Although men are more likely than women to be victims of violent crimes—61 per 1,000 for men, 42.6 per 1,000 for women (Bastian, 1995)—patterns of victimization differ. Women are far more likely than men to be victimized by an intimate partner (Kilpatrick et al., 1992; Bachman, 1994; Bachman and Saltzman, 1995). In fact, about three-quarters of all lone-offender violence against women in 1993 was perpetrated by someone known to the woman, compared with one-half of lone-offender violence against men (Bachman and Saltzman, 1995). It is important to note that attacks by intimates are more dangerous to women than attacks by strangers: 52 percent of the women victimized by an intimate sustain injuries, compared with 20 percent of those victimized by a stranger (Bachman and Saltzman, 1995). Women are also significantly more likely to be killed by an intimate than are men. In 1993, 29 percent of female homicide victims were killed by their husbands, ex-husbands, or boyfriends; only 3 percent of male homicide victims were killed by their wives, ex-wives, or girlfriends (Federal Bureau of Investigation, 1993).1
Women are more likely to be victimized by male offenders than by female offenders; about three-quarters of violent crimes against women are committed by males (Bachman, 1994). In one urban emergency room, violence was the most common cause of injury to women between the ages of 15 and 44 and the second most common cause of injury for all women (Grisso et al., 1991). Finally, women are far more likely than men to be sexually assaulted. The National Crime Victimization Survey (NCVS) found women were 10 times more likely to be raped or sexually assaulted than were men (Bastian, 1995). The annual rate of rape is estimated to be 7.1 per 1,000 adult women, and 13 percent of all women will experience forcible rape sometime during their lives (Kilpatrick et al., 1994).
The exact dimensions of violence against women are frequently disputed, yet even conservative estimates indicate that millions of American women experience violent victimization. The fear of violence, in particular the fear of rape, affects many more, if not most, women (Gordon and Riger, 1989). A few researchers have even suggested that learning to cope with the threat of violent victimization is a normative developmental task for females in the United States (Gilfus, 1995).
In spite of the attention that has been paid to violence against women in recent years, the research endeavor is relatively young, and much remains unknown. There really is no one field focused on violence against women per se. For example, studies on rape and sexual assault are distinct from those on intimate partner violence, which is distinct from the nascent study of stalking. And all this research is separate from that on violence in general. Many of the studies in this newly emerging field of research on violence against women are at an early stage of scientific rigor. The methodological weaknesses in the research on battering and rape have been discussed at length in other documents (Rosenbaum, 1988; Gelles, 1990; Koss, 1992, 1993; Rosenfeld, 1992; Smith, 1994). Definitions differ from study to study, making comparisons
difficult. Much of the research on both victims and perpetrators is based on clinical samples, samples of convenience, or other nonrandomized samples, so one cannot draw general conclusions. Sample sizes are often quite small. Only recently have sophisticated statistical analyses been used. Yet in spite of all the shortcomings, a lot has been learned about the extent of violence against women, about perpetrators of violence, and about the effects on victims.
What Is Violence Against Women?
The term violence against women has been used to describe a wide range of acts, including murder, rape and sexual assault, physical assault, emotional abuse, battering, stalking, prostitution, genital mutilation, sexual harassment, and pornography. There is little consensus in the still evolving field on exactly how to define violence against women. The major contention concerns whether to strictly define the word ''violence" or to think of the phrase "violence against women" more broadly as aggressive behaviors that adversely and disproportionately affect women.
Researchers in such fields as sociology and criminology tend to prefer definitions that narrowly define violence, definitions that can be operationalized. For example, Gelles and Straus (1979) defined violence as "any act carried out with the intention of, or perceived intention of, causing physical pain or injury to another person." Similarly, the National Research Council (NRC) report Understanding and Preventing Violence (Reiss and Roth, 1993) limited its definition to "behavior by persons against persons that intentionally threatens, attempts, or actually inflicts physical harm." The 1993 NRC study deliberately excluded behavior that inflicts harm unintentionally, while the Gelles and Straus definition includes behaviors that may be unintentional but are perceived by the victim to be intentional. The 1993 NRC study also specifically excluded from its definition of violence such events as verbal abuse, harassment, or humiliation, in which
psychological trauma is the sole harm to the victim. However, in its consideration of family violence and sexual assault, the report did include the psychological consequences of threatened physical injury.
