Chapter I
Background and Overview
Rationale for the Study
This report focuses on recent efforts by child welfare agencies to take account of battering experienced by mothers in cases of child abuse and neglect. These efforts are in their infancy. The oldest of them, in Massachusetts, dates back only about six years, and others are still more recent. A few model programs have been developed (Schechter 1994), and several states have already embarked on training activities for child welfare workers. In many cases these training efforts have been supported by federal grants from the Office of Community Services' Family Violence Prevention and Services Program in the Department of Health and Human Services. Because the field is so young, far more questions than answers exist regarding what to do, where to start, how to proceed, and even what the goals of the effort should be.
A common goal of many studies is to describe "best practices" so that others interested in pursuing similar activities may profit from these experiences. The same motive prompted the Office of the Assistant Secretary for Planning and Evaluation (ASPE) and its partner agencies, the Office of Community Services and the National Institute of Justice, to undertake this study, though they recognize that definitive findings must wait until the field is more developed. The issues are of great importance to children's well-being and to child welfare agencies' ability to protect it. Many child welfare agencies are beginning to understand this importance and are looking for guidance before undertaking significant changes in their policies and case practice.(1) This report helps provide some of this guidance.
The study's goal is to document what child welfare agencies in five state and local communities are doing in attempting to integrate domestic violence(2) concerns into their agency routines, and to identify and highlight issues that a child welfare agency would do well to consider before proceeding with similar efforts. This report should be read with the understanding that it presents few answers or "best practices." It focuses instead on raising issues that, on the basis of fieldwork, seem critical for child welfare agencies to consider as they make changes to address the problem of domestic violence among families in their caseloads.
The rest of this chapter examines (1) the evidence for the co-occurrence of woman-battering and child abuse and neglect, and (2) the areas of potential involvement for child welfare agencies that guided the fieldwork of this study. It also describes our site selection criteria and procedures. Subsequent chapters (II through VI) describe the different approaches agencies have taken in the five sites visited and the community context in which these efforts operate. The responsibility for addressing family violence is shared by many and how child welfare agencies address domestic violence is strongly affected by the availability and quality of other services and supports in the community. The final chapter of this report summarizes our findings and presents implications for child welfare agencies.
Background
Until recently, programs and policies for family violence or abuse have responded to its two forms--child maltreatment and woman battering--through two entirely different service systems, child protective services and domestic violence programs. This separation is due, in part, to differences in when these service systems were established and how they developed over time. The child welfare system is by far the older, dating back to early in this century. Child welfare agencies have tended to view the mother's role in child abuse perpetrated by a male partner as "failure to protect" the child, rather than acknowledging that the child's safety might depend on addressing a situation that endangers both mother and child. Emergency shelters and other services for battered women first emerged in the mid- and late 1970s. Their focus has been on helping battered women. Services directed specifically toward the children who accompany their mothers into these shelters are very recent and remain limited in many communities. Relations between the two systems have at times been strained, since a primary focus on helping the mother and a primary focus on protecting the child have not always been seen as compatible. Working out mutually supportive relationships between the systems is an important part of the effort to change child welfare practice around issues of domestic violence.
Evidence of Overlap Between Child Maltreatment and Domestic Violence
Before a public agency undertakes any major change in orientation or practice, it needs evidence that such a change is warranted. The degree of overlap between domestic violence and child maltreatment can be documented by population surveys or case record reviews. Early efforts to examine the prevalence of child maltreatment and domestic violence within the same family, summarized by Magen and colleagues (1995), confirm that the overlap can be substantial, whether approached from the point of view of the child or that of the mother. Between 11 and 45 percent of children who are abused or neglected have a mother who is being abused, and between 37 and 63 percent of abused women have children who are being abused or neglected. The studies on which these ranges are based differ in many ways (by, for example, employing different methodologies and definitions of abuse and looking at different populations of interest), but consistently report a high level of overlap. The data include national population surveys examining family violence, reports by battered women, state child fatality reviews, and studies of child protection cases.
