Fixing Only Half the Problem:
Public Policy and Batterer Intervention
by John Hamel, LCSW Copyright, 2004

481 Via Hidalgo #240, Greenbrae, CA 94904 (415) 472-3275
angercounseling@aol.com

When one hears the term “domestic violence,” what comes to mind is usually a woman, eyes blackened from a recent beating, cowering in a corner of her home, arms wrapped protectively around her frightened children. Towering above her, fist clenched tight, the raging husband scowls menacingly. This is the image from television movies-of-the-week, and from posters in mental health clinics and battered women shelters.

Prior to the advent of the shelter movement in the 70's, the plight of these women was given scant consideration. Men who beat their wives often were not arrested, and their victims sometimes were blamed for the abuse. The situation is much different today. Although shelters for abused women remain underfunded, and far too many women continue to be assaulted and killed by their partners, society has clearly begun to take the problem seriously. Assaultive men are being arrested at increasingly greater rates, and are court-mandated in most states to complete specialized batterer treatment groups.

The prevailing view, reflected in the movie-of-the-week image, is that while physical assault rates for women may be comparable to those for men in cases of minor assaults, such as grabbing and pushing, men perpetrate the great majority of severe assaults - what is generally known as “battering.” When serious, violence by women is assumed to be in self defense, “expressive,” or symptomatic of underlying emotional issues. Not so with men, who are assumed to be the initiators in most cases, and whose violence is viewed as “instrumental,” and driven by misogynistic, patriarchal attitudes (Walker, 1979, 1983; Dobash and Dobash, 1979; Brygger and Edleson, 1987) . Clinicians in the mental health field, from county agencies and Child Protective Services to therapists in private practice, routinely refer abusive women to individual psychotherapy, or to support groups for battered women.

Neither does the judicial system, on the whole, view women as perpetrators. Presumed to be victims, women are rarely convicted of domestic violence or mandated to batterer treatment. For example, the number of women arrested for domestic violence in California is close to 20% of the total, but the number mandated to offender groups, including the ones at our own treatment center, is only about 5%. Public policy, in this perspective, does not reflect any anti-male bias, but merely addresses the realities of partner violence. The fact that women, in some localities, are being arrested at increasingly greater rates, is regarded as a disturbing trend (e.g., Das Dasgupta, 2001), one that adds “insult to injury” by further victimizing abused women.

A dissenting view holds that severe assaults by women are more prevalent than is generally believed, that much of partner violence is bi-directional, and that violent men and women have in common similar motives and etiological roots (Cook, 1997; Fontes, 1998.) Dissenters maintain that violence is always harmful, regardless of the perpetrator’s sex. They cite one set of studies showing that violent men and women are equally likely to have witnessed mother hit father as children, as opposed to father hitting mother (Straus, et al., 1990; Sommer, 1994; Langhinrichsen-Rohling, et al., 1995); and another body of research indicating that over 8 million children each year are exposed to violence between their parents (Straus, 1991). Dissenters acknowledge that children are more frightened by father’s assaults, because of men’s ability to cause greater injury, but argue that assaults by both parents have serious, long-term consequences on children, putting them at a 40% greater risk than those from non-violent homes to develop emotional, behavior and academic problems (Holden, et al., 1998; Wolak & Finklehor, 1998; Johnston, 1995). Finally, many who work directly with perpetrators, such as our colleagues in the San Francisco Bay Area (FAVTEA, 2002), are beginning to question the assumption that intervention programs are exclusively - or even mostly - comprised of “batterers.” In sum, the dissenting view holds that current policies have hindered, rather than helped, efforts at reducing partner violence

This paper will show how public policy on batterer intervention is woefully inadequate because it fails to take into account the high rates and seriousness of assaults by women, holds to a simplistic, unfounded causal view, and supports treatment models that ignore the complex etiology of partner violence. We will attempt to shed light on this debate, by providing answers to three fundamental questions:

1. How serious is violence perpetrated by women upon their male partners?
2. To what extent is “patriarchy” a cause of partner violence, compared to other factors?
3. How useful are batterer intervention programs, and do they actually treat the populations we assume they are treating?

Literature Review

Assault Rates and Context

In intimate partner relationships, men are assumed to perpetrate physical assaults far more often than women. A rate of 95% for male-perpetrated violence is widely assumed, cited by therapists, law enforcement and in “fact” sheets distributed by battered women shelters. In reality, this is a preposterous distortion. Martin Fiebert’s (1996) exhaustive annotated bibliography, containing 70 empirical studies and 15 literature reviews, indicates that women assault men as often, or more often, than the reverse; and Archer’s recent meta-analysis (2000) yields similar results. Reviews by Archer (2000) and Steinmetz (1981), indicate that although men are indeed more physically abusive in underdeveloped countries, and in countries with strong patriarchal systems, physical assault rates in liberal, western industrialized nations are comparable between the sexes, just like in the United States.

Survey samples based on crime statistics, such as the National Crime Victimization Surveys, as well as hospitalization records and interviews with battered women, do indicate higher rates for men; but this is due to the restricted samples used (Fontes, 1998; Straus, 1999). National population surveys provide less data about specific populations, but their findings can more readily be generalized. Drawing on a representative sample of 6,000 couples across the United States (Straus, 1990), the National Family Violence Surveys of 1976 and 1985 indicate an overall equal rate of assaults by gender. The National Violence Against Women Survey reveals a ratio of male-perpetrated partner aggression of 1.5:1 over that perpetrated by females (Tjaden, et al., 1998a). However, because of its similarity to crime surveys, which frame their questions in the context of victimization and safety issues, the NVAWS tends to limit the amount of information obtained, underestimating total assault rates while artificially skewing those across gender (Straus, 1999). Men simply refuse to be perceived as victims. Follingstad, et al. (1991) found this to be true even when they acknowledge having been physically assaulted. In the same study, nearly twice as many women as men reported to having been assaulted, but 250% more women said they had perpetrated violence. This is a crucial finding, considering that the NVAWS figures were based solely on victimization rates.

The vast majority of assaults, about 94%, are comparable between the sexes. According to the women interviewed in the NFVS, women perpetrate 52% of minor assaults (push, grab, slap), and 50% of serious assaults (punch, kick, bite, choke, hit with an object, threaten to use a weapon). Approximately two-thirds of the most severe violence, including fatal assaults, are committed by men. Men tend to beat up their partners far more often than the reverse, but women use weapons at equal or greater rates. These findings are supported by Archer (2002) and by crime studies of large urban areas (Mann, 1988). Surveys by the Bureau of Justice Statistics (1992) indicates that 72% of intimate partner homicides are perpetrated by men. However, female-perpetrated homicides may be higher. Farrell (1999) points out that gender rates provided by the Bureau of Justice Statistics are limited to cases where there the assailant, and the cause of death, are known. Women use poison, and other hard-to-detect methods, far more often than do men. They are also far more likely to enlist the services of a third party to carry out the killing. In such cases, the female is not identified as having perpetrated the crime.

But what about context? When research on domestic violence first began to proliferate, coinciding with the growth of battered women shelters in the seventies and early eighties, researchers were at pains to explain the high rates of physical assaults by women. Not wanting to engage in victim blaming, researchers were content to accept studies based on interviews with battered women (Saunders, 1984), which suggested that most female-initiated violence was in self-defense. Even Murray Straus, who carried out the NFVS, was initially willing to accept this view (Straus, 1980). It was only after conducting the second NFVS in 1985 that he began to question this assumption. Using data from the wives, he found that women strike the first blow 53 % of the time and men 42% (Gelles, 1996). The women reported that in the past year, the violence had been bi-directional 50% of the time. In fact, throughout the course of most violent relationships, both parties have, at one time or other, initiated violence. For instance, Langhinrichsen-Rohling’s (1995) study of military men arrested for domestic violence, which utilized in-depth interviews with both partners, determined that 83% of the violence involved mutual assaults. Since the second NFVS, several large-scale studies of both sexes have been conducted, which indicate that women hit in self-defense only 10% to 20% of the time (e.g., Carrado, et al., 1996; Sommer, 1994; Fiebert, 1997). Trying to “get through” to one’s partner and ”getting his/her attention” are the most common reasons men and women give for hitting their partners, offered in half the cases.

Seventeen years after the second NFVS, some researchers persist in ignoring the data, and argue that most intimate partner violence perpetrated by women is indeed defensive. Some limit their research to self-selected samples, such as interviews with battered women, and make unwarranted generalizations; while others engage in questionable research practices to make their point. An example is Das Dasgupta (2001), who cites several studies, including the one by Saunders (1986), and confidently declares: “These studies find that self-defense is the most common reason for women’s use of violence towards their intimate male partners.” However, at least two of the studies he cites - one by Vivian and Langhinrichsen-Rohling (1994) of couples seeking counseling in a New York clinic; the other by Straus (1999) - draw no such conclusions. The first suggests that most partner violence is mutual, but that women experience more physical, and somewhat more psychological, injury than men. With respect to the second, Das Dasgupta appears to be crossing that fine line of research ethics, between taking data out of context, and manipulating it. Citing a number of self-defense studies, including the British study by Carrado, et al. (1996), Straus states the very opposite of what Das Dasgupta reported. Here is what Straus actually said:

In my early research on domestic assaults, it seemed so obvious that women were injured more than men, and that domestic assaults by women were primarily in self defense, that I did not collect data on injury and self defense. I simply asserted it as a self-evident fact...So, when, in the 1985 National Family Violence Survey, I did ask who was the first to hit, I was surprised to find that half of the women respondents reported they had hit first...Several other studies...also found about equal rates of initiation by men and women (p.28).

