Australian Domestic & Family Violence CLEARINGHOUSE · 2016-10-24 · 2 Australian Domestic &...

ISSUES PAPER 7

IntroductionDeveloping a response to men who use violence andabuse in their intimate relationships2 is one of themost controversial issues facing the field. Debatetends to centre most strongly around the issue oftreatment/education programs for perpetrators, oftenwith people adopting polarised positions either ‘for’or ‘against’ the development of such programs. Manyof these debates can be understood as being about atwhich level – the individual, the institutional, or thesocial – efforts to end violence against women andchildren, are best directed. Are efforts best directedat the men who use violence? At the institutions,such as the criminal justice system and mentalhealth services, which have not always respondedappropriately to the crime of violence and theimpacts of violence on women, or to developing asociety that does not tolerate violence and thatpromotes gender equity?

Debates also centre on the possible ‘unintendedconsequences’ that may arise from efforts directed atany one of these levels of intervention as theyreverberate through the other levels (Laing, 2001).Does working with individual men suggest that the

1

Australian Domestic & Family Violence

Issues Paper 7 2002

CLEARINGHOUSE

problem lies solely with one identified group of‘deviant’ individuals and obscure the role of socialstructures and attitudes which perpetuate violenceagainst women (Ashcroft, 2000)? Do perpetratorprograms take scarce resources away from servicesfor women and children? Do they lull women into afalse sense of hope that the men who abuse themcan change, and thus work against the core goal ofenhancing women’s safety? (Feder and Dugan, 2002)Early research by Gondolf and Fisher (1988) withwomen in refuges in the US, for example, found thattheir partner’s participation in perpetratorcounselling was the most influential factor inwomen’s decisions to return to the relationship. In arecent longitudinal study of four US perpetratorprograms, Gondolf (2002a, p. 105) found the womenpartners’ hopes for the outcomes of their partners’participation to be ‘somewhat inflated and overlyhopeful’, despite having received advice of theprograms’ limitations, and despite histories of severeabuse over prolonged periods prior to their partners’arrest and mandated program participation.

At a time of growing interest in the development ofperpetrator programs in Australia, this issues paperreviews the evidence of their effectiveness, basedlargely on recent international research. It begins,however, with a discussion of the dominant ways inwhich the causes of domestic violence areunderstood, and the types of intervention withperpetrators that flows from these perspectives.

Theories and interventionsApproaches to intervention with perpetrators arebased on differing theoretical explanations for theirabusive, violent and controlling behaviour. Theseperspectives, in part, reflect different emphases ondifferent levels of intervention. This section of thepaper describes three explanatory perspectives onmale violence in intimate relationships, and thetypes of interventions which they underpin.

Through a socio-political lensThe recognition of domestic violence as a serioussocial problem is an achievement of second wave

Responding tomen1 whoperpetratedomestic violence: Controversies,interventions andchallenges

Dr Lesley Laing

2 A u s t r a l i a n D o m e s t i c & F a m i l y V i o l e n c e C l e a r i n g h o u s e I s s u e s P a p e r 7

feminism, a social movement originating in the late1960s and early 1970s. Feminism applied a socio-political framework to understanding domesticviolence, a vastly different perspective from the(then) prevailing medical model which saw thecauses of domestic violence as lying within thepathology of the individuals involved. A socio-political approach to understanding domesticviolence seeks answers at a social or group level,rather than at the level of the individual man.‘Instead of examining why this particular man beatshis particular wife, feminists seek to understandwhy men in general use physical force against theirpartners.’ (Bograd, 1988, p. 13) Because domesticviolence is a common (Australian Bureau ofStatistics, 1996), rather than a rare event, the focusis on the social conditions which support it, ratherthan on attempting to identify the characteristics ofa small ‘deviant’ group of men.

Key to this perspective on domestic violence are theconcepts of gender and power: men as a social grouphave greater power than women and violence is animportant way by which men maintain theirdominant position. Rather than a ‘safe haven’, ‘thefamily as a social institution mediates betweenoppression at the broadest social level and thepersonal relationships of intimate adult partners.’(Bograd, 1988, p. 14)

From a socio-political perspective, violence againstwomen can only be understood in its social context:

Men are violent to their women partners in a widercontext of family, friends, and the general culturaland institutional settings in which such behaviorand accompanying attitudes are more or lesscondemned or condoned. The messages and responsesare often mixed and ambivalent, showing support formen’s authority over wives, boundaries of‘appropriate’ behavior for women in the role of a wife,and more or less tolerance for the use of violenceunder certain circumstances. Sanctions for the use ofviolence are often weak or nonexistent and men incurfew if any costs for its use. (Dobash, Dobash, Cavanagh, & Lewis, 2000, p. 31)

Increasingly, feminist writing and action hasattempted to grapple with the ‘intersection’(Crenshaw, 1991) of gender with race, class andethnicity in order to more fully understand all thedimensions of the socio-political context in whichviolence against women occurs (e.g. Bograd, 1999;Hanmer, 2000). Also arising from a socio-politicalperspective is work which explores the ways inwhich ‘masculinity’ is socially constructed andrelated to violence. (e.g. Dobash & Dobash, 1998;Frey & Bellotti, 1995)

A focus of feminist intervention has been onunderstanding and validating the experience ofwomen and exploring the ways in which they have

coped with the terrifying and oppressive situationswith which they live (e.g. Campbell, Rose, Kub, &Nedd, 1988). It is sometimes assumed, therefore,that the voice of men who use violence is ignored bythose working from this perspective. However, someresearchers operating from a socio-political perspec-tive have explored men’s accounts of their violencethrough qualitative methodologies. For example,Dobash et al. (1998) used in-depth interviews tocompare the accounts of men and women of theviolence that resulted in the men’s arrest andreferral to a perpetrator program. The men’s andwomen’s accounts revealed ‘profound’ differences intheir reports of the nature and extent of men’sviolence, controlling behaviours, and the injurieswhich result: ‘women provide much more detailedand lengthy accounts than men, usually enter thenarrative at a much earlier point in the violentevent, and extend the narrative beyond acts ofviolence to include injuries and other consequences.’(Dobash et al., 1998, p. 407) In another study whichsought to understand abuse from the man’s perspec-tive, the men’s narratives were similarly noted to benarrow and thin (Stamp & Sabourin, 1995).

Cavanagh et al. (2001) used Goffman’s concept of‘remedial work’ and his three related devices of‘accounts’, ‘apologies’ and ‘requests’, to analyse theaccounts of abusive men. They found that men’saccounts seek to mitigate their culpability yet, at thesame time, paradoxically seek ‘absolution’ andforgiveness from the women they abuse. They arguethat the men’s accounts of their violence reveal theintentionality of the violence: ‘Through theseaccounts which dominate their reports, men seek notonly to neutralise and eradicate women’s experiencesof abuse but also to control the ways in whichwomen themselves might interpret and respond tothe violence.’ (Cavanagh et al., 2001, p. 711)

Interventions shaped by a socio-political perspectiveThe socio-political perspective has not generatedconsensus on the most appropriate response toviolent men. On one hand, some who explaindomestic violence from this perspective stronglyoppose intervention with individual men, a positionexemplified by the following remark:

If you look at people who have brought about majorsocial changes in our country and other men…peoplelike Nelson Mandela, Martin Luther King or Ghandi.Can you imagine anger management classes forwhite supremacists? ‘How did you feel right beforeyou burnt the cross?’ ‘Can you recognise your angercues?’ This is just something we would not do. (Pence1990, cited in Townsend, 1991, p.2)

An alternative position is that a socio-politicalexplanation for violence can be incorporated into

3A u s t r a l i a n D o m e s t i c & F a m i l y V i o l e n c e C l e a r i n g h o u s e I s s u e s P a p e r 7

work with individual men as part of the largerprocess of social change:

Even the more modest agenda of turning down thevolume of violence for increasing numbers of menand turning up the volume of safety for increasingnumbers of women is not only a small undertaking,it is a necessary and inevitable part of the pathway tothe elimination of all violence against women.(Dobash et al., 2000, p. 40).

Those adopting this position, however, advocate thatthe work done with men who use violence mustaddress as its core the social context of genderedinequality. ‘Pro-feminist’ or ‘gender-based’ groupwork, discussed in more detail below, has been themost common response to working with men whouse violence in their intimate relationships,stemming from a socio-political understanding ofdomestic violence.

It is paradoxical that, despite the wide acceptance ofsocio-political explanations for domestic violence,most interventions with men who perpetrateviolence in intimate relationships have been individ-ualised (Blagg, 2001). When their review of interven-tions with men who use violence identified nointerventions addressing the sociocultural level,Eisikovits and Edleson speculated that: ‘It may beeasier and less threatening to society to targetindividuals and families for change rather than thenorms or values that are part of an intricate web ofsocial order.’ (1989, p. 407)

A recent Australian example of an intervention thataims to bring about change in men’s behaviour bytargeting the broader social context is the NSW‘Violence Against Women – It’s Against all the Rules’community education campaign (Cheetham, 2001;Violence Against Women Specialist Unit, 2000). Thistargets young men via their interest in sport, andinvites them to be agents of social change inchallenging the use of violence against women.Approaches such as these which give the strongmessage from the community that violence is notacceptable, represent creative ways of underminingthe socio-cultural support for violence against women.The many creative programs working with youngpeople in schools to foster the development of egalit-arian and non-abusive relationships are also examplesof work addressing the social and cultural roots ofviolence (e.g. Sidey & Lynch, 2001; Zuchowski, 1999).

The development of programs for men whouse violence, based on a socio-politicalframework

In North America, intervention with men utilisinggender-based (sometimes termed ‘pro-feminist’),cognitive behavioural group work is currently themost commonly used approach stemming from asocio-political understanding of domestic violence.The first such group was established in 1977, at theurging of advocates working with women escapingviolence, and involved voluntary participants(Adams, 2000). Similarly, in Australia, some women’sservices have argued the need for involvement in thedevelopment of such programs (e.g. Taylor, 2000;Woodbridge, 2000):

The drive by a women’s service to take a lead in theprovision of men’s programs may be seen by some ascontentious and controversial. However, DVS remainsconvinced that the safety of women and childrendemands of us to be involved. For men’s programs tofunction in isolation to women’s services and vice-aversa is fraught with danger. (Taylor, 2000)

From the mid 1980s, the ‘mandatory’ arrest policiesin the US presented the practical problem of how thecriminal justice system could deal with the increasednumbers of men arrested for domestic violence.Gondolf (2002a) suggests that referral to perpetratorprograms was a ‘convenient option’ for the criminaljustice system, but notes that there was often poorfollow through and few consequences if men failed toattend or dropped out of the programs.

Because Australia has not gone down the US path ofadopting ‘mandatory’ arrest policies, there has notbeen the same impetus to develop mandatedprograms in concert with the criminal justicesystem. A comprehensive audit of Australianprograms for National Crime Prevention (KeysYoung, 1999) found that Australian programs havetended to develop in an ‘ad hoc’ manner. Factorssuch as lack of funding for the development ofspecialist perpetrator programs, increasing identifi-cation of violence as an issue in many presentationsto generalist relationship counselling services andrequests from women to provide counselling optionsfor their partners, encouraged the development ofvoluntary programs (Shaw, Bouris, & Pye, 1999).The National Crime Prevention audit identifiedconsiderable resistance to the concept of courtmandated, rather than voluntary, programs forperpetrators. Nevertheless, many of the programssurveyed claimed to have been influenced by thewell known Duluth Domestic Violence InterventionProject in the United States (Pence & Paymar,1993), despite the fact that:

…a core aspect of the Duluth program is that it existsin the context of an integrated community response inwhich a strong pro-arrest policy and mandatory

‘It may be easier and less threatening

to society to target individuals and

families for change rather than the

norms or values that are part of an

intricate web of social order.’


intentional leads to a moral discourse identifyingmen as responsible and accountable for their acts.(Dobash & Dobash, 1992, p. 248)

Key characteristics of these groups include thefollowing:

• Most importantly, perpetrator groups are notseen on their own as an adequate response todomestic violence. Rather they are but onecomponent of a co-ordinated communityresponse 5 involving, at a minimum, the criminaljustice system and services for abused women.(Adams, 2000; Dobash et al., 2000; Gondolf,2002a; Murphy, Musser, & Maton, 1998; Tolman& Bennett, 1990; Woodbridge, 1998)

• The safety of women and children is the primarygoal (Woodbridge, 2000): ‘Victim safety is the“gold standard” for batterer interventionprograms, the primary criterion by whichprogram effectiveness will be judged.’ (Bennet &Williams, 1998, p. 1)

• Wider accountability to women (e.g. Boyle, 2001;Hall, 2001; Laming, 2000; Mullender, 1996) ‘Indoing this work it is vital that the program usedis accountable and transparent to women’s livedexperience of domestic violence.’ (Woodbridge,2000, p. 9)

• Limited confidentiality. In contrast to the normsof conventional therapeutic encounters, theconfidentiality offered to participants is sharplycurtailed, with the program given permission tocontact partners and the criminal justice systemin order to ensure victim safety and offenderaccountability. (Mederos, 1999)

• Respectful interventions. In Australia, the workof Jenkins (1990) and White (1989) has beenparticularly influential in developing approachesthat respectfully invite men to address theirviolent behaviour as the focus of work, butwithout unproductive confrontation. Based onnarrative therapy (White & Epston, 1989), theseAustralian developments are different from someof the ‘educational’ group work approachesdeveloped in North America. This approach is

attendance at group education programs as part ofsentencing are key features. These features, ingeneral, appear to be notably absent from mostprograms currently provided in Australia. (Keys Young, 1999, p. 49)

In the time since the Keys Young audit, the situationin Australia has begun to change. More perpetratorprograms are being developed in co-ordination withthe criminal justice system. Such programs nowexist in Western Australia (Kraszlan & RebeccaWest, 2001), South Australia (Central ViolenceIntervention Program, 2002; Mort, 2001),Queensland (Eglington, 1999; Taylor, 2001;Woodbridge, 1998), the Northern Territory (Demos,Rudd, Gzik, & Griffiths, 1998) and the ACT (KeysYoung, 2000), with a pilot program underway inNSW. Some policy differences have emerged. Forexample, the NSW Government, through its Councilon Violence Against Women (1999) supports thedevelopment of mandatory3 rather than voluntaryprograms, while the Victorian Government supportsprograms being made available to men who self-referas well as for those who are directed to attend andfor those within the criminal justice system (Office ofWomen’s Policy Victoria, 2001).

Gender-based, cognitive behavioural groups

Socio-political perspectives on domestic violencehave resulted in the development of a specialisedtype of group work intervention, commonly termed‘gender-based, cognitive behavioural’ group work.These groups can be differentiated from moretraditional group work in several key respects.Emphasis on social context results in a focus oneducation4 of men about gender inequality and thetactics of power and control in relationships.Stopping the violence, abuse and controllingbehaviours is the goal of intervention, rather thanproviding a therapeutic response to the man’sindividual psychological problems such as ‘low self-esteem’ or ‘poor impulse control’.

