Countering Confusion about the Duluth Model
By Michael Paymar and Graham Barnes, Research Specialists with the Battered Women’s Justice Project, a program of Domestic Abuse Intervention Programs
Recently, the “Duluth Model” of working with men who batter has received serious criticism, despite being the pre-eminent model internationally. Much of the criticism is based on flawed research that is contradicted by other better-designed, more comprehensive studies. Here the authors respond directly to misinformation generated by 1) a 2003 study from the National Institute of Justice of the U.S. Department of Justice; 2) an analysis of batterer treatment models by Babcock, Green, and Robie (2002); and 3) unsubstantiated criticism of the philosophy underlying the Duluth Model by Dutton and Corvo (2006). The authors, both of whom have worked with the Domestic Abuse Intervention Project in Duluth, Minnesota, rebut inaccurate assumptions and myths about what the model is and how it works and challenge the ethics, as well as the efficacy, of avoiding criminal justice sanctions and relying solely on resource-intensive mental health treatments for batterers. Changing historically entrenched battering behavior is difficult, the authors explain, but the Duluth Model prioritizes victim safety and autonomy.
Critical Review of National Institute of Justice Research
In 2003, the National Institute of Justice (NIJ) released Batter Intervention Programs: Where Do We Go from Here? a report suggesting that there was “no benefit from batterer counseling” and that the Duluth Model “did not work.” Despite these claims, the NIJ researchers admit to limitations in their study, and other researchers have roundly criticized it as seriously flawed. Nevertheless, critics of the Duluth Model have used the study to question the effectiveness of batterer intervention programs, most specifically the curriculum Creating a Process of Change for Men Who Batter which was developed by the Duluth Domestic Abuse Intervention Project (DAIP).
The NIJ research focused on recidivism rates and behavioral changes of male offenders at sites in Broward County, Florida, and Brooklyn, New York, which were ostensibly using the Duluth Model. In each of these sites, half of the men were ordered to participate in counseling programs using the Duluth curriculum Creating a Process of Change for Men Who Batter, and the other half comprised a control group that received supervised probation without counseling.
Following are some of the reasons the NIJ study has been widely criticized:
Although it uses the term “Duluth Model,” the NIJ report examines only the “batterers’ treatment” component. The researchers do not indicate to what extent, or even whether, the two treatment sites were part of a coordinated community response (CCR), a core element of the Duluth Model. Another key component of the Duluth Model is swift, consistent consequences (jail or return to the program) for noncompliance with conditions of probation, civil court orders, or program violations, e.g., missing groups and further acts of violence. This did not occur consistently at either NIJ site. In fact, at no stage did the NIJ study establish that the programs being studied were operated in accordance with Duluth Model practice. “Testing how faithful programs are to the models on which they are based requires process evaluations, which, to date, few evaluations have incorporated” (Jackson et al.
2003). Yet the researchers claim that the Duluth Model batterer programs “had little or no effect” (Jackson et al. 2003). Such conclusions are not only scientifically impossible to draw but ethically questionable.
In the NIJ’s Brooklyn site, a major change in methodology occurred during the study. Defense attorneys objected to some offenders being mandated to attend counseling while others were placed on probation. Rather than discontinue the study, researchers offered an accelerated eight-week class to the program group rather than adhering to the Duluth curriculum, which requires 26 weeks of classes. A change in methodology of such magnitude seriously weakens the validity of the research.
The researchers’ ability to assess batterers’ changes in attitude was compromised. They admit this, stating, “In both studies, response rates were low, many people dropped out of the program, and victims could not be found for subsequent interviews. The test used to measure batterers’ attitudes toward domestic violence and their likelihood to engage in future abuse was of questionable validity. In the Brooklyn site study, random assignment was overridden to a significant extent, which made it difficult to attribute effects exclusively to the program” (Jackson et al. 2003).
The NIJ research didn’t analyze different treatment programs or models. The researchers simply concluded that offenders’ rates of recidivism and attitudes about women were basically the same for offenders who went into counseling programs using the Duluth curriculum and for those that were placed on probation.
In light of their conclusions, NIJ researchers recommended that a cognitive-behavioral approach be used in batterers’ intervention programs. They cite the Emerge and Amend programs as cognitive-behavioral alternatives equivalent to the Duluth curriculum. As a matter of fact, the basis of the Duluth curriculum is cognitive- behavioral. It is very similar to the curriculum offered by Emerge and Amend. One difference is that Duluth focuses less on psychological assessment and more on how power relationships and entitlement are reflected in individuals, families, communities, and different cultures. However, these distinctions are not as significant as the similarities in the programs’ approaches and philosophy.
Other research on batterer intervention programs has reported outcomes very different from those of the NIJ study. For example, researchers in Scotland found that offenders ordered to counseling using the Duluth curriculum who were threatened with immediate consequences for failure to participate—the Duluth Model—had a success rate of 73 percent as opposed to only 33 percent for those offenders who were simply placed on probation (Dobash and Dobash 1999). This study directly contradicts reports that claim no program effect.
Significantly, the Scotland researchers were able to determine that the program using the Duluth Model did in fact include the essential elements of the model before they compared its results to those of the control group whose members received no educational intervention.
