
2.41
n = 1,665
3.55
n = 4,259
14.12*
n = 13,392
9.43*
n = 29,511
0.74
n = 1,208
17.92*
n = 3,156
0.46
n = 5,989
3.34
n = 19,158
*p 0.5, two-tailed test.
Note: Standard errors in parentheses.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 633
In sum, the bivariate patterns presented in Table 2 do little to support
hypotheses derived from Felson’s (2006) theory on the functionality of gang
membership. These bivariate models, however, are poorly suited for evaluating
the hypotheses derived from Felson’s (2006) theory. First, they fail to utilize
one of the major advantages of the NCVS—numerous situational characteristics—
to test the relationship between gang status and the outcome variables net
offender, incident, and victim characteristics. Literature clearly identifies many
correlates related to the outcome variables including the number of offenders
(Ullman, 2005), weapon presence (Hart & Miethe, 2008; Marshall & Webb,
1992), victim’s gender and race (Baumer, 2002; Hart & Rennison, 2003; Marshall
& Webb, 1992), victim and offender relationship (Bachman, 1998; Felson et al.,
2000; Laub, 1997; Shotland & Straw, 1976; Skogan, 1984), and victim injury (see
e.g., Gottfredson & Gottfredson, 1980; Gottfredson & Hindelang, 1979; Laub,
1981; Skogan, 1976, 1984). Failure to account for these correlates may act to
suppress the relationships suggested by Felson (2006). Therefore, in the next
section we estimate the effects of gang status on victim resistance, bystander
intervention, and police reporting while controlling for situational, offender,
and victim characteristics.
Multivariate Results
Table 3 presents findings from a series of multinomial regression models that
evaluate whether the effects of gang status on victim resistance persist after
controlling for victim, offender, and incident characteristics. A total crime
model as well as a model for each of the four disaggregated crime types is
presented. In general, findings offer limited support of the victim resistance
hypothesis. In support of the victim resistance hypothesis is the finding that
violence committed by a perceived gang member decreases the odds of forceful
victim resistance during a simple assault (AOR = 0.80). In contrast to expecta-
tions, however, aggravated assault committed by a perceived gang member
increases the odds of forceful victim resistance (AOR = 1.37). Results from the
other models show that the perceived gang status of the offender is unrelated
to non-forceful resistance regardless of the type of crime, and is unrelated to
forceful victim resistance for total violence, rape/sexual assault, and robbery.
Our second hypothesis posits that gang violence is associated with a reduced
likelihood of bystander intervention. Multivariate regression results from models
for each of the five crime categories are presented in Table 4. Findings fail to
support this hypothesis. In fact, several findings are contrary to expectations.
When the offender is thought to be a gang member, the odds of bystander inter-
vention increase for total violence (AOR = 1.23), aggravated assault (AOR = 1.25),
and simple assault (AOR = 1.27). An examination of rape/sexual assault and
robbery reveals no influence of perceived gang status on bystander intervention.
Findings from a set of multivariate models that investigate the influence of
perceived gang membership of the offender on police reporting are provided in
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
634 MELDE AND RENNISON
Table 3 Multinomial regression coefficients of victim resistance by type of crime
a. Total violence Rape/sexual assault
Forceful resistance Non-forceful resistance Forceful resistance Non-forceful resistance
Coef SE AOR Coef/SE AOR Coef/SE AOR Coef/SE AOR
Gang membership 0.01
0.07
1.01 0.09
0.07
1.09 0.35
0.70
0.70 0.26
0.77
0.77
Situational characteristics
Rape/sexual assault 0.67*
0.09
1.96 0.30*
0.10
1.35
Robbery 0.33*
0.06
1.40 0.20*
0.07
0.82
Aggravated assault 0.65*
0.07
1.91 0.41*
0.07
1.51
Serious injury 0.31*
0.08
1.36 1.16*
0.10
0.31 0.84*
0.43
0.43 1.22*
0.30
0.30
Minor injury 1.05*
0.04
2.87 0.53*
0.05
0.59 0.26
1.30
1.30 1.02*
0.36
0.36
Number of offenders 0.09
0.06
1.09 0.24*
0.06
1.28 0.93*
0.39
0.39 0.11
0.90
0.90
Firearm 1.32*
0.08
0.27 0.36*
0.08
0.70 0.84*
0.43
0.43 1.12
0.33
0.33
Knife 0.03
0.08
0.97 0.02
0.08
0.98 0.12
0.88
0.88 0.56
0.57
0.57
Other 0.39*
0.06
0.67 0.14*
0.05
1.15 0.41
0.66
0.66 0.25
1.28
1.28
Bystander 0.26*
0.04
1.30 0.01
0.04
1.01 0.05
0.95
0.95 0.01
1.01
1.01
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 635
Table 3 (Continued)
a. Total violence Rape/sexual assault
Forceful resistance Non-forceful resistance Forceful resistance Non-forceful resistance
Coef SE AOR Coef/SE AOR Coef/SE AOR Coef/SE AOR
Offender Characteristics
Female 0.04
0.05
0.96 0.22*
0.05
0.81 0.10
1.10
1.10 0.39
1.47
1.47
Group 0.21*
0.10
0.81 0.08
0.09
0.92 0.86
0.42
0.42 0.51
1.66
1.66
Black 0.09*
0.05
0.91 0.04
0.05
0.96 0.12
0.89
0.89 0.06
1.07
1.07
Other 0.05
0.06
1.05 0.00
0.06
1.00 0.08
1.08
1.08 0.07
1.07
1.07
Group 0.19*
0.07
0.83 0.21*
0.07
0.81 0.07
0.93
0.93 0.20
1.22
1.22
19 to 29 0.05
0.05
1.05 0.17*
0.05
1.18 0.05
1.05
1.05 0.00
1.00
1.00
30 Plus 0.18*
0.06
0.83 0.23*
0.05
1.26 0.45
0.64
0.64 0.17
0.84
0.84
Group 0.14*
0.07
0.87 0.09
0.07
1.10 0.04
0.96
0.96 0.68
0.51
0.51
Intimate 0.16*
0.06
1.17 0.11
0.06
1.12 0.42
0.66
0.66 0.74*
0.48
0.48
Family 0.23*
0.08
1.26 0.03
0.08
0.97 0.04
1.04
1.04 0.30
0.74
0.74
Friends 0.01
0.04
1.01 0.10*
0.04
0.91 0.15
0.86
0.86 0.31
0.73
0.73
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
636 MELDE AND RENNISON
Table 3 (Continued)
a. Total violence Rape/sexual assault
Forceful resistance Non-forceful resistance Forceful resistance Non-forceful resistance
Coef SE AOR Coef/SE AOR Coef/SE AOR Coef/SE AOR
Victim Characteristics
Female 0.50*
0.04
0.61 0.32*
0.04
1.38 1.01*
2.76
2.76 1.11*
3.04
3.04
Age 0.01*
0.00
0.99 0.00
0.00
1.00 0.01
0.99
0.99 0.00
1.00
1.00
Non-Hispanic black 0.24*
0.06
1.27 0.07
0.06
0.94 0.74*
2.10
2.10 0.07
0.94
0.94
Non-Hispanic other 0.16
0.09
0.85 0.15
0.09
0.86 0.73
2.07
2.07 0.50
1.64
1.64
Hispanic, any race 0.05
0.06
1.05 0.11
0.06
0.90 0.30
1.35
1.35 0.08
1.08
1.08
Annual household income 0.02*
0.00
0.98 0.01
0.01
1.01 0.01
0.99
0.99 0.01
1.01
1.01
Never married 0.07
0.05
0.94 0.17*
0.04
0.85 0.11
0.89
0.89 0.22
0.80
0.80
Widowed 0.08
0.15
0.92 00.20
0.13
0.82 0.55
1.74
1.74 0.49
1.64
1.64
Divorced 0.12
0.07
1.13 0.03
0.06
0.97 0.19
1.21
1.21 0.15
0.86
0.86
Separated 0.10