Education groups focus on male and female socialand cultural roles. If it is accepted that abuse stemsnot from anger but from a belief system wherein menare convinced they have the right to dominate andcontrol, and men force their relationships to becomedeeply embedded in such assumptions, then thatbelief system has to be confronted for abusing men.(Orme, Dominelli, & Mullender, 2000, p. 97)

The notion of the man accepting responsibility(Boyle, 2001; Hall, 2001; Jenkins, 1990) for theabuse, for its effects and for stopping violence andcontrolling behaviour, is central to this form ofintervention:

Violence is seen as intentional behaviour chosen bymen as a tactic or resource associated with attemptsto control and dominate women…defining violence as

‘... If it is accepted that abuse stems not

from anger but from a belief system

wherein men are convinced they have

the right to dominate and control, and

men force their relationships to become

deeply embedded in such assumptions,

then that belief system has to be

confronted for abusing men.’


reflected in the following excerpt from a groupwork manual reflecting the narrative therapyapproach: ‘It is (therefore) important thatworkers do not reproduce abusive behaviours byusing their powers to impose opinions and valueson clients. This includes the use of “beneficialbullying” to break down a man’s denial of theproblem.’ (Northern Metropolitan CommunityHealth Service, 1997, p. 33). The respectfulstance from which the worker operates makes itpossible to acknowledge factors such as childabuse which may be part of the man’s history,while maintaining a clear focus on stoppingviolence.6

The Emerge program in Boston provides an exampleof the operationalisation of the principle of account-ability to the women (ex)partners of men in itsprogram. ‘Emerge views its primary mission asbuilding accountability, as opposed to changingbatterers.’ (Adams, 2000, p. 318) Therefore, theprogram devotes considerable attention todocumenting the men’s violence and non-compliancewith the program requirements, since their experi-ence indicates that many men will drop out ofprograms or will deny previous admissions ofabusive behaviour once away from the program. Thisdocumentation of the man’s violence includesbehaviour acknowledged by the man and otherinformation from the victim, police and child protec-tion records. This documentation is made availableto women on request, for example in matters of childcontact. ‘The report may be the only documentationof the batterer’s problems which is independent ofthe victim’s allegations.’ (Adams, 2000, p. 319)

As work with men has developed, many educationalprograms have also incorporated cognitivebehavioural interventions, which are based on sociallearning theory. Socio-cultural explanations forviolence can incorporate social learning theorythrough their emphasis on gender role socialisationand acknowledgment that violence is learned andsustained in many social contexts: friends andrelatives, peers, the response (or non-response) ofthe justice system and the popular media. Cognitivebehavioural approaches aim to change behaviourthough challenging the thinking on which abusivebehaviour is based – ideas, for example about maleentitlement and victim blame. The following is atypical description of gender-based cognitivebehavioural groups:

This approach confronts men with the consequencesof their behavior, holds them responsible for theirabuse, confronts rationalizations and excuses, andteaches alternative reactions and behaviors. Itcombines components such as skill training andanger management with education and confrontationabout power and control issues. (White & Gondolf, 2000, p. 468)

For some, the incorporation of behaviouraltechniques such as assertiveness training and angermanagement strengthens the intervention byproviding men with skills for developing non-violentrelationships. For others, including such componentsis seen as a diversion from addressing the genderissues that underlie violence against women. Theirinclusion implies that violence arises from skillsdeficits, rather than from a sense of entitlement andbelief in male authority to decide on female standardsof behaviour and to enforce these standards. Fromthis perspective, including behavioural techniquessuch as these, represents a movement away from asocial to a psychological framework in responding tomen who abuse (Gondolf, 2002a). The key issueappears to be whether such elements can be incorpo-rated within a broader framework which challengesmen’s domination of women.

While there is disagreement about whether or notsuch components have a place in group work withmen who use violence, there is more consensus thatthe use of such behavioural interventions alone,outside of a socio-political perspective on domesticviolence, is an incomplete and inadequate responseto men’s violence (Orme et al., 2000). This debate isheard most strongly with respect to the use of ‘angermanagement’ as the sole intervention. ‘Stand alone’anger management programs are regarded asinappropriate interventions because, for example,they may imply that the victim bears some responsi-bility for ‘provoking’ the anger and thus the abuse;because they fail to address the premeditated,controlling behaviours which are a core part of theabuse; and because they fail to address the socialsupports for wife abuse (Gondolf, 2002a).

Individual/psychological perspectivesAnother body of literature, contrasting sharply withsocio-cultural perspectives, focuses on understandingthe characteristics of individual men who are violenttowards their partners. This literature follows one oftwo lines of inquiry. The first compares domesticviolence perpetrators to men who do not perpetrateviolence in their intimate relationships, while thesecond attempts to identify the different types ofdomestically violent men (Hamberger, Lohr, Bonge,& Tolin, 1996).

Attempts to Identify the characteristicsof domestic violence perpetratorsIn a comprehensive review of the literaturecomparing ‘maritally violent’ and ‘maritally non-violent’ men, Holtzworth-Munroe et al. (1997)conclude that maritally violent men show more:psychological distress, personality disorders, attach-ment/dependency problems, anger/hostility and


alcohol problems, than do non-violent men. They alsofound that violent men ‘may hold attitudes (towardsviolence against women) and make attributions(regarding both wife behavior and their ownviolence) that increase their risk for using physicalaggression. They may lack resources or feelpowerless.’ (Holtzworth-Munroe et al., 1997, p. 94).Men who perpetrate violence in their intimaterelationships are also more likely to have experiencedviolence in their family of origin.

Among these characteristics of violent men, thepsychological variable of adult attachment style isattracting increased theoretical, research andtreatment interest (e.g. Dutton, 1995; Scott & Wolfe,2000; Stosny, 1995). Mauricio and Gormley (2001, p.1069) describe the concept of adult attachment styleas ‘the propensity to interact in patterned ways inclose adult relationships’. It is proposed that thequality of the infant-caretaker relationship providesa ‘blueprint’ for the individual’s later interpersonalrelationships. This theoretical perspective proposesthat: ‘…excessive interpersonal dependency amongabusive men is…a consequence of insecure attach-ment in childhood’. (Buttell & Jones, 2001, p. 376)However, a recent study which compared the level ofinterpersonal dependency of a sample of men whowere court mandated to domestic violence treatmentwith a comparison group of non-violent men did notfind that the domestic violence perpetratorsexhibited greater interpersonal dependency than thecomparison group (Buttell & Jones, 2001), incontrast to two earlier studies which had found someevidence of a relationship between high interpersonaldependency and violence perpetration.

Interest in the impact of trauma symptoms in adultperpetrators of violence is another, connected area ofemerging interest. Dutton (2000) suggests that bothexperiencing child abuse and witnessing violencebetween parents are risk factors for the developmentof post-traumatic stress disorder (PTSD), and thatthis is associated with violence perpetration: ‘…PTSDmay be a link or mediating variable betweenchildhood abuse victimization and adult perpetrationof intimate abuse.’ (Dutton, 2000, p. 301) Thistheoretical approach has promoted studies whichexplore the existence of trauma symptoms inperpetrators compared to non-violent men (e.g.Dutton, 1995) and which explore the extent to whichperpetrators of domestic violence evidence symptomsof dissociation7 (Simoneti, Scott, & Murphy, 2000). Ithas been hypothesised that dissociative copingmechanisms facilitate ‘the extreme detachmentnecessary to engage in severe violence towards anintimate partner’. (Simoneti et al., 2000, p. 1264)This perspective, however, does not address thegender difference in violence perpetration: both maleand female children experience violence in theirfamilies of origin and may dissociate in response tothese traumatic experiences, yet domestic violence is

perpetrated by men in the majority of cases (Taft etal., 2001).

Identifying typologies of abusive menHoltzworth-Munroe et al. (2000, p. 1000) argue that:

…the understanding of marital violence is morelikely to be advanced by drawing attention to (these)differences (between violent men) than by continuingto treat all violent husbands as one homogenousgroup…the identification of batterer subtypes opensthe possibility of patient-treatment matching thatmay increase intervention effectiveness.

In a widely cited paper, Holtzworth-Munroe andStuart (1994) developed a batterer typology based onthe three dimensions of:

• the severity and frequency of the violence,including psychological and sexual abuse

• the domain of the violence (i.e. family only orextra-familial violence and other criminalbehaviour)

• the batterer’s psychopathology or personalitydisorders. (Holtzworth-Munroe et al., 2000, p. 1000)

These dimensions generated three sub-types ofperpetrators: the ‘family only (FO)’ perpetrators, whoit was hypothesised would engage in the least severedomestic violence, the least violence outside thefamily and be engaged in less criminal behaviour;the ‘borderline-dysphoric (BD)’ group, whose abuse oftheir partners would be moderate to severe, whomay exhibit some violence outside the family andwho ‘would be psychologically distressed, evidencingborderline personality characteristics and someproblems with substance abuse’ (Holtzworth-Munroeet al., 2000, p. 1000); and the ‘generally violent-antisocial (GVA)’ group whose abuse of partnerswould be moderate to severe and who would engagein the most extra-familial violence and criminalbehaviour. This last group would have antisocialpersonality disorders and problems with substanceabuse.

Drawing on a number of theoretical approaches,Holtzworth-Munroe and Stuart (1994) thensuggested a developmental model for identifying riskfactors on the basis of this typology. Their modelincorporated both ‘distal-historical correlates’ suchas growing up with child abuse and parentalviolence and associating with violent peers and‘proximal correlates’ such as impulsivity, social skillsand attachment and dependency (Holtzworth-Munroe et al., 2000). A number of studies havesubsequently been undertaken to test this theoret-ical model (Hamberger et al., 1996; Holtzworth-Munroe et al., 2000).


In a review of batterer typologies, Holtzworth-Munroe et al. (2000) find that many ‘fit’ with thedirection of the Holtzworth-Munroe and Stuart(1994) typology. For example, Jacobson and Gottman(1998) identified ‘cobras’ (similar to the GVA group)and ‘pitbulls’ (similar to the BD group) while Tweedand Dutton (1998) identified ‘instrumental’ (similarto the GVA group) and ‘impulsive’ batterers (similarto the BD group).

Linking psychological characteristicsto treatmentHoltzworth-Munroe et al. (2000) argue that futurework needs to explore how men in the differentperpetrator sub-types respond (or not) to differentforms of treatment. For example, they postulate thata sub-group of the ‘GVA’ men may be psychopaths forwhom treatment is likely to be ineffective. In asimilar vein, Huss and Langhinrichsen-Rohling(2000) suggest that it would be useful for thedomestic violence field to become aware of the litera-ture on psychopathy in addressing the policy andtreatment implications arising from identifying thissub-group of batterers.

Despite the burgeoning interest in typologies, andconviction that their development will lead to amovement away from a ‘one size fits all’ approach toperpetrator treatment, there has to date been littleresearch which addresses the differential benefits oftreatment approaches for different perpetrator sub-types. An exception is a study by Saunders (1996)which randomly assigned 218 perpetrators (themajority court mandated) to one of two treatmenttypes: a feminist-cognitive-behavioural group(FCBT) or a process-psychodynamic group (PPT).The study aimed to evaluate the effectiveness ofthese two treatment models in preventing reassaultand to test whether offenders with particularpersonality traits and developmental histories wouldhave better outcomes depending on the type oftreatment received. Drawing on the Holtzworth-Munroe and Stuart typology (1994), it was predictedthat the feminist-cognitive-behavioural treatmentapproach of skills training and attitude changewould be best suited to men with antisocial person-

ality traits, and that, the more dependent theabuser’s personality, the more he would benefit fromthe less tightly structured process psychodynamicgroup (PPT) which concentrated on building trustand uncovering and reconnecting with childhoodtraumas. The study found that recidivism rates werealmost identical between the two treatmentconditions. The findings regarding the interaction ofperpetrator traits and intervention type were in thedirection predicted.

Those who understand domestic violence as linked tochildhood experiences of abuse and witnessingviolence, argue that treatment for perpetratorsshould include addressing and resolving childhoodtrauma in the sub-group of perpetrators who havesymptoms of PTSD. The issue of how to addressthese issues for men within a framework that doesnot undermine their acceptance of responsibility forthe abuse which they have perpetrated, is atreatment issue which many grapple to address (e.g.Hall, 2001). The main concern which is voiced inrespect to these perspectives is that a focus onproviding treatment for men’s psychologicalproblems ‘can too easily reinforce the sense ofentitlement, self-righteousness, and narcissism sooften associated with men who batter women.’(Gondolf, 2002a, p. 11)

The recent large multi-site, longitudinal study ofperpetrator program outcome (Gondolf, 2002a) whichis discussed in detail later in this paper 8, offers someinteresting findings about the utility of attempts toidentify the personality profiles of abusive men, thedevelopment of “batterer typologies’ and the link tointervention. Using the Millon Clinical MultiaxialInventory (MCMI), the instrument that has beenwidely used in previous studies of perpetrators, thisstudy attempted to verify the findings of previousstudies which had indicated that: ‘(1) the vastmajority of batterer program participants showevidence of psychopathology, and (2) batterers maybe distinguished by their personality tendencies anddisorders.’ (Gondolf, 1999c, p. 2) The MCMI-III wasadministered to 840 men, 82 per cent of whom werecourt mandated, at entry to perpetrator programs infour cities.

In contrast to previous studies, the findings revealed‘less pathology’ in this sample of men, with the mostnotable trend being towards narcissistic or antisocialpersonalities. Gondolf suggests that this finding isconsistent with characterisations of perpetrators asacting with a sense of entitlement, dominance andself-centredness. Further, the study found nosupport for the proposition that perpetrators arecharacterised by borderline tendencies: ‘…there islittle evidence for a prevailing “abusive personality”typified by borderline personality tendencies, andlittle support for a preponderance of post-traumaticstress disorder among our sample, in contrast to

Those who understand domestic

violence as linked to childhood

experiences of abuse and witnessing

violence, argue that treatment for

perpetrators should include addressing

and resolving childhood trauma in the

subgroup of perpetrators who have

symptoms of PTSD.


findings of a Canadian study (Dutton, 1995).’(Gondolf, 1999c, p. 13) Gondolf concludes, on thebasis of this study:

These findings raise caution to characterizations thatmay overly ‘pathologise’ batterers and battering.(1999c, p. 15)

Gondolf also explored the notion that identifyingperpetrator typologies would allow a better ‘fit’between the man and the treatment approach.Proponents of individual/psychological approacheshave argued against the dominance of feminist-informed cognitive behavioural approaches, on thegrounds that a ‘one size fits all’ approach cannotrespond to the treatment needs of different sub-groups of perpetrators (eg. Holtzworth-Munroe et al.,2000). White and Gondolf (2000) drew a randomsample of 100 perpetrators from the larger study.Based on MCMI profiles for the sample, the studyidentified three levels of personality pathology: low(56 per cent of the sample); moderate (29 per cent ofthe sample) and severe (15 per cent of the sample).Six major personality groupings of perpetratorprofiles were identified which were consistent withprevious typologies. Further examination of theprofiles, however, led to the identification of anothertrend: ‘…a continuum of narcissistic and avoidanttendencies that cut across the groupings’. (White &Gondolf, 2000, p. 483) The authors argue thattreatment recommendations associated with theMCMI profiles identified in this study suggest thatcognitive behavioural treatment approaches areappropriate for most of the men in the sample. Thereis a small group of men with severe personalityproblems or psychopathology who may need referralfor additional psychiatric help or close monitoringfor violations of limits. The authors conclude, on thebasis of this study, with respect to the link betweenpersonality types and treatment, that:

The prevailing gender-based, cognitive behaviouralgroup treatment may, however, be appropriate formost men referred to batterer programs. It appearsthat although one size does not fit all, one sizeappears to fit most. (White & Gondolf, 2000, p. 486)

Systemic perspectivesSystemic theories focus on the patterns of interac-tion between couples and underlie interventionswith the man and woman as a couple. Reviewingapproaches in the 1980s, Eisikovits and Edleson(1989, p. 388) noted that: ‘The language thatsystems-oriented practitioners use often replaces“victim” and “abuser” labels with such terms such as“abusive or violent couples”. Such approaches havebeen vigorously critiqued on the grounds that couplecounselling may jeopardise the woman’s safety,because she is made vulnerable to retaliationthrough disclosures made in the therapy situation;

that it implies that the problem is mutual and that,as a contributor, the woman is expected to change;and that the focus will be on saving the relationshiprather than addressing the violence and coercivecontrol exercised by the perpetrator (Lipchik, Sirles,& Kubicki, 1997). In most states in the US, this formof intervention is regarded as inappropriate anddangerous (Austin & Dankwort, 1999b) and istypically limited until the man has completed aperpetrator program and has been ‘violence free’ fora prescribed period of time (Trute, 1998).

However, Lipchik, Sirles and Kubicki (1997) arguethat there should be a place for conjoint therapy in aco-ordinated interagency response to domesticviolence, because 50-75 per cent of couples ‘continuein the relationship despite the best efforts of police,prosectors, shelters and advocates’, and becausethere is as yet little evidence of the effectiveness ofperpetrator groups, the most popular form ofintervention. Over the last decade, a number ofapproaches which attempt to address the concernsand risks of conjoint therapy have been described inthe literature (Goldner, 1999; Goldner, Penn,Scheinberg, & Walker, 1990; Lipchik et al., 1997;Shaw et al., 1999).

Summary – perspectives andintervention approachesFrom the previous discussion it can be seen thatapproaches to understanding and responding toperpetrators of domestic violence have tended to bepolarised, emphasising either an individual/psycho-logical or a socio-political perspective (Goldner,1999). Mankowski, Haaken and Silvergleid (2002, p.172) describe the dilemma as a struggle overwhether the response to perpetrators should be‘therapeutic treatment to overcome a psychologicalproblem or rather, directive re-education, andpunishment to interrupt criminal behavior.’ Thispolarisation leaves each side with concerns about theother. For those who emphasise the individual andpsychological context:

…focusing on the political context and ignoring theindividual context seems dehumanizing and dismissive of men’s experience, which often includeshistories of abuse or neglect. For proponents of thepower and control model, however, ignoring thepolitical context is oppressively dismissive of women’sexperience. (Mankowski et al., 2002, p. 173)

It seems unlikely that this debate will be easily orquickly resolved. Resolution will involve welldesigned studies which evaluate the outcomes ofdifferent types of intervention. The next section ofthis paper turns to the current evidence for theeffectiveness of perpetrator programs.