Reviews of case records completed before an agency made any explicit effort to build screening for domestic violence into case practice, reveal that domestic violence is clearly present in a large share of families in need of child protection services. A 1993 annual review of child fatalities in Oregon found evidence of mothers being battered in 41 percent of the case records (Oregon Department of Human Resources 1993, in Schechter and Edleson 1994), even before the state began to focus on the issue of overlap and make a point of recording such information. A similar review of 67 child fatalities in Massachusetts found that the mother had disclosed being abused herself in 43 percent of cases (Felix and McCarthy 1994, cited in Edleson 1995b). This high level of overlap is not limited to child fatality cases. In a review of 200 cases of substantiated child abuse, the Massachusetts Department of Social Services found some mention of adult domestic violence in 30 percent of the case records (Herskowitz and Seck 1990). As in Oregon, this figure is likely to be a substantial underestimate of the true incidence of domestic violence because the review was done before child protection workers were required to ask about it as part of their investigation.
Existing studies of the overlap between domestic violence and child maltreatment are mostly limited to child abuse rather than neglect. This means that we do not have adequate information about the relationship of domestic violence to child neglect. This gap in our knowledge may have implications for child welfare agencies, since neglect is far more prevalent in their caseloads than is abuse.
The dynamics of violence within families can be very complex. Drawing on a national survey of more than 6,000 American families, Straus and Gelles (1990) found that half of men who frequently abuse their wives also frequently abuse their children, and the more severe and frequent the violence against the woman, the more likely it is that the children are also being abused. Children need not be the primary target of a father's violence in order to get hurt. Blows directed at the mother may land on children, and children may be harmed when trying to intervene or defend a parent. Similarly, a mother may endure blows from her partner in an effort to deflect his attention from the children (McKay 1994).
Knowing which partner in a domestic violence situation is the perpetrator of abuse against the children can help guide child welfare agency decision-making. According to the American Humane Association, close to half (44.1 percent) of reported child abusers are men (1988), despite the fact that men spend far less time with children than women do. Reviews of other studies suggest that children are more likely to be physically abused by their fathers than by their mothers and that the most severe forms of child abuse are committed by men (McKay 1994; Pecora et al. 1992). On the other hand, the Giles-Sims' (1985) study of battered women found that over half (56 percent) of the women sampled recounted having been violent toward their child(ren). These women also reported that 63 percent of their battering partners had been abusive toward their children, and that the abuse of children by these men was six times more frequent than that by the women themselves. Battered women are much more likely than their violence-free counterparts to abuse their children. In their national study, Straus and Gelles (1990) found the rate of child abuse by mothers who were beaten to be "at least double that of mothers whose husbands did not assault them" (p. 409).
These findings confirm those of earlier studies documenting the relationship of domestic violence to child abuse, but fail to capture many of the complex interconnections between various forms of family violence. As McKay (1994) explains:
. . . battered women are not a homogeneous group. They range from women of great strength and coping capacity to women who experience a wide spectrum of mental health difficulties. Given the stress associated with a violent household, therefore, abusive behaviors by mothers are likely to emerge from those at risk of such reactions. In the context of domestic violence, the battered women may view themselves as being more in control of their anger and the level of punishment of their children than their partner. In reality, their frustration with their situation combines with the stresses of parenthood to set the stage for physical abuse of children. Finally, some battered women are so fearful of their partner's response to the children that they overdiscipline them in an effort to control their behavior and protect them from what they perceive as even greater abuse. (p. 30)
Clearly, no single pattern can account for the co-occurrence of domestic violence and child abuse (and neglect) within a family. The overlap takes many different forms and involves different family members as offender and/or victim.
Witnessing Domestic Violence
Simply witnessing violence may be detrimental to children, but "witnessing" abuse has many dimensions. Children who do not see their mother being abused may hear screams, crying, degrading language, or objects being thrown and broken. Children may also witness the aftermath of an abusive incident, including blood, bruises, torn clothes, broken glass, a police officer's presence, or an arrest. Finally, most children sense tension in the home and their mother's fear or apprehension when the abuser enters a room (National Center on Women and Family Law, n.d.). While studies of the effects of witnessing violence are very limited and the exact effects examined vary considerably (McCloskey, Figueredo, and Koss 1995), effects on children may include a host of behavioral, emotional, and cognitive problems. Children who are both abused and witness their mother's abuse generally have the greatest variety and most intense symptoms. Finally, there is some evidence to suggest that children raised in violent homes are more likely to become perpetrators and victims of violence as adults, although more research is needed in this area. For reviews of the limited number of studies of the effects on children of witnessing violence, see Kolbo, Blakely, and Engleman (1996), Petchers (1995), and Edleson (1995b).
