search | site index | faq | about us | what's new
Publication Date: December 1995
Whether it's called "therapy", "treatment", "counseling", "education", "intervention" or simply "programs", support for perpetrators of wife abuse presently benefits from a relentless lobbying effort, one that rekindles confidence and expectations with the general public, with government and judicial decision-makers and, eventually, with abused women themselves, in spite of their own perceptions and choices. In times of financial crisis and anguish over any demise of The Family, government officials are easily interested by the apparent savings of dejudiciarizing the crimes that seem most easily amenable to psychologizing "explanations" and measures. Yet, in the words of R. Karl Hanson and Liz Hart, authors of The Evaluation of Treatment Programs for Male Batterers ( 1991 ):
If important decisions are going to be made based on whether a batterer attends treatment (e.g., partner stays or leaves, sentenced to jail or probation), then it becomes crucial to know the effectiveness of the treatment. (INTRODUCTION, p. i) 1
It is to this end that Hanson and Hart recently organized a conference, bringing together---with the help of Solicitor General Canada---most of the premier researchers in this field. The document quoted above offers this conference's proceedings. Montreal Men Against Sexism has undertaken to present in synthetic form significant quotes from these proceedings, in order to let conference participants attest in their own words to the limits and risks presented by such programs at this stage. [Emphasis is ours wherever sentences appear in bold.]
It will readily appear that most of the speakers concur that a strong and systemic judicial intervention and a victim-based mode of support prove to be more efficient means of achieving the prime objective of any intervention aimed at the perpetrator, e.g. getting him to stop battering and controlling that woman.
It is thus as much in the interest of men than in that of women that our society must show both realism and accountability in its response to the claims and experimental models put forward by the dejudiciarization lobby, under the guise of a psychologizing intervention. As newspaper headlines remind us daily, human lives are at stake.
Robert J. Brown, psychologist, Calgary General Hospital: ...private and government resources for dealing with family violence are extremely limited. As a result, it has become increasingly important to ensure that we use those resources in the very best places and in the most efficient ways possible. (73-4)
Finally, if we speak of a "psychologizing" approach, it is because a realistic psychological exploration of misogyny and of men's will to control remains generally absent from the "treatment" debate, because it is censored as a feminist analysis. What now presents itself as psychological theory and practice eludes the true dynamics of the situation and is much closer to masculinist politics than to a realistic and progressist analysis of the dynamics of sexist violence.
Martin Dufresne, for Mtl Men Vs Sexism
The issue of how efficient the various forms of psychological "treatment" offered to wife batterers actually are cannot be eluded or reduced to some standard give-it-a-further-try optimism. We hope that this will become self-evident as the readers ponders the various problems raised by program providers and analysts themselves.
Such is the judgment of James Browning, who produced in
1984 the first overall assessment of the limits and
contradictions of Canadian treatment programs for
batterers:
It should be noted that a proper evaluation is beyond the
resources of most individual programs, which are
struggling to simply provide a clinical service. It is
therefore incumbent upon government to provide resources
for an evaluation component... (94)
Daniel J. Saunders: ...testing treatment integrity is a major research investment: a coding scheme must be developed, coders need to be trained to reliability, and the coding itself is time-consuming. (6)
James Browning: In addition to the foregoing shortcomings in the evaluation literature, there have been few successful attempts to create an evaluation system for a series of programs. (94)
This lack of common ground renders empirically unjustifiable any appeal to "success statistics" in support of the general notion of "treatment" for batterers, especially in comparison to more efficient proven approaches.
Jurgen Dankwort, a Montreal social worker, has looked
closely at a vast sample of treatment programs for
batterers, both in Quebec and throughout Canada. Here are
a few of the conclusions of this study which he
communicated to conference participants:
Contrary to expectations, the research did not identify
techniques that would transmit different messages about
either the causes of violence, the reasons why men resist
changing, or the remedies required to stop the abuse.
(47-8)
Eisikovits and Edleson ( 1989 ) recently concluded, in their extensive review of existing studies, that to answer the pressing question of whether or no treatment for wife abusers "works", research must attempt to link outcome, theory and technique in a much more unified manner... (39)
The findings suggest that much needs to be done to improve group counselling to achieve the goals the program themselves identified. Furthermore, the findings clarified why the fundamental problems identified in this study are so important to developing intervention that is congruent with the goals of the criminal justice system and in harmony with the objectives of victim's advocates. The numerous inconsistencies offer significant support for the need to develop program standards or guidelines to ensure that the safety and freedom of women is the primary goal of intervention with male perpetrators. (50)
In the present state of counselling, just about anyone can claim to offer "therapy" for wife batterers. Back in 1984, Browning already pointed to the multiplication of highly speculative and mutually contradictory streams of "explanations" offered for wife abuse by a growing variety of service providers.
The groups and services vying for government contracts to "deal with" perpetrators of wife abuse - or with "violent couples" - can be masculinist self-help organizations, local social service delivery centres, recycled family or conjugal counsellors, so-called "mediators", private sector clinicians, transition houses, family service associations, clergymen, etc. Each of these speak of "therapy" but generally bypass the professional requirements and accountability criteria touted by most participants during Hanson & Hart's conference. Indeed, would the psychologizing approach apply so generally and appear so effective in cutting back government expenses if one were to insist on such requirements and make this type of "therapy" a truly professional practice?
Daniel J. Saunders: ...we need to make sure that the leaders are competent. Competency involves both background knowledge and therapy skills. Background knowledge must include a high level of awareness of the causes of domestic violence and the impact that it has on the victim. Knowledge of the many ways that offenders minimize and rationalize their behaviour is crucial... The importance of training is exemplified in the research of others... Therapist training, however, can be a major investment. The National Institute of Mental Health depression study spent many thousand dollars per therapist in the training. (6)
Further, in a discussion of some counter-productive factors of the psychologizing approach, we will outline to what extent the 1991 Ottawa conference participants proved aware of how competency lacks and biases of current program leaders had the effect of impeding any improvement of perpetrators' behaviour, to the point of actually worsening wife abuse during and after "treatment" as the Urban Institute's extensive Baltimore survey pointed out in late 1991.
But first, let's review the fundamental problems discussed by conference participants. One of the requirements of any competency assessment has to be a consensus on the dynamics of wife abuse. The existence of any valid therapy requires a consensus on a pathology and a typology of affected subjects. Yet, after twenty years of research and ceaseless efforts to arrive at these, these pursuits are at a dead-end in the very words of these specialists. The hypothesis that wife battering is some kind of disease amenable to therapy simply doesn't hold water on the face of the available data.
The organizations and individuals vying for research grants and service delivery contracts generally claim that their approach is inclusive of every dimension of wife abuse and will benefit the whole community. Yet, Hanson & Hart acknowledge at the onset that...
...none of the existing programs are successful for all batterers, and there is only limited information on who is likely to benefit from treatment. (p. i)
Barbara Pressman, Adjunct Faculty at Wilfrid Laurier
University's Faculty of Social Work, validates this
assessment, while focusing on one of the risks created by
the inconsistency in the field:
...Currently, there is not universal agreement on the
causes [of wife assault]. Consequently, some treatment
agencies are employing theoretical models that blame
victims, trivialize the problem, and fail to appreciate
the impact of violence on victims. (17)
One can even wonder whether the problem of wife abuse
can be assimilated to a mental health issue for men.
Pressman demonstrates that it isn't :
...the rate of abuse of women in the home is so pervasive
(encompassing all economic, cultural and religious groups)
and so extensive (1 in 8) that one cannot explain
behaviour of such epidemic proportions as intrapsychic
phenomenon or relationships and interactional patterns
gone awry. (...) It is the described societal context that
must inform our treatment of abusing men. (17, 19)
Jurgen Dankwort demonstrates this, offering a
statistical assessment:
...as some researchers have noted, it is puzzling that
most empirical investigation of wife battering has
examined the intra-personal level of wife abuse even
though it is doubtful that more than 2 or 3% of all
wife-beating can be attributed to purely inter-personal
characteristics. (38)
Oto Cadsky, of Edmonton's Forensic Assessment and
Community Services, is adamant:
Treating individual men in individual treatment programs
is not going to make the slightest difference in the large
scale. (117)
The failure of the psychologizing approach is especially obvious in the failure of attempts to identify any pathology specific to wife batterers, which would justify designing the intervention along a mental health approach and letting perpetrators avoid sanctions.