Evaluating the effectivenessof perpetrator programs

IntroductionBecause of the controversy surrounding responses toperpetrators of domestic violence, the question of theeffectiveness of perpetrator programs has beenextensively canvassed. Yet it is clear that there is noeasy answer to the question as to whether or notperpetrator programs ‘work’ and that teasing out thecomplexities which underlie this apparently simplequestion is a challenging endeavour.

The earliest efforts to evaluate outcomes ofperpetrator programs were plagued by an array ofmethodological problems, including, for example:

• Lack of agreement about what constitutes‘success’ or ‘effectiveness’ (Muller, 1997). Is it, forexample, a total cessation of all forms of abusive,coercive and controlling behaviour; cessation ofphysical abuse; or a reduction in the amount ofphysical abuse? (Edleson, 1995; Gondolf, 1997a)

• Small sample sizes. This is common in theAustralian research. (e.g. Hewitt & Cavanagh,1998; Nankervis & Donne, 1993; urbis keysyoung, 2001a)

• Disagreement over what constitutes an adequatefollow-up period. This relates to concerns thatmen may revert to violent and abusive behaviourwhen no longer under the scrutiny of theprogram and/or the legal system.

• Reliance on self reports of change by the men orrearrest records as outcome measures, both ofwhich under-count re-offending. (Palmer, Brown,& Barrera, 1992)

• Low response rates in follow-up because of highrates of program drop out and difficulty intracing participants over extended follow-upperiods. For example, in one Australian study, 69of 86 men completed a 12 week program; only 23men and their partners participated in follow-upat the six month point, and 20 at 12 months.(Poynter, 1991)

• Difficulty of involving partners in follow-up, forexample because of the risks of infringing ontheir privacy, jeopardising their safety, or otherreasons. (Frances, 1996; Gondolf, 2000c; Palmeret al., 1992; Taylor, Davis, & Maxwell, 2001)

• Inclusion of only initial partners, leavingunanswered the question of whether men simplytransfer their abusive behaviour to new partners.

• No control or comparison group used so that anychanges identified cannot necessarily be attrib-uted to the program. Much of the international

‘first generation’ research in the field was of thiskind (Eisikovits & Edleson, 1989) as is thecurrent Australian research evaluation literature.(e.g. Baum, Brand, Colley, & Cooke, 1987;Poynter, 1991; urbis keys young, 2001a)

• Evaluations conducted by staff with a biastowards demonstrating ‘success’. (Gondolf, 2002a)

• Problems in operationalising outcomes for womenwhich aim to increase their safety and well being.(Austin & Dankwort, 1999a)

More recently, a number of studies have attemptedto address some of these methodological challenges.Other areas of debate have also emerged, such asthe debate about which type of research design canprovide the best evidence for perpetrator programoutcomes. There are some who contend, for example,that nothing less than evidence of effectiveness fromevaluation of perpetrator programs using an experi-mental research design is required to validate theiroperation. Experimental designs randomly assignparticipants to two groups – a control group, whichdoes not receive the intervention being tested, and atreatment group which does receive it. The resultsfor the two groups are then compared, to establish ifthe intervention has had an impact:

…random assignment ensures that the groups beingcompared are similar prior to the implementation ofthe experimental stimulus. As such, random assignment provides the most unambiguous resultsconcerning the efficacy of treatment… No othermethod is better able to rule out competing explanations for changes in the dependent variable…(Feder, Jolin, & Feyerherm, 2000, p. 381)

However, random assignment can be difficult toimplement in field situations, as evidenced in anumber of the studies to be discussed shortly.

Another type of research design, ‘quasi-experi-mental’, uses a comparison group, but withoutrandom assignment. In perpetrator program evaluations, for example, the comparison group maycomprise program ‘dropouts’, rather than menrandomly assigned to a ‘no intervention’ group. Theobvious problem here is that the comparison groupsmay be different - perhaps the men who drop out aremore dangerous and disturbed, hence any differencesbetween the groups are due to the characteristics of

...It is clear that there is no easy answer

to the question as to whether or not

perpetrator programs ‘work’ and that

teasing out the complexities which

underlie this apparently simple question

is a challenging endeavour.


the two groups of participants rather than to theimpact of the program. Statistical techniques areemployed to attempt to control for such differencesbetween the groups.

‘Non experiments’ measure participants before andafter intervention, and have no comparison group.Hence is it not possible to attribute identifiedchanges to the program since they may be due toother, intervening factors. Non experimentsdominated the early attempts to evaluateperpetrator programs and as noted above, this is thedesign used to date in all Australian evaluations.

While some argue that it is unethical to legallymandate treatment that has not been provedeffective using an experimental design, others arguethat in a complex area such as domestic violence,where programs for perpetrators are but onecomponent of a coordinated system of intervention,experiments are ‘artificial’ and ‘naive’ about programcontext (Gondolf, 2001). For example, experimentalprograms tend to treat men who drop out as part ofthe treatment group, when the ideal within a co-ordinated community response would be thatdropping out should lead to consequences such asjailing or more intensive probation supervision(Gondolf, 2002a).

Evaluations which use anexperimental research designIn the first evaluation to utilise an experimentaldesign, Palmer, Brown and Barrera (1992) compared59 court ordered perpetrators who participatedunder probation in a 10 week psycho-educational,client-centred group program with a control groupassigned randomly to probation only. Follow-upquestionnaires were mailed to both groups 12months after program completion. Although theresearch design aimed to collect outcome data fromthe men, their partners and from police reports, thelow response rates, particularly for the women9, ledto reliance on police data. Recidivism, measured bypolice records, was significantly higher for thecontrol group than for the program group. However,this study suffered from small numbers and relianceon police data.

Using a random group assignment of men who hadphysically assaulted their wives, Dunford (2000)compared four, 12-month interventions with navalbase personnel: a men’s cognitive behavioural group;a couples’ counselling group10 using a cognitivebehavioural approach; rigorous monitoring of themen; and no intervention with the men (althoughsafety planning was implemented with the womenpartners). Four outcome measures were used: selfreport by the men and their partners of three levelsof abuse (fear of abuse; physical abuse; physicalinjury); the modified Conflict Tactics Scale; police

and court reports; and date of the first repeat case ofspouse abuse. The men and their partners wereinterviewed four times at six monthly intervals:before treatment; at the end of six monthstreatment; and twice more at six monthly intervals 11.This study found no difference in outcomes betweenthe treatment conditions during a year long follow-up. Dunford (2000, p. 475) concludes that: ‘Theinterventions of the cognitive-behavioural modelfailed to produce meaningful changes in thebehaviour they were designed to impact.’

In many ways, the men in this study were nottypical of other men referred to perpetratorprograms: all were married, were much younger, andhad shorter abuse histories and less drug andalcohol problems than non-military perpetrators.Dunford argues that the failure to demonstrate theeffectiveness of cognitive behavioural groupintervention in this ‘optimal’ setting sounds a note ofcaution for programs run under more typicalcommunity settings.

Others (e.g. Bennet & Williams, 1998; Gondolf,2002a) have interpreted these results differently.Bennet and Williams note that the overall recidivismin the study was low (30 per cent on the women’sreport and four per cent by arrest) and attribute thisto the fact that the men, by virtue of their being inthe navy, were subject to more extensive supervisionand decisive sanctions (loss of job and housing) thanare men in the general community. Hence theyargue that this study:

…while questionable as an indicator of battererprogram effectiveness, is nevertheless useful as anindicator of coordinated community intervention…Ifcommunities take a proactive response to domesticviolence, including assertive probation work,sanctions for non-compliance, victim safetymonitoring, and batterer intervention programs, theywill reduce the incidence of repeat violence.(Bennet & Williams, 1998, p. 6)

In New York, 376 men were randomly assigned12 to asix month Duluth-style perpetrator program or to sixmonths community service (Taylor et al., 2001).During the evaluation, some men were assigned toan alternative treatment program of two months

‘… If communities take a proactive

response to domestic violence,

including assertive probation work,

sanctions for non-compliance, victim

safety monitoring, and batterer interven-

tion programs, they will reduce the

incidence of repeat violence.”


duration, with two sessions a week rather than one.The control group comprised the men assigned tocommunity service. Outcomes were measured byinterviewing the women partners of the men atsentencing, and again six and 12 months later andby police record search 12 months after sentencing.Completion rates for victim surveys, however, werelow: 51 per cent at intake; 48 percent for the second,and 50 per cent for the third interview. Thirty fiveper cent of victims could not be contacted at anytime during the follow-up period.

The evaluation found that the men in the longer (sixmonth program) were significantly less likely to bearrested for domestic violence than the men in thecontrol (community service) group and also men inthe shorter treatment group. This treatment effectdid not diminish significantly between the six and12-month follow-up periods. However, when theresearchers looked at the victim reports of newincidents, there was no statistically significantdifference between the groups. As mentionedpreviously, criminal justice recidivism data tends toidentify lower rates of re-offending than do victims.However, the low victim response rates achieved inthis study mean that it is possible that the criminaljustice recidivism data is more reliable and that thefindings on victim report measures are due toresponse bias or to other, intervening variables(Gondolf, 2001). This study also encountered difficul-ties in implementing random assignment: judgesover-rode the random assignment to communityservices in 14 per cent of cases, and the shorterprogram was introduced to appease defence lawyers.

In Broward County, Florida, all men (404) convictedof misdemeanour domestic violence in two specialistdomestic violence courts over a five month periodwere randomly assigned to either one year’sprobation with 26 weeks’ court mandatedperpetrator counselling13, or to one year’s probation,with no perpetrator program (Feder & Dugan; 2002;Feder et al., 2000). Data was collected frombatterers, victims and criminal justice records. Themen and their initial partners were interviewed atadjudication, six months later, and the women 12months after adjudication. Tools measuring attitudestowards women and wife abuse, the criminalisationof domestic violence, responsibility for domesticviolence, the likelihood of hitting/being hit again andthe revised Conflict Tactics Scale, were administeredto the men and women. Probation and arrest recordswere also checked at 12 months post adjudication.

No evidence was found that the men in the programdid any better either behaviourally, or in attitudinalchange, than the men who only received probation.

The results show that, in this county, there were noclear and demonstrable positive effects of this court-mandated SAAP [spouse abuse abatement] programon the offenders’ attitudes, beliefs, and behaviours.

An analysis of the offenders’ self-reported andvictims’ reports of psychological and physical abuse,using the CTS2, suggested that the offenders’behavior did not change over time. Of note, there wasstill evidence of physical abuse 6-12 months post-sentencing. Analyses failed to uncover differencesbetween control and experimental subjects in theirlikelihood of reoffending and being arrested duringthe follow-up period.(Feder and Dugan, 2002, p. 371)

The researchers in this study have documented thedifficulties which they encountered in implementingthis randomised field trial within the criminaljustice system, in a climate in which the use ofmandated perpetrator programs enjoyed widesupport (Feder et al., 2000). One consequence of thiswas limited access to victims14. However, theresearchers argue that the low victim response ratewas compensated for by the use of multiple sourcesof data. Opponents of the experiment believed thatwomen were being placed at greater danger if theirpartner were not mandated to treatment, yet theresearcher’s position was that:

…it was unethical to mandate an intervention thathad not been rigorously tested such that personsunderstood the possible positive and negativeconsequences of the treatment.(Feder et al., 2000, p. 384)

Summary – experimental studiesIt was hoped that outcome studies using an experi-mental design would unequivocally answer thequestion as to whether or not perpetrator programsare effective. However, reviews of the results of theseexperimental studies have led to varying conclusionsby different researchers. For example, Feder andDugan (2002, p. 351) review these studies andidentify a number of problems in their implementa-tion, concluding that: ‘…there are inconsistenciesand (that) questions remain about the integrity ofthese experiments and the generalizability of theirresults.’ They argue that experimental studies ofperpetrator programs need to address three issues(Feder and Dugan, 2002, p. 351-352):

• They must include men from the general popula-tion so that the results can be generalised (aweakness of the study which included only navypersonnel)

• The sampling frame must not be limited to menwho volunteer or who are assessed by others assuitable to attend a perpetrator program (as inthe New York study)

• Participants must receive ‘the same things in thesame amount – other than the experimentalstimulus…men in both experimental and controlconditions would have to receive the same


amount and duration of criminal justice supervi-sion, monitoring and nontreatment contact sothat the only difference between the two groupswould be that one group was court mandated intocounselling and that the other was not.’ (p. 352)

Feder and Dugan outline the ways in which theBroward County study met these criteria and argue,therefore, that their findings suggest that:

…an unquestioning acceptance of domestic violencebatterers’ intervention needs to be challenged. Thereis evidence from multiple sources that it did not workin Broward County. (Feder and Dugan, 2002, p. 372)

Bennet & Williams (1998) conclude that the evidenceon program effectiveness to date is inconclusive.They argue it is both difficult and undesirable to tryto distinguish the effects of perpetrator programsfrom the impact of the co-ordinated responses withinwhich they are located. Gondolf makes a similarpoint in his review of the experimental studies:

The evaluations together suggest that the effectivenessof batterer programs alone is not readily apparent orrather weak and that claims of overwhelming successshould be regarded with suspicion. They alsoindirectly imply that more attention needs to be givento program context. The encompassing interventionsystem of arrest, court action, victim services, andprobation monitoring may substantially affectprogram success. (Gondolf, 2001, p. 87)

It is this exploration of batterer programs withintheir context which is addressed in Gondolf ’s ownmulti-site study, discussed in the next section of thispaper.

Recent quasi-experimental evaluation studiesRussell and Rebecca Dobash and colleaguescompared the effects of two Scottish court mandatedmen’s programs with alternative, traditional formsof criminal justice dispositions such as fines,probation or imprisonment (Dobash et al., 2000;Dobash & Dobash, 1997; Dobash & Dobash, 2000;Dobash, Dobash, Cavanagh, & Lewis, 1999). Thestudy compared two naturally occurring groups: the‘Men’s Program Group’ comprising men who weresentenced to and who completed one of two abusergroups as a condition of their probation; and the‘Other Criminal Justice Group’ (Other CJ)comprising men sentenced by the courts for adomestic violence offence, who received some otherform of sanction (Dobash et al., 2000, p. 72)15. Datawas gathered via an initial in-depth interview withthe men and women and by postal questionnaire attwo follow-up times, three and 12 months afterinitial contact. The methodology included using bothmen and women partners as respondents and thedevelopment of four new tools to study ‘violence’,

‘injuries’, ‘controlling behaviour’ and ‘quality of life’for men and women. In this way, the researchersattempted to move beyond using re-assault as thesole measure of outcome.

The study found that very few men in either theprogram group or the comparison group werecharged with further violence towards theirpartners, indicating no difference on the outcomemeasure of criminal recidivism. In contrast to therecidivism data, the women’s reports revealed muchhigher rates of re-abuse for both groups. However,the program group were more successful at reducingtheir violent and controlling behaviour. Threemonths after the initial interview, 62 per cent of the‘other’ (non program) group had perpetrated at leastone act of violence, compared with 30 per cent of the‘program‘ group, a statistically significant difference.At the 12 months follow-up, 75 per cent of thecomparison group had re-assaulted, compared to 33per cent of the ‘program’ group. Further, at the 12-month point, 37 per cent of the women partners ofmen in the comparison group reported frequentviolence compared with seven per cent of womenpartners of men in the program group (Dobash &Dobash, 1997).

Women living with men in the program group alsoreported significant reductions in controllingbehaviours. With respect to findings about quality oflife, women partners of men in the program groupwere:

…much more likely than women in the Other CJgroup to say they were happy, more relaxed and lessfrightened than before the intervention…Anoverwhelming majority of women in the ProgrammeGroup also indicated that their partner was lesslikely to try to restrict their lives and more likely totake responsibility for their violence.(Dobash & Dobash, 1997, p. 251)

An important feature of this study was its use of thewomen’s reports as its main indicator of outcomeand in the development of new tools whichattempted to more adequately capture all aspects ofabuse – the violence, its impact, and the coredynamic of coercive control. In order to gather asrich a picture as possible, the interviews began withopen-ended questions before the more specificmeasurement tools were introduced. In theinterviews, respondents were shown numbered cardsso they could identify abuse which they had experi-enced or perpetrated; this technique introduced toaddress the impact on reporting of victim shame andthe tendency for perpetrators to minimise theirabusive behaviour.