0.09
0.90 0.20*
0.09
0.82 0.35
1.43
1.43 0.43
1.54
1.54
Employed 0.21*
0.04
1.23 0.13*
0.04
1.14 0.11
0.89
0.89 0.07
0.93
0.93
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 637
Table 3 (Continued)
a. Total violence Rape/sexual assault
Forceful resistance Non-forceful resistance Forceful resistance Non-forceful resistance
Coef SE AOR Coef/SE AOR Coef/SE AOR Coef/SE AOR
Owned 0.00
0.04
1.00 0.04
0.03
0.96 0.20
1.22
1.22 0.21
0.81
0.81
Suburban 0.03
0.04
1.03 0.00
0.04
1.00 0.15
1.17
1.17 0.20
1.22
1.22
Rural 0.03
0.06
1.03 0.04
0.06
0.96 0.19
0.83
0.83 0.25
1.29
1.29
Intercept 0.32
0.12
0.60*
0.12
0.55 0.98
0.77
2.66 0.02
0.80
1.02
Model chi-square 4006.85*
n = 29,511
210.35*
n = 1,208
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
638 MELDE AND RENNISON
Table 3 (Continued)
b. Robbery Aggravated assault
Forceful resistance Non-forceful resistance Forceful resistance Non-forceful resistance
Coef/SE AOR Coef/SE AOR Coef/SE AOR Coef/SE AOR
Gang membership 0.24
0.17
1.27 0.13
0.19
0.88 0.32*
0.11
1.37 0.08
0.10
1.08
Situational characteristics t
Rape/sexual assault
Robbery
Aggravated assault
Serious injury 0.74*
0.19
2.10 0.65*
0.25
0.51 0.65*
0.13
1.91 1.20*
0.17
0.30
Minor injury 0.83*
0.12
2.29 0.49*
0.13
0.61 0.92*
0.10
2.51 0.76*
0.11
0.47
Number of offenders 0.34*
0.12
0.71 0.25*
0.13
1.28 0.33*
0.11
1.39 0.44*
0.10
1.56
Firearm 1.57*
0.16
0.21 0.73*
0.15
0.48 0.65*
0.21
0.52 0.25
0.26
1.29
Knife 0.03
0.15
0.97 0.15
0.17
0.86 0.46*
0.19
1.59 0.41
0.26
1.50
Other 0.29*
0.13
0.75 0.16
0.14
1.17 0.10
0.20
1.11 0.49
0.26
1.62
Bystander 0.21*
0.10
1.24 0.04
0.12
1.04 0.19*
0.09
1.21 0.21*
0.08
0.81
Offender characteristics
Female 0.22
0.20
1.25 0.00
0.22
1.00 0.07
0.12
1.07 0.17
0.13
0.85
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 639
Table 3 (Continued)
b. Robbery Aggravated assault
Forceful resistance Non-forceful resistance Forceful resistance Non-forceful resistance
Coef/SE AOR Coef/SE AOR Coef/SE AOR Coef/SE AOR
Group 0.09
0.25
0.92 0.35
0.26
0.70 0.32
0.20
0.73 0.20
0.17
0.82
Black 0.15
0.12
0.86 0.11
0.13
0.89 0.09
0.11
0.91 0.05
0.10
0.90
Other 0.15
0.17
1.16 0.15
0.18
0.86 0.10
0.12
0.91 0.05
0.10
0.95
Group 0.04
0.17
0.96 0.34
0.22
0.71 0.31*
0.16
0.73 0.30*
0.14
0.74
19 to 29 0.08
0.16
1.09 0.04
0.17
0.96 0.22*
0.11
1.25 0.27*
0.11
1.31
30 Plus 0.32
0.19
0.72 0.11
0.19
0.90 0.10
0.11
1.10 0.31*
0.11
1.36
Group 0.08
0.17
0.92 0.05
0.19
0.95 0.05
0.13
0.95 0.09
0.13
1.09
Intimate 0.28
0.22
1.33 0.98*
0.25
2.67 0.38*
0.16
1.46 0.21
0.15
1.23
Family 0.15
0.30
0.86 0.55
0.37
1.73 0.46*
0.17
1.59 0.17
0.17
1.18
Friends 0.16
0.13
0.85 0.17
0.14
1.19 0.27*
0.09
1.32 0.13
0.08
1.13
Victim characteristics
Female .037*
0.11
0.69 0.04
0.12
1.05 0.66*
0.09
0.52 0.13
0.08
1.14
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
640 MELDE AND RENNISON
Table 3 (Continued)
b. Robbery Aggravated assault
Forceful resistance Non-forceful resistance Forceful resistance Non-forceful resistance
Coef/SE AOR Coef/SE AOR Coef/SE AOR Coef/SE AOR
Age 0.00
0.00
1.00 0.00
0.00
1.00 0.01
0.00
0.99 0.01
0.00
0.99
Non-Hispanic olack 0.24
0.15
0.78 0.35*
0.14
0.71 0.07
0.13
1.07 0.11
0.12
0.90
Non-Hispanic other 0.53*
0.25
0.59 0.06
0.22
1.06 0.22
0.20
0.80 0.29
0.19
0.75
Hispanic, any race 0.40*
0.14
0.67 0.56*
0.16
0.57 0.01
0.11
0.99 0.06
0.10
0.95
Annual household income 0.00
0.01
1.00 0.02
0.01
1.02 0.02
0.01
0.99 0.00
0.01
1.00
Never married 0.14
0.14
1.15 0.07
0.15
1.07 0.29*
0.11
0.75 0.37*
0.10
0.69
Widowed 0.19
0.31
0.83 0.21
0.36
0.81 0.54
0.36
0.58 0.73*
0.27
0.48
Divorced 0.06
0.16
0.95 0.10
0.17
0.91 0.08
0.12
1.09 0.04
0.13
0.96
Separated 0.28
0.29
0.76 0.43
0.28
0.65 0.47*
0.19
0.62 0.22
0.18
0.80
Employed 0.28*
0.11
1.33 0.27*
0.11
1.31 0.26*
0.09
1.29 0.05
0.08
1.05
Owned 0.14
0.11
1.15 0.01
0.11
1.01 0.08
0.09
0.92 0.04
0.08
0.96
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 641
Table 3 (Continued)
b. Robbery Aggravated assault
Forceful resistance Non-forceful resistance Forceful resistance Non-forceful resistance
Coef/SE AOR Coef/SE AOR Coef/SE AOR Coef/SE AOR
Suburban 0.22*
0.11
1.25 0.04
0.10
1.04 0.05
0.08
1.05 0.06
0.08
1.06
Rural 0.13
0.20
1.14 0.05
0.20
0.95 0.01
0.12
1.01 0.16
0.12
0.85
Intercept 0.21
0.34
1.23 0.64
0.34
0.53 0.69
0.33
0.50 0.37
0.35
0.69
Model chi-square 596.59*
n = 3,156
1060.19*
n = 5,989
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
642 MELDE AND RENNISON
Table 3 (Continued)
c. Simple assault
Forceful resistance Non-forceful resistance
Coef/SE AOR Coef/SE AOR
Gang membership 0.23 *
0.10
0.80 0.12
0.09
1.13
Situational characteristics
Rape/sexual assault
Robbery
Aggravated assault
Serious injury 29.94 *
0.61
0.00 29.94 *
0.68
0.00
Minor injury 1.16*
0.05
3.20 0.45 *
0.06
0.64
Number of offenders 0.18*
0.08
1.20 0.25*
0.07
1.28
Firearm
Knife
Other
Bystander 0.32*
0.05
1.38 0.06
0.04
1.06
Offender characteristics
Female 0.04
0.05
0.96 0.25 *
0.05
0.78
Group 0.23
0.13
0.80 0.03
0.11
0.97
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 643
Table 3 (Continued)
c. Simple assault
Forceful resistance Non-forceful resistance
Coef/SE AOR Coef/SE AOR
Black 0.11
0.06
0.90 0.00
0.06
1.00
Other 0.06
0.07
1.06 0.02
0.08
1.02
Group 0.22 *
0.10
0.80 0.19*
0.08
0.83
19 to 29 0.03
0.06
0.97 0.16*
0.06
1.17
30 Plus 0.26 *
0.07
0.77 0.22*
0.06
1.25
Group 0.25 *
0.10
0.78 0.10
0.09
1.11
Intimate 0.19*
0.08
1.21 0.01
0.07
1.01
Family 0.24*
0.10
1.28 0.20 *
0.09
0.82
Friends 0.02
0.05
0.98 0.20 *
0.04
0.82
Victim characteristics
Female 0.52*
0.05
0.59 0.40*
0.05
1.50
Age 0.01*
0.00
0.99 0.00
0.00
1.00
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
644 MELDE AND RENNISON
Table 3 (Continued)
c. Simple assault
Forceful resistance Non-forceful resistance
Coef/SE AOR Coef/SE AOR
Non-Hispanic black 0.40*
0.08
1.49 0.05
0.09
1.05
Non-Hispanic other 0.10
0.12
0.91 0.19
0.12
0.83
Hispanic, any race 0.16*
0.08
1.18 0.03
0.08
0.98
Annual household income 0.02*
0.01
0.98 0.00
0.01
1.00
Never married 0.06
0.06
0.94 0.14*
0.06
0.87
Widowed 0.04
0.21
0.96 0.05
0.17
0.95
Divorced 0.13
0.09
1.14 0.02
0.07
0.98
Separated 0.01
0.12
0.99 0.21*
0.11
0.81
Employed 0.22*
0.05
1.25 0.15*
0.05
1.16
Owned 0.00
0.05
1.00 0.04
0.04
0.96
Suburban 0.02
0.05
0.98 0.04
0.05
0.96
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 645
Table 3 (Continued)
c. Simple assault
Forceful resistance Non-forceful resistance
Coef/SE AOR Coef/SE AOR
Rural 0.02
0.08
1.02 0.03
0.06
0.97
Intercept 0.33
0.16
0.72 0.73*
0.14
0.48
Model chi-square 2470.81*
n = 19,158
*p .05, two-tailed test.