In a summation of program efficiency assessment studies
written in 1990 and appended to the conference
proceedings, the well-known specialists Richard M. Tolman
and Larry W. Bennett write:
The evidence presented thus far heavily suggests
heterogeneity of characteristics of men who batter. Given
the heterogeneity, the question remains whether there are
identifiable clusters of attributes of relevance to
practice... Typologies may be harmful if they cloud a
societal role or overemphasize spurious differences... The
heterogeneity of behavioral and psychological
characteristics suggests that no one pathology can be
linked to battering. (135, 137)
Psychologist Robert J. Brown confirms this assessment:
Overall, the research field has not had much success in
using psychological testing to provide either definitive
data or answers about the psychological profile of the
wife batterer. Even the performance of psychology's
assessment flagship, the Minnesota Multiphasic Personality
Inventory (MMPI;
Hathaway & McKinley,
1983
), has been very disappointing... (75)
...our large sample of batterers, on average, came right down the middle of the general population on personality traits and clinical factors. Even a much closer look at the suspiciousness and paranoia scales showed nothing that would set this sample apart from the general population. (76)
Given these data does it even make sense to go on speaking of treating "the" male batterer?
This is the conclusion reached by James Browning:
However, when we examine research that has attempted to
demonstrate differences between abusive and non-abusive
men, we find that many of the client characteristics we
thought were associated with violence do not differentiate
when demographic variables are controlled. (89-90)
This assessment is validated by conference organizers
Hanson & Hart:
Based on research such as Brown's, we now know that
batterers are not characterized by high levels of
psychiatric disorders - as traditionally defined. (p.
iii)
This emerges clearly from Tolman & Bennett's
summation of the existing literature:
Therefore, the overrepresentation of psychopathology in
clinical battering populations does not clearly implicate
psychopathology as a causal factor in battering. (127)
The high depression level often noted in wife batterers
held accountable for their assaults has often been
presented as a explanatory factor of the violence. Tolman
& Bennett make the following point:
...the Hamberger and Hastings (1988) finding that
batterers in treatment were more depressed than a
nonviolent comparison group, but
nonidentified batterers in the community were
not
, supports the contention that it is the
consequences
of battering that cause depression. (131)
More often than depression, it is "poor anger control" that is most often targeted as a cause of wife battering by program providers. Whatever the specific theoretical line emphasized by program leaders in what a conference participant referred to as a "grab-bag of treatment approaches" (123), it remains a fact that "anger management" of perpetrators is almost universally the practice implemented by Canadian treatment providers.
Yet, this theory suffers diminishing credibility. As
Tolman &Bennett write, among other considerations :
Although some programs operate on the assumption that
anger is central to battering and that anger control is
sufficient to stop battering, others have suggested that
anger is merely one rationalization for use of aggressive
control (
Gondolf & Russell,
1986
)... Studies utilizing the Novaco Anger Scale (
Novaco, 1975
) suggest no anger problems for men who batter (
Hastings & Hamberger,
1988
,
Telch & Lindquist, 1984
). ...the degree to which men may self-generate such anger
to serve as a rationalization for coercive behavior is
unresolved by current research. (129-30)
The theory referring to some stress experienced by the
spouse as an explanation for his assaults on his wife is
also invalidated by the available data. Tolman &
Bennett write :
On most of the measures [
Straus, 1980
;
Hotaling & Sugarman,
1986
;
Barling & Rosenbaum,
1986
], however, work stress was unrelated to battering.
MacEwen and Barling (
1988
) studied working couples at 6 and 18 months. The
researchers found that work and life stress was not
predictive of abuse... These two studies, and those that
preceded them, do not make a strong case for the role of
external stress in battering. ...although some men who
batter may use the notion of a stress/battering link to
excuse their aggressive behavior, stress is probably not
directly linked to woman battering. (133-4)
Much pop psychology pundits would have us believe that
batterers are merely redirecting outwards abuse they have
suffered as children. Many self-help group leaders and
traditional clinicians have latched on to this sweeping
generalization to focus empathetically on their charges'
unhappy childhood. But this deresponsabilizing hypothesis
remains empirically unfounded and actually invalidated by
available data, according to Tolman & Bennett :
Kantor and Straus (
1989
), using data from the 1985 survey, found that observation
of parental battering in husband's family of origin
predicted minor marital violence but did not predict
severe marital violence. (134)
The link between childhood experience and adult
behaviour appears to be one of modelling rather than of
one of psychopathology requiring treatment, for, as Tolman
and Bennett point out :
Observing violence in the family of origin has more
consistently predicted violence by men toward their
partners than has abuse as a child (
Hotaling and Sugarman,
1986
) (134)
The "anger management" approach, feeding on our society's reluctance to acknowledge family men's choice to repeatedly batter wives and children, has led to a vision of assaultive husbands as devoid of certain skills. This is not the experience of the women, and even of some program leaders, who deal with these men's lies and evasive behaviour.
In a book published in 1993, Next Time She'll Be Dead:
Battering and How to Stop It, sociologist and
criminologist Ann Jones writes :
It's vital to understand that battering is not a series of
isolated blow-ups. It is a process of deliberate
intimidation intended to coerce the victim to do the will
of the victimizer. The batterer is not just losing his
temper, not just suffering from stress, not just
manifesting "insecurity" or a spontaneous reaction
"provoked" by something the victim did or (as
psychologists put it) "a deficit of interpersonal skills"
or an "inhibition in anger control mechanisms." These are
excuses for violence, popular even among therapists who
work with batterers; yet we all know aggrieved, insecure,
stressed-out people with meagre interpersonal skills who
lose their temper without becoming violent. We assume,
then, that the grievances of the violent man must be
worse, and that under extreme stress he has spun out of
control. He looks it, and that's what he says: "I wasn't
myself." "I was drunk." "I went bananas." "I lost it." "I
went out of my mind." It's lines like these that provide a
public excuse and deceive a battered woman into giving one
more chance to the so-called real, nonviolent men
underneath. But in fact that violence is himself,
perfectly in control and exercising control.(Next Time,
She'll Be Dead, Ann Jones, Beacon, Boston, 1989, pp.
88-89)
Nevertheless, behaviourist program providers go on
presenting their services to the State as an opportunity
to "solve" wife abuse by providing wife batterers with
skills to offset the lacks they are said to suffer. A
conference participant quotes Don Dutton on this issue :
Don Dutton has told me repeatedly... that a lot of these
men do not lack skills, they lack something else. They
know how to say the right answers and they know what
people are looking for, but they do not do it. It is as if
they know what is right but they just do not do what is
right. (121)
In the second part of this paper, we will look more closely at the repercussions of anger management ideology on the partners of batterers, and specifically at the effects of the new skills taught to these men to better exert control in the conjugal sphere.
Objective observers of the field acknowledge that, for lack of theoretical coherence and appropriate methods, the psychologizing approach fails not only to prevent but even to predict recidivism in "treated" subjects.
This was apparent at the outset of a July 1991
synthetic review of Canadian treatment programs: the Abt
Institute's Treatment Programs for Men who Batter: A
Review of the Evidence of their Success. James Browning
summarizes one of the report's conclusions, echoed by many
conference participants in their own research reports :
Burns, Meredith and Paquette (
1991
) point out that we know very little about which client
characteristics predict recidivism. (90)
Although the psychologizing approach attempts to
identify attitude changes that would somehow translate
into an end to batterers' violence, or at least into a
reduction of these or their transposition at a verbal,
"managed" level, the following exchange between Don Dutton
and a conference participant demonstrates that researchers
remain in the dark about just what mechanisms would allow
such a transition:
[Question:] ...more generally, when you take measures
before, during and after treatment and you get some
movement, what do we know about how that movement relates
to [battering or to] future battering treatment? ...is
there any systematic, empirical evidence that movement of
a particular type, as a function of participation in
therapy, is associated with reductions in battering?
[Dutton:] As far as I know, it has not been done. (108-9)
The following comments were exchanged after James Browning very realistic account of methodology problems in the design of treatment programs and in the accounting of so-called success rates.
[Comment:] One of the other concerns is that some of the men in these programs are or become involved in multiple relationships over the course of treatment. No one has really addressed this in terms of program development. Often, there are cases where men are abandoning their partners or their partners are leaving and they are immediately getting involved with another woman in an abusive kind of relationship...
[Comment:] The other side of that problem is when a man gets involved in a new relationship and the honeymoon period of the new relationship just enables him to blame his ex-partner all the more for what happened. You must try to work with him in that context.
[Comment:] We see a lot of that in our program: men who are just coming out of a relationship and are blaming what happened on their ex-partner completely. (95-6)
Behind closed doors, specialists acknowledge the extent to which this complete theoretical anomy keeps them from guaranteeing society even a modicum of program efficiency.
