Summary of Discussion on Safety and Accountability Audits: Workshop for Future Initiatives

Washington, DC
October 26, 2004

Table of Contents

Sarah V. Hart, Director, National Institute of Justice, U.S. Department of Justice

Director Hart welcomed the participants and said that she had been troubled in the past, as a prosecutor, at the "narrow" way that different parts of the criminal justice system viewed what they considered to be "success." For example, police would look at crime clearance rates; prosecutors would look at convictions (only things that went to trial); and courts would look at case backlogs and speedy trial guidelines. The public is interested in reasonable decisions before an "independent fact-finder" in the courts, and they are not so concerned about rates of convictions. She thought that viewing success from a connected, systems point of view could help to close the "justice gap" between crimes committed and offenders held accountable. The UK is currently working with a project along similar lines ("closing the justice gap"), using analysis and measurements from a systems point of view.

In the area of victim safety in domestic violence, there are difficult questions and trade-offs. An offender may "plead down," and then be released; or he may be given a "shot before the jury" in which the outcome is hard to predict. She noted that people in this conference have the experience and research caliber to advise NIJ on these difficult victim safety issues.

Back to Top

Diane M. Stuart, Director, Office on Violence Against Women

Director Stuart also thanked the participants for taking time from busy schedules to assist NIJ. The VAW Act is about coordinated response on both state and national levels. This requires understanding each job and viewpoint. She told of her experience working in rape crisis centers and invited policymakers to come "walk in the shoes" of her job at the time. They sat in the intake center and heard the thoughts and questions about child welfare concerns, for example. She also attended law enforcement ride-alongs to understand their experience, and she has been in the court for cross training. The study of safety audits emphasizes the effort to look at the system from the eyes of those doing particular jobs and those who have been victimized. How well are things working?

Hennepin County provides technical assistance to 15 Family Justice Centers across the United States. These centers have been struggling to find ways to co-locate services needed by victims of domestic violence. OVW and NIJ are learning about effective methods through these Centers. What inter-organizational memoranda of understanding (MOUs) and collaborative tools help to achieve better success in terms of victims' safety? She asked participants to please share what they have been learning about the implementation and process for safety audits.

Back to Top

Bernard Auchter, Program Manager, Violence and Victimization Research Division, Office of Research and Evaluation, National Institute of Justice

Mr. Auchter described the meeting agenda, which included the opportunity to hear from five experienced practitioners on safety and audit concepts. The group planned to brainstorm during the afternoon, further suggestions for victims' safety and offender accountability, based on these approaches.

Back to Top

Ellen Pence, Executive Director, Praxis International, Duluth, MN

Ms. Pence has been working in the Duluth Coordinated Community Response program and offering technical assistance on the safety audit concepts. She noted that whether an abused woman is protected or not often depends on the "mindset of the practitioner" she meets in the justice system. The idea is to look at the system from the point of view of those women whose lives are being managed. She found that her own thinking was challenged when she met Dr. Dorothy Smith, especially with regard to understanding the importance of particular coordination of workers in a system. Cases or activities are always linked, typically, by "scripted reactions" or structures; these tell the line-level officials how they should act.

Agency goals, such as "get a prosecution," could conflict with better methods for being accountable to the victim and holding the batterer accountable. Accountability means correct relations: dispatch to officer, probation to community corrections, child protective services to victims, etc. Structures are linked certain ways in the system, sometimes with unknown or undesirable effects. In the safety audit concept, these structures or "scripts" are examined from the viewpoint of those "whose lives are being managed," the victim of the violence.

Stopping the violence and victim safety should be the primary considerations. For women in violent relationships where children are affected, persons working on the case must ask, "Will this action enhance or diminish the relation of the mother to her children?" Ms. Pence noted that this topic will never be neutral. For example, often when the battered woman calls the police for help, the relationship to the children "goes downhill." Women sometimes use force with the idea that they are making a "preemptive strike," to keep family violence from continuing. She might find herself arrested and prosecuted, when she has considered the action necessary to defend herself or the children.