Dobash et al. (1999) assert that their results point tothe need for mandated treatment, since theprograms studied suffered little attrition, a commonproblem for many programs. While this is a


commonly held position, Feder and Dugan (2002)point out that, in the Broward County studydiscussed earlier, offenders’ behaviour wasmonitored by probation officers and they weresanctioned when they did not attend. Nevertheless,they point out that it was still not possible to ensurethat all the men attended all the required sessions.This study explored the impact on perpetratortreatment outcomes of ‘stake in conformity’variables, such as age, employment status, residen-tial stability and marital status, as identified in theoriginal Minneapolis Spouse Abuse Experiment andthe replication studies about the relationshipbetween arrest and repeat domestic violenceoffending.16 They found that:

Despite the monitoring and sanctions, the men stillself-selected into compliers (treatment completers)and non compliers (treatment dropouts). Additionalanalyses indicated that stake-in-conformity variablespredicted both whether a man would comply with histerms of probation (specifically the court mandate toattend his SAAP program) and whether he wouldreoffend during the follow-up period.(Feder and Dugan, 2002, p. 371)

In Seattle, Babcock and Steiner (1999) conducted astudy that examined recidivism within a co-ordinated community response involving mandatoryarrest, the courts, probation officers and domesticviolence and chemical dependency service providers.Recidivism was measured by police records twoyears following the initial domestic violence incident,approximately one year after treatment completion.Despite the association between domestic violenceand substance abuse (Bennet, 1997), the authorsnote that this was the first study to explore theimpact of chemical dependency treatment ondomestic violence recidivism. Controlling fordemographic differences and differences in priorcriminal history, the study compared 387 domesticviolence perpetrators who were court mandated toeither domestic violence group treatment, domesticviolence treatment plus chemical dependencytreatment, or incarceration. Thirty one per cent ofthe men mandated to domestic violence completed it,as did 43 per cent of men mandated to chemicaldependency treatment. The results suggest that

completing domestic violence treatment is related toa statistically significant reduction in criminalrecidivism during a two year follow-up. Those whowere mandated to treatment but who failed tocomplete it, were more likely to reoffend. The studyfound little evidence that chemical dependency treat-ment is related to a reduction in domestic violence.

In the USA, a large multi-site evaluation using aquasi-experimental design was funded by theNational Centre for Injury Prevention and Control,Centres for Disease Control and Prevention. Itsfindings have been released in a large number ofjournal articles (e.g. Gondolf, 1997a, 1997b, 1999a,1999b, 1999c, 2000a, 2000b, 2000d; Gondolf, Chang,& Laporte, 1999; Gondolf & White, 2000).17

In designing the study, the research team aimed toaddress many of the methodological problems identi-fied in earlier outcome studies and to be sensitive tothe concerns of those working with perpetrators andwith abused women. Key to the approach adoptedwas the recognition that perpetrator programs donot operate in isolation, but rather, are embedded ina broader context:

…batterer intervention programs are part of abroader intervention system. They depend on – or atleast are related to – arrest practices, courtprocedures, probation supervision, battered-women’sservices, and other community services. (Gondolf, 2002a, p. 2)

Hence, the evaluation is of the ‘batterer interventionsystem’ (Gondolf, 2002a, p. 2), rather than ofbatterer ‘programs’. The importance of this approachis emphasised by a number of evaluation studieswhich find that men can fail to attend courtmandated programs, with no consequence. Forexample, Palmer, Brown and Barrera, (1992) whoevaluated a court mandated treatment program,comment on their finding that men could opt out ofthe treatment with impunity:

The lack of pursuit of those subjects who failed toattend the required number of group sessions…raisesquestions about the legal system’s failure in responsi-bility to society in general and to the abused partnersof these men in particular. (Palmer et al., 1992, p. 282)

The Gondolf research design comprised severalinnovations, compared to previous studies. Theseincluded:

• Use of ‘naturalistic comparison’. Four programswere examined as conducted within theircommunities. Site visits by members of theresearch team explored the program approachand contextual changes over time.

• A multisite study. Research participants wererecruited from four sites in different geographical

An important feature of this study was

its use of the women’s reports as its

main indicator of outcome and in the

development of new tools which

attempted to more adequately capture

all aspects of abuse – the violence, its

impact, and the core dynamic of

coercive control.


areas and with different client racial and ethnicmixes – Pittsburgh, Houston, Dallas and Denver.

• Sites were selected with different formats andlengths. They ranged from a three month, pre-trial program to a nine month, post-adjudicationprogram with additional specialist componentssuch as personality assessment at intake, alcoholtreatment and a women’s services co-ordinator.However, the four programs had a commonstructure and format, meeting criteria for being‘well established’: compliance with the relevantstate standards, collaboration with the batteredwomen’s services in their communities, use of acognitive-behavioural approach; and beingoperational for five years or more with at least40-50 referrals per month.

• Uniform intake procedures were implemented ateach site. Each site used the same assessmenttools and compiled similar records about issuessuch as attendance and drop outs, contributing tothe large sample in the study.

• Longitudinal follow-up. Phone interviews wereconducted every three months with the men andtheir initial partners, initially for 15 months fromprogram intake, with a subsequent extension tofour years. The researchers argue that thefrequent follow-up assisted with locating studyparticipants, increased the accuracy of the datathrough less reliance on recall over long periodsand assisted to build rapport which facilitateddisclosure by the women.

• New partners were included in the follow-upwhere they were identified. This study found thatat least 20 per cent of men in the sample had anew partner for some period during the 48 monthfollow-up period.

• Both qualitative and quantitative data werecollected.

• The main measure of reassault was the women’sreports. The study aimed to achieve a 70 per centresponse rate18. These reports were furthersupplemented by arrest records, the men’s reportsand medical records of a sub-group of the women.

• Intervening variables were assessed, includingnot living together, no contact, shelter contexts,legal action, counselling, drug and alcoholtreatment, economic assistance, informal help

• Large sample size (840) compared to manyprevious studies. Of this sample, the majoritywere court mandated (82 per cent), rather thanvoluntary (Gondolf, 2002a, pp. 65-68).

In this study, the main outcome measure selectedwas reassault. Reasons for this included that this isthe focus of criminal law and that stopping theviolence is the key goal of perpetrator programs. The

use of reassault as an outcome measure has beenchallenged, particularly by women’s advocates, astoo narrow, in that it fails to address the fact thatdomestic violence represents a pattern of coercivecontrol achieved through the combination of tacticssuch as social isolation, threats and economicdeprivation (Almeida & Durkin, 1999). Whileacknowledging this concern, Gondolf argues thatassault is highly correlated with these other forms ofabusive behaviours, and with the severity of theimpact of abuse. A complex array of data wascollected in order to determine reassault. Themultiple measures of reassault, providing bothqualitative and quantitative data, provide a richerand more detailed picture of the women’s experiencethan the identification of ‘reassault’ as primaryoutcome measure initially suggests.

The cumulative reassault rates were calculated forall court mandated men, whether they completed theprogram or dropped out. Based on the women’sreports, 32 per cent of the perpetrators reassaultedduring the first, 15-month period of the follow-up.This increased to 37 per cent at the 30 monthsfollow-up and to 42 per cent by the 48 month follow-up. When adjusted for under-reporting by thewomen (Gondolf et al., 1999), using men’s reportsand arrest data, the percentages who reassaulted atthe three points in the follow-up were: 40 per cent at15 months; 45 per cent at 30 months; and 48 percent at 48 months (Gondolf, 2002a). Sixty per cent ofthe men who reassaulted a partner inflicted a bruiseor injury, and of the injured women, one in fivesought medical assistance. Two men murdered theirpartners, and one committed suicide.

From the cumulative reassault rates, it can be seenthat almost half the sample committed at least onereassault. However, a different slant on theoutcomes can be seen when the trends in reassaultover time are identified. This different view ispossible because of the longitudinal research design.This makes it possible to test the hypothesis thatmen may reduce their violence while under thesupervision of the program and the legal system (incases of mandated attendance), but that after this

The use of reassault as an outcome

measure has been challenged, particu-

larly by women’s advocates, as too

narrow, in that it fails to address the

fact that domestic violence represents a

pattern of coercive control achieved

through the combination of tactics such

as social isolation, threats and

economic deprivation.


period of surveillance they again employ theirabusive behaviours. In this respect the study’sfindings are illuminating:

Overall, the cumulative reassault rate showed that asubstantial portion of the men reassaulted, but thetrend of the reassault revealed a de-escalation overtime and a sustained cessation of violence for the vastmajority of men. (Gondolf, 2002a, p. 113)

The results of this study suggest that the majority ofmen in the programs ‘eventually do stop theirviolence, apparently for long periods of time.’(Gondolf, 2002a, p. 123)

• At 30 months, ‘more than 80% of the men hadbeen violence-free for at least a full year’. (p. 122)

• At 48 months, ‘nearly 90 per cent of the batterershad reportedly not reassaulted a partner in theprevious year; and three quarters had not beenassaultive for the previous 2.5 years’. (Gondolf,2002a, p. 122)

The different findings from looking at cumulativerates of reassault and trends in reassault over timeinitially appear contradictory. They suggest that ahigh proportion of men commit at least onereassault, but that much of this reoffending occursearly in their involvement with the program. This isunderstandable given the serious nature of theproblem which perpetrator programs are addressing,at a point at which the men have had only minimalexposure to the program. The finding that the risk ofreassault is highest when men are first in theprogram, has implications for the intensity of theprogram offered, the intensity of the legal supervi-sion of the men, and the level of support offered towomen in the early stages of perpetrator programs.Gondolf suggests, for example, that men might berequired to attend three or four times weekly for thefirst month or so.

The study identified a group of men who reassaultedrepeatedly and at dangerous levels throughout theentire follow-up period. The research team devotedconsiderable effort to establishing whether thisdangerous group of men could be identified, bylooking at the utility of commonly used risk assess-ment tools, personality profiles, and batterer typolo-gies. However, the predicative power of these toolswas found to be weak.

The repeat reassaulters – the most dangerous men inour evaluation – were not as readily distinguishableas we had thought and hoped. Remarkably, themajority appeared to be acceptable candidates forconventional batterer counselling. Most did not appearpathological, and only a small portion appeared to betruly psychopathic. The mode of violence among thesemen was also not substantially different from thoseof other men. The one striking difference was thattheir partners were less assertive and additional

intervention was less certain. Little was done toapprehend or stop the men from continuing theirviolence. (Gondolf, 2002a, p. 191)

What emerges about this group of men is that theywere able to continue to reassault and inflict seriousharm because of a system’s failure: they reassaultedand faced no consequences. Based on these findings,Gondolf suggests that, in contrast to currentpractice, where risk assessment is undertaken at thebeginning of intervention, risk assessment needs tocontinue throughout the intervention period, sincerisk markers such as drunkenness and the women’sperceptions of dangerousness emerged during thefollow-up periods.

A commonly voiced concern about perpetratorprograms is that the men may simply substitute andincrease other, non-physical forms of abuse tomaintain their coercive control over their partner(Gondolf, 1997a). In this study, the women wereasked about these other forms of abuse. Instrumentsused included an inventory of controlling behaviours(which included items such as kept from talking onthe phone, kept from spending time with friends,taking money from the person); verbal abuse;property damage; and threats (e.g. threatened to killany person, threatened to take away or harmchildren) (Gondolf, 1997b). Based on the women’sreports, all forms of non-physical abuse reduced overtime, along with reassault. However, they remainedat relatively high levels, and Gondolf recommendsthat programs develop better methods of addressingthese other forms of abuse. Notably, the researchersdid not ask women about sexual violence19.

The research also aimed to address a commoncriticism of perpetrator program evaluations: thatthey pay insufficient attention to the subjectiveexperience of the victims. Is a program successful,for example, if the man’s violent and abusivebehaviour decreases or stops, yet the womancontinues to live in fear? This issue was addressedby exploring the women’s subjective view of theirown well being using a ‘quality of life’ inventory. Thistool was developed through consultation withwomen’s advocates and women who had experienceddomestic violence (Gondolf, 1997b). The researchfound that: ‘The majority of women in our evalua-tion…indicated that they were “better off” or “feltsafer” and their experience corresponds with thedecreases in reassault and abuse.’ (Gondolf, 2002a,p. 127) The area of greatest concern reported by thewomen was financial, with over a quarter of thewomen reporting increasing financial problems atthe 15-month follow-up point. At the 48 monthfollow-up, 85 per cent of initial partners said thatthey felt ‘very safe’, and 84 per cent indicated that itwas ‘very unlikely’ that their partners would hit them.Unfortunately, little detailed information about theway in which ‘safety’ was measured, is supplied.20

demographic and behavioural differences between‘dropouts’ and ‘completers’, was undertaken todetermine whether these differences, rather thanthe program, may have accounted for differences inoutcome. Finally, the researchers conducted qualita-tive research on the perceptions of the men and thewomen about the program and about the process ofchange (Gondolf, 2000b). Ninety per cent of menreported using techniques learned in the programsto avoid reassault. They reported that they usedbehavioural techniques (e.g. time out) mostfrequently (50 per cent). Given the programs’ goalsin terms of attitudes about gender equality,disappointingly only five per cent of the menreported utilising notions of respect and empathy fortheir partners (Gondolf, 2000b). Qualitative dataalso showed that around fifty per cent of womenattributed changes in their partners to their partici-pation in the program. Another study (Heckert &Gondolf, 2000) found that perception of the likeli-hood of criminal justice sanctions on their own didnot prevent reassault or dropout. This leads to thefollowing conclusion:

Our analysis of the program effect suggested arelatively moderate-sized effect for the programsinvolved in our multisite evaluation. Attendingbatterer counselling as opposed to dropping outappeared to reduce the probability of reassaultaccording to cross-tabulation, effect-size analysis,logistic regression, and structural equation model.Thus, we might assume that the programs addedsomething beyond the practice of men just beingarrested and put on probation. (Gondolf, 2002a, p. 159-160)

One interesting and unexpected finding of this studywas that the outcomes were relatively equivalentacross the four program sites included in the study,despite the fact that the programs varied in length(from three to nine months) and that the longestprogram offered an additional array of specialistservice components. One possible explanation is thatthis finding could be due to the quality of programimplementation at the different sites. However,Gondolf (1999b) suggests that the equivalent successof the shortest program (Pittsburgh) may be due tocharacteristics of the broader system in which it islocated. This is a three month, pre-trial program,


Gondolf also notes that, corresponding with thegroup of men who assaulted at frequent and harmfullevels, there was a group of women who were ‘worseoff ’ (12 per cent at 15 months and six per cent at 30months) following their partner’s participation in aperpetrator program:

These women told the kind of horror stories that havegiven batterer programs a bad name. Their experi-ences are the ones that raise concerns about battererprograms among some battered-women’s advocatesand confront batterer program staff with the limita-tions of their efforts. (Gondolf, 2002a, p. 128)

The study identified some interesting outcomesabout matters which have frequently been cause forspeculation. On the issue of whether men go on toassault new partners, it found that:

The new partners in our follow-up were half as likelyto be reassaulted than the initial partners, and mostof the men who did reassault new partners had alsoreassaulted their initial partners. Rather thanreassaulting a new partner instead of the initial one,some men simply continue reassaulting from onepartner to the next. (Gondolf, 2002a, p. 118)

It is frequently assumed that men who attendprograms voluntarily will be more motivated tochange. However, in this study, the voluntary partic-ipants (approximately 18 per cent of the sample)were almost ‘twice as likely to drop out as the court-referred men (61% vs. 33%), and they reassaultedtheir partners at a significantly higher rate at the 15month follow-up (44% vs. 29%).’ (Gondolf, 2002a, p.119). He postulates that ‘coerced participation’ may,in fact, be essential for program completion.

Although only one of the four programs in this studyoffered culturally specific services, African Americanand Latino men in the study were found to havesimilar rates for reassault (with Latino men lowest)and similar rates for women’s perceptions of safety,as ‘Anglo’ participants.21

Since this study did not use an experimentalresearch design, it faces the challenge ofdemonstrating that the reduction in assaults andother forms of abuse is due to the men’s participa-tion in the programs. For example, it could beargued, that it is the intervention of the criminaljustice system (arrest, court appearance and disposi-tion) which is associated with the outcomes. In orderto test this, the study adopted three approaches. Inthe first instance, it used men who dropped out ofthe program (‘dropouts’) as a ‘quasi-control group’(Gondolf, 2002a, p. 136). This group’s reassault rateswere compared with those of men who attended forat least two months (‘completers’). It found that thecumulative reassault rate at 30 months was 55 percent for ‘dropouts’ and 36 per cent for ‘completers’.

Secondly, a complex statistical analysis of

It is frequently assumed that men who

attend programs voluntarily will be more

motivated to change. However, in this

study, the voluntary participants

(approximately 18 per cent of the

sample) were almost ‘twice as likely to

drop out as the court-referred men’.


with swift court involvement and regular courtreview: ‘In sum, the men were held more quickly anddecisively accountable for their behavior.’ (Gondolf,2002a, p. 203)

In summing up the findings of this comprehensiveresearch project, Gondolf (2002a, p. 199) concludesthat the major implication of the study is that thesystem matters. He argues that, rather than a refine-ment of therapeutic/educational techniques used inperpetrator programs, what is most needed issystem development, echoing the calls for thedevelopment of co-ordinated or integrated responsesto domestic violence (Holder, 2001).