[Comment addressed to Oto Cadsky:] The bottom line is the issues related to the selection of individuals for treatment. The big question is what kinds of treatment are best for which kinds of individuals? If we go for a global approach that mixes up a whole range of offenders, when we do not have a good idea as to what program content or process is about, then we may not be that far ahead... We should also look at the issue of program integrity as an integral part of any evaluation. We need an evaluation which is going to provide us with some valuable information for bettering treatment programs so that they become more effective.
[Response (Cadsky)]: I think that is "pie in the sky". You are not going to get it... Those of us who have been working in the area can no longer predict, even on clinical grounds, who does well and who does badly. I do not think any of your psychological measures are going to be able to find them. I think you are kidding yourself if you think you will find anything that will predict who does well and who does badly... I doubt that is achievable. (122-3)
Jurgen Dankwort and James Browning tried at the
conference to bring to light the major deontology issues
created by a discourse that attempts to create a
pseudo-uniformity around the concept of "treatment" for
batterers :
Dankwort: Most researchers and policy makers, however,
continue to behave as if programs for wife abusers are
uniform. They seem to prefer avoiding discussions that
risk shifting research focus towards such normative issues
as practice philosophy. The meaning of that concept, and
how we are going to operationalize "philosophy" in program
evaluation, has been all but ignored. (37-8)
And yet this silence, combined with an opportunistic
eclecticism end up undermining any possibility of
achieving a truly efficient protocol, as another
conference participant pointed out:
[We cannot]...just open up every program and mix a variety
of offenders with a grab-bag of treatment approaches. If
we do that, we are not going to be able to find out that
sort of thing. (123)
There are very real risks in such an uncontrolled
approach both for "treated" men and for their partners, as
Browning, among others, reminded conference participants:
Self-help groups may degenerate into hate sessions,
individual therapy does not provide opportunities for
breaking social isolation, and couples therapy risks a
reinforcement of the blaming dynamic. (91)
We shall return shortly to these very real risks intrinsic to the various forms of the psychologizing approach.
Tolman and Bennett: No evidence suggests that alcohol treatment in itself will be effective in changing abusive behavior, but alcohol and drug problems no doubt interfere with the process of such change... Alcohol-abusing batterers are more likely to have antisocial or sociopathic profiles, predicting a poorer treatment outcome compared to nonalcoholic batterers. (128)
Why is this selection process being implemented? In order to identify the "poor subjects" to be avoided.
Daniel J. Saunders: We can also try to uncover the abuser types that consistently fail to benefit from treatment. Preliminary, cross-sectional studies give some evidence of those most likely to re-offend: the alcoholic batterer, the narcissist, those with longer histories of abuse, and those who witnessed parent abuse... (3)
This is quite a group! Nevertheless, under pressure to provide society with proofs of their programs' efficiency, program leaders make every effort to limit their intake to those assaultive husbands most likely to "succeed", based on current hunches. How can one justify this weeding out of "poor prospects" in order to upgrade program's outcomes with subjects carefully selected to best match the therapist's pet hypothesis, at the same time as therapy is presented as a universal method?
Approximately half of candidates are thus summarily rejected despite the confidence and financial resources invested by decision makers in the psychologizing approach, at the expense of judicial sanctions and of resources for victims, solutions whose efficiency is not in question (when they are available).
Don Dutton: The biggest drop off point in the criminal justice system is police arrest rates, which are still probably too low. Based on the studies I reviewed, in cases where there is prima facie evidence for assault, arrests were only happening 21% of the time. There is tremendous variability based on the attitudes of police officers. Police officers need remedial attitudinal adjustment... In a study I did with Kirk Williams, Les Kennedy and Steve Hart, we found that one of the side effects of arrest is that both the man and the woman tell more people about the assault and, therefore, it is not quite as isolated an event... The woman tends to come out of the closet more and she may begin to stop blaming herself for the violence. (107)
So, just how available is an intervention whose efficiency is thus validated? Dutton reminded participants that "the studies that have taken police reports and wife reports (Sherman and Berk was one of them) show that, for every eight assaults that wives reported, one got picked up by the police." (108)
It is with regard to this context that one should question the ideology and practice of the continuing non-judiciarization of patriarchal crimes, which must be identified as an attempt to re-privatize this crucial problem. One should also question the unrealistic efficiency rates touted by treatment programs in order to justify their public funding.
Daniel J. Saunders: We know that in working with men who batter we are only reaching a very small percentage. (2)
The main illusion surrounding alleged therapies for wife batterers is probably the importance they are given. One must know how very few batterers are ever influenced by them. Jurgen Dankwort spoke at the conference of...
"...the almost insignificant number of men who actually go through their 15 or 20 sessions. With at least one million Canadian women being beaten each year, the estimated 150 programs across Canada do not hold out much hope for change... In fact it begs the question of why one would want to either continue investing money in this project or continue studying what appears from the outset to be a useless endeavour. (40)
How can one explain this disproportion between, on one hand, the incoherence and marginality of the psychologizing approach and, on the other, the budgets and limelight it goes on appropriating?
Without daring to say that it may be because this approach comforts a traditional priority to men's interests and a just as traditional judicial immunity for intra-familial abuse, one can at the very least look to what Jurgen Dankwort calls the "optimistic and idealistic" discourse used and abused to promote batterers' treatment on every podium over the last fifteen years.
And yet, the closer one looks at the success figures presented to the media and to funding agencies, the less these seem representative of real-life "treatment", inasmuch as the studies offered in support of these claims suffer major methodological problems that render them most of them quite invalid.
Hanson and Hart's conference participants proved quite forthright about these major drawbacks invalidating the most wildly optimistic assessments of their works. (One hopes that they show just as much scepticism in their requests for further Satte funding of treatment programs...)
Hanson and Hart: ...all treatment programs have large drop-out rates from the point of initial contact, to the end of treatment, to follow-up. (p. iii)
Tolman and Bennett: Men who batter prematurely drop out of intervention at very high rates. Based on a national [U.S.] survey of batterers' programs, Pirog-Good and Stets (1986) placed the attrition rate at 40%. (137)
Efficiency assessments of programs are based on the men who complete them. But are these men truly representative of the assaultive spouses that are referred to such programs?
The following data, provided to participants by Cadsky (p. 118 bar graph), from his own Edmonton program, seem typical enough: Out of a hundred men referred to the program either by court order or by their wife, 80 survive the waiting period, 60 are accepted, 30 start the program, 20 are left after 12 weeks and a mere 5 complete the required 30 weeks.
This is to say that, even with the elimination of as
many "bad prognosis" subjects as possible, programs fail
at least 94% of men referred to them, those that do not
attend or do not complete treatment. This factor can only
bias the representativeness of the remaining sample, as
Browning told participants:
Attrition from treatment and follow-up can skew
chronological comparisons in a positive direction (as
presumably poorer prospects tend to drop out) and little
is known about these dropouts. (93)
Jurgen Dankwort offers an obvious summation of this
problem:
We have mounted long-term studies that attempt to convince
us that the effects of treatment endure, but we rarely
discuss the fact that the declining sample size very
likely means that we are interviewing those couples who
would report the least amount of repeat violence because
the unsuccessful cases have disappeared from our
longitudinal study. (39)
In other words, and to quote Saunders as he questions
the alleged success of this or that "approach":
The treatment may have a higher success rate only because
it finishes with the most self-motivated men or those who
are least violent. If differential attrition occurs, the
randomization of the experiment is jeopardized. (5)
Don Dutton argues that what we call treatment ends up
not even reaching the most dangerous of perpetrators,
although these still exploit the dejudiciarization ethos
to bypass any sanction for their assaults:
There are different pockets of batterers who get into
different streams. The worst ones are the men who drive
their wives and girlfriends into transition houses. They
seem to be the ones that are the most violent. They have
substance abuse problems. They also seem to have
personality disorders. A lot of the time, they dodge the
law.
The ones who seem to need it the most are the ones
who get away with it. If, however, you go and sit in a
treatment group, the men you see in there are not like
that. (108)
Measuring any modicum of success would appear to
require quality follow-up studies. Dankwort points out
that program leaders responding to his survey had no
empirical basis for their highly unrealistic efficiency
assessments:
Informants appeared
idealistic and optimistic
about the effectiveness of their programs. While about 80%
of counsellors reported their programs were "effective"
(as opposed to "ineffective" or "don't know"), the same
respondents also reported that 50% of their clients fail
to complete treatment.
Only one program has conducted an empirical study
that shows (with arguable validity) reduced violence for
post-treatment subjects...