Although the number of children who witness domestic violence is not known, some researchers have tried to estimate this figure by using national surveys of the number of women who are beaten (or severely beaten) in a given year (Straus, Gelles, and Steinmetz 1980; Straus and Gelles 1990) and the expected number of children in these women's homes. Depending on how domestic or family violence is defined, this approach suggests that between 3.3 million (Carlson 1984) and 10 million (Straus 1991) children witness domestic violence each year. Even if only half of these children are themselves abused, as some studies suggest, the numbers would be high.
Child Welfare Services
This report focuses on child welfare services offered to children and their families to protect children from abuse and neglect. Services are provided to strengthen families, enable children to remain safely in their homes, remove children temporarily from parental custody if there is imminent risk to them, or pursue termination of parental rights if the custodial family cannot be preserved without serious risk to the child(ren). Many states and child welfare professionals use the term "child protective services" to include only the intake, investigation, initial assessment, and referral functions of the child welfare system, whereas others use the term more broadly and include in-home protective services, foster care, and adoption. Unless specified otherwise, we use the term "child protective services" to include the wider array of services.(3)
As a way to develop an appropriate set of questions to explore during site visits, we reviewed the available literature on domestic violence issues in child welfare practice and spoke with experts on the overlap of child abuse and domestic violence. We also conducted a series of informal interviews with child welfare administrators around the country to further develop our focus and to identify potential sites to visit. As would be expected with a very new field, most of the available literature is more normative than empirical, discussing issues and problems rather than presenting implementation experiences, case practice, or effectiveness data. In the following section, we provide a summary of the issues that the literature and our initial phone conversations prompted us to examine during our site visits.
Mission and Goals
The child protection system exists both to ensure the safety of children and to promote improved family functioning (Pecora et al. 1992). The overall mission is to protect children from harm, but keeping children with (or returning them to) their families is preferable to placing and maintaining children in foster care if their safety can be reasonably assured. Addressing domestic violence in a family may make the difference in achieving this end. For example, in an evaluation of 62 families in Indiana known to have a child re-entering foster care, Hess, Folaron, and Jefferson (1992) found "violence between adults" to be a contributing problem to disruption of the family's reunification in 35 (or 56 percent) of the cases. In many of these cases, child protective services failed in its initial case planning to identify the violence between adults in the family or to provide adequate resources to address the problem, and thus may have made the wrong decision in allowing children to return home.
Many liken the current situation of CPS agencies and domestic violence to where the field was ten years ago on the subject of child sexual abuse. Acceptance is spreading that child protection agencies would serve children better if they were able to handle situations involving domestic violence, but many professionals are still searching for the best ways to do so. Because domestic violence is a complex issue in its own right, and most communities have people and agencies with experience in helping its victims, child welfare agencies can benefit from working with these services. Often a community's domestic violence specialists and child protection specialists have operated in mutual ignorance, and sometimes in mutual hostility stemming from a lack of understanding of the roles and priorities of each. Both types of agencies, along with other community stakeholders, can gain from sharing their respective knowledge bases and coordinating their agencies and service delivery systems to protect children and serve families more effectively.
Domestic violence services usually take the mother's well-being and safety as their primary concern, while child protection agencies are legally obligated to focus on the child's safety and well-being. These goals need not be incompatible. Keeping the mother and child together and safe from the batterer often may be a better solution for both than removing the child. Removing children from their home traumatizes most children and families. If the child's mother is also being battered, such a move can further isolate and endanger her (Stark and Flitcraft 1988, in McKay 1994). If the child's father or mother's partner is abusive, it should be possible to remove him from the home. Holding him accountable for his actions rather than charging the mother with failure to protect is also more likely to change his behavior and less likely to victimize her further. Most important, the mother and child(ren) may be more likely to remain safe over the long term (DePanfilis and Brooks 1989). Protecting the mother from further abuse may also help her to become a more effective parent (Bograd 1990, in McKay 1994). Identifying domestic violence as soon as possible within a family that comes to the attention of CPS will help workers plan appropriate interventions.
