Batterer intervention programs (BIPs) are one of several types of interventions designed to prevent the onset or continuation of intimate partner violence (IPV). Other interventions include (a) arrest, prosecution, sentencing, and probation of the offender; (b) services for victims of IPV, including counseling, crisis intervention, advocacy, children’s programs, and shelter; (c) couples groups; and (d) individual counseling. Couples groups and individual counseling are less often utilized due to concerns about the safety and blaming of victims in couples treatment and concerns about reinforcing the batterer’s code of secrecy in individual counseling. Nevertheless, both couples groups and individual treatment are viable interventions for other populations, and their application to batterers, with proper criteria, increases the intervention options for a very diverse group of people.
Although criminal justice actions and services for victims are not usually thought of as interventions for batterers, BIPs are now part of a larger community system of violence prevention in which criminal sanction and victim services are pivotal. Unlike mental health services, BIPs are not designed to be free-standing interventions, but a local node in a community antiviolence network.
BIPs are intended for people (usually men) arrested for domestic violence, for people who would be arrested if their actions were public, or for people who believe their aggressive behavior toward partners or ex-partners is troubling in some way. Men from this latter category of self-referred batterers are often dubbed wife referrals by practitioners who doubt the true motivation behind a man’s self-referral to a BIP. One of the unintended consequences of BIPs is that a man’s participation may support his belief that he is changing his behavior but his partner is not changing hers, therefore increasing his risk for IPV. Research suggests that self-referred batterers are more likely than court-referred batterers to drop out of the BIP and to reoffend.
BIPs usually consist of a short evaluation followed by anywhere from 3 to 12 months of weekly groups. These groups may be educational, treatment oriented, or focused on personal growth, but there are usually elements of all three in a BIP, in varying combinations. BIPs may also include other intervention elements, such as personal counseling, case management, addiction treatment, parent education, mentoring, or programming drawn from cultural and ethnic traditions. BIPs may be focused on partner violence by men or by women, by heterosexuals or by people in same-sex relationships, but groups are usually not mixed by gender or sexual orientation. BIPs are often housed in nonprofit or private agencies, and less frequently in the criminal justice system or in public institutions. The details of conducting batterer intervention programs are readily available in a number of topics and papers. Most states and provinces require that BIPs meet standards, and most standards require that the staff of BIPs meet specific educational and training requirements.
The current focus is on group-based, same-sex groups for men. There are two theoretical perspectives that, although seemingly in conflict, are usually combined in practice to form what is called the standard model BIP. The original BIPs emerged from the women’s movement of the 1970s and suggested that men’s violence against women was socially supported as a means of maintaining male dominance of women. The function of a batterer program drawn from this tradition is to help men change their minds about male dominance through a process of psycho-education and community activism. The Domestic Abuse Intervention Program in Minnesota is the most widely known of the psycho-educational approaches, and a sizable proportion of BIPs identify their program as a Duluth model. The Duluth “power and control wheel” is ubiquitous in BIPs, regardless of theoretical orientation.
The second perspective on BIPs is based on cognitive-behavioral (CB) treatment principles. In a “CB” group, the emphasis is on learning new skills, including identifying triggers for violence, interrupting the escalation process, managing anger, and substituting prosocial behaviors for controlling behaviors. In practice, Duluth-type programs engage in CB treatment and CB group leaders are often feminists, so the distinction between CB and Duluth-type approaches is fuzzy; in fact, the thoughtful combination of these approaches forms a more complete explanation of battering and batterers. The standard model BIP in the United States at the present time is best characterized as a profeminist CB psycho-educational program.
The typical batterer program accepts both voluntary and court referrals, although since domestic violence is a crime, most programs prefer that men are referred as a condition of their prosecution or probation. Same-sex offender groups are usually the preferred modality because they allow for peer feedback and reduce the isolation and private behavior common to batterers.
Nobody knows how many BIPs there are. At this time, there is no viable national organization of BIPs, nor are BIPs registered at the federal level, so all the information about them must come from state networks or licensing bodies. Despite their growth in the past 20 years, BIPs serve far fewer batterers than programs for battered women serve victims. In Illinois, for example, researchers estimated that approximately 12,000 batterers were in BIPs at some point during 1998, which was less than a third of the number of victims served by Illinois victim service agencies during that same year.
