• Center for Problem oriented policing

previous page next page

Responses to the Problem of Domestic Violence

Your analysis of your local problem should give you a better understanding of the patterns of domestic violence cases and calls in your jurisdiction. Once you have analyzed your local problem or important aspects of it and established a baseline for measuring effectiveness, you should consider possible responses to address the problem.

These strategies are drawn from a variety of sources, including descriptive materials, research studies, and police reports. It is critical that you tailor responses based on reliable analysis. In most cases, an effective strategy will involve implementing several different responses. Law enforcement responses alone are seldom effective in reducing or solving the problem. Do not limit yourself to considering what police can do: carefully consider others in your community who share responsibility for the problem and can help police better respond to it. In some cases, the responsibility of responding may need to shift toward those who have the capacity to implement more effective responses. (For more detailed information on shifting and sharing responsibility, see Response Guide No. 3, Shifting and Sharing Responsibility for Public Safety Problems).

General Considerations for an Effective Response Strategy

1. Implementing a comprehensive and collaborative response strategy. Comprehensive and collaborative approaches to reducing domestic violence are more likely to succeed than piecemeal approaches, but they require significant commitment from all participating. One research team wrote, "[a]n offense as complex as domestic violence is unlikely to be prevented by a single measure."[53] A thorough assessment of the current overall response to domestic violence can identify critical gaps in strategies, resources, and response protocols.

Although some communities have adopted a more integrated approach engaging advocates, police, and the criminal justice system,† for the most part, recidivism remains high. In the small studies of these integrated domestic violence approaches, there is evidence that victim satisfaction is high but insufficient evidence that recidivism and revictimization rates have decreased. One commentary suggests that the small core of persistent batterers (who are violent toward others as well as their intimates) are perhaps resistant to even highly coordinated efforts.[54]

† For a fuller description of the different types of integrated approaches adopted and their studies, see Buzawa and Buzawa, chapter 15 (2003).

To improve the likelihood that a comprehensive approach reduces recidivism and victimization requires a continuum of responses depending on the most reliable research and covering the different points in time most important to reducing domestic abuse: before an incident to keep it from occurring, during an incident to stop the immediate violence, and after an incident to reduce or prevent revictimization. It involves responses that focus on victims and potential victims and strategies that focus on offenders and potential offenders. As well, it involves the improved identification and reporting of cases of abuse between current and former intimates and dating partners.

The matrix below may help you organize the strategic focus, goals, and timing of your responses to domestic violence. This section is followed by information on the impact of specific responses to domestic violence strategies. Many of the responses will require the support and collaboration of other governmental agencies and community nonprofits.

Figure 2. Matrix of Responses to Domestic Violence
Strategic FocusStrategic Times for ResponsesGoalPolice RoleOther Agencies, Organizations, Group
At-risk populationBefore incidentsPrevention; persuade those at risk that, if abused, call the policeAlert and educate at-risk victim population; educate/warn at-risk offending populationPublic health organizations; domestic violence coalitions; schools and educators; medical professionals
Peers and neighbors of at-risk individualsOngoingGetting peers and neighbors to call the police if they learn of domestic abuseEducate these groups about the importance of calling the police to reduce the violencePublic health organizations; domestic violence coalitions; educators
Injured women and menDuring medical careScreen the injured for domestic violence; raise awareness of available services; provide medical careEngage the medical profession and link medical professionals with appropriate referral organizationsMedical professionals
Individual incidentDuringViolence cessationStop the violence; identify primary aggressor; accurately identify abuse historyMedical and public health professionals
Immediately after incidentAfter; ongoingPrevent revictimizationAssist with victim safety; develop tailored strategies for victim and offender based on risk/physical violence history; increase focus on high-risk offenders; ensure victim is linked with needed resources; increase focus on high-risk victims; ongoing monitoringDomestic violence victim advocates, victims' friends and family, shelters, victim services, criminal justice system, treatment services
 2. Educating collaborative partners. Each partner in a domestic violence reduction collaborative brings a unique perspective and body of knowledge. It is important for members of domestic violence collaborations to operate with precise and accurate information about what does and does not work and about the dimensions of the problem in your community. Do not assume that other professional groups participating in the collaborative have the most up-to-date knowledge about the problem or are following prescribed protocols.