Child protective workers may be the first outside service providers with access to a family experiencing domestic violence (Cummings and Mooney 1988). Batterers often isolate their immediate families from other family members, friends, and supports. Because CPS is often involved with families in which the woman does not otherwise seek services on her own, a child protection investigation can provide the mother with her first opportunity to confide about the battering and learn about the domestic violence services and shelter(s) in the community (Cummings and Mooney 1988). CPS workers cannot assume that the domestic violence will be addressed or identified through avenues other than CPS intervention, and can build assistance to the mother into plans designed to assure the child's safety.
Investigations
Given its dynamics of secrecy, fear, and control, domestic violence often
goes undetected by traditional CPS assessment methods (McKay 1994; DePanfilis
and Brooks 1989; Whitney and Davis 1992). Massachusetts, for example,
discovered that 70 percent of cases referred to a high-risk assessment unit
involved domestic violence but that intake and investigations workers had
identified this problem in less than half of these referrals (Schechter and
Edleson 1994).
Many standard methods used in child protection investigations may be inappropriate for families experiencing domestic violence. If a protective services investigator interviews a child's parents together, for example, the mother is not likely to disclose that she is being battered. To begin to identify the presence and effects of domestic violence in their caseloads, child welfare agencies may need to change case practice in a number of ways. New screening tools, including specific questions to ask, effective ways to ask them, and easy access to civil and criminal history records and police reports, are necessary to do a better job of identifying domestic violence in families already active in the CPS system.
It is also important to develop methods to investigate allegations of child abuse that do not further victimize or endanger a mother who is herself in an abusive situation. Threatening to remove a child or charging a mother with failure to protect does not empower the mother, nor does it help her change her situation or place the responsibility for the abuse on the perpetrator.
Finally, workers need knowledge regarding domestic violence in order to assess risks to a child more accurately. Decisions regarding whether to remove (or return) a child should be made with as much information as possible, including knowledge regarding the potential for further violence in the home. Proper training of workers, on-site domestic violence advocates, and knowledge of community resources enable workers to be more confident that when domestic violence is present in a family, appropriate plans are in place to protect the child.
Services
Once an investigation has determined that risk to the child exists, that domestic violence is present, and that the case warrants intervention, case managers' awareness of domestic violence dynamics is critical if they are to recommend appropriate services and interact with family members in safe and effective ways. Some family interventions may be inappropriate when CPS must work with both adults in the abusive relationship. Use of family or marital therapy involving both parents simultaneously is generally considered inappropriate in cases of domestic violence because it does not take into account the power differential between the victim and the abuser (Bograd 1984, Erickson 1988, and Goldner 1985 in McKay 1994). Some child welfare agencies routinely use joint meetings with all family members to discuss how the family can reach agreed-upon goals for child safety. These meetings may also include members of the family's support network. This approach may be appropriate for some families. When there is domestic violence, however, the family's existing support network may be more supportive of the abuser because the partner often isolates the woman from friends and family, especially those who question his behavior. In these cases, the mother will not have a fair chance of sharing her perspective, and the safety of the mother and child may be jeopardized.
To develop appropriate services for families affected by domestic violence, CPS will need to work closely with other professionals in the domestic violence community. These include local battered women's services, batterer intervention programs, and other agencies involved with victims or perpetrators. By developing relationships with batterer intervention programs or adult probation, for example, case workers can better assess perpetrators' compliance with CPS case plans.
Finally, it is essential that foster care and adoptive homes be adequately screened for domestic violence. This is especially critical when CPS is considering placing a child with the offending parent's relatives, who might allow the offender unauthorized access to the child or work against the mother's efforts to protect the child. Child welfare agencies and workers may want to develop new interventions or alter existing ones when working with families for whom domestic violence has been identified. Or, agencies might change case practice to be safe for all families under the assumption that domestic violence exists in many families and is often undetected.
Worker Safety and Burnout
Promoting the safety of workers and reducing staff stress and turnover is an important concern for child welfare agencies. Cases involving domestic violence can severely strain workers' personal resources. In the extreme, these cases can place workers directly at risk of violence from the abuser when, for example, they deny the abuser contact with a child or a partner as part of a safety plan. With proper training, child welfare workers are more likely to understand the dynamics of domestic violence and refrain from actions that place mothers, children, and themselves at risk. Trained workers may feel more comfortable working with each member of an abusive family, even in the family's home. Family preservation workers generally spend as much as 10 to 20 hours in a family's home and are on call 24 hours a day. These workers may find themselves in potentially dangerous situations or may be called by an abused mother at a time of crisis. Workers who see family members primarily in an office setting may also receive threats or have direct personal confrontations with violent partners. In addition, partners may try to manipulate the worker, deny abuse, and paint the mother as mentally ill or out of control. For their own safety and that of mothers and children, workers need to know how to recognize and respond to both manipulation and to a volatile, or potentially volatile, situation.
