Facing Up to Family Violence
For an alarming number of Americans, the family is a source of fear and physical violence. The 1996 National Incidence Study of Child Abuse and Neglect found 2.8 million reported cases of child maltreatment, a rate of 41.9 per 1,000 children. The rate of domestic violence, according to the most recent National Crime Victims Survey, was 9.3 cases per 1,000 adults. These cases involve both physical and psychological injuries that can extend long beyond the violent events themselves; family violence also contributes to the development of other social problems such as alcoholism, drug abuse, delinquency, crime, teenage pregnancy, and homelessness.
Society's traditional unwillingness to intervene in family matters has given way in the past few decades to a host of efforts to support and protect victims and to deter and rehabilitate offenders. We have witnessed the birth of shelters for battered women, special police units focused on domestic violence, victim advocates in the court system, guidelines for health care providers who see evidence of family violence, and a wide range of other services.
Many of these programs deal with the acute injuries and crisis nature of family violence, when the victim's needs are clearly visible. More recently, however, service providers and policymakers have been exploring how to develop prevention programs or treatment approaches that can address the long-term consequences of family violence. This shift from acute response to problem prevention and treatment is difficult, especially because the state of research knowledge about family violence and the effectiveness of interventions is not well developed, and service providers seek to balance multiple goals in meeting the needs of children and adults.
As society comes to understand the complex nature of family violence, it is becoming increasingly clear that simply identifying problems and responding to crises will not solve the problem. Many aspects of family violence need more attention. Parenting practices have proven to be highly resistant to change. When alerted to reports of child abuse or neglect, for example, social service and law enforcement agencies must struggle with how to respond to the safety and developmental needs of the child, maintain and support families challenged by instability and stress, and keep their services efficient and inexpensive. In addressing domestic violence, law enforcement agencies and others are looking closely at the comparative benefits of treatment and deterrence, and are often uncertain about how to balance victim safety, fairness for the accused, and community support. Elder abuse has also raised difficult questions about the benefits of separating family members when abuse or neglect has been reported, especially when the abused person is dependent on the relationship for important benefits.
But although serious gaps in our understanding of the origins and causes of family violence still exist, opportunities now exist to improve the nature and effectiveness of interventions in health care, social service, and law enforcement settings. As program sponsors, policy officials, and service providers move forward in building victim support and offender deterrence efforts, a series of challenges needs to be overcome to create an effective service delivery system and an adequate knowledge base that can guide policy and practice.
Problem discovery and crisis response
The origin of many of the current shortcomings in research can be traced to the way in which the problems of child maltreatment, domestic violence, and elder abuse emerged as a source of social concern. Traditionally defined as a social and legal problem, the issue of child abuse emerged as a topic of major medical concern in the early 1960s, when health professionals became aware that they could detect signs of chronic and traumatic abuse in young children in the form of healed injuries that were no longer visible. With this knowledge, physicians and other child advocates encouraged the federal government to develop a national child abuse reporting system so that health care workers, teachers, and others would be required to report suspected cases of child maltreatment to local social service agencies. Mandatory reporting became the first component of child protection policy.
Defenders of battered women brought the needs and concerns of their clients to the attention of legal officials throughout the states in a different way during the past few decades. Recognizing that violence among intimates was often trivialized or treated as a civil disorder rather than a criminal action, police officials, prosecutors, and judges were urged to reform their response efforts to offer more forceful protection to victims involved in domestic violence cases. With federal resources, victim advocates were supported within law enforcement systems to counsel victims, predominantly women, who often were not aware of their options in bringing charges against a batterer or protecting themselves and their children from assaults and unwanted intrusions. Protective interventions thus became a distinctive component of law enforcement interventions for domestic violence.
Elder abuse, which has never received the level of national attention given to child maltreatment and domestic violence, has nonetheless been a topic of recurring concern among social service and health care agencies. Local agencies responsible for services for the aging have long been troubled by the cases of elder neglect that come to their attention. Service providers consistently search for solutions that can meet the needs of the vulnerable adult, preserve the older person's autonomy and decisionmaking authority, and retain family support systems where appropriate.
The caseloads associated with these early response efforts quickly became overwhelming. Child protective service and other social workers were assigned dozens of reports of families in trouble. Courts became backlogged as they tried to sort out the intricacies of evidence and the appropriateness of sanctions. Intimate partners or abused elders would cry for help from police one day, only to drop charges the next. Families reported for abuse would be monitored for months without incident, then they would move or disappear within the community, only to surface again as new allegations emerged. Law enforcement officials and social service personnel became frustrated and disillusioned by the lack of success of their efforts as they witnessed patterns of abuse and violence that were seemingly resistant to social controls and counseling.