It is often argued that self-defense accounts for most cases of female-perpetrated intimate partner homicide. In Mann’s study, 60% of the women killers claimed self-defense. Mann, however, had reason to be skeptical. Overall, 58% of the murders were determined to have been premeditated.. Furthermore, 30% of the women killed their partners when they were incapacitated - either drunk, bound or asleep; yet, the majority of this subgroup (60%) also claimed self-defense. Cases of actual self-defense are in fact much lower, although the rates may be proportionately higher for women than for men. In Felson and Messner’s (1998) analysis of 2,000 intimate partner homicides, self-defense, defined as protecting oneself from bodily harm, accounted for 9.6% of female-perpetrated killings, but only .5% of male. An expanded definition, to include previous physical attacks, with or without a self-defense motive, yielded rates of 46.2% and 11.1%. But in Jurik’s (1989) review, only 21% of the women who killed their partners claimed to have experienced prior abuse, or the threat of abuse. Regardless of which studies one accepts, it is clear that most intimate partner murders, committed by either sex, are not in self-defense, according to even the most liberal interpretation of the term.

Battering and Domestic Violence: the Politics of Definition

Beliefs that minimize female violence against males exist largely because public policy on domestic violence is influenced by the women’s shelter movement, which tends to disseminate information supportive of their views; and because of the debatable assumption, accepted by many researchers, that talking about male victims will lessen funding for battered women services. Much of the confusion also relates to how the terms “domestic violence” and “battering” are defined. As we know, men perpetrate the great majority of the latter. The assaults by many of these men, labeled “Cobras” and “Pit Bulls” by Jacobsen and Gottman (1998), are particularly vicious, sometimes deadly. Naturally, they receive the most attention. Not surprisingly, these types have come to represent all violent males in the popular imagination.

Sociopathic batterers can terrorize their mates without necessarily having to beat them up. Shouldn’t a pattern of intense emotional abuse that includes highly controlling behavior and threats of serious violence be included in the definition of “battering,” even though the actual assaults are less than those of the “very serious” kind? Some, however, expand the definition of “control” beyond physical intimidation to include psychological abuse and jealousy. In addition, men are assumed to enjoy greater control because of their presumed dominant positions of power in the household. Thus, the vast majority of men who don’t use physical intimidation but have perpetrated any violence, even minor, tend to be characterized as batterers; but not so women who do the same, because they are less able to physically intimidate, and are assumed to have less power. But women utilize similar control tactics and wield comparable power, in the majority of violent homes, as their partners. What we have, in essence, is guilt-by-association, and guilt-by-assumption, methods of defining a batterer. Thus the arrest policies currently in place, and the tendency to brand all male offenders as batterers - practices that are sexist, demeaning and clinically unsound.

Defined solely on frequency of assaults, battering is perpetrated at equal rates by men and women. But assaults by men, because of their superior strength, result in more physical injuries. An 8:1 ratio for women versus men victims is often quoted, but this is a misleading statistic, based on reports of doctor visits in the NFVS. Men, who are notoriously unwilling to be perceived as weak, downplay ailments of all kinds, and are especially reluctant to admit they have been injured by a woman. According to the NVAWS, 41% of partner assaults on women result in injuries, and for men the figure is 19% - an injury ratio of slightly higher than 2:1. In Langhinrichsen-Rohling’s military study, and Johnston’s (1995) research on divorcing couples, it was determined that the 2:1 ratio only applies at lower levels, and that the discrepancy widens with more serious assaults. This trend was supported by the NVAWS, in which the proportion of injured women needing medical attention was 33% higher than for men. In Archer’s review (2000), however, the total number of women injured by their partners, averaged across the 19 studies analyzed, was 62%, very similar to the NVAWS; but the proportion requiring medical attention was only slightly higher, at 65%. Two conclusions can thus be drawn. First, even at the adjusted 2:1 rate, it is clear that women suffer more injuries than men. The second, which runs directly counter to the claims of feminists, is that a substantial minority of men are also injured, many seriously. “It is therefore not the case,” writes Archer, “that women’s violence towards men severe enough to cause physical injury is negligible or nonexistent.” (p. 665).

The assumption that there is no such thing as a “level playing field” in partner violence, because men are bigger and stronger and able to inflict greater physical damage, is partly true. Men, however, are constrained both by their conscience and by the law, and often “pull their punches”out of fear of seriously harming their partner. Women don’t have to hold back; and, as mentioned above, they make up for their physical limitations by using objects and weapons. They are also more likely to carry out their assaults when their partners are in a vulnerable position (McLeod, 1984; Steinmetz, & Lucca, 1988; Shupe et al., 1987.) As we have seen, 30% of the female-perpetrated spousal murders in Mann’s study were committed when their partners were drunk, bound or asleep.

Because they suffer the greater share of physical injuries, women may in fact suffer a comparably higher rate of psychological symptoms. But it would be a mistake to presume that the psychological impact of domestic violence on men is negligible. Women interviewed in the National Family Violence Surveys reported a higher level of psychological distress, in the form of psychosomatic symptoms, anxiety and depression, compared to men. Except for depression, however, the differences were not statistically significant (Straus, et al., 1990). A study by Vivian and Langhinrichsen-Rohling (1994) of 57 couples seeking counseling for marital violence, 32, or 56%, engaged in mutual violence, with low to moderate rates of victimization on several indices, including frequency and severity of assaults, injuries and psychological distress. In 25 cases, the violence was primarily unilateral, perpetrated by the husband in 15, and by the wives in 10. In both groups, levels of victimization were equally high on all indices, for both husbands and wives. However, the overall levels of psychological distress across all three groups was higher for the wives. Also, perpetrator husbands reported to have experienced hardly any psychological distress, but perpetrator wives did, and to the same extent as their victims. From this the authors conclude that female perpetrators of serious, unilateral violence feel more guilty about their behavior than their male counterparts, and that theirs is a different kind of violence - “expressive” (arising from the heat of an argument), as opposed to “instrumental” (strictly to dominate the other).

These conclusions are probably unwarranted, in light of the fact that men are conditioned to suppress feelings, particularly those that render one vulnerable. Any competent psychotherapist - or frustrated wife - can attest to this. We have previously discussed men’s aversion to being perceived as victims, a phenomenon that has been well documented (e.g., Mihalic and Elliott, 1997). In Archer’s meta-analysis of research utilizing the Conflict Tactics Scales (1999), men were found to underreport their own victimization even when the questions were not framed in the context of a crime. It would not be much of a leap to suggest that men also minimize the impact of those assaults they acknowledge to have been perpetrated against them. In fact, studies by Linda Marshall, drawing from samples of college students and community residents in Texas, indicate that men minimize both the physical, and the emotional harm, of the assaults against them. The impact weights for every one of the 46 items of violence, or threat of violence, on her Violence Against Men Scales (Marshall, 1992a) were lower than the corresponding items on the Violence Against Women Scales (Marshall, 1992b). One could argue that men know there is a lesser likelihood of physical injury, which might cause them to be less distressed about the assault. However, it remains unexplained why the impact weights of emotional harm for being burned, or shot with a gun, should be less for men than for women. Unlike being punched or pushed, these assaults that are not dependent on physical size. The emotional toll from having a pot of scalding water poured over one’s head, or suffering a gunshot wound to the abdomen, ought to be equally devastating to a petite woman or a burly, 300 pound man.

Simonelli & Ingram (1998) examined the effects of physical, verbal and emotional abuse on a sample of college men. Physical and verbal abuse were based on the CTS, and emotional abuse was measured according to a revised, gender-inclusive version of Tolman’s Psychological Maltreatment of Women Inventory, which included subscales for social control, diminishment of self-esteem, jealousy and withdrawal. Effects were based on a 30-item General Health Questionnaire (GHQ). 29% had been the targets of severe physical aggression. Received physical abuse predicted 37% of the variance in depression. Received emotional and verbal abuse were also significantly related to general distress and depression. Emotional abuse, for example, predicted 14% - 33% of the variance in depression.

Johnson (2000-a) has proposed a typology of partner violence consisting of four categories, according to control, actual use of violence, and whether or not the assaults are mutual. “Intimate terrorism,” in his estimation, is perpetrated 97% of the time by men, and accounts for11% of all couple violence. Unfortunately, only women were asked to participate, and they were drawn from a data sample comprised mostly of shelter residents and crime victims. Swan and Snow (2002) attempted their own typology. But again, only women were interviewed. And, like Johnson’s, the questionnaire used to elicit information on control was skewed in favor of typically male tactics.

