Citation
Female domestic violence arrests in a pro-arrest jurisdiction

Material Information

Title:
Female domestic violence arrests in a pro-arrest jurisdiction
Creator:
Guerrero, Catherine L
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
44 leaves : ; 28 cm

Subjects

Subjects / Keywords:
Husband abuse ( lcsh )
Wife abuse ( lcsh )
Marital violence ( lcsh )
Abused husbands ( lcsh )
Abused wives ( lcsh )
Arrest ( lcsh )
Abused husbands ( fast )
Abused wives ( fast )
Arrest ( fast )
Husband abuse ( fast )
Marital violence ( fast )
Wife abuse ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 40-44).
Thesis:
Public administration
General Note:
School of Public Affairs
Statement of Responsibility:
by Catherine L. Guerrero.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
53987678 ( OCLC )
ocm53987678
Classification:
LD1190.P86 2003m G83 ( lcc )

Downloads

This item has the following downloads:


Full Text
FEMALE DOMESTIC VIOLENCE ARRESTS IN A PRO-ARREST
JURISDICTION
by
Catherine L. Guerrero
BA.., George Mason University, 1992
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Public Administration
2003


Thesis for Master of Public Administration
degree by
Catherine L. Guerrero
has been approved
by
Barbara Paradiso


Guerrero, Catherine L. (M.P.A., Graduate School of Public Affairs)
Female Domestic Violence Arrests in a Pro-Arrest Jurisdiction
Thesis directed by Associate Professor Mary Dodge
ABSTRACT
Data for this study were collected in Boulder County, Colorado by the
Domestic Abuse Prevention Project. Data were coded from police reports gathered
from 8 law enforcement agencies operating in Boulder County. One thousand nine
hundred eighty eight cases involving female suspects were compared to 6964 cases
involving male suspects to determine differences and similarities in factors
contributing to police arrest decisions. A Pearsons chi square analysis was run to
identify important relationships. A subsequent logistic regression was run to further
investigate the effects of variables on arrest. A formal police response occurred in
78.7% of cases.
The data shows differences in the victimization experiences of women
arrested for domestic violence as compared to male suspects. Fewer than 15% of the
male victims of female suspects had prior arrests for domestic violence, compared to
only 10% of the victims of male suspects. Almost 25% of female suspects had prior
victimization history, compared to only 9% of the male suspects.
Multivariate analysis was run on 239 cases involving a female suspect and
584 cases involving a male suspect. The presence of a weapon increased the
likelihood of arrest for both males and females. The presence of children increased
the likelihood of arrest for female suspects, as did a prior history of domestic
violence. Police were less likely to arrest male suspects if the couple was still in a
relationship at the time of the arrest.
Other research and anecdotal evidence for womens different experiences and
motivations for violence, as well as findings about the relationship between womens
victimization and their use of self defensive violence, support the possibility that
many of the women being arrested are actually victims of domestic violence. The
current study shows significant differences in the way that police officers in Boulder
County, Colorado, respond to male and female suspects in domestic violence cases.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
Signed <
in


CONTENTS
Tables....................................................vi
CHAPTER
1. INTRODUCTION............................................1
Current Debate.......................................3
Womens Use of Violence........................4
Mandatory Arrest...............................5
Unintended Consequences of Mandatory Arrest..........7
2. REVIEW OF THE LITERATURE................................9
Arrests of Women.....................................9
Factors Affecting Police Arrest Decisions...........12
Analysis of Police Reports....................12
Officer Survey Studies........................15
Descriptive Field Studies.....................19
Studies Using Combined Methodology............22
3. METHODOLOGY............................................23
4. FINDINGS...............................................26
Pearsons Chi-Square................................28
Multivariate Analysis...............................33
iv


5. CONCLUSION................................37
RESOURCES.........................................40
v


TABLES
Table
4.1 Victim Offender Characteristics.............................................27
4.2 Incident Characteristics....................................................28
4.3 Victim Offender Characteristics: Chi-Square Test............................29
4.4 Incident Characteristics: Chi-Square Test...................................31
4.5 Logistic Regression Predicting Arrest.......................................34
vi


CHAPTER 1
INTRODUCTION
Until recently, research concerning arrest in cases of domestic violence and
inquiry into womens use of violence in intimate relationships has followed
parallel tracks. As more jurisdictions report staggering increases in the number of
women arrested for domestic violence, there has been an increased focus on the
response of the criminal legal system to womens use of violence in intimate
relationships.
The majority of these studies have focused only on women arrested as part
of a dual arrest, in which both male and female parties are presumably the co-
combatants or perpetrators of domestic violence. These studies share the
common hypothesis that women who are arrested as part of a dual arrest are more
often than not self-defending victims. The underlying assumption is that high rates
of dual arrest indicate poor policing, whereas low rates of dual arrests indicate
effective investigation techniques and appropriate response. -It is important to
understand in what circumstances inappropriate arrests of self-defending victims
are being made. It is necessary, however, for researchers to look deeper into the
nuances of these arrests to understand all of the circumstances in which women are
subject to criminal legal sanctions for domestic violence. Arguments for a
1


contextualized approach to womens use of violence should be applied to the
arrests of all women, and the context should be expanded to include the response
of the criminal legal system. Failing to apply a gender analysis to our
investigations into what factors influence officer decisions to arrest in cases of
domestic violence diminishes the ability of scholars and advocates to understand
fully how the criminal legal system is responding to womens use of violence.
Integrating our growing knowledge of how and why women use violence
into public policy discussions has laid the groundwork for a challenge to the idea
that women are as violent as men. Without the inclusion of a gender analysis in
evaluations of public policies such as mandatory and pro arrest laws, however,
these arguments are doomed to partiality. An increasing sophistication in our
conversations about how women are affected by interpersonal, as well as, state
violence is necessary if we are to create a public policy response to domestic
violence that can work to increase the safety of many more victims.
This is not to say that the arrests of women are not justified in any case, or
that women are always victims in cases of domestic violence. A fully contextual
understanding of the effects of violence requires consideration of the interaction of
interpersonal violence and structural oppression in analyses of womens actions in
ways that are different from mens. The research and anecdotal evidence for
womens different experiences and motivations for violence, as well as findings
about the relationship between womens victimization and their use of self-
2


defensive violence, support investigation into the growing number of women
arrested for domestic violence (DeKeseredy & Schwartz, 1998). Saunders (1986)
found that 71% of battered women arrested for domestic violence used violence in
self-defense.
The current study uses a quantitative approach to investigate the arrests of
women in a jurisdiction with a strong mandatory arrest policy. Characteristics of
the victim-offender relationship and characteristics of the incident that resulted in
arrest are considered. A Pearsons Chi-Square test was used to identify significant
relationships, and a logistic regression was then used to further investigate the
nature of those relationships. Results for women arrested for domestic violence
were compared to results for men in order to gain an understanding of the
differences and similarities between police response to the two groups.
Current Debate
Debate surrounding womens use of violence in intimate relationships has
centered on two major themes over the last thirty years. First, concern about
whether women, in general, are as violent and aggressive in their intimate
relationships as men are has produced a large body of scholarship describing
gender differences in motivations, perceptions, and outcomes of violence in
intimate relationships. A second, concurrent debate has focused on the high
3


numbers of women arrested as a result of dual arrests, or as single suspects, has
been raging among police and advocates (Das Gupta, 2002).
Womens Use of Violence
Many studies have concluded that women use violence in intimate
relationships at the same or higher rates than men (Steinmetz, 1977-1978; Straus &
Gelles, 1986,1990; Strauss, Gelles, & Steinmetz, 1980). Most advocates and
scholars, however, disagree with the idea that women are as violent as men, and
argue that simply counting up the number of hits can mislead our attempts to fully
understand the experience of violence in intimate relationships (Das Gupta, 2002;
DeKeseredy & Schwartz, 1998). Without a full accounting for context, including
the motivations and outcomes of violent incidents, only a partial understanding of
violence in intimate relationships can be attained (DeKeseredy & Schwartz, 1998).
Gender differences in motivation for the use of violence are well
documented in the literature. Women report using violence in self-defense at
higher rates than men do, and their use of violence in self-defense goes up in
relation to their rate of victimization (DeKeseredy & Schwartz, 1998). Women
often use violence in an attempt to regain some control over a situation and out of
frustration and anger (Saunders, 1986). The categorizing of womens violence as
self-defense, retaliation, or coercive by researchers has not been representative of
many womens experiences; simply describing womens use of violence as self-
4


defense may be an oversimplification. Often, women do not distinguish between
self-defense and fighting back, but instead describe complex responses that
include reactions to the immediate situation, as well as resistance to their
experiences of domination over time (Saunders, 1986). Scholars and advocates
also point to the presence or absence of fear as an important indicator of
culpability in violent relationships (Hamburger & Guse, 2002).
Mandatory Arrest
More recently, debate about the high numbers of women arrested for
domestic violence has grown out of conflicts over the efficacy of arrest and the
implementation of pro- and mandatory arrest laws. It is well documented that
woman battering was traditionally seen as a private matter, outside the domain of
public intervention. Police were taught not to arrest, but instead to attempt to de-
escalate the situation by walking the batterer around the block, attempting to
mediate for the couple, or removing either the batterer or the victim from the
scene. Victims often were actively discouraged from seeking arrest or prosecution
(Zorza & Woods, 1994). When officers did arrest the batterer, it was usually in
response to aggression towards the officers, an act seen as a threat to police
authority. Arrest was used as a tool for the officer to maintain control of the
situation, rather than as a response to the assault itself (Buzawa & Buzawa, 2000).
5


A number of social and legal challenges in the 1970s and 1980s led to
changes in police response to domestic violence, including a broader public
understanding of the issue, and greater judicial and legislative action. Activists and
advocates began public education campaigns that focused attention on the
experiences of battered women. As an outgrowth of feminist activism in the
United States, thfe battered womens movement was able to capitalize on quickly
changing views about womens roles in society (Zorza & Woods, 1994).
Activists also focused public pressure on legislatures and police
departments to better protect battered women. Initial laws that focused specifically
on the police response to woman battering were introduced in the early 1970s, and
in 1977 the first law imposing a mandatory duty to arrest in cases where probable
cause existed was passed in Oregon (Zorza & Woods, 1994).
Feminist calls for police accountability also laid the groundwork for a
series of court cases, culminating in the highly publicized case of Tracy Thurman,
a woman who was severely injured by her husband while police were on the scene
(Zorza & Woods, 1994).. As huge sums of money were awarded to battered women
and their attorneys, police departments and municipalities began to recognize the
need to change their policies in order to protect themselves from lawsuits.
(Belknap, 1996; Buzawa & Buzawa, 2000; Zorza, 1992)
In 1984, amid debates about police response to woman battering,
researchers in Minnesota released results from the Minneapolis experiments that
6


claimed arrest was the most effective police response in deterring further assaults
(Sherman & Berk, 1984). Six subsequent replication studies offered mixed results
compared to the Minneapolis findings, although most seemed to show that arrest
was an effective deterrent in some cases (Zorza, 1992). Following the release of
the Minneapolis arrest study, police departments around the country began to
change their policies to encourage or require arrest in domestic violence cases
(Zorza & Woods, 1994).
Unintended Consequences of Mandatory Arrest
The accomplishments of the battered womens movement over the last 30
years have been significant. Profound changes have occurred in public awareness,
legal recognition, police response, and funding availability. It is clear, however,
that these shifts in public response to woman battering have not come without
costs. Mandatory and pro-arrest policies have been criticized on a number of
fronts.
Many advocates and researchers have criticized mandatory arrest policies
on the basis that they take decision-making power away from battered women in
ways that are ultimately disempowering and patronizing (Buzawa & Buzawa,
2000). There is also concern that men of color and poor men may be
disproportionately arrested under these laws (Coker, 2001; Miller, 1989). Poor
women and women of color call the police more often to resolve conflicts and
7


problems than do white middle and upper class women (Miller, 1989). Women
with more resources may be able to use friends and hotels as temporary shelter,
and may be able to leave relationships in ways that allow them to avoid criminal
legal intervention (Miller, 2001). The economic consequences of arrest, such as
the loss of a job and the burden of court costs and fines, may also have a more
profound effect on poor families (Coker, 2001; Miller, 1989).
The drastic increase in the numbers of women arrested for domestic
violence in many communities has been cause for concern among domestic
violence advocates. Although the arrest of batterers increased after the
implementation of these laws, so did the number of women being arrested as
batterers, either alone or as part of dual arrests (Jones & Belknap, 1999; Zorza &
Woods, 1994).
8