Most notably, the NIJ-sponsored Broward County experiment clashes with a much more comprehensive study by Edward Gondolf that included a seven-year, multi-site evaluation (Gondolf 2003). In 2002, Gondolf concluded that “well-established batterer intervention programs with sufficient reinforcement from the courts do contribute to a substantial decline in re-assault.” In 2004, he reported that “at the 30-month follow-up, less than 20% of the men had re-assaulted their partner in the previous year; at the 48-month follow-up, approximately 10% had re- assaulted in the previous year. Moreover, over two-thirds of the women said their quality of life had improved and 85% felt very safe at both these follow-up points.” Gondolf’s study was much more carefully designed and carried out than the NIJ study, with larger samples, four sites, and better-informed attention to program design, implementation, and the context of re-offense. It provides a benchmark against which other research can be considered.
We agree with the NIJ researchers that changing offenders’ attitudes towards women is extremely difficult. We still live in a sexist society where women are devalued, where many men believe they are entitled to be in control in an intimate relationship, and where men who batter believe they have the right to use violence. While it is a goal to change the attitudes of men who batter, the ultimate goal of the Duluth Model has always been to ensure that victims are safer by having the state intervene to stop the violence and address the power imbalance inherent in relationships where one person has been systematically dominated and subjugated by another.
Duluth’s experience is that many battered women want their partners to go to counseling. They want the violence to stop. They want their abusive husbands to change. While we can argue the relative success of all treatment
programs, we believe that as a society we should offer men who batter an opportunity to change. If they don’t change, we believe that the state has an obligation to increase sanctions to stop the violence.
Critical Review of Research by Babcock, Green, and Robie
In 2002, Julia Babcock, Charles Green, and Chet Robie published “Does Batterers’ Treatment Work? A Meta- analytic Review of Domestic Violence Treatment” in the Clinical Psychology Review. As with the NIJ study, some practitioners in the field have been using this research to criticize the Duluth education model, cognitive- behavioral therapy, and other treatment programs. In their article, the researchers examined the findings of 22 studies evaluating batterer treatment efficacy. They essentially concluded that the Duluth approach, cognitive- behavioral therapy, and other models yield mixed results, and therefore the criminal justice system should consider alternative treatments.
We agree with these researchers that the criminal justice system should consider alternative responses when offenders aren’t amenable to treatment. In such cases, we believe that community service, probation, or jail might be more appropriate. We also believe that offenders with diagnosed mental health problems might benefit from different kinds of treatment, resources permitting. A major dilemma for the courts and human service providers is whether to conduct a comprehensive assessment of offenders to determine which treatment approach would be most effective. This process can be very expensive, and its potential benefits need to be balanced against other options, for example, providing more resources to victims and non-offending family members. Similarly, the safety of victims must be considered if the court begins to order offenders to treatment programs using restorative justice and couples' counseling while a batterer is still exercising power and control over his partner. Finally, we have never supported state statutes that mandate a particular treatment model, even though some states have adopted standards that dictate treatment approaches (Dutton and Corvo 2006). We do advocate for state standards that require batterer intervention programs to keep victim safety central to their programming.
We also acknowledge that not every person who has used physical force against a partner is what we would describe as a batterer. A person who batters is one who uses a pattern of intimidation, coercion, and violence against a partner. It is unusual for men to be arrested for assault in cases where there has been no such history. Women call the police because they are afraid. Neighbors call because the violence is alarming. Children call because they are trying to help their mothers. While some therapists may encounter “walk-in” clients whose use of violence does not constitute a pattern of abuse, these are rarely the cases coming through the criminal justice system, and when they do, they rarely result in a conviction. This article and the Duluth Model address a public response to battering rather than isolated incidents of domestic violence.
Babcock, Green, and Robie do state that we shouldn’t abandon traditional treatment programs if there is a strong coordinated community response, the centerpiece of the Duluth Model: “Based on partner reports, treated batterers have a 40 percent chance of being successfully nonviolent, and without treatment, men have a 35 percent chance of maintaining nonviolence. Thus, the five percent increase in success rate is attributable to treatment. A five percent decrease in violence may appear insignificant: however, batterer treatment in all reported cases of domestic violence in the U.S. would equate to approximately 42,000 women per year no longer being battered.” However, none of the major, methodologically sound research studies that examine recidivism supports such a paltry claim of change.
“Policymakers should not accept the null hypothesis and dismiss the potential for batterers’ intervention to have an impact on intimate partner abuse. Results showing a small effect of treatment on violence abstinence do not imply that we should abandon our current battering intervention programs. Similar small treatment effects are found in meta-analyses of abstinence abuse treatments when abstinence from alcohol is the outcome of interest. Yet, some people are able to dramatically transform their lives following substance abuse or battering interventions. Given what we know about the overall small effect size of batterers’ treatment, the energies of treatment providers, advocates, and researchers alike may best be directed at ways to improve batterers’ treatment. Because no one model or modality has demonstrated superiority over the others, it is premature for states to issue mandates limiting the range of treatment options for batterers” (Babcock, Green, and Robie 2002).
Nowhere in their research are there claims that other treatment methodologies work better than the Duluth approach. In fact, they state that programs using the Duluth curriculum performed slightly better than the alternatives studied in their research.
At the DAIP in Duluth, 100 offenders who had been ordered to complete a 26-week education program were subsequently tracked over a five-year period. The researchers found that most of these participants used less violence less frequently. Forty percent of the court-mandated offenders recidivated at least once during the five years (Shepard 1992). Extrapolating from that number to account for those that didn’t get caught, we would conclude that about 60 percent of court-mandated offenders will use some violence again, but a much lower number will continue to batter their partners on an ongoing basis. This finding is consistent with Babcock, Green, and Robie’s research.