3. Tailoring the police response on the basis of offender and victim risk. Some researchers advocate a graded response to domestic violence. They view batterers along a continuum—some are easily dissuaded from rebattering, others require increased actions—and a graded or tiered approach to control offender behavior can be effective. For instance, we know that a percentage of batterers is deterred from rebattering simply by having the police called on them, so encouraging reporting is essential; others may be deterred with the additional application of a restraining order. More is required to keep other batterers from rebattering, which suggests a need to refine assessments about who these batterers are.[55] Some researchers also suggest a graded approach to victim safety for similar reasons: some victims are at low risk of being revictimized and some are at higher risk.

Graded approaches to both victim and offender can be effective ways to reduce revictimization. British researchers conducted several evaluations and found that significant reductions in repeat victimization are achievable.[56] The key to reduction is that, at each level, police (and others) must focus equal attention on the victim and the offender. The victim and the offender must know about the actions police have and will take in relation to each other. Graded approaches must be applied quickly because the highest risk period for further assault is within the first four weeks of the last assault.[57]

Under a graded approach victims are assigned to one of three follow-up response levels based on the following:

  • The harm the abuser caused or threatened
  • The victim's account of previous incidents not reported to the police
  • All incidents recorded in police databases involving the same victim with the same offender†
  • The offender's previous violence to other victims.[58]

† Incidents other than domestic violence are reviewed because they may be indicators of escalating aggressiveness (Hanmer et al. 1999).

Each of the three levels to which a victim is assigned requires some follow-up. There is in an increase in the variety and intensity of safety measures used to protect the victim and to restrict the offender from rebattering if the offender's violence and criminal history indicates he is at an increased risk to re-offend.[59] Evaluations of graded approaches found reduced domestic violence calls and increased time intervals between violence.[60] The methods to protect the victim or deter the offender can include a variety of situational crime prevention opportunity blocking mechanisms, such as the following:

  • Increased police surveillance of victims' homes
  • Greater coordination with other parts of the criminal justice system
  • Pendant alarms for at-risk victims
  • Video cameras placed in high-risk victims' homes
  • Cocoon watch over victims†
  • Target hardening of victims' vulnerable properties
  • Police watch of offenders
  • Police opposition to bail
  • electronic ankle bracelet monitoring of high-risk released offenders
  • alarm-activated recording devices with two-way speech capability (allowing victims to speak directly to the police, and vice versa).‡

† With a victim's permission, neighbors, relatives, friends, or all three are asked to look out for the victim and immediately call if the offender returns. 

‡ To view the 25 techniques of situational crime prevention, see www.popcenter.org/25techniques/ in the Center for Problem-Oriented Policing.

Tailoring police responses to particular offenders based on the seriousness and frequency of their offenses has been successfully applied in the context of conventional crime and may be as useful for dealing with domestic batterers.§ This is likely because much domestic violence is committed by repeat and chronic offenders who may be particularly vulnerable because they often are under various forms of legal supervision due to past offending.[61]

§ See Kennedy, Waring, and Piehl (2001) for a description of the pulling levers/focused deterrence application in the homicide context, and Spelman (1990) for a discussion of repeat offender programs. See Kennedy (2002) for a discussion of the application of the approach to domestic violence.

According to Buzawa and Buzawa, "The criminal justice system must develop the capabilities to identify those batterers for whom normal deterrence can be effective, perhaps the majority in terms of numbers of incidents… [i]t should also be able to differentiate, segregate, and incapacitate batterers who must be deterred by special approaches."[62]

Specific Responses to Domestic Violence

4. Educating potential victims and offenders. Some police agencies participate in domestic violence awareness campaigns and school programming, such as classroom instruction to teens about dating violence and ways to handle conflict. Domestic violence prevention messages may target the general population or specific populations. For example, campaigns may be designed to encourage victim reporting, deter potential offenders, or raise the consciousness of potential witnesses of abuse (neighbors, friends, relatives).† However, the effect of these prevention strategies is unknown.[63] For instance, few of the programs developed to reduce teen dating violence have been evaluated, and of those that have, there have been mixed results.[64] Although some report an increase in knowledge in the targeted population and greater familiarity with available resources to help victims, this does not necessarily translate into a reduction in the incidence level of dating violence.