CHAPTER 2
REVIEW OF THE LITERATURE
There is a growing body of research that is focused on the factors that
affect the arrest decisions of police officers, although only a handful of these
specifically focus on the arrest of women. This research offers a unique
opportunity to understand how police officers operationalize mandatory arrest
policies on scene, and how battered womens advocates and criminal justice
personnel view the arrests of women.
Arrests of Women
Saunders (1995) surveyed 111 officers from three urban and seven rural
police departments in Wisconsin. The survey was conducted prior to the passage
or implementation of mandatory arrest laws or policies. Officers were given two
vignettes and asked to choose from ten responses that included: referral of each
party, warning one or the other party, and arrest of one or both of the parties. In the
first scenario presented to officers, the argument ended when the police arrived on
scene. In the second scenario, the couple continued to argue after the officers
arrived on scene. Officers were much more likely to arrest the victim in the
9


scenario in which the argument continued. Those officers who responded that they
would arrest the woman were more apt to believe that a man is justified in
assaulting a woman if she cheats on him and less likely to believe that women stay
in violent relationships because of practical barriers to leaving. These officers also
reported a higher level of discomfort when interacting with victims. Comments
written by officers in response to the second vignette support other research that
suggests an affront to the officers authority, such as a continuing argument, may
be a greater indicator of arrest than the legal characteristics of the situation.
Miller (2001) also explored the arrest of women for domestic violence. She
interviewed 37 professionals that included battered womens advocates from a
local shelter, victim advocates affiliated with a local police department, police
officers, treatment providers, probation officers, prosecutors, public defenders, and
family court advocates. Miller explored the questions of whether the increase of
arrests of women in a Delaware county might mean that women had a greater
willingness to use violence in their personal relationships than was previously
thought, or whether strict adherence to pro-arrest laws meant police were ignoring
the context of the domestic violence cases and arresting women who were acting
in self defense. Police departments included in the studies operated under a pro-
arrest policy that had been in place since 1998.
Miller (2001) found that none of the respondents believed that the use of
violence by women was actually increasing. Most dismissed the idea of mutually
10


violent couples and identified multiple reasons why women might use violence in
their relationships. Many felt that the change in policy now directed officers to
make arrests, and that officers were more focused on the violent incident itself,
ignoring the context within which it occurred. Even prosecutors and social workers
in the prosecutors office recognized these women as victims, and not as offenders.
Respondents in Millers study often felt that police were enforcing the letter of the
law and ignoring the spirit of the law.
Victim advocates and prosecutors claimed that a high number of the
women they interacted with had histories of victimization. A number of the
professionals interviewed in Millers (2001) study had either seen or heard about
batterers who had previously been arrested, and now used their experiences to
manipulate the system by behaving calmly when the police arrived, or by injuring
themselves and claiming it was the victim. Therapists reported hearing men in
perpetrator treatment groups state the importance of being the first to call the
police, which would increase the likelihood of the victims arrest, as a means to
further their own control of her. These men often threatened to call 911 again or
make a false report to the victims probation officer that she had violated her
probation.
Miller (2001) also found that police officers were often arresting both
parties to avoid doing a lengthy investigation; thus placing the burden on the
prosecutors office to sort out culpability. Many women were unfamiliar with the
11


legal system, and more likely to take a plea bargain without a full understanding of
the implications to pleading guilty to a charge of domestic violence. Advocates
relayed their experiences with women who often felt betrayed by lawyers who
seemed more interested injudicial efficiency than in defending the women they
were representing.
Factors Affecting Police Arrest Decisions
Other studies have excluded the arrests of women altogether or included
these arrests with only minimal analysis of their meaning or implications. This
research offers an opportunity to understand how mandatory arrest policies are
operationalized by police officers on scene. Researchers have looked at the
demographic, attitudinal and situational variables that have affected arrest, and
have employed analyses of police report data, surveys, and descriptive field
studies, in attempts to understand police response to domestic violence (Robinson
& Chandek, 2000).
Analysis of Police Reports
In their 1994 study, Bourg and Stock examined domestic violence arrest
reports from a sheriffs department that did not participate in any coordinated
community response to domestic violence and engaged in only minimal training
on the subject. Over a twelve-month period, 1,870 reports of domestic violence
12


resulted in 538 arrests (28.8%). The study found that felony assaults were most
likely to result in arrest, although only 34.7% of felony assaults ended in arrest. Of
the 28 sexual assault cases reported to the Sheriffs Department, only five ended in
arrest. AH those arrested for sexual assault were boyfriends; no husbands or ex-
husbands were arrested for sexual assault.
Bourg and Stock (1994) included the most in-depth look at gender and
arrest compared to other studies that have examined police report data. They found
that women were significantly more likely to be arrested on felony charges than
were men. Black women were significantly more likely to be arrested for felonies;
84.6% of Black women arrested were arrested on felony charges. White males, on
the other hand, were significantly less likely to be arrested for felony assault.
Bachman and Coker (1995) analyzed data from the National Crime
Victimization Survey for the years 1987-1992. Probability of arrest was affected
by victim injury, marital status, and race. Assaults involving Black victims and
Black offenders were almost twice as likely to result in arrest as white on white
victimizations. Surprisingly, offenders with prior histories of violence were less
likely to be arrested than offenders with no history of violence. Bachman and
Coker suggested this might be due to uncontrolled factors such as offender
aggression towards police. Batterers who have experienced prior police contact
may have learned how to behave appropriately.
13


In a study of police reports filed by 28 police departments in a Midwestern
county, Erez (1986) found that officers were more likely to make an arrest in cases
that involved ex-spouses compared to cases that involved married or dating
couples. This finding was surprising in light of further conclusions that a higher
severity of assault and injury was found in married and dating couples.
Burris and Jaffe (1983) investigated the effect of a newly instituted pro-
arrest policy on the behavior of officers in a Canadian police department.
Expectedly, they found a significant increase in the number of arrests after the
policy was instituted. The severity of police charges were significantly related to
whether or not the male party had been drinking, the seriousness of injury to the
victim, the need for hospital treatment of injury, and the criminal history of the
batterer.
In a review of 1,521 police reports, Jones and Belknap (1999) examined
factors related to the arrest and jailing of batterers in Boulder County, Colorado.
Almost 75% of the sample studied received some formal police response. This was
much higher than findings from any of the other studies, and may be because
Boulder County has been progressive in its response to domestic violence. Most
agencies in the county have had a pro-arrest policy in place since 1986, and an
oversight agency works to coordinate policy and provide feedback to departments.
Jones and Belknap (1999) found that the presence of drugs or alcohol,
children, or weapons increased the likelihood that police would make an arrest.
14


The police were less likely to make an arrest if there was or ever had been a formal
relationship between the couple. Dual arrests were made in 8.7% of the cases
included in the sample.
Eigenberg, Scarborough, and Kappeler (1996) examined police reports
over a five-year period in a mid-sized university town in order to compare police
response to domestic violence assaults with police response to non-domestic
assault (180 domestic violence assaults and 335 non-domestic assaults were
compared). Officers were significantly less likely to arrest in domestic assaults
than they were in non-domestic assaults. Arrest in cases of domestic assault was
significantly related to weapon use, the presence of witnesses beyond the victim,
the suspects presence on scene, and the victims request for arrest. The presence
of injuries was not significantly related to arrest.
Officer Survey Studies
A number of researchers have surveyed police officers, asking them to
describe how they would respond to vignettes describing an assault. Officers were
asked if th^j- would arrest or not, and what factors they would consider in making
their decisions. Other studies asked officers to rank, in order of importance, factors
effecting their decisions to arrest or not arrest when responding to a domestic
violence assault. Additionally, Gandolf and McFerron (1989) interviewed battered
women who had experienced police intervention.
15


Feder (1997) surveyed 297 officers in a Florida police department with a
newly implemented preferred arrest policy. The survey collected demographic
information on officers, officers attitudes towards women, officers use of
violence in their own homes, officer response to scenarios, and officer perceptions
about the effectiveness of police response in domestic assault cases. The surveys
revealed strong support for police intervention in domestic violence cases. Both
support for pro-arrest laws and the officer s use of violence in the home were
significantly related to their score on the attitudes towards womens roles section
of the survey.
Feders (1997) questions on officer attitudes towards women sought to
measure officers beliefs about womens roles at home and in public. Officer
responses were ranked on a continuum from traditional to liberal. Questions posed
to officers included: Women police are as good as men? Boys should be taught
more independence? And, better for man to be achiever? The majority of
officers disagreed with all three questions.
When responding to the scenarios, officers were least likely to make an
arrest when the victim had been cheating on the batterer or when the batterer had
recently lost his job. The officers attitudes towards women, knowledge of the
departments policy, and belief in the effectiveness of police intervention were
significantly related to their decisions to arrest. Although they were more likely to
16


arrest than not arrest in each of the scenarios, officers were most likely to arrest
when the batterer was drunk.
Holmes (1993) used in-depth interviews with officers, police call logs, and
information from supplementary homicide reports to reconstruct police response to
domestic violence in Massachusetts. The study found that the most frequent
outcome of police response was to advise the victim of her rights (19.6%). Arrest
took place in only 7.7% of cases. The speed with which the victim contacted the
police and the number of officers on scene were also related to arrest of the
offender. Violation of a restraining order had the strongest relationship to arrest in
this study. A history of police calls to the same address, the presence of a weapon,
the race of the offender, injury, victim preference, and the presence of children, all
had minimal, if any, impact on the decision to arrest.
Dolan, Hendricks, and Meagher (1986) surveyed police officers from three
Midwestern police departments-. Officers were asked to identify what factors they
considered important when making a decision to arrest, what other alternatives to
arrest they considered effective^ how they used referrals to social service agencies,
and how they evaluated the effectiveness of those referrals. Police ranked eleven
factors in order of influence on their decision to make an arrest: Factors that
indicated a direct threat to officers, such as use of violence toward an officer,
ranked highest. Indicators of a high level of seriousness of the assault and of a high
possibility of future assault ranked next. The commission of a felony, use of a
17


weapon, serious injury, and the likelihood of future assault were also ranked high
by police officers. The use of alcohol, disrespect to officers, history of assault, and
victim preference for arrest were ranked low on the list. Officers were also asked
to rank factors that played a role in their decision not to make an arrest. Officers
ranked twelve factors in order of importance. The top factors influencing the
officers decisions not to arrest were focused on the victims behavior, either
victim refusal to press charges at the scene, or the perception that victims often
drop charges later. After victim behavior, officers ranked the lack of serious injury,
commission of a misdemeanor, jail overcrowding, and intoxication of the victim
and assailant as no-arrest influences.
In both surveys, the size of the department made a difference in the
officers rankings. Officers from the larger department said they were much more
likely to make an arrest if there was a history of police response to the couple and
if the batterer was intoxicated. They were much less likely to arrest if the jail was
overcrowded and the victim was also intoxicated (Dolan, et al.,1986).
Gandolf and McFerron (1989) interviewed over 6,000 battered women who
had received shelter services in Texas. Nine categories of police response
possibilities were identified, including: did nothing, mediated, referred victim,
provided transportation, advised victim of legal rights, warned the batterer, forced
the batterer to leave, and arrested the batterer. Women who reported that no police
action was taken were actually more likely to have a weapon used against them
18


and to have their lives threatened, and their level of injury and experience of abuse
was similar to women whose batterers were arrested. Gandolf and McFerron
(1989) attribute this difference to the anti-social behavior of the batterers
themselves. More aggressive batterers were more likely to be arrested than other
batterers regardless of the severity of the specific assault police were responding
to. Batterers in the sample who exhibited aggressive behavior also had other
arrests, alcohol abuse issues, and were more likely to severely injure their victims.
Descriptive Field Studies
Other studies have involved the observation of police officers by trained
observers. Recording what officers actually do on scene has highlighted some
notable differences compared to survey research. Although officers ranked use of
alcohol, disrespect for officers, and victim preference low in surveys, these factors
often played an important role in their decision to arrest as illustrated by their
behavior on scene.
Ferraro (1989) observed officers from the police department in Phoenix,
Arizona, shortly after the department instituted a presumptive arrest policy.
Observers found that police were less likely to make an arrest when the offender
and victim were not married. Ferraro attributes this to the fact that the new pro-
arrest policy did not cover cohabitants. She did note, however, that even in cases
where one or more arrestable crimes not dependent on the relationship of the
19


victim and offender had been committed, such as destruction of property, the
police still did not make an arrest. In cases where the victim and offender were
married, police made arrests in a minority of the cases. Ferraro found that when
looking for probable cause to make an arrest, police often used felony standards of
evaluation; minor injury and property damage were not considered probable cause
for an arrest.
In spite of the presumptive arrest policy implemented by the Arizona police
department, legal criteria were not found to be the most important considerations
in police decision-making. Observers found ideological, practical, and political
concerns to be more powerful predictors of arrest. Police beliefs about deviant
citizens versus normal citizens (heterosexual, white, and English-speaking),
stereotypes of battered women as likely to drop charges and as masochistic, and
the idea that a mans home is his castle, supported the officers belief that arrest
was an invasion of privacy and that the new policy was an illogical policy
(Ferraro, 1989).
Practical considerations and convenience also played a role in officer
decisions to arrest. Batterers that fled the scene, uncooperative victims, and the
need for investigation to establish a crime all played a part in officer decisions not
to arrest. Having to involve mental health professionals and social services
because of the presence of children were seen as barriers to arrest. Although
officers in Arizona had incentives to respond with seriousness and intensity to
20


other non-domestic crimes, no professional benefits, such as promotions or
commendations, were attached to compliance with the arrest policy in cases of
domestic violence. This lack of positive incentive also undermined officer
valuation of the policy itself, with some seeing it as a move by police
administration to appease political supporters of the policy more than as an effort
to support victims (Ferraro, 1989).
Smith and Klein (1984) observed officers in 24 police departments
operating in three metropolitan areas. Interviews with neighborhood residents were
also completed. Cases selected by Smith and Klein included both domestic and
non-domestic assaults. The strongest indicator of arrest was a victims statement
that they wanted the offender arrested. Intoxication of the offender, the level of
aggressiveness toward the officers, and a history of police response were related to
arrest. Although higher levels of violence increased the likelihood of arrest, neither
the presence of injury nor the use of a weapon were significantly related to arrest.
Race was not found to have any relationship to the officers decision to arrest,
although arrests were more frequent in lower socio-economic neighborhoods.
Police were more likely to respect victims wishes for arrest in higher SES areas
compared to lower SES areas. Police were less likely to make arrests in domestic
disputes versus non-domestic disputes in higher SES neighborhoods. The overall
likelihood of arrest in lower SES areas was higher, and the rates of arrest for
domestic and non-domestic assaults were about the same.
21