Gondolf’s multi-site evaluation of programs that are similar to the Duluth Model, both in curriculum and criminal justice system accountability, showed “a clear de-escalation of re-assault and other abuse, the vast majority of men do reach sustained nonviolence, and about 20% continuously re-assault. The prevailing cognitive–behavioral approach appears appropriate for most of the men… Program effectiveness depends substantially on the intervention system of which the program is a part” (Gondolf 2004).
Critical Review of Dutton and Corvo
In their 2006 article, “Transforming a Flawed Policy: A Call to Revive Psychology and Science in Domestic Violence Research and Practice,” Donald Dutton and Kenneth Corvo grossly misrepresent the Duluth curriculum when they write “according to the Duluth Model, all [men] must be treated as patriarchal terrorists regardless of differences in how the violence developed.” They further state that “essentially, the Duluth Model views every man convicted as equivalent to the worst man convicted without gradations or nuance,” and “the primary goal of this model is to get male clients to acknowledge ‘male privilege’ and how they have used ‘power and control’ to dominate their wives.” Interestingly, Dutton and Corvo place quotation marks around power and control and male privilege as if they don’t exist, but that is for a later discussion.
The underpinnings of the Duluth curriculum do come from a historical analysis. When Europeans came to this continent, they brought religion, laws, and economic systems that institutionalized the status of women as the property of men through marriage. From the church to the state, there was not only acceptance of male supremacy, but also an expectation that husbands would maintain the family order by controlling their wives. Various indiscretions committed by wives were offenses to be punished by husbands. This system of male dominance (like any social structure where one group oppresses another) was perpetuated by: a) a belief in the primacy of men over women; b) institutional rules requiring the submission of women to men; c) the objectification of women which made violence acceptable; and d) the right of men to use violence to punish with impunity (Dobash and Dobash 1983).
The status quo of male domination remained fully intact until the early twentieth century when state legislatures began to make wife beating unlawful. However, the practice of men using violence to control women didn’t diminish. In the late 1960s, the Women’s Movement began challenging the state to intervene in domestic violence cases and women and some men began to confront the concept of male supremacy in the home. In the 1970s, the Battered Women’s Movement emerged as the voice of victims and advocates to challenge psychological theories about the causes of violence and explanations of why victims often stayed in abusive relationships.
Do all men who batter want to dominate women? This is a complicated question. Clearly, many men who batter believe that women should be submissive to men and there are others who share a variation of these sexist beliefs—“The man is the head of the household” or “You can’t have two captains of one ship.” However, there are other men who batter that don’t believe that their wives or girlfriends should be subservient because of their gender, but they still batter. These men use violence to control their partners because they can and violence works. Violence ends arguments. Violence is punishment—it sends a powerful message of disapproval.
Some mental health practitioners are now repackaging old psychological theories in opposition to analyses that indicate that culture and socialization shape the way men who batter think and act in intimate relationships.
Although there is much of value in mental health theories that can assist the healing of victims and perpetrators
alike, we do not see men’s violence against women as stemming from individual pathology, but rather from a socially reinforced sense of entitlement. We believe that the beliefs and attitudes possessed by men who batter can be changed through an educational process.
Dutton and Corvo claim “the Duluth Model maintains an ineffective system where resources are diverted from other potential program responses, e.g., joint treatment of violence and chemical dependency or mutuality of partner violence.” They go on to state “mandatory arrest policies are a product of the ideologically driven view that since domestic violence is always strategic, always intentional, always unidirectional, and always with the objective of female domination by men, that it must be contravened by the power of the state. Once one removes this ideological presumption, the rationale for mandatory arrest disappears.”
They cite research on mandatory arrest policies in Milwaukee and other cities that apparently fails to demonstrate that mandatory arrest reduces recidivism (Shepard 1992). What they don’t tell us is that the cities chosen for this research had very poor prosecution rates and lacked a tight coordinated community response, which is the cornerstone of the Duluth Model. It is axiomatic that arrest without prosecution, meaningful sentencing, jail, the threat of jail and counseling will usually be less effective in reducing violence because offenders will get the message that the criminal justice system and society as a whole don’t take domestic violence seriously. And, although arrests alone may not reduce recidivism to the level we would hope for, critics of pro-arrest policies would never argue that we should stop arresting rapists or thieves who assault their victims if arrests don’t measurably reduce recidivism for those violent offenders.
From a public policy perspective, not arresting batterers essentially decriminalizes domestic violence and condemns a victim to either live with the violence or (as in the “bad old days”) be forced to press charges against an abusive spouse. Doing away with pro-arrest policies targeting the predominant aggressor (a core component of the Duluth Model) reduces the total number of arrests but increases the proportion of dual arrests. Dual arrests have proven ineffective in stopping violence, and they also have the unfortunate consequence of making victims more reluctant to call the police when further acts of violence occur.
Dutton and Corvo also contend that the “best designed studies” (Ehrensaft, Moffitt, and Caspi 2004; Moffit et al. 2001) indicate that intimate partner violence is committed by both genders with equal consequences. They point to these studies to buttress their argument that marriage counseling is an appropriate treatment to end domestic violence. We respond that, even if surveys comparing rates of perpetration by gender are accurate, proponents of the argument that women are as violent as men fail to account for the impact of the violence (Langhinrichsen- Rohling et al. 1995; Cascardi and Vivian 1995), the severity of the violence (Hamberger and Guse 2002), the level of fear experienced by the person being assaulted (Barnett and Thelen 1995; Hamberger and Guse 2002), or motivation for the violence.