Note: Standard errors are located below their respective regression coefficients.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
646 MELDE AND RENNISON
Table 4 Binomial regression coefficients of bystander intervention by type of crime
a. Total violence Rape/sexual assault Robbery
Coef/SE AOR Coef/SE AOR Coef/SE AOR
Gang membership 0.21*
0.07
1.23 0.53
0.49
0.59 0.21
0.16
1.24
Situational Characteristics
Rape/sexual assault 0.00
0.12
1.00
Robbery 0.12
0.06
0.88
Aggravated assault 0.31*
0.07
1.37
Serious injury 0.08
0.09
1.08 0.20
0.34
1.22 0.39
0.21
1.48
Minor injury 0.23*
0.04
1.26 0.47
0.34
1.59 0.30*
0.14
1.35
Number of offenders 0.25*
0.05
1.28 0.09
0.46
1.10 0.28*
0.14
1.32
Firearm 0.53*
0.08
0.59 0.52
0.69
1.68 0.51*
0.16
0.60
Knife 0.05
0.09
0.95 0.86
0.99
2.36 0.06
0.20
0.94
Other 0.21*
0.06
0.81 0.58
0.39
0.56 0.09
0.15
0.92
Offender characteristics
Female 0.08
0.05
1.09 0.76
0.73
0.47 0.17
0.18
1.18
Group 0.01
0.09
1.01 0.60
0.73
0.55 0.00
0.26
1.00
Black 0.00
0.05
1.00 0.09
0.35
0.92 0.23
0.14
0.79
Other 0.04
0.06
1.04 0.16
0.39
1.17 0.01
0.19
1.01
Group 0.04
0.07
1.04 0.74
0.60
0.48 0.28
0.21
0.76
19 to 29 0.31*
0.05
1.37 0.35
0.37
0.71 0.06
0.16
1.07
30 Plus 0.24*
0.05
1.27 0.20
0.43
1.22 0.02
0.18
0.98
Group 0.10
0.07
1.10 0.62
0.51
1.86 0.01
0.19
0.99
Intimate 0.09
0.08
0.91 0.26
0.45
0.77 0.37
0.30
1.44
Family 0.01
0.08
0.99 0.47
0.83
1.61 0.18
0.30
1.19
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 647
Table 4 (Continued)
a. Total violence Rape/sexual assault Robbery
Coef/SE AOR Coef/SE AOR Coef/SE AOR
Friends 0.05
0.04
0.95 0.11
0.29
0.89 0.27
0.15
0.77
Victim characteristics
Female 0.10*
0.04
1.10 0.02
0.40
0.98 0.01
0.12
1.01
Age 0.00*
0.00
1.00 0.01
0.02
0.99 0.00
0.01
1.00
Non-Hispanic black 0.08
0.07
1.08 0.74
0.44
2.10 0.28
0.16
1.32
Non-Hispanic other 0.07
0.10
1.07 0.31
1.09
0.73 0.16
0.30
1.18
Hispanic, any race 0.03
0.06
0.97 0.50
0.35
1.65 0.25
0.18
0.78
Annual household income 0.01*
0.00
1.02 0.03
0.04
1.03 0.02
0.02
1.02
Never married 0.01
0.05
0.99 0.09
0.39
1.10 0.11
0.17
1.12
Widowed 0.09
0.17
1.09 0.94
1.04
2.57 0.11
0.40
1.11
Divorced 0.07
0.06
1.08 0.04
0.46
0.97 0.17
0.21
1.19
Separated 0.04
0.10
0.96 0.90
0.60
2.46 0.18
0.28
1.19
Employed 0.11*
0.04
1.11 0.43
0.28
1.53 0.15
0.12
1.16
Owned 0.02
0.04
1.02 0.23
0.34
1.26 0.08
0.12
0.92
Suburban 0.04
0.04
1.04 0.14
0.30
1.15 0.14
0.11
1.16
Rural 0.05
0.07
0.95 0.49
0.42
1.62 0.25
0.22
1.29
Intercept 0.67*
0.12
0.51 0.71
1.07
0.49 0.80
0.40
0.45
Model chi-square 306.68*
n = 19,668
34.71
n = 352
64.98*
n = 1,665
b. Aggravated assault Simple assault
Coef/SE AOR Coef/SE AOR
Gang membership 0.22*
0.11
1.25 0.24*
0.08
1.27
Situational characteristics
Rape/sexual assault
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
648 MELDE AND RENNISON
Table 4 (Continued)
b. Aggravated assault Simple assault
Coef/SE AOR Coef/SE AOR
Robbery
Aggravated assault
Serious injury 0.02
0.12
1.02 0.48
1.24
1.61
Minor injury 0.20*
0.09
1.22 0.23*
0.05
1.26
Number of offenders 0.31*
0.09
1.36 0.22*
0.07
1.25
Firearm 0.41*
0.20
0.66
Knife 0.01
0.20
1.01
Other 0.15
0.20
0.86
Offender characteristics
Female 0.21
0.12
1.24 0.05
0.05
1.05
Group 0.07
0.16
1.07 0.00
0.11
1.00
Black 0.01
0.10
0.99 0.04
0.06
1.04
Other 0.18
0.12
0.84 0.11
0.07
1.12
Group 0.03
0.14
0.97 0.10
0.08
1.10
19 to 29 0.25*
0.11
1.29 0.37*
0.06
1.45
30 Plus 0.26*
0.11
1.30 0.25*
0.06
1.28
Group 0.00
0.13
1.00 0.11
0.09
1.12
Intimate 0.24
0.17
0.78 0.11
0.10
0.90
Family 0.13
0.15
1.14 0.08
0.09
0.93
Friends 0.08
0.08
1.09 0.08
0.04
0.93
Victim characteristics
Female 0.02
0.07
1.02 0.14*
0.05
1.15
Age 0.00
0.00
1.00 0.01*
0.00
0.99
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 649
Table 5. It was hypothesized that violence by perceived gang members would be
associated with reduced police reporting. Findings offer mixed support for this
hypothesis. In support of the hypothesis is the finding that robbery committed
by a perceived gang member significantly reduces the odds of police reporting
(AOR = 0.68). In contrast, the odds of police reporting are greater when a
perceived gang member commits total violence (AOR = 1.15) or simple assault
(AOR = 1.38). The perceived gang membership of the offender did not influence
police reporting for rape/sexual assault or aggravated assault even when rele-
vant controls were included in the models.
Our research also investigated whether the effects of offender’s gang statuson
victim resistance, bystander intervention and police reporting were conditioned
Table 4 (Continued)
b. Aggravated assault Simple assault
Coef/SE AOR Coef/SE AOR
Non-Hispanic black 0.01
0.12
1.01 0.04
0.08
1.04
Non-Hispanic other 0.18
0.19
1.19 0.01
0.13
1.01
Hispanic, any race 0.12
0.11
0.88 0.05
0.07
1.05
Annual household income 0.01
0.01
1.01 0.02*
0.01
1.02
Never married 0.02
0.10
0.98 0.01
0.06
0.99
Widowed 0.11
0.34
0.90 0.14
0.23
1.15
Divorced 0.21
0.13
1.23 0.02
0.07
1.02
Separated 0.33
0.18
1.39 0.25*
0.12
0.78
Employed 0.01
0.08
0.99 0.13*
0.04
1.14
Owned 0.03
0.08
1.04 0.02
0.05
1.02
Suburban 0.02
0.08
0.98 0.04
0.05
1.04
Rural 0.04
0.12
0.96 0.08
0.08
0.92
Intercept 0.41
0.30
0.66 0.70*
0.15
0.50
Model chi-square 82.30*
n = 4,259
189*
n = 13,392
*p .05, two-tailed test.