The safety audit is an analytical tool to find out if things are having undesirable effects. The audit brings together a team of different kinds of practitioners and domestic violence prevention advocates to analyze processes in the jurisdiction to prevent a prosecution or an arrest from unintentionally making a victim less safe. However, Ms. Pence said, "Don't tell police what to do unless you have gone around with them." The way the official reports tell about a violent event in the home often "brings the woman into the institutional context," instead of considering her perspective and different sides of her life. The audit team tries to be unaffected by turf issues or "personalities." Pence noted that, "you are not looking for bad behavior," rather you are standardizing those institutional methods that work better. In many jurisdictions the Local Coordinating Council has been helped by the audit process to make changes and move forward in areas where it "was stuck."

Methods of the safety audit include collecting data from interviews, drive alongs, observations of investigations, and analysis of administrative texts of many kinds. The program looks at all regulatory text and writing produced by responding officers or other contacts. The audit looks at appropriateness of solutions in areas where the "system seems not to work." In the effort to see what problems the woman has encountered in her experience of the system, the research often finds natural solutions.

Sometimes remedial measures are taken with the woman rather than the man, simply because he is too hard to "reach." Or child protective services might be unaware that violence is occurring in a household, and it makes too difficult demands on the woman. As an example, Ms. Pence described a woman who was sent to parenting classes as part of a service plan, without recognition of the fact that her partner was battering her. She continued failing in her plan, worrying about the children, and still getting beaten as well. The audit team looks for gaps like this and frames a set of possible solutions.

Back to Top

Dorothy E. Smith, Professor Emerita, University of Toronto, Vancouver, British Columbia, Canada

Institutional ethnography is part of sociology and has been used in the education and health fields. The study tries to document how an institution is working "as is." It does not study people's behavior or try to change things. Rather, it starts where people are and draws on their everyday experience concerning what they do and how it is coordinated. She encourages participants in audits to "use a generous conception of work." In the example of housework, analysis would include such activities as driving to work, getting kids off to school, or taking clothes to the cleaners. Ethnographic study thinks concretely about what people have to do (also outside the institutional framework).

Coordination, illustrated by how institutions standardize, write, and computerize in text, shows the way details are managed from a distance. Institutional procedures are always connected to text of some kind. Texts represent how the institution "acts" on persons' lives and time; and what efforts, risks, and conditions (like financial fines etc.) are involved. Work text covers and standardizes work sequences, personnel hierarchies, charts, and schedules. In order to find the texts, the ethnographic interviewer at the organization needs to assume the posture of "ignorance," simply listening or recording, and to avoid discussing things or arguing. They try to capture concrete examples, in plain language, to build the ethnographic analysis.

Back to Top

Questions and Discussion

One person pointed to the question, "Whose safety?" There is no "universal battered woman." Cultural aspects of the woman's life must be considered. For example, her extended family may be either a strength or a risk, and there may be immigration issues.

One participant asked whether the audit process interviews are voluntary. Ms. Pence said the study tried to recruit competent people and gather sufficient descriptions of their experience. The examiner is particularly interested in documenting institutional things that work better (and then helping the organization to make these a standard).

Currently, safety audit procedures are too dependent on "who is at the table." Not everyone on an audit team will agree concerning the work: shared definitions help with cultural influences. It is useful to have. Ms. Pence recommended considering the example of pre-sentence investigations. An auditor might check with the woman about all aspects of her experience: what she decided to wear; what technology or transportation she has had access to, and what she did to be safe. How can accountability be created? Have psychological evaluations been done? Has a guardian ad litem been appointed; and what impacts on the mother-child circumstances will that have? Police too often have little to say in such hearings, but their "piece of the picture" can teach a lot. During hearings about bail decisions, who does most of the talking? Ms. Pence said that you will usually hear about 90 percent from the defense attorney, but not enough from the victim, the advocate, or even the prosecutor. Dispatchers have an important connection to domestic crises also. Sometimes they need to keep someone on the line, and sometimes it may be dangerous to do so.

One participant suggested looking at work that women do to be safe while still living with a batterer (apart from the institutional process). The responses need a lot of individualization. In past focus groups on domestic violence survivors, about half of the measures taken to keep victims safe would not be considered safe for other victims. The key is to find out safety implications of different actions from the woman herself. Professor Smith asked why violence prevention advocates could not take care of this, using the woman's particular experience? Advocates are not trained to serve a "mediating response," and they usually have no background or administrative support to perform this function.

