National domestic violence data collection

Material Information

National domestic violence data collection quantifying the oppression of women
Portion of title:
Quantifying the oppression of women
Groginsky, Elizabeth Ann
Publication Date:
Physical Description:
vi, 52 leaves : illustrations ; 29 cm


Subjects / Keywords:
Family violence -- Statistics -- United States ( lcsh )
Family violence ( fast )
United States ( fast )
Statistics. ( fast )
bibliography ( marcgt )
statistics ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )
Statistics ( fast )


Includes bibliographical references (leaves 49-52).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Social Science.
General Note:
Department of Humanities and Social Sciences
Statement of Responsibility:
by Elizabeth Ann Groginsky.

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:
39729956 ( OCLC )
LD1190.L65 1997m .G76 ( lcc )

Full Text
Elizabeth Ann Groginsky
B.A. University of Maryland, 1988
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Social Science

This thesis for the Master of Social Science
degree by
Elizabeth Ann Groginsky
has been approved
Jan Mikish

Groginsky, Elizabeth Ann (Master of Social Science)
National Domestic Violence Data Collection: Quantifying the Oppression of Women
Thesis directed by Jana Everett
This research provides a feminist critique of current domestic violence data collection
systems that are based primarily on police and hospital data. The sole reliance on
officially reported data limits our understanding of the problem and understates the
incidence of domestic violence. This thesis argues that it is imperative for the
battered womens movement to begin to develop its own data collection system
before the criminal justice system co-opts the measurement, definition and eventually
the policy decisions related to domestic violence. Without this data collection effort,
battered womens programs risk having their so-called inflated numbers ignored or
even challenged in policy decisions. The battered womens movement must work to
build consistency and integrity into the data they collect from women every day. This
research is a wake-up call to the battered womens movement to become involved in
defining the measurement of domestic violence occurrence before it is too late.
This abstract accurately represents the content of the candidates thesis. I recommend
its publication.


A special thanks to my wonderful husband, Scott Groginsky, who provided me with
hours of emotional and moral support as well as invaluable editing. I would also like
to thank my mother-in-law, Bobbie Alexander, a very special woman who provided
unconditional support and made it possible for me to complete my Masters degree.

^vjtn miN i a
1. INTRODUCTION.........................................I
Purpose of the Research............................5
Arrangement of the Thesis..........................6
Historical Framework ..............................9
Uniform Crime Report............................18
National Crime Victimization Survey ............20
Violence Surveys................................24
Intimate Partner Surveillance System............26
Rhode Island Pilot Project....................29
Michigan Pilot Project .......................32
Barriers to Collecting Domestic Violence Program Data .... 39
Reliability and Validity of Data from Programs.41
5. CONCLUSION.........................................43


In the past 20 years, domestic violence has become a recognized social
problem. Domestic violence is sometimes referred to as family violence, spouse abuse,
domestic abuse or battering. The problem of domestic violence in heterosexual
relationships is overwhelmingly represented by male violence against females.
According to the National Crime Victimization survey, men are victims of intimate
violence only 8 percent of the time (Bachman 1994). Some feminists define domestic
violence as an act or threatened act of violence or intimidation that is used by men to
maintain power and control over woman they are in an intimate relationship with
(Dobash and Dobash 1992, Zorza 1992, Yllo and Bograd 1988). This definition is
grounded in the feminist theory that intimate relationships are based on a patriarchal
assumption that men have permission to dominate women. This assumption is based
on another patriarchal notion: the individual with the most power is able to dominate
and control others. While this definition emphasizes male violence against women,
same-sex relationship violence can also be understood in these terms since all people
who grow up in a patriarchal society learn the values and traditions of that society.

A feminist theory of domestic violence is supported by a historical analysis of
laws regarding the rights of women within relationships to receive protection from
violence (Pagelow 1981, Zorza 1992, Dobash and Dobash 1992). Until the early
1970s a woman could receive a restraining order only if she filed for divorce or legal
separation (Zorza 1992, Fagan 1995). There were relatively minor sanctions imposed
against most men who beat their wives until the early 1980s (Zorza 1992, Dobash and
Dobash 1992).
Beginning in the late 1980s, the majority of states began passing legislation that
defined domestic violence as a crime (Zorza 1992). The National Council of Juvenile
& Family Court Judges (NCJFCJ) published a Model State Code for Family Violence
and in it defined domestic violence as any act that:
attempts to cause physical harm to another family or
household member; placing a family or household member
in fear of physical harm; or causing a family or household
member to engage involuntarily in sexual activity by force,
threat of force, or duress. (NCJFCJ, 1994:1)
How a problem is defined often dictates how the problem is solved (Lerman
1992). A feminist definition of domestic violence suggests that the solution lies in
changing the nature of social and political relationships, particularly within the
institution of marriage and/or conjugal relationships (Dobash and Dobash 1992,
Pagelow 1981). A criminal-law definition suggests that police and prosecutors are the
answer to solving the problem of domestic violence (Lerman 1992, Fagan 1995).

Both approaches are important in order to protect women and children, but the
criminal definition merely responds to the symptoms of domestic violence without
addressing the systemic issue of gender inequality.
In order to adequately address the problem of domestic violence, it is
important to understand its occurrence and prevalence. The lack of good data has
thwarted appropriate public policy (Besharov, 1990). Its very difficult to develop
good public policy and good programs if you dont have good data on what is going
on at the local level, said Anne Menard, director of the National Resource Center on
Domestic Violence. Hard national data on domestic violence are scarce (Meckler
Due to the nature and complexity of domestic violence, discovering the
incidence and prevalence is a difficult and complex process. Because strong emotional
ties usually exist between the victim and the perpetrator, the victim often feels
protective of the perpetrator (Fagan 1995; Walker 1980). The victim is frequently in
denial about the abuse and may be financially dependent upon the perpetrator (Walker
1980). For example, a victim may be afraid to report an incident because the
perpetrator could lose his job or be unable to pay the rent. Victims are also often
afraid of retaliation if they report the violence. Moreover, society stigmatizes and
shames women who are battered (Pagelow 1981, Yllo 1988). The too often repeated

phrase why does she stay? is an example of this stigma, contributing to a woman not
reaching out and telling someone about her experience.
Domestic violence is not necessarily a single incident occurring at one point in
time. It can be multiple acts of violence that are repeated weekly, monthly or yearly
(Fagan 1995). Quantifying this kind of experience is challenging. In addition,
domestic violence is clearly an underreported crime. According to the National Crime
Victimization Survey, women who are victimized by an intimate partner report half of
their victimizations to the police (U.S. Department of Justice 1994). Other researchers
estimate that as few as 10 to 20 percent of cases are reported to the police (Fagan
1995; Walker 1980). Police come in contact with only a small portion of the women
who are being battered (Debair 1997, Keith 1997). Our national domestic violence
data collection effort must go beyond official reported data or acts defined as criminal.
Stanko (1988:41) correctly points out that
Crime, once defined as such by official agencies, takes on an
apparently objective quality divorced from eveiyday reality. Thus
serious physical and sexual assaults, for example, are defined by the
criminal statutes in a way far removed from the social context in which
they occur.
To truly understand the incidence of domestic violence, battered womens
program data needs to be organized in a meaningful way. Every day over 1,500
hundred programs provide a safe place for battered women to get support, but
there has been no attempt to compile it on a statewide or national basis.

Battered womens programs consist of shelters, outreach agencies, and/or
hotlines. Their mission is to support and empower women, which is rooted in
feminist philosophy, and they are therefore in touch with the daily lives of
battered women. The kind of data these programs collect would help provide a
clearer picture of the scope, frequency, and severity of domestic violence.
Purpose of the Research
This research is exploratory because national domestic violence data collection
has only been discussed in the last three years. The research explores how domestic
violence is currently being quantified. It also examines who collects domestic violence
data, how the data is defined and measured, and the extent to which domestic violence
services program data are included in these efforts. The collection of data is a political
issue. Policy makers and legislators look to the data when determining the seriousness
of the problem and the appropriate policy response. Who defines the problem is then a
critical issue. This research will provide an analysis of who defines the problem and
the limitations of the definition. The results of the research will provide advocates,
researchers, and policy makers with critical information about the status of national
domestic violence data collection. There are many issues and complexities involved in
defining, measuring, and compiling domestic violence data and this research hopes to
open more fully the dialogue and research in this area.

Arrangements of the thesis
Chapter two provides the theoretical perspective of the researcher and a
historical analysis of the battered womens movement. Chapter three describes the
current data systems that are used to quantify the problem of domestic violence.
Chapter four examines eight domestic violence programs in Colorado that collect data
and discusses the strengths and limitations of domestic violence program data.
Chapter five discusses where we go from here and includes recommendations for the
future of domestic violence data collection.