In a replication of Berk and Losekes (1981) earlier study of police
response to domestic violence, Worden and Pollitz (1984) found that the strongest
indicator of arrest was whether or not a victim agreed to sign a complaint. The
probability of arrest was also increased if the batterer had been drinking and if the
victim made an allegation of violence. The likelihood of arrest increased
significantly, if the batterer was disrespectful towards the police officer.
Differences in officer perceptions about their roles as crime fighters or problem-
solvers was also important in how they evaluated the situation at hand, and showed
some evidence of effecting arrest patterns.
Studies Using Combined Methodology
In an attempt to create a more contextualized picture of police response to
domestic violence, Robinson and Chandek (2000) combined methodological
strategies to investigate factors influencing police arrest decisions. This study
attempted to clarify inconsistencies in variable definitions and the results of earlier
studies. Case summary reports and a supplemental survey that was completed on-
scene were combined with demographic data collected by the responding officers.
Data were collected over a 5-month period in a medium-sized metropolitan police
department. The presence of witnesses at the scene and co-habitation increased the
likelihood of arrest. If the suspect left the scene, the odds of arrest decreased
significantly. In fact, suspect presence had the most powerful effect on arrest.
22


CHAPTER3
METHODOLOGY
Data for the present study were collected from Boulder County, Colorado.
This study uses data from the same source used by Jones and Belknap (1999).
Police incident reports were reviewed and coded by the Domestic Abuse
Prevention Project, a progressive oversight agency that organizes the coordinated
community response in Boulder County.1 The sample is composed of 8,952 total
incidents responded to by eight law enforcement agencies in Boulder County
between January 1994 and January 2000.
Data for 32 cases where the sex of either the suspect or the victim was
unknown were excluded, as was data for incidents involving same-sex couples. It
is important to note, however, that in the six-year period covered by the data set,
only 80 cases involved female same-sex couples, compared to only 43 cases
involving male same-sex couples. It is unclear whether these numbers are high or
low because no reliable population numbers exist for the community. Given the
estimated rates of violence in same-sex couples, these numbers may signal a
dangerous and disheartening state of affairs for the safety of all victims of same-
1 For a more complete discussion of the history and function of DAPP see Jones, D.A., and
Belknap, J. (1999). Police responses to battering in a progressive pro-arrest jurisdiction. Justice
Quarterly. 12:2
23


sex intimate violence in Boulder County. It is likely that police responded
to more cases of same-sex violence than are included in this data set, but if those
cases were not coded as domestic violence they would neither fall into the purview
of the Domestic Abuse Prevention Project nor be given any legal or social
consideration as domestic violence cases.
The arrests of men were compared to the arrests of women to identify
differences and similarities in the factors that may affect police arrests in cases of
domestic violence. Factors were divided into two categories: victim-offender
characteristics and incident characteristics.
Victim-offender characteristics include demographic information for the
suspects, the suspects prior interactions with police in Boulder County, and the
victim-offender relationship. The data for prior police contact covers only prior
arrests in Boulder County. There are conceivably offenders with prior arrests
outside of Boulder County that would not be reflected in this data set. The victim-
offender relationship is coded in two variables. The variable formality refers to
the length and legal recognition of the relationship, while the measure of
togetherness refers to whether or not the victim and offender were together at the
time of the incident. (Jones & Belknap, 1999). Victims and offenders that were
married in the past or lived as common law spouses were coded as formal, while
ex-boyfriends or ex-girlfriends who had never been married were coded as
informal. Togetherness was coded as together if the couple was living together,
24


married, or dating at the time of the arrest, while those coded as not-together were
separated, divorced, or had formerly dated.
Incident characteristics include the presence of weapons during the
incident, which party contacted the police, whether or not children were present,
whether or not drugs or alcohol were present, and whether the assault was a
felony. Injury data was not available in this data set.
25


CHAPTER4
FINDINGS
Male suspects in the study were typically white, in their early thirties, and
had no prior police contact. Females were typically white, in their early thirties,
and also had no prior formal police contact. The majority of female suspects
(76.6%) and male suspects (70.4%) were involved in relationships that were
ongoing at the time of arrest. Similar numbers of male and female suspects had
formal relationships with their victims, and the majority were either married or
living together.
Similar numbers of female suspects (32.4%) and male suspects (33.3%)
were using drugs and alcohol at the time of the arrest. Female suspects used a
weapon in 14.7% of the incidents, while male suspects used a weapon in less that
10% of the incidents. Felony charges were filed in 5.8% of the cases involving a
female suspect, and 5% of the cases involving a male suspect. Of the 8952 total
cases, 7044 (78.7%) resulted in arrest, while 1908 (21.3%) did not result in arrest.
These numbers are higher than those numbers cited by other researchers, with the
exception of Jones and Belknap (1999) who studied this same jurisdiction. The
difference can most likely be attributed to the length of time a mandatory arrest
policy has been in place in these departments, and the presence of a coordinated
26


community response that includes support of the administration of the departments
in the county.
Table 4.1
Victim Offender Characteristics
Female Suspect Male Suspect
N % (n) N % (n)
Suspect Race Black 1988 1.80 36 6964 4.20 292
Hispanic 6.10 122 13.30 924
White 86.40 1717 78.80 5490
Asian 109.00 38 1.20 83
American Indian 0.20 4 0.20 16
Other 0.30 7 0.30 20
Unknown 4.00 64 1.80 125
Victim Race 1988 6964
Black 2.80 55 1.00 72
Hispanic 6.60 131 6.00 419
White 60.40 1201 52.00 3621
Asian 0.90 17 1.00 72
American Indian 0.00 0 0.20 11
Other 0.30 5 0.30 13
Unknown 24.70 492 34.60 2407
Victim-Offender Rel. Married 1959 31.50 627 6816 31.20 2170
Common Law Married 1.60 32 1.00 70
Dating 13.80 274 13.40 930
Divorced 4.00 79 4.00 277
Ex-dating/Ex-living Together 12.80 254 17.80 1242
Living Together 28.60 568 23.40 1628
Separated 6.30 125 7.20 499
Couple together? 1959 . 6816
Yes 75.50 1501 68.90 4798
No 23.00 458 29.00 2018
27


Table 4.2
Incident Characteristics
Female Suspect N % (n) Male Suspect N % (n)
Law Enforcement Response 1988 6964
Arrested 80.20 1594 78.30 5450
Not Arrested 19.80 394 21.70 1554
Dual Arrest? 1641 6964
Yes 77.30 1269 6.60 384
No 22.70 372 93.40 5469
Suspect using drugs? 1988 6964
Yes 32.40 644 33.30 2316
No 23.80 474 19.80 1381
Unknown 43.80 870 46.90 3267
Victim using drugs? 1988 6964
Yes 8.30 165 4.80 332
No 20.20 401 18.30 1272
Unknown 71.50 1422 76.90 5360
Weapons Used? 1988 6964
Yes 14.70 292 8.90 617
No 18.10 360 18.60 1294
Unknown 67.20 1336 72.60 5053
Felony Charges? 1988 6964
Yes 5.80 116 5.00 347
No 57.70 1147 22.50 1566
Unknown 36.50 725 72.50 5051
Children Present? 1988 6964
Yes 27.10 538 29.80 2076
No 47.10 937 40.70 2832
Unknown 25.80 513 29.50 2056
Pearsons Chi-Square
Cross-tabular analysis of factors affecting arrest supported the findings of
many studies that some extra legal variables play an important role in police
28


officers decisions to arrest. Table 4.3 shows victim-offender characteristic
variables related to arrest for both male and female suspects.
Table 4.3
Victim Offender Characteristics: Chi-Square Test
N Not arrested Arrested x2
% n % n
Prior Formal Relationship With Victim Female SusDects No Prior Formal Relationship with Victim 1959 16.90 (185) 83.10 (911) 12.54 ***
Prior Formal Relationship with Victim 23.30 (201) 76.70 (622)
Male Susoects No Prior Formal Relationship with Victim 6816 20.20 (769) 79.80 (3031) 12.33 ***
Prior Formal Relationship with Victim 23.80 (717) 76.20 (2299)
Couple Together at Time of Arrest Female SusDects Couple Not Together at Time of Anest 1959 24.50 (112) 75.50 (346) 8.53 **
Couple together at Time of Arrest 18.30 (274) 81.70 (1227)
Male SusDects Couple Not Together at Time of Arrest 6816 28.80 (581) 71.2 (1437) 82.142 ***
Couple Together at Time of Arrest 18.90 (905) 81.100 (3893)
* p < .05
** p < .01
*** p < .001
In this study, the sex of the suspect was not significantly related to arrest.
Female suspects were as likely as male suspects to be arrested. The race of the
suspect was not significant!y- related to arrest for either males or females.
Both male and female suspects who had some prior formal relationship
with their victims were significantly less likely to be arrested than those suspects
that did not have a prior formal relationship with their victims. This is consistent
with prior findings (Jones & Belknap, 1999). Suspects who were in couples that
29


were together at the time of arrest were more likely to be arrested than those
suspects not together with their victims.
Consistent with other studies, suspects using drugs or alcohol at the time of
the incident were significantly more likely to be arrested than those that were not.
This was true for both male and female suspects. Interestingly, suspects whose
victims were using drugs at the time of arrest were also significantly more likely to
be arrested. This was true for both male and female suspects.
Weapons were significantly related to arrest for both male and female
suspects. Weapons in this study included anything other than the suspects hands
or feet. Obvious weapons such as guns and knives, as well as, weapons of
convenience such as ashtrays, glasses, or golf clubs that may have been thrown or
used to hit were coded.
Surprisingly, the presence of children on scene was only significantly
related to arrest for male suspects. This is an interesting result given the
predominant social view that women are primarily responsible for the safety of
their children. It would be expected that women using violence in front of their
children are deemed more responsible than men who use violence in front of their
children. The local battered womens shelters have worked closely with police
departments and the department of social services to raise awareness about
working with children exposed to domestic violence. This may mean that police
officers who are noting the presence of children on scene are better educated about
30


domestic violence in general, and working with child witnesses in particular. They
may be doing a more thorough investigation in order to protect the children in the
long run. These officers may also be better at interviewing children, and may be
gathering evidence from them that implicates the male party. Further analysis
would be necessary to fully understand what is behind this result.
Table 4.4
Incident Characteristics: Chi-Square Test
N Not arrested % n Arrested % n x2
Victim Using Alcohol/Drugs at Time of Arrest
Female Suspects 566 5.91 **
Victim Not Using Alcohol/Drugs 27.9 (112) 72.1 (289)
Victim Using Alcohol/Drugs 18.2 (30) 81.8 (135)
Male Suspects 1604
Victim Not Using Alcohol/Drugs 30.3 (385) 69.7 (887) 12.44 **
Victim Using Alcohol/Drugs 20.5 (68) 79.5 (264)
31


Table 4.4 (Continued)
Incident Characteristics: Chi-Souare Test
N Not arrested Arrested x2
% n % %
Suspect Using AIcohol/Drugs at Time of Arrest Female Susoects Suspect Not Using AIcohol/Drugs 1118 24.3 (115) 75.7 (359) 18.01 ***
Suspect Using AIcohol/Drugs 14.3 (92) 85.7 (552)
Male Susnects Suspect Not Using AIcohol/Drugs 3697 27.1 (374) 72.9 (1007) 93.87 ***
Suspect Using AIcohol/Drugs 14.2 (328) 85.8 (1988)
Presence of Children at the Scene Female SusDects Children Not Present 1475 21 (197) 79 (740) 3.024
Children Present 17.3 (93) 82.7 (445)
Male Susnects Children Not Present 4908 23.4 (662) 76.6 (2170) 31.08 ***
Children Present 16.9 (350) 83.1 (1726)
Weapon Present Female SusDects No Weapon 652 71.7 (258) 28.3 (102) 286.62 ***
Weapon 5.8 (17) 94.2 (275)
Male SusDects No Weapon 1911 67.3 (871) 32.7 (423) 586.58 ***
Weapon 8.1 (50) 91.9 (567)
Suspect Prior Open Cases Female SusDects No Open Cases 1970 18.8 (327) 81.2 (1414) 4.157*
Open Cases 24.5 . (56) 75.5 (173)
Male Susnects No Open Cases 6899 20.5 (1223) 79.5 (4751) 23.24 ***
Open Cases 27.5 (254) 72.5 (671)
* p < .05
**p<.01
*** p < .001
32


