2
Family Violence and Family Violence Interventions

In the 1960s, physical child abuse and child neglect were recognized as significant problems that required identification and intervention. In the 1970s, similar attention was called to domestic violence. In the 1980s, child sexual abuse and elder abuse gained attention. Each type of family violence developed its own set of definitions, research, and interventions. In recent years, researchers have begun to examine family violence across the life span, seeking commonalities among various forms of child maltreatment, domestic violence, and elder abuse (Barnett et al., 1997). Although the very nature of family relationships makes it reasonable to look for common risk and protective factors and interactions among the forms of family violence, debate continues about what determines the appropriate unit of analysis: the individuals who perpetrate or are victims of violent behavior, the family, the community, or broader social-cultural norms.

How the nature of family violence is conceptualized has important implications for the ways in which interventions are structured and outcomes are measured in evaluating them. This chapter examines similarities and differences in research on child maltreatment, domestic violence, and elder abuse in definitions, measurement, risk factors, and interventions. Each area has developed its own approach to the problems under study, resulting in tremendous variation in theoretical frameworks, research instrumentation, scholarly journals, and funding sources. Although studies of the field of family violence are now emerging, research in this area has traditionally been fragmented into separate areas of inquiry.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 31
--> 2 Family Violence and Family Violence Interventions In the 1960s, physical child abuse and child neglect were recognized as significant problems that required identification and intervention. In the 1970s, similar attention was called to domestic violence. In the 1980s, child sexual abuse and elder abuse gained attention. Each type of family violence developed its own set of definitions, research, and interventions. In recent years, researchers have begun to examine family violence across the life span, seeking commonalities among various forms of child maltreatment, domestic violence, and elder abuse (Barnett et al., 1997). Although the very nature of family relationships makes it reasonable to look for common risk and protective factors and interactions among the forms of family violence, debate continues about what determines the appropriate unit of analysis: the individuals who perpetrate or are victims of violent behavior, the family, the community, or broader social-cultural norms. How the nature of family violence is conceptualized has important implications for the ways in which interventions are structured and outcomes are measured in evaluating them. This chapter examines similarities and differences in research on child maltreatment, domestic violence, and elder abuse in definitions, measurement, risk factors, and interventions. Each area has developed its own approach to the problems under study, resulting in tremendous variation in theoretical frameworks, research instrumentation, scholarly journals, and funding sources. Although studies of the field of family violence are now emerging, research in this area has traditionally been fragmented into separate areas of inquiry.

OCR for page 31
--> Definitional Issues How family violence is measured and studied and who is identified to receive interventions depend on its definition. The multiple dimensions of family violence make definition tasks both difficult and controversial, and definitions have evolved somewhat differently for child maltreatment, domestic violence, and elder abuse. Although precise definitions are required for research and data collection, the situational context of many cases of family violence often challenges efforts to establish consistent criteria that can govern case reporting and selection decisions. The types of cases deemed appropriate for intervention may differ by service sector as well as within service settings. Despite these differences, common concerns have emerged that influence efforts to define family violence: a trend toward broadening the set of behaviors that are viewed as abuse or neglect, the role of cultural norms, the role of intentionality, the balance of power in the offender-victim relationship, and the influence of service-sector resources on definitions. Broadening of Definitions Early definitions of child maltreatment were narrowly focused on issues of physical harm and endangerment. They were often based on a medical diagnostic approach, emphasizing the more serious forms of physical abuse and neglect that can result in physical injury and poor health. More recent definitions focus on the impact of maltreatment on development, expanding the category of abuse and neglect to include actions viewed as harmful to children, such as emotional abuse or educational or medical neglect, even though they are not necessarily associated with immediate physical injury. The negative impacts of exposure to domestic violence on children have recently been recognized. Some perpetrators of domestic violence have been charged not only with assault of their wives, but also reported as child abusers for exposing their children to their violent behavior. The National Center on Child Abuse and Neglect (1988) recognizes six major types of child maltreatment: physical abuse, sexual abuse, emotional abuse, physical neglect, educational neglect, and emotional neglect. Similarly, initial definitions of domestic violence focused on acts of physical violence directed toward women by their spouses or partners (Gelles and Straus, 1974; Martin, 1976). Survey research found women also committed acts of physical violence against their husbands, suggesting that the initial definition might be too restrictive (Straus et al., 1980; Gelles, 1987; Gelles and Straus, 1988; Straus and Gelles, 1986). Further research broadened the definition to include sexual violence, marital rape, and acts of emotional or psychological abuse. Feminist scholars conceptualize domestic violence as coercive control of women by their partners (Yllo, 1993); the coercion can be physical, emotional, or sexual.