From a feminist theoretical perspective, domestic violence is understood as a
problem associated with living in a patriarchal society and state (Yllo and Bograd
1988, Dobash and Dobash 1992). A patriarchal society accepts male violence and the
use of violence to control and dominate women (Yllo and Bograd 1988). A
patriarchal state institutionalizes the acceptance of male violence. Feminist theorists of
domestic violence challenge theorists who choose to understand domestic violence as
individual pathology or understand violence within the home only as part of the
violence in the larger society (Dobash & Dobash 1992; Lerman 1992). Some feminists
see violence in the context of the social and political nature of social relationships that
are based on power and domination (Murray 1988). Domestic violence is
unquestionably a sophisticated tool used by men in a patriarchal society to maintain
power and control over the lives of women and children.
Instead of examining individual acts of domestic violence, feminists ask why
men use violence against their partners and what function this serves for a society in a
specific historical context (Walker 1980, Pagelow 1981, Yllo and Bograd 1988).
Feminists are interested in exposing how marriage, as a patriarchal institution, provides

a legitimate rationale for domestic violence, instead of merely learning why one man
may abuse his wife (Yllo and Bograd 1988).
The increased awareness in the early 1970s about the problem of domestic
violence resulted directly from consciousness-raising groups in the feminist movement.
When women began to speak out about the violence in their lives, the battered
womens movement emerged. The personal is political had significant meaning for
battered women. The states response toward domestic violence was that the home
was a private domain and the state had no legitimate right to intervene. Battered
womens advocates made battering a political issue and insisted that the state not only
had a right, but a duty to intervene on behalf of battered women (Zorza 1992, Dobash
and Dobash 1991, Fagan 1995). At that time, the response considered appropriate for
a police officer called to the scene of a domestic disturbance was to calm things
down and not make arrests (Dobash and Dobash 1992, Zorza 1992, Fagan 1995).
The policy of many police departments to not arrest was interpreted by battered
womens activists as denying women the right to equal protection under the law
(Dobash and Dobash 1992). This response reinforced the feminist assumption that the
state was patriarchal.
Some feminists argue that the state is patriarchal because the laws of the state
are based on the way men view and treat women. Objectivity, a male value is
considered the norm for the state (MaKinnon 1989). For battered women, the law of

privacy translated into a sphere of personal freedom for men, while it left women in a
private sphere of intimate violence and abuse (Mackixmon 1989). State laws have
evolved over the past 20 years in regard to providing women with legal rights and
protections, but male power is systemic. For example, mandatory arrest laws were
enacted so battered women could receive police protection. Now, women are being
arrested at alarming rates under these same laws that were originally written to protect
them. According to feminist scholar Catharine MacKinnon, State power, embodied
in law, exists throughout society as male power at the same time as the power of men
over women throughout society is organized as the power of the state (MacKinnon
Historical Framework
Feminists began organizing in the late 1970s to change the laws concerning
protection for battered women (Fagan 1995, Zorza 1992, Dobash and Dobash).
These women activists realized that the existing legal system provided protection to
men who beat their wives, while leaving many women terrified, permanently disabled
and some even dead (Dobash and Dobash 1992).
Resistance by law enforcement agencies to change their policies mobilized
battered womens advocates to use civil law to force changes in the way police
responded to domestic violence incidents (Zorza 1992, Dobash and Dobash 1992,

Fagan 1995). In October 1976, the case of Scott vs. Hart was filed against the
Oakland police. The complaint was multifaceted, ranging from enjoining the police
from refusing to respond adequately to battered womens calls to ordering the police
to establish a shelter for women and ordering the police to train officers to handle
domestic violence cases (Zorza 1992:55).
Two months later, the case of Bruno vs. Codd was filed against the New York
City Police (Dobash and Dobash 1992, Zorza 1992). The litigation charged police
and courts with gross failure to comply with New York State laws. The case was
brought on behalf of 12 women who received no action from police called to respond
to a physical attack upon the women by their male partners (Dobash and Dobash
1992). The New York City Police Department settled out of court in 1978. Without
admitting guilt, police department officials agreed to change their practice (Dobash
and Dobash 1992). These lawsuits made it clear that police departments would be
held liable if they failed to protect the rights of battered women (Zorza 1992).
In 1984, Tracy Thurman brought a successful class action lawsuit against the
city of Torrington, Connecticut because police had failed to respond to her calls for
help even though she had an order of protection (Dobash and Dobash 1992). Police
officers actually sat in a patrol car outside her house as her estranged husband attacked
her (Dobash & Dobash 1992). The case was clearly a victory for battered women and
battered womens activists because the decision addressed the root causes of domestic

violence. First, a federal grand jury awarded Thurman and her son $2.3 million
because the police were negligent in failing to protect them from her abusive husband.
Second, the court ruled that the police response amounted to sex discrimination
(Zorza 1992). Because of the resulting liability issues, Connecticut lawmakers enacted
mandatory arrest legislation as the judicial branch validated the existence of sexism in
the criminal justice system (Dobash and Dobash 1992). Mandatory arrest policies
required police officers to arrest anyone in a domestic violence incident whom the
officer had probable cause to believe had committed an assault or placed a victim.. .in
fear of imminent physical injury (Zorza 1992). Over the next decade, the issue of
mandatory arrest became the focal point of domestic violence policy reform. The
other key systemic issues, such as training for police, sexism in the system, and telling
women their rights, which were brought out in these civil cases, were virtually ignored
(Lerman 1992, Zorza 1992, Dobash and Dobash 1992).
At the time of the Thurman lawsuit, a criminology researcher, Larry Sherman,
was conducting a study in Minneapolis, Minnesota and determined that arrest in
domestic violence cases was a more effective deterrent than mediation (Zorza 1992,
Dobash & Dobash 1992, Fagan 1995). Although mandatory arrest policies were
already being adopted, this research became the stated impetus for sweeping changes
in law enforcement policies across the nation (Fagan 1995, Besharov 1990). Policy

makers turned to the criminal justice system for the solution to domestic violence, and
later turned again to the system to help define and measure the extent of the problem.
At the time that Shermans findings were published, advocates across the
country had long been arguing for a more appropriate police response to battered
women. Their recommendations included training for law enforcement officers on
how to handle domestic violence cases, the police explanations to women about their
legal options, and police arresting the abusive partner if the woman requested it (Zorza
1992). Law enforcement agencies, however, pointed to Shermans limited study as
their sole guide for policy changes. The result was the narrow solution of mandatory
arrest. This action ignored the requests of activists and battered women who had
articulated a much more complex solution (Zorza 1992, Lerman 1992, Dobash &
Dobash 1992, U.S. Department of Justice 1996). Shermans study, rather than the
body of case law initiated by activists, became the basis for the change in police policy
(Besharov 1990; Lerman 1992; Fagan 1995).
For more than 20 years, the battered womens movement has been
collaborating with the criminal justice system to improve the options of women living
with violence. Many of these changes have been instrumental in reducing domestic
violence fatalities and in providing needed resources to women and children.
Unfortunately, the institutionalized sexism of the past is still prevalent in the criminal
justice system, including law enforcement agencies, prosecutors offices, and judges

chambers. Women are not only being arrested under mandatory arrest laws, but those
women who refuse to testify against their perpetrators are often charged with false
reporting. Recent studies show that women who kill their current or former partners
receive 15 to 30 years in prison, while men who kill their current or former partners
receive 5 to 10 years. Battered womens advocates are just beginning to compile
evidence on the increasing numbers of women arrested under mandatory arrest laws
(Dobash and Dobash 1992, Zorza 1992, Fagan 1995). Police still deny women
protection and information about their rights when responding to calls, and the male
perpetrators of violence are punished minimally for their crimes. In Colorado, a man
who is arrested and convicted of assaulting, harassing, or threatening his current or
former partner receives, on average, one year of probation and 36 weeks of domestic
violence counseling. Some feminists have argued that the battered womens
movement should be wary of its relationship with this patriarchal institution because
the criminal justice system is not concerned about the equality of women or in
stopping male domination and oppression, but rather with expanding regulation of
society. Dobash & Dobash (1992:209) succinctly describe the dynamic between the
criminal justice system and the battered womens movement:
Once encapsulated within criminal justice, reform movements must
inevitably serve the needs of the system and potentially assist its expansion
into more areas of civil society. The criminal justice system may embrace
the movement and yet subvert is demands, particularly within a law and
order agenda, as a means of gaining greater resources and of widening the

net of its response. An absorbent state may be able to incorporate the
pragmatic concerns of the movement while ignoring or even rejecting its
politics (209).
Although there are decisionmakers in the criminal justice system that support and/or
subscribe to a feminist definition of battering, their ability to make substantial changes
is limited by the institutionalized patriarchy in the criminal justice system.
In 1994, the federal Violence Against Women Act (VAWA) was enacted,
again as the result of years of lobbying and educating by battered womens activists
about the problem of violence against women (U.S. Department of Justice 1996). The
VAWA legislation was a comprehensive approach combining a broad array of legal
and practical reforms (U.S. Department of Justice 1996). The legislation was aimed
at improving the response of police, prosecutors, and judges. It increased funding for
battered womens shelters and state coalitions and strengthened temporary restraining
orders (U.S. Department of Justice 1996).
VAWA also mandated that a study be conducted to find out how states collect
and analyze domestic violence and sexual assault data and provided funding for states
to establish intimate partner violence surveillance systems (Title IV-Violence Against
Women Act 1994). Three states are currently piloting systems and all are relying on
officially reported data either to police, courts (restraining order registries) and/or
hospitals (Verhoek-Oftendahl 1997, Largo 1997). This thesis will argue, that VAWA
funding could contribute to a better understanding of the frequency of domestic
violence by helping battered womens programs automate and organize the key data