In contrast to those definitions, researchers in such fields as psychology, mental health, and social work frequently consider "violence" to cover a wider range of behaviors. The Committee on Family Violence of the National Institute of Mental Health (1992) included in its definition of violence "acts that are physically and emotionally harmful or that carry the potential to cause physical harm … [and] may also include sexual coercion or assaults, physical intimidation, threats to kill or to harm, restraint of normal activities or freedom, and denial of access to resources." The Task Force on Male Violence Against Women of the American Psychological Association defined violence as ''physical, visual, verbal, or sexual acts that are experienced by a woman or a girl as a threat, invasion, or assault and that have the effect of hurting her or degrading her and/or taking away her ability to control contact (intimate or otherwise) with another individual" (Koss et al., 1994). Those who argue for these broader definitions suggest they more accurately represent the experiences of victims, who often say they find verbal and psychological abuse more harmful than actual physical abuse (Walker, 1979; Follingstad et al., 1990; Herman, 1995).
In the field of intimate partner violence or battering, the problem of violence against women is frequently characterized as one of coercive control that is maintained by tactics such as physical violence, psychological abuse, sexual violence, and denial of resources. The concern is with the array of behaviors that are used to dominate women. Physical violence need not be used often to be effective: "In fact, abusers may regret resorting to violence, but may perceive themselves as 'driven to it' when their other methods of enforcing subordination are insufficient" (Herman, 1995:2). In the field of rape, fear is a key element; it is an overriding concern for many women (Warr, 1985; Gordon and Riger, 1989; Klod-
awsky and Lundy, 1994). Even though women are less frequently the victims of violent crime than men, women fear crime more (Federal Bureau of Investigation, 1991) and this fear appears to be largely based on their fear of rape (Riger et al., 1981). Many feminist theorists contend that this fear of rape serves to intimidate and control all women (e.g., Griffin, 1971; Brownmiller, 1975; Dworkin, 1991).
Although research would benefit from more unified definitions, the panel understands the difficulty of reaching agreement on definitional issues in light of the many complex behaviors that are involved. The panel held lengthy discussions on defining violence against women, focused on the key issue of whether psychological abuse should be included. The panel concluded that it could not resolve a question that is so open among researchers and that a global definition was not necessary for carrying out the task of reviewing what is known and recommending needed research (see below). Thus, the panel agreed that this study would be primarily a review of the literature on intimate partner violence (battering), rape, and sexual assault. The study does not include violence that occurs in conjunction with other crimes, such as robbery, burglary, or car theft. Nor does it include prostitution, sexual harassment, or issues such as genital mutilation, dowry murders, and trafficking in women that are more relevant internationally than in the United States.
Whether one uses a narrow definition confined to physical and sexual violence or one accepts a broader definition of violence against women, definitional debates also surround each of the individual components. For example, how does one define rape or sexual assault? Should all physical aggression or use of force be considered violent? What constitutes psychological abuse? These questions affect both the research that is done and how much it can be generalized.
Rape and Sexual Assault
Although all definitions of rape, sexual assault, and re-
lated terms include the notion of nonconsensual sexual behavior, the definitions used by researchers have varied along several dimensions. These include the behaviors specified, the criteria for nonconsent, the individuals involved, and who decides whether rape or sexual assault has occurred (Muehlenhard et al., 1992; Koss, 1993).
Many data sources and some researchers rely on legal definitions of rape, but those definitions differ from state to state and change over time. In common law, rape was traditionally defined as "carnal knowledge [penile-vaginal penetration only] of a female forcibly and against her will" (Bienen, 1980:174). The FBI's Uniform Crime Report (1993) still uses this narrow definition of rape even though most states have reformed their rape laws during the past 20 years. There have been three common reforms:
- broadening the definition to include sexual penetration of any type, including vaginal, anal, or oral penetration, whether by penis, fingers, or objects;
- focusing on the offender's behavior rather than the victim's resistance; and
- restricting the use of the victim's prior sexual conduct as evidence.