In an emergency room study, gender differences among a cohort of injured patients found that men initiated violence in far greater numbers than women (Phelan et al. 2005). One hundred percent of female respondents versus 39 percent of the male respondents reported being injured in a domestic violence incident. Thirty-six percent of women reported being intimidated by their partner’s size while none of the men reported being intimidated by their partner’s size. Seventy percent of women reported that they were very strongly afraid during partner-initiated violence while only one man reported experiencing this degree of fear. The majority of men (85 percent) reported not being afraid at all when their female partners initiate violence. From experience, police officers and domestic violence advocates have long understood the contextual differences between the violence of men and women, yet proponents of the “women are as violent as men” agenda downplay the significance of gender differences in the ways that men and women use violence (Dutton and Corvo 2006; Mills 2003).
Opponents of a feminist analysis of domestic violence continue to argue their theory that women are as violent as men and that the level of mutual violence calls out for changing arrest and prosecution policies as well as advocating for marriage counseling to stop the violence. This may be an attractive theory to some in the mental health field and “men’s rights” activists. The problem is that practitioners who endorse couples’ counseling while one person is still intimidating or using violence against another ignore the very real risk of violent assaults following counseling sessions. Most psychologists and therapists who have knowledge of domestic violence dynamics would concede that marriage counseling is ineffective if one party is a batterer and has power over the
other. How can a victim be honest about what is happening in the relationship or talk about the violence when she fears physical retribution?
Dutton and Corvo’s notion that women are as violent as men or that most domestic violence cases are mutual assaults simply distorts the reality that any law enforcement officer, emergency room nurse or doctor, or domestic violence advocate can validate from their experiences with victims. In 1998, national statistics showed that women were the victims in nearly three out of four of the 1,830 murders attributable to intimate partners in the United States (Rennison and Wechans 2000). That year, women experienced about 900,000 violent offenses by an intimate partner, and men, 160,000. Even if there is some underreporting by men, the claim that men and women commit assaults in equal numbers and with equal severity defies common sense. It is simply untrue. We acknowledge that women use aggression and violence in intimate relationships and not always in self-defense.
But we also contend that relying on family conflict studies that utilize the Conflict Tactics Study (CTS)—which simply counts acts of violence without accounting for the circumstances under which these acts occur and the size and strength of the people involved—is deceptive (Kimmel 2002). In a conflict study, a push in response to a beating would be scored as one conflict tactic for each party.
While there are women who kill their male partners for reasons other than self-defense, the numbers pale in comparison to men who kill their female partners when the female partner is trying to end an abusive relationship. In 2006, the U.S. Department of Justice reported that the number of men murdered by an intimate since 1976 had dropped by 70 percent. One third of female murder victims were killed by an intimate, while only three percent of male murder victims were killed by an intimate.
Stating that domestic violence is gender neutral is not only disingenuous, but also has serious public policy implications. How often do you read about a woman killing her husband and the couple’s children because the man is trying to leave the relationship? How many men are raped by an abusive woman as an act designed to punish or retaliate? Yet in city after city, male batterers stalk and murder their partners. Men and women use violence in very different ways, and therefore our response must be different.
Responses to Other Criticisms of the Duluth Model
Following is our response to some of the other criticisms of the Duluth curriculum and model:
The Duluth Curriculum Is Shamed-BasedThe Duluth curriculum does not use shame as a technique. Some critics mistakenly believe that because we have stated that it is important for facilitators and counselors to challenge sexist comments and an offender’s justifications for his use of violence that we are setting up a confrontational dynamic that shames group members. The dialogical process used in the Duluth curriculum does challenge men who batter to think more critically and reflectively about their beliefs, but this is done in a very respectful manner. We recognize that facilitators and counselors can use any program model, including Duluth’s, in shaming ways, and we have observed that some do. Our training strongly discourages facilitators and counselors from using confrontation in a shaming way.Conversely, our training instills facilitation skills that encourage critical thinking. Paulo Freire has heavily influenced our work. Freire’s pedagogical approach poses questions in a way that fosters dialogue (Freire 1990; Freire and Faundez 1989). While this may be challenging for group members, this process is neither shaming nor blaming. Facilitators and counselors using the Duluth curriculum are trained to be genuinely inquisitive about how and why a group member thinks the way he does. The deeper the dialogue travels, the greater the opportunity for a group member to recognize how his beliefs lead to self-defeating behavior and what he can do to change that.We have no doubt that some practitioners have inappropriately used the curriculum, but the criticism that the Duluth curriculum is confrontational and shame-based is simply false. Our experience is that when offenders increase their empathy, guilt may be a healthy by-product. We do not see this guilt as a basis for long-term positive change, but rather it is part of the change process.