that they already collect. According to eight programs in Colorado, compiling
domestic violence program data in a meaningful way would be very important (Powell,
Sweet, Cameron, Mitchell, Moore, Olguin, Simmons, Miller 1997). Since public
funding is already provided to police, courts and hospitals to collect data, using
VAWA funds to make domestic violence program data accessible would effectively
complement existing officially reported data.
Battered womens programs experiences and collective knowledge are left out
of data collection design systems (Largo 1997, Verhoek-Oftendahl 1997).
Researchers too often look with skepticism at the data collected by battered womens
programs. Instead they turn to hard data such as police and emergency room
records to quantify how many women are abused. Researchers often regard analyses
provided by shelter workers and formerly battered women as nonscientific,
subjective, and political and not comparable to analysis based on hard data (Yllo
and Bograd 1988:43). These researchers do not address explicitly the fact that
medical professionals and criminal justice system officials have a political agenda and
are not free of bias.
It is valuable to compile and organize domestic violence program data in a
meaningful way to realize a deeper understanding of the characteristics, incidence,
frequency and severity of domestic violence. The majority of battered womens
programs track detailed information about the nature of battering relationships. The

data from these programs will increase our understanding of the social/political reality
of violence against women. Including information from agencies in the community
who are providing safety and support to battered women will also more completely
reveal the problem of domestic violence. Program data offers a more comprehensive
perspective for several other reasons. First, these programs often serve women who
have never reported abuse to law enforcement agencies or medical professionals
(Walker 1980, Keith 1997, Debair 1997). Second, while crime victimization survey
data can help uncover unreported crimes, the surveys domestic violence findings are
limited for a variety of reasons. Third, abused women are likely to report more
honestly the frequency and severity of abuse to a domestic violence program because
they feel less afraid, ashamed or stigmatized by a shelter or womens advocate than a
police officer, doctor, or researcher (Smith 1994). These issues are more fully
addressed in Chapter Three. Fourth, domestic violence programs capture the reality of
living with domestic violence on a daily basis. Chapter Four more fully discusses the
strengths and limitations of program data.

There are two primary measurements of criminal victimization in the United
States (U.S. Department of Justice 1994). The oldest is the Uniform Crime Report
(UCR), which measures the number of arrests reported by participating law
enforcement agencies. The other is the National Crime Victimization Survey (NCVS),
which is designed to measure reported and unreported crime from the perspective of
the victim (National Institute of Justice 1994, Biderman 1991). The Federal Bureau of
Investigation (FBI) compiles the UCR and the Bureau of Justice Statistics collects the
NCVS (National Institute of Justice 1994; Biderman 1991). Both measure the
incidence and rate of crime. Although neither specifically collects data on domestic
violence, these numbers can be extrapolated from the NCVS reports based on victim-
offender relationships. The UCR collects victim-offender relationship data only on
homicide cases (National Institute of Justice 1994). Violence against women surveys
and family violence surveys are other tools used by researchers to understand the
incidence and prevalence of violence in the United States. This chapter will discuss
how these different research methods measure the incidence of domestic violence and
the limitations of the resulting data.

The Uniform Crime Report
Since 1929, the Uniform Crime Report (UCR) has been the official source of
information of crimes reported to the police. The most commonly used series of the
UCR is the Crime Index (Biderman 1991). The Crime Index is a compilation of seven
major crime categories and is used as an indicator of the amount of serious crime
occurring in the U.S. (Biderman 1991, U.S. Department of Justice 1996). The UCR
consists of voluntary monthly reports from police and sheriffs offices across the
country. Data are collected on characteristics of individuals who are arrested. The
UCR collects data on crimes that are prevalent, serious, and well reported to the
police, (Biderman 1991:) but the UCRs Crime Index is inadequate as a
comprehensive measurement of domestic violence. This inadequacy presents a major
data problem, especially because statistics indicate that domestic violence-related calls
comprise a significant percentage of police calls (Zorza 1992 Yllo and Bograd 1988).
The current UCR system is unable to differentiate domestic violence calls from other
acts of violence because it collects victim-offender relationship data only for homicides
(U.S. Department of Justice 1994, Biderman 1991). In order to determine whether a
simple assault or even a burglary was domestic violence-related, the UCR would have
to include relationship of victim-offender in all of its criminal data. Despite this
substantial limitation, UCR data are frequently cited for national, state, and local crime
trends (U.S. Department of Justice 1996).

In order to address this limitation, federal officials are in the process of
converting the UCR into a more comprehensive and detailed database known as the
National Incident Based Reporting System (NIBRS) (U.S. Department of Justice 1995
and 1994). This conversion represents a significant shift in the way crime statistics are
compiled. There are six categories of data being compiled in the NIBRS data
collection system: 1) Administrative data, 2) offense data, 3) property data, 4) victim
data, 5) offender data, and 6) arrestee data. NIBRS has classified 22 categories of
offense as Group A offenses, which exclude domestic violence (U.S. Department of
Justice 1995). According to the NIBRS categories, Family offenses, nonviolent is
included in Group B offenses along with bad checks, disorderly conduct, peeping tom,
and drunkenness (U.S. Department of Justice 1996). This categorization is another
example of criminal justice system administrators placing a low priority on domestic
violence cases. It raises key data collection questions. When is a family offense
nonviolent? Why is it included with minor criminal acts?
Moreover, the NIBRS data will fail to capture the vast amount of crime that is
never reported to the police. Kidd and Chayet (1984) argue that there are three
primary reasons why victims fail to report their victimization to the police. First,
victims are afraid of their perpetrators response. In the NCVS, researchers found that
six times as many women victimized by intimates as those victimized by strangers did
not report the violence because they feared retaliation from the offender (U.S.

Department of Justice 1994:5). Second, victims feel helpless and perceive the police
to be powerless. Third, victims worry about the threat of further victimization from
authorities. When a victim has to retell her story to police, victim advocates, district
attorneys, and judges, it can be very re-victimizing. These three factors are almost
always present in domestic violence cases.
Official statistics need to'be considered within this reporting climate and in
light of changing laws (Gartner 1995). According to Gartner (1995), the official rate
of domestic violence may reflect changes in victims willingness to report. She notes
that Citizens report crime at different rates in different localities and police record
events differently in different localities (Gartner 1995:396). Gartners research
provides further evidence that victims of violence, and especially victims of intimate
violence, rarely resort to the criminal justice system. Gartner (1995:394) insightfully
asserts that:
Because women are most likely to be victimized by people they
know well, and because people victimized by those they know
well are less likely to inform authorities; official information on
violence against women inevitably underestimates its prevalence
and misrepresents its character.
National Crime Victimization Survey
When U.S. government leaders declared a war on crime in the mid 1960s,
officials wanted to go beyond official reported crime to what they call the dark figure
of crime (Stanko 1988). Thus, the National Crime Victimization Survey (NCVS) was

developed to uncover the prevalence of unreported crime (Stanko 1988, U.S.
Department of Justice 1996, Biderman 1991). The primary task of the NCVS is to
measure the same kind of crime by the same definitions as the UCR data measure,
but to include unreported crimes as well as those that are reported (U.S. Department
of Justice 1996, Biderman 1991). The NCVS also collects data about victims and the
consequences of crime (Biderman 1991). The U.S. Bureau of Justice Statistics (BJS)
operates the National Crime Victimization Survey (NCVS), but the U.S. Census
Bureau personnel conduct the interviews (U.S. Department of Justice 1996). Data
from the NCVS are based upon a nationally representative sample of 50,000
households. U.S. Census Bureau staff interview all household members age 12 and
older. The NCVS gathers data on the month, time, and location of the crime, the
relationship between victim and offender, self-protective actions taken by the victims
during an incident, and consequences of the victimization, including injury or property
loss (U.S. Department of Justice 1994). The NCVS uses particular data elements
within the incident report data to define the incident as a crime. If an incident does not
satisfy particular criteria, it is omitted from the crime count (U.S. Department of
Justice 1994). The fundamental unit of analysis in crime surveys is the victimization
instead of the person arrested (Skogan 1984, Maguire 1995).
The NCVS was redesigned in 1992 to better capture incidents involving
intimates or other family members and to improve its accuracy and use of crime

measurement (U.S. Department of Justice 1994). It now includes explicit questions
about incidents involving family members, friends, and acquaintances. The new
screening questions also include multiple references to acts of domestic violence to
encourage respondents to report incidents they may not typically define as crimes
(U. S. Department of Justice 1994). The definition of intimate partner in the NC VS
includes spouses or ex-spouses, boyfriends and girlfriends, or ex-boyfriends and ex-
girlfriends. (U.S. Department of Justice 1995). According to a 1994 BJS report,
women experience more than ten times as many violent incidents then men. (572,000
violent victimizations per year by an intimate, compared to 49,000 incidents per year
against men by intimates) (U.S. Department of Justice 1994). These numbers
reinforce the theory that in heterosexual relationships domestic violence is a problem
of male oppression of women.
Because battered women are also likely to underreport incidents to survey
researchers, the NCVS is limited in its effectiveness to estimate the rate of intimate
violence against women (Dobash and Dobash 1995, Smith 1994, Kidd 1984). Many
factors affect whether a woman will report her victimization to an interviewer. Similar
to a victims potential concern with police interaction, survey data factors include the
private nature of the event and the perceived stigma associated with victimization
(U.S. Department of Justice 1995, Stanko 1988).