We propose a simpler, more straightforward scheme for defining intimate partner violence (see table 1). From statistics provided by the NFVS, the NVAWS, Archer’s 2000 meta-analysis and the Department of Justice, roughly 70% of spousal assaults involve lesser violence not leading to injury. This type of violence, which we may designate as high conflict violence, is perpetrated primarily by women. If a battering incident is defined as one leading to physical injuries, we may then distinguish between one type of battering, which leads to minimal or moderate degrees of physical injury, and a second type, which involves severe, sometimes fatal injuries. Men perpetrate approximately two-thirds of the first type, which might be termed common battering, and three-fourths of the second type, or severe battering. In general, the greater the severity of physical assault, the greater the levels of emotional abuse and controlling behaviors, but there is not a perfect correlation. Some domestic violence, therefore, may be considered “battering” regardless of physical injury, when the non-verbal abuse has reached extreme levels. This type of abuse, perpetrated at approximately equal rates by men and women, can also be characterized as emotional battering.

Table 1. Categories of Domestic Violence

Domestic Roles, Patriarchy, and the Politics of Power and Control

In their meta-analytic review of 29 studies on patriarchy, Sugarman and Frankel (1996) measured patriarchal norms along three dimensions: attitudes about violence, attitudes about gender roles, and gender schemas (extent of “masculine” vs. “feminine” traits). The researchers found a significant correlation between attitudes supportive of violence against women and the use of such violence. However, the effects of traditional gender role attitudes (e.g., that a woman shouldn’t work outside the home, let the man make the decisions, etc.) were weak and nonsignificant; and, contrary to expectations, the violent men actually measured lower on dimensions of masculinity - defined, at least in part, as a tendency towards instrumentality versus expressiveness.

There is no evidence that society is more accepting of violence by men than by women. In fact, the opposite is true (Follingstad, et al, 1991; Straus, et al, 1997). Furthermore, women are victimized as much in lesbian relationships, where the patriarchal structures should not exist, than in heterosexual ones (Lie, 1991). Lesbians who batter, Claire Renzetti found (1992), represent all types - “feminine” as well as “butch.” Also, lesbian victims tend to wield more power in the relationship, in terms of income level, job status and education, than perpetrators. This dynamic appears to mimic the often-cited one in heterosexual relationships, where unemployed males are thought to batter their employed, better-educated female partners in an attempt to regain “male privilege.” One could argue that this type of behavior is, in heterosexual relationships, another example of the influence of male patriarchy. But what are we to make of the same dynamic in lesbian relationships - an attempt to regain “female privilege?” It might be more plausible to suggest that certain individuals, regardless of gender or sexual orientation, are simply envious and insecure, and use violence to feel better about themselves. In this view, violence by men is not symptomatic of sexism per se, or some pervasive cultural misogyny. According to author Christina Hoff Sommers (1994), it is rather “a pathology of intimacy, as frequent among gays as among straight people.”

We have seen that partner violence has a greater physical impact on women because of the higher injury rates that they suffer. Some argue that assaults against women are different than assaults against men for other reasons as well. One assumption is that men have more power in intimate relationships. However, results from the NFVS indicate that women wield comparable levels of power in the home, in terms of “who has the final say” in decision-making (Straus, et al., 1990). Also, power can be found in other resources besides status and economic power, including communication skills (Babcock, et al., 1993) and the allocation of love and affection (Teichman & Teichman, 1989). Another assumption is that violence is but one method, or tactic, with which men seek to keep male privilege and maintain domination over their partners; and women, unlike men, are subjected to a “system of victimization,” which includes physical intimidation and threats, sexual abuse (including rape), stalking, and whole range of control tactics that include emotional and economic abuse, and behaviors designed to isolate them from the world outside the domicile. The evidence, however, does not support these conclusions.

According to the NVAWS (Tjaden, et al., 1998-b), 0.26% of men are stalked each year by a current or former intimate, and 0.77% of women, a ratio of 3 women for each man victimized. Unlike the NVAWS, most other studies drawn from community samples do not frame the questions within the context of a crime survey, and they generally indicate much more comparable rates, depending on how “stalking” is defined.. Sptizberg and Rhea (1999) examined a variety of stalking subtypes, collectively known as obsessive relational intrusion (ORI). Results from their sample of college students in Texas revealed a 54% rate of male-perpetrated ORI’s, versus 46% for females. Langhinrichsen-Rohling, et al’s (2000) college survey asked respondents to report on their own ORI behavior, as well as incidents of victimization. There were no overall differences in stalking rates. “As a whole,” the authors write, “these results suggest that unwanted pursuit behaviors may occur in a relatively gender-neutral manner” (p. 86). A major difference between the sexes was that men made more unwanted visits to homes and apartments, whereas women left the greater share of unwanted phone messages. Women were also four times as likely to report having been physically threatened. The review by Davis and Frieze (2000) echoes these findings. Considering that women are both injured and murdered twice as often as men, it is not surprising that they report higher degrees of fear, and seek the vast majority of protective orders in stalking cases.

In the NVAWS (Tjaden, et al., 1998-a), less than 0.1% of the men reported to having been raped the previous year. More than twice as many women (0.2%) said that this had happened to them. As with stalking rates, however, rates for coercive sexual behaviors narrow considerably between the sexes when an expanded definition is employed, and when the interview is not framed within the context of a crime survey. Muehlenhard and Cook’s (1988) college study revealed that men, more often than women, engage in unwanted sexual intercourse, at rates of 63% versus 46%. Being taken advantage of when intoxicated was reported by 30.8% of the men, and 21.0% of the women. 13.4% of the men and 11.5% of the women said they had been verbally coerced. The rates were 5.7% for men subjected to nonviolent coercion (e.g., blocking the door, holding the person down), compared with 5.4% for the women. Coercion involving physical assaults (e.g., slapping, punching) was experienced by 1.4% of the men and 2.7% of the women. A later study by Waldner-Haugrud and Magruder (1995) asked a similar population about a range of coercive tactics. In the previous year, the men had an average of 2.26 incidents perpetrated upon them, and the women 2.86. Persistent touching was common, reported by 51% of males and 70% of females. Men were twice as likely to report blackmail (8.5% versus 4.2%); women reported a higher incidence of manipulative guilt (30.1% versus 22.5%). With respect to physical coercion, the women were twice as likely than men to be restrained or detained, and more threatened with physical force (6.9% to 6.0%). However, three times more men than women had weapons used against them (4.5% verus 1.4%).

The high rates of stalking by women, and the degree to which they use coercion to obtain sex is not surprising, unless one regards them as docile Madonnas - weak, sexless creatures devoid of passion and desire. But, as the literature on anger and general aggression indicates (Frodi, et al., 1977; Eagly & Steffen, 1986; Averil, 1983), women are every bit as angry as men and, in many ways, just as aggressive. There is no reason to think that women should restrict their aggressive impulses to physical assaults, or that they would be less inclined than men, on the whole, to impose their will on their partner. What about other forms of coercion? A list of power and control tactics, formulated from interviews with 200 battered women, frequently appears at seminars and in literature provided by women’s shelters. Supporters of a patriarchal explanation for violence assume that these tactics are used only by men. The list, known as the “Power and Control Wheel,” lists the following tactics: using (physical) intimidation; using emotional abuse; using isolation; minimizing, denying and blaming children; using children; using male privilege; using economic abuse; and using coercion and threats. It has found widespread acceptance because it reflects the experiences of many abused women, and because men’s batterer treatment providers have identified those behaviors in some of their clients. However, our experience with violent men and women suggests that such tactics, with the notable exception of physical intimidation, are in fact utilized by both sexes.

Although not much attention is paid to the plight of battered men, victimization accounts indicate the use of a variety of control tactics by women (Shupe, et al, 1987; Steinmetz & Lucca, 1988; Cook, 1997). Patricia Pearson (1997) recounts the case of Steve Easton, director of the Easton Alliance in Toronto, one of the few shelters in the world that serves both male and female victims:

His partner, an exotically beautiful woman from upstate New York, had seen her mother abuse her father. Ursula approached her lover the same way. She called him “cock sucker” and “prick.” She chose what clothes he could wear to work, arguing that certain ties or shirts would attract his female colleagues. If he disregarded her choices, he came come to find his wardrobe burned to ashes. She insisted...that he couldn’t go out with his friends. If he did, she locked him out of the house for the night. He wasn’t permitted to read the Toronto Sun, because the tabloid carries daily photos of a woman in a bikini - the “Sunshine Girl” - and that was evidence that he lusted after other women. When she started a fight, she would follow him from room to room in their house, keeping up all night: “I’m not finished with you!” Exhausted, he came late to work too many times and got fired (p. 124).

Stories such as these are helpful, but there is a dearth of quantitative data in the literature on control tactics. As discussed in an earlier section, Michael Johnson (2000-a) has developed a typology of partner violence that includes such behaviors. A limitation of his initial sample was that it used only female respondents, heavily drawn from crime victim samples. In a second study, Johnson (2000-b) analyzed data from the 1998 NVAWS. Again, he found evidence of two broad types of violence: common couple violence, and what he calls “intimate partner terrorism,” characterized by more serious assaults, higher levels of control and psychological abuse, and greater overall impact on victims. This is an important distinction, which ought to help clinicians develop more precise guidelines for assessment and treatment. Unfortunately, Johnson again chose to focus exclusively on female reports.

