Chapter VII

Findings and Observations

This project began with a fundamental premise--that child welfare agencies need to take some action to incorporate considerations of domestic violence into their case practice. Staff from the Office of the Assistant Secretary for Planning and Evaluation (ASPE) and other sponsoring agencies viewed the evidence for co-occurrence of physical violence toward mothers and child abuse, reviewed in Chapter I, as compelling enough to warrant an examination of current activities around the country. While many of these efforts are in their infancy, sponsors felt that interest was strong enough and the issues important enough to justify this preliminary exploration and report.

With the need for action as a given, the focus of fieldwork was to discover what different agencies were doing, what lessons they had learned, and what issues agencies need to consider in order to make these new efforts work well. The following themes can be distilled from our fieldwork:

Child welfare agencies have begun initiating changes from different organizational points within their agencies and have taken different approaches to changing case practice. Each starting point has advantages and disadvantages. Agencies need to think through which approach makes sense for them.

Child welfare agencies have experience acting to protect children but are breaking new ground when they attempt to address domestic violence. These agencies cannot make appropriate changes without major and continuing collaboration with community stakeholders who work with domestic violence victims and perpetrators and know the issues involved. There are complicated policy and practice issues that can only be handled appropriately if child welfare agencies work together with people specializing in domestic violence services. Chief among these issues is the need to refrain from actions that increase danger to mothers and their children.

Changes to child welfare agency practice around domestic violence will also benefit from collaborative policy development with police, civil and criminal courts, corrections (probation and parole), the schools, and local clinics and hospitals.

This chapter summarizes and integrates findings from the site visits and literature review. Reflecting the still early and developing state of the field, we do not provide definitive examples of successful practice. Instead we review issues to consider and resolve. Our "findings" are that some challenges are common to most communities but that different agencies approach them in different ways. There is no universal resolution or "right" approach for all agencies.

The chapter first looks at approaches to changing case practice within child welfare agencies including where within the agency to start; how to expand; issues of staff motivation, understanding, and commitment; and issues of resources and tools internal to the agency. We then turn to the community context and the need to coordinate with other agencies and service providers. In the case of organizations experienced in working with victims of domestic violence, such collaboration is essential for shaping changes in child welfare agency policy and practice. Other cross-agency collaborations are critical for ensuring that new approaches to the co-occurrence of domestic violence and child maltreatment are successful, including approaches that leverage the investigative powers of the police or the enforcement powers of courts and corrections. We conclude by reviewing several complex policy issues for child welfare agencies. Many of these complexities reflect the challenges involved in balancing multiple goals: helping battered women help their children, holding perpetrators of domestic violence responsible for their actions, and working with batterers who continue to be involved in children's lives.

Approaches to Addressing Domestic Violence Issues

within Child Welfare Agencies

The communities visited as part of this study have adopted different approaches to addressing domestic violence among families involved in the child welfare system. Communities in two states--Massachusetts and Oregon--have sought to change case practice throughout all of CPS, taking an agency-wide approach from the outset. In San Diego, by contrast, the Children's Services Bureau established a newly formed unit specifically for cases active to both CPS and adult probation. The Bureau focused its initial efforts on changing practice for these cases and then training other CPS workers on how to handle domestic violence in the regular CPS caseload. In Michigan, the link between CPS and the domestic violence community was forged through the state's family preservation program, and has focused on families just below the threshold for out-of-home placement. Finally, in Hilo, Hawaii, efforts to understand and serve families affected by domestic violence have been concentrated among the CPS agency's intake unit staff.

Starting with the Whole Agency

Massachusetts and Oregon have state-administered child welfare systems in which the state agency initiated changes intended to affect all public child welfare workers. Massachusetts' efforts began in the late 1980s when staff in the state's Department of Social Services (DSS) began to meet with battered women's organizations as part of a collaborative planning effort for a federal Family Violence Prevention and Services grant. During these meetings, the "disconnect" between the child protection and domestic violence communities became apparent. DSS staff believed that their mission to protect children overrode concerns about mothers and that battered women's advocates wrongly emphasized women's rights to self-determination while ignoring children's needs. Domestic violence service providers felt that battered women were re-victimized by DSS case workers who were insensitive to the presence and nature of domestic violence and often forced women to choose between their children and the batterer. Over the next several years DSS began to redesign its approach to families affected by domestic violence. It set out to shift its clinical practice in such a way as to "protect children by protecting their mothers." DSS's Domestic Violence Program is based on the principle that the best interests of children in families with domestic violence cannot be separated from the best interests of their mothers.

In 1990 DSS hired its first domestic violence advocate. Her job was to train staff in how to identify domestic violence, explore safe interventions, and find appropriate resources in the community. DSS's Domestic Violence Program has evolved into a separate unit, based in the central office and staffed by 11 domestic violence advocates or "specialists" and two full-time staff who supervise and support the advocates and help formulate agency policy on domestic violence. In two of its local offices, DSS has piloted interagency teams that bring the strength of a multi-disciplinary approach (including domestic violence) to case consultations. Successful approaches to managing cases involving domestic violence were identified and compiled into a "domestic violence protocol" that has since been adopted agency-wide. DSS believes, however, that without the clinical support of the domestic violence specialists, the protocol alone would not be terribly effective. All DSS workers in Massachusetts have been trained in how to recognize domestic violence in their cases and how to work with these families. All newly hired social workers are trained in domestic violence as part of their pre-service training.

Oregon's efforts began in 1994 with a grant from the Office of Community Services' Family Violence Prevention and Services Program to conduct agency-wide training. Oregon's State Office for Services to Children and Families (SCF) worked with domestic violence advocates to develop and present a training curriculum in eight regional conferences throughout the state. Every local child welfare office was invited to send staff for training and most did.

Oregon also used the OCS grant to encourage dialogue and collaboration at the local level by funding seven meetings around the state focusing on building collaboration between SCF branch managers, SCF supervisors, and domestic violence program directors. The meetings included small group discussions about specific cases that involved domestic violence--what current casework practice is and how SCF might respond differently. The groups identified various ways their localities could collaborate further including holding joint staffings and training, and developing support groups. SCF is encouraging its branch staff to sit on their local domestic violence councils and invite domestic violence participation on multi-disciplinary teams.

Many local offices in Oregon are beginning to incorporate these suggestions into practice. As noted in Chapter V, however, the training proved to be only a first step. Most workers found they could not incorporate what they learned into case practice without more intensive and more structured assistance. SCF was able to provide this assistance in two counties by using leftover grant money from several federal sources to support part-time domestic violence advocates to work directly with the local child welfare workers. In both counties this training made a great difference to changing workers' understanding of the issues and their actual practice. With a second grant, Oregon is continuing the approach of locating domestic violence advocates directly in four local child welfare branch offices.

Both Massachusetts and Oregon have thus come to the same point by different routes and attest to the effectiveness of having domestic violence expertise onsite to help implement and solidify the lessons learned in training.

Starting with a Service Focus

Two interesting alternatives to beginning with an agency-wide training approach were found in San Diego County and Michigan. San Diego's Children's Services Bureau was able to build on the county's tradition of interagency collaboration by establishing a new unit (the Family Violence Project) that bureaucratically bridges CPS and adult probation. This unit only serves families active to both of these agencies (a fairly small proportion of either agency's caseload). From the perspective of CPS, a strength of this model is that it takes advantage of the offending parent's probation status to hold him more accountable than would otherwise be possible, since the CPS agency's only real sanction is to remove a child from his or her home, whereas probation can put the offender in jail. The linkage with adult probation has another important advantage in that it deals with some of CPS's hardest-to-serve families and most violent perpetrators. Even if other CPS agencies in the country do not set up such a formal arrangement with adult probation, they could gain many of the same advantages by collaborating closely with adult probation. For example, the two agencies could work together to ensure that CPS's service and treatment requirements are explicitly incorporated into the perpetrator's probation terms.⁽¹⁾ Such a collaboration can also save public resources by eliminating duplication of effort around drug testing and other monitoring activities.