Back to Top

Randall H. Carroll, Chief of Police, Bellingham Police Department, Bellingham, Washington

Chief Carroll described background for the safety and accountability audit performed in his jurisdiction, in the northwest corner of Washington State, 20 miles from the Canadian border. The population is approximately175,000 and growing; they soon will have a 7th city incorporated in the county, and a couple tribal nations are in the county jurisdiction. The police department has 105 sworn officers and 56 supporting staff. Employees in the department mirror the ethnic make-up of the community, and about 25 percent of the sworn officers are women.

Chief Carroll liked the concept of doing the audit but had to struggle for funding. He looked at the background of batterers in his area and found that most of them were clearly substance abusers. In light of this, he decided to use some funding from drug seizure accounts and to seek the cooperation of the sheriff to support the audit.

Chief Carroll said that, due to limited resources, the project had to leave out several areas that have important roles: prosecutors' offices, tribal law enforcement, judicial setting, probation, advocate programs, and batterers' programs. However, the study was able to complete the audit on the victim's "entry portal" to the system, the 911-center, the jail, the Bellingham Police Department and the Whatcom County Sheriff's office. He provided updates to the City and County concerning the audit process; they have oversight for the agencies and are concerned about accountability in the system.

Chief Carroll said the audit helped to show the shortcomings and needs in the 911-process for the county. Law enforcement typically think that they are "doing things right." His agency is State accredited and had already been through other kinds of audits before. There was some apprehension about "opening the door" to the community. Through past experience, however, he has seen that unless community is included in design of service, they are likely to find fault. There are different understandings about what victim's safety means. Sometimes law enforcement service had been "considered haphazard" due to the mindset of a dispatcher or a particular officer who had been sent out. Greater knowledge among entry-portal staff benefited the responses to (sometimes delicate) safety is sues in the community.

"Mini-audit" procedures have been good for basic problem solving. Law enforcement is very "steeped in tradition," and officers have had difficulty, sometimes, to adjust their usual perceptions of how to articulate a threat to domestic situations. Inspection from the safety audit; and encouraging officers to "ask specific threat assessment questions," has been beneficial across the board, improving responses in the field and information gathered for the reporting process.

Back to Top

Sue Parrott, Director, Bellingham-WhatcomCounty Commission Against Domestic Violence, Bellingham, Washington

Ms. Parrott noted that the work for the audit is complex, with a variety of challenges and successes. It took them about a year to move from planning to implementation, and they were not done with the implementation by the end of two years. She emphasized the value of having a committed leader (like Chief Carroll) in the community. It is important to communicate expectations to all "players" and to the community concerning audit activities. Establishing the Memoranda of Understanding (MOUs) is a large project in itself. Important questions include:

Who will make sure the audit is implemented in a way that has integrity?

What will you do with the audit when it is finished?

Ms. Parrott said people need a focus on safety outcomes, program integrity, and a community base for the program. In that way, the tool builds community capacity for change and gets people to think about abused victims' safety, rather than "the way we always do things." In Bellingham, the community really embraced the audit process. The report has enhanced community relations and been a good tool for policy, risk assessment, and focus groups with victims and survivors. The audit supported formalization of the victim's input and a process for taking "an extra look" in certain kinds of cases where women are arrested.

Back to Top

Questions and Discussion

One person asked about the use of confiscated drug funds for supporting the program. Chief Carroll said he had no trouble with approval to use this money to support the safety audit. Another person wanted to know why the courts and judges were not also audited.

Chief Carroll said that there were a variety of personalities among the judges: some were young and innovative while other refused to have anyone come to the courtroom and "audit their judging." One city even had a Chief of Police that said "we do not have any domestic violence" and a judge that would not consider the audit process. The diversity of reporting systems [in the courts] is not conducive to research from the law enforcement position. It is "scary," he noted, simply to open records and methodologies to outside inspection. Other agencies have been less willing to participate.

Another person asked about the sheriff's office in Whatcom County. What has been convincing to the sheriff's department? Chief Carroll noted that the sheriff's department has had "less ownership" of the program and been slower to implement change. However, collaboration with treatment service providers and probation/parole has provided records of victims' circumstances (questions and responses), which law enforcement can use in strategies to keep the person safe. The department keeps statistics on safety measures implemented and on advocacy group assistance when appropriate. The victim is contacted the next day, or as soon as the victim can be re-contacted to check on his/her status.