(Among the many thick volumes of published data from the NVASW, there is not one word on how the men answered the seven questions regarding control tactics used against them. In personal communication with this author, the survey director, Patricia Tjaden, said she disregarded the men’s responses because she wished to examine correlations between control behaviors and stalking, and men’s stalking victimization was “minimal” compared to the women. This is a puzzling explanation. A 25% rate is hardly minimal. Moreover, the apparent objective of the NVAWS was to demonstrate the systematic victimization of women. A plethora of data was disseminated on physical assaults, injury rates, sexual victimization and stalking, all of it compared along gender lines. Why, then, so little on control, and no gender comparisons? One is tempted to conclude that the data was ignored because the results might have run contrary to the project’s thesis, potentially embarrassing to the authors. One wonders, too, about Johnson’s motives; certainly, he could have retrieved the information from the raw data, as he did for the women’s responses, but chose not to do so.)

In the study by Swan and Snow (2002), also mentioned in a previous section, 108 women who had perpetrated intimate partner violence in the past year were questioned about their abuse experiences. Although the women admitted to having been more physically assaultive, and more emotionally abusive, than their partners, the authors nonetheless determined that only 12% of these women were the aggressors! This dubious conclusion was possible only because the authors accorded equal weight to isolation-type control tactics as it did to the acts of physical violence. As reported by the women, the men used such tactics 75% more often than they did. Amazingly, the women’s use of violence and emotional abuse was not regarded as “coercive.”

The authors acknowledge that the instrument used to gauge coercive control, the Psychological Maltreatment of Women Inventory (Tolman, 1999), was designed to measure men’s behaviors (e.g., “get upset if housework was not done when you wanted,” and “demand partner stay home and take care of the children”), and recommend that “a new scale particular to women’s violence is needed” (Swann & Snow, 2002, p. 312). Despite this caution, they nevertheless maintain that the women were victims of abuse at a rate three times that of men. Women who assaulted and emotionally abused their mates were deemed “violent resisters” to male abuse. Self-defense is assumed after the fact; yet, the women weren’t asked about it during the interview process. Essentially, the study provides an excuse for women’s violence - more elaborate and couched in the terminology of science, but ultimately as disingenuous as those of men who excuse their violence with such claims as, “she kept nagging me and wouldn’t shut up.” The authors point out that, although these women were selected because they had been violent, the men were also quite violent, and they decree: “Women’s violent behavior can only be understood when placed in the context of their male partner’s violence against them” (p.310). In a large sense, they are correct. Recall the study by Langhinrichsen-Rohling (1995), in which 83% of the men arrested for battering were found to have been involved in mutually-violent relationships. With their research, Swan and Snow merely confirm the fact that most partner violence is mutual.

When men are asked about controlling behaviors used against them, they, too, report high rates of both emotional abuse and isolation tactics. In the first NFVS (Straus, et al., 1980), women were found to have engaged in a higher degree of yelling and swearing than men, and to have smashed or broken things more often. In a survey of college students using the revised version of the conflict tactics scale, or CTS-2 (Straus, 2001), the men reported to have used psychological aggression on their partner an average of 15.1 times in the year. The women said they had done so an average of 16.0. Mirroring the emotional abuse scale in Johnson’s questionnaire, psychological aggression was defined as verbal abuse (swearing, calling partner “fat or ugly,” accusing partner of being a “lousy lover”), threats of physical harm, and symbolic abuse (destroy something belonging to partner). The other items in Johnson’s questionnaire had to do with possessive behaviors, similar to the items in Swan and Snow’s study (e.g., “jealous or suspicious of partner’s friends” and “monitor partner’s time, make partner account for whereabouts.”) Johnson may have chosen to ignore the male data, but a survey of men in a domestic violence treatment diversion program (Shupe, et al, 1987) yielded the following results:

Two-thirds reported that their mates regularly went through their pockets and billfolds, not so much looking for money as for telephone numbers of possible girlfriends.

Three fourths said the women closely clocked them while they were outside the home.

One-third of the women tried to censor the men’s telephone calls and other communications with family and friends.

Two-thirds said the women withdrew sex as punishment when they resisted being monitored or misbehaved somehow (p. 59).

“Withdrawing sex” is not an item in Johnson’s research. Neither is remaining helpless or chronically depressed. They are not to be found in the Swan and Snow research, nor in the “Power and Control Wheel.” The kind of behavior described by Steve Easton, of being harassed all night by a girlfriend who “wasn’t finished” with him, is also absent in the various questionnaires. These tactics are typically utilized by women against men, rather than the other way around. We know about the psychological and emotional toll that women suffer when victimized by a male “intimate partner terrorist.” But consider the impact of having vicious rumors circulated about you in the community, or in your workplace, or being the recipient of false domestic violence charges. A counselor working with male offenders in Austin, Texas gives the following account of one man who has experienced the maternal form of intimate partner terrorism (Shupe et al., 1987):

Jerry is a 34-year old construction contractor who recently went to court for the tenth time on an assault charge brought against him by his ex-wife. On each charge he has pleaded not guilty, and each time his wife has failed to show up at the trial, therefore the charges against him have been dropped. But each time nevertheless he has had to hire a lawyer, taken time off from his job, and spent many hours trying to explain to his current girlfriend that he has not been violent against his ex-wife...He no longer knows what to do. He has been to the police department and has been told that there is nothing they can do...

He has been divorced from his wife for about a year and a half. It was a bitter divorce with a child custody case that he won. His ex-wife told him repeatedly that she would make life miserable for him and eventually would get the children from him. Jerry once told the counselor when speaking about her, “You want to see violence? I’ll show you violence!” He showed a recent cut on his forearm. “This is what she did the last time she got angry with me.” When the counselor asked why he did not file charges against her, Jerry flatly said that he was a man and that he would not ever call the police on a woman. His lawyer told him that there is little he can do except file a civil suit against her since she most recently has taken to harassing him at work. She also has promised to get him fired from his job and many times has shown up at job sites screaming accusations at him and telling his co-workers how he has beaten her (pp. 54 - 55).

Studies on emotional abuse and control that examine the behavior of both men and women are rare. In addition to the NFVS research, and the research by Straus mentioned above, this author found only three others. One, by Felson & Messner (2000) analyzed the National Crime Victim Survey for the years 1992 through 1994. Of the 2,597 cases involving a single assailant on a single victim, 22.7% of the antagonists were comprised of an intimate couple. Physical assaults were preceded by a verbal threat most often in cases of male-on-female abuse (54.6%), followed by female-on-female lesbian abuse (33.5%), female-on-male (27%) and, lastly, male-on-male gay abuse (18.2%).

Kasian & Painter (1992), using a gender-inclusive version of the Psychological Maltreatment of Women Inventory, investigated emotional abuse/controlling behavior in a population of 1, 625 college students. The different categories of abuse, and the most statistically significant items in each, are listed below:

Isolation and emotional control

My partner tried to keep me from seeing or talking to my family
My partner tried to turn my family and friends against me
My partner tried to keep me from doing things to help myself
My partner interfered in my relationship with family members
My partner threatened to have an affair with someone else

Diminishment of self-esteem

My partner treated me like I was stupid
My partner treated me like I was an inferior
My partner ordered me around
My partner treated me like his/her personal servant
My partner insulted or shamed me in front of others

Jealousy

My partner was jealous and suspicious of my friends
My partner was jealous of other men/women
My partner monitored my time and made me account for my whereabouts
My partner accused me of seeing another man/woman

Verbal abuse

My partner swore at me
My partner yelled and screamed at me
My partner called me names

Withdrawal

My partner sulked and refused to talk about a problem
My partner withheld affection from me
My partner gave me the silent treatment

The men reported a higher frequency of such abuse upon them than did the women , for the following categories: isolation and emotional control, jealousy, verbal abuse and withdrawal. Men and women were subjected to equal amounts of diminishment of self-esteem.

The third gender-inclusive study is a recent one from England, by Graham-Kevan and Archer (2002), who challenge the assumption that control is exclusively a feature of men’s violence. Unlike the Kasian and Painter study, which utilized only college students, theirs drew on a diverse population. Data was collected from samples of shelter women (n = 43), male prisoners (n = 108) and students of both sexes (n =113). Each respondent was asked about physical violence and control tactics, as used by themselves and by their partners. Control tactics were measured on the 24-item Controlling Behavior Scale, divided into five sub-categories from the “Power and Control Wheel”: “using economic abuse,” “using coercion or threats,” “using intimidation,” “using emotional abuse,” and “using isolation.” Results showed a significant, positive relationship between control tactics and physical aggression, for both sexes - and the strength of the correlation was the same for men and women. Intimidation was the most significant predictor of aggression for men, and the best predictor of female aggression, as well. Although the levels of control and physical aggression were higher overall among the shelter women’s male partners (“intimate terrorists”), compared to the college students (“common couple violence”),use of control was strongly associated with aggression in both groups. This finding runs counter to Johnson’s theory, which characterizes “common couple violence” as non-escalating aggression, with expressive anger and loss of control.