An important limitation of such an approach is that only a small share of CPS cases affected by domestic violence is likely to be eligible for the program. To serve all families affected by domestic violence more effectively, CPS needs to go beyond a narrowly focused unit such as the Family Violence Project. San Diego is doing this by building on Family Violence Project staff's expertise and knowledge to develop training curricula and protocols for all CPS agency case workers. Family Violence Project staff (along with community groups such as the police department and battered women's advocates) train other CPS workers directly and consult with workers on cases involving domestic violence. This model of first establishing a specialty unit that is ultimately expected to help modify case practice throughout the entire agency, is an interesting alternative to trying to change case practice through agency-wide training alone.

Michigan's initial collaborative efforts focused solely on its family preservation program, Families First, for several reasons. Both Families First and domestic violence programs share the goal of violence-free families, an empowerment philosophy, and strength-based service interventions. The family preservation model appears to be ideally suited to working with women and their children who are in or recovering from an abusive relationship, because it focuses on strengths within the family and empowering the family to use those strengths to change.

Administrators have taken very seriously the task of training all Families First workers, including those who receive referrals from CPS and those who receive referrals from shelters to respond to domestic violence. In conjunction with Washington state's HomeBuilders program, Michigan's Families First worked with the Family Violence Prevention Fund in San Francisco to develop a domestic violence training curriculum specifically for family preservation workers (Schechter and Ganley 1995).⁽²⁾ Elaborate training on domestic violence allows Families First workers to respond more effectively to the needs of families referred by CPS. Training is also important for safety reasons. Family preservation workers in Michigan were among the first to request training because in spending 10 to 20 hours in a family's home each week, domestic violence issues in the family threatened not only children's safety but workers' safety as well.

Michigan's Families First is the largest family preservation program in the country, offering services to almost 40 percent of the state's CPS caseload. The state's focus on training Families First workers as a starting point may benefit almost half of CPS families. Still, Michigan has begun the process of expanding this collaboration beyond family preservation and is currently in the process of developing a training curriculum and other policy changes for all CPS workers in the state.

Another aspect of Michigan's collaboration that began as a demonstration but has since been institutionalized and expanded provides funding for some domestic violence shelters to hire their own Families First staff. These shelters can refer to their own Families First worker any families at risk of homelessness and those living in abusive or neglectful environments that pose a potential danger to a child. However, abuse or neglect of children that meets mandatory reporting laws must still be referred to CPS.

It is worth noting that both agencies that adopted relatively narrower approaches to integrating domestic violence within child welfare (San Diego and Michigan) are working toward broadening their efforts within their agencies. Sometimes this expansion is informal, as when a specialized worker shares his or her newly developed expertise with other child welfare workers on a personal basis. Sometimes the expansion is formal and involves training and support for all CPS staff. This movement suggests that wherever one starts within the system, a comprehensive approach to domestic violence within the child welfare system is ultimately needed. It is important to start somewhere.

Prerequisites for Success within CPS

A key lesson from this study is that even with the investment of financial and other resources, changing CPS case practice around domestic violence requires a great deal of time and commitment. Key elements for success are:

Basic levels of awareness, understanding, and motivation around making needed changes, and

The tools and other resources needed to act safely and effectively on this awareness and understanding.

All of these elements are necessary. The elements of the first group--awareness, understanding, and motivation--are closely related to one another but are not the same. Awareness refers to having a basic knowledge of the problem of domestic violence, including knowing that it may be present in a family and needs to be addressed. Understanding domestic violence implies a deeper and more thorough comprehension and familiarity with the many dimensions of domestic violence, its underlying dynamics, and the many different manifestations of the problem. A full understanding of domestic violence also means that one knows safe and effective ways of working with families affected by such violence. Finally, motivation refers to a willingness to do something about domestic violence once one is aware of it. One can be aware of a problem without fully understanding it or being motivated to do anything about it. Similarly, one can be motivated to address a problem one is aware of without fully understanding many of the complexities involved.

Tools and resources are two additional distinct elements with more concrete characteristics. Tools include specific screening, case management, and other types of protocols, as well as techniques for investigation, assessment, safety planning, and referral. These types of tools are often included in CPS domestic violence protocols like those adopted in Massachusetts and San Diego County. These protocols offer case workers concrete guidance on how to interview family members, (e.g., they should be interviewed separately and in a certain order), specific questions designed to uncover domestic violence (e.g., children may be asked "When mommy and daddy fight, do they fight with words or with their hands?"), and advice on how to assess the level of risk to the mother and child(ren), the offender's lethality, and the impact of exposure to violence on children. Case workers also need tools to design appropriate service plans, to help mothers and children develop safety plans, and to document information in case records and other official records in ways that are thorough yet safe and confidential. Resources are both internal and external to CPS and include access to information on previous civil and criminal charges involving family members, effective treatment options, probation and court monitoring of perpetrators, and other supports such as domestic violence specialists. These resources benefit both case workers (by giving them options and increasing their flexibility in meeting family needs) and families (by providing the most appropriate services and treatment alternatives).

Interplay Among the Five Elements for Success

Many CPS agencies have succeeded in raising awareness about the need to screen cases for domestic violence. This is often accomplished through pre-service or in-service training. Frequently, however, workers receiving training are not given the necessary tools and resources to change case practice for families affected by domestic violence. These tools include well-developed screening questions, case management protocols, knowledge of domestic violence laws and services, and safety planning techniques. Important resources are experts or other knowledgeable individuals whom case workers can consult when faced with dangerous, unusual, or unexpected situations; and resources outside the agency including appropriate treatment options for batterers, advocacy and supports for survivors and children, and an effective civil and criminal justice response to the crime of domestic violence. A critical part of the training process must be educating workers about the existence of these outside resources and how they and the families they work with can access them. If critical external resources do not exist, CPS can be instrumental in collaborations that will develop them.

CPS agencies that appear to have high levels of understanding and motivation but lack needed tools or resources include Hilo and one local site in Oregon. CPS workers in the Oregon community reported that the danger they faced had increased since they began focusing on domestic violence in their caseload. These workers, who were aware of the importance of domestic violence in their caseload and were motivated to respond, were not initially provided with the depth of knowledge or appropriate procedures and protocols to address the issue adequately once it was identified in a family. In addition, their training did not cover how to work with batterers without being manipulated. Despite agreeing that working with these families was important, workers in this office described the situations they found themselves in as very dangerous. They said they were hesitant to take these cases and staff turnover had increased with the new policy.

Examples of what happens when one has many tools and resources but little motivation were found within some local CPS offices in Massachusetts. All local agencies were relatively resource-rich in that each had an in-house domestic violence specialist who was available on a part-time basis to provide ongoing education and training. However, some CPS case workers and even their supervisors lacked awareness and understanding about domestic violence, and thus were not motivated to consult their in-house specialist. Many domestic violence specialists reported that much of their initial work at the local agency involved "drumming up business" by conveying to workers why and how they should use a domestic violence specialist. Not surprisingly, local CPS agencies and individual case workers varied greatly in their level of understanding and acceptance of these issues. Cultivating awareness and understanding of an issue takes time; one should not be discouraged if motivation among CPS workers is not universally high at the outset. In other local agencies in Massachusetts where workers' motivation was very high, the resources needed to support the workers were readily available. Not only did battered mothers and their children benefit from improved services, but the CPS social workers were much more confident that they knew how to handle such cases well.

Fostering awareness, understanding, and motivation among case workers and their supervisors and providing them with useful tools and resources requires both initial education and training and ongoing technical assistance and support. Initial and ongoing training help to establish an awareness and understanding of the complex issues involved and to maintain the motivation needed to succeed. Similarly, providing line workers and their supervisors with technical assistance (e.g., on how to use new protocols and how to access needed resources) helps them implement changes by giving them the specific tools they need to approach potentially complex and dangerous situations safely and effectively. Workers' continued motivation to approach domestic violence cases in new and creative ways is further enhanced as they use these tools and observe their benefits directly.