A history of domestic violence arrests was not significantly related to arrest
for either female or male suspects, although the existence of open cases was found
to decrease the likelihood of arrest for both female and male suspects. This
finding was surprising, but may be related to officer perception that the victim is
not willing to separate from the suspect. Officer surveys found that victim
behavior was a strong influence in decisions not to arrest (Dolon et al., 1986).
Officer perception that some victims are not willing to help themselves by leaving
a relationship may be linked to assumptions about whether the victim will follow
through with the case.
The commission of a felony was not significantly related to arrest in this
study. This is inconsistent with other studies and may be related to limitations with
the data. The information on charges in this data set was limited and felony arrests
made up a very small number of overall arrests.
The data in this study does show significant differences in the victimization
experiences of men and women suspects. Just fewer than 15% of the male victims
of female suspects had prior arrests for domestic violence, compared to only 10%
of the victims of male defendants. Almost 25% of the female suspects had a prior
victimization history compared to only 9% of the male suspects.
33


Multivariate Analysis
Table 4.5 shows the result of a multivariate analysis using a logistic
regression model. Arrest was used as the dependent variable in the model. The
sample size for the multivariate analysis was decreased considerably because of
the amount of missing data. A considerable amount of felony data was missing, so
this variable was excluded from the analysis. Analysis was done on 239 cases
involving a female suspect and 584 cases involving a male suspect.
Table 4.5
Logistic Regression Predicting Arrest
Female Suspects Exp(B) S. E. Male Suspects Exp (B) S. E.
Incident Characteristics
Use of a Weapon 49.66 0.52** 24.98 0.29**
Drug/AJcoh'ol Use by Suspect 0.89 0.38 1.48 0.23
Children Present 2.60 0.44* 1.16 0.24
Who Contacted Police 0.99 0.39 0.37 0.23**
Victim-Offender Characteristics
Suspect Prior Victim 1.03 0.53 1.61 0.57
Suspect Prior DV 3.05 0.50* 1.58 0.25
Suspect Prior Open Case 0.56 0.64 0.77 0.31
Suspect Race 0.89 0.56 0.80 0.26
Victim Race 0.98 0.14 1.04 0.07
Formal Relationship 1.18 0.38 1.00 0.23
Couple Together 1.48 0.44 0.54 0.28*
* p < .05
** P < 01
*** p< .001
34


As expected, the presence of a weapon increased the likelihood of arrest for
both male and female suspects. If the couple were together at the time of the arrest,
the police were less likely to charge the male suspect. This finding is consistent
with other studies that found ex-spouses more likely to be arrested than married or
dating couples. The difference in results for male and female suspects may be
related to victim-blaming biases held by officers about why battered women do or
do not leave an abusive relationship. This finding was also consistent with officer
survey studies that identified the female victims behavior as a strong influence in
police decisions not to arrest (Dolon, et. al, 1986).
The presence of children on scene was found to significantly increase the
likelihood of arrest for female suspects, but did not remain significant for male
suspects. This result more accurately reflects societal attitudes that hold women
responsible for violence perpetrated against children by their chosen intimate
partners. More research would need to be done to fully understand the relationship
between the presence of children and arrest.
The use of drugs and alcohol did not remain significantly related to arrest
for either male or female suspects, but was approaching significance for males. A
criminal history of domestic violence was significantly related to an increase in the
likelihood of arrest for female suspects, but not for males. This result may be due
to social views of men and women. Violence is often a celebrated cultural value
35


for men in the United States, while women who use violence deviate profoundly
from social norms.
36


CHAPTER 5
CONCLUSION
Assertions that women use violence in interpersonal relationships as much
or more than men do have often been met with a challenge to contextualize the
violence being studied. Only a qualified understanding of the violence in a
relationship can be obtained by counting the number of physical assaults. A large
body of research has now been produced that illustrates profound differences in
the ways women and men use and experience violence. The purpose of this
research study has been to argue for a similar contextual understanding of
domestic violence arrests. Data collected from Boulder County, Colorado,
illustrate the differences in factors that increase the likelihood of arrest for male
and female suspects. It is clear from this gender analysis that profound differences
exist in police decision-making regarding arrest. This study demonstrates that
women experience state intervention differently than men; they are at greater risk
of being arrested if their children are on the scene, or if they have a history of
police intervention for domestic violence. Women who are victims of domestic
violence are less likely to be protected by the police if there is an open case
pending against their batterers.
37


Understanding the why and how police officers make decisions to arrest is
important for advocates engaged in systems change work and public policy
refinement. As the numbers of women arrested for domestic violence increases,
advocates, researchers, and policy makers must try to understand why and how
these arrests are happening, and to integrate this knowledge with what we are
learning about gender differences in the use and meaning of violence.
The differences in offending and victimization histories of the men and
women in this data set give some urgency to refining research on criminal legal
interventions into domestic violence to include an effective gender analysis. It will
be important for future research to include voices of women that have been
arrested, especially when advocates identify these women as victims of domestic
violence, and not as perpetrators.
Researchers and advocates who have begun to build bridges between the
womens anti-violence movement and the prison abolition movement have laid the
groundwork for a more complex understanding of the intersection of interpersonal
and state control in womens lives. The number of women arrested for domestic
violence is increasing in many jurisdictions at the same time as the general
womens prison population is increasing. Given that arrest for domestic violence is
a relatively more recent way for women to enter the criminal justice system, and
that a womans history of domestic violence increases the likelihood for
38


subsequent arrest, this issue will become an increasingly important consideration
for criminologists.
39


RESOURCES
Bachman, R., & Coker, A. (1995). Police involvement in domestic violence: The
interactive effects of victim injury, offenders history of violence, and race.
Violence and Victims, 10, 91-106.
Belknap, J., & McCall, K. D. (1994). Woman battering and police referrals.
Journal of Criminal Justice, 22, 223-236.
Belknap, J. (1996). Law enforcement officers attitudes about the appropriate
responses to woman battering. In The Invisible Woman: Gender, Crime,
and Justice. Belmont, CA: Wadsworth.
Berk, S., & Loseke, D. R. (1981). Handling family violence: Situational
determinants of police arrest in domestic disturbance! Law and Society,
15(317-346).
Berk, R., & Newton, P. J. (1985). Does arrest really deter wife battery? An effort
to replicate the findings of the Minneapolis spouse abuse experiment.
American Sociological Review, 50,253-262.
Bhattachaijee, A. (2001). Whose safety? Women of color and the violence of law
enforcement: American Friends Service Committee, and the Committee on
Women, Population, and the Environment.
Bourg, S., & Stock, H. V. (1994). A review of domestic violence arrest statistics in
a police department using a pro-arrest policy. Journal of Family Violence,
9(177-189).
Bowker, L. H. (1982). Police services to battered women: Bad or not so bad?
Criminal Justice and Behavior, 9,476-494.
Bowman, C. G. (1992). The arrest experiments: A feminist critique. Journal of
Criminal Law and Criminology, 83, 201-209.
Brown, S. E. (1984). Police responses to wife beating: Neglect of a crime of
violence. Journal of Criminal Justice, 12, 277-288.
40


Burris, C. A., & Jaffe, P. (1983). Wife abuse as a crime: The impact of police
laying charges. Canadian Journal of Criminology, 25.
Buzawa, E., & Austin, T. (1993). Determining police response to domestic
violence victims:The role of victim preferences. American Behavioral
Scientist, 36(5), 610-623.
Buzawa, E. S., & Buzawa, C. G. (2002). Domestic violence: The criminal justice
response (3rd ed.). Thousand Oaks, CA: Sage.
Das Gupta, S. D. (2002). A framework for understanding women's use of nonlethal
violence in intimate heterosexual relationships. Violence Against Women,
5(11), 1364-1389.
DeKeseredy, W. S., & Schwartz, M. (1988). Woman abuse on campus: Results
from the Canadian National Survey. Thousand Oaks: Sage.
Dobash, R. P., Dobash, R. E., Wilson, M., & Daly, M. (1992). The myth of sexual
symmetry in marital violence. Social Problems, 59(71-91).
Dolon, R., Hendricks, J., & Meagher, M. S. (1986). Police practices and attitudes
towards domestic violence. Journal of Police Science and Administration,
14,187-192.
Eigenberg, E. M., Scarborough, K.E., Kappeler, V.E. (2001). Contributory factors
affecting arrest in domestic and non-domestic assaults. In H. M. Eigenberg
(Ed.), Woman Battering in the United States (pp. 298-318). Prospect
Heights, Illinois: Waveland Press.
Erez, E. (1986). Intimacy, violence, and the police. Human Relations, 39,265-281.
Feder, L. (1997). Domestic violence and police response in a pro-arrest
jurisdiction. Women and Criminal Justice, 8, 79-97.
Ferraro, K. J. (1989). Policing woman battering. Social Problems, 36, 61-74.
Gandolf, E. W., & McFerron, J. R. (1989). Handling battering men: police action
in wife abuse cases. Criminal Justice and Behavior, 16, 429-439.
41


Gelles, R., & Loseke, D. (1993). Current controversies in family violence.
Newbury Park, CA: Sage.
Gondolf, E. W. (1988). Battered women as survivors: An alternative to treating
learned helplessness. New York: Lexicon Books.
Hamberger, L. K., Guse, C.E. (2002). Men's and women's use of intimate partner
violence in clinical samples. Violence Against Women, 5(11), 1301-1331.
Hart, B. (1993). Battered women and the criminal justice system. American
Behavioral Scientist, 83, 73-119.
Hirschel, D., & Buzawa, E. (2002). Understanding the context of dual arrest with
directions for future research. Violence Against Women, 5(12), 1449-1473.
Holmes, W. (1993). Police arrests for domestic violence. American Journal of
Police, 12(4), 101-125.
Hooper, M. (1996). When domestic violence diversion is no longer and option:
What to do with the female offender. Berkeley Women's Law Journal, 11,
168-181.
Johnson, M. (1995). Patriarchal terrorism and common couple violence: two forms
of violence against women. Journal of Marriage and The Family, 57, 283-
294.
Jones, D. A., & Belknap, J. (1999). Police responses to battering in a progressive
jurisdiction. Justice Quarterly, 16(2), 249-273.
Manless, M. (1996). Mandatory arrest: A step toward eradicating domestic
violence, but is it enough? University of Illinois Law Review, 2, 533-587.
Martin, M. (1997). Double your trouble: Dual arrest in family violence. Journal of
Family Violence, 12,139-157.
Miller, S. L. (1989). Unintended side effects of pro-arrest policies and their race
and class implications for battered women: A cautionary note. Criminal
Justice Policy Review, 3(3), 299-317.
Miller, S. L. (2001). The paradox of women arrested for domestic violence.
Violence Against Women, 7, 1339-1376.
42


Mills, L. G. (1999). Commentary: Killing her softly: Intimate abuse and the
violence of state intervention. Harvard Law Review, 113 Harvard Law
Review 550.
Oppenlander, N. (1982). Coping or copping out. Criminology, 20,449-465.
Pleck, E. (1987). Domestic tyranny: The making of social policy against family
violence from colonial times to the present. New York: Oxford University
Press.
Ritchie, B. E. (1996). Compelled to crime: The gender entrapment of battered
black women. New York: Routledge.
Ruttenberg, M. H. (1994). Feminist critique of mandatory arrest: An analysis of
race and gender in domestic violence policy. Gender and Law, 2, 171-199.
Saunders, D. G. (1995). The tendency to arrest victims of domestic violence.
Journal of Interpersonal Violence, 10,147-158.
Saunders, D. G. (1996). When battered women use violence: Husband abuse or
self-defense? Violence and Victims, 1,47-60.
Sherman, L. W., & Berk, R. (1984). The specific deterrent effects of arrest for
domestic assault. American Sociological Review, 49, 261-272.
Sherman, L. W. (1992). The influence of criminology on criminal law: Evaluating
arrests for misdemeanor domestic violence. Journal of Criminal Law and
Criminology, 83, 1-45.
Smith, D. A., & Klein, J. R. (1984). Police control of interpersonal disputes. Social
Problems, 31, 469-481.
Stanko, E. A. (1989). Missing the mark? Policing battering. In J. Hanmer, J.
Radford & E. A. Stanko (Eds.), Women, policing, and male violence (pp.
46-69). London: Routledge.
Stark, E. (1993). Mandatory arrest of batterers: A reply to its critics. American
Behavioral Scientist, 36, 651-680.
43


Stark, E. (1995). Re-presenting woman battering: From battered woman syndrome
to coercive control. Albany Law Review, 58, 973-1026.
Steinmetz, S. K. (1977-1978). The battered husband syndrome. Victimology: An
International Journal, 2,499-509.
Stets, J. E., & Straus, M. A. (1990). The marriage license as a hitting license: A
comparison of assaults in dating, cohabitating, and married couples. In M.
A. Straus & R. J. Gelles (Eds.), Physical violence in American families (pp.
227-244). New Brunswick: Transaction Publishing.
Straus, M. A., Gelles, R., & Steinmetz, S. K. (1980). Behind closed doors:
Violence in the American family. New York: Anchor.
Straus, M. A., & Gelles, R. (1986). Societal change and change in family violence
from 1975 to 1985 as revealed by two national surveys. Journal of
Marriage and The Family, 48, 465-479.
Waaland, P., & Keeley, S. (1985). Police decision-making in wife abuse: The
impact of legal and extra legal factors. Law and Human Behavior, 9, 355-
366.
Worden, R. E., & Pollitz, A. A. (1984). Police arrests in domestic disturbances.
Law and Society Review, 18,105-119.
Zorza, J. (1992). The criminal law of misdemeanor domestic violence, 1970-1990.
The Journal of Criminal Law and Criminology, 55(1), 46-72.
Zorza, J., & Woods, L. (1994). Mandatory arrest: Problems and possibilities.
Washington DC: Center on Women and Family Law.
44