Many states have also removed the marital exemption from their rape laws. Some states and the U.S. Code (18 U.S.C. § 2241-2245) have replaced the term "rape" with terms such as "sexual assault," "sexual battery," or "sexual abuse'' (Epstein and Langenbahn, 1994). Many laws now have a series of graded offenses defined by the presence or absence of aggravating conditions, making sexual assault laws similar to other assault laws. For example, the U.S. Code uses the categories aggravated sexual abuse when someone "knowingly causes another person to engage in a sexual act by using force against that other person, or by threatening or placing that other person in fear that any person will be subjected to death, serious bodily injury, or kidnapping" or by knowingly causing
another person to become incapable of giving consent by rendering them unconscious or administering intoxicants. Sexual abuse involves lesser threats or engaging in sexual acts with a person who cannot give consent.
The definition of rape or sexual assault used in a research study has an effect on who is counted as a rape victim. The type of screening questions, the use of the word rape versus the use of behavioral descriptions, and other considerations all affect the research results (Koss et al., 1994). Higher rates of rape and sexual assault are found when behavioral descriptions and multiple questions are used than when surveys ask directly about rape or sexual assault. Women may not label experiences that meet the legal definition of rape or sexual assault as such, particularly if the perpetrator was an intimate partner or an acquaintance. The use of behavioral descriptions in studies assures that what is being measured are experiences rather than an individual's conceptions of the words rape or sexual assault.
In this report, rape means forced or coerced penetration—vaginal, anal, or oral; "sexual assault" means other forced or coerced sexual acts not involving penetration; and "sexual violence" includes both rape and sexual assault.
Although defining physical violence would seem to be more clear-cut, there are disagreements both over definitions and measurement. As noted above, some researchers include only acts that were intended to cause physical harm or injury (Reiss and Roth, 1993); others argue that intentionality may be difficult to ascertain, and therefore physical violence should also include acts that are perceived as having the intention of producing physical harm or injury (Gelles and Straus, 1979). Akin to intentionality is the consideration of the context of the act. For example, should an action taken in self-defense be considered violent? Should an act be considered violent only if an injury occurs, or is the potential for
injury sufficient? Some definitions of physical violence, following legal models of assault, include threats of physical harm; others consider that threats fall under verbal or psychological abuse (Straus, 1990a). There is disagreement about whether behaviors such as slapping a spouse should be equated with more severe acts such as kicking or using a weapon. How violence is defined and measured influences the rate of violence found in a study: all else being equal, the broader the definition, the higher the level of violence reported (Smith, 1994).
Physical violence is most commonly measured by the Conflict Tactic Scales (Straus, 1979, 1990b) or some modification of it. Such scales ask about the occurrence of various representative behaviors. For example, the Conflict Tactic Scales list nine physical violence items:
- threw something at you;
- pushed, grabbed, or shoved you;
- slapped you;
- kicked, bit, or hit you with a fist;
- hit or tried to hit you with something;
- beat you up;
- choked you;
- threatened you with a knife or gun; and
- used a knife or fired a gun.
The last six behaviors in this list are considered to be "severe" physical violence.
In this report, "physical violence" refers to behaviors that threaten, attempt, or actually inflict physical harm. The behaviors listed in the Conflict Tactic Scales, while not all inclusive, typify the type of behaviors meant by physical violence. In this report, "severe" violence refers to the type of behaviors typified by the severe violence items on the scales.
Psychological abuse (also refered to as psychological maltreatment or emotional abuse) has received less research attention than physical or sexual violence, and hence there have been fewer attempts to define it. At a minimum, psychological abuse refers to psychological acts that cause psychological harm (McGee and Wolfe, 1991). It has been argued that separating physical and psychological conditions "overly simplifies the topic and denies reality" (Hart and Brassard, 1991:63): physically violent acts can have psychological consequences and psychological acts can have physical consequences. The difficulty of separating physical violence and psychological abuse is exemplified by the treatment of threats of physical violence, with researchers split over whether to classify such threats as physical violence or psychological abuse. As with physical violence, there is debate about intentionality, that is, must the offender intend harm for an act to be considered abuse? Deciphering the intention of a psychological act may be even more difficult than for a physical act, and so intention is generally not included in defining psychological abuse.