A branch manager in Oregon who oversees two counties--only one of which had an on-site advocate for six months--noted a dramatic difference in the ability of the respective line staff to integrate information from training into their work with families. The office with the on-site advocate made much greater strides in understanding, motivation, and creative case planning. In both Oregon and Massachusetts, the presence of an on-site advocate had two major benefits: knowledge and time. The domestic violence advocate worked to promote understanding and awareness among the workers and was able to change the way specific cases were handled. Through case staffing and informal conversations, she also facilitated the process of translating general knowledge from the training to actual changes in practice. In addition, the advocate spent a great deal of effort directly supporting and counseling women in abusive situations, an activity that the CPS workers themselves rarely have the time to do.

Family preservation workers in Michigan receive in-depth training that includes specific tools for working with families in domestic violence situations. Workers receive training both to work with women who are in the process of leaving their batterer and establishing new households, and to work with families where domestic violence is an active issue and the batterer is still in the home. One intent of the training is to increase workers' ability to effectively solve problems together with the woman in creating a safety plan for her and her children. One Families First worker explained how one of her clients, a victim of domestic violence, had no phone at home. As part of her safety plan, the woman worked out an agreement with some of her neighbors to call the police if she signaled she was in danger by turning the porch light off and on. Workers also provide women with information on available resources in the community, identify supportive friends and family to whom she can turn for help, and conduct role playing. Families First workers also receive ongoing support from trainers knowledgeable about the issue in order to work effectively with these families. In addition, these workers have an important resource that traditional CPS workers do not: time to work with families intensively for up to 20 hours per week.

Who Should Be Involved

Effectively serving families impacted by domestic violence requires the involvement and commitment of individuals at all levels within CPS. Some examples of what can be initiated at various organizational levels are shown in Exhibit VII.1 (for CPS Administrators/ Managers), Exhibit VII.2 (for CPS Supervisors), and Exhibit VII.3 (for CPS Social Workers/ Caseworkers). Most of the examples at the higher administrative and managerial levels within CPS relate to agency policy. These include allocating resources, developing new criteria for screening, assigning, and investigating cases, and adopting new case management protocols. The agency's overall philosophy toward domestic violence and battered women is also very important, and can strongly affect the success of other efforts within the agency to help families in need.

Exhibit VII.1

Examples of What CPS Administrators/Managers Can Do to

Integrate Domestic Violence Issues into CPS Case Practice

Exhibit VII.2

Examples of What CPS Supervisors Can Do to

Integrate Domestic Violence Issues into CPS Case Practice

Exhibit VII.3

Examples of What CPS Social Workers/Caseworkers Can Do to

Integrate Domestic Violence Issues into CPS Case Practice

Efforts by CPS supervisors can span both policy and practice areas (see Exhibit VII.2). They set the tone for their unit of caseworkers, can encourage and support caseworkers to take advantage of training opportunities, and can act as an important back-up for caseworkers through case consultations and in other advisory and supervisory processes. Supervisors can also advocate on behalf of caseworkers and the families they are helping by communicating training and other resource needs to higher level managers. Finally, there are a number of changes in case practice that social workers and other line workers can make to improve services to families affected by domestic violence (see Exhibit VII.3). These include becoming informed about domestic violence, learning how to screen for and identify it, placing a high priority on the safety of mothers as well as that of children, and understanding that in some cases CPS can best help children by helping their battered mothers.

The efforts underway in the five communities visited as part of this study range from very low-cost options (adopting new questions and techniques when interviewing mothers and children) to more expensive ones (hiring full-time domestic violence specialists to consult with CPS caseworkers and model best-practice approaches). Some of the most important and effective prerequisites to bridging child welfare and domestic violence services are relatively low-cost. These include participating in community-wide interagency forums, learning and raising others' awareness about the overlap between child maltreatment and domestic violence, and supporting other agencies' efforts to improve services for families experiencing violence or abuse.

Low-cost activities are a way to begin changing CPS practice. But as Massachusetts' experience makes clear, accomplishing system-wide change will require substantial resources. It may be worthwhile to develop methods to justify these resources, such as documenting the number of families in need, tracking these families over time, and observing if they are more likely to re-enter the child welfare system because of unaddressed domestic violence concerns. Some of these methods are fairly easy to do and may help attract the additional resources needed for more costly options such as hiring in-house domestic violence specialists.

There is no single "right" approach; each one has advantages and disadvantages. The appropriateness and success of any one approach strongly depends on the conditions and circumstances within a particular agency and, as the next section makes clear, the larger community. Agency- and community-level factors that should be considered include whether CPS is administered at the state or county level, degree of local agency/individual unit autonomy, ease of establishing and monitoring the effort, willingness of staff at various levels to take on new initiatives, previous experiences with new case management approaches and innovation, the community's overall level of responsiveness to domestic and family violence, and the level of resources available to devote to the effort. Other important considerations are the expected longevity of the initiative and, in the case of efforts that are limited to one part of the agency, the potential to impact other portions of CPS.

The Importance of the Larger Community

Many agencies and individuals may be involved in responding to family violence, providing many opportunities for cooperation and collaboration. This study found many examples of such collaborations. The Family Violence Project in San Diego was the result of a formal agreement between the adult probation department and CPS. This link also gave regular CPS staff access to critical information on criminal histories from the criminal justice system. CPS and domestic violence agencies in several Oregon counties developed extensive collaborative linkages including training, case consultation, and therapeutic services for children. In another Oregon community, CPS and a local mental health agency collaborated to develop a batterer intervention program in a community that did not have one. The existence of this new service, in turn, stimulated judges to include participation in the treatment program as a condition of restraining orders and probation. Since batterers had to pay for participation themselves, this further increased service availability. In Hilo, the judge screens cases on the domestic violence docket and refers them to child welfare if children appear to be in danger. Below we review the many community-wide issues that CPS and other agencies involved in helping families affected by domestic violence will need to consider. We examine various aspects of the linkage between CPS and the domestic violence service community, as well as other community-wide collaborations, especially those that foster information sharing and expand access to treatment services. We conclude with a discussion of more general community characteristics that may shape efforts to coordinate services for CPS-involved families affected by domestic violence.

The Domestic Violence Community

To change policy and case practice around domestic violence safely and effectively, CPS must consult with and involve battered women's advocates and other experts from the domestic violence community. Without the input of domestic violence specialists, CPS is likely to make serious and potentially dangerous mistakes. This input must be sought despite the history of hostility and antagonism between child protection workers and domestic violence advocates in many communities. Bridging this gap is an important part of coordinating services for families. Professionals from both the child welfare and domestic violence communities have much to learn from one another. As they begin to communicate and build trust, both sides are likely to evolve and benefit. Child welfare workers can come to appreciate that in many cases children can be protected best by protecting their mothers. Battered women's advocates can come to appreciate that CPS has many legal obligations affecting its work with families, and that it must act in the best interest of children when there are conflicts with that interest.

As they work together, the two communities can recognize that while they see many of the same families, there are also important differences in the populations being served. As discussed below, domestic violence specialists based in CPS agencies have found that many cases are much more complex and dangerous than those they encountered in battered women's shelters. Once CPS and the domestic violence community begin to share their concerns, experiences, and perspectives, they can begin to make progress in coordinating efforts to serve all families in need.

An important issue that arose in several sites concerned the appropriateness of existing battered women's service providers for helping CPS-involved battered women. Communities with well-established child welfare-domestic violence collaborations, including several in Massachusetts and Oregon, provided anecdotal information about the differences between CPS-involved battered women and women who traditionally seek help from battered women's shelters. Most battered women who have turned to emergency shelters and other support services have done so on their own initiative. In contrast, CPS-involved women are generally receiving treatment as part of an externally imposed service plan to keep custody of their children, and may not have reached the point of acknowledging the danger to themselves and their children. If they do seek shelter, it is often because CPS has told them they must do so or lose their children. Often these women are not willing or ready to end their relationships with their partners.⁽³⁾

Some communities have found that CPS-involved women are more likely than women traditionally served by battered women's shelters to have active substance abuse and serious mental health problems. They are also more likely to have older children and children with special needs. Finally, some of these mothers are teen mothers or may not speak English well.⁽⁴⁾

One key question that arises from these observations is whether or not existing battered women's shelters are able or willing to accept CPS referrals. Many shelters have requirements such as sobriety and participation in group counseling that a newly referred CPS mother may not satisfy. Group counselors at shelters in Massachusetts, for example, found that the mandated participation of some CPS-referred mothers was undermining their effectiveness with other shelter victims. Some of these CPS-referred mothers were unwilling to admit that their partners had been abusive. As a result, a separate off-site support group designed especially for CPS-involved mothers was established. It is co-led by a shelter worker and a CPS domestic violence specialist.