If control tactics can co-exist with violence at all levels, from minor to very serious, what about the function of the violent acts themselves? Recall the brief discussion, in the first part of this paper, about “instrumental” and “expressive” violence. Determining what is done purely to dominate is difficult to tease out from what is done reflexively in the heat of a conflict, because of poor impulse control - or, for that matter, due to relationship dynamics. Small scale studies are contradictory. In a study of dating couples (Follingstad, et al., 1991), 250% more women than men reported that controlling their partner was a motive for their violence. In sample of college students, on the other hand (Makepeace, 1986), three times more men than women reported intimidation as a motive for their assaults, while three times more women than men said their assaults were intended to cause harm. In a large scale, national British study (Carrado, et al, 1996), an equal amount of men and women, about 25%, stated that they had assaulted their partner “to make him/her do what I wanted.” But twice as many of both sexes said they were either trying to “get through” to the other or else were retaliating for something the other had said or threatened to do.

The assertion that a number of abusive men harbor strong patriarchal beliefs is not in dispute. What is debatable, however, is the extent to which such beliefs, at an individual level, are causally related to patriarchal structures and norms at the macro-level of society. “Patriarchy,” writes Dutton (1994), “does not elicit violence against women in any direct fashion. Rather, it may provide the values and attitudes that personality-disordered men can exploit to justify their abuse of women. This distinction is an important one: It explains why the majority of men remain nonviolent and how they differ in at least one essential and nontautological aspect from violent men (p. 176).”

Furthermore, if patriarchal violence is defined as behavior used to enforce traditional sex roles, then such behavior can be in the service of either male or female interests. Balancing the wish by some men to maintain “head of the household” status is the fact that a woman’s identity is more deeply anchored in family than a man’s, so there is a greater need for her to defend her interests and reputation. (Straus, 1999). Whereas some men may use isolation tactics to keep “their woman” in the home, cooking and caring for the children, some women may ally with those children, act helpless, and make financial demands on the husband with the intent of reinforcing his role as provider. Women are far less likely to use physical aggression outside, but in the home might feel a need to use violence when negotiation and talking don’t work.

And yet, women wield comparable decision-making power in the household (Straus, et al. 1990); and we have seen that their use of coercive control tactics is as extensive as men, and that members of both sexes may be subject to a “system of victimization.” We have also seen that violent men do not harbor more traditional sex-role beliefs than men who are nonviolent. Therefore, power and control is probably grounded less in “patriarchy” than basic human desire. Desperate attempts to communicate, or to inflict retribution, are the more frequent precipitants to violence. One could, of course, argue that trying to “get through” to one’s partner also contains an element of power and control, but if it does, it applies equally to both sexes. The line between “expressive” and “instrumental” may not be so clear; screaming at someone may be due to poor impulse control, and thus “expressive,” but additionally serves the instrumental goal of getting heard. In most cases, couples fight because their wants and needs are incompatible and because they don’t have the will, or the ability, to negotiate mutually acceptable solutions. Thus, for treatment to be effective, it cannot be based solely on a feminist analysis. As Dutton (1994) aptly puts it, “If patriarchy “causes” violence, how can we hold men individually responsible for their violence?” (P. 177). And of course, the same question can be asked about women.

Relationship Dynamics

Throughout this discussion, we have focused on the various characteristics and causes of partner violence as they pertain to individuals. We have seen how women engage in partner assaults at rates comparable to men, and that much of it is mutual rather than unidirectional. It is widely assumed that female recipients of male assaults are passive victims, suffering from “battered woman’s syndrome ” (Walker, 1979, 1983), and “traumatically bonded” to their perpetrators due to the power discrepancy of their relationships and the intermittent nature of the abuse (Dutton & Painter, 1993). This is certainly the case with some women victims, but not with others.

Despite its enduring popularity, there are serious flaws in Walker’s conceptualization of the battered woman’s syndrome, including her misapplication of learned helplessness theory, as well as the scant empirical support for her three-stage theory of partner violence (Faigman, 1986). Elaborating on Walker’s three-stage theory, Deschner (1984) proposed a seven-stage cycle that more closely accounts for the realities of partner violence, including negative interchanges, and it allows for more complex relationship dynamics. A number of laboratory-based research projects using direct observation of subject couples have yielded a wealth of additional information about these dynamics, and their role in partner violence. Negative communication by moderately abusive husbands, according to Margolin (1988), are reciprocated with negative communication by the wives. Similar results were found by Cordova, et al (1993), in a study focused on severe assaults:

Women were every bit as inclined toward negative reciprocity as the men, even in the DV condition...There is virtually no evidence from these interaction sequences that battered women placate their husbands or attempt in any way to neutralize their aversive behavior. The behavior of DV wives in this sample does not suggest passivity, docility, or surrender. Rather, the women are continuing the conflict engagement, even though they have histories of being subjected to physical abuse. (P. 563)

In a study by Burman, et al. (1992), wives of abusive husbands responded to both negative-offensive statements (e.g., criticism, insult) and negative-defensive statements (disagreement, “yes/but”) with negative-offensive statements of their own. The husbands, however, typically responded to negative-offensive statements with negative-defensive ones. These findings indicate that abusive couples engage attack/defend cycles, characterized by a tendency for men to become more despairing and to withdraw as the conflict escalates. Babcock, et al. (1993) found that violent husbands and their wives were equally likely to make demands, or to withdraw in response to their partner’s demands, thus ensuring continuation of the conflict, further resentments and power struggles. And in yet another study (Jacobsen, et al., 1994), husbands were observed to be more domineering and defensive, but the wives were rated as more angry, belligerent and contemptuous. All the male subjects had perpetrated either several acts of serious violence, or at least one act of very serious violence, in the past year. In spite of the self-selective nature of the sample, the authors later asserted that approximately half the wives would have qualified for inclusion into a batterer treatment group themselves.

There were additional gender differences worth noting in the Jacobsen study previously discussed. As would be expected from a study on male batterers, the wives expressed a much higher degree of fear than their partners, as well as more sadness. There was also an important difference in how the violence escalated. Nothing a wife did, including withdraw from the conflict, could effectively stop the husband’s violence once it began, whereas the wive’s violence escalated only in response to the husband’s violence or emotional abuse and would desist once he withdrew. A re-analysis of the first NFVS by Straus (1980) showed that when violence escalates to mutual assaults, men may be more likely than their wives to meet minor violence with serious violence. In personal communication with this author, Professor Straus points out that the data is not conclusive, because the time order of reported assaults was never determined, and because the results have never been replicated. Even if the findings are correct, the same study acknowledges that men and women engage in approximately equal rates of both minor and serious violence. Therefore, if men are indeed more likely to respond to minor assaults with serious assaults, how do we account for the remaining serious assaults by women? This author speculates, and Professor Straus concurs, that in a large number of abusive relationships, men probably respond to major assaults by their wives with minor violence of their own (e.g., grabbing her to prevent a punch). It may be that some men, whose personalities and childhood of origin experiences predispose them to violence, find assaults threatening and escalate in response, whereas others, not so threatened, are able to restrain themselves.

However these findings are interpreted, it remains that relationship dynamics are an inherent component of partner violence. And according to Burman et al., the dynamics of violent couples bear far more resemblance to other high-conflict, nonviolent couples than to happily married ones. Earlier research by Telch and Lindquist (1984) had come to the same conclusions:

Several findings suggest that couples in battering relationships are similar to other couples in treatment. They have low self-esteem; they have greater difficulty with communication; and they experience greater dissatisfaction and disagreement in their marriages than satisfied spouses. Therefore they are likely to need and respond to interventions such as communications training, negotiating, and other skills typically used by marital therapists. Emphasis to V couples that the content of their problems is no different from most marrried couples seems valuable for their self-esteem and for rapport...Our violent couples were found to exhibit more passive and aggressive behaviors and less assertive behaviors than the comparison couples. This patterns suggests that the V couples have difficulty in expressing wants, needs, and feelings directly, and that instead they operate in a passive-aggressive manner. The violent outburst presumably results in part from the inadequacy of this pattern in terms of getting one’s needs and wants met (p. 247).

Beyond Patriarchy

Table 2 summarizes the research on the etiology of partner violence. The most significant factors for men, according to the NFVS (Straus, et al., 1990), Hotaling and Sugarman’s (1986) comprehensive analysis, and the Canadian study by Reena Sommer (1994), are the following: having witnessed violence in one’s family of origin, perpetrating violence against one’s children, working-class occupational status, low income and alcohol abuse and, to some extent, low educational level. Personality factors, according to Holtzworth-Munroe and Stuart (1994), include genetic predisposition to aggression, attachment disturbance, impulsivity, poor social skills, negative attitudes towards women and positive attitudes towards violence. O’Leary’s (1988) social learning model adds stress and relationship dissatisfaction to the mix. It takes into account relationship dynamics, and allows for a variety of abusive interactions, including unilateral battering by women or mutual abuse.