OCR for page 31
--> The recognition of elder abuse through adult protective services has led to definitions similar to those of child maltreatment. Elder abuse is generally defined to include physical and sexual violence, psychological abuse, and neglect. However, elder abuse definitions frequently go beyond those of child maltreatment to include financial abuse or exploitation. Elder abuse may also be a continuation of domestic violence in the relationship, in which case the issue of coercive control is pertinent. Definitions of family violence have also broadened with respect to who is considered family. For domestic violence, the trend is to include all intimate relationships, such as cohabiting couples, same-sex couples, and ex-spouses, as well as currently married couples. Dating couples and ex-boyfriends/girlfriends may also be included. For child maltreatment, a distinction is made between abuse and neglect; the definition of neglect is generally limited to acts by a child's parents or legal caregiver. For child physical and sexual abuse, no consensus exists as to how broadly the set of potential abusers should be defined (National Research Council, 1993a), although, in the context of family violence, strangers are not included. For elder abuse, family members as well as legal or informal caregivers can be involved, and self-neglect has sometimes been counted in elder abuse data. In addition, violence between siblings and violence by children directed at their parents may be included in family violence. The broadening of definitions of the various forms of family violence has been criticized by some on the grounds that it leads to a diffusion of effort within service agencies that have limited resources. Others have noted, however, that because emotional abuse and emotional neglect are often associated with other forms of maltreatment, the broader definitions create opportunities for early interventions that may be successful in preventing more serious and lethal forms of violence. How broadly or narrowly family violence is defined has an obvious impact on its prevalence. Wyatt and Peters (1986) demonstrated empirically the impact of variations in definitions of sexual abuse on the estimated prevalence rates reported in a number of studies. Table 2-1 presents the rates of violence from various studies broken down by physical violence, sexual abuse, emotional abuse, and neglect. Physical violence is the most commonly measured form of maltreatment of adults and is second only to neglect in the maltreatment of children. Although clinical reports suggest that emotional abuse (such as yelling, criticism, ridicule, and threats) may be pervasive and central components of family violence, no rigorous measures currently exist to assess the extent of this type of behavior. The Role of Culture Definitions of family violence vary not only over time, but also across cultures. Behavior that may be considered the norm by some groups may be considered

OCR for page 31
--> TABLE 2-1 Past Year Rates of Family Violence (per 1,000 persons) Study and Author Sample Size All Maltreatment All Physical Violence Severe Physical Violence Sexual Abuse or Marital Rape Emotional or Psychological Abuse Neglect Fatal Abuse Children NIS-3 (National Center on Child Abuse and Neglect, 1996a)   41.6 (reported) 9.2 — 4.4 7.9 20.1   NCANDS (National Center on Child Abuse and Neglect, 1996b)   43 (reported) 3.4 — 2 <0.1 6 — National Committee to Prevent Child Abuse (1996)   47 (reported) 16 (substantiated — — — — — 0.0019 1985 National Family Violence Survey (Straus and Gelles, 1988) 3,232 — 498 23 — 63.4 — — Finkelhor et al. (1990) Females 1,374 — — — 270a — — — Males 1,252 — — — 160a — — — Siblings 1975 National Family Violence Survey (Straus et al., 1980)   — 800 530 — — — — Adult Women 1985 National Family Violence Survey (Straus and Gelles, 1988) 6,002 — 116 34 12 74 — — National Crime Victimization Survey (Bachman and Saltzman, 1995) ˜50,000 — 9.3 — 1 — — — Adult Men 1985 National Family Violence Survey (Straus and Gelles, 1988) 6,002 — 124 48 not asked 74.8 — — National Crime Victimization Survey (Bachman and Saltzman, 1995) ˜50,000 — 1.4 — —b — — — Elderly Adults Pillemer and Finkelhor (1988) 2,020 32 20 — — 11 4 — NOTE: Dash indicates absence of data. a Prevalence rates of sexual abuse before age 18. b Ten or fewer sample cases. SOURCE: Committee on the Assessment of Family Violence Interventions, National Research Council and Institute of Medicine, 1998.

OCR for page 31
Complete table on previous page.

OCR for page 31
--> abusive by others. Debate continues over the extent to which definitions of family violence can accommodate cultural variability without promoting different thresholds for intervention on the basis of race, ethnicity, or socioeconomic status. Agreement across cultural groups may be easy to reach on severe forms of family violence, such as beating a family member to death. Less severe behaviors are more subject to cultural differences. For example, a number of child advocates, researchers, and some parent groups believe that corporal punishment is abusive and harmful to children and should be illegal (Straus, 1994). In contrast, many working-class parents believe in strict child discipline practices and favor its use. Role expectations derived from traditional family relationships (which commonly viewed children and wives as property of the head of the household) can also contribute to conflict and violence when major changes occur in the status of women and children within a society. Pervasive gender inequality may be embedded in cultural norms and may influence patterns and rates of victimization. As U.S. society becomes more diverse, cultural practices such as ''coining" (a curing ritual that involves the forceful pressing of coins on a child's body, resulting in bruises) among Vietnamese populations and female genital mutilation among some African and Middle Eastern populations will increasingly be the focus of debate. Although such practices may meet criteria for abuse in the United States, it is important to understand their social and cultural contexts, at least in the perceptions of the families involved, in order to influence behavioral changes. American society lacks a clear definition of caregiver responsibility for elders comparable to parental responsibilities for children, and conflicts may arise in certain cultures, classes, and social groups as to what types of arrangements constitute maltreatment. Recently, some researchers have questioned the legal and clinical definitions of elder abuse, suggesting that it is the older person's perception of particular behavior, influenced by culture and tradition, that should be the salient factor in identification and intervention (Gebotys et al., 1992; Hudson, 1994). The Role of Intentionality Although some definitions of family violence require that acts be intentional (see, e.g., Gelles and Straus, 1979; National Research Council, 1993b), recent trends in research definitions have been to focus on acts that are harmful or potentially harmful rather than on intent (Zuravin, 1991). Although intent is often difficult to determine, it is often important in determining the type of intervention necessary when decisions are made to investigate a family. Neglect of children and the elderly may result from ignorance or lack of resources rather than malevolence. Parental knowledge of children's developmental abilities (especially feeding and toilet training), parental abilities to supervise