The main challenge appears to be in making theexisting components of intervention work togethermore decisively and consistently. They need to holdmen accountable for their behavior…the notion of‘accountability’, so frequently urged in batterercounselling, warrants some reinforcement. (Gondolf,2002a, p. 218)

The failure of a prompt and certain response by thecriminal justice system to reassault, with its implica-tions for the safety of women, was identified in thisstudy as the main characteristic of situations wheremen continued to abuse and to inflict the greatestharm of their victims. Hence we turn now to adiscussion of studies which seek to evaluate theeffectiveness of the co-ordinated intervention systemas a whole.

Studies of perpetrator programswithin a co-ordinated response to domestic violenceShepard (1999) points out that, despite the emphasisin recent years on developing co-ordinated responsesto domestic violence, most evaluation effort to datehas focussed on the individual components of thatresponse, such as perpetrator programs, or arrest.However, a body of research which attempts to focusmore broadly is emerging.

Findings from an early study which explored theimpact of a community intervention project on re-offending, suggested that lower recidivism wasassociated with ‘police making arrests on first visitswith the use of mandated treatment by the courts asa consequence. The strength of this finding appearsto increase the longer men are monitored.’ (Syers &Edleson, 1992) More recently, a study found thateach additional intervention in a co-ordinatedcommunity response increased the reduction indomestic violence, measured by criminal justice re-offences (Murphy et al., 1998). Lower recidivism wasfound to be associated with the cumulative effects ofsuccessful prosecution, probation monitoring,receiving a court order for counselling, attendingcounselling and completion of counselling. The

offender’s prior domestic violence history did notpredict recidivism: ‘In brief, more serious offenderstended to experience more intervention; yet greatersystem intervention was associated with lowerrecidivism.’ (Murphy et al., 1998, p. 279) However,the researchers note that only four per cent of casesin the study were subject to the full range criminaljustice system responses.

The Seattle study described earlier (Babcock &Steiner, 1999) highlights the challenges inimplementing co-ordinated responses to domesticviolence. Less than a third of the men courtmandated to domestic violence treatment completedit. Fifty eight per cent of program ‘non-completers’ infact did not attend any sessions; and only 37 percent of men who did not attend mandated treatmenthad their probation revoked. The majority – 64 percent – received no legal consequence for failing toattend court-mandated treatment:

The failure of the majority of batterers to completetreatment despite the fact that it was court-mandatedrenders questionable the potency of the coordinatedcommunity response to batterers’ treatment.Treatment may be most effective when offered as onecomponent in a well-orchestrated coordinatedcommunity response with consistent, legalconsequences for noncompliance with treatment.(Babcock & Steiner, 1999, p. 55)

A recent study (Shepard, Falk, & Elliott, 2002)evaluated the effectiveness of a project which aimedto enhance the Duluth Domestic Abuse InterventionProject (DAIP). Despite the project’s internationalrecognition, which often focuses on its educationprogram for perpetrators, the authors stress that:

…the core of the DAIP has always been focused oninstitutional change to effectively coordinatecommunity responses to domestic violence. Thephilosophy has been that communities, rather thanindividuals, must be responsible for holding abusersaccountable for their violence and ensuring the safetyof victims. (Shepard et al., 2002, p. 552)

The study evaluated the impact of the enhancedDAIP (EDAIP), which involved criminal justicepersonnel and victim advocates collecting andsharing risk assessment data which was used todetermine the level of sanctions applied to domesticviolence perpetrators. Recidivism rates, measured bycriminal justice data, were compared for maleoffenders across several years: 1994 (prior toimplementation of the EDAIP), 1996 (pilot year forEDAIP), 1997 and 1998. Recidivism rates weresignificantly lower for 1997 and 1998, compared with1994 and 1996. The authors conclude that: ‘Theresults are encouraging in that there was evidence ofreduced recidivism rates with enhanced coordinatedresponses.’ (Shepard et al., 2002, p. 568)


This study also highlighted the challenges inimplementing coordinated responses to domesticviolence. Even when the EDAIP was to have beenfully operational, police followed EDAIP protocols inonly 37 per cent of cases and probation officers in 55 per cent of cases.

To date, the only Australian research which hasattempted to address the wider context of theintervention system is the evaluation of the ACTInteragency Family Violence Intervention Program(Keys Young, 2000; urbis keys young, 2001b). TheFVIP is a co-ordinated criminal justice approachwhich emphasises ‘improved investigative practiceimbedded within a project comprising detailedtraining, close management and monitoring, newtechnology, and closer working relationships betweenthe Office of the DPP and the non-governmentDomestic Violence Crisis Service (DVCS)’. (Holder,2001, p.14) The Perpetrator Education Program(PEP) component of the FVIP began in September1998. Although the evaluation does not look at thelink between coordinated intervention and recidi-vism, it does provide a blueprint for systemiccomponents which need to be explored in evaluationswhich take a broader view than focusing on aperpetrator program in isolation. Among theemerging outcomes, it has been noted that:

The arrest rate of all incidents in the pilot patrolarea has increased from 16% to 27%. In Phase I ofthe FVIP about 70 per cent of criminal familyviolence matters proceeded to full hearing. In PhaseII, this has almost reversed with 60-70 per cent ofdefendants entering a guilty plea at the mentionstage (the first or second court appearance). Thissaves the victim from the stress of waiting time andappearance at court. (Holder, 2001, p. 14)

Some challenges forintervention withperpetrators

Avoiding a narrow focusSome commentators have argued against theemphasis on perpetrator programs as the keyresponse to men who use violence against theirpartners. For example:

While, at best, they [perpetrator programs] may stopindividual men using violence against familymembers, this does not equate, in our view, with theprevention of men’s violence. If we accept the analysisthat men’s violence is socially constructed, as well asindividually willed, strategies to prevent men’sviolence would need to be more oriented towardsstructural and social change rather than individualbehaviour change alone. We need to be alert to thedanger that men’s behaviour change programs may

be seen as ‘the solution’. (Pease & Fisher, 2001, p. 50)

In a similar vein, Bennett and Williams (2001) arguethat men’s programs should ally themselves withthose parts of the broader men’s movement whichare working to eliminate violence against womenand children and that they should be involved incommunity education to raise awareness about thesocial conditions that support men’s abusivebehaviour and which are expressed in a range ofways including bullying, sexual harassment anddating violence. They argue that there is currently animbalance in community education activities, withmost of this work left to women’s advocates: ‘Anunintended effect of this imbalance is an overem-phasis on the question “Why does she stay?” and anunderemphasis on the question “Why does hebatter?” and equally as important, “Why do we lethim batter.”’(Bennett and Williams, 2001, p. 275)

In arguing for a broader role for perpetratorprograms in the movement against domesticviolence, Bennett and Williams also suggest thatthis agenda should include programs moreadequately addressing what they term ‘responsiblefatherhood’ (p. 273). They suggest that modules inperpetrator programs on parenting are too often an‘afterthought’ and that an emphasis on genericparenting skills is less important than assisting aman to understand the impact of his violence on hischildren:

A responsible father is a man who, in addition tocaring for and guiding his children, is a coequal tohis children’s mother, even if he does not live withher; contributes to the support of his children regard-less of where they live; and accepts responsibility forhis controlling behavior and its effects on hischildren and his children’s mother. (Bennett &Williams, 2001, pp. 273-274)

This highlights the fact that the growing volume ofliterature about the links between child abuse anddomestic violence and about the impact of livingwith violence and abuse on children, focuses almostentirely on mothers (Edleson, 1998). While collabora-tion between child protection services and women’sdomestic services is regularly urged to address thisissue (e.g. Fleck-Henderson, 2000), Bennett andWilliams break new ground in urging that suchcollaborations be developed between perpetratorprograms and child protection services in order toidentify and respond to men whose behaviourthreatens the safety of children. In a similar vein,the Victorian framework for developing best practicein programs for men who use violence towardsfamily members, explicitly identifies the childprotection system as one of the pathways via whichmen may be directed to participate in a program:

Men may be directed to be assessed for theirsuitability to attend a program as part of a case plan


developed by a regional child protection unit, or asordered by the Children’s Court. Depending on theassessment, a man may then attend a MBCG,individual counselling, or be referred to other serviceoptions. A comprehensive report on the man’sattendance and participation in the program(s)would be made available by the program to DHS.(Office of Women’s Policy Victoria, 2001)

Developing responses toIndigenous family violenceWhen saying that ‘the system matters’, it isimportant to consider whether this system needsnecessarily to be the criminal justice system. Forgroups such as Indigenous people, who suffer racismand discrimination within the criminal justicesystem, urging an increased criminal justice role,and programs only within that context, may seemlimited and insensitive to the historical and socialcontexts in which family violence has developed as awidespread problem (Thompson, 2000). In theinternational literature, similar concerns have beenraised about building criminal justice based systemsfor men in an environment which ‘already overlycriminalizes men of color’. (Mederos, 1999) Mederosraises the possibility of other types of systems forbuilding accountability and safety:

Do we need to think more carefully of communityinterventions that are educational and cultural – ofintegrating discussions and awareness of domesticviolence in community settings such as schools,churches, social service agencies, and similar sites –rather than marginalizing those discourses toshelters and batterer intervention programs.(1999, p. 143)

A criminal justice based approach is seen by some asinconsistent with the community based responseemphasising family and community healing whichhas been identified by some Indigenous groups as apreferred way to develop a response to familyviolence. Such an approach more readily acknowl-edges the contributions of colonisation, racism anddispossession (Aboriginal and Torres Strait IslanderWomen’s Task Force on Violence, 2000). Forexample, Atkinson & Atkinson (2000) argue againstthe use of the terms ‘victim’ and ‘perpetrator’ in

addressing Indigenous family violence. High levels ofviolence are seen as reflecting the intergenerationaltransmission of trauma, to which a criminal justiceresponse represents a continuations of the damage,dispossession and trauma imposed by the process ofcolonisation:

In colonised societies there have been multiple layersof both acute and overt acts of violence, and theestablishment of chronic and covert conditions ofcontrol…In many instances the pain has beeninternalised into abusive and self-abusivebehaviours, often within families and communities.The rage is not only turned inwards, but compoundsacross generations and becomes complex. Thebureaucratic interventions of the state, the processesof law, social welfare and health care have notaddressed the core issue of human traumatisation…On the contrary, by their interventions, they haveacted to increase and compound the trauma bycreating and increasing victimisation anddependency on the State, criminal incarceration, andescalating violence against members of our ownfamilies and communities. (Atkinson & Atkinson,2000, pp. 143–144)

From this perspective, restorative justice 22

approaches have been proposed as an alternativeresponse to using the criminal justice system (e.g.Braithwaite & Daly, 1994). Restorative justiceapproaches draw on a range of processes such asmediation (regarded as inappropriate in situations ofdomestic and family violence where there are powerdisparities between participants) and the disputeresolution processes used by the Maori in NewZealand. The essence of this approach is that afteran offence where guilt is admitted, victims, offendersand their supporters meet together with a facilitatorand are encouraged to work out a plan for repairingthe damage and minimising harm from the incident:‘These interventions promise social justice throughhealing encounters between victims and offenders,sponsored by community members.’ (Presser &Gaarder, 2000, p. 175)

Despite the apparent ‘fit’ between restorative justiceapproaches and notions of community healing,serious questions about the applicability of restora-tive justice in situations of domestic and familyviolence have been raised (e.g. Greer, 2001; Martin,1996; Stubbs, 1997). Restorative justice approachesdeveloped in the field of juvenile justice. Familyviolence is a very different context in that theviolence is part of an ongoing pattern of behaviour,rather than a one-off crime committed by a stranger.Restorative justice assumes that hearing the victim’sperspective will promote shame in the offender, andpromotes the concept of the offender’s apology. Yetmany women living with domestic and familyviolence are familiar with ‘apologies’ as a part ofwhat Walker (1977–78) identified as the ‘cycle ofabuse’ (Stubbs, 1997). As Stubbs points out, it cannot

‘If we accept the analysis that men’s

violence is socially constructed, as well

as individually willed, strategies to

prevent men’s violence would need to

be more oriented towards structural and

social change rather than individual

behaviour change alone.’


be assumed that a woman who has been subjected toviolence and abuse will be willing or able to talkabout that abuse in a conference, and indeed it maybe unsafe for her to do so. Nor can it be assumedthat the ‘community’ will totally condemn violence,rather than minimise it and join with the victimblaming by the offender. Martin cautions about theconsequences of restorative justice approaches in thefollowing terms:

We must be careful that proposals to have familyviolence dealt with by the ‘community’ or criticisms ofpolice arrest policy do not result in the decriminalisa-tion of family violence and a return to the viewing offamily violence as a private matter or ‘just adomestic.’ (Martin, 1996, p. 59)

At present in Australia, programs for Indigenousand non-Indigenous men appear to be developingalong very different lines. Writing of the situation inAotearoa/New Zealand, Balzar (1999) describes theprocess for Maori activists in the refuge movementto support the pilot of a ‘Duluth’ type program inHamilton. The socio-political perspective informingthe Duluth program was attractive, not solelybecause it challenged victim-blaming, individualisticperspectives, but because:

…it also allowed an analysis that accounted for theimpact of colonisation on Maori life and violenceagainst Maori women by Maori men…The Duluthreport had widespread appeal to both Maori andnon-Maori women within the refuge movement as itprovided a framework that explained the currentstatus of women and Maori within New Zealand, astatus that evolved through the culturally supportedsubjugation of both groups. (Balzar, 1999, p. 240)

Other Maori groups, in contrast, were concernedthat the approach would legitimate increasedinvolvement of the State in the lives of Maori. Whatemerged from the debates, was a clear imperativethat a ‘Maori-controlled, Maori-centred program beoffered as part of the overall project’. (Balzar, 1999,p. 250)

Including cross culturalperspectivesPerpetrator programs are frequently criticised forfailing to address cultural diversity in theircurricula, staffing and understandings of domesticviolence (e.g. Bennett & Williams, 2001; Gondolf,1997a). Taft et al. (2001) note that despite theconsiderable interest in the issue of dropout rates(e.g. Gerlock, 2001; Rondeau, Brodeur, Brochu, &Lemire, 2001; Rooney & Hanson, 2001) fromperpetrator programs – estimated at 40-60 per cent– racial differences in treatment dropout haveattracted little research attention. In a study whichexplored the differences between African Americans

and Caucasians in attendance at a domestic violenceprogram, taking into account other demographicfactors which might account for differences, ‘racewas the strongest predictor of treatment dropout andnumber of treatment sessions completed byindividual members... Only 55% of African Americanclients, versus 79% of Caucasian clients, attendedthree fourths or more of scheduled sessions’. (Taft etal., 2001, p. 395-396) Similarly, in their evaluation ofthe Seattle Co-ordinated Community response,Babcock and Steiner (1999) noted that ethnicminorities comprised a high (but not quantified)proportion of program non-completers. They suggestthat this finding points to a need to make programsmore culturally sensitive, or to tailor treatmentspecifically to the needs of culturally and linguistically diverse groups.

Since research indicates that maintainingattendance in programs is linked with more effectiveintervention (Gondolf, 2002a; Taft et al., 2001),program failures to engage men from marginalisedgroups have important implications for the safety ofwomen. Further, the over-representation of menfrom minority cultural groups in the criminal justicesystems suggests that mandated court programsshould actively pursue ‘cultural competence’23

(Bennett & Williams, 2001).

This aspect of work with perpetrators is similarlyunder-developed in Australia. A literature reviewundertaken by the Clearinghouse in 2001 identifiedonly two programs specifically working with menfrom non English speaking backgrounds.Bhattacharjee (2000) writes about working with menfrom the Lao, Vietnamese, Khmer and Chinesecommunities on issues of child abuse and domesticviolence and argues the need for a differentapproach from that adopted by mainstream services,such as the use of conjoint counselling. An ethno-specific group for men who abused their partnerswas developed and delivered in cooperation with theVietnamese community (Melvin, 1998; Melvin,Muller, Chapman, Shine, & Edwards, 1999). In theAustralian literature, the utility of simply ‘adapting’mainstream programs, as opposed to working withcommunities to develop suitable programs, has beenquestioned (McCallum, 2000).