† The Lancashire (United Kingdom) Police Constabulary placed messages about domestic violence on police vehicles, beer glass coasters in bars, utility bills, and lampposts, and used radio advertising to increase awareness of domestic violence.

As a rule, prevention is more likely to work if highly targeted. General campaigns are not typically effective. Highly targeted campaigns that focus on a specific target group or geographic area can have some impact. Offender-oriented campaigns, which are designed to raise potential offenders' perceptions that there will be meaningful consequences to battering, are more likely to be effective than campaigns that appeal to potential offenders' morals.‡

‡ For more in-depth information about prevention campaigns and the conditions under which they are most likely to be effective, see Response Guide No. 5, Crime Prevention Publicity Campaigns.

Prevention efforts targeting potential victims should focus on those at higher risk, such as young women ages 16 to 24, as they experience the highest rates of intimate violence. Special efforts should be made to reach the poorest women in this age group as they are at an even higher risk. In addition, some recent immigrant communities, depending on the laws and privileges in the home country, may show a high level of domestic abuse, particularly if there is a lack of familiarity with assault laws in the adopted country.† One of the reasons crime prevention campaigns have had limited success is that potential victims do not see themselves as such; victim-oriented prevention campaigns must overcome this threshold issue.

† For a review of the research pertaining to domestic violence and immigrant populations, see Buzawa and Buzawa (2003). See also Overcoming Language Barriers: Solutions for Law EnforcementPDF

5. Encouraging domestic violence victims and witnesses to call the police. Police and other members of a domestic violence reduction collaborative should encourage people to call the police if they are victims of, witnesses to, or know a victim of domestic violence. Prevention and education efforts should include this as a core message. A study of more than 2,500 domestic violence victims concluded that calling the police had a strong deterrent effect on revictimization, even when the police did not make an arrest, when the offender had a prior history of violence against the victim, and when the assault was sexual. Calling the police was beneficial even when the violence was severe. In addition, offender retaliation did not appear to be more likely even when a victim rather than a third party called police.[65], †

‡ The study, which examined felony and misdemeanor violence, male and female offenders, and couples in different types of relationships, tracked victims for three years.

6. Encouraging other professionals to screen for domestic violence victimization and make appropriate referrals. The American Medical Association adopted domestic violence screening and referral guidelines for medical practitioners.[66] Physicians should screen injured women patients to determine if domestic violence was the cause of the injury.§ Medical professionals should also discuss domestic violence with pregnant patients during prenatal checkups. Physicians' documentation of specific incidents of domestic abuse can be critical to the successful prosecutions of batterers.¶

§ The American Academy of Pediatrics, the American Medical Association, and the American College of Obstetricians and Gynecologists all endorse screening.

¶ See Isaac and Enos (2001) for guidelines for proper medical documentation of battered patients.

In spite of these professional recommendations, most physicians are reluctant to routinely screen women for domestic violence, citing a lack of training in how to conduct screenings and insufficient knowledge of appropriate responses and referrals when a patient discloses domestic violence.[67]

7. Providing victims with emergency protection and services after an assault. Battered women's shelters protect women from further harm after an assault, sometimes on referral from the police and sometimes not. Typical services include a domestic violence hotline, temporary housing, information and referrals to other social services, safety planning, victim advocacy for emergency benefits or at court proceedings, and referrals for legal services.†

The first shelter for battered women and their children opened in London in 1972. There are currently more than 2,000 shelters in the United States.[68] In most communities, shelters raise community funds for operation; in some communities, police contribute a portion of their budget to aid in shelter operation. Shelters often rely on volunteers and a few paid personnel to provide round-the-clock assistance to battered women and their families. Little is known about the number of repeat victims served annually, the length of average stays, or the effectiveness of shelter services in preventing subsequent violence.