Since the NCVS survey questions are framed with reference to experiences of
crime, the instrument typically discourages respondents from thinking about intimate
violence as relevant to the survey. This occurs because victims of intimate violence
often exclude their incidents from the definition of a criminal act. Moreover, most of
these surveys are not designed to take into account the sensitive nature of intimate
violence, therefore a respondent may wish to conceal these experiences in an interview
(Gartner 1995). Victimization surveys not only capture a smaller proportion of
victimizations by known offenders than stranger offenders, they also report an
unrepresentative sample of these victimizations. Victims of less serious attacks appear
more willing to reveal these incidents in surveys if the offender was a stranger (Gartner
1995). This circumstance results from a specific construction of knowledge about
intimate relations and a more general construction of reality that are both male-
dominated. Thus, women and men view violence by intimates as less serious than
violence by strangers, but not because stranger violence is inherently more serious
(Gartner 1995).
The UCR and the NCVS are limited because they both define domestic
violence as a criminal event. Gartner and MacMiliians (1995) study reinforces
criticisms expressed by criminologists concerning data analysis from victimization
surveys and official sources. Both conventional victimization surveys and official
crime statistics undercount violence by known offenders to a greater degree than

violence by strangers. Victimization surveys, as they have typically been conducted,
inadequately address biases in official statistics which is what these surveys were
initially designed to accomplish.
Violence Surveys
Murray Straus and Richard Geiles National Family Violence Survey is the
most widely cited study on family violence (Yllo and Bograd, Lerman 1992). It was
conducted in 1975 and again in 1985. This survey used the Conflict Tactics Scale
(CTS), an instrument developed by Straus and Geiles (Straus 1986). A national
probability sample of 2,143 currently married or cohabiting persons was interviewed.
The interviewers asked respondents to think of situations in the past year when they
had a disagreement or were angry with a family member and indicate how often they
engaged in each of the acts listed in the CTS (Straus 1986:467). This study was the
first attempt to quantify the incidence of domestic violence.
Both studies are notorious for finding that men and women are equally violent
in relationships (Strauus 1986). In 1975, the study found that there was husband-to-
wife violence in 121 of every 1,000 couples and that wife-to-husband violence was
present in 116 of every 1,000 couples (Straus 1986). The 1985 results revealed 113
and 121 respectively, indicating a reduction in husband-to-wife violence and an
increase in wife-to-husband violence in the second study (Straus 1986). Feminists
criticized this studys method because it fails to capture the lethality or context of a

violent act (Lerman 1992, Pagelow 1981). For example, a man who hits a woman
with a closed fist and a woman who hits a man when he is attacking her would both be
considered equally violent acts. The limitation of this study is based on the fact that
Strauss and Gelles view domestic violence as a symptom of living in a violent culture
instead of a patriarchal culture (Pagelow 1981).
In 1996, Patricia Tjaden, Ph.D. developed and conducted a national telephone
survey, titled Violence and Threats of Violence Against Women in America. The
survey involved interviews with a national probability sample of 8,000 Spanish-and
English-speaking women age 18 and older residing in the U.S. (Tjaden 1996). The
research questionnaire was designed to elicit information about fear of violence,
power, and emotional abuse, and actual incidents of or threatened violence by
strangers and non-strangers (Tjaden 1996). The questionnaire included sexual assault,
physical assault, and stalking as types of violence. As of November 1997, only the
report on stalking has been disseminated. The other reports are scheduled to be
published in 1998 (Tjaden 1996). Preliminary findings indicate that women are three
times more likely to be assaulted in some way be a male partner and 17 times more
likely to be badly beaten over a lifetime than men are by women (USA Today 1997).
Unfortunately, Dr. Tjaden was unsure of why her study found such different findings
than the Straus and Gelles study and said, that is the million dollar question (USA

Today 1997). Perhaps Tjadens inclusion of power and emotional abuse, which more
accurately encompass the definition of domestic violence, accounted for the difference.
Intimate Partner Violence Surveillance System
The Violence Against Women Act of 1994 states:
The Secretary of Health and Human Services, acting through
the Centers for Disease Control Injury Control Division, shall
conduct a study to obtain a national project of the incidence
of injuries resulting from domestic violence, the cost of
injuries to health care facilities, and recommend health care
strategies for reducing the incidence and cost of such injuries
(Title TV, VAWA, 40293(a)).
As a result of this legislation, The National Center for Injury Prevention and Control
(NCIPC) received federal funds to convene a panel of experts in 1995 to explore the
possibilities of developing a national surveillance system for intimate partner violence.
In 1996, the NCIPC funded three pilot projects. The background material, which
explains how the panel was formed, the scope of the panels work, and how the pilot
sites were chosen, was unavailable for public dissemination at the time this paper is
written. Linda Saltzman, Ph.D., a member of the Family and Intimate Violence
Prevention Team at the Centers for Disease Control and Prevention, however,
provided the list of data elements and definitions that were developed by the panel
(Appendix A).
The panel developed a minimum data set, which includes general
demographics of the victim and a victims history of physical, sexual, and

psychological/emotional abuse, including threatened physical or sexual violence. The
data set also includes information on the most recent violent episode perpetrated by
an intimate partner. These variables include details related to the episode,
consequences for the victim, demographics of the victims most recent violent
intimate partner, and perpetrator drug and weapon use. For example, item 38
provides an eight-item checklist to measure physical consequences to the victim as a
result of the most recent violent episode: pregnancy, pregnancy ended, disability,
STDs, HTV/AIDS, injuries, death or fatal injuries, and other medical conditions.
Psychological consequences are measured using a four-item choice: yes, no, not
applicable (due to death of victim), and Unknown. There are a total of 50 data
elements included in this minimum data set, ranging from date of birth to substance
abuse and mental health services received by the victim. It is evident from many of
these data elements that the medical and psychology profession were involved in
developing this data set because the nature of the questions seem more directed
towards doctors, counselors, and social workers.
The answers to the questions are collected in an investigative manner rather
than giving a survey or questionnaire to a victim. In addition, the data elements can be
gathered from any source. The sources of data that are defined in the minimum data
set are: hospital, other medical facility/clinic, emergency medical service, death

certificate, police, courts, prosecutor, social service agency, mental health agency,
shelter for battered women, rape crisis center, hotline, directly from victim, or other.
In order to pilot the data set, three states, Massachusetts, Michigan, and Rhode
Island, were awarded cooperative agreements to implement statewide surveillance
systems. The Massachusetts Department of Public Health, Division of Womens
Health received funding for its WATCH: Women Abuse Tracking in Clinics and
Hospitals program. The Michigan Department of Public Health Domestic Violence
Prevention Unit established the Michigan Prevention of Violence Against Women
Program. The Rhode Island Department of Health, in collaboration with the Family
Violence Research Program, Attorney General, Rhode Island Medical Society, Office
of the Medical Examiner, Joint Legislative Commission to Study Womens Health
Issues, and the American College of Emergency Physicians, received a grant to design
a surveillance system.
In order to understand the implementation of these pilot projects, I contacted
and interviewed site coordinators in Rhode Island and Michigan. These states were
chosen primarily because they responded to my inquiry and also represent regionally
diverse areas of the country. I also interviewed the executive directors of these states
domestic violence coalitions in order to leam about their involvement in these data
collection efforts. In Rhode Island, I interviewed Wendy Verhoek-Oftedahl of the
Rhode Island Department of Health and Deb Debair, Executive Director of the Rhode

Island Coalition Against Domestic Violence. In Michigan, I interviewed Tom Largo
from the Michigan Department of Health and Mary Keith, Executive Director of the
Michigan Coalition Against Domestic Violence.
Rhode Island Pilot Project
Rhode Island project representatives are compiling data from five different
sources (Verhoek-Oftendahl 1997). The primary source is the police reporting form,
Domestic Violence Sexual Assault 1 (DVSAl). Police officers complete this form
when they respond to a domestic violence situation. Three factors in using this form
definitely lead to undercounting the incidence and prevalence of domestic violence
(Verhoel-Oftendahl 1997). First, filling out the form is voluntary if no arrest is made.
Second, police often incorrectly interpret the policy to mean they must fill out one
report per household rather than every time an incident is reported (Verhoel-Oftendahl
1997). Third, police have not generally filled out the form in cases of restraining order
violations (Verhoel-Oftendahl 1997).
The second source of data is Rhode Islands Court Case Tracking System
(PROMIS), which tracks cases through the court system in seven out of eight districts
in the state. This database is maintained by the district attorneys office and is
designed to collect data on cases through sentencing of the offender (Verhoel-
Oftendahl 1997). The third source of data is the Restraining Order/No Contact
Registry (RONCO), which compiles data on the number of individuals receiving

