Enhancing Safety and Rehabilitation in Intimate Violence Treatments: New Perspectives

Over the last 30 years, safety has been the driving concern in developing treatments for intimate violence and has significantly shaped systematic responses to this problem. One approach has been to develop coordinated community response models that encourage close collaboration among system and service providers to offer more comprehensive, wrap-around services to better guarantee safety.¹ Yet another approach has been to legally or physically separate batterers from victims. Pro-arrest policies, which were first introduced in the 1980s, address the safety of female victims by ensuring male batterers are held legally accountable for their offenses and are separated, at least briefly, from their victims.² Restraining or no-contact orders are also designed to legally separate batterers from victims, while the shelter system seeks to provide women victims a safe place away from home, should they need it. In time, women victims expressed their desire for treatment programs for their partners rather than arrest and incarceration.³ Courts began referring offenders to Batterer Intervention Programs (BIPs) such as the Duluth Model, where male batterers were treated in groups in isolation from their victims.³ This rehabilitation strategy has reinforced the trend to separate female victims from abusive male partners to address each party’s legal and treatment needs individually.

Such strategies have been designed almost exclusively to address male violence perpetrated against females and ensure female victims’ safety, while ignoring the fact that intimate violence afflicts both women and men.^4,5 As a result, much of the relevant theory and research has been similarly focused on a gendered conception of intimate violence. While this article reviews and answers the concerns of this literature, it underscores as well that new, more inclusive treatments may be applied to the complex range of intimate violence cases.

It is also important to acknowledge that criminal justice strategies often overlook the fact that many couples remain inextricably bound for a variety of reasons, regardless of intervention or divorce.^4,5 Despite no-contact orders or the threat of future violence, offenders and victims often have continued contact during or after state interventions.^4,6 Although no consistent evidence exists on how many women stay or leave, perhaps as many as half of victims persist in their relationships—and if they do leave, research shows that this process unfolds over time.^7–10 Many other victims, even if they have left the relationship, remain connected to their abusers through children in common, or other family and community attachments.

Safety planning, then, must recognize that some women will continue to cohabitate with an abuser, that others will reunite after an arrest or jail time, and that couples will continue to raise children together, despite the threat or presence of violence. As Pennell and Francis have argued, “many women who were abused saw their social connections at the heart of their decisions [to stay in or leave the relationship]. They knew some bonds endangered their lives; however, they also would never feel safe and empowered without links to others.”¹

With this reality in mind, Davies, Lyon, and Monti-Catania¹¹ examined the gap between battered women’s services and women’s specific needs and interests, including that some will stay in their abusive relationships. Davies et al. underscore the importance of engaging battered women in designing their own safety plans insofar as they, with their proximity to the batterer and experience in past incidents, are in a unique position to assess increasing danger and take part in implementing a more comprehensive safety plan.¹¹

In practice, most BIPs and probation services have inconsistent contact with victims and often rely on a one-time contact. Rarely do they conduct more comprehensive safety assessments that might include Campbell’s Danger Assessment Instrument,¹² primarily focused on lethality, or the Spousal Assault Risk Assessment (SARA) instrument,¹³ which predicts physical injury. The literature suggests that these assessment tools may be helpful in guiding professionals, but should not be the only measures used to gauge the likelihood of violence or its magnitude.¹⁴ Practitioners have advocated for ongoing risk assessment and case management support from BIP programs for the batterer, as well as the victim, whose appraisal of her safety has been shown to be a significant predictor of additional violence.^1,11,14,15 Those closest to the victim and offender may also be important resources in predicting, and possibly preventing, future violence.¹

Given safety concerns and, at the same time, the propensity of many couples to reunite after a domestic violence incident, communities must recognize the importance of engaging those closest to the violence and their social supports in order to enhance safety and treatment. This article surveys the evidence of safety and rehabilitative effect of two treatment approaches: BIPs, which are widely used in the U.S., and restorative justice-based programs, a controversial alternative that allows women victims to participate in face-to-face meetings with their abusers, thereby challenging accepted notions of safety.