The Duluth Curriculum Doesn’t Account for Women’s ViolenceAs earlier stated, there is a growing movement of practitioners who maintain that women are as violent as men or that women share responsibility for the violence. These practitioners often insist that domestic violence is a relationship problem and that marriage counseling should be an option for couples.The Duluth curriculum is designed for male perpetrators. In Duluth, a separate court-deferral program called Crossroads was designed for women who use illegal violence against the men who batter them (Asmus 2004). Most women arrested in Duluth have been able to document to the court a history of abuse against them by the person they have assaulted (past calls to 911 for help, protection orders, previous assaults, etc.). Those women who use violence against a partner with no history of that partner abusing them are not eligible for the Crossroads diversion program, but face the same consequences as male offenders after a conviction, i.e., a jail sentence or counseling in lieu of jail. The vast majority of women arrested in Duluth for domestic assaults are being battered by the person they assault. Most, but not all, are retaliating against an abusive spouse or are using violence in self- defense. The notion that battered women share responsibility for the violence used against them because of provocative words or actions is a dangerous form of collusion with men who batter (Mills 2003). We do not accept that these women should complete a batterers’ program. We do agree that there are a small number of women who use violence resulting in police action against their partners without themselves being abused. This is not a social problem requiring institutional organizing in the way that men’s violence against women is. For these women, a separate gender-specific counseling program may be appropriate.
The Duluth Model Is Not Culturally Appropriate for Marginalized Groups, and It Relies Too Much on the Criminal Justice SystemThe population of Duluth, Minnesota, is over 90 percent white, and the next largest cultural group (Native American) makes up less than five percent of the population. However, in developing the Duluth Model, the DAIP partnered with Native Americans and individuals from other marginalized communities to assess what effect greater state intervention would have on marginalized groups. As in most communities, the criminal justice system in Duluth most dramatically impacts low-income people and communities of color. It is the responsibility of program managers to work with law enforcement, the courts, and victim advocates to ensure that unintended negative effects of state intervention are addressed by community monitoring. The DAIP is not a state agency, so is better able to highlight and take leadership on how a CCR is disproportionately affecting marginalized people. Over the last 20 years, Duluth has modeled the development of culturally specific programs and designed a tracking and monitoring role that identifies and advocates for those having the least power within the community.Over-reliance on the criminal justice system is a legitimate criticism in light of the reality that, in many jurisdictions, a disproportionate number of men of color are arrested for domestic violence. In fact, when officers exercised total discretion on which offenders to arrest prior to the adoption of the pro-arrest policy, most men arrested in Duluth were Native American. The first year of the arrest policy saw arrest rates for whites rise from 11 percent to 81 percent of arrested offenders. Enforced arrest will result in a high number of men being arrested for crimes police previously discharged through practices like advising and separating. There are four questions regarding this issue that need to be addressed: Is there an over-reliance on the criminal justice system? What is the responsibility of the state to protect women of color who are battered? What are the alternatives to calling the police when a victim is being assaulted? How can men who batter be held to account consistently if we choose not to use the criminal justice system?The Duluth curriculum’s central focus is exploring and understanding power relationships and the effects of violence and controlling behavior on domestic partners. Dialogue about how more powerful people use their belief in entitlement and tactics of control to dominate the less powerful is a very useful comparison for all men, and especially men from marginalized groups. Some men may legitimately claim they are less powerful in our society, but they cannot use that experience to justify their abuse of women and children. Program counselors and facilitators who are members of marginalized groups themselves have found the tools of the curriculum can both liberate and hold men accountable. Few other curriculums make use of the parallels between personal power relationships and wider socio-cultural power relationships.We support alternative efforts by different cultures to confront violence against women in meaningful ways that will be accepted by men of color who batter. Community institutions such as religious organizations andcommunity groups can intervene at many levels. Men of color can mentor boys to not view girls and women as objects to denigrate, can model respectful relationships, and can take leadership by speaking out against violence against women.The Duluth Model—swift consequences for batterers, interveners that don’t collude, meaningful sanctions for offenders, consequences for further acts of violence, victim empowerment, counseling that focuses on stopping violence and changing beliefs, and community-wide expectation of accountability—will do more to deter domestic violence in all our communities. There may be value in providing culturally specific groups for men who batter, however, experimental evaluations of batterer programs showed no difference in outcomes for the culturally focused approach vs. more conventional approaches(Gondolf 2005). Research and dialogue on intervention and prevention strategies in communities of color should continue.
The Duluth Model Works Counter to Restorative JusticeThe development of the DAIP came from a framework where women’s safety and men’s accountability were foundational to practice. The field of restorative justice did not inform the DAIP’s work. We believe it is inappropriate to advocate for restorative practice when it puts women and the men who batter them together for purposes of healing or restoration.Generally, we do not advocate a “restorative justice” response to domestic violence because the risks of placing the victim of abuse in a setting where she may be re-abused by the restorative justice process are very high.Restorative justice is an increasingly popular field, and the term includes a wide variety of interventions, many adapted from pre-colonial indigenous practices. However, many restorative justice models are strikingly similar to mediation and couples’ counseling which have been shown to further endanger victims. This approach is promoted by practitioners who are not grounded in an understanding of the unique power inequities that characterize domestic assault crimes. We do not advocate for practices that frame a batterer’s change process around healing instead of changing his beliefs about violence and entitlement.The Duluth Model does have elements of restorative justice. It places victim safety at the center of institutional reform and, as the DAIP developed, many of our interventions turned out to be restorative in nature. However, group or couple meetings where victims confront their offenders (often a centerpiece of restorative justice) require huge resources to create a change of community and family culture that repudiates domestic violence and provides an effective means to monitor offenders and individual restorative interventions. Although the Duluth Model represents restorative justice in its broadest sense, it does not include victim-offender mediation and other similar face-to-face interventions because they do not adequately ensure safety for victims. In their 2003 article “The Role of Restorative Justice in the Battered Women’s Movement,” Loretta Frederick and Kristine Lizdas examine how restorative justice practices can increase risk for battered women:Because of the power balance implications of battering, it is critical that all current restorative justice practices should include screening for and exclusion of cases involving domestic violence. Any process that places the battered woman in a negotiating relationship with her source of fear offers her a false promise of hope and might, therefore, place her in danger. To date, the restorative justice movement has failed to adequately address these concerns.