OCR for page 31
--> young children, and parental expectations regarding the needs and behavior of young children have been identified as key areas in which significant variation exists within neighborhoods and between social classes and cultures. These discrepancies may be particularly difficult to resolve in identifying cases of maltreatment that occur in settings characterized by extreme poverty, substance abuse, community violence, and transience (including immigration and homelessness). Elder abuse may also be unintentional because of the caregiver/perpetrator's ignorance of the needs of the elderly or because of his or her own infirmity. The lack of a clear definition of caregiver responsibility for elders also makes determining intentional withholding of care difficult. Whether behavior is labeled as abusive or neglectful may depend on its frequency, duration, intensity, severity, and consequences. Dependence and Power Running through discussions of child maltreatment, domestic violence, and elder abuse is the idea of unequal power in the relationship between abuser and victim. In all three domains, abusers may use violence to control the victim. It is expected that children are dependent and that parents have more power than their children and will exert control over them. However, if the parent abdicates responsibility for caring for the child, neglect can occur. If the parent uses excessive physical force in exerting control, abuse can occur. For the incapacitated elderly, it is also expected that they are dependent on family or a caregiver; the typical elder abuse victim was thought to be frail and dependent. Although some research has found that elder abuse may be more likely when an adult child remains financially dependent on his or her parent rather than when the elderly parent has become dependent (Pillemer and Suitor, 1992), the adult child may still have power over the elderly parent by virtue of size, strength, and mobility. The feminist analysis of domestic violence posits that physical violence is but one tactic used by abusers to exert control over their partners. In this paradigm, physical violence, emotional abuse, sexual violence, social isolation, and withholding of financial resources all serve to undermine a woman's autonomy and limit her power in the relationship. Feminist theory focused on gender inequalities offers a framework for a broad set of interventions designed to expose the pattern of male dominance and control in violent relationships and the ways in which society tolerates and legitimizes social inequalities that can contribute to entrapment. Influence of the Service Sector on Definitions There is an interplay between the conceptualization of the types of family violence and the service sector in which interventions take place. For example, the identification of the "battered child syndrome" in the health care system at

OCR for page 31
--> first focused the definitions of child maltreatment on physical injury (Kempe et al., 1962). As the legal and social service systems became involved and the emphasis in services became that of protecting the child, acts that threatened the child's well-being became a relevant factor in defining maltreatment; the definitions were subsequently broadened to include emotional abuse, sexual abuse, and neglect (National Research Council, 1993a). It is important to note that legal definitions encompass stringent standards in each category of abuse focused on thresholds for legal intervention that may not be useful in other service systems. Similarly, the emphasis in domestic violence was first on physical injury. Even the term "battered wife" echoed the earlier use of "battered child," and much attention was focused and continues to be focused on the identification and documentation of physical injury in medical settings. Also for domestic violence, an emphasis on physical injury carried over to early legal interventions. For example, evidence of physical injury was often deemed necessary before an arrest was made. As a result of research studies and programs to change community attitudes toward the needs of battered women, definitions broadened beyond physical injury to include sexual and emotional abuse and threats of harm. Recent legislation now recognizes the stalking of estranged or former partners as domestic violence. The recognition of elder abuse grew out of adult protective services—that is, services for adults who, for whatever reason, are unable to care for themselves. The similarity to child protective services heavily influenced definitions of elder abuse. It has been suggested that there has been an overreliance on the child abuse model. Recent findings suggest that domestic violence may be a more useful framework for study and intervention, since the individuals involved are legally independent adults. To some health researchers, however, the family violence paradigm is also not suitable. They recommend that elder abuse be considered from the perspective of "inadequate care," since it is easier to measure unmet needs than inappropriate behavior (Fulmer and O'Malley, 1987). Measurement Issues Even if there were consensus on definitions, estimates of the scope of family violence would vary because of the different methods used to measure its incidence and prevalence. Detailed discussions of measuring the scope of family violence are available elsewhere (National Research Council, 1993a,b; 1996). Data on family violence are available from clinical data, administrative data, and social surveys. Each type of data has its own strengths and weaknesses and is useful for different purposes. Clinical studies, carried out by psychiatrists, psychologists, and counselors, continue to be a frequent source of data on family violence. This is primarily due to the fact that these investigators have the most direct access to cases of family violence. The clinical or agency setting (including hospital emergency departments