Note: Standard errors are located below their respective regression coefficients.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
650 MELDE AND RENNISON
Table 5 Binomial regression coefficients of police reporting by type of crime
Total violence Rape/sexual assault Robbery Aggravated assault Simple assault
Coef/SE AOR Coef/SE AOR Coef/SE AOR Coef/SE AOR Coef/SE AOR
Gang membership 0.14*
0.06
1.15 0.01
0.37
0.99 0.39*
0.13
0.68 0.15
0.09
1.16 0.32*
0.08
1.38
Situational Characteristics
Rape/sexual assault 0.68*
0.08
0.51
Robbery 0.41*
0.06
1.51
Aggravated assault 0.28*
0.05
1.32
Serious injury 1.13*
0.08
3.10 0.36
0.20
1.43 1.90*
0.21
6.69 1.24*
0.13
3.44
Minor injury 0.70*
0.03
2.02 0.82*
0.21
2.27 0.75*
0.10
2.11 0.40*
0.09
1.49 0.77*
0.04
2.16
Number of offenders 0.27*
0.04
1.31 0.50
0.39
0.61 0.19*
0.11
1.21 0.29*
0.09
1.34 0.33*
0.06
1.39
Firearm 0.76*
0.07
2.13 1.87*
0.41
6.51 1.11
0.12
3.05 0.48*
0.20
1.62
Knife 0.21*
0.07
1.24 1.43*
0.44
4.19 0.04*
0.13
0.96 0.10
0.20
1.11
Other 0.11*
0.05
1.11 0.37
0.25
1.45 0.22
0.12
1.24 0.04
0.19
0.96
Bystander 0.34*
0.03
1.40 0.16
0.16
1.17 0.18
0.09
1.20 0.54*
0.07
1.72 0.33*
0.04
1.39
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 651
Table 5 (Continued)
Total violence Rape/sexual assault Robbery Aggravated assault Simple assault
Coef/SE AOR Coef/SE AOR Coef/SE AOR Coef/SE AOR Coef/SE AOR
Offender Characteristics
Female 0.03
0.04
0.97 0.75
0.61
0.47 0.19
0.17
0.83 0.08
0.09
0.92 0.03
0.05
1.03
Group 0.42*
0.07
1.52 0.72
0.62
2.06 0.24
0.19
1.27 0.56*
0.15
1.75 0.40*
0.10
1.49
Black 0.04
0.04
1.04 0.49*
0.21
1.63 0.15
0.12
1.16 0.12
0.09
1.13 0.03
0.05
0.97
Other 0.07
0.05
0.93 0.07
0.23
1.07 0.15
0.14
1.16 0.13
0.10
0.87 0.09
0.06
0.91
Group 0.19*
0.07
0.83 0.34
0.46
0.71 0.14
0.17
0.87 0.15
0.12
0.86 0.21*
0.09
0.81
19 to 29 0.60*
0.05
1.81 0.10
0.27
1.10 0.56*
0.14
1.75 0.50*
0.10
1.66 0.62*
0.06
1.86
30 Plus 0.70*
0.06
2.02 0.37
0.30
1.44 0.55*
0.16
1.73 0.71*
0.11
2.03 0.72*
0.07
2.05
Group 0.58*
0.06
1.79 0.74
0.41
2.09 0.49*
0.17
1.64 0.47*
0.12
1.60 0.62*
0.08
1.86
Intimate 0.16*
0.06
1.18 0.87*
0.23
0.42 0.62*
0.18
1.86 0.07
0.13
1.07 0.24*
0.07
1.27
Family 0.14
0.07
1.15 0.20
0.40
1.23 0.17
0.33
1.19 0.04
0.16
1.04 0.13
0.08
1.14
Friends 0.18*
0.03
0.84 0.52*
0.17
0.59 0.19
0.12
1.21 0.21*
0.06
0.81 0.20*
0.04
0.82
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
652 MELDE AND RENNISON
Table 5 (Continued)
Total violence Rape/sexual assault Robbery Aggravated assault Simple assault
Coef/SE AOR Coef/SE AOR Coef/SE AOR Coef/SE AOR Coef/SE AOR
Victim Characteristics
Female 0.34*
0.03
1.41 0.00
0.33
1.00 0.65*
0.09
1.92 0.35*
0.07
1.42 0.27*
0.04
1.31
Age 0.01*
0.00
1.01 0.02
0.01
0.98 0.01*
0.00
1.01 0.01*
0.00
1.01 0.01*
0.00
1.01
Non-Hispanic black 0.15*
0.05
1.16 0.51*
0.25
0.60 0.12
0.14
0.89 0.23*
0.11
1.26 0.19*
0.07
1.21
Non-Hispanic other 0.04
0.08
0.97 0.41
0.41
0.66 0.07
0.18
0.93 0.18
0.16
0.83 0.01
0.11
1.01
Hispanic, any race 0.08
0.05
1.08 0.25
0.23
0.78 0.31*
0.12
0.73 0.14
0.10
1.15 0.16*
0.06
1.17
Annual household income 0.02*
0.00
0.98 0.02
0.02
0.98 0.01
0.01
1.02 0.01
0.01
0.99 0.02*
0.01
0.98
Never married 0.38*
0.04
0.68 0.45*
0.22
0.64 0.34*
0.13
0.71 0.54*
0.09
0.58 0.35*
0.05
0.70
Widowed 0.17
0.13
0.84 1.01
0.73
2.74 0.06
0.30
0.95 0.00
0.31
1.00 0.34
0.18
0.71
Divorced 0.08
0.05
0.92 0.11
0.28
1.12 0.21
0.15
0.81 0.21*
0.11
0.81 0.05
0.07
0.95
Separated 0.20*
0.08
0.82 0.60
0.35
0.55 0.52*
0.24
0.60 0.30
0.16
0.74 0.11
0.09
0.90
Employed 0.06
0.04
1.06 0.61*
0.16
0.55 0.14
0.09
1.15 0.11
0.07
1.12 0.07
0.05
1.07
Owned 0.05
0.03
1.05 0.29
0.17
1.33 0.37*
0.11
1.45 0.06
0.07
1.06 0.00
0.04
1.00
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 653
Table 5 (Continued)
Total violence Rape/sexual assault Robbery Aggravated assault Simple assault
Coef/SE AOR Coef/SE AOR Coef/SE AOR Coef/SE AOR Coef/SE AOR
Owned 0.05
0.03
1.05 0.29
0.17
1.33 0.37*
0.11
1.45 0.06
0.07
1.06 0.00
0.04
1.00
Suburban 0.10*
0.04
1.10 0.02
0.18
1.02 0.16
0.0
0.86 0.14*
0.07
1.15 0.11*
0.04
1.11
Rural 0.14*
0.05
1.15 0.11
0.22
1.12 0.06
0.21
1.06 0.12
0.10
1.13 0.15*
0.06
1.16
Intercept 1.88*
0.10
0.15 0.69
0.73
1.99 1.77*
0.29
0.17 1.42*
0.28
0.24 1.94*
0.12
0.14
Model chi-square 3604.19*
n = 29,511
164.61*
n = 1,208
525.18*
n = 3,156
683.88*
n = 5,989
1768.84*
n = 19,158
*p .05, two-tailed test.
Note: Standard errors are located below their respective regression coefficients.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
654 MELDE AND RENNISON
by other factors. Given findings in the empirical literature, it seemed plausible
that support for hypotheses derived from Felson’s (2006) theory may be apparent
only among particular victims, offenders, or crime contexts. For instance,
perhaps only male victims would be more likely to forcefully resist when the
offender is thought to be a gang member. Or perhaps reporting is diminished
when the victim and offender were known to one another and the likelihood of
retaliation is greater. To evaluate these possibilities, the models shown in Tables
3-5 were re-estimated with product terms representing two-way interactions
between potential moderator variables and the focus variable, perceived gang
membership. Potential moderator variables included victim’s gender, victim’s
race, victim and offender relationship, and offender’s race.
Using the hierarchical testing technique discussed in Jaccard (2001) two
significant moderators were found.
7
First, victim and offender relationship
moderates the influence of gang status on police reporting for the total violence
model. Findings from models with these interaction terms demonstrate no
significant difference in police reporting of violence by a perceived gang
member when the offender is an intimate, a stranger, or a family member.
However, the police are more likely to be contacted regarding a violent victim-
ization when the offender is a gang member as well as a friend (AOR = 1.44).
The second moderator term reveals that the influence of perceived offender
gang status on bystander intervention is moderated by the offender’s race
during a violent victimization. Results from models indicate no significant
difference in the odds of bystander intervention during violence committed by a
gang member when the offender is some “other” race, or the race of the
offender is unknown. In contrast, the odds of bystander intervention during
perceived gang member perpetrated violence increase when the offender is
white (AOR = 1.36) or black (AOR = 1.39).
Discussion and Conclusions
Research demonstrates that gang members are involved in a disproportionately
large number of violent acts. What is less clear is how gang membership shapes
the incident-level characteristics of crime. Though a few studies have compared
the situational characteristics of gang and non-gang homicide, little is known
about the dynamics of non-lethal gang versus non-gang crime (for an exception
see Melde & Rennison, 2008). As McGloin (2007) suggested, a focus on the
microlevel context of gang crime may aid our understanding of the mechanisms
through which gang membership leads to greater levels of violent offending.
Felson (2006) provides a theoretical explanation of the microlevel functions of
gang membership by focusing on the unique manner in which gang members use
outward signs of affiliation to intimidate—and thus control the behavior of—
those around them. Specifically, Felson argues that victims are less likely to
7. Results not shown but available upon request.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 655
resist, and bystanders are less likely to intervene in gang crime for fear of retal-
iation by the gang. Also, Felson argues that victims and bystanders are less
likely to cooperate with the police after crimes involving gang members for fear
of gang reprisal.
The current study compared the microlevel interaction of victims, bystand-
ers, and offenders in gang and non-gang, non-lethal violent crime incidents. To
do so, we relied on data from the 1992 to 2005 NCVS, which includes a measure
of the victim’s perception of the offender’s gang membership. The NCVS is
ideally suited for the present study, as the victim’s perspective in such criminal
incidents, due to the focus on victim–offender dynamics, provides the most
appropriate and direct information on victim behavior (Block, 1981). Further,
the NCVS includes a number of relevant control variables across a large nation-
ally representative sample of criminal incidents. In all, we examined over
29,000 incidents of non-lethal violence, of which over 2,200 were perceived to
be committed by a gang member. Results provide little support for Felson’s
propositions regarding the functionality of gang membership for the commission
of crime.
Some support obtains in findings from multivariate models that show that
violence committed by a gang member decreases the odds of forceful victim
resistance during a simple assault. In contrast, gang member violence increases
the odds of forceful victim resistance during an aggravated assault.
8
Perceived
gang member offending did not significantly influence forceful victim resistance
for overall violence, rape/sexual assault, or robbery, nor did it influence non-
forceful victim resistance for any of the violent crime categories investigated.
Turning to the hypothesis that gang violence is associated with a reduced
likelihood of bystander intervention, our research offers almost no support.
Multivariate analyses demonstrate that when the offender is thought to be a
gang member, the odds of bystander intervention increase for total violence,
aggravated assault, and simple assault. Further, we found no influence of the
perceived gang status of the offender for rape/sexual assault or robbery.
However, additional analyses did show that perceived gang membership’s influ-
ence on bystander intervention is conditioned by the offender’s race for total
violence. The odds of bystander intervention during gang member perpetrated
violence increase when the offender is white or black. Still, no significant
difference in the odds of bystander intervention is found during total violence
committed by an offender thought to be a gang member when the offender is
some “other” race, or the race of the offender is unknown.