In the international literature, Almeida and Durkin(1999) are among the few to address the issue ofcultural diversity in perpetrator programs. They

...The over-representation of men from

minority cultural groups in the criminal

justice systems suggests that mandated

court programs should actively pursue

‘cultural competence’.


describe what is termed the ‘cultural context model’for intervention. This work is unique in describingan approach with men who use violence which workssimultaneously at the individual and socio-culturallevels of intervention. Links are formed withrelevant communities to recruit ‘cultural consul-tants’ who participate in the intervention system(1999, p. 319). Socio-cultural education process viaseparate ‘culture circles’ for women and men are thefirst stage of intervention. This model addressesintimate partner abuse within a context which alsoexplores the contexts of migration history, colonisa-tion and racism and so provides an example of anapproach which addresses ‘multiple oppressionswhile addressing personal responsibility’. (Almeida& Bograd 1991 cited in Bograd, 1999, p. 284)

Increasing the safety of women andchildrenSince most programs state that increasing the safetyand well being of women and their children is theirkey goal (Keys Young, 1999), developing methodolo-gies to assess the extent to which this is achieved, isclearly an important priority. Two of the studiesdiscussed in this paper addressed this goal by usingwomen’s reports as their main outcome measure ofprogram effectiveness. Both studies (Dobash et al.,1999; Gondolf, 2002a) gathered qualitative data andalso developed tools for collecting standardised dataabout women’s feelings of safety and quality of lifefollowing their partners’ participation in a program.

Austin and Dankwort (1999a) and Gregory and Erez(2002) note that women’s reports are often includedin studies of perpetrator programs with the limitedpurpose of verifying the men’s self reports of re-assault and change. They suggest that the percep-tions of partners of men in perpetrator programsshould be studied with a broader focus to betterunderstand women’s perspectives on these programs.

Austin and Dankwort interviewed 25 women withina year of their partners completing a CanadianBatterer Intervention Program (BIP) which wasattached to a women’s refuge and based on themodel developed by Jenkins (1990). All but three ofthe men participated voluntarily in the program. Atthe time of the interview, 16 women were living withtheir partners, and nine were separated from them.The results indicated that most of the women experi-enced some positive changes in their lives after theirpartners completed the program. Although most ofthe women indicated that they felt safer:

…most women said that they still remainedsomewhat fearful of their partners, even if they feltsafer. More than half the women stated that theywould likely never feel completely safe for tworeasons. First they feared that changes their partnershad made would not last, indicating that feeling

safer might be only a temporary condition….Second,the women feared that a reduction of the abuse didnot mean it would ever cease completely, introducingthe idea that a relative change in safety was notadequate for the security and well-being theyrequired. (p. 33)

Only two women reported that they felt no increasein personal safety with their partners. In both cases,the women said that they were separating from theirpartners and that they had been supported in thisdecision by the program counsellors. These womenreported that the program had helped them to makethe decision not to continue the relationship. Onesaid: ‘The program saved my life. I would have takenhim back, and taken him back, and eventually, Iwould have been killed’. (Austin & Dankwort, 1999a,p. 34) This outcome demonstrates the complexity ofevaluating program effectiveness. These two menwere judged by the program counsellors to havemade no positive changes, yet the outcome reportedby the women was extremely positive in addressingthe goal of increasing safety.

Three quarters of the respondents reported a senseof enhanced well-being after their partner’s involve-ment in the program, this being associated withfactors such as increased positive self-esteem andfeelings of empowerment or relief that the programwas now shouldering some of the burden of support-ing their partners. Other benefits described by thewomen were the validation which they received fromthe program counsellors, the fact that they were notblamed for the abuse they had experienced, and theinformation which they received about the abuse,such as the abusive nature of forms of behaviourother than physical assault. The authors conclude:

These findings additionally illustrate how BIPs canbe experienced as a positive resource by batteredwomen, even if their partners make few, if any,changes towards non-violence. For example, women’sfeelings of validation and their increased knowledgeof abuse were evidence of the strategic role that BIPscan play in providing battered women with crucialinformation, validating their realities of abuse, andassisting them in acquiring a sense of trust in theirown capability to make decisions about theirlives…once their male partners are unequivocallyidentified as solely responsible for their abusivebehaviour by program facilitators, battered womencan feel a greater sense of self-worth and gainempowerment. (Austin & Dankwort, 1999a, pp. 38-39)

Gregory and Erez (2002) conducted a similarqualitative study, interviewing 33 women about theirreactions to their partners’ participation in aperpetrator program. However, in this study, themen’s participation in the program was courtmandated. A majority of the women reported positiveoutcomes from their partner’s participation. Forexample, 55 per cent of the women reported that the


program completely eliminated physical violence.Nevertheless, in a finding similar to the previouslydescribed study, ‘most women remained cautious anddid not completely rule out further abusiveoutbursts’. (p. 227) Given that the men’s participa-tion was court mandated, it is important to note thefinding that over three quarters of the respondentsreported experiencing fear and anxiety at the timethe man commenced the program, because of hisanger at being compelled to attend. Hence, theauthors argue that safety planning and supportservices need to be actively offered to women in theinitial stages of their partner’s involvement in aperpetrator program.

The evaluation of the second phase of the ACTFamily Violence Intervention Program (urbis keysyoung, 2001b) includes a survey of victims. Thesurvey explored the nature of the abuse experiencedby the women, their perceptions of the policing andprosecution of their case and their views about theoutcome of their case. It also explored the women’sperceptions of safety from further family violenceafter the completion of the case:

Most victims (57%) felt either ‘very safe’ (18%) or‘fairly safe’ (39%). However, a substantial proportion(41%) reported feeling ‘not very safe’ (23%) or ‘not atall safe’ (18%). (urbis keys young, 2001b, p. 83)

The finding of the Gondolf study that very fewwomen partners of men in the perpetrator programsstudied accessed assistance for themselves – despiteexperiencing long-term, severe violation and abuse(Gondolf, 2002b) – raises many challenges andquestions for perpetrator programs and for women’sservices. Each of the programs in the study was partof a coordinated community response to domesticviolence and it was anticipated that women’s helpseeking would be enhanced through their contactwith victim advocates in the criminal justice system.Gondolf notes that the extensive literature onwomen’s help seeking has not to date specificallyaddressed the women partners of men in perpetratorprograms. At the 15-month point in the follow-upstudy, Gondolf found that only eight per cent of thewomen partners had contact with a women’sprogram during more than one of the three monthfollow-up periods. Further, this contact was‘primarily in response to repeated reassaults…mostly “reactive” as opposed to preventive’. (Gondolf,2002b, p. 219) In exploring the reasons for this, thestudy found that 59 per cent of respondents reportedthat they did not perceive a need for additionalservices, and almost half of these reported that theyused support from family and friends and churchmembers rather than from specialist domesticviolence services. Further, most of the women (70per cent) reported that they did not want additionalservices from the perpetrator programs.

A number of explanations are canvassed for the low

service utilisation by women partners of men inperpetrator programs, given the severe and long-term nature of the abuse most had experienced priorto their partner’s referral to the programs. It mayreflect the reduction in the men’s violence and abuse,the effects of having been subjected to severe abuse,or the women’s sense of self determination andagency.24 Only one of the four sites (Denver) in theGondolf study employed women’s service co-ordina-tors at the time of the study. Women partners of menin this program were found to be twice as likely toseek additional counselling or support services(Gondolf, 2002a). Consequently, Gondolf recommendsthat increased resources be made available foroutreach services to the women partners of menreferred to perpetrator programs.

The Central Violence Intervention Program (VIP) inAdelaide is an example of an Australian coordinatedresponse to domestic violence which includesoutreach to women (The Central ViolenceIntervention Program, 2000). The VIP is anintegrated, collaborative and co-ordinated intera-gency response which includes the criminal justicesystem. Key aims are to promote the safety ofwomen and children and to challenge men’s use ofviolence against women and children (CentralViolence Intervention Program, 2002). Assistance isoffered to women on referral of their (ex)partner bythe court, and a service is provided whatever theoutcome of their partner’s assessment for, or partici-pation in, the program. A recent evaluation (Power &Kowanko, 2001) demonstrated the potential forwomen’s programs which are part of a co-ordinatedresponse to achieve their goals in relation to the coreprinciples of promoting safety, responsibility, respectand empowerment. However, the evaluationhighlighted the necessity for this intensive workwith women to be adequately resourced. This isparticularly important given the finding that 50 percent of the women who received a service from theCentral Violence Intervention Program had notpreviously had contact with a domestic violenceservice, highlighting the importance of outreach:

Most of the women were directly contacted by theCVIP women’s worker, and acknowledged that they

The finding of the Gondolf study that

very few women partners of men in the

perpetrator programs studied accessed

assistance for themselves – despite

experiencing long-term, severe violation

and abuse...raises many challenges and

questions for perpetrator programs and

for women’s services.


would not have been assertive or confident enough tomake the initial contact themselves. (Power &Kowanko, 2001, p. 18)

The impact of perpetrator programs in women’s livescan be powerful, beyond the changes that theirpartners may make. For example, women who havebeen holding onto hope for change in their partnerreceive new information about his commitment totaking responsibility through the quality of hisparticipation in the program. For some women, thesupport offered to their partner by the programprovides a ‘breathing space’ in which to considertheir needs and to take stock of the relationship.One perpetrator program evaluation suggested thatfor some women, the man’s participation in aprogram may provide a way for her to leave therelationship safely (Shaw et al., 1999). Clearly,further research is required to better understand therange of impacts for women of their partners’ partici-pation in a perpetrator program.

ConclusionCurrently in Australia there is considerable interestin developing programs for perpetrators of domesticand family violence. Working with men who useviolence is one of the priority areas in phase two ofthe Partnerships Against Domestic Violence initia-tive, and policy and best practice frameworks arebeing developed by a number of States andTerritories (e.g. Domestic Violence Prevention Unit,2000; Golding, 2001; Office of Women’s PolicyVictoria, 2001). For many years it has been said thatthe effectiveness of such programs is unknown. Therecent international evaluation studies cited in thispaper provide sound guidelines about the ways inwhich evaluation strategies could be built into newlydeveloping programs to address the shortcomings ofprevious evaluations and to focus research onoutcomes for women and children.

It is also important to identify the ways in whichAustralian programs are developing which aredifferent to overseas models. An example is men’sgroup work using narrative models of therapy whichoffer new possibilities for engaging men beyond‘confrontation’ and ‘education’. (NorthernMetropolitan Community Health Service, 1997). Inthe Australian context, division of programs into‘mandatory’ and ‘voluntary’ does not encompassprograms which take referrals from the criminaljustice system, but which only accept men whochoose to participate following their referral andparticipation in an assessment process (e.g. CentralViolence Intervention Program, 2002).

The key message from the international research onprogram effectiveness is that, as Gondolf states soclearly, ‘the system matters’. Programs for perpetra-tors are but one component of the co-ordinated

response which is required to hold men accountableand to enhance the safety of women and children.This presents challenges for programs which operateoutside of the criminal justice context to buildalternative systems of accountability and to developlinks with women’s and other services. Whiledebates have focused on the relative value ofinterventions based on socio-political and psycholog-ical approaches, as outlined earlier in this paper,Gondolf (2002) suggests that the major focus ofefforts to intervene with men who use violence mustbe on building strong systems within which theseprograms are located.

At the start of this paper, debate about the levels ofintervention was discussed. In their recent book,Dobash et al. (2000, p. 5) address the controversyabout levels of intervention by framing efforts to endmen’s violence towards women as a ‘transformativeproject’, which is, they argue, ‘of necessity, bothbroad and narrow in focus and both general andspecific in nature’. They go on to outline this project:

The transformative project contains at least threegeneral arenas that must be effectively addressed inseeking a solution. It is simultaneously a project ofpersonal transformation for those who use violenceand for those who are its victims; it is a project ofinstitutional transformation for organizations thatprovide assistance to victims of violence or intervenewith respect to those who use violence; it is a projectof social and cultural transformation of publicorientations to this violence and the tolerance ofit…While this three-part vision may not be embodiedin every specific intervention or response, all threeparts must be present in the overall complement ofresponses that make up the wider transformativeproject the goal of which is to end this form ofviolence… (Dobash et al, 2000, pp. 5-6)

Feminists, who, through the refuge movement,began the modern movement against domesticviolence, provided the template for such a transfor-mative process. While they responded to the needs ofindividual women for safety and support, they linkedthis work with individual women to the need forchange at the broader societal level. They focused ongender inequity, the attitudes which support it andat the legal and institutional changes necessary toname domestic violence as a crime and to hold menaccountable and to make services responsive towomen’s needs for safety. It might well be asked ofany program that works with individual men toassist them to take responsibility for their violence,how well it articulates the connections of this workat the individual level with efforts at the institutional and social levels.


ReferencesAbel, E. M. (2001). Comparing the Social Service

Utilization, Exposure to Violence, and TraumaSymptomatology of Domestic Violence Female‘Victims’ and Female ‘Batterers’. Journal of FamilyViolence, 16(4), 401-420.

Aboriginal and Torres Strait Islander Women’s TaskForce on Violence. (2000). Aboriginal and TorresStrait Islander Women’s Task Force on ViolenceReport. Queensland: Department of AboriginalTorres Strait Islander Policy and Development.

Adams, D. (2000). The Emerge Program. In J. Hanmer& C. Itzin (Eds.), Home Truths About DomesticViolence: feminist influences on policy and practice:a reader (pp. 310-322). London: Routledge.

Almeida, R. V., & Durkin, T. (1999). The culturalcontext model: therapy for couples with domesticviolence. Journal of Marital & Family Therapy,25(3), 313-324.

Ashcroft, C. (2000). Naming Knowledge: A Languagefor Reconstructing Domestic Violence and SystemicGender Inequity. Women and Language, 23.

Atkinson, J., & Atkinson, C. (2000). Talking about‘Perpetrator Programs’. In R. Thompson (Ed.),Working in Indigenous perpetrator programs:proceedings of a forum, Adelaide, August 1999.Darwin: Ministerial Council for Aboriginal andTorres Strait Islander Affairs.

Austin, J. B., & Dankwort, J. (1999a). The impact of abatterers’ program on battered women. ViolenceAgainst Women, 5(1), 25-42.

Austin, J. B., & Dankwort, J. (1999b). Standards forBatterer Programs: A Review and Analysis.Journal of Interpersonal Violence, 14(2), 152-168.

Australian Bureau of Statistics. (1996). Women’sSafety Survey (Catalogue No. 4128.0.). Canberra:Australian Government Publisher.

Babcock, J. C., & Steiner, R. (1999). The RelationshipBetween Treatment, Incarceration, and Recidivismof Battering: A Program Evaluation of Seattle’sCoordinated Community Response to DomesticViolence. Journal of Family Psychology, 13(1), 46-59.

Balzar, R. (1999). The Hamilton Abuse InterventionProject: The Aotearoa Experience. In M. Shepard &E. Pence (Eds.), Coordinating community responsesto domestic violence: lessons from Duluth andbeyond. Thousand Oaks, California: SagePublications.

Baum, F., Brand, R., Colley, D., & Cooke, R. (1987).Preventing Family Violence: The Evaluation of aGroup for Men who are violent towards theirpartners. Australian Journal of Sex, Marriage andFamily, 8(4), 173-183.

Bennet, L. (1997, 23/9/97). Substance Abuse andWoman Abuse by Male Partners. NationalElectronic Network on Violence Against Women.

Available:http://www.vawnet.org/VNL/library/general/AR_substanc.html [2002, 31/05/02].

Bennet, L., & Williams, O. (1998, 18/04/02).Controversies and Recent Studies of BattererIntervention Program Effectiveness. ViolenceAgainst Women Online Resources. Available:http://www.vaw.umn.edu/Vawnet/AR_bip.htm13/05/02].

Bennett, L. W., & Williams, O. J. (2001). InterventionPrograms for Men Who batter. In C. M. Renzetti &J. L. Edleson (Eds.), Sourcebook on ViolenceAgainst Women (pp. 261-277). Newbury Park:Sage.

Bhattacharjee, S. (2000). Multicultural men andfamily relationships services – challenges andsuccesses. Family Services Australia Newsletter, 23,38-43.

Blagg, H. (2001). Pilot Counselling Programs forMandated and non-Mandated Indigenous Men –Research and Program Development. Perth: WADomestic Violence Prevention Unit.

Bograd, M. (1988). Feminist Perspectives on WifeAbuse: An introduction. In K. Yllo & M. Bograd(Eds.), Feminist Perspectives on Wife Abuse.Newbury Park: Sage Publications.

Bograd, M. (1999). Strengthening Domestic ViolenceTheories: Intersections of Race, Class, SexualOrientation, and Gender. Journal of Marital andFamily Therapy, 23(3), 275-289.

Boyle, M. (2001). Men, violence and change. Seekingsolutions: Australia’s inaugural domestic violenceand sexual assault conference, Combined Women’sCrisis Services Gold Coast Inc. Available:http://www.austdvclearinghouse.unsw.edu.au/PDF%20files/Boyle,Mick.pdf [2002, 25/06/02].