Table 2

ETIOLOGY OF PARTNER VIOLENCE

Unfortunately, research has almost exclusively focused on male violence. Few studies have been conducted on women perpetrators, despite the fact that female assaults, as we have seen, account for approximately 50% of all intimate partner violence and at least one-third of injuries. Other than alcohol abuse, the few studies we do have (e.g., Kalichman, 1988; O’Leary, 1988; Shupe, 1987; Johnston and Campbell, 1993; Coleman, 1994; Sommer,1994), have found far more similarities between violence by men and women than differences. It appears that the most violent individuals, regardless of sex, share in common borderline personality traits, and other characteristics of severe psychopatholgy. The lower-level, and far more common types of violence, involve less severe pathology, and are influenced by stress, impulse control and attitudinal factors and relationship dynamics, including communication style. The two major gender differences in partner violence – i.e., greater rates of very severe violence by men, and higher rates of physical injuries suffered by women - are a function of size and strength. Men and are less frequently injured not because women are less aggressive, but because they are usually physically weaker. Pearson (1997) sums it up this way:

On the whole, men do indeed have a more powerful left hook. The problem is that the dynamic of domestic violence is not analogous to two differently weighted boxers in a ring. There are relational strategies and psychological issues at work in an intimate relationship that negate the fact of physical strength. At the heart of the matter lies human will. Which partner - by dint of temperament, personality, life history - has the will to harm the other? (p. 117).

Batterer Intervention and Public Policy

Public attitudes about partner violence influence policy at all levels, and no doubt play a role in the small number of women arrested and sent to batterer treatment - from state laws regulating arrest procedures and batterer intervention, to how counties obtain funding and set up certification criteria for such treatment, to the officers who make the arrests and the district attorneys who decide on who gets prosecuted. These attitudes are reflected in county-wide police criteria for determining who is the primary aggressor in a domestic violence incident. Some, including “who is afraid,” “ability,” and “strength of parties,” encourage male arrests (ACAD, 2000). Revised versions of the procedures (Contra Costa, 2001), have deleted “strength of parties,” but “fear of physical injury” and “history of violence” are retained. Men, however, are reluctant to express fear. And regardless of what the written guidelines stipulate, men’s superior strength almost certainly prejudices officers when assessing risk for abuse. Unless the officer can conduct a thorough, accurate, psychosocial history on the scene, he/she is likely to make the arrest based on the potential for the man to cause greater harm, even if the man is a dedicated pacifist. Although often called for, dual arrests are discouraged under California law (California, 2001). The operating principle within many law enforcement circles seems to be, “when in doubt, arrest the man.”

A study of 3,300 domestic violence cases in Edmunton, Canada (Remington, 2002) yields some disturbing findings. When only the woman was injured, charges were filed in 91% of cases; but when only the man was injured, charges were filed at the much lesser rate of 60%. One might wonder if such discrepancies were due to the women suffering more severe injuries. However, in incidents involving minor injuries to women, male assailants were charged 88% of the time; whereas, when men suffered similar injuries, the female perpetrators were charged in only 72% of cases. The data shows evidence of a pervasive gender bias by prosecutors in that city.

Responsibility for the low rates of female arrests must not, however, be placed entirely at the feet of law enforcement. There is another problem, a significant and rather daunting one, with no immediate solution. As discussed in a previous section, men under-report assaults against them and are loathe to be perceived as victims. Recall the example of Jerry, the construction worker, who had been involved in a bitter custody dispute with his abusive ex-wife. The case of Ken is equally troubling (Shupe, et al., 1987):

Ken was a 28-year old man who appeared in court on an assault charge brought against him by his former live-in girlfriend. He pleaded not guilty and flatly denied ever having been violent with her. He said he decided to move out of this relationship because she had an uncontrolled drinking problem, became violent whenever she drank to excess, and refused to seek any kind of help.

Since leaving he had to move twice because she came to his apartment and, if he refused to let her in, would yell threats, break windows, and scream until neighbors called the police. About a month ago she came to his new apartment and talked a new roommate of his into letting her into the apartment while he was sleeping. She came into his room and stabbed at his groin with a pair of scissors, puncturing his scrotum. He had to be hospitalized after being taken to a hospital emergency room. Since that time he has had all four of his car tires slashed. Yet Ken refused to file any kind of charges against her, or take out peace bonds or any protective orders, because she is a woman (p. 55.)

Such bravado may also explain much of the resistance among therapists and shelter workers to the topic of female violence. It is a basic human trait to generalize from one’s personal experiences. Therapists rarely hear about female-perpetrated assaults in their clinical practice, and shelter workers may view all partner violence through the particular prism of that environment. Unless the individual personally conducted the survey or meta-analysis, they would be understandably skeptical of findings from, say, the NFVS or from Archer, if they spent each day offering support to abused women, many of them victims of serious, life-threatening violence. Still, no matter how understandable and sincere such predispositions may be, public policy must not rest exclusively upon them.

After being charged, offenders who are found guilty or (more likely) plea bargain their case, are often mandated to batterer treatment, in lieu of a jail sentence. In California, the same-sex group format is not simply a treatment option, but the only legally sanctioned one. Individual therapy, crucial in helping individuals to heal from childhood trauma and serious psychopathology, is expressly prohibited, as it is in many states. Couples work, which may be the treatment of choice in most cases (e.g., Geffner, 1989; Neidig & Friedman, 1984), is likewise prohibited, based on the erroneous assumption that all violence by men is unilateral, that women are always passive victims, and that couples work would only serve to legitimize the violence perpetrated upon them. Same-sex group is the preferred format in 90% of states that have domestic violence standards, and in 81% of them couples work is discouraged or prohibited (Austin, J. & Dankwort, J., 1998). In California, couples work is prohibited in all cases, regardless of any mediating variables, such as severity of the behavior, level of pathology, motivation, the extent to which the violence was mutual, or the willingness of the partner to participate.

Considering that the effectiveness of group has received only lukewarm empirical support (National Research Council, 1998), the law’s restrictions are indeed unfortunate. Guerney, et al. (1987) stress the importance of conjoint sessions in helping couples overcome the isolation associated with battering relationships. The Victim’s Information Bureau of Suffolk (VIBS), a battered woman’s center in New York State, had assigned its perpetrator clients - mostly men arrested on domestic violence charges - to traditional group treatment, until it became evident to the counselors that many of the victimized women would never leave the relationship (Geller and Wasserstrom, 1984). The center then switched over to a process-based, systems-oriented, marital therapy model. The experiment was an enormous success. From 1976 until 1980, 250 couples went through the program. The majority stayed together, and reported increased satisfaction with their relationship. More importantly, the physical abuse ceased for every couple who completed their 2-year commitment to treatment.

The Family Preservation Project (Geffner, et al, 1989) operated a similar model with great success, treating court-mandated male batterers and their wives at the East Texas Crisis Center, a battered women’s shelter in Tyler, Texas. As with the VIBS in New York, the program was initiated after it was found that many of the women refused to leave their partners, as the staff recommended, and sought assistance in working things out. Counselors in the men’s treatment program had been concerned about the limitations of working only with perpetrators, who routinely reported continued problems at home, even as they made efforts to change. The wives sometimes resisted the new behaviors their partners had learned (e.g., misinterpreting time-outs as evidence of abandonment, blocking the door to keep them from leaving, giving themselves permission to retaliate for past abuse), and would consciously or unconsciously sabotage their partner’s progress in other ways. Many of these problems were overcome when the couples began to be treated conjointly and introduced to the same materials. Also, working with the couple together reduced much of the mistrust that is present when the individuals go to separate counseling.

In the majority of states, “patriarchy” is regarded as the most important factor causing, and helping to maintain, violence by men against women, and standards for treatment are modeled on this erroneous assumption (e.g., Florida, 2002.) The most well-known batterer intervention program, known as the “Duluth” model, is used by a third of all programs (National Research Council, 1998). Although California law is written in a gender-neutral fashion, and does not specifically refer to “patriarchy” per se, this author’s experience with law enforcement and community domestic violence organizations is that this is indeed the preferred theoretical model, which many treatment programs have gladly adopted, and which have been incorporated into the certification standards of many counties. Santa Clara, for instance, proscribes a number of possibly useful interventions, including anger management, “fair fighting” techniques, approaches that use family systems models, or psychodynamic ones that link violence to unconscious processes or early childhood trauma (Santa Clara, 1997).

Furthermore, the probation departments who certify and supervise batterer programs in California depend on certain statewide agencies, such as the Office of Criminal Justice Planning, for funding. Some of this funding comes from the Violence Against Women Act. Local batterer treatment programs are thus required to sign operational agreements stipulating that they and the probation departments “work together toward the mutual goal of eliminating incidents of violence against women and children throughout the county” (Contra Costa, 2002). The omission of male victims is striking; one can only wonder how funding agreements such as these influence treatment. Finally, none of the standards in existence, including those in California, address the specific needs of gay and lesbian populations (Dankwort, J. & Austin, J., 1999.) From the perspective of a licensed therapist, committed to fashioning treatment in line with client needs, the laws and policies on domestic violence have been a clinical straightjacket.