Respondents did not qualify or explain their hopeful or
idealistic inclinations. (49)
It should be noted that the results of this
cross-Canada study proved similar to those of a 1989
Quebec survey conducted by Gilles Rondeau, Monique Gauvin
and Dankwort among community batterer treatment programs
active in Quebec at the time:
Of the 15 programs, only 2 collected data on the results
achieved. However, these were incomplete data, collected
using available resources... We will therefore treat them
as estimates. In 9 of the remaining 13 programs, the
counsellor and/or program leader shared with us his
guesstimate of the results achieved, based on his
knowledge and experience of participating clients. In
summary, we can offer no data based on hard numbers. (Les
programmes quebecois d'aide aux conjoints violents,
Planification-Evaluation Sante Services sociaux, 1989, p.
102)
Tolman and Bennett point out the same failing in the
studies most often quoted to establish treatment
efficiency:
The percentage of program participants actually contacted
for follow-up studies represents another critical
evaluation issue. Some studies reach only a small
percentage of participants; others fail to report at all
how many participants were actually reached.
Nonparticipants in follow up are probably more likely to
be abusive (
DeMaris & Jackson,
1987
), and therefore success rates reported in the literature
are probably inflated. ...the number of participants
contacted usually decreases with longer follow up.
(139)
For a program provider, the quickest and most expedient way to elicit information concerning a client or ex-client's recidivism is of course to ask him directly. As could be expected, such self-assessments have produced quite bombastic results. How can self-reports be treated as a credible source when it is well recognized that minimization is a major problem with perpetrators?
Don Dutton: When a man comes into the program, he is going to report one half as much violence as his partner does. This is a bench-mark of the average minimizing an assaultive husband does. (112)
Tolman and Bennett confirm this methodological problem:
Several studies depend only on male self-report, which is
problematic in that evidence suggests that men deny and
minimize their own level of abuse (
Edleson & Brygger,
1986
). (139)
Under fire from shelter-based women's advocates, researchers have come to recognize partner reports as much more trustworthy indicators of recidivism.
Tolman and Bennett: Percentages of successful outcome ranged from 53% to 85%. Lower percentages of success tended to occur in programs with lengthier follow up and when success was based on women's reports rather than on arrests or self-reports. (140)
Jurgen Dankwort: ...the most reliable source for reports of recidivism is now held to be the man's partner. (39)
At present, most program leaders look to partners for a confirmation of subject self-reports. But, as Richard Tolman pointed out in his presentation, even such partner reports can provide misleading data, especially when the partner feels responsible for her partner's (and his therapists's) success...
Partners' corroboration is critical, and any study that depends only on men's self-reports or on police data would grossly underestimate recidivism ( Edleson & Brygger, 1986 ). However, there are also times when the woman, too, may not report abuse that has occurred; for example, if she is fearful, if she does not know who is on the other end of the telephone, or if she feels dissonant about staying in an abusive situation. She may even deny the level of abuse she is experiencing as a way of coping with an abusive situation. ...the most effective solution is to combine self-reports, partner reports and police data. ...An index that combines reports of abuse from any source is the most conservative estimate of re-offence in terms of physical abuse. ( Hamberger & Hastings, 1988 ). (59)
This is precisely the triple-source approach chosen by The Urban Institute's Adele J. Harrell for her ground-breaking December 1991 study of treatment efficiency for 81 subjects in 3 different programs over a two-year period in the Baltimore area. Her data point to a counter-productivity of treatment in terms of recidivism and many attitudinal factors. This study will be reviewed a bit further.
As Tolman pointed out above, one of the main reasons of what Don Dutton calls "false positives" lies with too short follow-up periods, if any follow-up is done.
Tolman and Bennett: Some portion of the success is probably attributable to periodicity of violence: the follow-up periods may not have been long enough to reveal the recidivism ( Dutton, 1988 ) (146)
The same authors also point out:
Most confidence can be placed in studies utilizing
lengthier follow-up periods: men who batter may give up
their abuse for a short time following intervention but
later reoffend. (139)
The problem of too short follow-ups is compounded by what James Browning identifies as a "honeymoon period" experienced by the subject as he renews a relationship. This specifically affects self-reports of improvement and tends to create much larger problems further on, problems not amenable to present treatment approaches.
[Two community-based demonstration projects run by Correctional Services of Canada in Edmonton and Ottawa] are finding that, very often, there is this "honeymoon period" when men are re-integrating back into the community. In some cases, there is an unwillingness to recognize that there are problems in relationships. This means that treatment is not kicking in until some time after release. Then problems get really wild and it is not something that can be dealt with within a 12 or 25-week program. Relapse prevention is also emerging as an issue to be dealt with for all of the reasons that the Jennings' ( 1990 ) article argues. (95)
Yet, not all subjects take up again with their former victim. When this separation has not been taken into account, there has been an overvaluation of treatment success.
James Browning: Some studies have failed to factor in degree of contact with the victim and have not excluded separated men who have a reduced chance of re-offending even without treatment. (93)
Strange as it may seem, for lack of the right follow up questions and of even considering partner issues, program leaders have systematically presented as "cured" those men who had merely lost the opportunity to contact their former victim and assault her once again, or who had not yet formed new abusive relationships.
Daniel Saunders: There are also some important control variables to consider when interviewing the women. As a simple example, we need to find out how much of the time after treatment they are together. If they are separated, are they still having contact? We can adjust our rates of violence by the length of time for which there has been no contact... Often the [victim] reports are not relevant because the couple is divorced and having no contact. ...Considering the number of couples who do not have contact with each other after treatment or who cannot be located, we need better post-treatment measures... more valid than their self-report measures. (5, 7-8)
That is to say that similar successes would have been recorded in the absence of any treatment, the couple's separation being the real solution! To complicate matters, it appears that intervention is proving much more difficult with men who no longer live with a partner.
Oto Cadsky: You now have to control for variables like the following: Is the partner there at the time, while he is attending the group? Are they with somebody? Did they break up? This is important because it is very hard to get at the problems if they do not have a partner: if they are not with a partner, they do not have any problems. (120)
It is also hard not to wonder whether the partners that stay do not find themselves cast in the role of guinea pigs testing their abuser's success... At any rate, there have been instances of "therapists" pleading with partners to come back or take him back, vouching that he had "changed", the very same ones who, in the face of recidivism, see fit to publicly blame women for having "unreasonable expectations" towards treatment...
The pattern outlined above by Cadsky could be a clue to
the fact that, rather than "treatment", the operative
factor in whatever positive change occurs may well be the
shift in power that occurs when the perpetrator's partner
is socially supported by the very resources challenged by
the treatment approach. Tolman subscribes to this opinion:
One important reason to look at women's actions is that
we may be misattributing gains made by men to
men's treatment programs, when these changes may have more
to do with the kinds of services women have received. She
may be involved in a shelter, she may have a legal
advocate, or she may have received many kinds of services
that trigger changes in her partner. If we look only at
his actions, we may believe the changes occurred because
of treatment. (63)
Whatever the cause, this factor points to the major problem of the lack of valid control groups in what purports to be a scientific methodology.
For a whole lot of ethical and structural reasons, it appears almost impossible to identify control groups to which to compare "treated" abusive spouses. This also impugns the validity of these studies' optimistic conclusions for who can tell whether it is "treatment" or some other factor or combination of factors involved (e.g., the partner's decisions, the support she is given, judicial intervention) that is to be commended for any positive change recorded?
Daniel J. Saunders hypothesizes what would be the ideal
control set-up... to immediately conclude that this
approach may also qualify as "treatment", and an
inexpensive one at that...
...A form of minimal treatment control group is probably
the best solution. This control group would be akin to
very strict monitoring by a probation officer, i.e. weekly
visits that would not involve treatment. The monitoring
would help detect the escalation of violence in case a
person has to be withdrawn from the control condition and
placed in a "separate" crisis" condition. The monitoring
might also be found to be an inexpensive form of treatment
with a good rate of success for some types of abusers."
(3)
The strangest consequence of current methodology
defects has to be the fact that researchers note almost as
much improvement in
untreated
control groups (Harrell even found
more
improvement) than they do in treatment completers. Given
that completers are generally associated with the best
prognosis, this brings Tolman and Bennett to write:
The relatively successful outcomes for men who do not
complete intervention suggest that other factors
contribute to cessation of abuse following intervention.