The Successes and Limitations of Batterer Intervention Programs

An estimated 300 judicial systems nationwide use special domestic violence courts to hold abusers accountable for their violence and employ court-ordered BIPs, judicial monitoring of defendants and probationers, and, when possible, “no-contact” orders to protect victims while the case is pending.¹⁶ The Duluth Model, one of the earliest and most influential BIP models in the U.S. today,³ uses a feminist, psycho-educational approach whereby men are taught that battering is part of a range of male behaviors used to control women.^17,18 This approach presumes that all abusive men are equally socialized, and that their partners, parents, and even a violent community are secondary to the central causes of abusive behavior.¹⁹

Accordingly, BIPs generally dismiss individual cultural, racial, or religious identity, substance abuse, or a childhood history of abuse as less relevant to the rehabilitation process. Indeed, many advocates have argued that diverting the batterer’s attention toward family history, cultural identity, and dysfunctional behaviors excuses men’s violence and deflects the impact of male privilege and sexism on battering.^20,21 Empirical studies, however, have shown a strong correlation between a childhood history of domestic violence, sexual abuse, and other familial neglect, and an increased risk of adult offending and victimization.^22,23 Alcohol abuse has been found to be a consistent risk marker for repeat assault.¹⁵ Although substance abuse or a history of childhood abuse does not relinquish domestic violence offenders from accountability, the research suggests that effective interventions must consider these factors.^4,24,25

The research literature on BIPs remains constrained by methodological limitations and it is still unclear if BIPs are an effective intervention.³ Initially, studies of BIP programs indicated high rates of success in reducing the frequency and severity of violence among offenders; however, methodological issues such as small sample size and a lack of appropriate comparison/control groups, rather than actual program success, have since been shown to be important factors in determining the reliability of this evidence.^3,17 Recent evaluations using more rigorous designs have found little or no reduction in battering.³ Low response rates and a failure to determine model fidelity also continue to hinder this emerging research.^3,17

A major challenge to measuring treatment outcomes of mandated BIPs has been high attrition rates, including attrition prior to entering a program and failure to complete it.^26–28 In addition, many BIPs have difficulty maintaining contact with victims, further limiting opportunities to measure change in violent behavior patterns over time, ensure safety, or assess a program’s effectiveness in reducing violence from the victim’s perspective.^3,15,17 High attrition rates and failure to maintain contact with victims compromises the scientific quality of such studies.^15,17

An evaluation of BIPs based on the Duluth Model that was conducted in two cities—Broward, Florida, and Brooklyn, New York—found this intervention largely ineffective.³ In Broward, there were no significant differences in recidivism between batterers who did or did not have treatment, no differences on violations of probation, and no indication that those who received treatment modified their attitudes toward domestic violence. In Brooklyn, no evidence was found that this court-mandated batterer intervention promoted lasting change or that participation altered batterer’s attitudes toward partner abuse. No differences were found among all the treatment groups on attitudes toward violence against women or ways of dealing with conflict. These findings suggest that BIPs may simply repress the violence during the course of treatment.

However, a large multi-site evaluation of batterer intervention systems presents evidence of moderate program effect.²⁷ Overall, the four-year study found that men who had completed at least two months or more were less likely to re-assault their partners than those who had dropped out within two months. Similarly, after reviewing five experimental or quasi-experimental studies and computing an average effect size, Davis and Taylor concluded that there is fairly consistent evidence that BIPs can have a substantial effect on reducing violence.²⁹

In their review of the BIP literature, Saunders and Hamill suggest that programs sensitive to historical and contemporary experiences of particular groups may increase motivation to change and decrease attrition rates.²⁸ Similarly, a 2001 study by Gondolf and Williams found that African American men arrested for domestic violence have higher completion rates when the counseling group is culturally-focused.¹⁹ However, Gondolf subsequently reported that “there was no apparent benefit from the all-African American groups with conventional counseling or culturally-focused counseling.”³⁰

Despite the success of some programs, this review of the research suggests that batterer treatments must address an array of logistical and behavioral problems to have any widespread impact on domestic violence offending. Accordingly, effective programs must assess for substance abuse and mental health problems, and make services for these problems available concurrently.^19,31 In addition, the most dangerous time for re-assault appears to be at the beginning of treatment, and it is now recommended that intensive counseling services be offered in tandem with BIPs to offset this threat.¹⁵ Ongoing risk assessment and case management services for batterers and their partners can enhance victim safety; risk management may occur through systematic contact with victims, periodic reevaluations of safety, and additional support services as needed.¹⁵ Finally, extending the length of the BIP does not appear to keep offenders in need of more comprehensive services from re-assaulting their partners or from dropping out of treatment.¹⁵ Indeed, the issue of attrition is perhaps the most significant problem with BIPs. All of these program weaknesses emphasize the need to develop and rigorously study new treatment approaches to domestic violence.