Complexities
Any recommendations for changing case practice must take into account the reality and complexity of the child welfare caseload. Families affected by domestic violence and child maltreatment are not a homogeneous group. Many factors other than the violence itself may also impact a child's safety. Severe substance abuse by either or both adults is prevalent and complicates CPS's ability to work with families. The mother's batterer may or may not be the father of the child and may or may not be living in the home; and the father of the child may or may not be interested in remaining involved in the child's life. The father may attempt to use visitation or custodial decisions to intimidate and control the mother. In order to ensure a child's safety at home, CPS must work with anyone likely to be involved in that household. CPS could, for instance, use the power of the courts to require the batterer to attend appropriate treatment services, and could work with other agencies in the community to hold him accountable for his actions. If these elements are incorporated into CPS's safety plan, further violence within the family may be reduced.
In some families, battered mothers may also be abusive or neglectful of their children. Sometimes this behavior may stem, at least in part, from the abuse perpetrated against them. Even when a mother's abusive behavior is the primary basis for CPS's involvement, workers should screen for domestic violence. If domestic violence exists between the mother and her partner, it may also pose risks for the children. "It is impossible to assess a mother's true capacity to care for her children while she is being battered or experiencing posttraumatic stress from the abuse" (Bograd 1990, cited in McKay 1994). Once an accurate assessment of the family is made, caseworkers can proceed to work more effectively with the family.
Domestic violence victims involved with the child welfare system may differ from women traditionally served by domestic violence shelters. CPS is usually involved with women who have not sought services on their own. While many CPS families have "voluntary" service plans, CPS is traditionally an uninvited presence in the lives of most families it serves, and the possibility of the children's removal is always present. While experience indicates that the abuse of children is frequently what precipitates a request for shelter or a protection order, not all women are able or willing to leave their batterer even when their children are endangered (McKay 1994). Shelters may restrict access for families with active substance abuse issues, serious mental illness, large numbers of children, or older boys; these restrictions limit the ability of some women to leave their batterer. A woman may be financially dependent on a partner because he has not allowed her to work or he controls the household finances. Concerns regarding how she will support her children are well-grounded (DePanfilis and Brooks 1989). Other women remain committed to their relationship with the batterer and may believe that he will change his behavior. Finally, for many battered women, leaving a partner poses additional risks--threats of serious or lethal violence often increase after separation (American Psychological Association 1996). A woman may feel that she is being more protective of herself or her children by staying, and that she is in the best position to determine the risks that leaving poses given her particular situation (DePanfilis and Brooks 1989). Recommendations for changing case practice need to allow workers space for flexibility and creativity to work with each family safely and effectively.
Working with Other Agencies
It is also important that child welfare agencies work with other community agencies in addressing domestic violence issues. Child welfare professionals do not work alone with families but interact with a number of other community players such as law enforcement, juvenile court, other community service providers, and sometimes the criminal justice system (for sexual abuse and severe physical abuse cases). When families also have domestic violence issues, different organizations or agencies may already be working with the family and can offer access to additional services. Effectively serving such families requires that all agencies involved think about this overlap and reexamine their treatment approach and service delivery in light of it. In addition to child protection agencies and domestic violence programs, many groups may be involved in a community's coordinated response to domestic violence (Clark et al. 1996). These include family preservation service systems and private child welfare programs. They also include religious and civic organizations and agencies or organizations working in social services, family services, medicine, mental health, education, prosecution, law enforcement, the judiciary, probation, substance abuse treatment, and employment and training.
It will be helpful for child welfare agencies to work with this larger community for several reasons. When a batterer is removed from a home, it is not CPS's responsibility to keep him away from the premises. Rather, law enforcement and the criminal justice system must respond to his criminal behavior. Child welfare workers should be aware of community resources and legal avenues for the battered woman and assist her in accessing them (McKay 1994). Child welfare agencies are unlikely to have the in-house resources or experience to address domestic violence issues by themselves. In working through an individual family's situation and in making broad agency policy, many opportunities exist for CPS to consult with battered women's advocates, batterer intervention programs, and domestic violence experts, as well as other community agencies with specialized expertise.