OCR for page 31
--> and battered women's shelters) provides access to extensive in-depth information about particular cases of violence. Studies of violence toward women have relied heavily on samples of women who seek help at battered women's shelters (Dobash and Dobash, 1979; Giles-Sims, 1983; Pagelow, 1984). Such samples are important because they are often the only way of obtaining detailed data on severely battered women. Such data are also necessary to study the impact of intervention programs. However, because they are based on small, nonrepresentative samples, such data cannot be used to estimate the incidence and prevalence of domestic violence. Administrative data collected from official sources, such as child protective services, police departments, and adult protective services, provide the bulk of the information from which estimates of the incidence and prevalence of family violence are made. There are a number of factors that influence how well these data accurately reflect true incidence and prevalence. First, the quality and amount of these data vary by type of violence. All states now mandate reporting of child maltreatment and most states mandate reporting of elder abuse, although the definitions of reportable abuse differ from state to state. Even with mandatory reporting laws, it is widely acknowledged that abuse is underreported (e.g., Kalichman, 1993; Widom, 1988). One study found that only 44 percent of cases known to community professionals were officially known to child protective services agencies (Sedlak, 1991). This discrepancy may be based on the unwillingness of professionals to report, contradictory community standards regarding maltreatment definitions, or screening practices in child protective services agencies (Downing et al., 1990; Zellman, 1992). Furthermore, the number of substantiated cases may vary with the amount of resources devoted to case investigation. There is no administrative data source for domestic violence that is comparable to the state child and elder abuse datasets, and surveillance efforts are often challenged by concerns about victim safety and confidentiality. Second, segments of the population at risk are subject to different levels of surveillance. Families who, because of their demographic characteristics (e.g., poverty, unemployment, single parenthood), have more frequent contact with public-sector services (e.g., welfare, housing) are more often exposed to mandated reporters and get closer scrutiny. Consequently, children who are poor or members of minority groups are more likely to be identified as maltreated than are more affluent white children (Newberger et al., 1977). To remedy underreporting, national incidence studies of child maltreatment include official reports supplemented by surveys of other professionals about suspected cases of child abuse that have not been reported. The most recent National Incidence Study of Child Abuse and Neglect (NIS-3) found 2.8 million cases of child maltreatment (National Center on Child Abuse and Neglect, 1996b), a significant increase over the 1.4 million cases known by the agencies surveyed in 1986 (Sedlak, 1991). An unknown part of this increase may be due to broader definitions and broader sampling frames in the survey sample. Based on the NIS

OCR for page 31
--> methodology, the National Center on Elder Abuse has launched a similar incident study of elder abuse, scheduled to be completed in 1997 (National Center on Elder Abuse and Neglect, 1995). Researchers often turn to survey research to overcome the problems associated with administrative data. Surveys often find higher rates of violence than do administrative data; however, surveys have their own drawbacks. They are constrained by the low base rate of most forms of abuse and violence in families, as well as the sensitive and taboo nature of the topic. Some investigators cope with the problem of the low base rate by employing purposive or nonrepresentative sampling techniques to identify cases. Another approach has been to use large groups of available subjects. For example, investigators of courtship violence have made extensive use of survey research techniques using college students as subjects (Henton et al., 1983; Laner, 1989a,b; Makepeace, 1981, 1983). Other sources of bias in social survey data on family violence include inaccurate recall, differential interpretation of questions, and intended and unintended response error (Weis, 1989). Survey results may also vary by question wording and context. For example, when the National Crime Victimization Survey revised its questionnaire to stress, among other things, that they wanted to know about incidents involving relatives and family members as well as acts committed by strangers, the reported violent attacks by family members nearly doubled from 5.4 per 1,000 for women in 1987-1991 to 9.3 per 1,000 in 1992-1993 (Bachman and Saltzman, 1995). This change most likely reflects the change in the survey and not a sudden increase in the rate of violence against women (National Research Council, 1996). The rates for domestic violence are lower than the rates found in other surveys that did not rely solely on crime reports. The 1985 National Family Violence Survey found much higher rates of domestic violence: 116 per 1,000 women reported experiencing violence during the past year; 34 per 1,000 women reported severe violence; and 12 per 1,000 reported marital rape (Straus and Gelles, 1988). In the 1993 Commonwealth Fund Survey of Women's Health, 70 per 1,000 women reported physical violence by an intimate partner. The source of data has an impact not only on measures of incidence and prevalence of family violence, but also on what factors and variables are identified as risk and protective factors. In an analysis based on clinical data or official report data, risk and protective factors are confounded with factors such as labeling bias and agency or clinical setting catchment area. Researchers have long noted that certain individuals and families are more likely to be correctly and incorrectly labeled as offenders or victims of family violence (Gelles, 1975; Newberger et al., 1977; Hampton and Newberger, 1985). Social survey data are not immune to confounding problems either, as social or demographic factors may be related to willingness to participate in a self-report survey and a tendency toward providing socially desirable responses.

OCR for page 31
--> Risk Factors Researchers have looked, unsuccessfully for the most part, for traits that predispose children or adults to being victims of family violence. Much of the early retrospective research lacked comparison groups and confounded traits that preceded abuse with those that resulted from it. Research next focused on the traits of perpetrators. Although there appears to be a relatively high incidence of psychological and emotional problems among perpetrators of family violence, no characteristic profile of a child abuser, batterer, or elder abuser has emerged. It appears that no single factor can explain family violence. Rather it seems most likely that the complex interaction of personal history, personality traits, and demographic factors with social and environmental influences leads to violence in the family. Understanding better the particular factors that are relevant in a given family, and the sequence in which they emerge, may be important for choosing the most appropriate intervention. In the sections that follow we review the most widely discussed risk factors in the study of family violence and, when appropriate, identify for which forms of violence and which types of relationships the factors are relevant. To overcome the limitations of studies of single factors, interactive models of risk and protective factors may further understanding of the etiology of family violence (Belsky, 1980; Garbarino, 1987; Lutzker et al., 1984; Malamuth et al., 1993). However, the complexity of analysis associated with these models and the difficulty of distinguishing causal effects from observational data have inhibited their testing and application (National Research Council, 1993b). Hence, research often focuses on risk factors that appear to be subject to change as a result of intervention or that are helpful in identifying at-risk populations. In research on domestic violence, several reviews of case-comparison studies have noted that potential risk markers are highly correlated with each other (e.g., family socioeconomic status and husband's occupational status, witnessing and experiencing violence as a child). This conceptual and statistical redundancy requires careful attention to disentangle risk factors that are directly related to the use of violence from other confounding factors (Hotaling and Sugarman, 1990). Risk Factors for Perpetrators Age One of the most consistent risk factors for perpetration is age. As in violence between nonintimates, family violence is most likely to be perpetrated by those between ages 18 and 30. For child abuse, the age of the mother at the birth of the abused child has been found to be related to rates of physical abuse, with younger mothers exhibiting higher rates of abuse (Kinard and Klerman, 1980; Connelly and Straus, 1992). An analysis of data from the 1985 National Family Violence