Although there are confidentiality issues to resolve or respect, police should seek to exchange information with domestic violence victim service providers as much as possible to learn more about the domestic violence victim population, some of whom do not seek out police assistance. This information exchange can aid in identifying the highest risk victims and offenders, targeting prevention efforts, designing safety plans, and learning more about the community's offenders.

† More than 60 U.S. law schools offer student advocacy services for domestic violence victims at court proceedings (Roberts, 2002b).

Recently emerging are family justice centers, which house domestic violence victim services in one location to increase victim survival, independence, and recovery. Formerly, victim services were scattered in different places, sometimes at opposite ends of cities. If victims followed up with these fragmented services, they too often experienced the frustration of retelling the story of violence to every individual provider.‡

‡ The San Diego Family Justice Center provides victims with advocacy, childcare, clothing, counseling, court support, deaf/hard-of-hearing assistance, emergency housing, food, forensic documentation of injuries, housing for pets, internet access, law enforcement, legal assistance, locksmith services, medical services, military assistance, phones, phone cards, restraining orders, support groups, safety planning, spiritual support, transportation, and victim compensation (San Diego Family Justice Center).

8. Assessing the threat of repeat victimization. Gathering accurate information about past abuse, including unreported incidents, is critical in assessing a victim's current risk and tailoring appropriate offender interventions. Although there is no foolproof profile of a repeat batterer or repeat victim, having a complete and accurate picture of both the victim's and the offender's history of abuse is useful for predicting future risk. You may need to gather records from other jurisdictions where the victim or offender lived. You should also be alert for other related behaviors such as threats of violence, harassment, trespassing, vandalism, stalking, protective order violations, and prior use of a gun, as these behaviors offer clues as to whether abusive behavior is chronic and/ or escalating.

You should supplement official records with the victims personal knowledge about the offender. Keep in mind that for a variety of social and psychological reasons, victims may be reluctant to reveal the extent of the battering, particularly to police.

Police may consider soliciting the assistance of trained medical professionals to help determine a victim's abuse history. A study of one initiative in which a doctor and a nurse (or paramedic) accompanied police on domestic violence calls found victims revealing much more about the extent to which their partner battered them than police typically elicit, suggesting that victims may feel more comfortable reporting repeat victimization to medical professionals. Even with the high levels of repeat victimization uncovered, few victims had sought counseling, shelter, or medical treatment for the prior assaults.[69]

There is great interest in developing an assessment instrument police can use to predict and help prevent domestic violence homicides. Although some such instruments exist, they tend to over-predict lethality. This is because only a very small portion of domestic violence victims are murdered and distinguishing between victims who will be murdered and those who will not remains elusive. For instance, even though offender unemployment is a risk factor, the vast majority of unemployed abusers do not murder their current or former intimates. Even when you combine unemployment with other risks it does not give you the profile of a murderer, but someone who is at an increased risk of battering.

9. Arresting offenders. Many U.S. police agencies adopted pro-arrest or mandatory arrest domestic violence policies in the 1980s and early 1990s. Propelling these policies were:

  • Legal decisions establishing civil liability against the police for failure to protect women victims of domestic violence[70]
  • The women's movements' advocacy and activism on behalf of domestic violence victims. The womens movement challenged the view of domestic violence as a family problem that could or should be handled privately or differently from stranger assault
  • Widespread dissemination of pro-arrest results of a misdemeanor domestic assault research study.[71]

† More than 300 newspapers reported the results of the study, unprecedented coverage for that time (Fagan, 1996).

Generally, pro-arrest laws and policies apply not only to spouses, but to unmarried partners, former intimates, and persons who had or raised a child together. In many jurisdictions the laws or policies apply to both heterosexual and homosexual relationships.