Although inadequate, these initial interventions into family violence are tremendously important. They represent the first stage of a long-term effort to address an important social problem. They have revealed the scope and multiple dimensions of family violence. They have suggested that numerous pathways lead to violence within the home and that violence that occurs in childhood and among adults may have important linkages. They have indicated that relationships among family members and contextual factors are often important forces in triggering as well as preventing child maltreatment, domestic violence, and elder abuse. And they have demonstrated that the recurrent nature of family violence makes it a problem that will not be "cured" with a single dose of services.
Understanding the problem
Before moving on to develop new approaches and improved practices, it is important to take stock of what we have learned from the experience with family violence interventions. But taking stock is not easy. Research in this field involves several disciplines and is fragmented by competing theories that have not yet explained the pathways to family violence in a way that could serve as a basis for effective interventions. Understanding the challenges to research on family violence interventions is an important step in developing a strategic plan that can help build the next wave of interventions in this field.
First, family violence has traditionally been defined as a problem of pathological or criminal behavior rather than one that involves a variety of contributing causes and several stages of disorders. For many years, researchers focused on individual risk factors such as psychological deficits or environmental causes, but they were unable to isolate a single trait or risk factor that characterized a batterer or offending parent. This lack of success suggests that more complex and interactive forces are at work that stimulate, sustain, or moderate violent behavior. But observing family or adult interactions or relationships that emerge over time involves more intensive, and more intrusive, measures and methods that require significant resources and effort.
Second, the absence of large-scale studies describing the distribution and patterns of violence in everyday relationships challenges the development of appropriate measures and methods in the field. Most of the cases that have formed the basis of scientific studies of family violence involve actions that have been reported to the courts, social service agencies, or health professionals. These cases often involve the most serious, and possibly most resistant, forms of violent or abusive behavior. Scientists in this field have generally not had access to large general population samples and appropriate comparison groups that could provide insight into the antecedents of violence in the home.
Third, the tendency to deal with family violence cases as separate incidents rather than as symptoms of a broader social pattern of disorder and dysfunctional behavior has also contributed to the difficulty of developing a research base to identify the origins of family violence. Research remains highly idiosyncratic, often focusing on clinical populations or retrospective memories rather than large research samples with control populations and direct observation.
Fourth, researchers who study family violence are hindered not only by a lack of resources but also by concerns about safety, privacy, ethics, and law. The stigma and bias associated with family violence also raise important concerns in developing large-scale population-based studies in this field. Research on the efficacy of treatment programs in addressing domestic violence, for example, can be problematic when offenders are randomly assigned to jail time, community service, or treatment. Questions about fairness and victim safety, in particular, complicate efforts to duplicate the clinical trial model in judicial settings. The involvement of vulnerable parties in service network and research studies requires consistent vigilance to ensure that appropriate protections exist to keep them from further harm.
Although a few well-defined and rigorous studies aimed at understanding the causes of family violence and evaluating the effectiveness of interventions have emerged, a large research literature consistent with high standards does not yet exist. The absence of this literature complicates the study of interventions, because it is difficult to identify for whom a particular approach might work, under what conditions, and for how long.
Learning from what we know
Studies of the effectiveness of family violence interventions commonly focus on the relative effects of individual programs rather than incorporating an appreciation of the complex, systemic, and pervasive nature of this problem within the lives of those affected by it. Although some victims experience acute trauma and other clinical symptoms, for many others the impact of victimization is diffuse and global, compromising their ability to react to stress or other situations that require self-control and the management of anger, fear, or uncertainty. This knowledge suggests that service interventions will need to address family violence as an integral part of the lives of the victims and offenders rather than an isolated component of their behavior or experience.
As they acquire greater awareness of the relationships between different types of abuse (for example, the connection between witnessing domestic violence in the home and using violence as part of adult relationships later in life), researchers and practitioners are hopeful that new opportunities will emerge to apply preventive interventions for children and adolescents who have been exposed to violence before it can become part of their own parenting or intimate behaviors.