Our final set of models investigated the influence of offender’s perceived
gang membership status on police reporting. Again, findings offer very limited
support for Felson’s (2006) propositions. Results demonstrate that the odds of
8. One potential explanation for this finding is that victim behavior may influence whether or not a
simple assault escalates to an aggravated assault, such that victim resistance may actually provoke
an offender to use more violent means to complete the crime. Thus, the negative relationship found
between simple assault and the perceived gang membership status of the offender may be
influenced by victim/offender dynamics.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
656 MELDE AND RENNISON
police reporting decrease when a gang member commits a robbery. However,
results also show that the odds of police reporting increase when a gang
member commits overall violence or a simple assault. The perceived gang
membership of the offender failed to influence police reporting for rape/sexual
assault or aggravated assault even when relevant controls were included in the
models. Further investigation revealed that for total violence, gang member-
ship’s influence on police reporting is conditioned by the relationship between
victim and offender. Models including interaction terms demonstrate no differ-
ence in the likelihood of police reporting when the gang offender is an intimate,
a stranger, or a family member. However, the police are more likely to be
contacted regarding a violent victimization when the offender is a gang member
and a friend. In the end, it seems that the common predictors of police report-
ing found in the literature, such as violence severity, victim injury, and weapon
presence (Gottfredson & Gottfredson, 1988; Gottfredson and Hindelang, 1979;
Laub, 1981; Skogan, 1976, 1984), are more consistently associated with this
behavior than perceived gang involvement.
From a policy standpoint, the current study provided a first step in the empirical
examination of factors associated with legislation that targets gang intimidation
practices. Specifically, the current study examined the incident-level character-
istics of crime to determine if victim and bystander behavior was systematically
altered by the victim’s perception that the offender was a gang member. Results
suggest that gang members do not systematically alter victim and bystander
behavior in a manner that allows gang members to offend with greater ease.
Legislation based on this premise, thus, should be reviewed to ensure that
enhancement penalties associated with these laws are not imposed based on
confirmed gang membership alone, but that clear evidence exists to suggest that
individual members used their gang association to further their criminal actions.
Importantly, this research paves the way for future research to empirically exam-
ine the ability of gangs to intimidate victims and witnesses into not cooperating
in the prosecution of crimes involving gang members. Even though prosecutors
have cited this as an issue in gaining cooperation from victims and witnesses
(Johnson et al., 1995), empirical verification of this process is lacking.
The decreased likelihood of reporting incidents of gang-involved robbery to
the police is not in line with previous research which concluded that people use
less discretion when deciding to report serious acts of violence to the police
(Baumer, 2002; Felson, Messner, & Hoskin, 1999). As Baumer (2002, p. 607)
stated: “Victims of serious crimes may use less discretion in general …, and they
may feel compelled to use formal legal channels to ensure the recovery of
monetary losses associated with their victimization and to minimize their likeli-
hood of repeat victimization.” Perceived gang involvement appears to be one of
the few factors that influence the victim’s discretionary calculus by impeding
their decision to call the police after being robbed. Unfortunately, the decision
to not report the event to the police is likely an indication of the victim’s confi-
dence in the police, and may indicate a general mistrust in the ability of police
to regulate gang behavior (Baumer & Lauritsen, in press). Underreporting of
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 657
robberies involving gang members may also lead to under-enforcement by
police officials in neighborhoods affected by such crime (Baumer, 2002;
Kennedy, 1997), as intelligence led policing efforts commonly use such calls for
service when determining the number of patrol units to place in a given area.
Identifying street gang members has proved troublesome for quite some
time. Law enforcement officials and researchers alike have been unable to
agree upon a universal definition of what a gang or gang member is, and thus
systematic research on, and the enforcement and prosecution of, gangs and
gang members is often hampered by the confidence one has in the gang
measure(s) employed. While the current research on victim behavior is not
hampered by the subjective nature of our gang measure, it does highlight the
need to better understand how citizens, law enforcement officials, and prose-
cutors determine gang membership. In Felson’s (2006) terms, what are the
important “signals” these groups use in identifying individuals as gang involved?
Most importantly, does the use of these signals provide an accurate reflection of
the gang status of the offender? Given the use of sentence enhancements for
gang crime (Finn & Healey, 1996; Johnson et al., 1995; Spadanuta, 2008), in
particular, documentation of the accuracy of gang identification could help
local and federal legislators draft laws in a manner that could diminish the
probability of Type I and Type II errors at the sentencing stage of prosecution.
In the end, however, gang membership does not systematically alter the inci-
dent-level characteristics of violence in a manner consistent with Felson’s
(2006) assertions. Rather, the effect of gang membership on victim and
bystander behavior is dependent upon crime type, as well as other situational
characteristics. In fact, in a number of models the influence of the perceived
gang membership status of the offender on victim and bystander behavior was
contrary to that hypothesized. For instance, perceived gang membership was
associated with an increase in the probability of forceful victim resistance in
acts of aggravated assault—although it was associated with a decreased
likelihood of forceful resistance in simple assault. Explaining these inconsistent
findings is difficult empirically, given limitations in the data. However, they are
consistent with other theories of victim action.
Richard Felson’s (1993) social interactionist approach to violent conflict
proposes that victim behavior may be conditioned by factors not measured by
the NCVS, and direct a victim to act in ways contradictory to Marcus Felson’s
(2006) predictions regarding gang violence. For instance, Felson (1993, p. 115)
proposes three factors that may lead to victim resistance in violent crime. First,
the attack produces a grievance between parties and a natural impulse to retal-
iate. Second, retaliation has the ability to deter the attacker from future
attacks. And third, retaliation can help the victim “save face” (see also Jacobs
& Wright, 2006). In fact, Richard Felson’s arguments regarding victim behavior
are in line with Anderson’s (1999) description of the code of the streets,
whereby victims are expected to retaliate in instances of assault in order to
maintain a certain level of respect in one’s community, and to prevent one’s
continued victimization due to the potential for being labeled an easy target.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
658 MELDE AND RENNISON
This seems especially likely in incidents where the victim is also a gang member,
as they are unlikely to call the police due to their own involvement in illicit
activities, and thus must use informal measures to protect themselves from
future attack (Jacobs & Wright, 2006; Wright & Decker, 1997). Unfortunately,
the NCVS does not allow one to model motivation for resistance such as
attitudes about violence, adherence to the street code, or the gang status of
the victim. Without these measures, deciphering the mental state of the victim
and why they behaved in the manner reported is not possible.
Future research on the effects of gang violence on victim behavior should
examine the various theoretical perspectives on victim behavior by collecting
data on more subjective and perceptual factors related to decision-making.
Thus, direct measures of victim attitudes related to gangs, community violence,
trust in the police, dispute resolution, as well as the perceived risk and fear of
future gang victimization may help further our understanding of this phenome-
non. This seems especially important when examining robbery, given these
offenses frequently target those also involved in the illicit street economy
(Wright & Decker, 1997). More importantly, it is through a better understanding
of how victim attitudes and motivations impact behavior both during and after
acts of violence that policies directed toward informing the public on the proper
response to violence can be developed.
Finally, while the nationally representative nature of the NCVS is a strength
in determining the impact of gang membership on victim behavior more gener-
ally, use of such data may also hide idiosyncratic relationships that may exist at
the local level. This becomes especially important given the number of states
which have enacted witness intimidation laws in response to concerns related to
the perceived tendency of gangs to intimidate witnesses into not cooperating
with the police. While findings from the NCVS do not support this presumption,
we cannot be sure that gang intimidation practices do not systematically change
victim and bystander behavior in some communities. Perhaps, witness intimida-
tion practices are more frequent in chronic gang cities that are home to more
organized street gangs than is typical across the USA (e.g., Chicago, Los
Angeles). For this reason, future research should be conducted at the city or
community level in order to determine whether or not gang offenders alter the
behavior of victims and bystanders in a manner consistent with Felson’s (2006)
hypotheses and those of state and local lawmakers. According to data collected
as part of the NCVS, however, we find that victim and bystander behavior is not
systematically altered in a manner conducive to the commission of violent crime
for those perceived to be gang involved.
References
Anderson, E. (1999). Code of the street: Decency, violence, and the moral life of the
inner city. New York: W.W. Norton.
Babbie, E. R. (1973). Survey research methods. Belmont, CA: Wadsworth.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 659
Bachman, R. (1998). The factors related to rape reporting behavior and arrest: New
evidence from the national crime victimization survey. Criminal Justice and Behavior,
25, 8-29.
Baumer, E. P. (2002). Neighborhood disadvantage and police notification by victims of
violence. Criminology, 40, 579–616.
Baumer, E. P., Felson, R. B., & Messner, S. F. (2003). Changes in police notification for
rape, 1973-2000. Criminology, 41, 841–872.
Baumer, E. P., Horney, J., Felson, R. B., & Lauritsen, J. L. (2003). Neighborhood disad-
vantage and the nature of violence. Criminology, 41(1), 39–72.
Baumer, E. P., & Lauritsen, J. L. (in press). Reporting crime to the police, 1973-2005: A
multivariate analysis of long-term trends in the NCS and NCVS. Criminology.