The Duluth DAIP Is Anti-MarriageThe Duluth DAIP never encourages victims to leave or divorce their partners. DAIP advocates do encourage women to consider their options, try to accurately describe the program that their partners will participate in, and provide a realistic assessment about his potential to stop battering. Many battered women desperately want to hear from us that their partners will change their behavior. Since we can never predict which offenders will recidivate, we have long maintained that we’re ethically bound to share that information with the woman and help her assess whether her partner is actually changing. It is totally up to that woman to decide whether she wants to stay in an abusive relationship, give him more time to change, or get out of the marriage. Further, the DAIP does not see safety for battered women necessarily being enhanced by leaving the man who is battering her. In the short term, a woman may be safer if she stays with her partner while she plans for her future safety.The Duluth DAIP will recommend marriage counseling for some couples, but only after an offender has at least completed a 26-week program offered by the DAIP. We make these referrals only when counselors, advocates, and court personnel are relatively sure that the violence has stopped, the victim is not being coerced or intimidated, and she is not fearful of her partner.
The Duluth Curriculum Discounts Anger as a Causal FactorThe Duluth curriculum has been criticized for claiming that anger doesn’t cause violence (Dutton and Corvo 2005). We have never claimed that anger isn’t an emotion that often precedes violence or that some men who batter (as is often claimed) don’t have elevated levels of both anger and hostility (Maiuro et al. 1988). What we continue to emphasize is that teaching a batterer to control his anger will not stop the violence if the intent of the batterer is to control or dominate a partner.Defenders of anger management programs believe that teaching batterers to recognize what triggers their anger will help reduce violent outbursts. We believe anger management skills have limited utility in groups for men who batter. If a man who batterers stops his violence and truly relinquishes his domination and belief in entitlement over his partner, skill-building techniques like “time outs” and “cool downs” may help men manage poor impulse control and aggression. However, teaching “time out” and other anger management skills to men who want to maintain the status quo will only give a batterer another tool to control and manipulate his partner. An angry batterer who takes a “time out” leaves his partner waiting on pins and needles for his return. His partner will rightly wonder whether he will be calm when he returns or will have worked himself into a rage and violently attack her.In the United Kingdom, the Home Office, which has oversight over all probation offices, has sent instructions to probation staff to discontinue using anger management groups for domestic violence offenders because they are “inappropriate and ineffective and are a means for perpetrators to manipulate the courts” (Guardian 2006).To buttress their argument about men’s anger, Dutton and Corvo cite research by Neil Jacobson (1994) in which physically aggressive couples were studied in a laboratory setting. Jacobson’s study included groups of batterers he called “cobras and pit bulls.” The cobras had a history of antisocial behavior and their heart rates actually dropped when they were arguing with their wives, which would appear to debunk the “anger causes violence” theory. The group he characterized as pit bulls, men insisting on total control in their marriages, had batterers as fathers and usually confined their violence to their relationship. What Dutton and Corvo fail to mention is that Jacobson concluded that “psychotherapy doesn’t work with batterers. We should put our money elsewhere, into treatment programs to rebuild women’s live and into education programs to alert them to the signs of domestic violence.” Jacobson’s findings about anger and the conclusions he drew about psychotherapy for batterers are the antithesis of what Dutton and Corvo advocate.
The Duluth Curriculum Ignores Psychological ProblemsDo we have the resources to implement comprehensive assessments to determine what treatment should be recommended for every batterer who enters the criminal justice system? Are these assessment models accurate enough to justify their cost? Do we have the resources to provide individual psychotherapy for offenders with mental health problems? And should we?Some court-ordered domestic assault offenders do have psychological problems. In Duluth, the probation department, court, and DAIP attempt to flag those with serious mental health problems and refer them for psychological evaluations. Offenders with severe mental health issues, those who have been sexually abused, who are disruptive, or aren’t able to participate in a group process are referred for treatment at mental health centers connected to the DAIP. Less than 10 percent of court-ordered men are screened out of the program because of mental illness.Many people in our society could be diagnosed with personality disorders. Many people who commit crimes have personality disorders. Do we ignore the criminal behavior of batterers by not arresting them because of their personality disorders? In offender groups, do we not challenge the thinking of men who batter who are violent and controlling because they have an attachment disorder? Do we not ask men who experienced brain trauma as children to examine their beliefs about male entitlement? While not designed to deal with personality disorders,the Duluth curriculum is flexible enough for mental health practitioners to include therapeutic opportunities without colluding or compromising the principles of the Duluth Model. Finally, there is no evidence that a Duluth-based curriculum has a negative impact on offenders who have attachment disorder, depression, chemical dependency problems, or are antisocial.