Police interventions in domestic violence incidents have expanded beyond merely separating and counseling the parties; they've become full-blown criminal investigations in which witnesses are interviewed, neighbors are canvassed, injuries are photographed, physical evidence is collected, future threats are assessed, and victims are referred to follow-up protective services and helped to plan for their future safety. In addition, some states permit police to seize firearms from alleged batterers,[72] and federal laws generally prohibit convicted misdemeanant batterers or those against whom there is a valid order of protection from possessing a gun.[73] All U.S. states now permit police to make warrantless arrests for both misdemeanor and felony assaults.[74]

The highly influential 1980s study of police interventions in domestic violence incidents in Minneapolis found deterrence value in arrest in misdemeanor domestic assaults, as compared to two other interventions—separation of parties or mediation of the dispute at the scene.[75] However, less well-reported replication studies in the late 1980s produced mixed results.[76] A more recent analysis combining five of the replication studies concluded there is only some modest deterrent effect from arrest.[77] However, even this modest effect should be viewed with caution for several reasons:

  • These studies considered only misdemeanor, not felony, domestic assaults.
  • Offenders' employment status appears to be an important variable: unemployed offenders have less at stake (they will not lose their jobs) if they re-offend and thus are less deterred by arrest.[78]
  • A prior arrest record for any crime, and intoxication at the time of the incident, increase the risk that batterers will re-offend after arrest.[79]
  • A small group of offenders, perhaps fewer than 10 percent, appear to continue battering regardless of the intervention, including arrest.[80], †

† Nearly ten percent of the more than 3,000 studied battered repeatedly regardless of the intervention. This group of 250 batterers accounted for 7,380 battering incidents in the six months after the initial intervention. Interventions, even arrest, did not deter this small but violent group. 

A more recent study of victims of both misdemeanor and felony assaults concluded the following:

  • Calling the police was a strong deterrent to repeat battering
  • The effects of arrest were small and statistically insignificant but could not be ruled out completely
  • If arrest does have an effect, it is likely to be modest, particularly in comparison to the effect of calling the police.[81], ‡

‡ The researchers found that batterers who battered again had an increased likelihood of battering a third time, and offenders who were under the influence of alcohol or drugs at the time of the incident were more likely to re-offend. The researchers did not find that marital status, poverty, race, education or gender improved the effect of arrest, but because they did not have access to employment information, they could not rule out that arrest deters employed offenders but not unemployed offenders. 

Victim advocacy groups have generally not been swayed by findings of little or modest effect of arrest. For many advocates, batterer arrest is seen as an important symbol of a woman's legal right to be free of intimate partner violence[82] and, moreover, argue that police continue to arrest other types of offenders without strong evidence of its effectiveness. Arrest is believed to be an important message to children that abuse of their parent is illegal, and perhaps also a deterrent to male children as they become men.§ This belief has not yet been properly studied, however.

§ Victims of domestic abuse have also called for police agencies to monitor, address, and more appropriately sanction officers engaging in domestic assault. Although some police agencies have been responsive, additional efforts are required for victims to have higher levels of confidence in police agencies' domestic violence response policies.

10. Issuing and enforcing restraining orders. Restraining orders (also known as "stay away" or protective orders) are intended to prevent offenders from further harassing, threatening or contacting the victims. Courts have made restraining orders widely available to domestic violence victims, whether or not they file a police report.† Courts may issue a temporary (time-limited) restraining order even when the "party being restrained" is not present or represented. Protective relief may be temporary or permanent.‡ Violation of these orders is now a criminal offense in all U.S. states.[83]

† In most jurisdictions these are obtained from civil court; however, some jurisdictions also grant concurrent jurisdiction to criminal court. Criminal courts can also issue these once a criminal proceeding begins. 

‡ Civil restraining orders were in fact developed to counter the reluctance of police, prosecutors, and criminal courts to treat domestic violence as a serious criminal matter (Buzawa and Buzawa, 2003).

Domestic violence restraining orders are frequently violated although some offenders may be deterred by them. Some research findings suggest that a victim is more likely to seek a protective order if the partner had a criminal history of violent offending, which may be why so many orders are violated; those with robust abuse histories may be the least likely to be deterred by written limits§ so police are advised that more must be done in these cases.

§ For a good review of the research about protective orders, see Buzawa and Buzawa (2003).