Bjerregaard, B., & Lizotte, A. J. (1995). Gun ownership and gang membership. Journal of
Criminal Law and Criminology, 86, 37–58.
Block, R. (1981). Victim-offender dynamics in violent crime. Journal of Criminal Law and
Criminology, 72, 743–761.
Block, R., & Skogan, W. (1986). Resistance and non-fatal outcomes in stranger-to-
stranger predatory crime. Violence and Victims, 1, 241–253.
Catalano, S. (2006). Criminal victimization 2005. Washington, DC: U.S. Department of
Justice, Bureau of Justice Statistics.
Curry, G. D., & Spergel, I. A. (1988). Gang homicide, delinquency, and community.
Criminology, 26, 381–403.
Esbensen, F. A., & Tusinski, K. T. (2007). Youth gangs in the print media. Journal of
Criminal Justice and Popular Culture, 14(1), 21–38.
Felson, M. (2006). Crime and nature. Thousand Oaks, CA: Sage.
Felson, R. (1993). Predatory and dispute-related violence: A social interactionist
approach. In R. Clarke & M. Felson (Eds.), Routine activities and rational choice
(pp. 103-125). New Brunswick, NJ: Transaction.
Felson, R. B., Baumer, E. P., & Messner, S. F. (2000). Acquaintance robbery. Journal of
Research in Crime and Delinquency, 37(3), 284–305.
Felson, R. B., Messner, S. F., & Hoskin, A. (1999). The victim-offender relationship and
calling the police in assaults. Criminology, 37, 931–948.
Finn, P., & Healey, K. M. (1996). Preventing gang- and drug-related witness intimida-
tion. Washington, DC: U.S. Department of Justice, Office of Justice Programs,
National Institute of Justice.
Garofalo, J. (1990). The National Crime Survey, 1973-1986: Strengths and limitations of a
very large data set. In D. L. MacKenzie, P. J. Baunach, & R. R. Roberg (Eds.), Measur-
ing crime: Large-scale, long-range efforts (pp. 75-97). Albany, NY: Status University
of New York Press.
Gottfredson, M. R., & Gottfredson, D. M. (1980). Decision making in criminal justice:
Toward the rational exercise of discretion. Cambridge, MA: Ballinger.
Gottfredson, M. R., & Gottfredson, D. M. (1988). Decision making in criminal justice (2nd
ed.). New York: Plenum.
Gottfredson, M. R., & Hindelang, M. J. (1979). A study of the behavior of law. American
Sociological Review, 43, 3–18.
Gottfredson, M. R., & Hirschi, T. (1990). A general theory of crime. Palo Alto, CA:
Stanford University Press.
Greenberg, M. S., & Ruback, R. B. (1992). After the crime: Victim decision making. New
York: Plenum.
Hart, T. C., & Miethe, T. D. (2008). Exploring bystander presence and intervention in
nonfatal violence victimization: When does helping really help? Violence and Victims,
23(5), 637–651.
Hart, T. C., & Rennison, C. M. (2003). Reporting crime to the police, 1992-2000.
Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
660 MELDE AND RENNISON
Heimer, K. (2000). Changes in the gender gap in crime and women’s economic marginal-
ization. In K. Heimer (Ed.), The nature of crime: Continuity and change (pp. 427-484).
Washington, DC: USGPO.
Hindelang, M. J. (1978). Race and involvement in common-law personal crimes. American
Sociological Review, 43, 93–109.
Hindelang, M. J. (1981). Variations in sex-race-age specific incidence rates of offending.
American Sociological Review, 46, 461–474.
Hosmer, D. W., & Lemeshow, S. (2000). Applied logistic regression (2nd ed.). New York:
Wiley.
Howell, J. C. (1998). Youth gangs: An overview. Washington, DC: Office of Juvenile
Justice and Delinquency Prevention, U.S. Department of Justice.
Hubble, D. L. (1995, August). The national crime victimization survey redesign: New
questionnaire and procedures development and phase-in methodology. Paper
presented at the annual meeting of the American Statistical Association, Orlando, FL.
Ihalainen, E., Lindstrom, L., & Mappes, J. (2007). Investigating Mullerian mimicry:
Predator learning and variation in prey defences. Journal of Evolutionary Biology, 20,
780–791.
Jaccard, J. (2001). Interaction effects in logistic regression. Newbury Park: Sage.
Jacobs, B. A., & Wright, R. (2006). Street justice: Retaliation in the criminal under-
world. New York: Cambridge University Press.
Johnson, C., Webster, B., & Connors, E. 1995. Prosecuting gangs: A national assessment
(Research in brief). Washington, DC: National Institute of Justice, U.S. Department of
Justice.
Kapan, D. D. (2001). Three-butterfly system provides a field test of Müllerian mimicry.
Nature, 409, 338–340.
Katz, J. (1988). The seductions of crime. New York: Basic Books.
Kennedy, R. (1997). Race, crime, and the law. New York: Vintage Books.
Klaus, P., & Rennison, C. M. (2002). Age patterns of violent victimization, 1976-2000.
Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics.
Lane, J. (2002). Fear of gang crime: A qualitative examination of the four perspectives.
Journal of Research in Crime and Delinquency, 39, 437–471.
Laub, J. H. (1981). Ecological considerations in victim reporting to the police. Journal of
Criminal Justice, 9, 419–430.
Laub, J. H. (1997). Patterns of criminal victimization in the United States. In R. C. Davis,
A. J. Lurigio, & W. G. Skogan (Eds.), Victims of crime (pp. 9-26). Thousand Oaks, CA:
Sage.
Lauritsen, J. L., & Rennison, C. M. (2005). The role of race and ethnicity in violence
against women. In K. Heimer & C. Kruttschnitt (Eds.), Gender and crime: Patterns in
victimization and offending (pp. 303-322). New York: New York University Press.
Lauritsen, J. L., & Schaum, R. (2004). Social ecology of violence against women.
Criminology, 42, 323–357.
Lizotte, A. J., Tesoriero, J. M., Thornberry, T. P., & Krohn, M. D. (1994). Patterns of
adolescent firearms ownership and use. Justice Quarterly, 11, 51–73.
Lueptow, L., Mueller, S. A., Hammes, R. R., & Master, L. R. (1977). The impact of
informed consent regulations on response rate and response bias. Social Methods and
Research, 6, 183–204.
Marshall, C. E., & Webb, V. J. (1992). Impact of gender and race upon armed victim
resistance: Some findings from the National Crime Survey. Criminal Justice Policy
Review, 6(3), 241–260.
Maxson, C. L. (2004). Civil gang injunctions: The ambiguous case of the national migra-
tion of a gang enforcement strategy. In F.-A. Esbensen, S. G. Tibbetts, & L. Gaines
(Eds.), American youth gangs at the millennium (pp. 375-389). Long Grove, IL:
Waveland.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 661
Maxson, C. L., Gordon, M. A., & Klein, M. W. (1985). Differences between gang and
nongang homicides. Criminology, 23(2), 209–222.
McGloin, J. M. (2007). The continued relevance of gang membership. Criminology and
Public Policy, 6(2), 231–240.
Melde, C., & Rennison, C. M. (2008). The effect of gang perpetrated crime on the
likelihood of victim injury. American Journal of Criminal Justice, 33, 234–251.
Melde, C., Taylor, T. J., & Esbensen, F. A. (2009). “I got your back”: An examination of
the protective function of gang membership in adolescence. Criminology, 47, 565–594.
Mitchell, R. W. (1996). The psychology of human deception. Social Research, 63, 819–861.
Planty, M. (2002). Third-party involvement in violent crime, 1993-99. Washington, DC:
U.S. Department of Justice, Bureau of Justice Statistics.
Pizarro, J. M., & McGloin, J. M. (2006). Explaining gang homicides in Newark, NJ: Collective
behavior or social disorganization. Journal of Criminal Justice, 34, 195–207.
Renn, O. (2005). Risk governance: Towards an integrative approach (White Paper
Series). Geneva: International Risk Governance Council.
Rennison, C. M. (2001). Criminal victimization 2000: Changes 1999-2000 with trends
1993-2000. Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics.
Rennison, C. M., & Rand, M. R. (2003). Criminal victimization, 2002. Washington, DC:
U.S. Department of Justice, Bureau of Justice Statistics.
Rennison, C. M., & Rand, M. R. (2007). Introduction to the National Crime Victimization
Survey. In J. P. Lynch & L. A. Addington (Eds.), Understanding crime statistics:
Revisiting the divergence of the NCVS and the UCR (pp. 3-13). Cambridge: Cambridge
University Press.
Rosenfeld, R. B., Bray, T., & Egley, A. H. (1999). Facilitating violence: A comparison of
gang-motivated, gang-affiliated, and non-gang youth homicides. Journal of Quantita-
tive Criminology, 15, 495–516.
Ruback, R. B., Greenberg, M. S., & Westcott, D. R. (1984). Social influence and crime-
victim decision making. Journal of Social Issues, 40, 51–76.
Sampson, R. J., & Lauritsen, J. L. (1997). Racial and ethnic disparities in crime and criminal
justice. In M. Tonry (Ed.), Ethnicity, crime, and immigration: Comparative and cross-
national perspectives (Vol. 21, pp. 311-374). Chicago: University of Chicago Press.