Restorative Justice and Why This May Be a Better Approach

Restorative justice (RJ) is a promising, but not yet rigorously evaluated, approach to domestic violence. RJ practices (which usually involve one or a series of conferences or meetings) restore victims, offenders, and communities through participation of a plurality of stakeholders in the process of recovering from crime.³² At its core, RJ emphasizes interdependence between citizens and families, and assumes that all cultures will find this approach more emotionally satisfying than retribution.³² Those who have something at stake in the events that occurred define what restoration means in a specific context; however, it generally encompasses what matters to the stakeholders, including restoration from injury or lost property, restoration of dignity, social support, security, and a sense of empowerment.^32,33 While the modern RJ movement began in the 1970s, restorative practices were the dominant model of justice in many cultures until the modern era.³² Many indigenous peoples have never stopped using it.^34,35

The theory underlying restorative approaches generally is that all cultures must adapt their restorative traditions in ways that are meaningful to them.³² This allows great flexibility to address the needs of each offender, victim, or crime. In addition, all participants are regarded as having equal voices in the justice process and equal opportunities to air their concerns and participate in the discussion, and possibly resolution, of an offense.^4,36–38 A final idea that has emerged from such approaches is that restoring the parties to one crime can have a ripple effect in the family and community surrounding them, and may potentially expose and mitigate past offenses or prevent new ones.^4,39 In other words, the driving presumption of RJ is that justice can be personal, interactive, egalitarian, and transformative.

RJ models include victim-offender mediation, community reparative boards, family group conferencing, and peacemaking circles; all are united in theory but vary in approach. The focus of this article is family group conferencing (FGC), which gathers family and supportive friends of the offender and the victim, together with a facilitator and relevant child welfare and criminal justice professionals, in a structured setting to hold the offender accountable for the harm done, ensure victim safety, facilitate open dialogue about the violence, and to develop a plan to rectify the problem. The FGC’s underpinning theory is that family and social networks are in a strategic position to encourage the offender’s reform⁴⁰ and oversee the family’s plan to stop abuse as well as to monitor safety, thus preventing future violence.^33,41 While some offenders may be moved to compliance with domestic violence laws and by the intervention of police and courts,⁴² RJ proponents argue that the most potent force in changing behavior lies with the family and friends the offender trusts and loves.^4,43,44

Although communities have increasingly applied FGC to family violence cases, no country has systematically used this or other RJ models to stem domestic violence. While feminist academics and victim advocates continue to debate the utility of RJ practices in domestic violence cases, they agree that more interventions are needed to address the complexities of this problem.^1,4,45 In addition, advocates recognize that whether or not a woman leaves her abuser, she may remain connected to him through her children and that interventions must address this reality and enhance the safety of both the woman and children.^1,4 The growing use of family decision-making conferences in child welfare settings and the well documented co-occurrence of child maltreatment and woman abuse⁴⁶ has sparked a new discussion on how to apply RJ practices safely in intimate violence cases.^1,4,45,47

Critics of RJ fear that such an informal justice process will revictimize women victims of male violence.⁴⁸ They contend that RJ is too “soft” on male offenders; to truly legitimize the criminalization of domestic violence, society must punish men who abuse women. They also worry that RJ approaches may re-privatize domestic violence by locating the solution in a potentially patriarchal family.⁴⁵

Proponents argue that RJ processes facilitate conversations between willing victims, offenders, their families, and support networks, and increase the chance of condemning the violence while permitting victims to express their needs and concerns.⁴⁵ They contend that RJ encourages admissions of offending rather than denial, validates the victim’s experience, and provides assurances that the victim is not to blame for the abuser’s violent behavior.^4,37 RJ also gives victims a meaningful role in legal and treatment processes.^36,37 RJ may also offer more options to victims who believe that prosecution does not meet their needs.³⁶ RJ, they argue, attends to the lay, rather than legal, perspectives of crime and encourages a holistic understanding of the offense.⁴⁵ Finally, some who promote RJ believe that it can better attend to the complexities of intimate violence, including when the victim is male or the violence is bi-directional.⁴