In other areas, stronger theories about the causes of family violence can help specify how the targets, scope, and timing of interventions should be designed. For example, improving the quality of or access to parenting education will not help prevent child maltreatment if a large number of offending parents abuse their children as a result of adult depression, anxiety, and social isolation or if parents lack social support in using nonviolent ways to discipline their children. Similarly, a husband who abuses his wife only when he is drunk may require a different type of service response than one who is habitually aggressive to friends and family alike. The latter case may be a strong candidate for law enforcement interventions designed to isolate and punish the offender, whereas the former might be more appropriate for a program that combines elements of marital counseling, anger management, and substance abuse treatment. If different types of offenders are placed in a treatment program that works effectively for only one segment of the total population, the results of the evaluation study will be marginal or mixed, even though the program may be highly effective for a small portion of the total study sample. And if the client base for the program is too small or the follow-up times are too short to demonstrate the scale of the effects that a specific intervention can achieve, promising approaches may be too hastily dismissed when the research evidence is simply insufficient to warrant their continuation.
Definitional and classification issues also reveal areas where research can help inform practice. At present, family violence is often grouped into separate categories that capture isolated and fragmentary parts of the experience of victims and offenders. The broad diversity of these categories makes it difficult to organize existing interventions within frameworks that could strengthen interactions among different service settings. Treatment, prevention, and deterrence interventions use different focal points (the victim, the offender, the family, or the community), different units of measuring behavioral or social change, and different outcomes in assessing whether a selected program is achieving its intended effect.
And there are subcategories within categories. Child maltreatment, for example, covers a broad array of offenses: physical child abuse, sexual child abuse, child neglect, and emotional maltreatment. Treatment or prevention interventions for each of these areas often emerge within frameworks that are differentiated by the types of abuse that are reported. Child sexual abuse cases, for example, are more commonly referred to the police or courts for action, whereas physical abuse and neglect cases generally become the responsibility of social service agencies.
These categories and subcategories reflect the details of a case when it is first recognized or reported, reflecting the emphasis of current interventions on problem identification and acute response. The categories are of little use in understanding the underlying causes or forces at work within the family. Although these arbitrary divisions may be appropriate for legal and social service interventions, they present major challenges to researchers who are more concerned with issues such as the frequency of events, levels of intensity, variety of triggers, nature of relationships, and patterns of response. As researchers move away from case records to population-based studies to determine how hidden dynamics influence the continuum of violence, new categories may emerge that could more appropriately guide service interventions in allocating resources, fostering service interactions, and applying appropriate responses to individuals and families.
Building partnerships between research and practice in the field of family violence requires opportunities for sustained collaboration that can allow scientists and service providers to understand more about the multidimensional nature and sequence of patterns of victimization, trauma, and violence. Service providers point to the absence of coordination and comprehensive approaches within their communities as an obstacle to gaining perspective on the multiple dimensions of family violence. Researchers have consistently expressed concern about the lack of resources invested in work necessary to build longitudinal studies. Although many health, law enforcement, and social service agencies support programs and research in family violence, these efforts are scattered across several sectors and lack sufficient strength to define program priorities, promising directions, and research needs. In the meantime, caseworkers, health care providers, law enforcement officials, and the general public must struggle with what to do when confronted with cases of family violence in their communities.
In the midst of this confusion and uncertainty, however, one fact looms above the others. We are entering a new stage in building the scientific knowledge base necessary to understanding and treating the problem of family violence. Although this research is not yet strong enough to provide clear guidance to those who must make policy or programmatic decisions about specific treatment and prevention interventions, a large set of descriptive and empirical studies has emerged that provides an important foundation for future scholarship and programmatic efforts. Like many of the family violence interventions themselves, scientific research in this field is still young and immature, consisting largely of studies that rely on reports of what was done in the intervention rather than probing more carefully into the outcomes that the treatment or prevention effort sought to achieve, the characteristics of the client base, the pathways by which clients were assigned to the program, and the implementation process of the program itself. Critical tools that provide the foundation of solid evaluation studies in other fields-strong theory, large longitudinal and follow-up studies, reliable measures, and consistent definitions and diagnostic criteria-are only beginning to develop in the field of family violence studies.
Collaboration and comprehensive services
If family violence were an infectious disease, an intensive research effort would gradually become an integral part of the design of treatment and prevention interventions, involving large-scale clinical trials designed to test the strength, efficiency, and effects of various interventions. Problems associated with theory, measurement, sample size, and comparison groups would gradually be resolved as researchers and practitioners acquired greater familiarity with the nature of the phenomenon under study and learned from the experience of selected groups. As the knowledge base expanded, service providers and researchers would find ways to communicate their findings and experiences with each other and discover ways to improve on the first generation of interventions to build stronger, durable, and more effective services. This type of infrastructure development and the integration of knowledge and practice have not occurred in family violence for the reasons discussed above. But we now have the opportunity to build such collaborative efforts, which can incorporate findings from existing research.