Sewell, W. H., & Hauser, R. M. (1975). Education, occupation and earnings: Achievement
in the early career. New York: Academic Press.
Short, J. F., & Strodtbeck, F. L. (1965). Group processes and gang delinquency. Chicago:
University of Chicago Press.
Shotland, R. L., & Straw, M. K. (1976). Bystander response to an assault: When a man
attacks a woman. Journal of Personality and Social Psychology, 34, 990–999.
Skogan, W. G. (1976). Citizen reporting of crime: Some national panel data. Criminology,
13, 535–549.
Skogan, W. G. (1977). Dimensions of the dark figure of unreported crime. Crime and
Delinquency, 23, 41–50.
Skogan, W. G. (1984). Reporting crimes to the police: The status of world research.
Journal of Research in Crime and Delinquency, 21, 113–137.
Spadanuta, L. (2008). Can I get a witness? Security Management (November), 51–58.
Speed, M. P., Alderson, N. J., Hardman, C., & Ruxton, G. D. (2000). Mullerian mimicry:
An experiment with wild birds. Proceedings: Biological Sciences, 267, 725–731.
StataCorp. (2005). Stata Statistical Software: Release 9.0 [Computer Software]. College
Station, TX: StataCorp.
Tark, J., & Kleck, G. (2004). Resisting crime: The effects of victim action on the
outcomes of crimes. Criminology, 42, 861–909.
Thornberry, T. P. (1998). Membership in youth gangs and involvement in serious and
violent offending. In R. Loeber & D. P. Farrington (Eds.), Serious and violent juvenile
offenders (pp. 147-164). Thousand Oaks, CA: Sage.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
662 MELDE AND RENNISON
Thornberry, T. P., Krohn, M. D., Lizotte, A. J., Smith, C. A., & Tobin, K. (2003). Gangs
and delinquency in developmental perspective (A. Blumstein & D. Farrington, Eds.).
New York: Cambridge University Press.
Ullman, S. (2005, November). Comparing gang and individual rape incident characteris-
tics in a large sample of urban women. Paper presented at the annual meeting of the
American Society of Criminology, Toronto, Canada.
U.S. House of Representatives. (1997a). Administration’s anti-gang and youth violence
initiative (105th Cong., 1st session). Washington, DC: Committee on the Judiciary,
Subcommittee on Crime.
U.S. House of Representatives. (1997b). Gang-related witness intimidation and retaliation
(105th Cong., 1st session). Washington, DC: Committee on the Judiciary, Subcommittee
on Crime.
U.S. Senate. (1994). The gang problem in America: Formulating an effective federal
response (103rd Cong., 2nd session). Washington, DC: Committee on the Judiciary,
Subcommittee on Juvenile Justice.
Vane-Wright, R. I. (1980). On the definition of mimicry. Biological Journal of the Linnean
Society, 13, 1–6.
Wright, R. T., & Decker, S. H. (1997). Armed robbers in action: Stickups and street
culture. Boston: Northeastern University Press.
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 663
Appendix: Description of Variables
Variables Description
Dependent variables
Victim resistance Victim resistance is based on several NCVS questions. First, it was
determined whether the victim resisted or not based upon
responses to two questions: “Did you do anything with the idea of
protecting yourself or your property while the incident was going
on?” and “Was there anything you did or tried to do about the
incident while it was going on?” Ascertaining how a victim
responded during a violent victimization and is based on responses
to several NCVS questions (See variables v4141 through v4160). In
these analyses, victim resistance is described using a set of three
dichotomous variables: forceful resistance, non-forceful
resistance, and no resistance. Forceful victim resistance includes
instances where the victim attacked, threatened to attack, or
attempted to apprehend the offender. Non-forceful resistance
refers to victimizations in which the victim’s response was to stall,
scream, argue, or cry for help. And no resistance indicates the
victim did nothing. No resistance served as the reference group in
the analyses.
Bystander
intervention
Models using Bystander intervention as the dependent variable are
restricted to victimizations in which a bystander was present. The
presence or absence of a bystander was ascertained via a question
in the NCVS in which the victim was asked “Was anyone present
during the incident besides you and the offender?” (See variable
v4184). Among the subset of victimizations in which a bystander
was present, the bystander intervention variable indicates
whether a bystander intervened or not. Intervention was
determined using a series of NCVS questions regarding whether
the actions of any bystander “helped or hurt the situation in any
way” (See variables v4184 through v4202). Bystander intervention
can take the form of helping (e.g., prevented injury, scared off
offender, helped victim escape), or hurting the incident (e.g., led
to victim injury, made offender more angry). If a bystander
engaged in any of these activities the variable was coded a “1”
(intervention). A situation in which a bystander failed to act was
coded a “0” (no intervention).
Police reporting A dichotomous measure that contrasts victimizations that were
not reported (not reported “0”), and those that were reported to
the police (reported “1”) based on the NCVS question “Were the
police informed or did they find out about this incident in any
way?” (See variable V4399).
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
664 MELDE AND RENNISON
Variables Description
Independent variables
Perceived gang
membership
Offenders were coded as a gang member or not based on the
victim’s response to the following question: “Was the offender a
member of a street gang, or don’t you know?” Gang membership
is measured as a dichotomous variable where offenders perceived
as gang members are coded as “1” and those who were not or if
the victim was unsure are coded as “0.” Those who reported that
they were unsure if the offender was a gang member or not were
coded zero in analyses reported in the current paper. We coded
these incidents in this manner because it is unlikely that victim or
bystander behavior would be altered by gang signs unless the
victim was more certain of this possibility. We conducted
supplementary analyses which excluded those cases where the
victim reported “don’t know” to the gang question, with
substantively identical results.
Control variables Research identifies several empirically relevant situational,
offender and victim characteristics associated with the dependent
variables used in this research (Gottfredson & Hirschi, 1990;
Hindelang, 1978; Hindelang, 1981; Klaus & Rennison, 2002;
Lauritsen & Rennison, 2005; Lauritsen & Schaum, 2004; Planty,
2002; Rennison, 2001; Rennison & Rand, 2003; Sampson &
Lauritsen, 1997). These correlates are described below.
Situational control variables
Type of violence A series of four dichotomous variables are used to account for four
crime types when applicable: rape/sexual assault, robbery,
aggravated assault, and simple assault.
Injuries Injuries sustained by the victim are included in using the three
dichotomous variables: no injury (no injury), minor injury (minor),
and serious injury (serious). Minor injuries include bruises, black
eyes, cuts, scratches, swelling, or any unknown injury resulting in
less than two days of hospitalization. Serious injuries include
completed rape, broken bones, lost teeth, internal injuries, loss
of consciousness, and any unknown injury resulting in two or more
days of hospitalization. No injury serves as the excluded reference
group.
Number of
offenders
The number of offenders present is measured using a dichotomous
indicator in which 0 = one offender and 1 = two or more offenders.
Weapons Weapon presence is accounted for in the analyses with a series of
dichotomous variables: no weapon, firearm, knife or other cutting
instrument (knife), and other weapon (other). No weapon is the
excluded reference category.
Third party
presence
The presence or absence of a third party is measured using a
dichotomous variable in which 0 = no third party and 1 = at least
one-third party present.
Appendix (Continued)
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
INTIMIDATION AND STREET GANGS 665
Variables Description
Offender characteristics
Gender To account for the offender’s gender, three dichotomous variables
are used: male, female, group/unknown. Male refers to a single
male or a group of male offenders. Female identifies a single
female offender or a group of all female offenders. Group/
Unknown indicates a group of offenders with both males and
females, or that the victim could not identify the gender of the
offenders. Male serves as the reference category.
Age Offender’s age is controlled in the models using four dichotomous
indicators: less than 18, 19 to 29, 30 plus, and group/unknown.
Less than 18 is used as the excluded reference group.
Race Four dichotomous variables are used to account for offender’s
race: white, black, other, group/unknown. White refers to a single
white offender or a group of white offenders. Black refers to a
single black offender or a group of black offenders. Other refers
to a non-white or non-black offender or a group of the same. And
group/unknown indicates offenders that the victim could not
identify. White serves as the reference group in the analyses.
Victim/offender
relationship
Four dichotomous variables are used to distinguish among four
victim and offender relationships: stranger, family member,
friend, and intimate. Intimate is defined as a current or former
spouse, boyfriend or girlfriend. Stranger serves as the reference
group.
Victim characteristics
Gender Males are more likely to be victimized than are females with the
exception of rape/sexual assault. To account for the relationship
between gender and victimization a dichotomous variable where
0 = male, and 1 = female.
Age To account victim’s age, a continuous measure variable ranging
from 12 years to 90 years is included. Persons age 90 or older are
included in the “90” category.
Race The race and Hispanic origin of the victim (race) is accounted for
in the analyses using four categories: non-Hispanic white (white),
non-Hispanic black (black), non-Hispanic other (other), and
Hispanic of any race (Hispanic). White is used as the reference
category in the models.
Annual household
income
Annual household income is measured using 14 categories of
unequal width. From lowest to highest they are: Less than $5,000,
$5,000 to $7,499, $7,500 to $9,999, $10,000 to $12,499, $12,500
to $14,999, $15,000 to $17,499, $17,500 to $19,999, $20,000 to
$24,999, $25,000 to $29,999, $30,000 to 34,999, $35,000 to
$39,999, $40,000 to $49,999, $50,000 to $74,999, $75,000 and
over. As has been done in previous work using this measure (see
e.g., Baumer, 2002), this variable is treated as a continuous
measure.