Braithwaite, J., & Daly, K. (1994). Masculinities,violence and communication control. In T.Newburn & E. A. Stanko (Eds.), Just boys doingbusiness?: men, masculinities and crime (pp. 189-213). London: Routledge.

Buttell, F. P., & Jones, C. (2001). InterpersonalDependency Among Court-Ordered DomesticViolence Offenders: A Descriptive Analysis. Journalof Family Violence, 16(4), 375-384.

Campbell, J., Rose, L., Kub, J., & Nedd, D. (1988).Voices of strength and resistance: A contextual andlongitudinal analysis of women’s responses tobattering. Journal of Interpersonal Violence, 13(6),743-762.

Cavanagh, K., Dobash, R. E., Dobash, R., P. , & Lewis,R. (2001). ‘Remedial Work’: Men’s StrategicResponses to Their Violence Against IntimateFemale Partners. Sociology, 35(3), 695-714.

Central Violence Intervention Program. (2002).Central Violence Intervention Program: ProgramManual. Adelaide: Central Violence InterventionProgram.


Cheetham, J. (2001). NSW Statewide campaign toreduce violence against women: ‘Violence AgainstWomen: It’s Against All the Rules’. Available:http://www.austdvclearinghouse.unsw.edu.au/PDF%20files/CheethamJacquie.pdf [2002, 20/06/02].

Crenshaw, K. (1991). Mapping the margins: intersec-tionality, identity politics and violence againstwomen of color. Stanford Law Review, 43, 1241-1299.

Demos, S., Rudd, P., Gzik, J., & Griffiths, P. (1998).Northern Territory court-mandated and court-referred programs for offenders of domestic andAboriginal family violence: background paper to thedevelopment of the program. Darwin: Office ofWomen’s Policy.

Dobash, R. E., Dobash, R., P., Cavanagh, K., & Lewis,R. (Eds.). (2000). Changing violent men. ThousandOaks, Calif: Sage Publications.

Dobash, R. E., & Dobash, R. P. (1992). Women, violenceand social change. London: Routledge.

Dobash, R. E., & Dobash, R. P. (1997). Men’s violenceand programs focused on change. Current Issues inCriminal Justice, 8(3), 243-262.

Dobash, R. E., & Dobash, R. P. (1998). Violent menand violent contexts. In R. E. Dobash & R. P.Dobash (Eds.), Rethinking violence against women(pp. 141-168). Thousand Oaks, CA, USA: SagePublications, Inc.

Dobash, R. E., & Dobash, R. P. (2000). EvaluatingCriminal Justice Interventions for DomesticViolence. Crime & Delinquency, 46(2), 252-271.

Dobash, R. P., Dobash, R. E., Cavanagh, J., & Lewis,R. (1999). A Research Evaluation of BritishProgrammes for Violent Men. Journal of SocialPolicy, 28(2), 205-233.

Dobash, R. P., Dobash, R. E., Cavanagh, K., & Lewis,R. (1998). Separate and intersecting realities: acomparison of men’s and women’s accounts ofviolence against women. Violence Against Women,4(4), 382-414.

Domestic Violence Prevention Unit. (2000). BestPractice Model for The Provision of Programs forPerpetrators of Domestic Violence in WesternAustralia. Perth: Domestic Violence PreventionUnit, Office of Women’s Policy.

Dunford, F. W. (2000). The San Diego Navy experi-ment: an assessment of interventions for men whoassault their wives. Journal of Consulting &Clinical Psychology, 68(3), 468-476.

Dutton, D. G. (1995). The domestic assault of women:Psychological and criminal justice perspectives (rev.& exp. ed.). Vancouver, BC: University of BritishColumbia Press.

Dutton, D. G. (1995). Trauma symptoms and PTSD-like profiles in perpetrators of intimate abuse.Journal of Traumatic Stress, 8(2), 299-316.

Dutton, D. G. (2000). Witnessing parental violence as

a traumatic experience shaping the abusivepersonality. Journal of Aggression, Maltreatment &Trauma, 3, 59-67.

Edleson, J. L. (1995). Do batterers’ programs work?Available:http://www.mincava.umn.edu/papers/battrx.htm[2001, 3 May].

Edleson, J. L. (1998). Responsible Mothers andInvisible Men: Child Protection in the Case ofAdult Domestic Violence. Journal of InterpersonalViolence, 13(2), 294-298.

Eglington, P. (1999, July). Can a coordinatedcommunity response to domestic violence assistwomen accessing the police in a domestic violencesituation? Paper presented at the SecondAustralasian women and policing conference, co-hosted by the Australasian Council of Women andPolicing Inc., and the Queensland Police Service,University of Queensland, Brisbane, Queensland.

Eisikovits, Z. C., & Edleson, J. L. (1989). Interveningwith Men who Batter: A Critical Review of theLiterature. Social Service Review, 63(3), 384-414.

Feder, L., & Dugan, L. (2002). A Test of the Efficacy ofCourt-Mandated Counseling for Domestic ViolenceOffenders: The Broward Experiment. JusticeQuarterly, 19(2), 343-375.

Feder, L., Jolin, A., & Feyerherm, W. (2000). LessonsFrom Two Randomized Experiments in CriminalJustice Settings. Crime and Delinquency, 46(3),380-400.

Fleck-Henderson, A. (2000). Domestic Violence in theChild Protection System: Seeing Double. Childrenand Youth Services Review, 22(5), 333-354.

Frances, R. (1996). Ethical, methodological, andgender issues in assessing intervention programsfor violent men. Psychiatry, Psychology and Law,3(2), 143-151.

Frey, R., & Bellotti, C. (1995). Pro-feministeducational programs run by men for maleoffenders: is it wise? is it possible? In Sandra JEgger & Duncan Chappell (Eds.), Australianviolence: contemporary perspectives 2 (pp. 139-153).Canberra: Australian Institute of Criminology.

Gerlock, A. A. (2001). A profile of Who Completes andWho Drops Out of Domestic ViolenceRehabilitation. Issues in Mental Health Nursing,22, 379-400.

Golding, S. (2001). Strategic Pathways for StoppingViolent Behaviour Programs in South Australia:Context Paper for the Implementation Plan.Adelaide: Stopping Violent Behaviour ProgramsWorking Group, Crime Prevention Unit.

Goldner, V. (1999). Morality and Multiplicity:Perspectives on the Treatment of Violence inIntimate Life. Journal of Marital and FamilyTherapy, 25(3), 325-336.

Goldner, V., Penn, P., Scheinberg, M., & Walker, G.


(1990). Love and Violence: Gender Paradoxes inVolatile Attachments. Family Process, 29(4), 343-364.

Gondolf, E. W. (1997a). Batterer programs – what weknow and need to know. Journal of InterpersonalViolence, 12(1), 83-98.

Gondolf, E. W. (1997b). Patterns of reassault inbatterer programs. Violence & Victims, 12(4), 373-387.

Gondolf, E. W. (1999a). Characteristics of court-mandated batterers in four cities: Diversity indichotomies. Violence Against Women, 5(11), 1277-1293.

Gondolf, E. W. (1999b). A comparison of four battererintervention systems: Do court referral, programlength, and services matter? Journal ofInterpersonal Violence, 14(1), 41-61.

Gondolf, E. W. (1999c). MCMI-III results for battererprogram participants in four cities: Less ‘patholog-ical’ than expected. Journal of Family Violence,14(1), 1-17.

Gondolf, E. W. (2000a). A 30-Month Follow-up ofCourt-Referred Batterers in Four Cities.International Journal of Offender Therapy &Comparative Criminology, 44(1), 111-128.

Gondolf, E. W. (2000b). How batterer program partici-pants avoid reassault. Violence Against Women,6(11), 1204-1222.

Gondolf, E. W. (2000c). Human Subject Issues inBatterer Program Evaluation. In S. K. Ward & D.Finkelhor (Eds.), Program Evaluation and FamilyViolence Research (pp. 273-297). New York: TheHaworth Press.

Gondolf, E. W. (2000d). Mandatory Court Review andBatterer Program Compliance. Journal ofInterpersonal Violence, 15(4), 428-437.

Gondolf, E. W. (2001). Limitations of experimentalevaluation of batterer programs. Trauma Violence& Abuse, 2(1), 79-88.

Gondolf, E. W. (2002a). Batterer Intervention Systems:Issues, outcomes and recommendations. ThousandOaks, CA: Sage Publications.

Gondolf, E. W. (2002b). Service Barriers for BatteredWomen With Male Partners in Batterer Programs.Journal of Interpersonal Violence, 17(2), 217-227.

Gondolf, E. W., Chang, Y. F., & Laporte, R. (1999).Capture-recapture analysis of batterer reassaults:an epidemiological innovation for batterer programevaluation. Violence & Victims, 14(2), 191-202.

Gondolf, E. W., & Fisher, E. R. (1988). Battered Womenas Survivors: An alternative to treating learnedhelplessness. Massachusetts/Toronto: LexingtonBooks.

Gondolf, E. W., & White, R. J. (2000). ‘Consumer’recommendations for batterer programs. ViolenceAgainst Women, 6(2), 198-217.

Greer, P. (2001). Pathways to safety: an interviewabout Indigenous family violence with Pam Greer,Indigenous training and development consultantAustralian Domestic & Family ViolenceClearinghouse, Issues Paper 5, UNSW, Sydney.

Gregory, C., & Erez, E. (2002). The Effects of BattererIntervention Programs: The Battered Women’sPerspectives. Violence Against Women, 8(2), 206-232.

Hall, R. (2001). Pitfalls and challenges in work withmen who use violence against their partners. In B.Pease & P. Camilleri (Eds.), Working with Men inHuman Services. Sydney: Allen and Unwin.

Hamberger, L. K., Lohr, J. M., Bonge, D., & Tolin, D.F. (1996). A large sample empirical typology ofmale spouse abusers and its relationship todimensions of abuse. Violence and Victims, 11(4),277-292.

Hamberger, L. K., Lohr, J. M., Bonge, D., & Tolin, D.F. (1997). An empirical classification of motivationsfor domestic violence. Violence Against Women,3(4), 401-423.

Hanmer, J. (2000). Domestic violence and genderrelations. In J. Hanmer & C. Itzin (Eds.), HomeTruths About Domestic Violence: feminist influenceson policy and practice: a reader (pp. 9-23). London:Routledge.

Heckert, D. A., & Gondolf, E. W. (2000). The effect ofperceptions of sanctions on batterer programoutcomes. Journal of Research in Crime &Delinquency, 37(4), 369-391.

Hewitt, L., & Cavanagh, J. (1998). Through the eyes ofthe children. Melbourne: Southern Family Life.

Holder, R. (2001). Domestic and Family Violence:Criminal Justice Interventions. AustralianDomestic and Family Violence Clearinghouse Issues paper 3, UNSW, Sydney.

Holtzworth-Munroe, A., Bates, L., Smutzler, N., &Sandin, E. (1997). A brief review of the research onhusband violence: I. Maritally violent versusnonviolent men. Aggression & Violent Behavior,2(1), 65-99.

Holtzworth-Munroe, A., Meehan, J. C., Herron, K.,Rehman, U., & Stuart, G. L. (2000). Testing theHoltzworth-Munroe and Stuart (1994) batterertypology. Journal of Consulting & ClinicalPsychology, 68(6), 1000-1019.

Holtzworth-Munroe, A., & Stuart, G. L. (1994).Typologies of male batterers: Three subtypes anddifferences among them. Psychological Bulletin,116, 476-497.

Huss, M. T., & Langhinrichsen-Rohling, J. (2000).Identification of the psychopathic batterer: Theclinical, legal and policy implications. Aggression &Violent Behavior, 5(4), 403-422.

Jenkins, A. (1990). Invitations to responsibility: thetherapeutic engagement of men who are violent andabusive. Adelaide: Dulwich Centre Publications.


Keys Young. (1999). Ending domestic violence?Programs for perpetrators: Full report. Barton,ACT: National Crime Prevention, AttorneyGeneral’s Department.

Keys Young. (2000). Evaluation of ACT InteragencyFamily Violence Intervention Program: FinalReport. Canberra, ACT: ACT Department of Justiceand Community Safety.

Kraszlan, K., & Rebecca West, R. (2001). WesternAustralia Trials a Specialised Court. AlternativeLaw Journal, 26(4), 197-198.

Laing, L. (2000). Progress, Trends And Challenges InAustralian Responses To Domestic Violence,Australian Domestic And Family ViolenceClearinghouse, Issues Paper 1, UNSW, Sydney.

Laing, L. (2001). Domestic violence: emergingchallenges. Paper presented at the NationalOutlook Symposium on Crime in Australia: newcrimes or new responses, Canberra, AIC.

Laming, C. (2000). The SHED Project: A ruralintervention project to prevent men’s violenceagainst women and children. In W. Weeks & M.Quinn (Eds.), Issues Facing Australian Families.Longman.

Lipchik, E., Sirles, E. A., & Kubicki, A. D. (1997).Multifaceted approach in Spouse Abuse Treatment.In R. Geffner & S. B. Sorenson & P. K. Lunberg-Love (Eds.), Violence and Sexual Abuse at home:Current issues in Spousal Battering and ChildMaltreatment (pp. 131-148). USA: The HaworthPress Inc.

Mankowski, E. S., Haaken, J., & Silvergleid, C. S.(2002). Collateral Damage: An Analysis of theAchievements and Unintended Consequences ofBatterer Intervention Programs and Discourse.Journal of Family Violence, 17(2), 167-184.

Martin, P. (1996). Restorative justice: a family violenceperspective. Social Policy Journal of New Zealand,6, 56-68.

Mauricio, A. M., & Gormley, B. (2001). MalePerpetration of Physical Violence Against FemalePartners: The interaction of dominance needs andattachment insecurity. Journal of InterpersonalViolence, 16(10), 1066-1081.

McCallum, S. (2000). Best Practice versusIntervention Research: The Northern Territory’sExperience. In R. Thompson (Ed.), Working inIndigenous perpetrator programs: proceedings of aforum, Adelaide, August 1999. Darwin, NT:Ministerial Council for Aboriginal and TorresStrait Islander Affairs.

Mederos, F. (1999). Batterer Intervention Programs:The past, and future prospects. In M. Shepard & E.Pence (Eds.), Coordinating community responses todomestic violence: lessons from Duluth and beyond.Thousand Oaks: Sage Publications.

Melvin, T. (1998, November). Family ViolenceIntervention and Research Project: preliminary

findings and service delivery implications. Paperpresented at the Changing Families, ChallengingFutures, 6th Australian Institute of Family StudiesConference, Melbourne.

Melvin, T., Muller, D., Chapman, A., Shine, R., &Edwards, R. (1999). A Study in Hope: The multi-site evaluation and development of a better practicemodel for family violence services. Canberra:Department of Family and Community Services.

Mort, F. (2001). Northern Violence InterventionProgram. Frontlines: a newsletter on violence,3(July), 6.

Mullender, A. (1996). Groupwork with Male ‘Domestic’Abusers: Models and Dilemmas. Groupwork, 9(1),27-47.

Muller, D. (1997). Aspects involved with evaluatingfamily violence programs. Domestic Violence andIncest Resource Centre Newsletter no, 1, 11-17.

Murphy, C. M., Musser, P. H., & Maton, K. I. (1998).Coordinated community intervention for domesticabusers – intervention system involvement andcriminal recidivism. Journal of Family Violence,13(3), 263-284.

Nankervis, J., & Donne, J. (1993). Men’s group forchange: challenging beliefs and behaviours. Paperpresented at the Australian Institute of FamilyStudies Fourth Australian Family ResearchConference, Manly NSW.

Northern Metropolitan Community Health Service.(1997). Stopping violence groups: a model of bestpractice for group work with men who wish to stopviolent and abusive behaviour towards their womenpartners and family. Adelaide: NorthernMetropolitan Community Health Service.

NSW Council on Violence Against Women. (1999).Position Paper on Programs for Perpetrators ofDomestic Violence. Sydney: Violence AgainstWomen Specialist Unit.

Office of Women’s Policy. (2000). Behaviour changeprograms for men who use violence towards familymembers: whole of government discussion paper.Melbourne: Victorian Office of Women’s Policy.

Office of Women’s Policy Victoria. (2001). TakingResponsibility: A Framework for developing bestpractice in programs for men who use violencetoward family members. Melbourne: Office ofWomen’s Policy Victoria and Partnerships AgainstDomestic Violence.

Orme, J., Dominelli, L., & Mullender, A. (2000).Working with violent men from a feminist socialwork perspective. International Social Work, 43(1),89-105.

Palmer, S. E., Brown, R. A., & Barrera, M. E. (1992).Group Treatment for Abusive Husbands: Long-Term Evaluation. American Journal OfOrthopsychiatry, 62(2), 276283.