Another feature that distinguishes CPS from voluntary domestic violence programs is that CPS must work with the child's (abusive) father if he lives in the child's home or if he wants to remain involved in the child's life. CPS must also work with any other battering partners who live in the child's home. This can be difficult if the mother is not prepared to leave her abusive partner. Anecdotal information suggests that CPS-involved perpetrators of domestic violence may differ from many of their non-CPS counterparts, just as their victims often do. In Massachusetts, the clinical director of a batterer intervention program that serves both CPS referrals and court-ordered perpetrators observed that CPS-involved men tend to be more pathological, dangerous, and generally lower functioning. He also remarked that they are younger and less educated than their non-CPS counterparts, and because they are not court-ordered into treatment, are much less likely to complete the program.

Differences between CPS-involved families and shelter-involved families extend to the children as well. Two CPS offices in Oregon collaborated with local domestic violence programs to develop therapeutic support groups for children who had been exposed to domestic violence. Both pilot counties had difficulty establishing these groups because the children were struggling with many problems in addition to domestic violence, including their own physical or sexual abuse. In addition, most of the children were in unstable living situations either because they were in temporary foster care or because they remained with their mothers in volatile and potentially unsafe environments. Coordinating transportation for the children in foster care from locations that were both geographically dispersed and temporary proved very challenging. One of the specialists also noted that the children needed individual therapy before they could benefit from group therapy, and that the group required almost one adult per child to keep things on track.

Communities may want to consider whether existing domestic violence programs can or should absorb battered women and their children identified by CPS or if existing agencies should work together to develop spin-offs or new services. If existing domestic violence services are already at capacity or are not always appropriate for families involved with CPS, the option of developing new services may need to be considered. Just as child welfare organizations in many communities created the demand for new approaches and services for child sexual abuse, CPS agencies should recognize that if they choose to reallocate funding they can also foster the development of domestic violence services appropriate for their families. It should be stressed, however, that whether they use existing or newly established services for battered women and their children, CPS should design these services in conjunction with battered women's advocates and other experts within the domestic violence community, to be sure that services do not further endanger or victimize mothers or children. Such a collaborative approach can also ensure that services are not duplicated and foster positive working relationships by sharing ownership with the domestic violence community and establishing a two-way education process.

By addressing domestic violence issues within the context of child welfare, CPS can be the impetus for improving community-based domestic violence services more generally. Some battered women's service providers may be unable or unwilling to serve CPS-involved women for a variety of reasons. Others may recognize that only a small share of battered women in need of help ever enter their shelter system, and that by coordinating with CPS and developing new approaches for this population, they can reach a larger group of battered women who would otherwise remain unserved.

There are many ways for CPS agencies to take advantage of existing domestic violence resources and expertise in the community. They can bring these resources directly into their agency, for example, either by hiring domestic violence specialists (as is being done in Massachusetts), by contracting with local domestic violence programs to provide staff to work with CPS workers and families (as in Oregon's pilot sites), or by inviting domestic violence advocates to sit in CPS offices for case consultations on a weekly or monthly basis. They can also encourage other private providers with whom they contract to address domestic violence issues (as Michigan is doing with its family preservation services). Having domestic violence specialists on staff demonstrates a high level of commitment to domestic violence concerns and builds bridges between the domestic violence and child welfare communities. The need for such bridges is clear: many domestic violence specialists hired by CPS initially experience tension from within the child welfare system and also from their former colleagues within the domestic violence community.

Other Options for Collaboration

Helping CPS-involved families affected by domestic violence cannot be done by CPS alone. Part of any effective intervention with families subject to domestic violence is to hold the perpetrators of such violence accountable for their actions and to provide the resources needed to promote the safety of adult victims. CPS has some role to play in this effort (e.g., by placing responsibility on the batterer rather than charging the battered woman with failure to protect her children). But offender accountability is primarily the responsibility of the criminal justice system, including law enforcement, prosecution, the courts, and corrections. CPS's only "stick" is the threat of removing a child from the home: if only the mother cares about losing her children, this threat is not a deterrent for the batterer. According to domestic violence advocates in several communities, batterers often use the threat of CPS action as another lever to intimidate and control their partners. To hold the ultimate perpetrator accountable, CPS must be able to depend on outside agencies to respond consistently and effectively to the batterer's criminal behavior. In addition, CPS agencies must be able to rely on other organizations in the community to provide many of the treatment services these families need.

Community Coordination and Information Sharing

In some communities an effective response to domestic violence exists, but CPS has not used the services of domestic violence programs or the criminal justice system to its fullest potential. In other places, existing services are not always appropriate or sufficient for CPS to refer all families in need. Because the families seen by CPS often have many problems, they may not fit the client profile accepted by existing community agencies. In any case, communication between the multiple agencies involved is critical to inform CPS of existing services and to enable workers to gain access to them, and for communities to develop new services to meet the unique needs of CPS families in domestic violence situations. Effective communication can begin to break down mistrust and misunderstandings. Many communities have built effective responses to domestic violence through formal task forces, but local child welfare agencies are rarely involved. In some cases they have not been invited and in other cases they have chosen not to attend. Conversely, few communities involve domestic violence representatives on their multidisciplinary child abuse and child fatality review teams, despite growing evidence that battering is often present in these cases.

Many communities have found formal taskforces and councils to be an effective means of beginning the process of collaboration. Some communities, however, have based their interagency collaborations on more informal connections, including individual-level relationships. Personal working relationships are an important element of successful collaboration, but enlisting the formal support of top-level administrators can accelerate the process of change and make it more consistent. Collaboration that increases the flow of information between agencies and develops and works toward shared goals can help different agencies send a consistent message to all family members.⁽⁵⁾ Consistency is critical, both to ensure that a perpetrator is unable to manipulate the system to his advantage and to assure a victim that her family will be safe after contacting the police or taking other official action. Prompt police response, prompt issuance and effective enforcement of restraining orders, and prosecution of violations of orders involve many players including law enforcement, prosecutors, judges, and probation officers. Furthermore, a coordinated community response to domestic violence allows families to access appropriate services regardless of how or where they enter the social service or justice system--by calling 911 or a domestic violence shelter or crisis line, applying for a temporary restraining order, or having contact with a child welfare agency (Clark et al. 1996).

CPS agencies in communities such as those examined in Hawaii, Massachusetts, and Oregon, have formal access to criminal justice information. Others rely on the benevolence of particular police officers or are limited to one computer terminal or one designated worker within the agency to obtain criminal histories. Other information-sharing arrangements were also found in the states and communities we visited, including sharing information between probation and CPS, and joint visits conducted by family preservation or CPS workers with law enforcement officers to discuss potential legal ramifications for the batterer. Many professionals throughout the country are mandated to report child abuse to their state child welfare systems. But opening lines of communication and building trust have inspired judges, prosecuting attorneys, hospital workers, and battered women's advocates in some communities to probe more deeply into the well-being of children in families where only domestic violence is immediately evident. If evidence is found that children are in danger, these authorities may then make a referral to CPS.