On the basis of descriptions of psychological abuse as reported by battered women, Follingstad et al. (1990) described the following categories of behavior as psychological abuse:
- verbal attacks such as ridicule, verbal harassment, and name calling, designed to make the woman believe she is not worthwhile in order to keep her under the control of the abuser;
- isolation that separates a woman from her social support networks or denies her access to finances and other resources, thus limiting her independence;
- extreme jealousy or possessiveness, such as excessive monitoring of her behavior, repeated accusations of infidelity, and controlling with whom she has contact;
- verbal threats of abuse, harm, or torture directed at the woman herself or at her family, children, or friends;
- repeated threats of abandonment, divorce, or of initiating an affair if the woman does not comply with the abuser's wishes; and
- damage or destruction of the woman's personal property.
Similar to measurements of physical violence, inventories or scales of representative behaviors are used to measure psychological abuse. The Conflict Tactics Scales subscale on verbal aggression (Straus and Gelles, 1990) measures some aspects of psychological abuse: items include "insulted or swore at you," "did or said something to spite you," "threatened to hit or throw something at you," and ''threw or smashed or hit or kicked something." Other measures that have undergone validity testing are the Psychological Maltreatment of Women Inventory, which consists of 58 behavioral items (Tolman, 1988) and the Abusive Behavior Inventory, which includes items on both physical and psychological acts (Shepard and Campbell, 1992).
Interviews with battered women have detailed clear-cut examples of extreme psychological abuse occurring between and in conjunction with physically violent episodes. Psychological abuse frequently occurs with physical violence (Walker, 1979; Browne, 1987; Follingstad et al., 1990; Hart and Brassard, 1991), and research has repeatedly shown a strong association between psychological abuse and physical and sexual violence (e.g., O'Leary and Curley, 1986; Margolin et al., 1988; Sabourin et al., 1993). Some battered women describe psychological abuse—particularly ridicule—as constituting the most paintful abuse they experienced (Martin, 1976; Walker, 1979, 1984; Follingstad et al., 1990). It has been suggested that ridicule may undermine a woman's self-worth, making her less able to cope with both physical violence and psychological abuse (Follingstad et al., 1990). Studies of child abuse have similarly shown that psychological maltreatment is present in most cases of physical abuse, and it predicts detrimental outcomes for children while severity of physical
abuse does not (Claussen and Crittenden, 1991; Hart and Brassard, 1991).
In this report, "psychological abuse" refers to the types of behaviors described by Follingstad et al. (1990) and listed above, with the exception of threats of physical violence, which this report considers under physical violence. There is no separate section of the report devoted to psychological abuse because it has received very little study in and of itself. Rather, it is considered to be part of the pattern of behavior of serious physical violence, psychological abuse, and sometimes sexual violence, between intimate partners that has been well described (e.g., Martin, 1976; Dobash and Dobash, 1979; Walker, 1979; Browne, 1987). This pattern of behavior has been referred to in many terms, including domestic violence, spouse abuse, battering, and wife beating. "Wife beating" and "spouse abuse" imply married couples, although all intimate relationships—cohabiting, dating, and lesbian and gay couples—are frequently meant to be included under these terms. "Domestic violence," although usually referring to violence between intimate partners, is sometimes used to mean all forms of family violence, including child abuse, spouse abuse, sibling abuse, and elder abuse. These conflicting and overlapping terms and their uses are confusing in the study of violence against women.
In this report, "intimate partner violence" and "battering" are used synonymously to refer to the pattern of violent and abusive behaviors by intimate partners, that is, spouses, ex-spouses, boyfriends and girlfriends, and ex-boyfriends and ex-girlfriend.2 The term batterer is used to mean the perpetrator of intimate partner violence, and battered woman, the victim.
In research studies, dating couples are sometimes considered as intimate partners and sometimes as acquaintances. "Acquaintance" generally refers to someone known to the victim but neither related nor an intimate. Particularly in crime data, it is not always clear what acquaintance means; it may include dating couples. Hence, date rape and dating
violence are sometimes included in crime data as violence by nonintimate acquaintances.