Process, instrumental, and Outcome Goals
The core issue addressed by this entry is the effectiveness of BIPs. Knowledge about batterer program effectiveness is important because courts now routinely refer men (and some women) convicted of domestic abuse to BIPs, suggesting a certain level of public confidence in the effectiveness of these programs. Is that confidence justified? Another reason to puzzle over BIP effectiveness is that the victims of domestic violence often want to remain in a relationship with their partner, and are looking for help in changing their partner’s violent and controlling behavior. Is that help reliable? A batterer’s seeking counseling is one of the strongest predictors that a woman will leave a domestic violence shelter and return to her batterer. Consequently, victim advocates and policymakers are justifiably concerned that BIPs not hold out a promise of help that may eventually become a vehicle for injury. A third reason to be concerned about BIPs’ effectiveness is that people who work with batterers are interested in outcomes so they can improve the level of program effectiveness. Is the research applicable to practice? For these people, the concern is often less about whether batterer programs work than how they work, for whom do they work best, and which elements of the program are most important. A final reason to question BIP effectiveness is that BIPs are increasingly likely to be funded by public dollars. Are these dollars well spent? Would these dollars be better spent on additional services for victims? Or, given the prevalence of substance abuse and mental illness among batterers, should public dollars be directed toward integrated programs for batterers and substance abusers or programs for mentally ill batterers? Researchers are in the early stages of answering these questions. Evaluation of BIPs using well-designed studies is relatively new, and the confidence researchers have in answering these questions is limited.
The first area to be addressed in any review of program effectiveness is: At what are they effective? BIPs have three orientations or sets of goals: (1) victim safety, (2) accountability and justice, and (3) rehabilitation. The achievement of the first goal, victim safety, is usually indicated by nonabusive behavior during and after a BIP. This is the standard indicator of BIP program effectiveness, usually measured by either victim report of IPV or criminal justice records of re-arrest.
Accountability and justice, the second possible goal for batterer programs, is usually a process or formative goal rather than an outcome goal. This goal asks to what extent batterers comply with program referrals, attend groups, and complete their probation requirements. Batterer programs, in their emphasis on accountability, are an extension of the criminal justice system. In the “New York model,” a popular accountability-based approach to BIPs, proponents argue that the batterer’s behavior in the program is less important than how the community responds when the batterer is noncompli-ant or reoffends. For advocates of this approach, the outcome to be measured is at the community level rather than at the individual level.
A final goal for BIPs is rehabilitation of and behavioral changes in the batterer, such as skill building, attitude change, and emotional development. These behavioral changes are viewed as instrumental in creating nonviolent behavior. In lieu of using more difficult measures from victims or the criminal justice system, some programs consider these instrumental variables as legitimate program outcomes. Changes in state or trait anger, misogynist attitudes, situational endorsement of violence, drug and alcohol use, gender-sensitive language, emotional expression, partner-specific assertiveness, social support, or other risk factors for abuse can be documented to measure immediate changes as a result of the program.
Goals for batterers programs can be process oriented (e.g., accountability), instrumental (e.g., attitude), or outcomes (e.g., recidivism). The fact that some of these goals are instrumental or process goals rather than outcome goals does not deter from their importance. A superior BIP evaluation would attend to all three kinds of goals. Understandably, for most evaluations of BIP effectiveness, the primary goal is recidivism: After admission to the BIP, is the batterer re-arrested, or does the batterer’s partner report physical or nonphysical abuse since he started the program? As it turns out, the “batterer’s partner” is usually a moving target. In a major study of over 800 batterers in four well-established BIPs in Pittsburgh, Dallas, Houston, and Denver, researchers found that 50% of the men in the study were not living with the index victim at the time of admission to the BIP, but 30 months after admission, 20% of these men had new partners, a quarter of whom had already been assaulted.
A number of issues complicate the question about whether BIPs are effective. Among these issues are the definition of what constitutes abuse, high rates of attrition from BIPs, cultural mismatching, and co-occurring problems such as substance abuse. One of the advantages of using victim reports to indicate outcome is that the victim can be asked about nonphysical forms of abuse. Nonphysical abuse, unless it involves threats, is legal, and will not come to the attention of law enforcement authorities. Therefore, batterers who recidivate with nonphysical abuse are often not counted as program failures. Some researchers argue that nonphysical abuse and control is a qualitatively different category of behavior from physical abuse, with different risk factors. Nevertheless, much of the content of contemporary batterer intervention programs is focused on learning noncontrolling behavior. A longstanding suspicion of advocates observing BIPs is that men may learn to avoid physical abuse by substituting more economical and legal forms of control such as intimidation and isolation. Consequently, ignoring nonphysical abuse overestimates the effectiveness of batterer programs.