OCR for page 31
--> degree of gender inequality in a relationship, community, and society, the higher are the rates of violence toward women (Browne and Williams, 1993; Coleman and Straus, 1986; Levinson, 1989; Morley, 1994; Straus, 1994; Straus et al., 1980). One review of 52 case-comparison studies did not find significant differences in measures of sex role inequality between violent and nonviolent couples (Hotaling and Sugarman, 1986). In a later analysis, the authors observe that expectations about division of labor in the household was one of four markers associated with a risk factor that they labeled "marital conflict." The other three markers were marital conflict, frequency of husband's drinking, and educational incompatibility (Hotaling and Sugarman, 1990). Presence of Other Violence A final general risk factor is that the presence of violence in one family relationship increases the risk that there will be violence in others. For example, children in homes in which there is violence between their parents are more likely to experience violence than are children who grow up in homes where there is no such violence. Moreover, children who witness and experience violence are more likely to use violence toward their parents and siblings than are children who do not experience or see violence in their homes (Straus et al., 1980; Straus and Gelles, 1988; Fagan and Browne, 1994). Risk Factors for Victims Early research in domestic violence and child maltreatment looked for factors that differentiated victims from nonvictims. It was suggested that personality or other personal traits of victims could provoke anger or aggression. Much of this research has been criticized on methodological grounds (e.g., Leventhal, 1981), and it has been suggested that personality traits of victims identified in some early studies were the result rather than the cause of the violence (Hotaling and Sugarman, 1990; Pittman and Taylor, 1992). Compared with research on offenders, there has been somewhat less recent research on victims of family violence that focuses on factors that increase or reduce the risk of victimization. Most research on victims examines the consequences of victimization (e.g., depression, psychological distress, suicide attempts, symptoms of post-traumatic stress syndrome) and the effectiveness of various intervention efforts. Children Early research suggested that a number of factors raise the risk of a child's being abused. Low-birthweight babies (Parke and Collmer, 1975), premature babies (Elmer and Gregg, 1967; Newberger et al., 1977; Parke and Collmer,

OCR for page 31
--> 1975; Steele and Pollock, 1974), and children with developmental or other disabilities (Friedrich and Boriskin, 1976; Gil, 1971; Steinmetz, 1978) were all described as being at greater risk of being abused by their parents or caregivers. However, a review of such studies calls into question many of these findings (Starr, 1988). One major problem is that few investigators used matched comparison groups. More recent studies did not find premature babies or children with disabilities as being at higher risk for abuse (Egeland and Vaughan, 1981; Starr et al., 1990). The very youngest children appear to be at the greatest risk of abuse, especially the most dangerous and potentially lethal forms of violence (Fergusson et al., 1972; Gil, 1971; Johnson, 1974). Not only are young children physically more fragile and thus more susceptible to injury, but also their vulnerability makes them more likely to be reported and diagnosed as abused when injured. Older children are most likely to be underreported as victims of abuse. Marital Partners Being female is the most consistent risk factor for being a victim of domestic violence (Hotaling and Sugarman, 1986). Early studies unsuccessfully attempted to find a psychological profile that put a woman at risk of being battered. Early descriptive and clinical accounts described battered women as dependent, having low esteem, and feeling inadequate and helpless (Ball, 1977; Hilberman and Munson, 1977; Shainess, 1979; Walker, 1979) and reported a high incidence of depression and anxiety among clinical samples (Hilberman, 1980). Later studies have questioned whether these victim characteristics were present before the women were battered or are the result of the victimization (Hotaling and Sugarman, 1990). Clinical studies often use small and selective samples and fail to have comparison groups. A comprehensive review of risk factors found that the only one consistently associated with being a victim of physical abuse was having witnessed parental violence as a child (Hotaling and Sugarman, 1986). As noted earlier, this finding was modified in a later review, and the authors attribute this modification to the use of multivariate analysis that can distinguish between minor and severe violence, marital conflict, and the use of violence in the home (Hotaling and Sugarman, 1990). Elders Research on elder abuse is divided on whether elder victims are more likely to be physically, socially, and emotionally dependent on their caregivers or whether it is the offender's dependence on the victim that increases the risk of elder abuse (see Pillemer and Suitor, 1992; Steinmetz, 1990). Conventional wisdom suggests that it is the oldest, sickest, most debilitated, and dependent