Assessing the effectiveness of treatment, prevention, and deterrence interventions for family violence, especially in open-service systems that have little control over which types of clients are assigned to experimental interventions or to standard service practices, requires creative strategies and close collaboration among researchers and practitioners. Yet caseworkers, who often have little opportunity to participate in the design of the evaluation studies of their services, can be understandably resentful of scientific evaluations that require randomized controls or experimental designs that require them to place vulnerable children, adults, or families in service settings that they "know" are not ideal. Likewise, they may be unwilling to collaborate with researchers if they believe that negative study results will undermine the funding base of their program in the future. Similarly, court officials who are concerned with the impact of sanctions and deterrent measures on future violent behavior may be reluctant to use scientific methods to compare the relative effectiveness of treatment and stiffer sanctions if concerns about fairness are not resolved.
An increasing emphasis on the need for integration of health care, social service, and law enforcement interventions in the area of family violence has now fostered an even more daunting task: how to measure the impact and cost-effectiveness of comprehensive community interventions that are designed for specific geographic regions and are affected by multiple service systems within their local settings. Yet despite the obstacles associated with their design and implementation, comprehensive interventions are revealing key opportunities where service providers concerned with family violence can interact more effectively with their counterparts in other service settings. For example, although studies have found a strong link between substance abuse and domestic violence, substance abuse treatment programs rarely address the problems of domestic violence, and batterers' treatment programs often lack the capacity to address the addictive behaviors of their clients. Comprehensive intervention seems to be a promising approach, but research to verify its effectiveness and identify the best mix of components will be difficult.
Despite these challenges, research studies are beginning to take a closer and more critical look at the knowledge base associated with family violence interventions. This research base, which includes more than 100 studies conducted over the past two decades that meet minimal standards of scientific rigor, is highlighting specific areas where evaluation studies have provided firm evidence of positive or negative effects of family violence interventions.
Although it is premature to offer policy recommendations for most family violence interventions in the absence of a more rigorous research base, a few lessons can be drawn from current studies:
Mandatory reporting laws for domestic violence should not be enacted until such systems have been tested and evaluated by research. In spite of extensive experience with mandatory reporting, no reliable evaluation of its effectiveness has been completed. Because a report of violence can sometimes diminish protections or precipitate more violence, particularly when there is no intervention, it is wiser to preserve the discretion of health professionals and other service providers in determining when to report troubled families. Such discretion is particularly important under circumstances where the provision of care is disrupted by the process of notification.
Early warning systems are needed in judicial settings to detect failure to comply with or complete treatment and to identify signs of new abuse or retaliation against victims. Research has indicated that reports of violent behavior often diminish while a batterer is in treatment, and victims are often more inclined to return to a relationship with a batterer if they know that he is enrolled in a program designed to curb violence in intimate relationships. But if the batterer drops out of the program or fails to enroll after court referral, the victim may be at great risk. Early warning systems would require court oversight to ensure offender compliance with the requirements of treatment referrals and should also address unintended or inadvertent results that may arise from the referral to or experience with treatment. For example, some treatment programs may simply involve marital counseling without ever addressing the violence that occurred in the relationships. Others may lack appropriate supervision within a group treatment program. As a result, a batterer may feel justified in using violence since others report that they act in a similar fashion with their partners. The absence of certification for treatment programs allows the use of practices that lack appropriate research support. Courts need to be vigilant to ensure that offenders are referred only to treatment programs that are known to be adequate.
Abuse and histories of family violence need to be documented in individual and group health care and social service records, but such documentation requires safeguards. If research continues to suggest that family violence is a significant contributor to health outcomes and family caregiving practices, the need to know about early and chronic histories of family violence will become stronger. This need will have to be balanced against individual privacy and confidentiality concerns. Health professionals are often reluctant to record incidents of violence, viewing such events as legal rather than medical matters. But patterns of violent injury should not be allowed to go unnoticed in medical histories, since they can provide important clues to future health disorders, especially in areas that involve chronic fear, stress, anxiety, trauma, or anger.