A second complicating factor for examining BIP effectiveness is program completion. On average, 50% of BIP participants never complete the program, regardless of whether or not a court ordered them to participate. Recidivism rates for men who drop out of BIPs are greater than for men who complete the program, so calculating recidivism based on the minority of men who complete the program results in an artificially low rate of recidivism. On the other hand, calculating recidivism based on all men referred to the program, regardless of whether they complete it or not, underestimates the impact of the program because dropouts did not get the full “dose.”
The next consideration related to BIP effectiveness is the frequent mismatch between the culture of the program, including the ethnicity of the group leaders, and the culture of the participants. For example, while African American men are overrepresented in BIPs, they do not fare as well in them as do other men. Culturally focused intervention is proposed as a specialized approach for ethnically homogenous batterer groups that focus on cultural issues linked to preventing violence. At present, the effectiveness of culture-focused programs over other forms of batterer treatment has yet to be firmly established, although the same judgment could be made about any approach to batterer treatment. In general, African American and Latino men have the same reassault rate and generate the same level of victim fear as Caucasians, despite not participating in special culturally focused programs. However, there is a significant difference in the dropout rates of these ethnic groups, with African Americans dropping out twice as often as Caucasians. A clinical trial of culturally focused counseling compared to conventional batterer counseling for both racially mixed and all African American groups found no between-group differences in partner-reported violence at follow-up. However, despite the no-difference finding, that clinical trial also found that men who scored high in cultural identification were more likely to complete the all African American groups. These findings provide support for continuation of cultural-specific programming, particularly for men to whom ethnic identity is important. Culture-focused programs, while not yet superior to other groups in terms of preventing recidivism, may be superior in preventing dropout.
At least half of batterers referred to BIPs through the courts have co-occurring substance abuse or mental illness issues. Some BIPs screen out substance abusers or men with serious mental disorders, but the current standard of practice is that batterers who have co-occurring substance abuse problems should be in a BIP and in substance abuse treatment at roughly the same time, either in separate programs coordinating their services or in an integrated program addressing the issues concurrently. Integrated substance abuse and BIP intervention has been found at follow-up to be more successful than traditional serial or parallel interventions at engaging offenders in treatment, maintaining offenders in treatment, and reducing re-arrest.
The bottom line for BIPs is whether they prevent future episodes of physical violence as measured by partner report and/or official records. Before considering the findings of quasi-experimental and experimental studies, it is necessary to note that the science of all the studies on BIPs to date is less than satisfactory, so caution must be exercised about the conclusions drawn. Experimental studies are always challenging to conduct in the field, outside of the controls afforded a laboratory setting. BIP experiments are especially challenged in three key areas: random assignment, subject attrition, and difficulty with victim contact. Several of the studies of BIPs also lacked a control group, which makes it impossible to attribute outcomes to the program rather than a number of other potential causes, even when batterers have been randomly assigned to treatment groups. Another issue is that random assignment may break down when officials change the assignment of a batterer from one experimental condition to another. Researchers then have to decide whether to consider the men as belonging to the group to which they were assigned or the group in which they actually participated.
The second problem with BIP experiments is attrition. On average, half of the participants in BIPs do not complete the program, regardless of whether or not they were court ordered to participate. Studies have found that the “dosing” effect of keeping men in programs longer may have a direct effect on outcome, even after such studies have controlled for other differences between dropouts and completers. The number of sessions attended is an important predictor of recidivism, and successful completion of all treatment sessions reduces the likelihood of re-arrest.
The third difficulty of BIP experiments, victim contact, is an issue because of the difficulty in contacting past and current partners of batterers, as well as potential problems for victims as a result of such contact, as noted by Edward Gondolf. Victim contact is the preferred data source for recidivism in BIP research because of the poor reliability of official records compared to the reports of victims. However, victims may be coerced to participate in BIP studies by both their partner and the researcher. Any form of coercion is unethical. Also, victims are very hard to locate, and for a good reason: their safety. Coercion and inability to contact victims make using victim reports a difficult job, and for those reasons, many studies use arrest records. Arrest is a much less sensitive indicator of reoffense than victim report. For example, one study found that the proportion of arrest to victim-reported abuse was 1 in 35; that is, for every reported arrest, there were 35 assaultive actions. A second problem with arrest is figuring out what it means: Is arrest an individual marker of recidivism or a systemic marker of accountability?