In many cases, decisions are not really based on what is safe for the woman. The safety audit procedure brings attention to areas like this, where a jurisdiction's procedures need better mediating strategies to keep a victim safe. For example, immigration is sues may arise during interviews. Victims with immigration difficulties respond better if they know that they will not be deported for speaking about an incident or crisis. The law in these and other situations tends to work against women who are poor and women of color. Chief Carroll related that his department's awareness about domestic violence crises was tragically heightened through the experience of a police chief in Washington State who killed his wife and committed suicide, in front of their child. He said people in organizations have different levels of openness for change. He took the safety audit idea to the Washington State Association of Sheriffs and Police Chiefs (WASPC) that meets twice per year. Through the focus groups on children's is sues at the conference, he was able to gain support and interest in the audit process.

Many kinds of government agency could benefit from the innovative influence of safety audits. Practitioner interviews are a big part of the safety audit. Although these are very time consuming, they tell much about professional relationships and how people handle their jobs. Confidentiality and development of trusting relationships among participants are critical to audit team success. Auditors must remember that they come "to find out about the work" and not "to criticize." The concept works best as a participatory process, driven by the agency staff members' own experience. They are not "being told what to do." At the end, a sense comes out about what things have been done well, and a renewed sense of direction. The process also serves as a help to budget planning and basis for requesting technical assistance from federal or state agencies. Chief Carroll noted that his agency had expanded the use of audits from (just) domestic violence to all acts of violence. He wants to incorporate better intervention and prevention methods. Regardless of the kind of institution being audited, good results require effort and commitment. It helps to talk things over with staff at each stage to make sure critical team members understand that they have to "show up." For best results, the team must be "energized about the process."

Chief Carroll recommended wider publishing about the benefits of the process. There are many sheriffs and police leaders who "don't like new things." Safety audits could be offered (for example) as a way to help departments' integration efforts. A web-based offering might describe information and best practices in safety/accountability audits, with a "jumping off point" that can lead other agencies wishing to try the method to appropriate technical assistance. Additionally the positive results of those agencies that have completed the audit process should be stressed and contact information provided.

Back to Top

Luncheon Discussion

Participants met in small groups and discussed the use of audit procedures during the working luncheon.

One speaker noted that advocacy organizations do not accept accountability to county organizations (as though they do not "buy-in" to the jurisdiction's audit effort). Another local official has been interested in using audit methods to delve into problems of juveniles of color, particularly to study neglect issues and questions of disproportionate contact with the criminal justice system. In neglect and abuse cases reviewed in one courtroom over the last 20 years, 75 percent were repetitive cases (parents abused more than once; and parents had been abused themselves). Some kinds of "criminogenic culture" are passed from one generation to the next.

Other attendees spoke together about unwritten "peer education that occurs in institutions." Police sometimes have "rookie policies" which are different standards for handling situations on the street than are included with formal agency policy. How do the audit methods find and record that kind of information? Dr. Smith noted that the audit process captures information about informal learning. It focuses on how people learn and how they teach others. Some problems come from too much erroneous sense of accountability or responsibility. She gave an example of a steel worker who was reluctant to leave an area when a "valve blew." Since he taught others and had fixed the valve before; he did not feel right about leaving, even though it was a life-threatening emergency.

Back to Top

Charles Garrity, Dispatch Supervisor, Communications Division, Massachusetts State Police

During 2003-2004, the jurisdiction received a grant to perform an audit and the police departments jointly decided to use six representative departments to study 911 dispatches (using police records). Cellular technologies and continued increase in distribution of wireless service are not necessarily a positive influence for victim safety; but VOIP is not highly regulated at this time. Massachusetts is beginning to receive approximate location information on cellular and wireless callers. By the end of 2006 most of the state's 288 dispatch centers should have new 911 call handling equipment including mapping displays showing an approximate location of the caller. This approximation translates to an area of about100 yards, which can be of great assistance in a rural environment. However 100 yards in an urban multi-story building while still helpful, is much less so. Presently most cellular calls only provide the name of the phone owner and the location of the tower the caller is activating. A VOIP phone may look and function just like a regular phone, however when a 911 emergency call is made they may route through the primary address of the internet service provider (ISP). For example a person could be in Japan and hooked into the internet. When an emergency occurs they dial 911 and the call routes back to the dispatch center in Worcester where the ISP is located. An IP may be coming from Worcester (nearby) or could be a laptop in Japan. The indicated path for the call may not allow dispatchers to locate callers. There are 15 state control points for coordinating responses in cases of terror, nuclear crisis, or natural emergency. The Shelburne Falls communications center is one of these control points.