restraining orders and no-contact orders (Verhoek-Oftendahl 1997). No-contact
orders include conditions placed on defendants arrested for domestic violence. The
fourth source is derived from the medical examiner records on homicide and suicide.
The fifth source of data is the Emergency Trauma Registry used by hospitals to
identify victims and refer them to an advocate (Verhoek-Oftendahl 1997). Victim and
perpetrator identifying information are encrypted to ensure confidentiality.
According to Ms. Verhoek-Oftedahl, the overall goal of the surveillance
system is to identify characteristics of domestic violence incidents. According to Ms.
Verhoek-Oftedahl, some members of the community want to use the data to determine
whether the judicial training developed under VAWA has increased the number of
severe sentences for domestic violence perpetrators. Other individuals are interested
in recidivism rates and the outcomes for court-mandated perpetrators. For example,
they ask: Are perpetrators who receive treatment re-offending less than those who
received no treatment? Ms. Verhoek-Oftedahl believes that these alternative agendas
create barriers in implementing the surveillance system (Verhoek-Oftendahl 1997).
When asked about the level of involvement of the state domestic violence
coalition, Ms. Verhoek-Oftendahl stated, it is politically unlikely that victim services
data would be included. For confidentiality reasons, Ms. Verhoek-Oftendahl believed
strongly that domestic violence program data would never be included in Rhode
Islands surveillance system (Verhoek-Oftendahl). Ms. Debair, executive director of

the Rhode Island Domestic Violence Coalition, confirmed that confidentiality issues
were why the coalition parted company with the surveillance system team (Debair
1997). The state coalition was very involved in the beginning stages of this
surveillance system design in 1995. Ms. Debair believes that including program data
was vital because women who turn to us turn to no one else (Debair 1997). During
the process, Ms. Debair notes that there was a breakdown in communication between
the state coalition and the surveillance system team. She said there was a lack of
understanding or respect for victim confidentiality (Debair 1997). Ms. Debair is
concerned that giving data to the statewide data collection system would be violating a
womans confidentiality.
Clearly the idea of integrating shelter data with police and prosecutor data
raises serious concerns among the domestic violence community. Although the
encryption system allowed for information to be reported without identifying
information being included, Ms. Debair emphatically believes that confidentiality
would be compromised in an integrated data collection system (Debair 1997). Women
in a shelter are guaranteed anonymity and confidentiality. Ms. Debair believes that it
would be possible for the encryption to be decoded by individuals in the military or
other government agencies and then confidentiality would be jeopardized for hundreds
of battered women. Ms. Debair (1997) stated, The first priority is safety and
protection of victims. If this is compromised, do we really need this kind of

database? This example demonstrates the seriousness of the confidentiality issue
within the battered womens movement.
Michigan Pilot Project
Michigan officials have developed an Adult Case Tracking System to
monitor the level of violence against intimates. In 1994, the Michigan legislature
enacted a law that mandated police officers to prepare a domestic violence report
after investigating or intervening in a domestic dispute or an incident involving
domestic violence.... (Largo 1997). Michigan law requires police officers to fill out
the form regardless of whether an arrest is made. The legislation clearly defined the
elements to be included (Largo 1997). The data include, but are not limited to,
general demographic information, alcohol or drugs involved, whether the victim
sought medical attention, amount of property damage, and a description of prior
domestic violence incidents. The law provides no specifics about how the data should
be compiled after it is collected (Largo 1997).
After the police fill out the form, the report is entered into a database at the
district attorneys office. The case is tracked all the way through conviction and
sentencing. Unfortunately, it has been difficult to achieve cooperation throughout the
state (Largo 1997). According to Mr. Largo (1997) of the Michigan Department of
Health, it was very difficult for every county to reach agreement on a common data
collection form. Therefore, members of a state advisory group decided to pilot the

Adult Case Tracking System in four counties. Mr. Largo believes that the more data
elements that are collected, the harder it will be to reach statewide consensus. He
said, If you try to do complete coverage, it is too massive. Everyone has their own
way of addressing data collection(Largo 1997). Mr. Largo believes that a
surveillance system should be stealth like so that the people collecting the data are
unaware they are collecting data (Largo 1997).
When asked about domestic violence program data, Mr. Largo said that he
believes they had too much on their plate. He said that they probably have some
representation on the advisory board (Largo 1997). Mary Keith, executive director of
the Michigan Coalition Against Domestic Violence, was unaware of this project. She
is, however, a member of the sexual assault data collection advisory committee.
Interestingly, the data collection system for sexual assault in Michigan primarily uses
service provider data (Keith 1997). Ms. Keith believes that data should focus on
information from service providers because many women do not go to police or
emergency rooms. This is only a very small piece of the pie (Keith 1997). The
Michigan state coalition has formed a technology committee, which is examining the
gaps in data collection, specifically funding for hardware, software, and training. They
have developed a five-year plan to provide all programs with computers so they can
begin collecting statistics in a uniform manner. One program in the state has already
developed a statistical tracking system for its data. Ms. Keith expects that the

Michigan domestic violence programs will have some good data on the numbers of
battered women in Michigan over the next five years (Keith 1997).
In both states, domestic violence program data are ignored. In Rhode Island,
the reason is political and in Michigan, it is an act of omission. Both examples,
however illustrate ways that domestic violence data collection systems exclude
domestic violence program data. The problem of domestic violence will continue to
be understood from the perspective of police and hospitals until the battered womens
movement organizes their data in a meaningful way.
The UCR, NCVS, and violence surveys provide some data about the incidence
of domestic violence, but each is limited in its ability to capture the scope of the
problem. The shame and stigma experienced by most victims of domestic violence
clearly will affect the ability of police, hospitals, and researchers to obtain an accurate
measurement of the occurrence of domestic violence. The Uniform Crime
Report/National Incident Based Reporting Survey provide data only on domestic
violence cases that are officially reported incidents of crime. The NIBRS system treats
domestic violence as a non-serious offense. The NCVS has made considerable efforts
to develop interview questions that help elicit more reliable information about
domestic violence. The limitation of the NCVS is that the behavior it measures still
needs to be considered criminal in order to be included in the official count. For
example, the acts of a man who forbids his wife to open the mail or go to the grocery

store alone would be excluded from the definition of domestic violence in the NCVS
because those acts are not criminal. Straus and Gelles family violence study is limited
because it lacks a social context for why and how the violence occurred. Straus and
Gelles take an objective look at violence in the home and do not address gender
inequality issues that are inherent in intimate relationships (Dobash & Dobash 1992,
Pagelow 1981). For example, a man hitting a woman and a woman hitting a man are
equal in their study. Their findings therefore are suspect and it is difficult to make
sense of the meaning. Hopefully the work of Dr. Pat Tjaden will shed more light on
the problem of violence against women, since her study used a broader definition of
violence. The next chapter will examine how domestic violence program data could
also be used to help define and measure the problem of domestic violence.

Data are collected by domestic violence programs every time program staff
talk with a woman on the phone or in person. Programs maintain statistics on the
demographics of the women and children they serve and the services that are provided.
Unfortunately, battered womens program have collected data based on federal, state,
and local funding requirements that may not reflect programs own needs. This
situation has created a reactive environment for domestic violence programs. As a
result, the battered womens movement has yet to internally struggle with defining and
developing a data collection system based on the needs of programs.
The data collected by domestic violence programs are often viewed as less
credible than police or hospital records (Yllo 1988), but domestic violence program
data should also be considered official reported data. Women call a hotline to report
their abuse just like they call the police or tell the emergency room doctor about their
abuse. The difficulty has largely been making sense of the numbers that domestic
violence programs report. Each program produces reports that are meaningful to its
particular funding community, but there is currently no ability to compile standard data
across programs.

In order to understand the scope of the information that domestic violence
programs are collecting, I conducted telephone interviews in the summer of 1997 with
individuals from eight different programs in Colorado who are responsible for
collecting and compiling data. I also examined their intake data collection forms. My
sample was regionally diverse. I chose two programs from each of the four regions
that comprise the Colorado Coalition Against Domestic Violence (CCADV). The
programs included were located in Boulder, Denver, Montrose, Alamosa, Greeley, Ft.
Collins, Durango, and Trinidad.
One criticism of domestic violence program data is that the number of women
served often contains a duplicate count of women (Largo 1997). Therefore, every
time a woman calls or comes into a program for help, she would be counted as another
client. Yet, all but one of the eight programs surveyed in Colorado track an
unduplicated count of women and children served (Cameron, Miller, Mitchell, Moore,
Olguin, Powell, Simmons, Sweet 1997). For most programs, a woman is counted as
new only when she has received no services within a year (Sweet, Cameron, Miller,
Mitchell, Moore, Simmons 1997). This would also be the same in other official data
since data is compiled on a yearly basis. One program used six months without
services as the criteria for when a woman would be considered a new client (Olguin