Appendix (Continued)
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
666 MELDE AND RENNISON
Variables Description
Marital status Victim’s marital status is measured using five dichotomous
variables: married, never married, widowed, divorced, and
separated. Married serves as the excluded reference category.
Employment
status
This concept is captured using a dichotomous measure in which 0
= unemployed and 1 = employed. Employed refers to those having
been employed during the last six months, while unemployed
indicates a victim who has not been employed during the last six
months.
Owned Extant research shows that whether one owns or rents their
residence is related to violence. This is accounted for by a
dichotomous measure in which 0 = rented, and 1 = owned.
MSA Location is also relevant to the behavior of violence. To account
for this characteristic, three dichotomous measures are included:
urban, suburban, and rural. Urban is the excluded category.
Appendix (Continued)
Downloaded By: [Michigan State University] At: 12:53 10 August 2010
... Differences in crime reporting may vary with trust in law enforcement across neighborhoods (Kirk and Papachristos 2011). It is not entirely clear that willingness to report crime varies much between gang-related and non-gang crime (Melde and Rennison 2010). In any case, as long as the underreporting probability is the same for triggering crimes (parents) and offspring, then estimates of triggering parameters are not biased (Daley and Vere-Jones 2003), or are impacted in predictable ways (Tucker et al. 2019). ...
Article
Full-text available
Objectives Gangs are thought to enhance participation in violence. It is expected then that gang-related violent crimes trigger additional crimes in a contagious manner, above and beyond what is typical for non-gang violent crime.Methods This paper uses a multivariate self-exciting point process model to estimate the extent of contagious spread of violent crime for both gang-related and non-gang aggravated assaults and homicides in recent data from Los Angeles. The degree of contagious cross-triggering between gang-related and non-gang violent crime is also estimated.ResultsGang-related violence triggers twice as many offspring events as non-gang violence and there is little or no cross-triggering. Gang-related offspring events are significantly more lethal than non-gang offspring events, but no more lethal than non-contagious background gang crimes.Conclusions Contagious spread of gang-related violent crime is different from contagion in non-gang violence. The results support crime prevention policies that target the disruption of gang retaliations.
... Although these models were originally designed to explain the relationship between gang membership and offending, they may also explain the gang-victimization link. For example, the selection hypothesis states that youth who join gangs are those most likely to be victimized in the first place-perhaps suggesting that gang membership is sought as a protective mechanism (but see Melde & Rennison, 2010). Thus, the gang is unlikely to have any direct effect on victimization. ...
Article
The current study represents the first attempt to examine how genetic and environmental factors work in concert to influence gang membership, victimization, and the effect of gang membership on victimization experiences. Using data drawn from the National Longitudinal Study of Adolescent Health, the present study employs behavioral genetic modeling (i.e., the ACE model and the DeFries-Fulker model) to examine these associations. The findings indicate that gang affiliation is influenced significantly by both genetic factors and environmental factors that are uniquely experienced by the individual. Controlling for heritable influences, gang membership increased the risk of victimization over time. The latter finding suggests that gang membership operates as a nonshared environmental influence on victimization.
... On that note, Cooney's theory and the findings of this paper have implications for another body of research. Studies in "self-protective behavior," or "SP" for short, are concerned with how and why people respond when being victimized, and the effect of reactions on offending outcomes (see, e.g., Block & Skogan, 1986;Kleck & DeLone, 1993;Melde & Rennison, 2010;Tark & Kleck, 2004). Thus, reflexive control and SP are different terms for the same behavior. ...
Article
Full-text available
There are a number of ways that victims of violence informally handle attacks as they unfold. Their responses range in severity from physical resistance, to talking it out with the offender, to running away, to cooperating. Why do victims respond in a more or less severe manner? Cooney (2009) suggests that social distance is part of the answer: the further the relational or cultural distance between offender and victim, the more severe the latter's response. Using National Crime Victimization Survey data, we test hypotheses derived from this theory and find oppositional findings. Specifically, results indicate that closer social distance predicts more severe responses. We conclude by discussing the implications of this finding for future work, especially as relates to the study of self-protective behavior.
Article
Data from the National Incident-Based Reporting System (NIBRS) are used to examine the tendency for victims of physical assault, sexual assault, and robbery to refuse to cooperate with the police (N= 3,856,171). Analyses of physical assaults involving homosexual and heterosexual couples did not support the hypothesis that women attacked by their male partners are less likely to cooperate than victims of other assaults. Analyses of violent offenses more generally showed that victims of violence were more likely to refuse to cooperate if they knew the offender in any way than if the offender was a stranger. In the case of physical and sexual assault, these effects were mainly observed for minor incidents. Finally, victims of sexual assault were more likely to cooperate with the police than victims of physical assault. The findings suggest the importance of comparing the victim's reactions to intimate partner violence and sexual assault to their reactions to other offenses. Aggr. Behav. 9999:1-12, 2015. © 2015 Wiley Periodicals, Inc.
Article
This paper concerns passivity. An array of research is reviewed that identifies factors that influence passivity and people’s failure to intervene in the face of social harm and injustice. The ways in which criminological theories could potentially explain the connections between passivity and harmful action are also discussed. Criminologists have tended to neglect passivity as either an independent or dependent variable, which we explain in terms of the generally individualistic orientation of the discipline of criminology and of the government agencies that support criminological research. Finally, we offer suggestions for effective intervention into passivity.
Book
The study of decisions in the criminal justice process provides a useful focus for the examination of many fundamental aspects of criminal jus­ tice. These decisions are not always highly visible. They are made, or­ dinarily, within wide areas of discretion. The aims of the decisions are not always clear, and, indeed, the principal objectives of these decisions are often the subject of much debate. Usually they are not guided by explicit decision policies. Often the participants are unable to verbalize the basis for the selection of decision alternatives. Adequate information for the decisions is usually unavailable. Rarely can the decisions be demonstrated to be rational. By a rationaldecision we mean "that decision among those possible for the decisionmaker which, in the light of the information available, maximizes the probability of the achievement of the purpose of the decisionmaker in that specific and particular case" (Wilkins, 1974a: 70; also 1969). This definition, which stems from statistical decision theory, points to three fundamental characteristics of decisions. First, it is as­ sumed that a choice of possible decisions (or, more precisely, of possible alternatives) is available. If only one choice is possible, there is no de­ cision problem, and the question of rationality does not arise. Usually, of course, there will be a choice, even if the alternative is to decide not to decide-a choice that, of course, often has profound consequences.
Chapter
Having presented the findings from 20 studies involving a variety of methodologies, we now introduce a theoretical framework that gives conceptual coherence to these findings. The proposed model, which emerged from our own empirical work and the work of others, offers several advantages. It provides an explanatory tool for comparing the decisions of various types of victims. In addition, it allows for the integration of the present findings with the existing literature on victim decision making. Further, the model suggests directions for future research, while at the same time laying the foundation for public policy decisions (see Chapter 10).
Article
If we knew more about the character of both offenders and victims, the nature of their relationships and the circumstances that create a high probability of crime conduct, it seems likely that crime prevention and control programs could be made much more effective. The President's Commission on Law Enforcement and the Administration of Justice, 1967a For more than 30 years, the National Crime Victimization Survey (NCVS) and its predecessor the National Crime Survey (NCS) have provided detailed information about the nature and extent of personal and property victimization in the United States. Much of our current understanding about crime, such as the most vulnerable segments of the population, the proportion of crime involving weapons, the types of injuries sustained in violent crimes, and the extent of reporting to police, is derived from the NCS and NCVS. The NCVS also serves as a model for victimization surveys implemented throughout the world because it incorporates many innovative methodological protocols that enhance its ability to produce reliable estimates of the nature and extent of criminal victimization. These methodological techniques combined with the large sample size and the extensive details collected by the survey on crime events are all reasons that the current NCVS remains the “most comprehensive and systematic survey of victims in the United States” (Mosher et al., 2002, p. 137). The NCVS and the Uniform Crime Reporting System (UCR) comprise the two ongoing national measures of crime in the United States.
Article
Admit it. When someone wrongs you, you want to get back at them. Despite Biblical injunctions to turn the other cheek, most of us are reluctant to do so. The urge to get even is so ingrained in the popular imagination that it has spawned a whole genre of Hollywood movies in which a peace-loving hero is driven to avenge the harm done to a loved one in an explosion of pent-up rage. Think of old classics like Death Wish, Billy Jack, or The Outlaw Josie Wales. Such films were popular because they tapped into a deep-seated human desire to see the good deliver justice to the bad, vanquishing evil once and for all. The real world of retaliation, however, is seldom as neat and clean as Hollywood would have us believe. It often is difficult to distinguish the good guy from the bad guy in disputes that take place beyond the reach of formal law. Short of death, few such conflicts are ever really settled for good, with each new strike generating a counter-strike in a deepening cycle of instability and violence. Indeed, formal law emerged in part to ameliorate the chaos engendered by retaliation, by replacing informal dispute resolution with a more institutionalized mechanism of social control. Street criminals, however, cannot realistically rely on formal law to settle their disputes. Despite being especially vulnerable to being preyed on, it is difficult for them to stake a legitimate claim to victim status.