Both critics and proponents of RJ recognize that attention to victim safety and offender accountability must remain a priority. Central to this debate is whether or not standards can be developed to make RJ models satisfactory and safe for domestic violence victims, while also considering just outcomes for offenders.^{38,44,49–51} Both groups recognize that the successful activism of some feminist advocates and academics has restricted application of RJ in domestic violence cases thus far; one consequence is the current dearth of scientific evidence to confirm or discount the critics’ or proponents’ claims.⁴⁵

Strong support for using RJ in domestic violence cases comes from Burford and Pennell’s study of a FGC-based approach to family violence in Canada.⁴² This study found a marked reduction in indicators of both child abuse/neglect and abuse of mothers/partners after the intervention, advancement in children’s development, and an extension of social supports.⁵² One year after the conference, the incidents of abuse/neglect were 50% less compared to the year before, while incidents increased significantly for 31 families in the control group, who did not participate in an RJ intervention.⁴⁷

Critics and proponents have expressed concern for victim safety and empowerment in RJ models where the victim and offender are both present.⁵³ Pennell and Francis¹ conducted a series of focus groups with survivors, shelter staff, and academics to discuss the use of “safety conferences,” which are drawn from the FGC model and gather together the victim and her social support network to make decisions regarding safety. Focus group participants viewed the safety conferences as an opportunity to eliminate the secrecy about the abuse, while also offering a comprehensive approach to planning for women’s safety using family support. Shelter staff pointed out that safety conferences have the potential to educate the family group about domestic violence, gather support for both parties to address the violence, and help remove the stigma of returning to or failing to leave an abusive relationship. All focus group members agreed that it might be too risky for some women, such as those residing in shelters or who lacked familial or emotional support, to participate in conferences; however, the survivors also felt that input from the abuser’s family, particularly from offenders’ mothers who may have also been abused, would be welcomed, and that the family could use FGC to exert control over the offender and to protect the victim. Finally, survivors agreed that, when possible, children should participate in conferences because they were aware of the violence and needed to be part of the deliberations.¹

FGC may actually provide greater victim empowerment and satisfaction than the criminal justice process alone. Despite court reforms to improve the legal experience of domestic violence and sexual assault victims, many remain dissatisfied and reluctant to participate in the criminal justice process.^44,54–57

Recent studies of FGC in Minnesota, Pennsylvania, Australia, and Canada have found higher rates of victim participation and satisfaction when compared to traditional approaches to crime.^42,44,54 While critics of FGC have often expressed concern over power and control dynamics during and after the conferences,⁴⁸ FGC partnership-building has been found to foster collaboration between family and service providers, and enhance safety and empowerment.⁵² Pennell and Burford found that despite concerns that the offender’s presence at the FGC would silence victims, female victims took leadership roles in deliberations and in developing plans to address offender accountability and problem-solving.⁵²

The paucity of research conducted so far on restorative justice interventions does not indicate a reason for the categorical exclusion of domestic violence from FGC.⁵⁸ Admittedly, FGC research is still in its infancy—especially in domestic violence where many programs exclude such cases in their referral criteria.^58,59 Several communities, however, are experimenting with RJ in gender crimes in Arizona, Hawaii, Canada, and Minnesota.^34,37,51 In Nogales, Arizona, for example, the community, together with the Santa Cruz County Court, has implemented Peacemaking Circles. This RJ-based domestic violence intervention draws on the lessons learned from FGCs and batterer treatment by improving standards and approaches to victim safety and enhancing the availability of supplemental services (substance abuse treatment, job training) to address related problems. Peacemaking Circles and a BIP, also offered in Nogales, are being compared in a randomized study currently underway at New York University, which has been funded by the National Science Foundation.⁶⁰ While this program and study signal an important new direction in domestic violence practice and research, the need to rigorously test RJ and other interventions remains.