11. Aggressively pursuing criminal prosecution of severe domestic violence cases and publicizing convictions. Police pro-arrest and mandatory arrest policies have generated significantly larger caseloads for prosecutors. Similarly, prosecution policies against dropping charges ("no-drop") even when the victim expresses such a desire (the functional equivalent of "mandatory arrest" for police) has further strained prosecutorial resources.[84], ¶ Although such police and prosecution policies can have the beneficial effect of reducing an offender's urge to retaliate against the victim because responsibility for the prosecution is no longer in the victims hands,** it is not yet clear whether such policies have limited further violence or have had the unintended consequence of discouraging some victims from calling police in the first instance.[85]

¶ The rationale for "no-drop" policies is that the state has an independent interest in seeing domestic violence offenders prosecuted because of the harm caused to victims, victims' children, and potential victims.

** Victim recanting remains high. As a result, many prosecutors rely on physical evidence such as photographs and medical reports of victims' injuries and out-ofcourt statements (e.g., 911 call tapes of in-progress assaults) to counter uncooperative and fearful victims. For an excellent review of prosecutorial response to the increased numbers of domestic violence cases and studies of prosecutorial case screening practices, see Buzawa and Buzawa chapter 11 (2003).

"No-drop" policies have some drawbacks. Victim discretion is further reduced, case backlogs increase, and time to disposition is lengthened, which can strain resources devoted to pretrial victim safety.[86] The increase in prosecutorial workload can force prosecutors to trade off the prosecution of other crimes for misdemeanor battering. It is now apparent that prosecutors in "no-drop" jurisdictions rarely prosecute all cases; they retain some level of discretion in case filing decisions, typically at the intake and case screening point.[87] Police must therefore help prosecutors identify the most severe and chronic cases from among the many arrests and encourage prosecutors to prosecute such cases vigorously, make special efforts to protect the victim,† and publicize convictions so as to maximize the general deterrent effect.

† See Problem-Specific Guide No. 42, Witness Intimidation, for further information on measures to protect victim-witnesses. 

12. Establishing special domestic violence courts. There are more than 200 domestic violence courts in the United States, and a growing number in the United Kingdom as well. The proliferation of these courts is part of a wider trend toward specialty courts: drug court, mental health court, drunk driving court, etc.‡ Advocates for specialty courts believe they result in improved outcomes: an increase in specialty knowledge critical to case handling (including the dynamics of the underlying crime/behavior, whether it is battering, drinking, or schizophrenia, depending upon the court), timely attention to the case, and a concentration of appropriate resources that traditional courts do not have that can lead to more effective case handling.

‡ Some of these courts are termed "problem-solving" courts, but the term in this context should not be confused with its meaning in problem-oriented policing. Problem-solving in the court context refers to solving the individual defendant's personal problems that contribute to his offending rather than to addressing an aggregation of incidents. See National Center for State Courts (n.d.).

Typically in specialty courts a single judge works with a community team to develop a case plan for the defendant and uses pending criminal sanctions to compel a defendant's compliance with treatment. The judge monitors compliance and imposes criminal sanctions if the defendant fails to keep to the case plan.

Early evaluations of domestic violence courts generally report on how these courts handle their workload, victim satisfaction, and issues of implementation. It remains unclear if these courts impact recidivism. Researchers who examined these courts in New York describe some of the more important unresolved issues:

Many domestic violence advocates are hesitant to embrace the idea that domestic violence courts are "problem-solving courts." There are substantial differences between domestic violence courts and other problem-solving courts. Many of these differences stem from how success is measured and to whom services are offered. Drug courts can easily look to see whether defendants are successfully completing their court-mandated drug-treatment programs. But domestic violence courts are not targeted at "rehabilitating" defendants. Indeed, services are offered primarily to help victims achieve independence. The primary "service" offered to defendants is batterers programs. But in New York, batterers programs are used by domestic violence courts primarily as a monitoring tool rather than as a therapeutic device. This approach is based on the research about batterers programs, which is extremely mixed. It is unclear whether these programs have any impact at all in deterring further violence.[88]

13. Providing treatment for batterers. Some batterer treatment programs are voluntary; others are court-mandated. In some jurisdictions, prosecutors recommend these programs as part of pre-trial diversion; in others they are part of court-ordered mandatory sentencing. Many states now mandate batterer treatment.[89]

Batterer treatment programs may take a variety of forms. Many offer group treatment with a focus on anger management. Others include individual assessments and individual counseling, and substance abuse and/or mental health treatment.[90]