Increased communication within the court system can also improve the response to victims in violent households. Families in the justice system due to both child maltreatment and domestic violence may be involved in several court cases simultaneously. A child maltreatment case may be active in both juvenile (civil) court and criminal court, while any of a number of courts may be involved with domestic violence depending on the nature of the action (civil or criminal), the level of offense if criminal (misdemeanor or felony), and locale (district versus municipality). Divorce and custody issues between parents are often settled in another court (civil family court). Each community has its own unique and often complex court system. More often than not there is little or no systematic means of communication among courts, or between the courts and CPS. This can cause a great deal of confusion for families, and may have harmful repercussions when orders conflict, as when custody and visitation provisions are incompatible with the provisions of restraining or stay-away orders.⁽⁶⁾ Small communities and those with more unified court systems are at an advantage because one judge often presides over several courts. In these situations, one judge may hear all cases involving a single family or the judge can call up information about other court cases when deciding on the matter before the court. It is still up to the individual judge, however, to access information from other court cases to inform his/her decisions. Some judges have reported that they intentionally try to keep cases coming before them quite separate in their minds despite hearing all of them, while other judges take full advantage of access to information even when they personally have not heard all of the various related cases.

The sharing of information among cases involving members of a single family is an important potential advantage to unifying all civil and criminal family cases into a unified family court. Many nationally recognized domestic violence experts, however, are concerned about the effects of unified family courts (Battered Women's Justice Project 1994). One important objection is that by moving domestic violence cases out of the regular criminal courts, some people may view domestic violence offenses as being less important or less serious than other criminal offenses. This concern, which was also raised by battered women's advocates in Michigan, is quite understandable given that much of the progress made in addressing domestic violence has been achieved through the criminal justice system (by ensuring that laws treat domestic violence as a serious crime). Many advocates at the national level also want to ensure that battered women continue to have available to them the full range of criminal and civil court options. Some fear that one side (civil or criminal) may suffer if unified within a single court. Finally, advocates have also noted that family violence is only a small share of all court-based family matters. By unifying all cases within a family court, domestic violence and other forms of family violence issues might get inadequate attention. In the case of this last concern, a unified domestic violence court (Merryman n.d.) is less problematic than a unified family court. These same advocates argue that before establishing a unified family court, communities should identify the specific problems that need to be resolved and explore the full range of solutions to that problem. In the case of inter-court information sharing, they note that there are many ways of addressing this issue without unifying the courts involved.

Treatment Services

A critical community resource is the availability of affordable, accessible, and appropriate and effective treatment services for battered women, children, and perpetrators. Without adequate referral options, CPS workers may even be reluctant to screen their cases for domestic violence. Documenting the presence of domestic violence in the CPS caseload may help agencies justify the need for new or expanded treatment options.

Many batterer intervention programs are available in this country, but there is little evidence that they have a long-lasting and significant impact on the behavior of batterers (see Edleson 1995a; Gondolf n.d.; Tolmon and Edleson 1995). Many of these programs base their approach on the theory that battering is a manifestation of male power and control supported by societal attitudes and structural arrangements. Others are based on simple "anger management" or behavior control principles that treat battering in the same way they would treat fear of heights or smoking cessation. One promising program that has not been adequately evaluated bases its approach on psychological principles of attachment, affect, and individuation theories (Stosny 1995). Some states have minimum requirements for the number of sessions in approved programs. California's is among the longest at 52 weeks. In other states, judges order offenders into programs as short as a single Saturday afternoon.

Community Characteristics

Every community is unique and has a variety of strengths on which to build. Just as efforts to change case practice for families affected by domestic violence can begin from a variety of points within the CPS agency, different starting points can also be found in the community at large. If a community already coordinates its efforts and collaborates in a given area, child welfare and domestic violence groups can begin by working within such forums. Hawaii, for example, has a well-established network of Child Advocacy Centers--child-friendly centers jointly supported by the state judiciary and private community donations. These centers were originally designed to help with the investigation and prosecution of child sexual abuse cases. They are also a key point of coordination for families involved in multiple court cases. Center staff are now hoping to attract cases other than sexual abuse. CPS and other community agencies may want to use such centers when working with families affected by child maltreatment and domestic violence.

Other community characteristics will also affect how various agencies will be able to come together to help families affected by both child maltreatment and domestic violence. These include a community's size and its urban or rural status. Service delivery systems in rural areas, for example, have several distinct features.⁽⁷⁾

Rural systems are likely to be informal and personal rather than formal or bureaucratic, even when a service agency is involved. Often agency staff and their clients know each other or their families quite well.⁽⁸⁾ Service providers are also likely to know one another and to have a long history of mutual referrals. Often, agency staff know which agencies have resources at the moment and what types of resources they have. There are also important informal linkages to religious organizations and other unofficial sources of community assistance. There are many advantages to this quality of small-town service delivery. One can build on the many informal supports present in rural communities.

Small communities may have to rely on the willingness and ability of one critical person, such as a judge, to cooperate in this effort. Once this person is on board, however, change can occur more quickly in smaller communities than in larger ones. Other members of the community can lay the groundwork to educate and encourage all higher officials to understand the importance of particular issues and to take advantage of new opportunities as they arise. Other factors such as isolation, large geographical distances, a lack of resources, local attitudes and politics (including a denial of social problems), and a lack of knowledge of successful efforts in other communities mean that rural service providers may find it difficult to organize an efficient service delivery network for families affected by violence.

Complex Policy Questions

The site visits conducted for this study uncovered a number of complex policy issues which communities find themselves addressing either explicitly or indirectly. These policy questions, which reflect many philosophical issues include: (1) Should CPS actively seek out domestic violence cases potentially involving child abuse/neglect, in addition to screening for domestic violence in its existing caseload? (2) How should CPS handle cases in which a child witnesses domestic violence, but in which direct evidence of physical abuse or neglect is lacking? (3) Should CPS mandate battered mothers into counseling or other treatment services related to domestic violence, especially if the mandate involves forcing her to leave her home (and leaving the batterer in it)?

There is no one right answer to these questions. What is right and makes sense for one family or community may not for another. What is important, however, is to include major community stakeholders in the decision-making process and to begin by building on the community's strengths. Working with the larger community means that everyone knows what has been decided, and may ensure that a family's service needs beyond the scope of CPS are met by other organizations in the community. As expert members of a recent policy roundtable on many of these issues advised, communities should move ahead one step at a time, blend policy and practice, and recognize that CPS is a finite resource.⁽⁹⁾ They also noted that it is important to stay grounded in reality and to be practical.

Identifying Cases Affected by Both Child Maltreatment and Domestic Violence

The first issue that CPS and other agencies will want to consider is how to identify families and children who are affected by both child maltreatment and domestic violence. As a first step CPS should screen its own new and existing cases for domestic violence, and be prepared to offer needed services accordingly. These services may be offered directly by CPS or through referral to experienced providers in the community.

Once CPS has established safe and effective screening procedures and service systems for its existing cases involving domestic violence, it may want to work with other service providers or agencies in the community to ensure that those families with domestic violence who are also maltreating children are referred to CPS promptly. One way to do this is to train organizations and individuals who see victims of domestic violence to ask about the presence of children in the household, and to get a sense of their exposure to the domestic violence and their risk of being harmed. Child welfare and battered women's experts, working together, should develop screening and interviewing questions to be used in these efforts. People to be trained might include domestic violence service providers, law enforcement officers, emergency room personnel, and judges or court clerks who review applications for protection orders. Many of these individuals are mandatory reporters who must report incidents of child abuse or neglect once they become aware of it. In Hilo, for example, the family court judge asked about the presence and whereabouts of children during the court proceedings on protection orders and their violations. He only referred cases where he thought the harm to the child was enough to warrant CPS involvement regardless of the domestic violence, and he uncovered these situations by probing beyond the standard practice of most courts. In Ann Arbor, the domestic violence program has teams on call to respond immediately to homes where a report of domestic violence has been made to the police and children are present. While one advocate talks to the adult victim, the other advocate speaks with the children. If a referral to CPS seems necessary, the advocate encourages the mother to contact CPS herself or makes the referral directly.

Some local communities have established systems in which law enforcement agencies automatically route all reports of domestic violence to CPS when children are in the home. CPS then screens the cases as it would any other CPS referral, and acts if the threat of harm is detected. While this approach may succeed in identifying families in real need of CPS intervention, CPS should expect a major increase in referrals and should carefully consider whether it can handle the volume of screening and is able to provide all such families with safe, effective services. In mid-1995, for example, referrals to San Diego County's Children's Services Bureau more than doubled after the City Police Department began referring all domestic violence cases with children in the home. The state of Massachusetts implemented this policy in 1989 but revoked it in 1990, after experiencing a dramatic increase in the number of reports and observing that battered women were less likely to seek help if their children were automatically reported to CPS. Massachusetts adopted its 1989 policy before providing appropriate training or clinical support and concluded that it was ineffective and potentially dangerous.