Full Text

PAGE 1

FEMALE DOMESTIC VIOLENCE ARRESTS IN A PRO-ARREST JURISDICTION by Catherine L. Guerrero B.A., George Mason University, 1992 A thesis submitted to the University of Colorado at Denver in partial fulfillment. of the requirements for the degree of Master ofPublic Administration 2003

PAGE 2

Thesis for Master of Public Administration degree by Catherine L. Guerrero has been approved by Barbara Paradiso {!]ltif l. d dV 5 7 Date

PAGE 3

Guerrero, Catherine L. (M.P.A., Graduate School of Public Affairs) Female Domestic Violence Arrests in a Pro-Arrest Jurisdiction Thesis directed by Associate Professor Mary Dodge ABSTRACT Data for this study were collected in Boulder County, Colorado by the Domestic Abuse Prevention Project. Data were coded from police reports gathered from 8 law enforcement agencies operating in Boulder County. One thousand nine hundred eighty eight cases involving female suspects were compared to 6964 cases involving male suspects to determine differences and similarities in factors contributing to police arrest decisions. A Pearson's chi square analysis was run to identify important relationships. A subsequent logistic regression was run to further investigate the effects of variables on arrest. A formal police response occurred in 78.7% of cases. The data shows differences in the. victimization experiences ofwomen arrested for domestic violence as compared to male suspects. Fewer than 15% of the male victims of female suspects had prior arrests for domestic violence, compared to only 10% of the victims of male suspects. Almost 25% of female suspects had prior victimization history, compared to only 9% of the male suspects. Multivariate analysis was run on 239 c;:tSes involving a female suspect and 584 cases involving a male suspect. The presence of a weapon increased the likelihood of arrest for both males and females. The presence of children increased the likelihood of arrest for female suspects, as did a prior history of domestic violence. Police were less likely to arrest male suspects if the couple was still in a relationship at the time of the arrest. Other research and. anecdotal evidence for women's different experiences and motivations for violence, as well.as findings about the relationship between women's victimization and their use of self defensive violence, support the possibility that many of the women being arrested are actually victims of domestic violence. The current study shows significant differences in the way that police officers in Boulder County, Colorado, respond to male and female suspects in domestic violence cases. This abstract accurately represents the content of the candidate's thesis. I recommend its publication. 111

PAGE 4

CONTENTS Tables ......................................................................................................... vi CHAPTER 1. INTRODUCTION ................................................................................... Current Debate ................................................................................. 3 Women's Use ofViolence .................................................. .4 Mandatory Arrest ................................................................. 5 Unintended Consequences of Mandatory Arrest ............................. 7 2. REVIEW OF THE LITERATURE ......................................................... 9 Arrests ofWomen ............................................................................ 9 Factors Affecting Police Arrest Decisions ..................................... l2 Analysis ofPolice Reports ................................................. 12 Officer Survey Studies ....................................................... l5 Descriptive Field Studies .................................................. .19 Studies Using Combined Methodology ............................. 22 3. METHODOLOGY ................................................................................ 23 4. FINDINGS ............................................................................................. 26 Pearson's Chi-Square ..................................................................... 28 Multivariate Analysis ..................................................................... 33 iv

PAGE 5

5. CONCLUSION ................ .......... ........................................................... 37 RESOURCES ........................................................................................................ 40 v

PAGE 6

TABLES Table 4.1 Victim Offender Characteristics ...................................................................... 27 4.2 Incident Characteristics .................................................................................... 28 4.3 Victim Offender Characteristics: Chi-Square Test .......................................... 29 4.4 Incident Characteristics: Chi-Square Test ....................................................... 31 4.5 Logistic Regression Predicting Arrest ............................................................ .34 VI

PAGE 7

CHAPTER 1 INTRODUCTION Until recently, research concerning arrest in cases of domestic violence and inquiry into women's use ofviolence in intimate relationships has followed parallel tracks. As more jurisdictions report staggering increases in the number of women arrested for domestic violence, there has been an increased focus on the response ofthe criminal legal system to WQmen's .use ofviolence in intimate relationships . J1le majority of these studies have focused only on women arrested as part of a dual arrest, in which both male and female parties are presumably the 'co combatants' or 'perpetrators' of domestic violence. These studies share the common hypothesis that women who are arrested as. part of a dual arrest are more often than not self-defending victims. The underlying assumption is that high rates of dual arrest indicate poor policing, whereas low rates of dual arrests indicate effective investigation teclmiques and appropriate It is important to understand in what circumstances inappropriate arrests of self-defending victims are being made. It is necessary, however, for researchers to look deeper into the nuances of these arrests to understand all of the circumstances in which women are subject to criminal legal sanctions for domestic violence. Arguments for a 1

PAGE 8

contextualized approach to women's use of violence should be applied to the arrests of all women, and the context should be expanded to include the response ofthe criminal legal system. Failing to apply a gender analysis to our investigations into what factors influence officer decisions to arrest in cases of domestic violence diminishes the ability of scholars and advocates to understand fully how the criminal legal system is responding to women's use of violence. Integrating our growing knowledge of how and why women use violence into public policy discussions has laid the groundwork for a challenge to the idea that women are as violent as men. Without the inclusion of a gender analysis in evaluations of public policies such as mandatory and pro arrest laws, however, these arguments are doomed to partiality. An increasing sophistication in our conversations about how women are affected by interpersonal, as well as, state violence is necessary if we are to create a public policy response to domestic violence that can work to increase the safety of many more victims. This is not to say that the arrests of women are not justified in any case, or that women are always victims in cases of domestic violence. A fully contextual understanding ofthe effects of violence requires consideration ofthe interaction of interpersonal violence and structural oppression in analyses of women's actions in ways that are different from men's. The research and anecdotal evidence for women's experiences and motivations for violence, as well as findings about the relationship between women's victimization and their use of self-2

PAGE 9

defensive violence, support investigation into the growing number of women arrested for domestic violence (DeKeseredy & Schwartz, 1998). Saunders (1986) found that 71% of battered women arrested for domestic violence used violence in self-defense. The current study uses a quantitative approach to investigate the arrests of women in a jurisdiction with a strong mandatory arrest policy. Characteristics of the victim-offender relationship and characteristics of the incident that resulted in arrest are considered. A Pearson's Chi-Square test was used to identify significant relationships, and a logistic regression was then used to further investigate the nature of those relationships. Results for women arrested for domestic violence were compared to results for men in order to gain an understanding of the differences and similarities between police response to the two groups. Current Debate Debate surrounding women's use of violence in intimate relationships has centered on two major themes over the last thirty years. First, concern about whether women, in general, are as violent and aggressive in their intimate relationships as men are has produced a large body of scholarship describing gender differences in motivations, perceptions, and outcomes of violence in intimate relationships. A second, concurrent debate has focused on the high 3

PAGE 10

numbers of women arrested as a result of dual arrests, or as single suspects, has been raging among police and advocates (Das Gupta, 2002). Women's Use of Violence Many studies have concluded that women use violence in intimate relationships at the same or higher rates than men (Steinmetz, 1977-1978; Straus & Gelles, 1986, 1990; Strauss, Gelles, & Steinmetz, 1980). Most advocates and however, disagree with the idea that women are as violent as men, and argue that simply counting up the number of hits can mislead our attempts to fully understand the experience of violence in intimate relationships (Das Gupta, 2002; DeKeseredy & Schwartz, 1998). Without a full accounting for context, including the motivations and outcomes of violent incidents, only a partial understanding of violence in intimate relationships can be. attained (DeKeseredy & Schwartz, 1998). Gender differences in motivation for the use of violence are well documented in the literature. Women report using violence in self-defense at higher rates than men do, and their use of violence in self-defense goes up in relationto their rate of victimization (DeKeseredy & Schwartz, 1998). Women often use violence in an attempt to regain some control over a situation and out of frustration and anger (Saunders, 1986). The categorizing of women's violence as self-defense,. retaliation, or coercive by researchers has not been representative of many women's experiences; simply describing women's use of violence as self-4

PAGE 11

defense may be an oversimplification. Often, women do not distinguish between "self-defense" and "fighting back," but instead describe complex responses that include reactions to the immediate situation, as well as resistance to their experiences of domination over time (Saunders, 1986). Scholars and advocates also point to the presence or absence of fear as an important indicator of culpability in violent relationships (Hamburger & 9use, 2002). Mandatory Arrest More recently, debate about the high numbers of women arrested for domestic violence has grown out of conflicts over the efficacy of arrest and the implementation of proand mandatory arrest laws. It is we11 documented that woman battering was traditionally seen as a private matter, outside the domain of public intervention. Police were taught.not to arrest, but instead to attempt to de escalate the situation by walking the batterer around the block, attempting to mediate for the couple, or removing either the batterer or the victim from the scene. Victims often were actively discouraged from seeking arrest or prosecution (Zorza & Woods, 1994). When officers did arrest the batterer, it was usually in response to aggression towards the officers, an act seen as a threat to police authority. Arrest was used as a tool for the officer to maintain control of the situation, rather than as a response to the assault itself (Buzawa & Buzawa, 2000). 5

PAGE 12

-A number of social and legal challenges in the 1970s and 1980s led to changes in police response to domestic violence, including a broader public understanding of the issue, and greater judicial and legislative action, Activists and advocates began public education campaigns that focused attention on the experiences of battered women. As an outgrowth of feminist activism in the United States, the battered women's movement was able to capitalize on quickly changing views about women's roles in society (Zorza & Woods, 1994). Activists also focused public pressure on legislatures and police departments to better protect battered women. Initial laws that focused specifically on the police response to woman battering were introduced in the early 1970s, and in 1977 the first law imposing a mandatory duty to arrest in cases where probable cause existed was passed in Oregon (Zorza & Woods, 1994). Feminist calls for police accountability also laid the groundwork for a series of court cases, culminating in the highly publicized case of Tracy Thurman, a woman who was severely injured by her husband while police were on the scene (Zorza & Woods, 1994).' As huge sums of money were awarded to battered women and their attorneys, police departments and municipalities began to recognize the need to change their policies in order to protect themselves from lawsuits. (Belknap, 1996; Buzawa & Buzawa, 2000; Zorza, 1992) In amid debates about police response to woman battering, researchers in Minnesota released results from the "Minneapolis experiments" that 6

PAGE 13

claimed arrest was the most effective police response in deterring further assaults (Sherman & Berk, 1984). Six subsequent replication studies offered mixed results compared to the Minneapolis findings, although most seemed to show that arrest was an effective deterrent in some cases (Zorza, 1992). Following the release of the Minneapolis arrest study, police departments around the country began to change their policies to encourage or require arrest in domestic violence cases (Zorza & Woods, ). Unintended Consequences of Mandatory Arrest The accomplishments of the battered women's movement over the last 30 years have been significant. Profound changes have occurred in public awareness, legal recognition, police response, and funding availability. It is clear, however, that these shifts in public response to woman battering have not come without costs. Mandatory and pro-arrest policies have been criticized on a number of fronts. Many advocates and researchers have criticized mandatory arrest policies on the basb that they take decision-making power away from battered women in ways that are ultimately disempowering and patronizing (Buzawa & Buzawa, 2000). There is also concern that men of color and poor men may be disproportionately arrested under these laws (Coker, 2001; Miller, 1989). Poor women and women of color call the police more often to resolve conflicts and 7

PAGE 14

problems than do white middle and upper class women (Miller, 1989). Women with more resow-ces may be able to use friends and hotels as temporary shelter, and may be able to leave relationships in ways that allow them to avoid criminal legal intervention (Miller, 2001). The economic consequences of arrest, such as the loss of a job and the burden of court costs and fines, may also have a more profound effect on poor families (Coker, 2001; Miller, 1989). The drastiC increase in the numbers of women arrested for domestic violence in many communities has been cause for concern among domestic violence advocates. Although the arrest of batterers increased after the implementation ofthese laws, so did the number of women being arrested as batterers, either alone or as part of dual arrests (Jones & Belknap, 1999; Zorza & Woods, 1994). 8

PAGE 15

CHAPTER2 REVIEW OF THE LITERA TORE There is a growing body of research that is focused on the factors that affect the arrest decisions of police officers, although only a handful of these specifically focus on the arrest of women. This research offers a unique opportunity to understand how police officers operationalize mandatory arrest policies on scene, and how battered women's advocates and criminal justice personnel view the arrests of women. Arrests of Women Saunders (1995) surveyed 111 officers from three urban and seven rural police departments in Wiscor:'sin. !he survey was conducted prior to the passage or implementation of mandatory arrest laws or policies. Officers were given two vignettes and asked to choose from ten responses. that included: referral of each party, warning one or the other party, and arrest of one or both of the parties. Ir. ihe first scenario presented to officers, the argument ended when the police arrived on scene. In the second scenario, the couple continued to argue after the officers arrived on scene. Officers were much more likely to arrest the victim in the 9