There is a lot of commonality among the programs regarding the data they
collect. Nearly all the programs collect information related to age, race, disability,
religion, number of children, employment status, last date of incident, number of times
police called, and number of times arrests were made for domestic violence or child
abuse. They also collect detailed information about the history of violence and the
needs of the woman.
Fifty percent of the programs use a computerized system to compile then-
monthly and quarterly statistics (Mitchell, Moore, Miller, Cameron 1997). Three
programs use a relational database and the other uses a spreadsheet software. The
other four programs calculate manually the number of women and children served and
the services offered (Simmons, Powell, Sweet, Olguin 1997). In 1996, CCADV
received funding from the state Public Utilities Commission to provide member
programs with Pentium computers, internet hardware and software, color printers, and
Microsoft Office software. These programs were also given training dollars, but many
program directors fail to realize that the Access database software on these computers
is all they need to begin computerizing and producing reports automatically. State
coalitions need to convene technology committees that could assist member programs
in becoming more sophisticated in their data collection.
All of the individuals interviewed believed that a national effort to collect
domestic violence program data would be useful. Several reasons were cited. First,

the effort would help provide a clearer idea of how many women and children
experience domestic violence. Second, this data would help secure funding. Program
representatives were asked if identifying information about victims and perpetrators
could be encrypted for security, would they be willing to integrate their data into a
state or national system. All eight programs responded affirmatively (Cameron, Miller-
Mitchell, Moore, Olguin, Powell, Simmons, Sweet 1997). The process of bringing
together diverse programs to discuss a common data collection form would also
encourage local battered womens programs to become more unified and organized.
The discussion of what data to collect and how to measure it would allow movement
leaders to begin defining for the movement what they would like to know about
domestic violence and its effects on the community.
Barriers to Collecting Domestic Violence Program Data
During my conversations, I asked what barriers these programs perceived in
including domestic violence program data in a national data collection system. Fifty
percent of them (four of eight) noted that confidentiality would be a significant issue to
overcome (Simmons, Sweet, Miller, Mitchell 1997). Two program representatives
identified funding for computers and hardware and software training for staff as a
barrier (Powell, Sweet 1997). Two programs, however, are currently working to
computerize their intake systems so advocates can enter the intake information directly
into a computer (Moore, Powell 1997). Another important issue that was raised was

the issue of duplication of women across programs. Women often move from one
county to another and may be served by multiple programs. One program
representative commented that it would be an education nightmare because people
define information so differently (Cameron 1997). Another significant barrier
identified is the time involved in designing a common data collection system (Simmons
1997). Because of apparent time constraints, many program staff may resist
committing to a computerized data collection system. Yet, if programs used a
computerized database, such as Access, staff could save significant time in the long
run which would enable them to spend more time serving women and educating the
community. Instead, each individual program spends one to three days per month
compiling statistics for funders that cannot be merged across programs. If they had a
system to integrate their data, this could serve as a way to collaborate and organize as
a collective voice.
Although there are significant barriers to domestic violence program data being
integrated into a unified data collection system, it is imperative that state and national
coalitions recognize the importance of the battered womens movement being part of
the definition and measurement of domestic violence. If the battered womens
movement stands idly by while police, hospitals, and researchers define, measure, and
compile statistics on domestic violence, the reality of battered womens lives will be
lost in the analysis.

Reliability and Validity of Data from Programs
A primary concern of the reliability and validity of domestic violence program
data is the category of women who are excluded from this data. Many women are
excluded from domestic violence programs because of discrimination. Therefore,
battered lesbians, prostitutes, women with disabilities, and women of color are all
underrepresented in domestic violence program data. Undoubtedly, they are also
underrepresented in criminal justice and hospital data. In addition, women of a higher
socioeconomic status typically use battered womens programs less and may be more
represented in police, court, and hospital records. As mentioned earlier, another
concern is that domestic violence programs fail to keep an accurate account of the
unduplicated number of women served within and between programs. A recent
comparison of official police numbers with domestic violence program numbers
demonstrates the difficulty of understanding the true scope of the problem.
The Colorado Bureau of Investigation (CBI) compiles the incident-based
reports of domestic violence calls to police departments. In 1996, the CBI report
indicates there were 9,128 domestic violence incidents reported to police, but the
report is missing data from 74 agencies, including at least four major police
departments in the Denver metro area (Colorado 1996). In comparison, the Colorado
Department of Human Services Domestic Abuse Assistance Program, released data
compiled from 39 of the 55 domestic violence programs, reporting a total of 28,554

crisis calls. It also reveals that 3,312 women were sheltered and 7,957 women were
turned away from shelters because of a lack of space (Colorado 1996). Without
reliable unduplicated data from shelter programs, it is impossible to know which of
these numbers is more accurate. One concern of the battered womens movement
should be that their inflated numbers may be ignored altogether once police and
hospitals develop scientifically reliable data.
In order to reconcile these numbers, domestic violence programs need to
develop more sophisticated ways of compiling and reporting data. If battered
womens program data lack sound methodology and organization, they may be
challenged and perhaps ignored. Instead, police, prosecutor, restraining order
registries, and emergency room data will be the only data that are considered. A
concerted national effort by state coalitions to increase funding for training and
technical assistance to their member programs would enhance their data collection and
reporting, while providing an important comparison to other officially reported data.

The criminal justice system is a patriarchal institution that provides limited
resources to women escaping domestic violence (Zorza 1992, Dobash & Dobash
1992, MacKinnon 1989). The number of battered women who rely on the criminal
justice system is a small and unique group of battered women (Keith 1997, Yllo and
Bograd 1988, Dobash and Dobash 1992). In contrast, battered womens programs
which are available 24 hours and provide anonymity and confidentiality, receive
reports from women who are afraid to call the police or tell their doctor. Residential
and non-residential programs have provided shelter, parenting support, and personal
and legal advocacy for more than 15 years. The first federal funds specifically aimed
at measuring and tracking domestic violence victims, however, are primarily directed
to the criminal justice system and the medical profession (Title IV 1994). The data
collection studies currently conducted are using only official data from law
enforcement, prosecution, and hospitals (Verhoek-Oftendahl 1997; Largo 1997).
In theory, the battered womens movement is a feminist movement. The focus
of data collection in these programs will focus on the perpetrators behavior, rather
than treating the victim and perpetrator as equals. A feminist analysis would take into
account the gender inequality inherent in male and female relationships. This kind of

analysis would provide a collective knowledge base, allowing the battered womens
movement to define itself internally and project a stronger identity externally.
Currently, there is only anecdotal information about battered womens
programswork. Designing an effective data collection system would promote a more
complete discussion of this work and help make sense out of all the data being
The battered womens movement will be celebrating its 20th anniversary in
1998. It marks an important time to reflect on the victories and the losses, as well as
develop a plan for the future. How we approach the next 20 years will be critical to
the survival of the movement. Many of the founders of the movement are still actively
involved in the struggle, but there will soon be a new generation of battered womens
activists. What will the priorities of the movement be? Will male domination and
female subordination continue to explain violence against women or will battered
womens advocates begin to view domestic violence strictly as a crime problem or a
health problem? The invaluable work of domestic violence programs in communities
across the United States should not be lost. The technology is available and it would
be shortsighted of the movement not to seize this opportunity.
Domestic violence programs currently collect and compile detailed, case-level
data on women and children living with domestic violence who receive services from
their program. Battered womens advocates should be included in the collection of

domestic violence data. This additional data would create a more accurate reporting
of the oppression of women and would better inform the development of national
The battered womens movement should organize statewide and national
campaigns in order to raise money for hardware, software, and training to be used by
domestic violence programs to establish their own data collection systems. Staff at
one Michigan program have already developed their own statistical software that will
be piloted in other programs. In Colorado, two of the eight programs interviewed are
working together to place their intake forms in a computer database. Yet, there is no
plan for those statistics to be compiled in a meaningful way across programs.
Just as the National Directory of Domestic Violence programs is a database of
programs, there should be a national database of statistics, which captures the number
of women being served by these battered womens programs. This type of database
would benefit both domestic violence programs and inform policy makers. Domestic
violence program staff spend many hours calculating data on the women and children
that they serve and they must collect it in various ways for various funders. A
statistical software program tailored to domestic violence programs would free up
staff time to provide intervention and prevention services. This data could also
provide more reflective, vital information to policy makers about the characteristics,

scope, and trends of domestic violence. Prevention efforts are better tailored to meet
the needs of families when there is an understanding of the level and type of violence
in families. Police will continue to collect crime statistics and the relatively small
number of domestic violence cases that come to their attention will be recorded, but
the data need to be complemented by a more accurate system of measurement that
includes domestic violence programs. While only some domestic violence programs
are feminist in their orientation, almost all programs collect critical data about the
frequency and severity of abuse, regardless of whether it was a crime or an injury.
More importantly, programs could consistently collect data on the number of women
who report this abuse to other agencies. The elaborate surveillance systems currently
being piloted are measuring only officially reported data to police, courts, and
hospitals. According to Kerti Yllo,
One of the deepest cleavages in the field is between mainstream
researchers and activists in the battered womens movement. I
believe that a crucial component of this division lies in the different
ways the two groups know about abuse and how those different
forms of knowledge are evaluated and acknowledged (Yllo
The adherence to the scientific method and the claim of objectivity by
mainstream researchers sets up a system in which the flow of knowledge is
unidirectional. Researchers gather and analyze data and then draw implications for
policy and practitioners. While this approach is not inherently wrong, the flow of

knowledge in the other direction is lacking (Yllo and Bograd 1988). The information
that battered womens programs are collecting is not informing research or policy.
A data collection system would involve individual programs sending their client
data to the state coalition without identifying information. The issue of confidentiality
could be minimized because these programs already have a relationship where they
share information about clients. A state would need to develop a unique identification
number assigned to each woman and a unique identification number assigned to each
intake form. State coalitions could develop an agreed-upon common code such as the
first three digits of a social security number and month of birth. Every program could
use this coding scheme and the state coalition would then gather the data and produce
accurate numbers on how many women were turned away from shelters, and how
many women called a hotline. Local programs would be more connected and the
actual need for shelter, housing, and services could be better ascertained.
State and national domestic violence coalitions need to form technology and
data collection committees and begin discussing common intake forms, definitions of
violence and abuse, software and hardware needs, and training. Movement leaders
must gain control of the measurement of, scope of, and solution to the problem and
not let it be defined using only criminal justice and health statistics. For the benefit of
battered women, the movement needs to begin the process of compiling its data in an
organized and meaningful way.