Why Domestic Violence Happens and How to Stop ItCriticism of the Duluth curriculum Creating a Process of Change for Men Who Batter is frequently used to support arguments for adopting different treatment approaches. Such alternative approaches include psychotherapy, recommended for offenders who experience mental illness or brain trauma; marriage counseling, suggested by practitioners who believe that both parties share responsibility for the violence; and restorative justice, championed by those who believe that having victims confront their abuser is healing. Others argue that treatment is simply ineffective and should be discontinued in favor of community service or probation. The controversy and divided opinions about “what works” are centered in myriad theories about why battering occurs and how it can be stopped. Some hypotheses about the underlying causes of domestic violence include:
- Anger and poor impulse control trigger violent responses to relationship problems.
- Violence is a manifestation of a dysfunctional relationship.
- Violence is a learned behavior emanating from the offender’s family of origin.
- Offenders have diagnosable psychological problems or personality disorders.
- Men are socialized to accept violence as a means to resolve conflicts.
- A culture steeped in sexism provides a blueprint for men to use violence to control their intimate partners.
Each of these theories will result in different interventions and treatment approaches by practitioners charged with trying to stop the violence. Before adopting any new approach, practitioners should seriously assess how effectively a proposed intervention or treatment model enhances or diminishes the safety of victims.
What the Duluth Model Is and What It Isn’t
The Duluth Model was designed in 1981 as a coordinated community response (CCR) of law enforcement, criminal and civil courts, and human service providers working together to make communities safer for victims. It was organized by the Domestic Abuse Intervention Project (DAIP), an independent entity that intervening agencies entrusted to oversee and guide the Duluth CCR (Pence 1996).
The Duluth Model is not a treatment program, but rather a coordinated response by community institutions that holds offenders accountable for their behavior while ensuring that victims are protected from ongoing violence. The core elements of the Duluth Model are:
- Written policies that centralize victim safety and offender accountability
- Practices that link intervening practitioners and agencies together
- An entity that tracks and monitors cases and assesses data
- An interagency process that brings practitioners together to dialogue and resolve problems
- A central role in the process for victim advocates, shelters, and battered women
- A shared philosophy about domestic violence
- A system that shifts responsibility for victim safety from the victim to the system
In 1981, the Duluth Police Department adopted one of the first mandatory arrest policies in the country, which resulted in an exponential increase in offenders entering the criminal justice system. The court concluded that it was impractical to incarcerate “first-time” misdemeanor offenders, so they requested that the DAIP work with mental health agencies to design a program for these offenders. Both the court and the DAIP wanted to implement a low-cost program that wouldn’t compete with the limited funds that were being used to support battered women’s programs. The DAIP has adhered to this philosophy since its inception.
The Duluth Approach to Working with Men Who Batter
The Duluth curriculum Creating a Process of Change for Men Who Batter has been denigrated for being confrontational, humiliating, and shaming to men. Others have criticized the theoretical underpinning of the model, saying that its feminist political ideology alleges that all men want to dominate women.
Let us set the record straight. There is a reason that the curriculum Creating a Process of Change for Men Who Batter is not only the most widely used batterer intervention model in this country, but has also been adapted internationally for use in many different cultures (Pence and Paymar 2003). The success of this program, or any other program for men who batter, relies on effective advocacy for victims, the cohesiveness of agencies in the criminal and civil justice systems in monitoring participants’ progress in group, and swift consequences for violating court orders, failing to comply with the program rules, or any further acts of violence. We believe the curriculum’s success is also due to its clear underlying philosophy and to the group process that reduces collusion and provides a clear path for batterers to change if they decide that they want to change.
The Duluth curriculum is an educational approach. The philosophical core of the model is the belief that men who batter use physical and sexual violence and other abusive tactics to control their partners. Men who batter use violence to stop arguments, to stop their partners from doing something, and to punish them for noncompliance.
The authors of the curriculum borrowed from the work of the late Brazilian educator Paulo Freire. Freire worked with impoverished, illiterate people in South America. He developed an education model that relies on dialogue and critical thinking rather than traditional learning (banking of knowledge) in which the teacher feeds the student information (Freire 1990). We adapted these proven educational methods in our work with court-mandated offenders. We didn’t want group participants to simply repeat back what they assumed counselors or facilitators wanted to hear, but rather we wanted men in the groups to genuinely struggle with their beliefs about men, women, relationships, and entitlement.
A central assumption in the Duluth curriculum is that nature and culture are separate. Men are cultural beings who can change the way they use violence in relationships because beliefs about male dominance and the use of violence to control are cultural, not innate. Facilitators engage men who batter in a dialogue about their beliefs.
Through curriculum exercises, group participants are immersed in critical thinking and self-reflection. Some of the men in our groups begin to understand the impact that their violence has had on their partners, children, and themselves.
A key teaching tool is the control log that helps group members analyze their abusive actions by recognizing that their behavior is intentional and inextricably tied to their beliefs. It further allows men in the groups to recognize that, while in the short run their violence gets them what they want, it is ultimately self-defeating. Like a prisoner of war, a victim of repeated assaults and humiliation will resist, sabotage, and if possible, escape or will acquiesce and adapt in order to survive. Either option produces a relationship devoid of intimacy and love. This is the point where men who batter must decide whether to try and maintain the status quo or take the necessary and often difficult steps to change. Group facilitators teach skills through role-playing and other exercises so that participants become aware of alternatives to violence. This is not a psychotherapy-based curriculum and thus does not require facilitators to have extensive mental health qualifications, although some mental health practitioners have successfully incorporated therapeutic techniques that address personality disorders into the Duluth curriculum structure (Ganley 2006).