Battered women who have left their batterers have described being stalked by the batterer (e.g., Walker, 1979). This behavior includes following and threatening the woman, repeated harassing phone calls, threatening her family, and breaking into her living quarters. Anecdotal evidence suggests that some batterers go to extraordinary lengths to track down their victims and that women who are stalked by expartners may be at high risk of being killed. Although descriptive information about stalking is available, few data exist.
The acknowledgment of stalking as a crime is a fairly recent phenomenon. California passed the first antistalking law in 1990 (Sohn, 1994); today, 48 states and the District of Columbia have passed antistalking statutes (Boychuk, 1994). Most state statutes define stalking as willful, malicious, and repeated following and harassing of another person. Many statutes include in the definition the intent to place the victim in reasonable fear of sexual battery, bodily injury, or death.
The Panel's Charge And Scope
In the Violence Against Women Act of 1994 (Title IV of P.L. 103-322, the Violent Crime Control and Law Enforcement Act of 1994), Congress directed the National Research Council to develop a research agenda on violence against women (Chapter 9, § 40291):
The Attorney General shall request the National Academy of Sciences, through its National Research Council, to enter into a contract to develop a research agenda to increase the understanding and control of violence against women, including rape and domestic violence. In furtherance of the
contract, the National Academy shall convene a panel of nationally recognized experts on violence against women, in the fields of law, medicine, criminal justice, and direct services to victims and experts on domestic violence in diverse, ethnic, social, and language minority communities and the social sciences. In setting the agenda, the Academy shall focus primarily on preventive, educative, social, and legal strategies, including addressing the needs of underserved populations.
In convening the Panel on Research on Violence Against Women, the National Research Council specifically charged the panel with the following tasks:
- synthesize the relevant research literature and develop a framework for clarifying what is known about the nature and scope of violence against women, including rape and domestic violence;
- supplement the research review with lessons learned by field professionals and service providers, including providers of services to ethnic, social, and language minorities; and
- identify promising areas of research to improve knowledge of the scope of the problem, and implementation and evaluation of preventive, educative, social, and legal interventions for dealing with violence against women.
In carrying out its charge, the panel limited its consideration to violence against women aged 12 and older. Child abuse and neglect and child sexual abuse were outside the purview of this panel and are covered by the report Understanding Child Abuse and Neglect (National Research Council, 1993), with a thorough research agenda.
The age of 12 was selected for several reasons. First, the types of violence to which teenage females are exposed are often more similar to violence directed at adult women than that directed at children. Second, sex offenders who prey on children seem to be quite different from those who target adolescent and adult women (Quinsey, 1984; Prentky, 1990).
Third, surveys on violence, such as The National Crime Victims Survey (NCVS), often include victims beginning at age 12. In addition, the highest rates of rape and sexual assault are found among women aged 12 to 24 years (Bachman and Saltzman, 1995): females in their teens and 20s are those most likely to be dating, and, therefore, subject to dating violence.
The panel's main task was to lay out a research agenda to improve understanding of violence and controlling that violence in the context of women's lives. This entailed reviewing the literature on intimate partner violence, rape, sexual assault, and stalking. The panel concentrated on studies published in peer-reviewed journals within the past 10 years, although very well-known or unique studies that were published earlier are also reviewed. The panel relied both on computerized literature searches, the expertise of various panel members, and monitoring a number of journals devoted to issues of violence. More than 300 journal articles and dozens of books were reviewed, many of which are cited in this report. The panel supplemented its literature review by holding a workshop of researchers and practitioners (see Appendix B).
The panel's review and analysis is divided into three topics: nature and scope, causes and consequences, and preventive and treatment interventions. Chapter 2 describes what the research shows about the nature and scope of violence against women. Chapter 3 discusses possible causes of violence against women and the consequences of violence to women and society. Chapter 4 examines preventive and treatment intervention efforts. Lastly, Chapter 5 discusses issues of research infrastructure and science policy on violence against women. Recommendations for research are discussed at the end of each chapter.