Child welfare and domestic violence workers alike are concerned that if automatic referrals of domestic violence cases to CPS are handled inappropriately, battered women may be placed in further danger or be less likely to seek help. For example, one local office in San Diego County was unable to respond in person to all the domestic violence police reports it received, and instead sent an information packet to each woman's home. Other child welfare staff worried that these packets might be seen by the batterer, who might then blame the mother for seeking assistance. Another child welfare caseworker was concerned that women would be less likely to seek help from police if they learned that they would be reported to CPS. Domestic violence workers in a community in Oregon shared these fears. These examples illustrate that even the simplest actions may have unexpected consequences, and need to be carefully considered.

Many families experiencing domestic violence do not meet legal definitions of child maltreatment and do not belong in the CPS system. Often, these families would be best served by other local agencies or organizations in the community. In general, CPS should work with battered women's organizations and others:

To determine the criteria under which children in families with domestic violence should be referred to CPS for investigation,

To establish how risk of abuse and neglect should be assessed in an environment of domestic violence, and

To develop appropriate services and service delivery strategies for these families in a variety of community agencies as well as CPS.

One factor that should be considered when determining how to help families affected by domestic violence--but in which there is no direct (threat of) abuse or neglect of the children--is the availability of supportive services for children and their mothers and sanctions against batterers other than those provided through CPS.

Does Witnessing Equal Abuse?

A second and related issue concerns whether or not witnessing domestic violence in and of itself constitutes child abuse that warrants CPS intervention. As CPS agencies and communities generally become more aware of children who live in families experiencing domestic violence, there is a natural inclination to include them in CPS protection efforts. States and localities must use extreme caution, however, in expanding legal definitions of child maltreatment based on new levels of awareness or current trends. What is essential is that the risk of harm to children be assessed properly and safely. Children attach different meanings to their experience of "witnessing" abuse. That experience is shaped by many factors such as the child's age, relationships with the offender and the non-offending parent, the chronicity and severity of the abuse, and the child's "witnessing" of the community's response.

Several communities visited for this study include domestic violence as one risk factor for the emotional abuse of a child. Within a general framework of risk, the legal burden of proof required to link violence between the parents to the emotional well-being or behavior of the child varies considerably from place to place. In one community in Oregon, child welfare workers found it impossible to use allegations of emotional abuse as a basis for intervening with a family. In another community in the same state, domestic violence to the mother, without evidence of direct physical harm to the child, was accepted as a primary basis for the child welfare agency's gaining jurisdiction over the family (under the premise that witnessing domestic violence emotionally harms the child). In this second community, the agency equated witnessing domestic violence with child abuse. According to staff at one battered women's shelter in that community, they and other mandatory reporters in the community were required to refer all battered mothers to CPS for screening.

Many experts in the field agree that domestic violence should not be equated with child abuse or neglect.⁽¹⁰⁾

Such a policy does not take into account variations in the severity of abuse within domestic violence cases, or the extent to which parents succeed in protecting their children from harm. Domestic violence should be one factor in assessing a child's risk but decisions should be made based on each family's unique situation. Establishing the broader criteria for CPS referral and screening should be discussed and decided at the local level with input from the wider community and the domestic violence community, in particular.⁽¹¹⁾

Finally, many of these complex issues highlight the need for prevention services. If CPS must limit the types of cases it can serve, one must ask what services should be available to more "borderline" families and who should provide them. Should CPS provide voluntary services to these families or should other organizations and agencies in the community consider doing so? Michigan's program allows domestic violence shelters to refer families who are at risk of child abuse or neglect to Families First, expanding the pool of families eligible for intensive services to families who might otherwise lie outside the child protection system. Rather than expand the definition of child abuse in order to serve children affected by domestic violence, Michigan has chosen to provide prevention services to these families through this alternative mechanism. Because they lack alternatives, other states might find it necessary to use threat of harm or emotional abuse as a basis for bringing families into the child welfare system and providing them with services.

Mandated Treatment

A third issue concerns mandating battered women into treatment for domestic violence-related issues. Virtually all domestic violence services rely on a woman's voluntary request for services as the philosophical starting point for her further empowerment. Requiring a woman to receive services, as CPS might do as part of its plan to maintain or reunify the family, is antithetical to the philosophical underpinnings of these programs.⁽¹²⁾ Many domestic violence advocates feel that mandating a battered woman into treatment further victimizes her. One director of a domestic violence program stated that in the past she has refused to take CPS referrals because of this concern. Most individuals we spoke with prefer that battered mothers choose willingly to participate in domestic violence-related treatment such as support groups and counseling.

Many respondents acknowledged, however, that in some cases CPS must require battered women to receive these services to protect children. This opinion was shared by child welfare workers, domestic violence advocates, and even several CPS-involved battered women. Several people also observed that by including non-residential domestic violence-related treatment in their service plan, battered women can justify their attendance to their partners by explaining that it is required of them. Under more voluntary circumstances, the partner might not allow the mother to attend treatment sessions.

Other battered women's service providers who had mixed feelings about accepting CPS referrals agreed to do so on the condition that they not be required to monitor the mother's behavior for the child welfare caseworker. Therapists and other providers of services to battered women are often very protective of the privacy and confidentiality of their work with battered women. This professional ethic of confidentiality is fundamental to the therapeutic approach, since the success of such an approach requires that a battered woman trust the therapist enough to disclose sensitive information. Therapists who work with battered women and their children often learn of information that is of great use to CPS and its work with the family, information that can be used to gauge if the family is following its service plan, and to assess if new approaches or services are needed. Exactly what information therapists are willing to share with CPS, however, varies. Some therapists may be unwilling or legally unable to report to CPS any information about battered women, while others may be willing to disclose helpful information with their client's permission. Some may be willing to report to CPS whether a mother is attending therapy sessions regularly. Others may be willing to alert CPS if they are concerned about the safety of the children (in situations which fall short of the threshold for a mandatory report of child abuse or neglect). Finally, some are willing to share information about a battered mother's progress in very general terms.⁽¹³⁾ As CPS and domestic violence service providers begin to work together more effectively, they can establish agreements that help each of them promote the recovery, safety, and well-being of mothers and children. Otherwise, relationships between CPS and the therapeutic community can become quite strained. In one community visited, a domestic violence counselor who was not affiliated with the long-standing domestic violence program in the area, reported back to child welfare workers on the content of her sessions with battered mothers as a condition of her service contract with CPS. This same counselor, however, was hesitant to hear sensitive information from her clients.

A more contentious issue for domestic violence providers is when CPS orders a woman and her child(ren) to leave their home and enter a shelter or safe house. When feasible, it is the batterer who should be required by the courts to leave the home. And CPS should work with law enforcement to prevent him from returning. If, however, CPS cannot be assured of the children's safety in the home, shelter may be the safest option for a mother and her child(ren). CPS must also work with mothers in identifying their options and making the safest choices. These scenarios are complicated when a mother does not want to leave the perpetrator (or does not support his leaving their home) or feels unable to do so because of safety concerns. These situations can best be resolved on a case-by-case basis by a domestic violence advocate working with CPS staff.

A Final Issue--Evaluation

A final important issue facing CPS administrators as they begin to change policy and practice regarding domestic violence is how to document the impact of their efforts. Strong evaluation results can assist administrators in determining which specific changes have made a positive difference and are worth maintaining. In addition, decisionmakers will be interested in longer run cost savings (e.g., from more effective case management, lower recidivism, improved worker safety) resulting from new specialized staff or new contracted services, and whether the increased commitment of resources pays off in improved child well-being.