(140)
Richard Tolman emphasized this point in his
presentation:
I am merely suggesting that criminal justice system
involvement can be a path towards cessation of direct
aggression in and of itself. Among the variables that did
not significantly predict recidivism was the number of
sessions attended. That leads me to think that criminal
justice system involvement in and of itself was a very
active ingredient of change for some men. (64-65)
This certainly validates the crucial role of judicializing sexist assaults. Still, the non-judiciarization ethos seemed central to the belief systems of most of program leaders interviewed by Dankwort in the research he presented at the Ottawa conference.
Since the majority of survey data are based on subject self-reports and on their answers to questionnaires that try to assess their attitudes, these results are wide open to a manipulation of these instruments by abusive men, according to what they know they should report.
Jurgen Dankwort: Finally, seemingly improved post-treatment psychological portraits of offenders have rarely been considered from the perspective of how social desirability has skewed the results from the outset. Yet men "caught" having committed reproachable behaviour will certainly attempt to mange their responses to solicit compassion when they participate in psychological tests or present their personal histories. (39)
Don Dutton is of the same opinion:
[As for] social desirability and how it relates to many of
the typical measures we use... the men know why they are
there; they know they are being tested because they have
assaulted their wives. They get into a stance about what
they want to present to the researcher or to the
therapist. ...whenever someone is a subject in an
experiment, they can tell what it is the experimenter
wants them to do or they can tell how the person who gives
them the questionnaire wants them to answer. The right
answers are transparent. (102-3)
It is not enough to try and solve this my merely
factoring in extra caution in the assessment process. The
issue is to what extent an attitudinal assessment process
- controlled by the subject himself - can even predict
recidivism. Don Dutton again on the limits of attitudinal
assessment, whether the subject is consciously or not
manipulative:
Frequently, the men know that they are not supposed to be
sexist and they know what the right answers are, but that
will not tell you very much about how they might behave.
...With the so-called pencil and paper measures you cannot
be sure they are actually going to tell you how the man
would behave when he is enraged, when he is angry, when he
is highly aroused, or when he is drunk. Men cannot always
tell you how they are going to behave under these
circumstances. We can ask them to project into a
situation, but they do not always know. ...There is also
the process called self-amplification, where they get
angry and begin to feed back off their own arousal and
their own behavour and their anger spirals upwards. ...It
is impossible to tap into that on self-reports, but it is
still something you want to know about. (103-4)
These two latter problems outline an idealistic bias in
the theory offered for wife abuse. Proposed "treatment"
approaches suffer from this bias and many abusive spouses
find themselves unable to relate to program design. Tolman
and Bennett point out that, according to some studies:
[
Pirog-Good & Stets, 1986
;
Grusznski & Carrillo, 1988
;
Saunders & Parker, 1989
;
Hamberger & Hastings,
1986
] ...these studies suggest that treatment may be less
likely to reach younger, less educated, lower income, and
minority men. (138)
This results in excessive and differential drop-out
rates along class lines, which render even more suspect
the glowing success rates advertised by treatment
programs. Therapists who look at attendance figures for
their program to survive find themselves forced to
jettison significant content material, despite its
relevance to their "theory":
Richard Tolman: I think that one of the reasons that we
see high attrition rates and poorer results for lower
educated and unemployed men is that we built these
programs for a middle class, highly educated audience (see
Saunders & Park, 1989
). What I am finding more and more in my own work is that
we keep stripping away layers and layers from the
intervention. ...we can't put in as overtly as we'd like
all the important ingredients, including material on male
sexual socialization and patriarchy and the kinds of
things we have included in the past.... The men who are
illiterate or undereducated, however, are still the most
likely to drop out. For these men, the idea of doing
homework is really difficult because of their lack of
skills. (65)
This suggests that the "treatment" approach is really only designed for a privileged category of males, those whom society refuses to sanction, as if their battering was less harmful.
Barbara Pressman: [These crimes] are no less serious when enacted by middle class professionals who are remorseful and would suffer financially, etc., if imprisoned. (19)
In order to maintain attendance by their
underprivileged charges - those that are unable to afford
the safe-conduct of having an individual clinician vouch
for them in court (as did O.J. Simpson) - and to justify
public funding through the social service delivery
network, program providers strip down their program to the
least disturbing of approaches, as we have seen above. But
even then the message doesn't seem to get across. Few
providers are as candid as Richard Tolman about how
inefficient content communication proves to be and how
irrelevant content itself may prove to be:
...we really only have six content sessions, about 45-50
minutes each, which are repeated about every six weeks.
The men are required to come to 18 weeks of full sessions,
in addition to a one-day workshop that starts the program
off. With this structure, they receive the same material
four times. Even on the fourth time, however, many men in
the group are still not mastering the content. ...Many
men, even men who have successful outcomes, are not
retaining the material in the fashion that we would
imagine they would need to if the key element in change
was really the skill itself or the content itself that we
were teaching. (67-8)
Daniel Saunders: ...For the severe abuser... we may need one to two years of treatment, rather than the five or six months that is typically given. (3)
James Browning: In my view, it is essential to cover [men's conditioning regarding sex roles and power] in the program as it relates directly to the men's reported anger. It makes sense conceptually and to ignore it gives the message that the therapist does not consider it important, thereby reinforcing the status quo. However, the chances of changing such strongly conditioned values in a short time with a relatively concrete group of men are poor. Even Gondolf's program, which emphasizes sex role re-conditioning, showed no change on this dimension ( Gondolf, 1988 ). (91-2)
Advertised success figures usually refer to assault recidivism. But even when the incidence of physical violence appears to be reduced - and this is enough for most providers - psychological violence doesn't necessarily go down.
Barbara Pressman reminded conference participants of
the central role of this more generalized type of
violence:
...it is very important to note that abusing behaviour is
not only physical abuse but also any behaviour which
demeans, degrades, intimidates, frightens, or limits
control and choices over another person's being. It is any
attempt to control another person's life or choices for
her own life. Although this would not constitute a legal
definition of abuse, it is critical for the work of ending
violence against women. Not all abusing physically abuse.
However, all abusing men engage in many behaviours to
control partners. Consequently, ending physical abuse will
not ensure that the other forms will be addressed.
Therefore, it is essential that any program treating
abusing men address all the forms of abusing behaviour,
including isolating partners, restricting mobility,
threatening to hurt a partner or children, controlling
finances so that a partner is financially dependent,
destroying objects or property, abusing pets, using
put--downs that erode self-esteem, using sexual coercion,
treating a partner as a servant, or using intimidating
gestures, looks or anger to frighten a partner.(20)
Jurgen Dankwort clearly identifies a central lack in
the current process of program assessment:
...we have measured program success by concentrating on
repeat physical violence. Yet, we have much anecdotal
evidence from battered women that psychological abuse, or
even the lingering threat of recurring violence in the
home, can be just as devastating as an actual slap, kick
or push. (39)
This "anecdotal" evidence is confirmed by Richard
Tolman more experimental protocol:
My own work on psychological maltreatment has given some
very tentative support to the notion that even when
physical abuse stops, levels of ongoing psychological
maltreatment will contribute to negative outcomes for
partners (
Tolman & Bhosley, 1991
). (62)
We will return shortly to this notion of a rise of psychological violence in the treatment context.
Not only is the slightest improvement treated as a
success...
[Question] "...they may go from a 100 % to an 80 % chance
of recidivating?" Saunders: Exactly. So treatment may
still have a good impact on them. (9)
... but some providers even count as non violent
spouses having benefitted from "treatment" victims who
have been brought into a new high-growth product, "couples
therapy":
Tolman and Bennett: ...some methodological concerns limit
confidence in these reports. Because victims and
perpetrators are seen in couples groups, it is critical to
know how success is determined, i.e. whether victims are
included in percentages of clients who are nonabusive
after treatment. But several studies neglect to make this
distinction. (145)
These couples groups are perhaps an indication of the continuing popularity of the violent or provoking partner in therapeutic circles, myths that can only comfort the perpetrator's denial of his responsibility.
Several conference participants acknowledged the impossibility of eliciting an adequate response from probation officers and courts in case of recidivism or program drop-out. The obvious consequence is that programs are much less efficient at getting perpetrators to face consequences of their violence, the stated objective.
Richard Tolman: ...the issues are much more complex: What is the nature of the [court] mandate? How is it enforced? What actions do the program or the criminal justice system take if the order is violated? ...The same program may not work in a context where there is no coordination between the program and the criminal justice system. ...we have to look at these meso-system issues: What is the relationship between the men's program and the probation department? When there is a re-offence, what happens? Dopes the probation officer take action? Does the program inform the court system? If action is taken, of what kind? What about linkages between the program and the man's partner? Is the woman reached out to? What happens if she's reached out to? What happens if she reports to the program that there's an offence? (64, 66)
Tolman actually answered some of these questions, with
an obvious embarrassment shared by the questioner:
[Question:] What happens to your court-mandated
participants who do not complete the program?