OCR for page 31
--> elders who are prone to the full range of mistreatment by their caregivers. However, Pillemer and Suitor (1992) have found that the victim's dependence was not as powerful a risk factor as perceived by clinicians, the public, and some researchers. A history of violence, particularly between spouses, may be predictive of elder abuse in later life (Lachs and Pillemer, 1995). Interventions A broad set of public and private initiatives at the national, state, and local levels has generated a complex array of institutions to address the consequences and the origins of family violence, but the dimensions of this effort are difficult to observe or measure (see Table 2-2). The large majority of legal, health, and social service interventions and programs relies extensively on community resources that reflect both regional strengths and limitations. In some cases, such efforts have emerged in the absence of national legislation to determine eligibility criteria or accountability standards as a basis for federal funding, and activities have evolved in health, social service, and legal settings (such as home visitations programs and adult protective services) that were not designed with the treatment or prevention of family violence as a primary goal. In other cases, federal and congressional initiatives sought to influence the design of interventions by establishing eligibility criteria, funding direct services, and specifying key elements of program design (such as the Family Violence Prevention and Services Act, the Child Abuse Prevention and Treatment Act, and the Violence Against Women legislation). As a result, the overall "system" of family violence interventions is highly disjointed, loosely structured, and often lacks central coordinating offices or comprehensive service delivery systems. Fragmentation of Services The fragmented nature of the categorical services involved in addressing different aspects of family violence has inhibited the development of a detailed picture and integrated datasets in this field. The fragmentation of services discourages research and cost analyses in assessing the impact and outcomes of family violence; it is therefore difficult to determine whether sufficient services exist in selected settings, the ways in which services address selected goals (such as identification, prevention, protection, or treatment), or the conditions under which persons in need of services do not have access to appropriate care. Victims of domestic violence, for example, may be seen by emergency department personnel, other health care providers, court officials, and battered women's shelter staff, but rarely do these service or agency staff members have an opportunity to collaborate, review, or understand the full dimensions of the victim's needs and experiences. Children who have been assaulted by relatives in their homes may have also witnessed incidents of violence between their parents, but the social

OCR for page 31
--> TABLE 2-2 Array of Services for Family Violence by Service Sector and Purpose Sector Prevention Case Identification/Risk Factor Detection Short-Term Victim Protection/Risk Assessment/Treatment Long-Term Intervention Social services Education programs; Service provider training programs; Community coordinating councils; Comprehensive community services; Community support groups Surveys; Case reports programs; Child protective services Shelters; Batterers' treatment; Family preservation services; Parenting practices and family support Peer support groups; Education and job training; Housing (transitional and permanent); Child and elder placement Health Service provider training programs Health reports; Emergency room procedures; Diagnostic protocols Home visitation and family support Mental health services for victims; Mental health services for offenders Law enforcement Service provider training programs Uniform crime reports orders; National crime victimization surveys Temporary restraining; Arrest procedures; Batterers' treatment programs; Victim advocates Offender incarceration; Sentencing guidelines; Prosecution procedures; Conditions of probation and parole   SOURCE: Committee on the Assessment of Family Violence Interventions, National Research Council and Institute of Medicine, 1998.

OCR for page 31
--> service agency and law enforcement personnel who respond to cases of child maltreatment may not be in a position to address other forms of violent behavior in the home. The fragmentation of the field has led some observers to question whether there is, in fact, a unified body of research and knowledge that can inform and integrate the different dimensions of this social problem. This complicated terrain has also discouraged and impeded efforts to determine the scope of federal, state, or local treatment and prevention programs, many of which are authorized and funded through a series of statutes that lack a coherent framework or strategy. Yet the broad array of services and resources, while poorly integrated, serves an extensive population. The lack of evidence and analysis about the effectiveness of the system of interventions as a whole does not mean that individual services are not doing an adequate job in addressing the needs of individual clients and communities. But the absence of coherence in the system of family violence interventions makes it difficult to observe or understand ways in which discrete parts of this system may interact, complement, or conflict with each other. In some cases, federal programs have provided resources to assist with organizing services, technical assistance, and data collection requirements that have helped to standardize a developing program or set of activities. Yet federal efforts lack a set of common goals or comprehensive strategy and often reflect administrative desires to meet the needs of child or adult victims in the context of existing service systems rather than initiating broad reforms focused explicitly on family violence. Furthermore, the role of the federal government has in large part been designed to strengthen and support local or state initiatives that reflect an enormous array of policy and program strategies, often through unrestricted block grants. The inherent flexibility in these types of funding mechanisms inhibits national or regional efforts to collect data on the clients, program outcomes, and measures of effectiveness. The broad scope and desegregated nature of federal efforts in the area of family violence interventions are reflected in a 1986 report prepared by the Office of the Assistant Secretary for Planning and Evaluation in the U.S. Department of Health and Human Services (DHHS) (U.S. Department of Health and Human Services, 1986). It is the last known federal interdepartmental review of programs focused on child abuse and neglect, domestic violence, and elder abuse; it identifies nine federal departments and three independent agencies that provide services or sponsor research relevant to family violence (see Table 2-3). The 1986 report was preceded by a detailed review of DHHS services for victims of domestic violence undertaken in 1979, when legislation for a categorical program of financial assistance for the provision of services to victims of domestic violence was being considered by the Congress (U.S. Department of Health and Human Services, 1981). The 1981 report noted that no information on services for domestic violence victims had been routinely collected in prior years, because no programs within DHHS had a specific mandate to serve these victims. This