Mr. Garrity said there are 288 dispatch centers in the state of Massachusetts, which has a total of 355 cities and towns. The majority of the dispatch centers in the state are single person dispatch centers which reduce the ability to properly assist callers. In contrast the facility under his supervision is the largest regional communications center in the state serving 25 communities. This facility always has at least 2 dispatchers on duty and often has 3 or more. While this facility serves as a regional communications center it receives most of its funding through the state police budget. Technology presents many challenges: cellular, VOIP phones, relay centers, phone number portability, and more may cause problems in helping victims. For about half of the calls, dispatchers are able to get locations presently.

An advocacy group participated and did the interviewing. To get started, they mapped the system, did interviews, and collected data from areas that had the most information. Focus groups were held in several categories: probation, advocates for victims, and victims. Meetings occurred about once per month. The audit team analyzed the material and made recommendations. Effective call handling by the dispatcher significantly influences the quality of service from the victim's perspective. Recommendations included additional training for dispatchers in "people skills" that might improve dispatcher services to victims. The 911 call answered by dispatchers are often the first step in an abused person seeking assistance and entering the Byzantine legal system. This pivotal first step relies on the knowledge, skill, and ability of the 911 dispatcher.

Back to Top

Nancy Ferron, Project Coordinator, Domestic Violence Unit, Northwestern Massachusetts

Ms. Ferron discussed the Victim First Project, which was a safety auditing initiative for law enforcement dispatchers. This was originally funded as part of the Violence Against Women Act's grants to encourage arrest policies. The two cities in her area, Northampton and Greenfield have, between them, a population of 270,000 and 47 police departments (including campus police). About 35 to 40 percent of the District Attorney's caseload has been domestic violence cases. The DA's active role has considerably helped the audit initiative. She characterized the community as mostly white, working class, and rural. There are some Latino and some Asian, particularly in the college areas.

The audit team examined and made recommendations regarding emergency dispatch, but also separately examined police response. The police and dispatch team members provided the team with agency information (structure, regulations, policies etc.), the domestic violence advocate team members interviewed dispatchers and police officers at their work. Audit members listened to 911 tapes and reviewed police reports as a team, making recommendations for change while in the process.

This work culminated in a Safety and Accountability Audit Report which was presented to all police departments and dispatchers at a jurisdiction-wide conference. The report contains the recommendations the audit team developed over the two year project. Recommendations focused on the relationship between thorough police investigations and detailed police reports and the benefits to victim safety and offender accountability. For dispatch, necessary skills included the ability to gather needed information from a caller, while developing a mutually respectful relationship that both calmed the victim and tuned the dispatcher in to the immediate safety concerns involved in the situation. Dispatchers and police officers were left with a deeper understanding of the relationship between their actions and the actions that take place before and after their actions.

Back to Top

Questions and Discussion

Participants spoke about "people skills" for dispatchers. How should call centers develop this? Mr. Garrity said there should be in-service practice for all domestic violence dispatch. Quality in the individual response to victims comes through the dispatchers. He noted that good first responses make more impact than the departmental report forms.

Mr. Garrity said that the consultant group (Praxis) helped the agency integrate complex parts of the study. Their expertise and the help of a DV Task Force person who was knowledgeable in computer-aided dispatch (CAD) helped them to gain budget resources. Some departments used the analysis to show that they needed technology or additional staff. The audit process helped them to support a claim for what they needed, including training and other things.

Back to Top

Nancy Halverson, Corrections Unit Supervisor, Adult Field Services, Hennepin County Department of Corrections, Minneapolis MN

The local Criminal Justice Coordinating Council paid for the safety and accountability audit in Hennepin County. Initially, the police department expressed the most interest because they wanted better guidance on release decisions after booking. Now, the program is in its fourth year; and prosecution in the county has gone through a safety audit, as did 911 operations, and probation.