Pease, B., & Fisher, S. (2001). Preventing Men’s


Violence: A response to the Victorian whole ofgovernment discussion paper on men’s behaviourchange programs. Women against violence: anAustralian feminist journal, 10, 49-53.

Pence, E., & Paymar, M. (1993). Education Groups forMen who Batter: The Duluth Model. New York:Springer Publishing Company.

Power, C., & Kowanko, I. (2001). Just don’t give up onus: an evaluation of the Central and NorthernViolence Intervention Programs. Adelaide: CentralViolence Intervention Program.

Poynter, T. L. (1991). An evaluation of a groupprogramme for male perpetrators of domesticviolence: a follow-up study. Australian Journal ofMarriage and Family, 12, 64-76.

Presser, L., & Gaarder, E. (2000). Can RestorativeJustice Reduce Battery? Some PreliminaryConsiderations. Social Justice, 27(1), 175-194.

Rondeau, G., Brodeur, N., Brochu, & Lemire, G.(2001). Dropout and Completion of TreatmentAmong Spouse Abusers. Violence and Victims,16(2), 127-143.

Rooney, J., & Hanson, K. (2001). Predicting AttritionFrom Treatment Programs for Abusive Men.Journal of Family Violence, 16(2), 131-149.

Saunders, D. G. (1996). Feminist-cognitive-behavioraland process-psychodynamic treatments for menwho batter: Interaction of abuser traits andtreatment models. Violence & Victims, 11(4), 393-414.

Scott, K. L., & Wolfe, D. A. (2000). Change AmongBatterers: Examining Men’s Success Stories.Journal of Interpersonal Violence, 15(8), 827-843.

Shaw, E., Bouris, A., & Pye, S. (1999). A comprehen-sive approach: the family safety model withdomestic violence. In J. Breckenridge & L. Laing(Eds.), Challenging silence: innovative responses tosexual and domestic violence. Sydney: Allen andUnwin.

Shepard, M. (1999). Evaluating CoordinatedCommunity Responses to Domestic Violence.National Electronic Network on Violence AgainstWomen. Available:http://www.vawnet.org/VNL/library/general/AR_ccr.html [2002, 31/05/2002].

Shepard, M., Falk, D. R., & Elliott, B. A. (2002).Enhancing Coordinated Community Responses toReduce Recidivism in Cases of Domestic Violence.Journal of Interpersonal Violence, 17(5), 551-569.

Sidey, M., & Lynch, T. (2001). Creating new choices: aviolence prevention project for schools in Australia.Geneva, Switzerland: International Bureau ofEducation.

Simoneti, S., Scott, E. C., & Murphy, C. M. (2000).Dissociative Experiences in Partner-AssaultiveMen. Journal of Interpersonal Violence, 15(12),1262-1283.

Stamp, G. H., & Sabourin, T. C. (1995). Accounting forviolence – an analysis of male spousal abusenarratives. Journal of Applied CommunicationResearch, 23(4), 284-307.

Stosny, S. (1995). Treating Attachment Abuse: ACompassionate Approach. New York: SpringerPublishing Company.

Stubbs, J. (1997). Shame, Defiance and ViolenceAgainst Women: A critical analysis of‘Communitarian Conferencing’. In C. S. Cook & J.Bessant (Eds.), Women’s Encounters with Violence:Australian Experiences (pp. 109-126). London: SagePublications.

Syers, M., & Edleson, J. L. (1992). The CombinedEffects of Coordinated Criminal JusticeIntervention in Woman Abuse. Journal ofInterpersonal Violence, 7(4), 490-501.

Taft, A., Hegarty, H., & Flood, M. (2001). Are Men andWomen Equally Violent to Intimate Partners?Australian and New Zealand Journal of PublicHealth, 25(6), 498-500.

Taft, C. T., Murphy, C. M., Elliott, J. D., & Keaser, M.C. (2001). Race and Demographic Factors inTreatment Attendance for Domestically AbusiveMen. Journal of Family Violence, 16(4), 385-400.

Taylor, B. (2000). ‘What’s a nice girl like you.....’: isthere a role for women’s services in the provision ofmen’s domestic violence programs? In PartnershipsAgainst Domestic Violence & Dept. of Family andCommunity Services (Eds.), Men and relationships:partnerships in progress conference. Sydney: Dept.of Family and Community Services, PartnershipsAgainst Domestic Violence.

Taylor, B. (2001). ‘Clearing pathways – reformingsystems’: a reflection of the achievements andexisting challenges of the Gold Coast domesticviolence integrated response. Domestic violence,action and resources, 11, 5-7.

Taylor, B. G., Davis, R. C., & Maxwell, C. D. (2001).The effects of a group batterer treatment program:A randomised experiment in Brooklyn. JusticeQuarterly, 18(1), 171-201.

The Central Violence Intervention Program. (2000).South Australia: Central Violence InterventionProgram: Information Bulletin. Adelaide.

Thompson, R. (Ed.). (2000). Working in Indigenousperpetrator programs: proceedings of a forum,Adelaide, August 1999. Darwin, NT: MinisterialCouncil for Aboriginal and Torres Strait IslanderAffairs.

Tolman, R. M., & Bennett, L. W. (1990). A Review ofthe Quantitative Research on Men Who Batter.Journal of Interpersonal Violence, 5(1), 87-118.

Townsend, L. (1991). Programs for perpetrators ofdomestic violence. Haymarket, NSW: NSWWomen’s Co ordination Unit.

Trute, B. (1998). Going beyond gender-specific


treatments in wife battering – pro-feminist coupleand family therapy. Aggression & Violent Behavior,3(1), 1-15.

Tweed, R. G., & Dutton, D. G. (1998). A comparison ofimpulsive and instrumental subgroups ofbatterers. Violence & Victims, 13(3), 217-230.

urbis keys young. (2001a). Evaluation of the Learningto Relate without Violence and Abuse Program:Final Report. Canberra: ACT Department ofJustice and Community Safety.

urbis keys young. (2001b). Evaluation of the ACTFamily Violence Intervention Program Phase11:Final Report. Canberra: ACT Department ofJustice and Community Safety.

Violence Against Women Specialist Unit. (2000).Violence against women: it’s against all the rules.Sydney: Violence Against Women Specialist Unit,NSW Attorney Generals Department.

Walker, L. E. (1977-78). Battered Women and LearnedHelplessness. Victimology, 2(3-4), 525-534.

White, M. (1989). The conjoint therapy of men who areviolent and the women with whom they live,Selected Papers (pp. 1-124). Adelaide: DulwichCentre Publications.

White, M., & Epston, D. (1989). Literate Means toTherapeutic Ends. Adelaide: Dulwich CentrePublications.

White, R. J., & Gondolf, E. W. (2000). Implications ofPersonality Profiles for Batterer Treatment.Journal of Interpersonal Violence, 15(2), 467-489.

Woodbridge, P. (1998). Dovetail: an integratedapproach to domestic violence. Domestic Violenceand Incest Resource Centre Newsletter, 2, 11-14.

Woodbridge, P. (2000). Working with men who useviolence: accountability to women? DomesticViolence and Incest Resource Centre Newsletter, 4,9-10.

Yllo, K. (1999). The Silence Surrounding SexualViolence. In M. Shepard & E. Pence (Eds.),Coordinating community responses to domesticviolence: lessons from Duluth and beyond.Thousand Oaks, California: Sage Publications.

Zuchowski, I. (1999). Catching a Dream and LettingPigs Fly: Schools respond to domestic violence.Domestic Violence and Incest Resource CentreNewsletter, 4, 25-27.

Endnotes1 This paper deals with men who are violent in

their intimate relationships since this is the mostcommon form of domestic violence (Taft, Hegarty,& Flood, 2001). An emerging body of literatureexplores the differences between male and femaleperpetrators of domestic violence and interestedreaders are referred, for example, to Hambergeret al. (1997). For initial research on differencesbetween women identified by the legal system as‘batterers’ and ‘victims’, see Abel (2001).

2 There is considerable debate about theterminology used to describe ‘domestic violence’,including use of the terms ‘victim’ and‘perpetrator’. For a discussion of these debatesand issues see Office of Women’s Policy Victoria(2001) and the introduction in Laing (2000). Insummary, the strength of the terms ‘victim’ and‘perpetrator’ is that they make clear the powerrelationship in which abuse occurs. However,they have the disadvantage of portraying womenas passive victims and men as abusers, therebynarrowing the perception of possibilities forchange. The term ‘perpetrator’ is the mostcommonly used one in Australia (Keys Young,1999), while in the USA, ‘batterers’ is commonlyused. This paper will use several termsinterchangeably as used in the references cited.

3 Generally the term ‘mandated’ refers toattendance at a program as part of an orderissued by a court, while ‘voluntary’ refers toattendance without legal requirement. InVictoria, the term ‘directed participant’ ispreferred to ‘mandated’ (Office of Women’s PolicyVictoria, 2001).

4 However, Mankowski, Haaken and Silvergleid(2002, p. 173) argue that the Duluth model is, infact, both therapeutic and skill building, ratherthan purely educational.

5 See Holder, R. 2001 ‘Domestic and FamilyViolence: Criminal justice Interventions’, IssuesPaper 3, Australian Domestic and FamilyViolence Clearinghouse, UNSW for a detaileddiscussion of co-ordinated community responses.

6 John Duncan, Central Violence InterventionProgram, personal communication.

7 Dissociation is a coping mechanism which isfunctional in situations of extreme traumabecause it enables detachment from feelingswhich would otherwise be overwhelming.However, what in childhood is a protectiveresponse in the face of traumatic events, becomesan unhelpful pattern of responding in adulthood.

8 See page 13

9 Response rates for men were low: 50% for thecontrol group and 61% for controls; and lower for


women: 30% of partners of men in the treatmentgroup and 11% of partners of controls.

10 There is contention about the use of this model.Participation rates by the women were very low.

11 Response rates at each of the four interviewtimes were: 86%; 82%; 78%; 75% (percentages forspouse or perpetrator were not suppliedseparately).

12 However, the study included only cases in whichthe prosecutor, defendant and judge agreed thattreatment was appropriate. Men enteringtreatment were not volunteers, as they wereoffered a plea agreement to enter treatment. Theauthors acknowledge that their recruitmentprocess probably filtered out men who wereunmotivated to attend a program. Feder andDugan (2002) point out that during the one yearperiod in which men were entered into this study,more than 11,000 domestic violence cases weredealt with in this jurisdiction, highlighting theimpact of this filtering on the sampling.

13 Men were referred to one of five battererprograms, each of which was county certified toprovide batterer treatment, and each of whichwas conducted in line with the Duluth model.

14 Survey completion rates for victims were 50% atadjudication; 30% for the first interview (sixmonths post adjudication); and 22% at 12months.

15 At intake, the ‘Men’s Program’ group comprised98 (51 men and 47 women) and the ‘Other CJ’group comprised 168 (71 men and 97 women).

16 See Holder, R. 2001 ‘Domestic and FamilyViolence: Criminal justice Interventions’, IssuesPaper 3, Australian Domestic and FamilyViolence Clearinghouse, UNSW for a detaileddiscussion of this body of research.

17 Summaries are available on the Internet, at theweb site of the Mid-Atlantic Addiction TrainingInstitute, Indiana University of Pennsylvania:www.iup.edu/maati/publications, and in a book(Gondolf, 2002a).

18 The study achieved a 68% response rate from thewomen at the 15-month mark, and a femalepartner of a man in the program was interviewedat least once during the 15-month follow-up in79% of cases. The extended follow-up (from 15-48months) included only the 618 court mandatedparticipants. The response rate for womenpartners of this group was 67% at 30 months and58% at 48 months.

19 The Duluth program has been critiqued forfailing to hold men accountable for sexualviolence, in the same way that physical violenceis addressed (Yllo, 1999).

20 For example, does safety refer to an absence ofdanger or do they look at the impact on women’slives and seek information on all aspects of thewoman’s safety including being able to haveopinions and being able to speak her mind?(Dallas Colley, personal communication).

21 Rates for reassault at 15 months post intakewere: African-American 32%, Anglo 31%, Latino,18%. Proportion of partners feeling ‘very safe’were Anglo, 74%; African American, 73% andLatino, 81%. The researchers noted that it wasdifficult to compare arrest rates because ofpolicing practices with African-American men atthe Pittsburgh site (Gondolf, 2002a, p. 120-121).

22 Restorative justice is known by a number of othernames including communitarian justice interven-tion, conferencing and family group conferencing.

23 The following definition is taken from the website of Centre for Effective Collaboration andPractice: ‘Cultural competence is defined as a setof congruent behaviors, attitudes, and policiesthat come together in a system, agency, or amongprofessionals and enables that system, agency, orthose professionals to work effectively in cross-cultural situations (Cross et al., 1989; Isaacs &Benjamin, 1991)…There are five essentialelements that contribute to a system’s ability tobecome more culturally competent. The systemshould (1) value diversity, (2) have the capacityfor cultural self-assessment, (3) be conscious ofthe ‘dynamics’ inherent when cultures interact,(4) institutionalize cultural knowledge, and (5)develop adaptations to service delivery reflectingan understanding of diversity between andwithin cultures. Further, these five elementsmust be manifested in every level of the servicedelivery system. They should be reflected inattitudes, structures, policies, and services.’http://cecp.air.org/cultural/ accessed 28/08/02).

24 The different views of abused women areexplored in Clearinghouse issues paper 4:‘Working with women: exploring individual andgroup work approaches’.

AcknowledgmentsThe following people’s contribution to this paper byreviewing the draft is gratefully acknowledged: JohnDuncan, Sarah Cartwright and Julie Felus, CentralViolence Intervention Program, Adelaide; PaulineWoodbridge, North Queensland Domestic ViolenceResource Service; Ros Adam, RelationshipsAustralia/WA; Elisabeth Shaw, RelationshipsAustralia/NSW; and Dallas Colley, Consulting &Training (Domestic Violence).

The author Dr Lesley Laing is Director of theAustralian Family and Domestic ViolenceClearinghouse, University of New South Wales.


ISSN: 1443-8496Published by the Australian Domestic and FamilyViolence Clearinghouse UNSW Sydney NSW 2052Ph: (02) 9385 2990 • TTY: 02 9385 2995 Fax: (02) 9385 2993 Email: [email protected] Website:http://www.austdvclearinghouse.unsw.edu.auClearinghouse Issues Papers are refereed publications.The views expressed in this newsletter do not necessarilyrepresent the views of the Commonwealth of Australia, theUniversity of New South Wales or the Partnerships AgainstDomestic Violence Taskforce. This work is copyright. Organisations have permission toreproduce parts or whole of this publication, as long as theoriginal meaning is retained and proper credit given.Whilst all reasonable care has been taken in the preparation ofthis publication, no liability is assumed for any errors or omissions.The Clearinghouse is linked to the Centre for Gender-RelatedViolence Studies, based in the University of New South WalesSchool of Social Work. Funded by Partnerships Against Domestic Violence, aCommonwealth Government initiative working with the States andTerritories and the community to find better ways of preventingand responding to domestic violence.

Des

ign:

ant

hous

e@ac

ay.c

om.a

u

Want to receive ClearinghouseNewsletters and Issues Papers?Fill out this form and we’ll put you on our mailing list.

Forthcoming in the Issues Paperseries in 2002:• Domestic Violence and Women’s Mental

Health

Previous Issues Papers:• Progress, trends and challenges in Australian

responses to domestic violence – IssuesPaper 1, 2000

• Children, young people and domestic violence– Issues Paper 2, 2000

• Domestic and Family Violence: CriminalJustice Interventions – Issues Paper 3, 2001

• Working with Women: Exploring individiualand group work approaches – Issues Paper 4,2001

• Pathways to Safety: An interview with PamGreer (Indigenous family violence) – IssuesPaper 5, 2001

• Violence against women in pregnancy andafter childbirth: Current knowledge andissues in health care responses – IssuesPaper 6, 2002

You can order copies of previous Issues Papersby contacting the Clearinghouse on:

Phone: (02) 9385 2990or email: [email protected] or fill in theform (right) to be placed on our mailing list forfuture publications.

✃■■■■ Please add my name to the mailing list to

receive Clearinghouse print publications

■■■■ Please send me information on entering myprogram on the Good Practice Database

Name

Position

Organisation

Address

Phone

Fax

Email

Please fill in and fax back to the Clearinghouse:

Fax: (02) 9385 2993

P/code

Australian Domestic & Family Violence CLEARINGHOUSE · 2016-10-24 · 2 Australian Domestic &...

Documents

Transcript of Australian Domestic & Family Violence CLEARINGHOUSE · 2016-10-24 · 2 Australian Domestic &...