In intimate partner relationships, assaults by women are clearly a major problem. However, as previously mentioned, women comprise only a small fraction of individuals referred to batterer groups. This might make more sense if these programs served only the most severe offenders, 72% of whom are men. The assumption is that the majority of male offenders do fit that definition. Studies of battered women in shelter settings do seem to support this proposition. However, there is almost no information in the literature on the wider spectrum of men arrested and sent to batterer treatment. Are all of these men severe wife beaters? This author’s experience providing treatment over the past 10 years, as well as those of treatment provider colleagues suggests otherwise. The “dirty secret” is that many, if not most of the men in our treatment programs, have engaged in lesser forms of violence, are involved in mutually abusive relationships with their partners, and are not batterers at all, certainly not the widely-accepted, cartoonish, movie-of-the-week type. A recent study by Apsler, et al. (2002), conducted in a Boston suburb whose population closely matches overall U.S. demographics, suggests that these impressions may be correct. Over a one year period, 95 female victims of partner violence who came to the attention of the police were questioned about their experiences. 48% of these women said that they were “not at all afraid” or only “slightly afraid” of their abuser. 61% said that future episodes of violence were either “not at all likely” or only “slightly likely.” These results, which indicate that many of the incidents are isolated, also bring into question Walker’s theory about a repetitive “cycle of violence.” The authors conclude:

Our results argue for a careful examination of the police response to domestic violence incidents. Perhaps greater efforts should be directed at tailoring the police response to the nature of the domestic violence incidents. It may be unreasonable to expect a universal strategy, such as mandatory arrest, to be effective when applied to fundamentally different types of domestic violence incidents. As we become more successful at classifying types of domestic violence victims, we will become better equipped at suggesting appropriate police responses (p. 453).

The Study

Method

In late winter and spring of 2002, written invitations were sent to the 12 batterer treatment programs certified in Contra Costa County, California, to participate in this study. Contra Costa County is a large, densely populated suburban area, located about 20 miles east of San Francisco. Six of the programs responded. For purposes of comparison, data from a program in Placer County, a rural area in the Sierra Nevada foothills, were also included.

Program supervisors were asked to select a few representative groups, and to determine where in the four categories described in table 1 each of their clients best fit, based on the offense they were charged with and assault history during the past year. They were also asked to indicate the number of “high conflict-no violence” participants who were abusive (e.g., broke things, made threats of bodily harm or engaged in stalking), versus those whose behavior was restricted to verbal aggression or simple disagreement (nonabusive).

Only men, and those who had been convicted of or plead guilty to a domestic violence crime, were included. Each supervisor was asked to base their decisions on the information gathered during the intake/assessment process, probation records, and what had been revealed in group.

Hypotheses

It was hypothesized that a majority of men mandated to attend a batterer intervention group do not fit the profile of a “batterer,” as described by the categories outlined in table 1, and that this holds for both the “severe” and “common” types. It was also hypothesized that a majority of participants included in this study would fit the “high conflict - violent” category.

Results

Both hypotheses were supported. Tables 3 and 4 indicate that only 20.9% of individuals currently enrolled in court-certified batterer treatment fit the profile of a “batterer.” Fully 79% have perpetrated either minor physical violence, symbolic violence, or none at all.

Although the survey targeted programs certified in Contra Costa County, two of the programs (A and B) provided data from groups physically located in adjacent Alameda County. Findings were based on a sample of 139 clients from 7 programs, with 17 groups conducted by 9 facilitators in several locations. Both Oakland and Richmond have large African-American populations, while Pleasant Hill, Walnut Creek and San Ramon are predominantly white and middle-class. Berkeley is home to an ethnically and socioeconomically diverse population.

Results were remarkably consistent across programs, with only a slight amount of variance. They indicate that only about 21% of men currently enrolled in local batterer groups have engaged in any recent pattern of battering. Only 4 perpetrators had exhibited severe battering, and 25 the more common type. The majority of men enrolled in 52-week groups fell in the “high conflict - violent” category. Individuals in this category, the reader will recall, perpetrate lower-level acts of violence, such as pushing and grabbing, cause slight or no injuries, and engage in only low to moderate levels of power and control. In an astounding 36 cases, 26% of the total number, there had not been any physical assaults at all.

The typical male client in a batterer intervention program is therefore one who has engaged in minor assaults, perpetrated in the general population at equal, or higher rates by women.

Table 3
CATEGORY TOTALS ACCORDING TO REFERRING PROGRAM

Table 4
CATEGORIES BY PERCENT OF TOTAL

Discussion

A possible limitation of the study is that the sample may not have accurately represented the actual perpetrator population in the counties sampled. Six of the agencies contacted, representing half the total number of agencies, chose not to participate, including one (STAND) which sponsors the largest men’s treatment program in the county. The sample was a self-selected one. Simply by chance, the perpetrator populations at those agencies might have been comprised of a higher number of batterers than the six programs who did take part. This is unlikely, but certainly possible.

We don’t know why any particular agency declined to be involved. Despite the clear criteria utilized, the process of assigning perpetrators to categories was to a large extent a subjective one. It is not unreasonable to wonder if the non-participants might have made different assignment decisions, based on their experience, political orientation and training. Finally, assuming that this is indeed a representative sample, we need to know more about batterer populations in other localities, inside and outside the state of California.

There is a need for further research, on arrest procedures and the characteristics of batterer treatment programs, and how each may be affected by public policy. We need to know more about the factors that contribute to gender disparities. To what extent are they related to pro-female bias, and how much to the unwillingness of men to seek help? Should we be only concerned about assaults that lead to serious injury, or intervene when lesser assaults are perpetrated? Are there better ways to assess partner violence, so the appropriate intervention can be made? What alternative treatment models, such as the VIBS and the Family Preservation Project, should we know about? The answers may never be certain, because the questions asked depend largely on the values, goals and philosophies of those conducting the research.

Our study, however, along with what we know about partner violence from the literature, makes it quite apparent that public policy regarding partner violence, at least in California, ought to be revisited. That is not to say that most of the men arrested don’t require some kind of treatment. All partner violence is harmful. It is preferable to intervene before the violence escalates to more severe, potentially fatal levels of battering. But if this is going to be our approach, then it must be a gender-inclusive one. Women can be as angry, vindictive, controlling, manipulative, verbally abusive and physically violent as men, and they cause a substantial portion of physical injuries. Far more women should be referred to anger management or batterer treatment, rather than be relegated to victim groups or prescribed standard psychotherapy.

Current policies are not only be biased against men, but remarkably short-sighted and inefficient as well. Unless they voluntarily seek help, untreated spouses are left to continue their assaults, adding stress to the family system, jeopardizing the perpetrator’s treatment and putting everyone - men, women, and children - at greater risk. We understand, and fully support, the “zero tolerance” policies currently in place. No one wants to return to the decades previous to the 70's, when domestic violence was not considered a problem worthy of serious consideration. However, for interventions to be effective, treatment must be based on an understanding of the dynamics of partner violence, including the factors that cause and maintain it, rather than on considerations of what is “politically correct.” Ignoring the problem of female partner violence is both insulting and dangerous, infantalizing women rather than empowering them to change; and it contributes to the perpetuation of violence in the family system, putting children at risk for developing emotional, behavioral and academic problems, and increasing the odds that it will be passed down to subsequent generations. As a result, we are only fixing half the problem

Accordingly, we offer the following recommendations:

1. Funding, public education, outreach and treatment efforts should address the broader problem of partner and family violence, rather than focus exclusively on male battering and its consequences.
2. At all levels of intervention, let’s hold every perpetrator accountable. Individuals perpetrate even minor partner assaults can benefit from anger management/domestic violence counseling, in addition to psychotherapy. Violent individuals all have issues, but excusing abusive behavior for psychological reasons is insulting to both sexes
3. 3. Providing shelter and support for assault victims ought to remain our number one priority. Since women are more often than men the victims of severe battering and incur a higher percentage of injuries, they will need the greater share of shelter resources.
4. We need to reach out to male victims. Out of fear of appearing “wimpy,” men rarely seek assistance from shelters. Public education campaigns ought to include male victims. Every one working with families in distress should make it a point to assess for possible female assaults on men. If men aren’t asked, they often won’t tell.
5. Distinctions between “high conflict,” “domestic violence” and the two categories of battering need to be understood. Too often, the term “battering” is used in reference to men who have engaged in minor violence but don’t in any way fit the profile of a controlling, instrumental batterer, and many of these individuals wind up in “batterer” treatment.
6. Domestic violence laws, when poorly written, can actually undermine efforts to reduce domestic violence. In many states, including California, changes need to be made both in arrest procedures, and in the program standards for batterer treatment:
   • (a) Arrests and treatment should be based on severity of assaults and rehabilitation potential, without gender bias.
   • (b) Current police procedures should be revisited, and more dual arrests ought to be made. If we are going to arrest men who engage in minor violence, let’s also arrest women who do the same. Better still, we need to provide alternatives to arrest, especially in cases with no clear perpetrators and victims. One such alternative might be a sort of citation system, which would mandate both parties to a more thorough assessment, and which would provide for restraining orders and other procedures to ensure victim safety prior to any charges being filed.
   • (c) Not everyone needs to complete a 52-week program, as required in California. Some would benefit from individual counseling. In addition, a great many individuals referred to batterer groups are in relationships where there has been an equal, or greater amount of assaults by the spouse. Couples therapy - particularly of the structured, multi-couples type - may be preferable for the majority of cases

There is no excuse for domestic violence , the bumper stickers proclaim. That, for sure, is a point on which we can all agree.