Minneapolis is a seven-county metropolitan area with a population of 2.6 million, which is expected to grow to 3.1 million. There are 46 municipalities and 21 percent of the population is non-white. In 90 percent of the areas, only 11 percent of residents were non-white in census counts, but the diversity is growing. Currently, about 18 percent are African American. Inside the city, 35 percent are nonwhite; and Minneapolis has, out of its total population, 18 percent African American.

An enormous immigration population has moved into the area, especially Hmong refugees and Somali immigrants (the largest of this group in the U.S.). Department of Corrections needs interpreters for over 65 languages. There have also been many Russian immigrants. Of the population growth generally, 80 percent of minorities are immigrants. Hennepin has batterer intervention for Russian or Hmong persons.

Minneapolis' inner city has about 400,000 people; and the police department receives about 21,000 "911-emergency" calls for domestic violence alone. Among the 3,200 domestic violence arrests, prosecuted cases had a very high dismissal rate (45 % convictions). There are 62 judges in five settings, of which three are in suburban locations. The safety audit did not include the suburban locations, which are very different from the main court.

Department of Corrections operates with an $86 million budget in Hennepin County. With juvenile and adult divisions together, 960 employees work for the DOC. Of full time employees, 185 are probation officers, working from 18 sites.

The safety audits have focused primarily on three things:

  1. Did partners of batterers who were on probation feel safer as a result of the partner's attendance at men's group?
  2. Is the court able to trace accountability using existing batterers' treatment programs for the probation conditions given to a batterer?
  3. Among the many (30 or more) batterers' programs, how should probation make recommendations for probationers? And, are the programs in compliance with state guidelines?

Looking at the text forms used during bail evaluation revealed that there was not much writing space to consider victim's information, which is important to safety issues. The probation office was trying to gather information to evaluate safety, but the form had left out room for the victim's viewpoint. This is a good example of (undesirable) form dictating function.

Ms. Halverson said about 1800 misdemeanor and gross misdemeanor domestic violence cases come through her office each year. A large, related topic is discrimination of the degree of risk, using appropriate assessment instruments for offenders prior to their release. Use of the assessment tools is "leadership independent," so middle managers need to be taught about their benefits and trained to use them. Also, many documents used to assess risk for criminals do not apply very well to domestic violence circumstances. Domestic violence offenders are often not very risky for other kinds of criminal offense. Some agencies do have good instruments to determine lower risk (the "DVSI" from Colorado was recommended), but the line managers are independent and might choose not to use them. Bail safety evaluations tend to look at whether the offender has a job or owns a house, but this does not measure something relating to victim safety in a domestic violence case. Halverson's staff used LSIR (level of service inventory) to test for "criminogenic factors" that might be addressed in probation. Advocates, experienced from working with battered women, can also articulate certain problems to practitioners.

Back to Top

Connie Sponslor, Military Projects Coordinator, Battered Women's Justice Project, Minneapolis MN

To set up the audit process, an agency needs considerable time (about six months) to work out "political juggling" among staff and leadership who will be affected. Ongoing consultation among related offices is important. She noted that the audit team should be sure to give credit for good things that are done and to keep a balanced view. The agency can also use resources from advocacy groups and domestic violence service providers in their area. These groups can serve on the guiding committees for the audit process. Ms. Sponslor's judicial audit involved multiple meetings with clerks, bailiffs, and judges. Focus groups were held with every "player in the court," to see what information they held or managed for the judge. Participants have to support the process and not, for example, just "decide not to do any ride-alongs that day" once things have begun. The victim is not usually present at the arraignment, and judges need to have observations concerning the "entire picture" of the victim's safety by that time.

During the first year, the police department and bail evaluators were assessed. Reports from the audits improved as the staff became more "invested." Content on the bail evaluation reports and police assessments went from about three written lines to a full report with observations. The study helped a variety of measurement activities, such as recording ratios of 911-calls relating to domestic violence, gathering data to sustain department programs, collecting observations about percentages prosecuted, and estimating program satisfaction or quality of service delivery.

Back to Top

Questions and Discussion

The process study in Minnesota also revealed that battered women are frequently very ignorant of the criminal justice process and staff roles. This can negatively impact their safety and possibly prevent favorable changes in their lives.

Ms. Halverson spoke about women who had been sentenced for perpetrating domestic violence. She said it is very hard to find women in this group who will speak. Using assistants from the Minnesota Advocates for Human Rights, the audit project was able to use a scripted list of questions with a small sample of these women who batter.