OCR for page 31
--> TABLE 2-3 Federal Programs That Provide Services or Sponsor Research Relevant to Family Violence Department of Health and Human Services Social services   Project SHARE   Community services block grant   Office of Human Development services   Coordinated discretionary funds program   Social services block grant   Child welfare services   Foster care and adoption assistance   Head Start program   Runaway and homeless youth program   Developmental disabilities program   Native Americans program   Older Americans program   Child abuse and neglect prevention and treatment program   National Center on Child Abuse and Neglect   Advisory Board on Child Abuse and Neglect Health   Medicaid (Title XIX)   Primary care block grant   Maternal and child health block grant   Alcohol, Drug Abuse and Mental Health Services block grant   Preventive Health and Health Services block grant   Centers for Disease Control and Prevention   Adolescent Family Life Program   Indian Health Service Program   National Institute of Mental Health   National Institute on Aging   National Institute on Alcohol Abuse and Alcoholism   National Institute on Drug Abuse   National Institute on Child Health and Human Development Income support   Aid to Families with Dependent Children   Supplemental Security Income Program Department of Agriculture   Food stamps   Extension Service Americorps   The Foster Grandparent Program   Retired Senior Volunteer Program   Volunteers in Service to America (VISTA)

OCR for page 31
--> Department of Defense Department of Education Department of Housing and Urban Development Department of the Interior Department of Justice   Violence against women grants   National Institute of Justice   Office of Victims of Crime grants Department of Transportation Department of the Treasury United States Commission on Civil Rights General Services Administration   SOURCE: U.S. Department of Health and Human Services, 1986. report describes ways in which 13 separate DHHS service programs had been used to provide assistance to victims of domestic violence, although the vast majority of the eligibility requirements for these programs did not include domestic violence. For example, through the Social Services (Title XX) program, states receive federal funding for their child protective services programs, which provide social services designed to protect children from abuse, neglect, and exploitation. Service providers who participated in the survey that was part of the 1981 DHHS report indicated that they refer victims of spouse abuse to other service providers (such as shelters or adult protective services programs) for direct help with the problem of domestic violence (U.S. Department of Health and Human Services, 1981). As a result, multiple agencies may become involved in meeting the needs of one family, and problems of coordination and integration of services become increasingly important with the emergence of specialized services. Impacts and Costs The costs associated with family violence include two key components: (1) direct costs, those of providing treatment and services, and (2) indirect costs, such as reduced productivity, diminished quality of life (including pain and suffering), and decreased ability to care for oneself or others. It is extremely difficult to

OCR for page 31
--> estimate the range of these costs. Efforts to collect data on the dimensions of the programs and services are impeded by their desegregated nature and their reliance on different reporting measures and units of analysis. Yet it is known that the injuries and mental health problems that occur in the wake of family violence have imposed a heavy burden on a broad range of service providers, including women's shelters, schools, hospitals, mental health clinics, police stations, and district attorney's offices. Responses to reports of domestic violence or the endangerment of children, for example, involve time-consuming and costly investigations to determine program eligibility by a broad range of social service programs, including child protective services, children and family resource programs, child welfare, and foster care offices. In cases in which injury has occurred and the victim or caregivers seek medical assistance, a wide range of health resources may be used, such as emergency department and trauma centers; various medical services, including pediatric, obstetric, and gynecological services; mental health services; oral health and nursing facilities; orthopedic, neurological, and radiological treatment programs; and community health centers. One recent study of the cost implications of treating children in a pediatric intensive care unit, for example, reported that children rendered critically ill from abuse differ markedly from other critically ill children in terms of age, severity of injury, mortality, and expenses for acute medical care (Irazuzta et al., 1997). This study further indicated that, despite the resource-intensive nature of their cases, children whose caregivers sought medical treatment for abuse-related injuries in an intensive care unit were at greater risk of death and severe residual morbidity, often because of irreversible brain injury, than the general patient population in the same unit within the time period of the study. The legal system has been profoundly affected by the problem of family violence, especially in handling cases involving decisions about child placement, termination of parental rights, and abuse by intimate partners. Massachusetts state courts, for example, issued almost 100,000 restraining orders during the period 1992-1994; on average, once every 10 minutes in Massachusetts a victim of domestic violence seeks a restraining order against an abusive defendant (Adams, 1994). The documentation and treatment of reports of family violence, especially abuse by intimate partners, presents a significant burden on local police agencies and state and municipal courts, including the criminal, family, and juvenile justice systems (Cochran, 1994). More recently, a broad range of voluntary and quasi-public programs have emerged to deal with the aftermath and prevention of family violence, including battered women's shelter programs, child fatality review teams, children's trust fund organizations, family preservation services, and local and state coordinating councils. The true range of costs associated with the immediate impact of family violence is simply unknown, but conservative estimates would suggest that the costs are quite large. Estimates vary widely depending on the definitions applied,

OCR for page 31
--> the measurement of both extent and service utilization, and assumptions about the cost of service utilization (see Table 2-4). The expenses of treatment programs for child maltreatment have been estimated to cost more than $500 million annually (U.S. General Accounting Office, 1991a); these costs cover only direct services (such as medical treatment, short-term foster care, and specialized education) and do not include mental health or educational services that may be required as a long-term consequence of child abuse or neglect. The General Accounting Office estimate also does not include additional costs associated with juvenile courts, longer-term foster care, drug or alcohol treatment, adult criminal activities, foregone future earnings, and potential welfare dependence, which have been acknowledged as consequences of child maltreatment but not quantified. TABLE 2-4 Estimated Annual Costs of Family Violence Study Costs Included Annual Cost Estimate (for U.S., unless otherwise noted) Straus, 1986 Data related to intrafamily violence $1.7 billion Daro, 1988 Medical costs; Rehabilitation and special education; Foster care; Lost productivity $20 milliona; $7 million; $7.1 billion; $658 million-$1.3 billion Meyer, 1992 Short- and long-term medical treatment and lost productivity $5-$10 billionb Dayaratna, 1992 National health care costs generated as a function of $326.6 million annual health care costs for Pennsylvania Blue Shield $6.5 billionb Zorza, 1994 National costs generated as a function of $506 million annual health care medical costs for New York City $31 billionb Miller et al., 1994 Medical bills; out-of-pocket expenses; property losses; productivity losses at home, school, work; pain, suffering, and lost quality of life $67 billionb a Includes only costs of child maltreatment. b Includes only costs of partner abuse. SOURCE: Committee on the Assessment of Family Violence Interventions, National Research Council and Institute of Medicine, 1998.