Several of the communities visited for this study are beginning to think about how they might evaluate the impact of the changes they have made. One researcher stressed the importance of thinking about evaluation early on and integrating it into the design of program changes. In particular, desired outcomes and the ability to measure them should be identified and developed at the outset. Many communities were hindered by information systems that did not automatically collect the relevant data. For example, computerized forms had no place to indicate the presence of domestic violence, and no questions on the screening protocol asked about domestic violence.

Policymakers will want to think carefully about the outcomes that most closely resemble their stated goals. Community involvement, particularly that of domestic violence advocates, may be desired in determining the goals of the changes and how to translate goals into specific indicators. Collaborators in Michigan determined that their primary goal is to empower women to make informed choices, regardless of what these choices may be. Outcome measures they are currently considering include: (1) a greater use of personal protection orders, (2) having safety plans in place, (3) lower levels of depression among women, (4) an increase in the social network of women, and (5) whether women feel safer and believe that they have more options. These measures help show whether a family is better off, regardless of whether the woman chooses to remain with or return to the batterer. Other indicators that may useful in making persuasive cases to budget-minded legislators will include changes in out-of-home placement of children (both frequency and length of stay), measures of children's well-being, recidivism of families, and the length of time their cases are open to CPS. Care must also be taken, however, to understand how better knowledge about domestic violence issues may appear to affect certain outcome measures negatively. These might include increases in caseload size and, given the added complexities and challenges of dealing with domestic violence issues, increases in the average length of time cases are open. These effects are not necessarily bad if the policy and practice changes also lead to improved safety and well-being among children and their battered mothers, and to fewer cases reappearing in the system at later dates. These more positive outcomes should also be measured and documented.

If a state agency is considering enhancing or changing its management information system, adding the ability to assess the prevalence of domestic violence in the caseload may provide the necessary preliminary evidence to persuade legislators or other decisionmakers of the seriousness of the issue, regardless of whether other policy or practice changes are being made. Note, however, that without training workers in appropriate techniques for screening cases for domestic violence, initial counts will likely be underestimated.

Conclusions

The fields of child welfare and domestic violence share one essential feature: they are both committed to ending violence in families. As the growing body of research documenting the overlap of child maltreatment and domestic violence within families makes clear, child welfare and domestic violence professionals must begin to cooperate more closely in their efforts to help families in need.

There are a number of ways in which CPS and domestic violence advocates can move forward together to serve families affected by multiple forms of violence. The complexity of this task is perhaps the main reason why existing services have not met needs. Requiring that CPS serve all families with abused and neglected children, regardless of whether the family wants CPS involved, raises different service delivery questions than those arising from families' a voluntary request for services, typical for domestic violence service providers. A woman's acknowledgment of the danger facing her and her child(ren) and her decision to leave or remain with the batterer do not alter the need for some mechanism to support and protect both the mother and her children. These factors do, however, affect how workers should proceed. Domestic violence can overlap with the abuse or neglect of children by either or both parents. Regardless of who is the offending parent, the larger context of spousal violence must be addressed. In particular, the degree of danger faced by children, mothers, and CPS workers creates an urgency for working with these families and a need to proceed with the utmost care, safety, and creativity.

Developing a system response able to deal with these complexities requires a number of elements. Because CPS alone cannot respond to domestic violence, substantial involvement of other community members is necessary to develop appropriate procedures. Existing providers of services to battered women have a great deal of knowledge that CPS agencies and workers need to incorporate into their practice. At the same time, CPS, domestic violence service providers, law enforcement officials, and others must work together to develop new services and service delivery strategies that stretch the existing boundaries of their professions. The responsibility to help families affected by both domestic violence and child abuse does not rest with either CPS or domestic violence advocates alone.

CPS agencies can also begin to take steps on their own. In addition to calling meetings to begin collaborative efforts, CPS can invite input on training curricula, hire domestic violence specialists to consult with case workers, and change the agency's philosophy to reflect the importance of the mother's safety as part of her ability to protect her child(ren). Our site visits provide many examples of ways in which CPS agencies have begun to enhance their ability to serve families in which the mother is being battered.

The site visits also revealed a number of other requirements that consistently appear to enhance CPS's ability to implement change. Staff awareness, understanding, and motivation coupled with tools and resources were all found to be necessary ingredients for sustained systems change. Without all of these ingredients, communities were much less likely to succeed. Providing workers with proper education and training, as well as ongoing technical assistance (e.g., in-house or contracted domestic violence specialists), created the understanding and motivation within individual caseworkers to improve their case practice. It also gave them the tools and procedures needed to realize these improvements. Ongoing technical assistance appeared to make a more lasting difference than initial training alone on workers' ability to put the content of training into practice on a daily basis.

The site visits revealed a number of interesting options for the location of initial change within CPS, but the need to reach out and collaborate with outside agencies was universal. As evidenced by the diversity of approaches undertaken by CPS agencies, individual child welfare workers, and domestic violence specialists, addressing domestic violence in CPS families often requires a great deal of commitment and innovation. Whether by thinking of new formal collaborations, finding ways to reach out to previously unserved families, or developing safety plans for mothers and their children, addressing domestic violence requires skill, creativity, and insight.

Finally, it is important to recognize that this process is neither easy nor swift. Any systematic change takes time, regardless of the nature of the issue. Historical tensions between the child welfare and battered women's communities, together with society's attitudes toward violence against women, further complicate the process. These complexities mean that members of each community must work together to design their own approach.

Several people interviewed for this study compared CPS's current response to domestic violence to that of the child sex abuse issue ten years ago. At one time, child sexual abuse was ignored due to its "controversial" nature. Now many communities provide support and compassion for the victims, and training on and services for child sexual abuse are an integral part of all CPS systems. Efforts to change the system's response to families affected by domestic violence will similarly take a great deal of time, effort, and resources, and must include a strong level of philosophical and financial commitment from CPS and other community partners.

References

American Humane Association, Highlights of Official Child Abuse and Neglect Reporting, 1987, Denver, CO, 1988.

American Psychological Association, Violence and the Family, Report of the American Psychological Association Presidential Task Force on Violence and the Family, Washington, D.C., 1996.

Battered Women's Justice Project, "Unified Courts: A Preliminary Discussion, June 1, 1994, Palmer House Chicago," Harrisburg, PA, 1994.

Bograd, M., "Why We Need Gender to Understand Human Violence," Journal of Interpersonal Violence, pp. 132-135, Volume 5, Number 1, March 1990.

Burt, Martha R., "Findings and Implications from RECD's Rural Homelessness Conferences," Report to Congress, Washington, D.C.: U.S. Department of Agriculture, Rural Economic and Community Development Administration, December 1995.

Carlson, B.E., "Children's Observations of Interparental Violence" pp. 147-167 in A.R. Roberts (ed.) Battered Women and their Families, Springer Publishing Company, New York, NY, 1984.

Center for Mental Health Services, "Taking Rural into Account: Report on the National Public Forum Co-Sponsored by the Center for Mental Health Services," CMHS/SAMHSA/PHS/DHHS, Rockville, MD, 1993.

Child Welfare Partnership, Cohort Two: Study of Families and Children Entering Foster Care 1991-1993, Portland State University, undated.

Clark, Sandra J., Martha R. Burt, Margaret M. Schulte, and Karen Maguire, "Coordinated Community Responses to Domestic Violence in Six Communities: Beyond the Justice System," The Urban Institute, October 1996.

Cummings, Nina and Andrea Mooney, "Child Protective Workers and Battered Women's Advocates: A Strategy for Family Violence Intervention, Response to the Victimization of Women and Children, Volume 11, Number 2, 1988.

DePanfilis, D. and G. Brooks, "Child Maltreatment and Woman Abuse, National Woman Abuse Prevention Project, Washington, D.C., 1989.

Edleson, Jeffrey L., "Do Batterers' Programs Work?" presented at the International Study Group on the Future of Intervention with Battered Women and their Families, Haifa, Israel, March 19-21, 1995a.

Edleson, Jeffrey L., "Mothers and Children: Understanding the Links Between Woman Battering and Child Abuse," presented at the Strategic Planning Workshop on Violence Against Women, National Institute of Justice, Washington, D.C. March 31, 1995b.