inti nK
If a
if a
d & at c ^
1/0 'i'kdVV' Y'r*'. i
L X SVt^-^Aw. ^/cd
data elements listed below are those recommended for inclusion in surveillance for intimate
er violence. Each element is numbered for convenience of presentation, and for easy
ence. They are not meant to be administered as a survey or a questionnaire, but instead are
it as information to be gathered from appropriate data sources in the jurisdiction conducting
ate partner violence surveillance. The elements can be gathered in any order, and can be
ained from one or more data sources for any given victim of intimate partner violence.
:tuse of this, skip patterns have not been used. Each element includes all possible answers,
larticular element might be inappropriate for certain victims of intimate partner violence (e.g.,
i iolent episode was fatal), there is an answer category such as not applicable that is
i ded.
Somfce of Information About Intimate Partner Violence
Viot m Demographics
Viqt ms Violence History
Physical Violence
Sexual Violence
Threat of Physical or Sexual Violence
Psychological/Emotional Abuse (Including Coercive Tactics)
Mosjt Recent Violent Episode Perpetrated by an Intimate Partner
Details Related to Episode
Consequences to Victim of Most Recent Violent Episode
Demographics of Victims Most Recent Violent Intimate Partner
Perpetrator Drug Use, Weapon Use
page 2
page 2
page 3
page 3
page 5
page 6
page 7
page 8
page 8
page 10
page 12
page 13

Case ID (a required, unique identification code assigned by the agency compiling data).
Source of information:
(Mark all that apply)
_____ Other medical facility/clinic
_____Emergency medical service (ambulatory care)
_____Death certificate
_____ Courts
_____Social service agency
_____Mental health agency
_____Shelter for battered women
_____Rape crisis center
_____Hotline/crisis line
_____Directly from victim
Birth date of victim (or age at time data were collected if birth date unknown) (or
estimated age at time data were collected if age unknown).
3A. Birth date:____
Unknown________(if birth date unknown, use 3B)
3B. Age _______
Unknown_______(if age unknown, use 3C)
3C. Estimated age
Sex of victim

Race of victim
(Mark all that apply)
_____American Indian/Alaskan Native
_____Asian or Pacific Islander
Hispanic ethnicity of victim
State of residence of victim at time the data were collected
___________(use Federal Information Processing Standards [FIPS])
County of residence of victim at time the data were collected
___________(use Federal Information Processing Standards [FIPS])
9) Marital Status at time the data were collected
Single, never married
Physical violence
If ar episode of physical violence also involved sexual violence or threats of violence, it should be
inc i ded in answers to questions relating to each of those types of violence, as well as being
inch ded under physical violence.
Has physical violence with ANY intimate partner ever occurred?

How many times has physical violence by an intimate partner ever occurred in the victims
lifetime? (If there has been more than one physically violent intimate partner, include the
total of all times for all of those partners.)
_____Not applicable (i.e., a no response on #10)
_____1-2 times
_____3-5 times
_____6-10 times
_____More than 10 times
How many times has physical violence by an intimate partner (current or former) occurred
in the 12 months prior to the date the data were collected?
_____Not applicable (i.e., a no response on #10)
_____0 (Has not occurred in the prior 12 months)
_____1-2 times
_____3-5 times
_____6-10 times
_____More than 10 times
How many times has physical violence involving the most recent violent intimate partner
occurred in the 12 months prior to the date that the data were collected?
____Not applicable (i.e., a no response on #10)
____0 (Has not occurred in the prior 12 months)
____1-2 times
____3-5 times
____6-10 times
____More than 10 times

Sexi al Violence
If £i episode of sexual violence also involved physical violence or threats of violence, it should be
inc i ded in answers to questions relating to each of those types of violence, as well as being
inch ded under sexual violence.
Has sexual violence with ANY intimate partner ever occurred?
(Mark all that apply)
_____ Yes, there was sexual violence involving physical force to compel the victim to
engage in sexual acts (attempted or completed)
_____ Yes, there was sexual violence involving the victim who, at the time, was unable to
appraise the situation, decline participation, or communicate unwillingness to
engage in a sexual act
_____ Yes, there was abusive sexual contact
_____ Yes, there was sexual violence, type unspecified
_____ No, there was no sexual violence
How many times has sexual violence by an intimate partner ever occurred in the victims
lifetime? (If there has been more than one sexually violent intimate partner, include the
total of all times for all of those partners.)
(Mark all that apply)
_____Not applicable (i.e., a no response on #10)
_____1-2 times
_____3-5 times
_____6-10 times
More than 10 times
How many times has any type of sexual violence with an intimate partner (current or
former) occurred in the 12 months prior to the date the data were collected?
_____Not applicable (i.e., a no response on #14)
_____0 (Has not occurred in the prior 12 months)
_____1-2 times
_____3-5 times
_____6-10 times
_____More than 10 times

How many times has any type of sexual violence involving the most recent violent intimate
partner occurred in the 12 months prior to the date that the data were collected?
____Not applicable (i.e., a no response on #14)
____0 (Has not occurred in the prior 12 months)
____1-2 times
____3-5 times
____6-10 times
____More than 10 times
Thn at of Physical or Sexual Violence
If i r episode of threat of physical or sexual violence also involved attempted or completed
phy: ical violence or sexual violence, it should be included in answers to questions relating to each
of tl ose types of violence, as well as being included under threat of physical or sexual violence.
Has threat of physical or sexual violence with ANY intimate partner ever occurred?
How many times has the threat of physical or sexual violence by an intimate partner ever
occurred in the victims lifetime? (If there has been more than one violent intimate
partner, include the total of all times for all those partners.)
_____Not applicable (i.e., a no response on #10)
_____1-2 times
_____3-5 times
_____6-10 times
_____More than 10 times

How many times has threat of physical or sexual violence with an intimate partner (current
or former) occurred in the 12 months prior to the date the data were collected?
_____Not applicable (i.e., a no response on #18)
_____0 (Has not occurred in the prior 12 months)
_____1-2 times
_____3-5 times
_____6-10 times
_____More than 10 times
How many times has threat of physical or sexual violence involving the most recent violent
intimate partner occurred in the 12 months prior to the date that the data were collected?
_____Not applicable (i.e., a no response on #18)
_____0 (Has not occurred in the prior 12 months)
_____1-2 times
_____3-5 times
_____6-10 times
_____More than 10 times
Psychological/Emotional abuse (Including Coercive Tactics)
Has psychological/emotional abuse with ANY intimate partner ever occurred?
Has psychological/emotional abuse with an intimate partner (current and/or former)
occurred in the 12 months prior to the date the data were collected?

In the 12 months prior to the date the data were collected, what proportion of the time did
the victim feel psychologically/emotionally abused by the most recent violent intimate
_____All of the time
_____Most of the time
_____Some of the time
_____None of the time
NotFor brevity, this term is shortened below to most recent violent episode, but be aware
that it refers to the most recent violent episode perpetrated by an intimate partner.
Details related to most recent episode
Nature of episode:
(Mark all that apply)
_____ physical violence
_____ sexual violence
_____ threat of physical or sexual violence
_____ psychological or emotional abuse
Age of victim at date when most recent violent episode began.
26A. Age _______
Unknown______(if age unknown, use 26B)
26B. Estimated age
Age of victim when services were first received for most recent violent episode.
27A. Age _______
Unknown_______(if age unknown, use 27B)
27B. Estimated age

State of occurrence of most recent violent episode.
____________(use Federal Information Processing Standards [FIPS])
County of occurrence of most recent violent episode.
____________(use Federal Information Processing Standards [FIPS])
Was victim pregnant at time of the most recent violent episode?
Was more than one perpetrator involved in perpetrating most recent violent episode?
Victims relationship to perpetrator of most recent violent episode. If a relationship has
changed over time (e.g., boyfriend/girlfriend, married, then divorced), their relationship
would be classified as the relationship at the time of the most recent episode of violence
(the one covered by the data collection). If there was more than one perpetrator, code
data on the most recent violent intimate partner only.
_____Spouse (legal)
_____Common law spouse
_____Divorced spouse
_____Former common law spouse
_____Separated spouse (legal or common law)
_____Former boyfriend
_____Same sex partner
_____Former same sex partner
Was victim cohabiting with the most recent violent intimate partner at the time of the most
recent violent episode?