PAGE 16

scenario in which the argument continued. Those officers who responded that they would arrest the woman were more apt to believe that a man is justified in assaulting a woman if she cheats on him and less likely to believe that women stay in violent relationships because of practical barriers to leaving. These officers also reported a higher level of discomfort when interacting with victims. Comments written by officers in response to the second vignette support other research that suggests an affront to the officer's authority, such as a continuing argument, may be a greater indicator of arrest than the legal characteristics of the situation. Miller (2001) also explored the arrest of women for domestic violence. She interviewed 37 professionals that included battered women's advocates from a -local shelter, victim advocates affiliated with a local police department, police officers, treatment-providers, probation officers, prosecutors, public defenders, and family court advocates. Miller explored the questions of whether the increase of arrests of women in a Delaware county might mean that women had a greater willingness to use violence in their personal relationships than was previously thought, or whether strict adherence to pro-arrest laws meant police were ignoring the context of the domestic violence cases and arresting women who were acting in self defense. Police departments included in the studies operated under a proarrest policy that had been in place since 1998. -Miller (2001) found that none ofthe respondents believed that the use of violence by women was actually increasing. Most dismissed the idea of mutually 10

PAGE 17

violent couples and identified multiple reasons why women might use violence in their relationships. Many felt that the change in policy now directed officers to make arrests, and that officers were more focused on the violent incident itself, ignoring the context within which it occurred. Even prosecutors and social workers in the prosecutor's office recognized these women as victims, and not as offenders. Respondents in Miller's study often felt that police were enforcing the letter of the law and ignoring the spirit of the law. Victim advocates and prosecutors claimed that a high number of the women they interacted with had histories of victimization. A number of the professionals interviewed in Miller's (2001) study had either seen or heard about batterers who had previously been-arrested, and now used their experiences to manipulate the system by behaving calmly when the police arrived, or by injuring themselves and claiming it was the victim. Therapists reported hearing men in perpetrator treatment groups state the importance.ofbeing the first to call the police, which would increase the likelihood of the victim's arrest, as a means to further their own control of her. These men often threatened to call 911 again or make a false report to the victim's probation officer that she had violated her probation. Miller (200 1) also found that police officers were often arresting both parties to avoid doing a lengthy investigation; thus placing the burden on the prosecutor's office to sort out culpability. Many women were unfamiliar with the 11

PAGE 18

legal system, and more likely to take a plea bargain without a full understanding of the implications to pl_eading guilty to a charge of domestic violence. Advocates relayed their experiences with women who often felt betrayed by lawyers who seemed more interested in judicial efficiency than in defending the women they were representing. Factors Affecting Police Arrest Decisions Other studies have excluded the arrests of women altogether or included these arrests with only minimal analysis of their meaning or implications. This research offers an opportUnity to understand how mandatory arrest policies are operationalized by police officers on scene. Researchers have looked at the demographic, attitudinal and situational variables that have affected arrest, and have employed analyses of police report data, surveys, and descriptive field studies, in attempts to understand police response to domestic violence (Robinson & Chandek, 2000). Analysis of Police Reports In their 1994 study, Bourg and Stock examined domestic violence arrest reports from a sheriffs department that did not participate in any coordinated comrimnity response to domestic violence and engaged in only minimal training on the subject. Over a twelve-month period, I ,870 reports of domestic violence 12

PAGE 19

resulted in 538 arrests (28.8%). The study found that felo.ny assaults were most likely to result in arrest, although only 34.7% of felony assaults ended in arrest. Of the 28 sexual assault cases reported to the Sheriff's Department, only five ended in arrest. All those arrested for sexual assault were boyfriends; no husbands or ex husbands were arrested for sexual assault. Bourg and Stock (1994) included the most in-depth look at gender and arrest compared to other studies that have examined police report data. They found that women were significantly more likely to be arrested on felony charges than were men. Black women were significantly more _likely to be arrested for felonies; 84.6% of Black women arrested were arrested on felony charges. White males, on the other hand, were significantly less likely to be arrested for felony assault. Bachman and Coker (1995) analyzed data from the National Crime Victimization Survey for the. years 1987-1992. Probability of arrest was affected by victim injury, marital status, and race. Assaults involving Black victims and Black offenders were almost twice as likely to result in arrest as white on white victimizations. Surprisingly, with prior histories of violence were less likely to be arrested than offenders with no history of violence. Bachman and Coker suggested this might be due to uncontrolled factors such as offender aggression towards police. Batterers who have experienced prior police contact may have "learned" how to behave 13

PAGE 20

In a study of police reports filed by 28 police departments in a Midwestern county, Erez (1986) found that officers were more likely to make an arrest in cases that involved ex-spouses compared to cases that involved married or dating couples. This finding was surprising in light of further conclusions that a higher severity of assault and injury was found in married and dating couples. Burris and Jaffe (1983) investigated the effect of a newly instituted pro arrest policy :on the behavior of officers in a Canadian police department. Expectedly, they found a significant increase in the number of arrests after the policy was instituted. The severity of police charges were significantly related to whether or not the male party had been drinking, the seriousness of injury to the victim, the need for hospital treatment of injury, and the criminal history ofthe batterer. In a review of 1,521 police reports, Jones and Belknap (1999) examined factors related to the arrest and jailing ofbatterers in Boulder County, Colorado. Almost 75% ofthe sample studied receivedsome.fonnal police response. This was much higher than findings from any of the other studies, and may be because Boulder County has been progressive in its response to domestic violence. Most agencies in the county have had a pro-arrest policy in place since 1986, and an oversight agency works to coordinate policy and provide feedback to departments. Jones and Belknap (1999) found that the presence of drugs or alcohol, children, or weapons increased the likelihood that police would make an arrest. 14

PAGE 21

The police were less likely to make an arrest if there was or ever had been a formal relationship between the couple. Dual arrests were made in 8.7% of the cases included in the.sample. Eigenberg, Scarborough, and Kappeler (1996) examined police reports over a five-year period in a mid-sized university town in order to compare police response to domestic violence assaults with police response to non-domestic assault (180 domestic violence assaults and 335 non-domestic assaults were compared). Officers were significantly less likely to arrest in domestic assaults than they were in non-domestic assaults. Arrest in cases of domestic assault was significantly related to weapon use, the presence of witnesses beyond the victim, the suspect's presence on scene, arid the victim's request for arrest. The presence of injuries was not significantly related to arrest. Officer Survey Studies A number of researchers have surveyed police officers, asking them to describe how they would respond to vignettes describing an assault. Officers were asked if th .... y would arrest or not, and what factors they would consider in making their decisions. Other studies asked officers to rank; in order of importance, factors effecting their decisions to arrest or not arrest when responding to a domestic violence assault. Additionally, Gandolf and McFerron (1989) interviewed battered women who had experienced police intervention. 15

PAGE 22

Feder (1997) surveyed 297 officers in a Florida police department with a newly implemented preferred arrest policy. The survey collected demographic information on officers, officer's attitudes towards women, officer's use of violence in their own homes, officer response to scenarios, and officer perceptions about the effectiveness of police response in domestic assault cases. The surveys revealed strong support for police intervention in domestic violence cases. Both support for pro;_arrest laws and the officer'.s use of violence in the home were significantly related to their score on the "attitudes towards women's roles" section ofthe survey. Feder's (1997) questions on officer attitudes towards women sought to measure officer's beliefs about women's roles athome and in public. Officer responses were ranked on a continuum from traditional to liberal. Questions posed to officers included: "Women police are as good as men?" "Boys should be taught more independence?" And, "better for man to be achiever?" The majority of officers disagreed with all three questions. When responding to the scenarios, officers were least likely to make an arrest when the victim had been cheating on the batterer or When the batterer had recently lost his job. The officer's attitudes towards women, knowledge ofthe department's policy, and belief in the effectiveness ofpolice interVention were significantly related to their decisions to arrest. Although they were more likely to 16

PAGE 23

arrest than not arrest in each of the scenarios, officers were most likely to arrest when the barterer was drunk. Holmes (1993) used in-depth interviews with officers, police call logs, and information from supplementary homicide reports to reconstruct police response to domestic violence in Massachusetts. The study found that the most frequent outcome of police response was to advise the victim of her rights (19.6%). Arrest took place in only 7. 7% of cases. The speed with which the victim contacted the police and the number of officers on scene were also related to arrest of the offender, Violation of a restraining order had the strongest relationship to arrest in this study. A history ofpolice calls to the same address, the presence of a weapon, the race of the injury, victim preference, and the presence of children, all had minimal, if any, impact on the decision to arrest. Dolan, Hendricks, and Meagher (1986) surVeyed police officers from three Midwestern police Officers wete asked to identify what factors they considered important when making a decision to arrest, what other alternatives to arrest they considered. effective, how they used referrals to social service agencies, and how they evaluated the effectiveness of those retertals. Police ranked eleven factors in order of influence on their decision to make an arrest; Factors that indicated a direct threat to officers, such as use of violence toward an officer, ranked highest. Indicators of a high level of seriousness of the assault and of a high possibility of future assault ranked next. The commission of a feloriy, use of a 17

PAGE 24

weapon, serious injury, and the likelihood of future assault were also ranked high by police officers. The use of alcohol, disrespect to officers, history of assault, and victim preference for arrest were ranked low on the list. Officers were also asked to rank factors that played a role in their decision not to make an arrest. Officers ranked twelve factors in order of importance. The top factors influencing the officer's decisions not to arrest were focused on the victim's behavior, either victim refusal to press charges at the scene, or the perception that victims often drop charges later. After victim behavior, officers ranked the lack of serious injury, commission of a misdemeanor, jail overcrowding, and intoxication of the victim and assailant as no-arrest influences. . In both surveys, the size ofthe department made a difference in the officer's rankings. Officers from the larger department said they were much more likely to make an arrest if there was a history of police response to the couple and if the batterer was intoxicated. They were much less likely to arrest if the jail was overcrowded and the victim was also intoxicated (Dolan, et al.,l986). Gandolf and McFerron ( 1989) interviewed over 6,000 battered women who had received shelter services in Texas. Nine categories of police response possibilities were identified, including: did nothing, mediated, referred victim, provided transportation, advised victim of legal rights, warned the batterer, forced the batterer to leave, and arrested the batterer. Women who reported that no police action was taken were actually more likely to have a weapon used against them 18

PAGE 25

and to have their lives threatened, and their level of injury and experience of abuse was similar to women whose barterers were arrested. Gandolf and McFerron (1989) attribute this difference to the anti-social behavior of the barterers themselves. More aggressive barterers were more likely to be arrested than other barterers regardless of the severity of the specific assault police were responding to. Barterers in the sample who exhibited aggressive behavior also had other arrests, alcohol abuse issues, and were more likely to severely injure their victims. Descriptive Field Studies Other studies have involved the observation of police officers by trained observers. Recording what officers actually do on scene has highlighted some notable differences compared to survey Although officers ranked use of alcohol, disrespect for officers, and victim preference low in surveys, these factors often played an important role in their decision to arrest as illustrated by their behavior on scene. Ferraro (1989) observed officers from the police department in Phoenix, Arizona, shortly after the department instituted a presumptive arrest policy. Observers found that police were less likely to make an arrest when the offender and victim were not married. Ferraro attributes this to the fact that the new pro arrest policy did not cover cohabitants. She did note, however, that even in cases where one or more arrestable crimes not dependent on the relationship of the 19

PAGE 26

victim and offender had been committed, such as destruction of property, the police still did not make an arrest. In cases where the victim and offender were married, police made arrests in a minority of the cases. Ferraro found that when looking for probable cause to make an arrest, police often used felony standards of evaluation; minor injury and property damage were not considered probable cause for an arrest. In spite of the presumptive arrest policy implemented by the Arizona police department, legal criteria were not found to be the most important considerations in police decision-making. Observers found ideological, practical, and political concerns to be more powerful predictors of arrest. Police beliefs about "deviant citizens" versus "normal citizens" (heterosexual, white, and English-speaking), stereotypes of battered women as likely to drop charges and as masochistic, and the idea that a man's home is his castle, supported the officer's belief that arrest was an invasion of privacy and that the new policy was an "'illogical policy" (Ferraro, 1989). Practical considerations and convenience also played a role in officer decisions to arrest. Batterers that fled the scene; uncooperative victims, and the need for investigation to establish a crime all played a part in officer decisions not to arrest. Having to involve mental health professionals and social services because of the presence of children were seen as barriers to arrest. Although officers in Arizona.had incentives to respond with seriousness and intensity to 20