References

ACAD (2000, October), Manual, Second Annual Training on Domestic Violence.

Apsler, R., et al. (2002), “Fear and Expectations: Differences Among Female Victims of Domestic Violence Who Come to the Attention of the Police.” Violence and Victims, 17 (4), pp. 445-453.

Archer, J. (1999), “Assessment of the Reliability of the Conflict Tactics Scale.” Journal of Interpersonal Violence, 14 (12).

Archer, J. (2000), “Sex Differences in Aggression Between Heterosexual Partners: A Meta-Analytic Review.” Psychological Bulletin, 126 (5).

Archer, J. (2002), “Sex Differences in Physically Aggressive Acts Between Heterosexual Partners: A Meta- Analytic Review.” Aggression and Violent Behavior, 7, 313-351.

Austin, J. & Dankwort, J. (1998), “A Review of Standards for Batterer Intervention Programs.” Violence Against Women Online Resources.

Averil, J.R. (1983), “Studies on Anger and Aggression: Implications for Theories of Emotion.” American Psychologist, (Nov.), pp. 1145-1160.

Babcock, J., et al. (1993), “Power and Violence: The Relation Between Communication Patterns, Power Discrepancies, and Domestic Violence.” Journal of Consulting and Clinical Psychology, 61 (1).

Brygger, M., & Edleson, J. (1987), “The Domestic Abuse Project: A Multisystems Intervention in Woman Battering.” Journal of Interpersonal Violence, 2 (3).

Bureau of Justice Statistics (1994), “Selected Findings: Violence Between Intimates.” NCJ-149259. U.S. Dpt. of Justice.

Burman, B., et al. (1992), “Observed Patterns of Conflict in Violent, Nonviolent, and Nondistressed Couples.” Behavioral Assessment, 14, pp. 15-37.

Carrado, M., et al. (1996), “Aggression in British Heterosexual Relationships: A Descriptive Analysis.” Aggressive Behavior, 22, pp. 401-415.

Coleman, V. (1994), “Lesbian Battering: The Relationship Between Personality and the Perpetration of Violence.” Violence and Victims, 9 (2).

Contra Costa (2001, April), “Contra Costa County Police Chief’s Association Policy Guideline, Domestic Violence.”

Contra Costa (2002), “Operational Agreement (Between Probation Department and Certified Batterer Programs”).

Cook, P. (1997), Abused Men: The Hidden Side of Domestic Violence.

Cordova, J., et al. (1993), “Negative Reciprocity and Communication in Couples With a Violent Husband.” Journal of Abnormal Psychology, 102 (4).

Dankwort, J. & Austin, J. (1999), “Standards for Batterer Programs: A Review and Analysis.” Journal of Interpersonal Violence, 14 (2).

Das Dasgupta, S. (2001) , “Towards An Understanding of Women’s Use of Non-Lethal Violence in Intimate Heterosexual Relationships.” Violence Against Women Online Resources

Davis, K. & Frieze, I. (2000), “Research on Stalking: What Do We Know and Where Do We Go?” Violence and Victims, 15 (4).

Deschner, J.P. (1984), The Hitting Habit: Anger Control for Battering Couples.

Dobash, R.E & Dobash, R.P. (1979), Violence Against Wives: A Case Against the Patriarchy.

Dutton, D.G. & Painter, S. (1993), “Emotional Attachments in Abusive Relationships: A Test of Traumatic Bonding Theory.” Violence and Victims, 8 (4).

Dutton, D.G. (1994), “Patriarchy and Wife Assault: The Ecological Fallacy.” Violence and Victims, 9 (2).

Eagly, A.H., & Steffen, V.J. (1986), “Gender And Aggressive Behavior: A Meta-Analytic Review of the Social Psychological Literature.” Psychological Bulletin, 100(3).

Faigman, D. (1986), “The Battered Woman Syndrome and Self-Defense: A Legal and Empirical Dissent.” Virginia Law Review, 72, pp. 619-647.

Farrell, W. (1999), Women Can’t Hear What Men Don’t Say.

Felson, R., & Messner, S. (1998), “Disentangling the Effects of Gender and Intimacy on Victim Precipitation in Homicide.” Criminology, 36 (2).

Felson, R. & Messner, S. (2000). The control motive in intimate partner violence. Social Psychology Quarterly, 63 (1), 86-94.

Fiebert, M. (1996), “References Examining Assaults by Women on Their Spouses/Partners: An Annotated Bibliography.” (Unpublished manuscript, California State University, Long Beach).

Fiebert, et al. (1997), “College Women Who Initiate Assaults on Their Male Partners and the Reasons Offered For Such Behavior.” Psychological Reports, 80.

Florida (2002), “Certification Procedures and Minimum Standards for Assessors and Batterers’ Intervention Programs.” Florida Department of Corrections.

Follingstad, D. (1991), et al., “Sex Differences in Motivations and Effects in Dating Relationships.” Family Relations, 40.

Fontes, D. (1998), “Violent Touch: Breaking Through the Stereotype.” Available at Safe4all.org.

Frodi, A., et al. (1977), “Are Women Always Less Aggressive Than Men?” Psychological Bulletin, 84.

Geffner, R., et al. (1989), “A Psychoeducational, Conjoint Therapy Approach to Reducing Family Violence.” In Caesar, P. & Hamberger, L., eds, Treating Men Who Batter: Theory, Practice and Programs.

Gelles, et al. (1996), Current Controversies on Family Violence.

Geller, J. & Wasserstrom, J. (1984), “Conjoint Therapy for the Treatment of Domestic Violence.” In Roberts, ed., Battered Women and Their Families: Intervention Strategies and Treatment Programs.

Graham-Kevan, N., & Archer, J. (2002), “Does Controlling Behavior Predict Aggression and Violence To Partners?” Unpublished manuscript, available online at: ngraham-kevan @uclan.ac.uk

Guerney, et al. (1987), “Wife Battering: A Theoretical Construct and Case Report.” The American Journal of Family Therapy, 15 (1).

Holtzworth-Munroe, A., & Stuart, G. (1994), “Typologies of Male Batterers.” Psychological Bulletin, 116 (3).

Hotaling, G.T. & Sugarman, D.B. (1986), “An Analysis of Risk Makers In Husband to Wife Violence.” Violence and Victims, 1.

Jacobsen, N. & Gottman, J. (1998), When Men Batter Women.

Jacobsen, N., et al. (1994), “Affect, Verbal Content, and Psychophysiology in the Arguments of Couples With a Violent Husband.” Journal of Consulting and Clinical Psychology, 62 (5).

Johnson, M. (2000a, November), “Conflict and Control: Symmetry and Asymmetry in Domestic Violence.” Presented at the National Institute of Justice Gender Symmetry Workshop - see: mpj @ psu.edu

Johnson, M. (2000b, July), “The Differential Effects of Patriarchal Terrorism and Common Couple Violence.” Tenth Annual Conference on Personal Relationships, Brisbane, Australia.

Johnston, J. (1995), “Domestic Violence and Parent-Child Relationships in Families Disputing Custody.” Australian Journal of Family Law.

Johnston, J. & Campbell, L. (1993), “A Clinical Typology of Interparental Violence in disputed-Custody Divorces.” American Journal of Orthopsychiatry, 63 (2).

Jurik, N.C., (1989), “Women Who Kill and the Reasonable Man: The Legal Issues Surrounding Female-Perpetrated Homicide.” 41 st Annual Meeting of the American Society of Criminology.

Kalichman, S.C. (1988), “MMPI Profiles of Women and Men Convicted of Domestic Homicide.” Journal of Clinical Psychology, 44 (6).

Kasian, M. & Painter, S. (1992). Frequency and severity of psychological abuse in a dating population. Journal of Interpersonal Violence, 7 (3), 350-364.

Langhinrichsen-Rohling, J., et al. (1995), “Violent Marriages: Gender Differences in Levels of Current Violence and Past Abuse.” Journal of Family Violence, 10 (2).

Langhinrichsen-Rohling, J., et al. (2000), “Breaking Up is Hard To Do: Unwanted Pursuit Behaviors Following the Dissolution of a Romantic Relationship.” Violence and Victims, 15 (1).

Lie, et al. (1991), “Lesbians in Currently Aggressive Relationships,” Violence and Victims, 6.

Makepeace, J. (1986), “Gender Differences in Courtship Violence Victimization.” Family Relations, 35, pp. 383-388.

Mann, C. (1988), “Getting Even? Women Who Kill in Domestic Encounters.” Justice Quarterly, 5, pp.33-51.

Margolin, G., et al. (1988), “Affective Responses to Conflictual Discussions in Violent and Nonviolent Couples.” Journal of Consul