[Tolman:] Not very much. That is a real problem.
[Question:] They do not have their probation breached or
anything like that?
[Tolman:] I find that, having come from Minnesota - the
fertile data of violence services in the United States -
Illinois is a little primitive in this regard... Right now
the situation is very haphazard; whether any action is
taken at all depends on which probation officer the men
are assigned to. Even with those programs in Chicago that
are run under the auspices of the court directly, if they
go back to the judges and say, "This man has violated. He
has re-offended. He has not come to session," the judges
will send the man back to treatment, as opposed to finding
some other penalty. The fact that there is not very good
integration in this area makes our intervention less
effective.(66-67)
In Canada, where "treatment" is mostly a health service issue, Don Dutton confirms this assessment: "...sometimes limits are imposed on [treatment groups] by the criminal justice system interface... If you do not have that leverage from the criminal justice system, it is hard to generate completion." (104)
Jurgen Dankwort has observed the same lack of probation follow-up in Quebec, but he lays the blame more squarely on the theoretical inconsistencies and the political choices of program providers themselves.
This raises the all-important issue of the counter-productive effects of so-called "therapies" for wife batterers. Justified and funded in a clear context of dejudiciarization, these therapies must be understood to be alternatives to justice and to more efficient victim support measures. Is this what we want?
The notion that an approach that alleges to be therapeutic can be not only inefficient but actually counter-productive may seem excessive at first, especially in a "better than nothing" assessment climate.
And yet the best empirical data available so far - in terms of sample dimensions, length of follow-up, number of sources and variety of treatment approach -, published shortly after Hanson & Hart's conference, reveal a higher rate of recidivism among treated men than in the control group of batterers arrested but not mandated into therapy. Although the arrest produced a marked reduction of physical abuse from both groups, this study found that only 57 % of treated perpetrators had ceased all violence, while 88 % of untreated men had shown the same improvement after a two-year period ( Harrell, 1991 )
Beyond such controlled data, a growing number of battered women have come forward and shared with shelter workers their experience of life with a batterer "in treatment". A few echoes of these testimonies transpired at the Hanson & Hart conference.
This problem is particularly acute given the control
skills taught to wife batterers in so-called "anger
management therapy":
[Comment from a participant:] ...anger control, as Gondolf
and Harris have pointed out, can also be used to improve
the man's ability to control and manipulate his wife.
...[EMERGE] do not use anger control at all... They
explained that they do not use [time-outs] because men can
say "time out" whenever they want to walk out on an
argument with their partners. Psychologically, this lets
the victim know, "You better not push me any further. I'm
doing a time-out, so if you don't want to get smacked
again you're going to respect where I'm at." In other
words, it involves her in his change process and, in doing
that, re-victimizes her.
...The problem is not his anger - anger is a universal feeling - the problem is his manipulation and his use of anger to control. (69)
And this is not only a theoretical possibility:
Jurgen Dankwort: All the anecdotal evidence - from
Battered but not beaten, Linda MacLeod's (
1987
) study of shelters in Canada, to the anecdotal evidence I
am getting from women who work in the shelters - indicates
that, while women using shelters may not necessarily be
physically abused to the same extent that they used to be,
they are now experiencing much more psychological abuse,
which is just as devastating. (52)
Robert J. Brown: We were noticing that when the batterer stopped physically assaulting his wife, his verbal assaults went way up for probably close to six months before he came back down to a normal level (whatever normal is). (86)
A revealing lapse: these "therapists" interpret as "normal" batterers' psychological violence. And of course, women pay the price.
False-positive results, i.e. overly optimistic efficiency assessments, have the effect of sidetracking society from the quest for real and lasting solutions to women's lack of safety.
Tolman and Bennett note the following:
...some studies consider
reduction
of violent behaviour a success while others set complete
cessation of violence as the criterion for success.
Viewing reduction as success is questionable; reduction of
violence may not end the terror that battered women feel
as a result of abuse (
Hart, Safety for Women: Monitoring
Batterers' Programs, 1988
), and even so-called minor violence like slapping,
pushing, and shoving may result in physical injury (
Rosenbaum, 1988
)... [We] believe that the most inclusive definitions of
violence (i.e., focus beyond physical abuse) provide the
highest level of program accountability as well as the
greatest validity in determining whether men have changed
their abusive behaviour. (139)
Jurgen Dankwort, a long-time practitioner and
researcher of Canadian programs, warned conference
participants:
We have to look at the
extent to which we are facilitating a
re-formulation of men's violence.
Women are saying, "Are you part of the problem, or part of
the solution? If you are part of the solution, then you're
going to address this issue. You are going to do something
so that physical abuse doesn't turn into psychological
abuse, which then is beyond the reach of the law. (52)
Similarly, Trudy Don, speaking for the Ontario Association of Interval and Transition Houses, expressed concern that teaching men to control their anger may actually assist the abuser to refine or re-adjust his abuse so that he may more effectively control his spouse psychologically, with equally harmful effects. (40)
This concern was validated by Richard Tolman in reply
to a participant's comment:
Men often misuse the pro-feminist aspects of the program
as well. They adopt a very sensitive stance,
but then they take the language of equality and
turn that back on their partners
, saying "Now you are no treating me equally; you are
psychologically maltreating me." Our sensitivity to
psychological maltreatment can backfire. They can say
"Well, my partner sometimes doesn't let me leave the
house. Sometimes she accuses me of having affairs.
Sometimes she berates me in front of my friends. She's
just as abusive as I am." That is the pro-feminist
sensitivity - the idea of psychological maltreatment, in
part, comes out of a pro-feminist framework - but yet it
can be misused by men. (70)
Dankwort quotes more of the women who work day to day
with victims of wife abuse and try to obtain
accountability
from perpetrators and the police/justice system:
Le Regroupement provincial des maisons d'hebergement et de
transition pour femmes victimes de violence conjugale au
Quebec complained that
these programs are being regarded as a "panacea"
and, as such, they are an unrealistic solution to the
violence. They stressed that these programs must not be
allowed to offer offenders refuge from culpability and
from criminal sanction.
(40)
Yet, every one in the field knows full well that this is exactly what is happening. "Therapy" is used daily as an alternative to judicial sanctions throughout the land.
Moreover, as Dutton points out, it may even come to
impede the possible intervention of the penal system:
Treatment groups came in to make judges feel better and to
give them some kind of viable sentencing option for wife
assault. (...) we need to look at the impact of treatment
groups on actual criminal justice system functioning. For
example, if a treatment group is installed, does it have a
ripple effect? Does it make judges more likely to convict?
Does it make crown prosecutors more likely to take the
charge seriously? To proceed with the charge? Does it make
the police more likely to arrest? (100-101)
The answer to all these questions may well be: NO. This is especially true in Quebec and other provinces where "therapies" are openly linked to a policy of dejudiciarization and where police officers are already being instructed to refer abusers to a support group, rather than arrest him.
In the conclusion of a literature review which
"attempted to provide an overview of existing outcome
evaluations on programs for abusive men", Burns, Meredith
and Paquette wrote in July 1991, a few months before the
conference:
... the existing data on the effectiveness of these
programs do not appear to justify their use to the
exclusion of other types of intervention.
Perhaps the most worrisome example of this is the
use of treatment as a diversion from prosecution. Given
the poor completion rates for the treatments themselves,
and the limited success of these treatments among those
men who do complete them, it seems ill-advised to place so
much reliance on these interventions.
(Treatment Programs for Men Who Batter: A Review of the
Evidence of their Success, p.56)
In Adele Harrell's definitive Baltimore study, it was noted that, even after having completed their program, perpetrators demonstrated a very high level of confidence that they would suffer no real sanction for repeat offenses which they themselves felt to be relatively probable.
Some therapy providers go as far as to petition the
State to fund their programs by pointing out just how much
money will thus be saved in reduced benefits to victims
and prosecutions... The therapy ideology adds to
traditional male priority the seduction of what appears to
be a "scientific" solution, one that promises to make the
problem "disappear":
Robert J. Brown, Calgary General Hospital: The Calgary
Region Domestic Violence Committee is also proposing a
study to look at the savings achieved by providing
effective treatment. The Committee will compare a group of
batterers and their victims with a group of non-violent
cohabiters. They will examine the billings to Alberta
Health Care, the costs of police car and ambulance
responses, court and probation costs, days lost at work,
and so on. We expect to find that the savings in these
areas will more than compensate for the costs of
treatment. This may be the information provincial
governments need in order to reallocate some of their
resources. (84)
Who can guarantee that the funding of women's vital front-line resources such as shelters and transition houses isn't threatened by this ominous "and so on"? In Quebec and elsewhere, one of their main conditions voiced by women's programs for any cooperation with batterers' programs was that these never compte for the budgets obtained through hears of tireless activism and consciousness-raising by women. This condition has never been respected and today, we find "therapy" providers trying to achieve an ever-growing portion of the State's global budget envelope, using premises as extraordinary as (battering) men's "right" to an equal share of State support...