The safety and accountability audit is a tool for the legal and judicial framework. The team does not try to fix everything, but brings parties into connection with each other to improve approaches. Ms. Pence referred to Native American customs and the way that tribal nations set up their justice system. The safety audit approaches have been influenced by these methods. In tribal settings, they ask, "Does this respect the woman?" and "Does it honor relationships (beyond the individual)?" Area project staff has continuing responsibility for monitoring the status of measures chosen for implementation after an audit.

One participant recommended looking for a better understanding of non-system actions taken by battered women to keep themselves safe. In many cases, these women have said that the criminal justice system would be their "last choice" when they needed help. Identification of women's successes would be useful, whether alone or with help from the system.

Back to Top

Discussion of Other Ideas and Directions for National Institute of Justice and Office on Violence Against Women

The group agreed to move forward with safety and accountability audits, attempting to create sustainability, and using in-depth analyses with community groups. Projects formed as individual academic research are not the goal in this instance, rather better safety management in the lives of women and children. This goes beyond counting a certain number of arrests or prosecutions. It would be also useful, as a side benefit, to be able to study or test impartiality in the criminal justice process, i.e., whether certain class or race codes in the system are creating unfair results for certain groups. If about 40 or 50 audits were done, the material could be reviewed to see how successfully these "achieved a community process" and whether they genuinely helped women who experienced domestic abuse.

Participants spoke briefly about who should receive reports on the outcome of safety and accountability audits. Best designs include dissemination plans in the program from the outset. Preparation of a video might be an effective way to get the concept out to people in local jurisdictions. In Ms. Pence's experience, reports have been sent, for example, to the audited agency itself, to the Governor, to the advocacy groups involved, to the media, and in brief form to the local legislators who work on health is sues. Disseminating the results makes their effect more concrete. Sustainability of programs or redesign of administrative forms (such as a bail hearing form) depends on professional circulation of the knowledge gained in the audit.

A presentation was recently held on safety and accountability audit for a meeting of state grantee administrators. The group discussed training for additional auditors in what has become known as the " Duluth model." Audits are good for finding and resolving systemic problems, but agencies often do not know about or plan for this support in grant applications. Examining the institution's work processes can particularly support organizational change. Also, quality technical assistance to local jurisdictions, in areas like child custody and visitation, can be guided more clearly after an audit process. When a program wants to assess victim safety, the safety audit analysis lets it use terms that OMB would accept as "good evidence."

One meeting attendee suggested preparing a brochure describing the challenges and advantages of doing a safety and accountability audit. A brochure could describe good implementation methods and the kinds of advantages to which this could lead. The audit has to be run by "outsiders," who understand the audit process and offer ongoing supportive training. The up-front investment in training, staff, and resources seem large, but over 97 percent of domestic violence reports now use threat assessments for the victim. This forward step was accomplished in two years and shows the community that they are "getting certain things they want."

Professor Smith recommended focusing on a limited part of the process in order to get material that will be practical and immediately useful. Depending on local resources, you could just look at courts perhaps, she suggested. The main thing is to bring people together and solve persistent problems in the community.

Chief Carroll noted that the integrity of the process must be preserved, whether the response concerns violence against women or all violence. It improves the "way we do business" by reducing (natural) avoidance of accountability. To bring change into law enforcement organizations, it is quicker and more comprehensive to convince a supportive mayor, sheriff, or leader who will support the concept to community groups.

One group's signed commitment with the National Center for Missing and Exploited Children has worked very well to support commitment and finish deliverables. Training for safety and accountability audits that are done twice a year in that area is very popular with many groups having to be turned away. A training group should not be larger than 45 people.

Ms. Pence noted that it is best to develop the technical assistance first and then prepare the demonstration site, rather than trying to do both at the same time. One speaker noted that, you have to sell the program to the community to achieve longer-term follow-up. Resistance in communities can prevent long-term effects. In the Battered Women's Justice Project, the safety audit served as a follow-up to a "death committee" already established in the community. The process has lived on, as an accountability arm for the system. When research and social action are combined in this kind of program, the design should preferably build evaluation planning into the project from the beginning, rather than tack on research at the end.

Back to Top

Date Created: November 28, 2007