Following are brief summaries of six experimental evaluations of BIPs. The studies are presented in the order they were published.
Minneapolis, Minnesota, randomly assigned 283 batterers to one of three programs (self-help vs. educational vs. combined) and one of two program intensities (weekly for 3 months vs. twice weekly for 4 months). A 6-month follow-up with 92 program completers and their partners found no significant differences between models or intensities. The main application of this study supports the contention that length of treatment is not an important consideration in BIPs.
Ontario, Canada, studied 59 men convicted of wife abuse, placed on probation, and randomly assigned either to a 10-week batterer program at a local family service agency or to probation with no batterer program. Three of the 30 men (10%) assigned to the batterer program reoffended, according to police records, compared to 8 of 26 men (31%) receiving probation only. The Ontario study provides support for the modest effectiveness of short-term BIPs.
Madison, Wisconsin, randomly assigned 218 batterers to cognitive-behavioral or process-psychodynamic group treatments. In a 18- to 54-month follow-up with program completers, there were no differences between the two treatment approaches in arrests or in victim-reported violence or fear of violence. However, men who had higher levels of dependency did better in the process-psychodynamic treatment, and men who had a more antisocial orientation did better in the cognitive-behavioral program. This study, currently being replicated, supports the suggestion that the “one size fits all” approach to BIPs may not be the best approach; matching batterers to program orientation may lead to a better fit.
Brooklyn, New York, reports the findings of a study of 376 men convicted of misdemeanor domestic violence and randomly assigned to 26 weeks of a Duluth model BIP, 8 weeks of a Duluth model BIP, or community service. At 12-month follow-up, men in the longer BIP were less likely to reoffend than men in the shorter BIP or men in the control condition, although partner report differences were not significant. A key finding of this study is its support for the value of longer-term programs over shorter-term programs.
San Diego, California, compared outcomes for U.S. Navy batterers randomly assigned to a 1-year cognitive-behavioral BIP, a 1-year couples group, a rigorous monitoring program similar to assertive probation work, or a safety planning condition approximating a control group. Men with substance abuse problems or mental disorders were excluded from the study. At 1-year follow-up, there were no differences in reoffense between the four groups. Unfortunately, the research protocol for this experiment not only excluded substance abusers and men with mental disorders but, due to the navy sample, also excluded men with prior criminal records, unmarried men, and unemployed men (i.e., most of the men who are seen in typical BIPs). While questionable as an indicator of normal batterers program effectiveness, the navy study serves as a useful indicator of the effects of assertive community intervention. The overall recidivism rate was 30% by spouse report and 4% by arrest. These figures compare very favorably with those of other interventions. We can conclude from the navy experiment that communities that take a proactive response to domestic violence—assertive probation work, sanctions for noncompliance, victim safety monitoring, and BIPs—are more likely to reduce the incidence of repeat violence.
Broward County, Florida, randomly assigned 404 male defendants convicted of misdemeanor domestic violence to either probation and 6 months of a Duluth model BIP or probation only. At 12-month follow-up, there were no differences between the BIP participants and regular probationers on measures of attitudes toward women, beliefs about wife beating, attitudes toward treating domestic violence as a crime, beliefs about the female partner’s responsibility for the violence, or estimated chance of hitting the partner in the next year, and official reports of recidivism. This study suggests that the hope of changing attitudes in BIPs may be misplaced. Another key finding of the Broward experiment is further support for the stake in conformity hypothesis: The men most likely to reoffend are those who have the least to lose, as measured by education, marital status, home ownership, employment, income, and length of residency.
In summary, the experimental studies of BIPs do not clearly answer the questions that have been posed about the programs’ effectiveness. In addition to the San Diego navy study, there is emerging evidence that coordinated community efforts in which the batterer program plays an integral role in violence prevention are more effective than situations in which the batterer program is viewed as the singular intervention for men who batter. The advantage of longer-term interventions over shorter-term interventions has not been clearly established, despite the trend for longer BIPs. Nor has the advantage of one program type over another program type been established. Men who are more vested in society do better in BIPs than men who are at society’s margins. The most concerning finding so far is that BIPs do not appear to have a clear edge over arrest and probation.