Even in the absence of such pathogenic effects, the therapeutic model is itself a travesty of the very nature of a form of violence wilfully chosen by batterers, for purely self-interested reasons. To deny this dynamic is to deny and silence women's experience.
Dankwort: (...) programs for men who batter may, in some circumstances, actually be counterproductive to the interests of women . There is real concern that, in stressing the clinical etiology of wife abuse and ignoring the utility of using force or threats to maintain privileges and entitlements in a patriarchal society, we are producing and reinforcing attitudes and messages pejorative to women. (40)
Barbara Pressman, the only woman invited to address the Conference, pointed out the victim-blaming nature of many of the theoretical hypotheses being bandied about. She is also wary of the theories that merely dilute the perpetrator's responsibility for his choice to batter, such as references to family "antecedents", especially when the perpetrator remains free to further assault his victim.
When I do healing work, it is always related back to the violence and it is in the service of ending abuse. As clinicians working with trauma survivors of child sexual abuse, we have a key rule regarding unravelling the history: we will not do it unless the person has stabilized or is working from strength. (...) Stabilization includes the fact that one is not abusing someone else or abusing substances... (32)
Similarly, the literature review offered by Tolman and
Bennett points out the counter-productive nature of the
attention given to perpetrators' feelings of depression:
(...) empathically supporting batterers for their
depression is counterproductive if it deflects
interventive focus away from their abusive behavior (
Adams, 1988
). (131)
The actual day-to-day practices used in either of the
many approaches described as "therapy" generally remain
vague or secret. Referring to self-help groups, someone
put the following comment to Daniel Saunders:
It is also extremely interesting to learn about the
attempts you are making to find out exactly what is going
on in these groups, i.e. the specific content of the
therapy. Most studies do not provide this information.
One sometimes wonders if anything was ever
implemented at all or whether or not the program was one
night out a week for the men to talk to friends.
I really applaud your efforts. I think this is a lesson
that we will have to keep hold of. (12)
This silence and the "empathy" that has Tolman and
Bennett concerned may be more than a mere grey area. In a
pan-Canadian study replicating one that he conducted in
Quebec with Rondeau and Gauvin in 1989, Dankwort recently
probed the attitudes and methods of therapy program
providers, specifically exploring certain contradictions
between their discourse, their current "theory", and their
practices. Allow us to quote him at length:
Whereas all programs indicated a preference for
sex-segregated format in their counselling work,
the reasons for this choice often reinforced
specific images of men, including those of men living in
isolation, being emotionally handicapped, and, in general,
being victims of their circumstances.
These are images that require groups to offer men support
and a secure place to "heal". (47)
(...) the findings did show inconsistencies between the ideas professed by counsellors and the programs' published philosophies. Lastly, and perhaps more importantly, there are apparent inconsistencies between the professed assumptions and the stated practices within programs. (...) the study's findings suggest that the strategy of attributing responsibility is often derailed by many factors that are proposed as contributing to or causing the violence. (...)Generally, respondents explain the etiology of wife abuse in terms of intrapersonal, interpersonal and social-structural factors, while overlooking the utility of using force. (...) This has the effect of minimizing the assertion that men who use violence are freely choosing this course of behaviour. (...) In many instances, this effectively removed agency from men's violent behaviour. Men's violence against women was portrayed, not as an instrumental act, but as an expressive act arising out of men's presumed deficiencies and their learned behaviours. Respondents portrayed men's violence as a situation allegedly aggravated by the harsh or stressful environment in which many men live and work. Consequently, intervention practice appeared to miss the target, addressing, for the most part, the probable sources of violence and neglecting the more likely source of the problem. (47)
Whether these problems are due to an "anything-goes"
level of theoretical improvisation or, more likely, to a
visceral hostility to the feminist analysis of male power,
they translate into the demonstrated ineffectiveness of
programs and counsellors, at a time where these should be
rejecting society's "explanations" for batterers' violence
and keeping them from male bonding and from using therapy
as a
reinforcement
for their misogynist attitudes and controlling strategies.
Dankwort notes an actual
solidarity
of many counsellors with perpetrators:
In a number of examples given during interviews however,
respondents appeared to overlook strategic
opportunities for confronting men on their manipulative
tactics.
Only two of the three programs in the pro-feminist cluster
expressed concern that anger-management techniques may be
counterproductive or contra-indicated. There were,
however, many occasions where counsellors acknowledged
that
wife abusers are manipulative and will collude
against women if given a chance to do so.
Most programs reported that their clients are much more
likely to collusively bond with one another than to be
aggressive and intimidating in mutual confrontation over
their past or present behaviours and attitudes towards
women. In some instances,
characterizations by counsellors about the origin
of the violence and about what has to be done to end it
appeared collusive with offenders. This presents a
possible risk of affording therapeutic legitimacy to men's
disavowal of the moral component in such
behaviour.
(48)
It seems logical to associate this collusion to the overt opposition of "therapists" to any substantive treatment of wife battering as a crime. At best, the judicial process is treated as a necessary-evil entry procedure to bring batterers - and, increasingly, their partners - to "therapists" offices. But any actual sanctions (e.g., losses of privilege) are not only frowned upon but actively opposed, for instance by a policy of secrecy concerning further assaults revealed in counselling. Far from working hand in hand with judiciary, program providers show, according to Dankwort's interviews, a clear and profound bias against any form or notion of retribution for the crimes committed. Their axiom seems to be that batterers are to be supported , not sanctioned, regardless of this method's effectiveness in ensuring women's safety.
(...) counsellors related difficulties in conceptualizing how criminal sanctions for assault might be congruous with counselling treatment. This led them to define legal and social intervention as distinct and, explicitly or implicitly, incompatible. (...) To most respondents, it seemed either unconscionable and/or counterproductive to punish the perpetrator even though they candidly admitted he had committed an act of assault . In many cases, respondents attempted to reconcile the antithetical nature of the social control demanded by victims' advocates, on the one hand, and the compassion counsellors were eager to provide on the other. (...) Other counsellors embarked on an equally problematic course of creating dichotomies that separate a "good" person from his "bad" behaviours. (48)
Given these circumstances, is it reasonable to expect from such mentors that they foster a taking of responsibility by wife batterers and to go on funding their male collusive experiments on this basis?
Most counsellors also appeared to ignore or disagree with the idea that men can use their power to exploit women to gain selfish personal advantages. They seem to perceive the dynamics of "conjugal violence" as mutually destructive. This view discounts the unequal power balance between the sexes and/or the destructive potential of using force in intimate relationships. (49)
Like many women's advocates before him, Dankwort
expresses concern about the silence and minimization that
surround the survival of such antiquated notions. He is
equally worried about the counter-productive methods used
by many therapists, especially in so-called "family
therapy":
I think we have a tendency to play down our differences
and there is a great risk in minimizing controversy. To
illustrate how serious the differences are, we can go back
to (...) issue number three,
"How is power seen?"
There are many therapists who do not recognize that power
can be exploited in a system by, for example, one member
of a family against another member of a family. This
does not even enter into their
definition.
Can we say that, for example, in Quebec, there are no more
family therapists with the Batesonian perspective? On the
contrary, the tendency is still as strong in Quebec as it
is in the rest of Canada and in North America. For
example, the Family Violence Bulletin recently mentioned a
study of psychotherapists and family therapists in the
United States (
Harway & Hansen, 1990
). It showed that
most family therapists either did not address the
violence, minimized the effect of the violence, or
intervened in ways that were inconsistent with current
minimal standards of practice.
These are very real concerns. When our particular
philosophical orientation does not allow us to recognize
that power can be exploited, it is a serious issue.