What was the length of relationship with the most recent violent intimate partner at the
time of the most recent violent episode?
____Less than 1 month
How long had the relationship between the victim and the most recent violent intimate
partner been violent at the time of the most recent violent episode?
____Less than 1 month
Describe the pattern of violence with the most recent violent intimate partner in the 12
months prior to data collection.
_____The relationship with the most recent violent intimate partner and the violence had
_____The violence increased over the prior 12 months
_____The violence decreased over the prior 12 months
_____No change in the violence over the prior 12 months
How many children under age 18 were living in victims household at the time of the most
recent violent episode?
Cor sequences to Victim of Most Recent Violent Episode Perpetrated by an Intimate Partner
Physical consequences to victim
(Mark all that apply)
_____Pregnancy attributed to most recent violent episode
_____Pregnancy ended due to most recent violent episode
_____Disability attributed to most recent violent episode
_____STDs attributed to most recent violent episode
_____HTV/AIDS attributed to most recent violent episode
_____Death or fatal injuries
_____Other medical condition attributed to most recent violent episode

Psychological consequences to victim
_____Not applicable (due to death of victim)
Medical health care received by victim for most recent violent episode:
(Mark all that apply)
_____Outpatient medical treatment not resulting in hospitalization (e.g., emergency
room, or physician office visit)
_____No medical care received
_____Not applicable (due to death of victim)
Substance abuse treatment received by victim for most recent violent episode:
(Mark all that apply)
____Alcohol treatment
____Alcoholics Anonymous (AA)
____Drug treatment
____Narcotics Anonymous (NA)
____No substance abuse treatment care received
____Not Applicable (due to death of victim)
Mental health care (other than substance abuse treatment) received by victim for most
recent violent episode:
(Mark all that apply)
_____In-patient care
_____Out-patient mental health services
_____No mental health care received
_____Not applicable (due to death of victim)

Deaths related to the victims most recent violent episode:
(Mark all that apply)
_____Victims death by homicide
_____Victims death, self-inflicted
_____Perpetrators death by homicide
_____Perpetrators death, self-inflicted
_____Death of children) in the household
_____Death of unborn child/children
_____Death of someone else
No deaths
Deniographics of Victims Most Recent Violent Intimate Partner
Sex of victims most recent violent intimate partner
Race of victims most recent violent intimate partner
(Mark all that apply)
_____American Indian/Alaskan Native
_____Asian or Pacific Islander
Hispanic ethnicity of victims most recent violent intimate partner

Birth date of victims most recent violent intimate partner (or age at the time of most
recent violent episode began, if birth date unknown) (or estimated age at time most recent
violent episode began, if age unknown).
47A. Birth date:____
Unknown ________(if birth date unknown, use 47B)
47B. Age at time of most recent violent episode
Unknown____________(if age unknown, use 47C)
47C. Estimated age at time of most recent violent episode
Pen etrator Drug Use, Weapon Use
How often is alcohol use by the victims most recent violent intimate partner associated
with the violence or abuse?
____All of the time
____Most of the time
____Some of the time
How often is drug use by the victims most recent violent intimate partner associated with
the violence or abuse?
_____All of the time
_____Most of the time
_____Some of the time

What weapons were used by the victims most recent violent intimate partner in the most
recent violent episode?
_____No weapons used
_____Long gun (e.g., shotgun, rifle)
_____Cutting or piercing instrument
_____Blunt object
_____Bodily force
_____Other weapon identified___________________

Bachman, Ronet. 1995. See U.S. Department of Justice 1995.
_______________. 1994 See U.S. Department of Justice 1994.
Besharov, Douglas. 1990. Family Violence: Research and Public Policy Issues for the
90s. Response 13, no 1:6-7.
Biderman, A.D., J.P. Lynch. 1991. Understanding Crime Incidence Statistics: Why
the UCR Diverges from the NCS. New York: Springer-Verlag.
Cameron, Colin, Operations, Manager Crossroad. 1997. Phone interview by author
17 September, Denver.
Colorado. Crime in Colorado, Department of Public Safety, 1996.
Colorado. Colorado Domestic Abuse Services Calendar Year Statistics 1996,
Department of Human Services.
Debair, Deb, Executive Director, Rhode Island Coalition Against Domestic Violence.
1997. Phone interview by author 15 October, Denver.
Dobash Russell P., R. Emerson Dobash. 1992. Women, Violence and Social ChangeA
London: Routledge.
_______________. 1995. Reflections on Findings from the Violence Against Women
Survey, Canadian Journal of Criminology, July 457-458.
Fagan, Jeffrey. 1995. See U.S. Department of Justice. 1995.
Gartner, Rosemary, Ross Macmillan. 1995. The Effect of Victim Offender
Relationship on Reporting Crimes of Violence Against Women, Canadian
Journal of Criminology, July, 393-429.
Keith, Mary, Executive Director, Michigan Coalition Against Domestic
Violence. 1997. Phone interview by author 24 October,

Kidd, Robert, Ellen F. Chayet. 1984. Why do Victims Fail to Report? The Psychology
of Criminal Victimization, Journal of Social Issues, 40 no 1:39-50.
Largo, Tom, Site Coordinator, Michigan Department of Health. 1997. Phone
interview by author 24 August, Boulder.
Lerman,Lisa. 1992. The Decontextualization of Domestic Violence, The Journal of
Criminal Law & Criminology, 83 no 1:217-240.
MacKinnon, Catharine. 1989. Toward A Feminist Theory of the Slate. Howard
University Press, Cambridge, MA.
Meckler, Laura, Lack of Data Hurts War on Home Violence, The Denver Post, 30
December 1995, 6a.
Miller, Jan, Administrative Assistance Montrose Resource Center. 1997. Interview
by author 20 September, Denver.
Mitchell, Sharon, Executive Director, A Womans Place. 1997. Phone interview by
author 17 September, Denver.
Moore, Terri, Director of Operations, Denver Safehouse. 1997. Interview by author
13 September, Denver.
Murray, Susan B. 1988. The Unhappy Marriage of Theory and Practice: An Analysis
of a Battered Womens Shelter, NWSA Journal 1 no 1:75-92.
National Council of Juvenile & Family Court Judges (NCJFCJ), Project of the Family
Violence Project, Advisory Committee of the Conrad N. Hilton Foundation
Model Code 1994. Family Violence: A Model Stale Code, Reno: National
Council of Juvenile & Family Court Judges.
Olguin, Eleanor, Victim Advocate, Advocates Against Domestic Assault. 1997.
Interview by author 13 September 1997.
Pagelow, Mildred. 1981. Woman-Battering Victims and Their Experiences^ Sage
Publications: Beverly Hills & London.
Powell, Vicki, Outreach Director, Boulder County Safehouse. 1997. Phone interview
by author 13 September, Denver.

Simmons, Carol, Executive Director, Alternative Horizons. 1997. Phone interview by
author, 13 September, Denver.
Skogan, Wesley. 1984. Reporting Crimes to the Police: The Status of World Research
Journal of Research in Crime and Delinquency, 21 no 2:113-137.
Smith, Michael D. 1994. Enhancing the Quality of Survey Data on Violence Against
Women: A Feminist Approach, Gender & Society, 8 no 1 87-105.
Stanko, Elizabeth, 1988. Hidden Violence Against Women. In Victims of Crime: A
New Deal? ed. Mike Maguire and John Pointing, 40-46. Open University
Straus, Murray and Richard Gelles. 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.
Sweet, Leah, Victim Advocate, Tu Casa. 1997. Phone interview by author 13
September, Denver.
Title IV-Violence Against Women Sections 40281-40703. 1994, accessed 29 March
1996, available from
Tjaden, Patricia. 1996. Violence and Threats of Violence Against Women in America:
Questionnaire Annotation. Center for Policy Research.
USA Today 1997. Partners unequal in abuse, 18 November 1997.
U.S. Department of Justice. Office of Justice Programs. Bureau of Justice Statistics.
1994. National Crime Victimization Survey: Questions and Answers about the
Redesign, Washington D.C.
U.S. Department of Justice, 1994. Technical Background on the Redesigned National
Crime Victimization Survey, NCJ-151172.
U.S. Department of Justice. Office of Justice Program. Bureau of Justice Statistics.
1994. Violence Between Intimates, Selected Findings, Washington D.C.

U.S. Department of Justice. Office of Justice Programs. Bureau of Justice Statistics.
1995. Violence Against Women: Estimates from the Redesigned Survey, by
Ronet Bachman, Special Report, Washington D.C.
U.S. Department of Justice. Office of Justice Programs. Bureau of Justice Statistics.
1994. Violence Against Women: A National Crime Victimization Survey
Report, by Ronet Bachman, Washington D.C.
U.S. Department of Justice. National Institute of Justice. Bureau of Justice Statistics.
1996. Domestic and Sexual Violence Data Collection: A report to Congress
under the Violence Against Women Act, Research Report, Washington D.C.
U.S. Department of Justice. Violence Against Women Office. Violence Against
Women Act Fact Sheet, 1996, accessed 9 November 1997 available from
http://www.usdoi. uov/vawo/vawafct. htm.
U.S. Department of Justice. National Institute of Justice. 1995. The Criminalization of
Domestic Violence: Promises and Limits, by Jeffrey Fagan. Research Report,
Washington D.C.
Verhoek-Oftendahl, Wendy, Site Coordinator, Rhode Island Department of Health.
1997. Phone interview by author 8 August, Boulder.
Walker, Lenore. 1980. Battered Woman, Harper Colophon Books:New York.
Yllo, Kersti and Michele Bograd. 1988. Feminist Perspectives on Wife Abuse, Sage
Publications, Inc.: Newbury Park, CA.
Zorza, Joan. 1992. The Criminal Law of Misdemeanor Domestic Violence, 1970-
1990, The Journal of Criminal Law & Criminology, 83 no 1:46-72.