Conclusion and Questions to Consider
We believe anyone seeking to implement an effective response to domestic assault offenders needs to critically examine the available research and choose approaches that address domestic violence as a social problem rather than an individual dysfunction. We do not believe in a “one-size-fits-all” approach. We do believe that different offenders may benefit from different treatments before, during, and after attending 26-52 classes or groups using the Duluth curriculum. Taking a public policy perspective, we must ask hard questions and be open to evaluation and critique, but not ignore the historical roots of violence against women as if something has magically happened in the last generation to make all domestic violence gender neutral.
Back in the 1970s, battered women’s advocates were rightly concerned about how the mental health community used psychological explanations to describe wife beating. They correctly worried that battered women would be labeled psychologically and that mental health practitioners would collude with men who batter by treating offenders’ personality disorders rather than working to change their beliefs and attitudes about women, men, and marriage. The advocates were justified in their concern that shelters and advocacy programs would be forced to
compete for limited government and nonprofit resources against counseling programs for men who batter. They worried about the impact on families (especially low-income) who are required to pay fees (in some states, over
$1000) for counseling, regardless of the model. And they were alarmed that treatment providers would cease to be accountable to the criminal justice system, battered women’s programs, and victims.
All stakeholders should be open to having a frank and ongoing dialogue about the future of batterer intervention programs and the ways law enforcement, the courts, and other community institutions intervene in domestic violence cases. Despite our differences, we believe common ground can be found in our work. But we remain unshakable in our belief that the safety of victims of violence should be the core principle that guides our work.
Asmus, M. 2004. At a crossroads: Developing a prosecution response to battered women who fight back. Praxis International.
Babcock, J., C. Green, and C. Robie. 2002. Does batterer treatment work? A meta-analytic review of domestic violence. Clinical Psychological Review.
Barnett and Thelen. 1995. Domestic violence among male and female patients seeking emergency medical services. Springer Publishing Company.
Cascardi, M., and D .Vivian. 1995. Context for specific episodes of marital violence: Gender and severity of violence differences. Journal of Family Violence.
Dobash, R.E., and R. Dobash. 1983. Violence against wives. The Free Press.
―――. 1999. Changing violent men. Sage Publications.
Dutton, D., and K. Corvo. 2006. Transforming a flawed policy: A call to revive psychology and science in domestic violence research and practice. Aggression and Violent Behavior.
Ehrensaft, M.K., T.E. Moffit, and A. Caspi. 2004. Clinically abusive relationships in an unselected birth cohort: Men’s and women’s participation and developmental antecedents. Journal of Abnormal Psychology.
Freire, P. 1990. Pedagogy of the oppressed. Continuum Publishing Co.
Freire, P., and A. Faundez. 1989. Learning to question. Continuum Publishing Co.
Frederick, L., and K. Lizdas. 2003. The role of restorative justice in the battered women’s movement. Battered Women’s Justice Project.
Ganley, Anne. Deconstructing Psychological Theories, audio conference recorded by Battered Women’s Justice Project, 2006.
Gondolf, E. 2002. Batterer intervention systems: Issues, outcomes and recommendations. Sage Publications.
―――. 2003. Questioning the Broward experiment. Domestic Violence Report.
―――. 2004. Evaluating batterer counseling programs: A difficult task showing some effects and implications. Aggression and Violent Behavior.
Guardian (London) 2006. Offenders' anger control classes help make some more dangerous. April 2006. http://www.guardian.co.uk/.
Jackson, S., L. Feder, R.F. David, R.C. Davis, C.D. Maxwell, and G.T. Bruce. 2003. Batterer intervention programs: Where do we go from here? National Institute of Justice.
Hamberger, L.K., and C.E. Guse. 2002. Men’s and women’s use of intimate partner violence in clinical samples. Violence Against Women.
Kimmel, M.S. 2002. Gender symmetry in domestic violence: A substantive and methodological research review. Violence Against Women.
Langhinrichsen-Rohling, et al. 1995. Violent marriages: Gender differences in levels of current and past abuse. Journal of Family Violence.
Maiuro, R. D., T. S. Cahn, P. P. Vitaliano , B. C. Wagner, and J. B. Zegree. 1988. Anger, hostility, and depression in domestically violent versus generally assaultive men and nonviolent control subjects. Journal of Consulting and Clinical Psychology.
Mills, L. 2003. Insult to injury: Rethinking our responses to intimate abuse. Princeton University Press.
Moffit, T.E., A. Caspi, M. Rutter, and P.A. Silva. 2001. Sex differences in antisocial behavior. Cambridge University Press. Phelan, M.B., L.K. Hamberger, C.E. Guse, S. Edwards, S. Walczak, and A. Zosel. 2005. Domestic violence among male and female patients seeking emergency medical services. Springer Publishing Co.
Pence, E. 1996. Coordinated community response to domestic assault cases: A guide for policy development. Minnesota Program Development Inc.
Pence, E., and M. Paymar. 2003. Creating a process of change for men who batter. Minnesota Program Development, Inc. Rennison, C. L., and S. Wechans. 2000. Special Report: Intimate Partner Violence. Washington, DC: Bureau of Justice Statistics, May 2000.
Shepard, M. 1992. Predicting batterer recidivism five years after intervention. Journal of Family Violence.
U.S. Department of Justice. 2006. Homicide trends in the U.S.