(55)
The corollary of the aforementioned collusive support
for perpetrators is a surprising lack of empathy for their
partners, one that compromises both their safety and their
entitlements. Jurgen Dankwort offers the following
conclusion to his research:
Limited comprehension of the principles of justice,
including the established idea of punishment or
retribution, also suggested that
counsellors still lack empathy for the situation
of victims
and their demands for justice. In their responses,
counsellors generally
failed to recognize how retribution might be
justified or how it might be integrated into the process
of attaining full offender accountability
. They appeared to pay little attention to the fact that
the criminal justice system has, in the past, often
treated wife abusers sympathetically or even exonerated
them altogether. Nor did they give any attention to the
effect this approach has had on the problem of woman abuse
generally. (...)
only a few respondents listed the safety and the
well-being of the man's partner as a program
objective.
(48-9)
These results are disconcerting because of the implications for victims who often hold high expectations about how their partners will change once enrolled. As a number of respected authors have pointed out, there is no conclusive evidence to support such expectations, and to encourage them may be counterproductive to the goal of improving victims' safety and women's well-being. (49)
Tolman and Bennett note:
As Gondolf pointed out ["Who are those guys? Toward a
behavioral typology of batterers", Violence and Victims,
3(3), 1988, 187-203], referring men whose violence may be
relatively intractable to counseling may
increase the likelihood of their partners'
remaining with them, raising their hopes with little real
possibility of change occurring. (...) Criminal justice
system involvement might be more effective for less-severe
batterers who don't have outside criminal
involvement.
(137)
A recent Canadian study by Meredith and Burns validates these concerns about the role of therapy in women's decision to remain in high-risk situations, a situation long-decried by shelter workers and by authors such as Ann Jones and Susan Schechter (When Love Goes Wrong, Harper-Collins: New York, 1992).
Richard Tolman: Is [the partner] more or less willing to call the police as the result of her partner's involvement in a program? That's something that was raised by Meredith and Burns ( 1990 ) in their study. That study found that some women reported a decrease in their willingness to use various kinds of sanctions toward their partners' violent behaviour, sanctions that they otherwise might have used if their partners were not in programs. (...) We must explore what might happen to her as a result of her partner's involvement in a program that decreases or increases her empowerment in the situation. (63)
A year later, Burns, Meredith and Paquette amplified
that concern:
It is discouraging that, as a result of treatment
programs, women are staying in these relationships which
seem only marginally improved in many cases.
(Treatment Programs for Men Who Batter, p.56)
It is easy, of course, for program providers to describe such outcomes as misguided in women, excessive expectations, self-induced illusions, etc. When confronted, they reject the notion that what they are offering is a universal remedy to wife battering. And yet their tireless promotion of alleged "therapies" and of speculative "explanations" - that reassure inasmuch as they deny batterers' self-interested agency - gives precisely that impression and seem to target specifically women, the new growth segment of programs' clientele.
As for Tolman and Bennett's suggestion of judiciarizing the cases of men who only assault their wife - and would be most receptive to sanctions -, recent justice statistics indicate that these perpetrators are the ones that are almost never sanctioned, but merely invited to attend "therapy".
Ironically, our "optimistic and idealist" discourse on
the prognosis for changing batterers' behaviour closely
resembles and unmistakably reinforces that which
perpetrators themselves use on their victims. As Richard
Tolman remarks:
The man's treatment may have a negative impact on his
partner. For example, a man's initial involvement in a
program may encourage her to return to him rather than
leave him. She may be less willing to call the police
because she wants "to cut him a break" while he is in
treatment. She may be overly hopeful, as Dr. Saunders
talked about.
She may have "learned hopefulness" in the
situation because of his promises to change
repeatedly.
(63)
Batterers and therapists both promise change. And who would pass up the opportunity to blame a woman for daring to doubt such fascinating men and male experiments?
For after all, her assaulter rarely matches the
dramatic "batterer profiles" which therapists speculate
publicly about to justify the psychologizing approach.
This disinformation process is another of the high-risk
repercussions of programs' total lack of accountability,
and of the theoretical confusion created by the
psychologizing approach and by the treatment of any
observed data as causal factors:
Tolman and Bennett point out
(...) a few potential pitfalls of the attempt to delineate
batterer "profiles" (...) One such pitfall, especially
important because intervention and prevention efforts may
be affected, is to mistake causes for effects or
vice-versa (
Stark & Flitcraft, 1988
). For example, one commonly noted trait of abusive men is
depression; but
does depression help cause battering, as some
assume, or does battering in fact contribute to depression
in the batterer?
Related pitfalls are to give a causal element more weight
than it deserves (
Gelles, 1980
;
Stark & Flitcraft, 1988
) or to mistake a benign element for a causal one (
Margolin, John & Gleberman,
1988
;
Stark & Flitcraft, 1988
). A final reservation:
looking for characteristics that differentiate men
who batter from other men may obscure more important
similarities or societal conditions that make battering
commonplace and tolerated.
(125)
One can also note that, for lack of suitable control groups, it is completely impossible to ascertain whether the "depression" observed in some batterers isn't linked to the program-perpetrator power differential, to the pressure of a recent or impeding separation or court procedure, to wanting to pass as the "real victim", or simply to having been "caught" and ever so slightly challenged.
Whether they do so in order to facilitate a partner's corroboration that has become essential to the credibility of their "success" figures, or because they believe in any of the theories that make wife abuse a systemic problem where both partners' are seen as "co-responsible" for male assaults, or because government money is much easier to obtain using what Brown calls "marital preservation" program goals, or because their background is in marriage counselling, or simply in order to buttress flagging attendance figures, many so-called therapists continue to - or are beginning to - make victims part of their aggressor's "treatment". Indeed, this formula has become the major growth strategy of self-help groups such as Quebec's PRO-GAM, by far the largest operator in the field, who rush to open up "partners' programs", oblivious to the well-documented hazards of such "couples intervention".
Tolman et Bennett: An issue generating great controversy is the practice of treating men who batter conjointly with their partners. The interventions utilized are similar to men-only groups (...) However, a number of additional criticisms have been levelled at conjoint work, including safety concerns and implicit or explicit victim-blaming. (...) the research on couples to date is limited to studies without comparison groups. Lindquist, Telch and Taylor ( 1984 ) reported that (...) at six-week follow up, half of the couples reported violent incidents, and at six months, all couples contacted reported further violence. (...) some methodological concerns limit confidence in [success] reports. Because victims and perpetrators are seen in couples groups, it is critical to know how success is determined (...) (145)
Robert J. Brown confirms this assessment:
(...) history has shown us that marriage counselling is
not typically an effective treatment for wife battering.
In fact, it can be a
very dangerous approach
if the batterer is still battering. (74)
But a social system that uses therapy to factually diminish men's responsibility and to keep women in marriages and the home has everything to gain by associating men and women as closely as possible in such schemes.
[Comment from a participant] There is a lot we can do. For example, an interesting reference for the assessment of education or counselling programs for batterers has been published by the Pennsylvania Coalition Against Domestic Violence [ Safety for Women: Monitoring batterers' programs, Barbara Hart, 1988 ] Ms. Hart lists 34 items of criteria, questions that should be used as a yardstick to measure programs (...) evaluating what and for whom? That is the bottom line. (123)
In her detailed agenda of the changes necessary to really end wife battering, Barbara Pressman offers two ground rules that can help us re-centre social intervention regarding wife battering, in a way not antithetic to really putting women's safety first.
Accountability to women and the women's movement through close affiliation and cooperation with shelters and other programs for battered woman is a critical component of responsible programming. (30)
An improved system of probation, including increasing and improving training programs for probation officers regarding wife abuse, maintaining contact with battered women, and returning men to court when failure to attend group or unsatisfactory participation is a breach of probationary orders. (20)
The efficiency of penal versus psychologizing
intervention is probably the best-grounded argument
holding back the ideology of non-judiciarization for
anyone truly concerned with women's safety and rights.
Here is how Tolman and Bennett conclude their literature
review:
Sherman and Berk (
1984
) studied police response to domestic violence calls and
found that, based on both police and victim reports,
arrest was a more effective deterrent of further violence
than was separation or mediation. Although the Sherman and
Berk (
1984
) study included domestic calls that were not between
cohabiting men and women, a reanalysis by Tauchen,
Tauchen, and Witte (
1986
) examined only those cases that involved men and women
who were in a relationship. Their results confirmed the
Sherman and Berk (
1984
) finding that arrest provided the most effective
deterrent. Sherman and Berk's (
1984
) results were also replicated by Berk and Newton (
1985
), who conducted a nonexperimental replication of the
Sherman and Berk (
1984
) study. Jaffe, Wolfe, Telford and Austin (
1986
) evaluated the effectiveness of a probable cause arrest
that police implemented in Ontario, Canada. The arrest
policy dramatically increased charges - from 3% of all
wife abuse occurrences to 67% of all wife abuse
o