Collaborative strategies among caseworkers, lawmakers, prosecutors, and judges have the potential to improve a batterer's compliance with treatment as well as to make sanctions more effective. Studies of police arrest practices, for example, have indicated that arrest has a greater deterrent effect on future violence than simple counseling or arbitration. Arrests without prosecution, however, can send messages to individuals and communities that family violence is not to be taken seriously. Although challenged by evidentiary standards and victim reluctance to press charges, courts need to find collaborative ways to improve the penalties for family violence and to ensure that adequate treatment programs exist for those who are motivated to change their practices.
Home visitation programs should be particularly encouraged as part of a comprehensive prevention strategy for child maltreatment, especially for first-time parents living in social settings with high rates of child maltreatment. An array of research studies has been conducted on home visitation services that offer promising findings, but they suggest that the greatest benefits of these services occur only with young, single, and poor mothers. But this group includes many who are highly mobile or reluctant to trust others offering them advice and guidance. We need to learn more about how to engage high-risk parents in prevention services designed to improve their child care-giving skills and also to improve the general quality of their own personal health care, education, and job training, which may require them to spend more time out of the home.
Although intensive family preservation services are an important part of the continuum of family support services, they should not be required in every situation in which a child is recommended for out-of-home placement. Community agencies need to be able to act quickly and decisively when children are endangered or neglected. The costs of delay can be enormous in short- and long-term consequences. Although some families can recover from difficulties if they are provided with appropriate services and guidance, many others are overwhelmed by their problems and simply cannot care for their children. When faced with recurring patterns of child abandonment and abuse, social service and law enforcement agencies should look for appropriate ways to provide reasonable care for vulnerable family members rather than working solely to preserve the family.
Does this analysis imply that the existing array of treatment and prevention interventions for family violence is misguided and a poor use of public funds? Not at all. It does suggest, however, that greater efforts need to be made to build an evaluation and research capacity that can help inform, question, and guide the development of service interventions. Improved evaluation studies can lead to better and more efficient service interventions that reflect an empirical and comprehensive understanding of the problem of family violence rather than a set of beliefs and anecdotes that lead to single-minded approaches and often result in inconsistent and piecemeal efforts to address the needs of victims and offenders. The development of family violence interventions needs to be seen as an iterative process in which services are initially put into place to respond to immediate needs but are refined and improved over time as a knowledge base emerges through collaboration between the service provider and research communities.
Building the next generation of evaluation studies and service interventions for family violence will require leadership and coordinated strategies within the policy, program, and research communities that cut across traditional disciplines and service settings. Efforts have already begun to integrate the network of health, social service, and law enforcement interventions at the local, state, and national policy levels, but many efforts remain disconnected. The Departments of Justice and of Health and Human Services are now collaborating in funding small-scale studies and the development of improved research measures and training programs, but new initiatives are rare because of the lack of new funds and the absence of innovative strategies that could strengthen interdisciplinary approaches to complex social policy concerns. It is far easier for Congress to appropriate funds for hundreds of community-based programs than to craft a bold research strategy that would integrate biological, psychological, social, behavioral, medical, and criminal justice research focused on family violence. But it must take the more difficult path. In the absence of new research-practice partnerships and an infrastructure capable of supporting long-term studies, we will continue to be faced in the next decade with a profusion of interventions and a limited capacity to examine their effects.
R. Chalk and P.A. King (eds.), Violence in Families: Assessing Prevention and Treatment Programs. National Research Council and the Institute of Medicine. Washington, D.C.: National Academy Press, 1998.
J. P. Connell, A. C. Kubisch, L.B. Schorr, and C.H. Weiss (eds.), New Approaches to Evaluating Community Initiatives. New York: The Aspen Institute, 1995.
N. Crowell and A. Burgess (eds.), Understanding Violence against Women. National Research Council. Washington, D.C.: National Academy Press, 1996.
G. Melton and F. Barry (eds.), Protecting Children from Abuse and Neglect. New York: Guilford, 1994.
National Research Council, Understanding Child Abuse and Neglect. Washington, D.C.: National Academy Press, 1993.
Schuerman, J.R., T.L. Rzepnicki, and J.H. Littell (1994). Putting Families First. New York: Aldine de Gruyter.
Rosemary Chalk was study director of the National Research Council Committee on the Assessment of Family Violence Interventions. Patricia A. King, the Carmack Waterhouse Professor of Law, Medicine, Ethics, and Public Policy at the Georgetown University Law Center, was chair of the committee.