PAGE 27

other non-domestic crimes, no professional benefits, such as promotions or commendations, were attached to compliance with the arrest policy in cases of domestic violence. This lack of positive incentive also undermined officer valuation of the policy itself, with some seeing it as a move by police administration to appease political supporters of the policy more than as an effort to support victims (Ferraro, 1989). Smith and Klein (1984) observed officers in 24 police departments operating in three metropolitan areas. Interviews with neighborhood residents were also completed. Cases selected by Smith and Klein included both domestic and non-domestic assaults. The strongest indicator of arrest was a victim's statement that they wanted the offender arrested. Intoxication of the offender, the level of aggressiveness toward the officers; and a history of police response were related to arrest. Although higher levels of violence increased the likelihood of neither the presence of injury nor the use of a weapon were significantly related to arrest. Race was not found to have any relationship to the officer's decision to arrest, although arrests were more frequent in lower socio-economic neighborhoods. Police were more likely to respect victims' wishes for arrest in higher SES areas compared to lower SES areas. Police were less likely to make arrests in domestic disputes versus non-domestic disputes in higher SES neighborhoods. The overall likelihood of arrest in lower SES areas was higher, and the rates of arrest for domestic and non-domestic assaults were about the same. 21

PAGE 28

In a replication ofBerk and Loseke's (1981) earlier study of police response to domestic violence, Worden and Pollitz (1984) found that the strongest indicator of arrest was whether or not a victim agreed to sign a complaint. The probability of arrest was also increased if the batterer had been drinking and if the victim made an allegation of violence. The likelihood ofarrest increased significantly, if the barterer was disrespectful towards the police officer. Differences in officer perceptions about their roles as crime fighters or problem solvers was also important in how they evaluated the situation at hand, and showed some evidence of effecting arrest patterns. Studies Using Combined Methodology In an attempt to create a more conteX:tualized picture of police response to domestic violence, Robinson arid Chandek (2000) combined methodological strategies to investigate factors influencing police arrest decisions. This study attempted to clarify inconsistencies in variable definitions and the results of earlier studies. Case summary reports and a supplemental survey that was completed on scene were combined with demographic data collected by the responding officers. Data were collected over a 5-month period in a mediuin-sized metropolitan police department. The presence of witnesses at the scene and co-habitation increased the likelihood of arrest. If the suspect left the scene, the odds of arrest decreased significantly. In fact, suspect presence had the most powerful effect on arrest. 22

PAGE 29

CHAPTER3 METHODOLOGY Data for the present study were collected from Boulder County, Colorado. This study uses data from the same source used by Jones and Belknap (1999). Police incident reports were reviewed and coded by the Domestic Abuse Prevention Project, a progressive oversight agency that organizes the coordinated community response in Boulder County. 1 The sample is composed of 8,952 total incidents responded to by eight law enforcement agencies in Boulder County between January 1994 and January 2000. Data for 32 cases where the sex of either the suspect or the victim was unknown were excluded, as was data for incidents involving same-sex couples. It is important to note, that in the six-year period covered by the data set, only 80 cases involved female same-sex couples, compared to only 43 cases involving male same-sex couples. It is unclear whether these numbers are high or low because no reliable population numbers exist community. Given the estimated rates of violence in same-sex couples, these numbers may signal a dangerous and disheartening state of affairs for the safety of all victims of same-1 For a more complete discussion of the history and function ofDAPP see Jones, D.A., and Belknap, J. (1999). Police responses to battering in a progressive pro-atrest jurisdiction. Justice Quarterly, 12:2 23

PAGE 30

sex intimate violence in Boulder County. It is likely that police responded to more cases of same-sex violence than are included in this data set, but if those cases were not coded as domestic violence they would neither fall into the purview of the Domestic Abuse Prevention Project nor be given any legal or social consideration as domestic violence cases. The arrests of men were compared to the arrests of women to identify differences and similarities in the factors that may affect police arrests in cases of domestic violence. Factors were divided into two categories: victim-offender characterj.stics and incident characteristics. Victim-offender characteristics include demographic information for the suspects, the suspect's prior interactions with police in Boulder County, and the victim-offender relationship. The data for prior police contact covers only prior arrests in Boulder County. There are conceivably offenders with prior arrests outside of Boulder County that would not be reflected in this data set. The victim offender relationship is coded in two variables. The variable "formality" refers to the length and legal recognition of the relationship, while the measure of "togetherness" refers to whether or not the victim and offender were together at the time ofthe incident. (Jones & Belknap, 1999). Victims and offenders that were married in the past or lived as common law spouses were coded as formal, while ex-boyfriends or ex-girlfriends who had never been married were coded as informal. Togetherness was coded as together if the couple was living together, 24

PAGE 31

married, or dating at the time of the arrest, while those coded as not-together were separated, divorced, or had formerly dated. Incident characteristics include the presence ofweapons during the incident, which party contacted the police, whether or not children were present, whether or not drugs or alcohol were present, and whether the assault was a felony. Injury data was not available in this data set. 25

PAGE 32

CHAPTER4 FINDINGS Male suspects in the study were typically white, in their early thirties, and had no prior police contact. Females were typically white, in their early thirties, and also had no prior formal police contact. The majority of female suspects (76.6%) and male suspects (70.4%) were involved in relationships that were ongoing at the time of arrest. Similar numbers of male and female suspects had formal relationships with their victims, and the majority were either married or living together. Similar numbers of female suspects (32.4%) and male suspects (33.3%) were using drugs and alcohol at the time of the arrest. Female suspects used a weapon in 14.7% ofthe incidents,_ while male suspects used a weapon in less that 10% ofthe incidents. Felony charges were filed in 5.8% of the cases involving a female suspect, and 5% of the cases involving a male suspect. Ofthe 8952 total cases, 7044 (78.7%) resulted in arrest, while 1908 (21.3%) did i1ot result in arrest. These numbers are higher than those numbers cited by other researchers, with the exception of Jones and Belknap (1999) who studied this same jurisdiction. The difference can most likely be attributed to the length of time a mandatory arrest policy has been in place in these departments, and the presence of a coordinated 26

PAGE 33

community response that includes support of the administration ofthe departments in the county. Table 4.1 Victim O(fender Characteristics Female Suspect Male Suspect N % (n) N % (n) Suspect Race 1988 6964 Black 1.80 36 4.20 292 Hispanic 6.10 122 13.30 924 White 86.40 1717 78.80 5490 Asian 109.00 38 1.20 83 American Indian 0.20 4 0.20 16 Other 0.30 7 0.30 20 Unknown 4.00 64 1.80 125 Victim Race 1988 6964 Black 2.80 55 1.00 72 Hispanic 6.60 I31 6.00 419 White 60.40 1201 52.00 3621 Asian 0.90 17 1.00 72 American Indian 0.00 0 0.20 I I Other 0.30 5 0.30 I3 Unknown 24.70. 492 34.60 2407 Victim-Offender Rei. 1959 6816 Married 31.50 627 31.20 2170 Common Law Married 1.60 32 1.00 70 Dating 13.80 274 13.40 930 Divorced 4.00 79 4.00 277 Ex-dating/Ex-living 12.80 254 17.80 1242 Together Living Together 28.60 568 23.40 1628 Separated 6.30 125 7.20 499 Couple together? 1959 .68I6 Yes 75.50 1501 68.90 4798 No 23.00 458 29.00 2018 27

PAGE 34

Table4.2 Incident Characteristics Female Suspect Male Suspect N % (n) N % (n) Law Enforcement Response 1988 6964 Arrested 80.20 1594 78.30 5450 Not Arrested 19.80 394 21.70 1554 Dual Arrest? 1641 6964 Yes 77.30 1269 6.60 384 No 22.70 372 93.40 5469 Suspect using drugs? 1988 6964 Yes 32.40 644 33.30 2316 No 23.80 474 19.80 1381 Unknown 43.80 870 46.90 3267 Victim using drugs? 1988 6964 Yes 8.30 165 4.80 332 No 20.20 401 18.30 1272 Unknown 71.50 1422 76.90 5360 Weapons Used? 1988 6964 Yes 14.70 292 8.90 617 No 18.10 360 18.60 1294 Unknown 67.20 1336 72.60 5053 Felony Charges? 1988 6964 Yes 5.80 116 5.00 347 No 57.70 1147 22.50 1566 Unknown 36.50 725 72.50 5051 Children Present? 1988 6964 Yes 27.10 538 29.80 2076 No 47.10 937 40.70 2832 Unknown 25.80 513 29.50 2056 Pearson's Chi-Square Cross-tabular analysis of factors affecting arrest supported the findings of many studies that some extra legal variables play an important role in police 28

PAGE 35

officer's decisions to arrest. Table 4.3 shows victim-offender characteristic variables related to arrest for both male and female suspects. Table4.3 Victim Offender Characteristics: Chi-Square Test N Not arrested Arrested xz % n % n Prior Formal Relationship With Victim Female Suspects 1959 12.54 *** No Prior Formal Relationship with Victim 16.90 (185) 83.10 (911) Prior Formal Relationship with Victim 23.30 (201) 76.70 (622) Male Suspects 6816 12.33 *** No Prior Formal Relationship with Victim 20.20 (769) 79.80 (3031) Prior Formal Relationship with Victim 23.80 (717) 76.20 (2299) Couple Together at Time of Arrest Female Suspects 1959 8.53 Couple Not Together at Time of Arrest 24.50 (112) 75.50 (346) Couple together at Time of Arrest 18.30 (274) 81.70 (1227) Male Suspects 6816 82.142 *** Couple Not Together at Time of Arrest 28.80 (581) 71.2 (1437) Couple Together at Time of Arrest 18.90 (905) 81.100 (3893) ** In this study, the sex ofthe suspect was not significantly related to arrest. Female suspects were as likely as male suspects to be arrested. The race ofthe suspect was not significantly--related to arrest for either males or females. Both male and female suspects who had some prior formal relationship with their victims were significantly less likely to be arrested than those suspects that did not have a prior formal relationship with their victims. This is consistent with prior findings (Jones & Belknap, 1999). Suspects who were in couples that 29

PAGE 36

were together at the time of arrest were more likely to be arrested than those suspects not "together" with their victims. Consistent with other studies, suspects using drugs or alcohol at the time of the incident were significantly more likely to be arrested than those that were not. This was true for both male and female suspects. Interestingly, suspects whose victims were using drugs at the time of arrest were also significantly more likely to be arrested. This was true for both male and female suspects. Weapons were significantly related to arrest for both male and female suspects. Weapons in this study included anything other than the suspect's hands or feet. Obvious weapons such as guns and knives, as well as, weapons of converiience such as ashtrays, glasses, or golf clubs that may have been thrown or used to hit were coded. Swprisingly, the presence of children on scene was only significantly related to arrest for male suspects. This is an interesting result given the predominant social view that women are primarily responsible for the safety of their children. It would be expected that women using violence in front of their children are deemed more responsible than men who use violence in front of their children. The local battered women's shelters have worked closely with police departments and the department of social services to raise awareness about working with children exposed to domestic violence. This may mean that police officers who are noting the presence of children on scene are better educated about 30

PAGE 37

domestic violence in general, and working with child witnesses in particular. They may be doing a more thorough investigation in order to protect the children in the long run. These officers may also be better at interviewing children, and may be gathering evidence from them that implicates the male party. Further analysis would be necessary to fully understand what is behind this result. Table4.4 Incident Characteristics: Chi-Square Test N Not arrested Arrested x' % n % n Victim Using Alcohol/Drugs at Time of Arrest Female Suspects 566 5.91 ... Victim Not Using Alcohol/Drugs 27.9 (112) 72.1 (289) Victim Using Alcohol/Drugs 18.2 (30) 81.8 (135) Male Suspects 1604 Victim Not Using AlcohoVDrugs 30.3 (385) 69.7 (887) 12.44 Victim Using AlcohoVDrugs 20.5 (68) 79.5 (264) 31

PAGE 38

Table 4.4 (Continued) Incident Characteristics: Chi-Sauare Test N Not arrested Arrested l % n % % Suspect Using Alcohol/Drugs at Time of Arrest Female Suspects 1118 18.01 .... Suspect Not Using Alcohol/Drugs 24.3 (115) 75.7 (359) Suspect Using Alcohol/Drugs 14.3 (92) 85.7 (552) Male Suspects 3697 93.87 Suspect Not Using Alcohol/Drugs 27.1 (374) 72.9 (1007) Suspect Using Alcohol/Drugs 14.2 (328) 85.8 (1988) Presence of Children at the Scene Female Suspects 1475 3.024 Children Not Present 21 (197) 79 (740) Children Present 17.3 (93) 82.7 (445) Male Suspects 4908 31.08 *** Children Not Present 23.4 (662) 76.6 (2170) Children Present 16.9 (350) 83.1 (1726) Weapon Present Female Suspects 652 286.62 *** No Weapon 71.7 (258) 28.3 (102) Weapon 5.8 (17) 94.2 (275) Male Suspects 1911 586.58 ... No Weapon 67.3 (871) 32.7 (423) Weapon 8.1 (50) 91.9 (567) Suspect Prior Open Cases Female Suspects 1970 4.157. No Open Cases 18.8 (327) 1:1.2 (1414) Open Cases 24.5 (56) 75.5 (173) Male Suspects 6899 23.24 ... No Open Cases 20.5 (1223) 79.5 (4751) Open Cases 27.5 (254) 72.5 (671) p::; .05 ** p::; .01 *** p::; .001 32