Shelters and other domestic violence programs may also be interested in working with CPS to offer more effective help to families in which both mother and children are victims of violence. Currently, domestic violence programs reach only a small proportion of women in need. By collaborating with CPS agencies, battered women's service providers may reach an important and sizeable pool of women who would otherwise remain unserved. The importance of reaching these women is great, both because they have children at risk and because they may be more isolated (and in greater danger) than many of their non-CPS counterparts who already reach domestic violence programs on their own.
Selecting the Five Communities for Study
We identified potential sites for this study through contacts with federal representatives in the offices of the Administration on Children and Families (ACF) and the National Center on Child Abuse and Neglect (NCCAN), national advocacy organizations, state and local officials known for their work in this area, and through calls to individual states and communities. The major criterion for selection was that the primary impetus for innovation and linkage come from within the child welfare agency and be directed to its own case practice. This eliminated as potential sites any efforts that came primarily from within domestic violence programs or hospitals to address the overlap between domestic violence and child abuse and neglect among their clients. Massachusetts and Michigan were included because of their long history in this area. The Massachusetts Department of Social Services began meeting regularly with battered women's organizations in 1987 and hired its first in-house domestic violence advocate in 1990. Michigan incorporated a domestic violence component into its family preservation program, Families First, in 1993.
Three other sites were selected to provide variation on the following characteristics: state versus local involvement in fostering linkages between child welfare and domestic violence agencies; the availability of additional funds and/or staff; the strength of the court system connection; and the presence of rural or other distinctive populations. In addition, the potential site had to be willing and able to host a site visit within the study's timeframe.
The three additional communities were San Diego County (California), Hilo (Hawaii), and the state of Oregon. In San Diego County, the county child welfare agency--the Children's Services Bureau--and Adult Probation together established a separate administrative unit (the Family Violence Project) to handle all cases active in both departments. Cases in the unit include some of the county's most violent families, who are managed by a two-person social worker-probation officer team. In Hilo, the East Hawaii CPS intake and investigative unit (the East Intake/Crisis/Investigative Unit) is committed to addressing domestic violence within its caseload and has established close relationships with the judiciary. The community has a very proactive judge who hears all temporary restraining order, divorce, juvenile justice, and child protection cases. (The state of Hawaii is distinctive in that all cases affecting a family can be combined within a single family court.) East Hawaii has a semi-unified family court; all family cases other than criminal cases are heard in a single court. The judge in this court actively screens restraining order petitions for child abuse/neglect and refers appropriate cases to CPS. Finally, Oregon's State Office for Services to Children and Families (SCF) is attempting to change case practice throughout the state by cross-training child protection workers and domestic violence workers about the relationship between the two forms of abuse. Oregon also recently ran pilot programs that placed domestic violence advocates in two local SCF offices, continues to encourage local-level dialogues on these issues throughout the state, and recently received continuation funding of its federal grant to expand its pilot efforts.
Individual descriptions of each of these five sites are presented in Chapters II through VI. Chapter VII reviews many of the crosscutting issues that arose out of these site visits, and presents the study's conclusions and policy implications.
NOTES:
1. The Family Violence Prevention Fund in San Francisco, California has published two child welfare and domestic violence curricula, both of which have been field tested in child welfare agencies: (1) Domestic Violence: A National Curriculum for Children's Protective Services by Anne Ganley and Susan Schechter and (2) Domestic Violence: A National Curriculum for Family Preservation Practitioners by Susan Schechter and Anne L. Ganley.
2. Throughout this report, the terms "domestic violence" and "spousal abuse" refer to a range of physical, sexual, and emotional maltreatment of an adult woman by her current or former partner, whether married or not. The term "family violence" is used to refer more generally to violence or abuse between any two members of the same family or household. While we also recognize that in some cases women batter male partners and that domestic violence is not confined to heterosexual relationships, the vast majority of batterers are men. Thus, throughout this report we assume the batterer is a man and the adult victim a woman.
3. In some of the communities we visited for this study, "child protective services" refers to a specific division within the child welfare agency that administers the more narrow set of services described above.