SPEAK OUT

TOWARDS STATISTICS
DATA SYSTEMS

Building Data Systems for Monitoring and Responding to Violence Against Women
October 27, 2000 / 49(RR11) Centers for Disease Control and US Department of Justice, as well as various legal and caregiver organisations

Violence against women (VAW) (i.e., both adolescents and adults) is a substantial public health problem in the United States. Law enforcement data indicate that 3,419 females died in 1998 as a result of homicide, and approximately one third of these women were murdered by a spouse, ex-spouse, or boyfriend. Data regarding nonfatal cases of assault are less accessible and are often inconsistent because of methodologic differences.

Recent survey data collected during 1995--1996 suggest that approximately 2.1 million women are physically assaulted or raped annually; 1.5 million of these women are physically assaulted or raped by a current or former intimate partner. Based on survey data from the Bureau of Justice Statistics' National Crime Victimization Survey, in 1998, women were victims in nearly 900,000 violent crimes committed by an intimate partner.

Some experts believe that statistics on violence against women underrepresent the problem; others believe that some studies overestimate the extent of violence against women. Such lack of consensus and confusion about the different findings from various data sources prompted the establishment of the Workshop in October 1998.

INTRODUCTION
In 1998, the U.S. Secretary of Health and Human Services and Attorney General held a briefing that focused on the nature and extent of VAW. Concerns were raised over differences among published estimates of rape, sexual assault, and intimate-partner violence and the difficulties for developing and implementing effective programs and policies.
In planning the Workshop, the Steering Committee

conceptualized VAW as encompassing many types of behaviors and relationships between victims and perpetrators. The Committee decided to focus on that subset of VAW categorized as intimate-partner violence and sexual violence by any perpetrator. Several issues were identified as needing to be addressed, including a) collection of national, state, and local VAW data from both public health and criminal justice sources to represent different perspectives; b) definitions and methodologies; and c) concerns about the availability of social services for VAW victims. The Steering Committee commissioned six background papers that targeted these issues. All Workshop participants were provided copies of these papers before the workshop. Each paper was presented at the Workshop, followed by comments from one or more respondents.

THE WORK GROUPS
Definitions of VAW should be established that are comprehensive enough to encompass women's physical and psychological experiences of violence, yet not so broad that they encompass behaviors that cannot be validly defined as VAW.

RECOMMENDATIONS
The following recommendations are categorized by several broad topics.

The term "violence and abuse against women" (VAAW) should become standard. The "VAAW" term can provide a middle ground between the desire not to muddle the generally understood meaning of the term "violence" (i.e. actions that cause or threaten actual physical harm) and the desire not to overlook psychological/emotional forms of abuse and the trauma and social costs they cause to victims. Continuing to use only the term "VAW" supports the misconception that a woman is only abused if she has broken bones or other physical injuries. Both practice guidelines and published research document he psychological and psychiatric sequelae of violence against women and the substantial use of mental health services by victims of intimate-partner violence.
The phrase "violence and abuse against women" should be used to refer to the combination of all five of the following major components of such maltreatment: physical violence; sexual violence; threats of physical and/or sexual violence; stalking; and psychological/emotional abuse.

The first three components --- physical violence, sexual violence, and threats of physical and/or sexual violence --- should comprise a narrower category of VAW. Accusations have been made that VAW statistics are falsely inflated with subjective measures of psychological abuse. With the recommended terminology and classification scheme, the first three categories can be combined and reported as VAW. All five components of maltreatment against women can still be used to represent a larger spectrum of behaviors harmful to women.

Stalking should be included as a component of VAAW.
Data should be collected on as many of the five major components of VAAW as possible, and data collection should allow for examination of the co-occurrence of the components.
Research, program, and public health surveillance data should report disaggregated statistics for each of the five forms of VAAW. Presentations of VAAW data should show cross-tabulations or Venn diagrams for all of the forms of maltreatment.
The use of common definitions and data elements should be encouraged. Uniformity of definitions and data elements will increase the reliability of VAW estimates across locale and time.
Guidelines for public health surveillance of intimate-partner violence are needed on local levels, potentially serving as a model for surveillance of other forms of VAW. Federal agencies (e.g., those responsible for addressing the legal or public health consequences of VAW) should jointly fund local surveillance efforts.
Personal interview surveys (national, state, and local) are a better tool for measuring the extent of VAW than record reviews (e.g., medical, crime, and other service delivery); however, no single or existing tool is sufficient to gauge and track all dimensions of VAW. Multiple data collection efforts and funding of health, criminal justice, and social services are needed.
Because no single measurement tool can capture all of the elements of VAAW, researchers and programs must continue drawing from existing tools and developing new measures.
Experts in several different disciplines should be encouraged to collaborate with researchers who specialize in VAW and to initiate similar research in their own fields. Disciplines that currently or could potentially conduct research on VAW include anthropology, business/management, criminal justice, demography, economics, education, epidemiology, geography, journalism/mass communication, philosophy/ethics, psychology, public health, social work, sociology, substance abuse, suicidology, system analysis/operations research, theology, urban/rural planning, and women's studies. Such collaboration might also include persons whose research areas focus on ethnicity, the behavior of boys and men, and research methodology (e.g., survey methodologists).
A chartbook or annual report should be produced to present the current available data regarding VAW. Such a report would help identify areas in the data systems that need improvement or areas in which more information is needed.

DATA SOURCES
Potential sources of local health data include emergency departments, hospital discharge records, mental health databases, medical examiner data, and trauma registries. Possible sources for local criminal justice data include databases for misdemeanors, restraining orders, court probation, and court-case tracking. Police departments, forensic labs, and district attorney offices may also provide local criminal-justice data. Service-provider data might be collected from battered women programs, rape crisis centers, protective-service programs, victim-witness advocates, teen dating violence prevention programs, child and family services, welfare offices, and school counselors.

Data used for monitoring should include past year prevalence, past year frequency, and lifetime prevalence. The lifetime prevalence calculation represents the physical health, mental health, and social consequences that can occur years after violence or abuse has stopped.
Settings and sources of information concerning VAW include employment locations; faith communities; health-care settings (e.g., emergency departments, migrant-health programs, community-health programs, maternal- and child-health programs, managed care programs, and military/veterans health services); community-based service agencies (e.g., welfare offices, child development and child care services, and day care centers); and programs for children (e.g., schools, Boys and Girls Clubs, gang programs, and programs for runaway children).

Other places where women and men congregate may provide venues for collecting information, including laundromats, hair salons, Internet chat rooms, and job training programs.
Because some victims and perpetrators of violence never seek violence-related services, monitoring systems should be implemented to estimate a) the prevalence and incidence of VAW in the general community and b) the number of persons in need of services who are not receiving them. Persons who seek such services are not likely to be representative of all victims or perpetrators of violence.

THOSE WHO CONDUCT RESEARCH
The safety of staff members who conduct research (e.g., interviewers) should also be considered. Study staff may suffer psychological distress after interviewing multiple violence victims or may fear attack from violent perpetrators.