PAGE 39

A history of domestic violence arrests was not significantly related to arrest for either female or male suspects, although the existence of open cases was found to decrease the likelihood of arrest for both female and male suspects. This finding was surprising, but may be related to officer perception that the victim is not willing to separate from the suspect. Officer surveys found that victim behavior was a strong influence in decisions not to arrest (Dolon et al., 1986). Officer perception that some victims are not willing to help themselves by leaving a relationship may be linked to assumptions about whether the victim will follow through with the case. The commission of a felony was not significantly related to arrest in this study. This is inconsistent with other studies and may be related to limitations with the data. The information on charges in this data set was limited and felony arrests made up a very small number of overall arrests. The data in this study does show significant differences in the victimization experiences of men and women suspects. Just fewer than 15% of the male victims of female suspects had prior arrests for domestic violence, compared to only 10% of the victims of male defendants. Almost25% ofthe female suspects had a prior victimization history compared to only 9% of the male suspects. 33

PAGE 40

Multivariate Analysis Table 4.5 shows the result of a multivariate analysis using a logistic regression model. Arrest was used as the dependent variable in the model. The sample size for the multivariate analysis was decreased considerably because of the amount of missing data. A considerable amount of felony data was missing, so this variable was excluded from the analysis. Analysis was done on 239 cases involving a female suspect and 584 cases involving a male suspect. Table4.5 Logistic Regression Predicting Arrest Female Suspects Exp(B) S.E. Incident Characteristics Use of a Weapon Drug/AJcohol Use by Suspect Children Present Who Contacted Police Victim-Offender Characteristics Suspect Prior Victim Suspect Prior DV Suspect Prior Open Case Suspect Race Victim Race Formal Relationship Couple Together p::;.05 ** p::;.ot *** p::: .001 49.66 0.52** 0.89 0.38 2.60 0.44* 0.99 0.39 1.03 0.53 3.05 0.50* 0.56 ... 0.64 0.89 0.56 0.98 0.14 1.18 0.38 1.48 0.44 34 Male Suspects Exp (B) S. E. 24.98 0.29** 1.48 0.23 1.16 0.24 0.37 0.23** 1.61 0.57 1.58 0.25 0.77 0.31 0.80 0.26 1.04 0.07 1.00 0.23 0.54 0.28*

PAGE 41

As expected, the presence of a weapon increased the likelihood of arrest for both male and female suspects. Ifthe couple were together at the time of the arrest, the police were less likely to charge the male suspect. This finding is consistent with other studies that found ex-spouses more likely to be arrested than married or dating couples. The difference in results for male and female suspects may be related to victim-blaming biases held by officers about why battered women do or do not leave an abusive relationship. This finding was also consistent with officer survey studies that identified the female victim's behavior as a strong influence in police decisions not to arrest (Dolon, et. al, 1986). The presence of children on scene was found to significantly increase the likelihood of arrest for female suspects, but did not remain significant for male suspects. This result more accurately reflects societal attitudes that hold women responsible for violence perpetrated against children by their chosen intimate partners. More research would need to be done to fully understand the relationship between the presence of children and arrest. The use of drugs and alcohol did not remain significantly related to arrest for either male or female su:spt!cts, but was approaching significance for males. A criminal history of domestic violence was significantly related to an increase in the likelihood of arrest for female suspects, but not for males. This result may be due to social views of men and women. Violence is often a celebrated cultural value 35

PAGE 42

for men in the United States, while women who use violence deviate profoundly from social norms. 36

PAGE 43

CHAPTERS CONCLUSION Assertions that women use violence in interpersonal relationships as much or more than men do have often been met with a challenge to contextualize the violence being studied. Only a qualified understanding of the violence in a relationship can be obtained by counting the number of physical assaults. A large body of research has now been produced that illustrates profound differences in the ways women and men use and experience violence. The purpose of this research study has been to argue for a similar contextual understanding of domestic violence arrests. Data collected from Boulder County, Colorado, illustrate the differences in factors that increase the likelihood of arrest for male and female suspects. It is clear from this gender analysis that profound differences exist in police decision-making regarding arrest. This study demonstrates that women experience state intervention differently than men; they are at greater risk of being a; rested if their children are on the scene, or if they have a history of police intervention for domestic violence. Women who are victims of domestic violence are less. likely to be protected by the police if there is an open case pending against their batterers. 37

PAGE 44

Understanding the why and how police officers make decisions to arrest is important for advocates engaged in systems change work and public policy refinement. As the numbers of women arrested for domestic violence increases, advocates, researchers, and policy makers must try to understand why and how these arrests are happening, and to integrate this knowledge with what we are learning about gender differences in the use and meaning of violence. The differences in offending and victimization histories of the men and women in this data set give some urgency to refining research on criminal legal into domestic violence to include an effective gender analysis. It will be important for future research to include voices of women that have been arrested, especially when advocates identify these women as victims of domestic violence, and not as perpetrators. Researchers and advocates who have begun to build bridges between the women's anti-violence movement and the prison abolition movement have laid the groundwork for a more complex understanding of the intersection of interpersonal and state control in women's lives. The number of women arrested for domestic violence is increasing in many jurisdictions at the same timt: as the general women's prison population is increasing. Given that arrest for domestic violence is a relatively more recent way for women to enter the criminal justice system, and that a woman's history of domestic violence increases the likelihood for 38

PAGE 45

subsequent arrest, this issue will become an increasingly important consideration for criminologists. 39

PAGE 46

RESOURCES Bachman, R., & Coker, A. (1995). Police involvement in domestic violence: The interactive effects of victim injury, offender's history of violence, and race. Violence and Victims, 10, 91-106. Belknap, J., & McCall, K. D. (1994). Woman battering and police referrals. Journal of Criminal Justice, 22, 223-236. Belknap, J. (1996). Law enforcement officers attitudes about the appropriate responses to woman battering. In The Invisible Woman: Gender, Crime, and Justice. Belmont, CA: Wadsworth. Berk, S., & Loseke, D. R. (1981).l:landling family violence: Situational determinants of police arrest in domestic disturbance: Law and Society, 15(317-346). Berk, R., & Newton, P. J. (1985). Does arrest really deter wife battery? An effort to replicate the findings of the Minneapolis spouse abuse experiment. American Sociological Review, 50, 253-262. Bhattacharjee, A. (2001). Whose saftty? Women of color and the violence of law enforcement: American Friends Service Committee, and the Committee on Women, Population, and the Environment. Bourg, S., & Stock, H. V. (1994). A review of domestic violence arrest statistics in a police department using a pro-arrest policy. Journal of Family Violence, 9(177 -189). Bowker, L. H. (1982). Police services to battered women: Bad or not so bad? Criminal Justice and Behavior, 9, 476-494. Bowman, C. G. (1992). The arrest experiments: A feminist critique. Journal of Criminal Law and Criminology, 83,201-209. Brown, S. E. (1984). Police responses to wife beating: Neglect of a crime of violence. Journal of Criminal Justice, 12, 277-288. 40

PAGE 47

Burris, C. A., & Jaffe, P. (1983). Wife abuse as a crime: The impact of police laying charges. Canadian Journal of Criminology, 25. Buzawa, E., & Austin, T. (1993). Determining police response to domestic violence victims:The role of victim preferences. American Behavioral Scientist, 36(5), 610-623. Buzawa, E. S., & Buzawa, C. G. (2002). Domestic violence: The criminal justice response (3rd ed.). Thousand Oaks, CA: Sage. Das Gupta, S. D. (2002). A framework for understanding women's use of nonlethal violence in intimate heterosexual relationships. Violence Against Women, 8(11), 1364-1389. DeKeseredy, W. S., & Schwartz, M: (1988). Woman abuse on campus: Results from the Canadian National Survey. Thousand Oaks: Sage. Dobash, R. P., Dobash, R. E., Wilson, M., & Daly, M. (1992). The myth of sexual symmetry in marital violence. Social Problems, 39(71-91). Dolon, R., Hendricks, J., & Meagher, M. S. (1986). Police practices and attitudes towards domestic violence. Journal of Police Science and Administration, 14, 187-192. Eigenberg, E. M., Scarborough, K.E., Kappeler, V.E. (2001). Contributory factors affecting arrest in domestic and non-domestic assaults. In H. M. Eigenberg (Ed.), Woman Battering in the United States (pp. 298-318). Prospect Heights, Illinois: Waveland Press. Erez, E. (1986). Intimacy, violence, and the police. Human Relations, 39,265-281. Feder, L. (1997). Domestic violence and police response in a pro-arrest jurisdiction. Women and Criminal Justice, 8, 79-97. Ferraro, K. J. (1989). Policing woman battering. Social Problems, 36, 61-74. Gandolf, E. W., & McFerron, J. R. (1989). Handling battering men: police action in wife abuse cases. Criminal Justice and Behavior, 16, 429-439. 41

PAGE 48

Gelles, R., & Loseke, D. (1993). Current controversies in family violence. Newbury Park, CA: Sage. Gondolf, E. W. (1988). Battered women as survivors: An alternative to treating learned helplessness. New York: Lexicon Books. Hamberger, L. K., Guse, C.E. (2002). Men's and women's use of intimate partner violence in clinical samples. Violence Against Women, 8(11), 1301-1331. Hart, B. (1993). Battered women and the criminal justice system. American Behavioral Scientist, 83, 73-119. Hirschel, D., & Buzawa, E. (2002). Understanding the context of dual arrest with directions for future research. Violence Against Women, 8(12), 1449-1473. Holmes, W. (1993). Police arrests for domestic violence. American Journal of Police, 12(4), 101-125. Hooper, M. (19%). When domestic violence diversion is no longer and option: What to do with the female offender. Berkeley Women's Law Journal, 11, 168-181. Johnson, M. (1995). Patriarchal terrorism and common couple violence: two forms ofviolence against women. Journal of Marriage and The Family, 57,283-294. Jones, D. A., & Belknap, J. (1999). Police to battering in a progressive jurisdiction. Justice Quarterly, 16(2), 249-273. Manless, M. (1996). Mandatory arrest: A step toward eradicating domestic violence, but is it enough? University of fllinois Law Review, 2, 533-587. Martin, M. (1997). Double your trouble: Dual arrest in family violence. Journal of Family Violence; 12, 139-157. Miller, S. L. (1989). Unintended side effects of pro-arrest policies and their race and class implications for battered women: A cautionary note. Criminal Justice Policy Review, 3(3), 299-317. Miller, S. L. (2001). The paradox ofwomen arrested for domestic violence. Violence Against Women, 7, 1339-1376. 42

PAGE 49

Mills, L. G. (1999). Commentary: Killing her softly: Intimate abuse and the violence of state intervention. Harvard Law Review, 113 Harvard Law Review 550. Oppenlander, N. (1982). Coping or copping out. Criminology, 20,449-465. Pleck, E. (1987). Domestic tyranny: The making of social policy against family violence from colonial times to the present. New York: Oxford University Press. Ritchie, B. E. (1996). Compelled to crime: The gender entrapment of battered black women. New York: Routledge. Ruttenberg, M. H. (1994). Feminist critique of mandatory arrest: An analysis of race and gender in domestic violence policy. Gender and Law, 2, 171-199. Saunders, D. G. (1995). The tendency to arrest victims of domestic violence. Journal of Interpersonal Violence, 10, 147-158. Saunders, D. G. (1996). When battered women use violence: Husband abuse or self-defense? Violence and Victims, 1, 47-60. Sherman, L. W., & Berk, R. (1984). The specific deterrent effects of arrest for domestic assault. American Sociological Review, 49, 261-272. Sherman, L. W. (1992). The influence of criminology on criminal law: Evaluating arrests for misdemeanor domestic violence. Journal of Criminal and Criminology, 83, 1-45. Smith, D. A., & Klein, J. R. (1984). Police control of interpersonal disputes. Social Problems . 31,469-481. Stanko, E. A. (1989). Missing the mark? Policing battering. In J. Hanmer, J. Radford & E. A. Stanko (Eds.), Women, policing, and male violence (pp. 46-69). London: Routledge. Stark, E. (1993). Mandatory arrest ofbatterers: A reply to its critics. American Behavioral Scientist, 36,651-680. 43

PAGE 50

Stark, E. (1995). Re-presenting woman battering: From battered woman syndrome to coercive control. Albany Law Review, 58, 973-1026. Steinmetz, S. K. (1977 -1978). The battered husband syndrome. Victimology: An International Journal, 2, 499-509. Stets, J. E., & Straus, M.A. (1990). The marriage license as a hitting license: A comparison of assaults in dating, cohabitating, and married couples. In M. A. Straus & R. J. Gelles (Eds.), Physical violence in American families (pp. 227-244). New Brunswick: Transaction Publishing. Straus, M.A., Gelles, R., & Steinmetz, S. K. (1980). Behind closed doors: Violence in the American family. New York: Anchor. Straus, M. A., & Gelles, R. (1986). Societal change and change in family violence from 1975 to 1985 as revealed by two national surveys. Journal of Marriage and The Family, 48,465-479. Waaland, P., & Keeley, S. (1985). Police decision-making in wife abuse: The impact of legal and extra legal factors. Law and Human Behavior, 9, 355366. Worden, R. E., & Pollitz, A. A. (1984). Policearrests in domestic disturbances. Law and Society Review, 18, 105-119. Zorza, J. (1992). The criminal law ofmisdemeanor domestic violence, 1970-1990. The Journal of Criminal Law and Criminology, 83(1), 46-72. Zorza, J., & Woods, L. (1994). Mandatory arrest: Problems and possibilities. Washington DC: Center on Women and Family Law. 44