OCR for page 31
--> Recent estimates place the annual cost of domestic violence in the United States at between $1.7 billion (Straus, 1986) and $140 billion (Miller et al., 1994). The variation among cost estimates stems from differences in the variables selected to generate cost figures; for example, Miller's estimate includes a high set of indirect costs associated with pain and suffering, which were not included in the Straus estimates. Another major source of difference is the definition used to determine the size of the victimized population. Depending on what prevalence estimates are used, the number of cases for which costs are estimated is affected. And the larger the prevalence rate used, the higher the annual estimate of total cost. One way to narrow the wide range of cost estimates would be to generate a national estimate of the prevalence or incidence of specific types of family violence. The Centers for Disease Control and Prevention, for example, is involved in an effort to define intimate violence in a form that could be used to establish national baseline rates. An interagency task force on child abuse and neglect is conducting a similar effort to define child maltreatment so that common data elements could be established in research studies in this field. The consequences of victimization are another variable that can generate vastly differing total costs. Cost estimates can include mental as well as physical health, long- as well as short-term treatment costs, family or social costs (such as lower productivity, absenteeism, high rates of turnover, and loss of earnings) as well as personal costs, indirect as well as direct costs, and costs of services for perpetrators as well as for victims. A model that includes emergency room costs only will generate a smaller total cost estimate than one that includes the cost of long-term mental health services to victims, which may not be accrued until months or years after an abusive incident. Once it has been determined how many cases of family violence should be counted and what costs will be measured, the source of cost data can affect total cost estimates. Conservative estimates may rely on cost data that are relatively easy to obtain, such as the cost of medical treatment; such estimates may thus be limited to expenses directly associated with service fees. In contrast, comprehensive cost estimates may include indirect costs for which no reliable estimate is available, such as the cost of diminished productivity and the costs incurred by the need to provide volunteer advocacy services. Large differences between total cost figures associated with estimating the impact of the problem of family violence can result from the inclusion or elimination of such indirect costs. The size of indirect cost measurements can be influenced by various estimating factors, such as the use of econometric forecasting techniques, including discount rates and future productivity costs. Applying annual productivity increase rates and discount rates will affect the total cost and result in a more sophisticated, and theoretically more accurate, estimate. The techniques used to estimate loss of productivity, usually human capital or willingness-to-pay approaches, also affect indirect cost measurements. The more comprehensive the model, the greater the cost estimate it yields

OCR for page 31
--> (Table 2-4). One comprehensive model for estimating the annual cost of domestic violence includes direct costs, such as health care costs, social service costs, and criminal justice costs, as well as indirect costs, such as morbidity and mortality costs, which measure lost output when a victim is incapacitated or killed (Rice et al., 1996). Several assumptions are implicit in the model: that direct cost components can be estimated with available data on charges and expenditures for goods and services; that a discount rate of 4 percent and an average annual increase in productivity of 1 percent are appropriate in generating future costs, and that certain other costs of health and social services, such as moral support and advocacy for victims, should be excluded. The costs of these may be significant, but no reliable data exist from which to estimate them. A cost estimate for child maltreatment for the year 1983 was developed by calculating the number of child abuse reports received, what percentage were substantiated, and what percentage actually received various types of services, including foster care (Daro, 1988). It estimated that the immediate cost of hospitalizing abused and neglected children was $20 million annually, rehabilitation and special education cost $7 million annually, and foster care costs were $460 million annually. Additional short-term costs include education, juvenile court and detention costs, $646 million for long-term foster care, and future lost earnings of abused and neglected children of between $658 and $1.3 billion. Extrapolating Daro's costs for 1994, Westman (1995) included estimates for hospitalization, rehabilitation and special education, foster care, social services case management, and court expenses. His cost estimate was between $8.4 and $32.3 billion each year, based on a range of $12,174 to $46,870 per maltreated child per year. Miller et al. (1994) estimate that personal crime costs Americans $105 billion each year. Including pain and suffering, the cost rises to $450 billion. Violent crime accounts for $426 billion of the total. The authors estimate that child abuse costs $67,000 per incident, sexual abuse $99,000, and emotional abuse $27,000 for an average of $60,000 per incident of child abuse. The authors estimated that child neglect costs $9,700 per incident. Thus, the total cost of child abuse and neglect per year is estimated to be $56 billion. Miller et al. (1994) do not include in their estimate the cost for sibling violence, noncriminal violence toward parents, or noncriminal elder abuse. Thus, their cost estimate of $77 billion for child abuse and domestic violence still underestimates the total costs of family violence each year. What is also unknown in reviewing these cost estimates is the extent to which existing expenses associated with health, social services, and legal services could be reduced if effective preventive interventions were in place. Reducing the scope of family violence and mitigating its consequences would have some impact on existing service expenses, most of which are borne by individuals or public agencies, but the size of that impact remains uncertain.