Felix, A.C. III and K.F. McCarthy, "An Analysis of Child Fatalities 1992," Massachusetts Department of Social Services, Boston, Massachusetts, 1994.

Ganley, Anne L. and Susan Schechter, Domestic Violence: A National Curriculum for Children's Protective Services, Family Violence Prevention Funds, San Francisco, CA, 1996.

Giles-Sims, S.J., "A Longitudinal Study of Battered Children of Battered Wives," Family Relations, pp. 205-210, Volume 34, Number 2, 1985.

Goldner, V., "Warning: Family Therapy Can Be Dangerous to Your Health," The Family Therapy Networker, pp. 19-23, Volume 9, Number 6, 1985.

Gondolf, Edward D., "Batterer Programs: What We Know and Need to Know," forthcoming in Journal of Interpersonal Violence.

Hangen, Eric, "DSS Interagency Team Pilot Project: Program Data Evaluation," Office of Management, Planning, and Analysis, Massachusetts Department of Social Services, Boston, Massachusetts, August 1994.

Herskowitz, J. and M. Seck, "Substance Abuse and Family Violence, Part 2: Identification of Drug and Alcohol Usage in Child Abuse Cases in Massachusetts," Massachusetts Department of Social Services, Boston, MA, January 1990.

Hess, Peg McCart, Gail Folaron, and Ann Buschmann Jefferson, "Effectiveness of Family Reunification Services: An Innovative Evaluative Model," Social Work, Volume 37, Number 4, July 1992.

Kolbo, Jerome R., Eleanor H. Blakely, and David Engleman, "Children Who Witness Domestic Violence: A Review of Empirical Literature," Journal of Interpersonal Violence, Vol. 11, No. 2, June 1996.

Magen, Randy H., Katherine Conroy, Peg McCart Hess, Ann Panciera, and Barbara Levy Simon, "Evaluation of a Protocol to Identify Battered Women During Investigations of Child Abuse and Neglect," paper presented at the 4th International Family Violence Research Conference, University of New Hampshire, Durham, New Hampshire, July 22, 1995.

McCloskey, Laura Ann, Aurelio José Figueredo, and Mary P. Koss, "The Effects of Systematic Family Violence on Children's Mental Health," Child Development, pp. 1239-1361, Volume 66, 1995.

McKay, Mary McKernan, "The Link between Domestic Violence and Child Abuse: Assessment and Treatment Considerations," Child Welfare, Volume 73, Number 1, 1994.

Merryman, Mithra, "Specialized Domestic Violence Courts: A New Means to Address an Age Old Problem," Battered Women's Justice Project, Harrisburg, PA, undated.

National Center on Women and Family Law, The Effects of Woman Abuse on Children: Psychological and Legal Authority, pp. 5-6, cited in "The Children of Domestic Violence," Massachusetts Coalition of Battered Women's Service Groups, Children's Working Group, forthcoming.

Pecora, Peter J., James K. Whittaker, Anthony N. Maluccio, with Richard P. Barth and Robert D. Plotnick, The Child Welfare Challenge: Policy, Practice, and Research, Hawthorne, NY: Walter de Gruyter, 1992.

Petchers, Marcia K., "Ethical Dilemmas in Services to Children in Battered Mothers' Households," paper presented at the 1995 NASW Annual Conference, Philadelphia, Pennsylvania, October 15, 1995.

Schechter, Susan, "Model Initiatives Linking Domestic Violence and Child Welfare," paper presented at a conference on Domestic Violence and Child Welfare: Integrating Policy and Practice for Families, Wingspread, Racine, Wisconsin, June 8-10, 1994.

Schechter, Susan and Jeffrey L. Edleson, "In the Best Interest of Women and Children: A Call for Collaboration Between Child Welfare and Domestic Violence Constituencies," paper presented at a conference on Domestic Violence and Child Welfare: Integrating Policy and Practice for Families, Wingspread, Racine, Wisconsin, June 8-10, 1994.

Schechter, Susan and Anne L. Ganley, Domestic Violence: A National Curriculum for Family Preservation Practitioners, Family Violence Prevention Funds, San Francisco, CA, 1995.

Stark, E. and A. Flitcraft, "Women and Children at Risk: A Feminist Perspective on Child Abuse," International Journal of Health Services, pp. 97-118, Volume 18, Number 1, 1988.

Stosny, Steven, Treating Attachment Abuse: A Compassionate Approach, Springer Publishing Company, New York, NY, 1995.

Straus, Murray A., "Children as Witnesses to Marital Violence: A Risk Factor for Life Long Problems Among a Nationally Representative Sample of American Men and Women," paper presented at the Ross Roundtable on Children and Violence, Washington, D.C., September 1991.

Straus, Murray A. and Richard J. Gelles, Physical Violence in American Families: Risk Factors and Adaptations to Violence in 8,145 Families, Transaction Publishers, New Brunswick, NJ, 1990.

Straus, Murray A., Richard J. Gelles, and S.K. Steinmetz, Behind Closed Doors: Violence in the American Family, Garden City, NY: Anchor Doubleday, 1980.

Tolman, Richard and Jeffrey L. Edleson, "Intervention for Men Who Batter: A Review of Research," pp. 262-273 in S.R. Stith and M.A. Straus, Understanding Partner Violence: Prevalence, Causes, Consequences, and Solutions, National Council on Family Relations, Minneapolis, MN, 1995.

Whitney, Pamela L. and Lonna Davis, "Child Abuse and Woman Abuse: Understanding the Connection and Designing Safe Interventions," Common Ground, Boston, Massachusetts, 1992.

Notes:

1. CPS workers in several communities noted that they work with clients' probation officers whenever possible. Instances of informal collaboration, however, were usually limited to individual workers and were not evident or feasible in all locations visited. The effectiveness and longevity of such relationships can be increased with the help of institutional support.

2. HomeBuilders is a model program of intensive family preservation services established in Tacoma, Washington in 1974.

3. Note that this is not unique to CPS-referred battered women. Many battered women choose to stay (or return) to their abusive partners. Some ultimately leave these partners and others do not. Unlike their non-CPS counterparts, however, many CPS-referred battered women have never independently sought help from a domestic violence program.

4. These differences do not apply to all communities, and will depend on many factors including the characteristics of CPS-involved and other battered women.

5. A major issue affecting interagency information sharing is the problem of confidentiality of client information. In the case of child welfare agencies and domestic violence service providers, some of these confidentiality constraints are legally mandated, while others are grounded in professional ethics and the need for trust as a basis for successful provider-client relationships. A full treatment of this issue is beyond the scope of this study.

6. In San Diego County, it is not uncommon for cases transferred to the Family Violence Project to have criminal and juvenile court orders that conflict with one another. In these cases the CPS social worker-probation officer team jointly reviews the orders, consults with their supervisors and, guided by a standard of what is best and safest for the children and family, goes back to one of the courts to have the order modified as needed.

7. Many of these findings have also been documented in studies of services for homeless people (Burt 1995) and people needing mental health services (Center for Mental Health Services 1993) in rural areas.

8. There can be disadvantages to this personal familiarity. Women may be less likely to seek help from someone they or their partners know. In addition, service providers may be less likely to act if they know the people involved (e.g., the police may fail to arrest someone they know).

9. "Domestic Violence-Children's Protective Services Roundtable, Ann Arbor, Michigan, June 21, 1996," draft summary prepared by Susan Kelly and Janet E. Findlater, Michigan Family Independence Agency, 1996.

10. Draft summary of a Domestic Violence-Children's Protective Services Roundtable, Ann Arbor, Michigan, June 21, 1996.

11. Just as domestic violence service providers can screen for risk of child maltreatment, incoming referrals to CPS may be one way of identifying families experiencing domestic violence early, especially when other efforts at help seeking have been ineffective or limited. After initial assessments by CPS and domestic violence service providers, many of these families can and should be helped by voluntary community-based services.

12. Not all child welfare service are mandatory. Participation in family preservation services, for example, is usually voluntary.

13. Note that like many other professionals, therapists who learn of child abuse or maltreatment are required by law to report this information to CPS.