Unfortunately, few batterer treatment programs have undergone thorough evaluation, and those that have show a mix of positive and negative results.[91] Court-mandated treatment is more likely to result in batterers completing programs,[92] "but there is little evidence to support the effectiveness of one batterer program over another in reducing recidivism."[93]

The quality of the evaluations of batterer treatment programs have improved over time but continue to encounter both methodological and programmatic challenges as illustrated by two recent studies of batterer treatment programs, one in Broward County, Florida and the other in Brooklyn, New York.[94] In both, offender treatment was based on the Duluth treatment model, which is the most commonly used.†

† The Duluth model suggests that batterers seek to control their partners (or ex-partners) and this must change for batterers' behavior to change (Pence and Paymar, 1993). The model "helps offenders to understand how their socialized beliefs about male dominance impede intimacy; that violence is intentional and a choice designed to control their intimate partner; that the effects of abusive behavior damage the family; and that everyone has the ability to change" (Minnesota Program Development, Inc.). 

The Broward evaluation found that treatment attendance did not reduce battering, but that offenders who were married, employed, or homeowners were less likely to batter again (that is, these offenders had a "stake in conformity"). Also, younger men, particularly those with no stable residence, were more likely to rebatter.[95] In Brooklyn, the evaluation showed minor improvement for some of the batterers (that is, some reduction in the number of battering incidents for those attending a 26-week treatment program rather than the same program condensed into an eight-week schedule). In neither case were batterers' attitudes toward domestic abuse changed. Even these evaluation results are not fully reliable because both studies experienced data collection challenges as a result of a high drop-out rate by offenders, difficulty finding relocated victims for follow-up interviews, and inadequate offender attitude assessment tools. In addition, judges sometimes overrode random assignment of batterers, thereby tainting the makeup of the different groups studied.[96] Evaluations of other types of treatment programs, including cognitive-behavior therapy (another widely used approach), have also suffered from similar methodological flaws.

Several experts suggest that greater refinement in assigning batterers to appropriate programs could improve results. The most chronic batterers should receive the most intensive treatment.[97] A "one-size fits-all" approach to batterer intervention cannot accommodate the diverse population of batterers entering the criminal justice system.[98] The different types of batterers—family-only, one who is generally violent even to others, dysphoric (mood-disordered)/borderline—may require tailored treatment.[99]

Experts recommend that treatment programs be designed around explicit theories. In other words, each intervention proposed should have a specific underlying theory. Outcomes expected from each of the interventions should be clearly defined and then evaluated for short- and long-term impact. Designing treatment programs that fit this model requires close collaboration between service providers and researchers. [100] In addition, the timing of treatment may be an important element to success. "Counseling ideally would begin almost immediately after a violent episode, when the offender feels most remorseful, most frightened of the criminal justice system, and most receptive to demands for change."[101]

Responses With Limited Effectiveness

14. Arresting both parties in a domestic violence incident. Arresting both participants in a domestic violence incident (so-called "dual arrest") under the principle of mutual combat is ineffective toward interrupting the pattern of violence between the two. In the context of the longer history of the relationship—as opposed to the one incident—there is nearly always one primary aggressor. When police respond to a domestic violence call, self-defense may look like "mutual combat" and only detailed interviews of the parties (their prior abuse history) and witnesses may reveal the primary aggressor. In addition, at the scene of domestic abuse, "a victim may feel safe to express anger against the batterer in the presence of the police… giving the impression that they are the perpetrators."[102] As a result, more than 20 states have enacted primary aggressor or "predominant aggressor" laws. Even these might not be enough to discourage dual arrest practices unless the law recognizes the importance of ascertaining the pattern/history and not just the aggression within a single incident. To complicate matters, there are couples in which both partners are violent. These are more likely to entail "recurrent acts of minor violence initiated by either party, but the type of violence generally seen by police (and in shelter and clinical samples) is more likely to involve serious and frequent beatings, as well as the terrorizing of women."†

† For a good discussion of "dual arrest" research and the complexities within the primary aggressor issue, see Buzawa and Buzawa (2003).