The Multlslte Study
With support from the Centers for Disease Control and Prevention, researchers studied 840 batterers and their partners in four cities (Pittsburgh, Denver, Houston, and Dallas) every 3 months for up to 4 years. Ignoring random assignment in favor of in-depth description and victim-sensitive follow-up, the research team interviewed not only initial but also subsequent partners; used funnel interviewing to increase response sensitivity; considered multiple outcomes, including a quality of life inventory for victims; analyzed numerous intervening variables such as shelter and counseling; used process measures such as program participation; collected counselor ratings; studied ethnic diversity; and conducted qualitative interviews with both batterers and victims. This study represents the most complete information about BIPs and batterer intervention systems that researchers have to date.
The researchers concluded that batterer intervention systems have a moderate effect on future violence. At 4 years after BIP intake, 11% of the men had been re-arrested for domestic violence, but according to partner report, 46% had been violent at least once. However, for participants whose partners were interviewed 4 years after the BIP intake, nearly 90% had been violence free in the past year, and three quarters of the men had not been violent at all, per partner report, for over years. These findings are supported by qualitative interviews with victims, with 85% of female partners saying they felt safe. One of the most striking findings of this study is that more than half (24%) of the 46% reoffense rate occurred during the first 6 months after intake, the time during which the man was still in the batterer program. This suggests that BIPs are not the short-term deterrent that some thought they might be, but may have a more far-reaching impact. The researchers also found that a small group of men, about one in five, reassault continually, including while they are in the BIP, and never desist. BIPs apparently have little effect on this group of dedicated offenders.
The best predictors of reassault in any follow-up period were drunkenness during that period and the woman’s prediction of her own safety and probability of reassault. Both of these predictors are dynamic and change over time, suggesting that BIPs should pay at least as much attention to changes in the ba tterer while he is in the program rather than relying on static predictors at intake such as personality, mental illness, or substance abuse diagnosis. Static predictors of reassault were prior arrest for crimes other than domestic violence, severe psychopathol-ogy, and severe levels of physical abuse. These predictors, coupled with the finding that one in five men in the program were constant and undeterred offenders, suggest that there is a subset of batterers in BIPs who should not be there or, alternately, that the standard model of BIPs does not impact this subset of offenders.
Multisite studies provide a richer perspective on BIPs than the experimental studies described earlier. Nevertheless, the multisite study is quasi-experimental and, lacking random assignment, is limited in the conclusions that can be drawn from it. Meta-analysis provides an additional perspective, combining the results of all BIP studies to look for an effect size by program type. An effect size is an estimate of the effect of participating in a program versus not participating in a program. Two recent meta-analyses have digested the results of 24 studies on BIP effectiveness using victim report and arrest data.
The first meta-analysis of 22 BIP studies found that without treatment, the proportion of batterers who reoffend was 21% based on police reports and 35% based on victim reports. The effect size for both police and partner reports was 0.18, a small but statistically significant effect. Effects were larger for studies using quasi-experimental designs (.23 and .34) than those using experimental designs (.12 and .09). No significant differences were found between Duluth-type programs and cognitive-behavioral treatment, and the researchers pointed out that the actual differences between these two approaches are minimal. The researchers concluded that a woman is 5% less likely to be battered by a man who was arrested, sanctioned, and attended a BIP than a man who was simply arrested and sanctioned, a figure that corresponds to approximately 42,000 women a year in the United States. The 5% improvement rate for participants in BIPs was compared with other meta-analyses’ findings, such as 16% improvement for treatment of aggressive adolescents and 12% for correctional treatment of adult prisoners.
The second meta-analysis was more restrictive in requirements for the analysis, including only 10 studies with a total of 19 BIP outcomes. Of the 19 outcomes, 13 were positive, but only 4 of the 13 outcomes were above the 0.5 level where the effect is considered to be of a moderate size. Like most meta-analyses, the researchers found that effect sizes from experimental studies were smaller than the effect sizes of quasi-experimental studies. They also found that studies using victim report had virtually no effect.
Batterer intervention programs are growing in number but still lack the necessary scientific support to be clearly established as the intervention of choice in all cases. BIPs are usually identified either as a form of cognitive-behavioral treatment or as profeminist psycho-education, but in practice BIPs do not differ that much from one other. Regardless of orientation, BIPs look to reduce violence against partners, teach new skills, and help men be more accountable for their behavior. In empirical studies, BIPs have a small, but statistically significant effect. As BIPs improve their response to attrition, cultural issues, and co-occurring disorders, they will become a better fit for participants. BIPs are a necessary but not sufficient response to intimate partner violence.