Chapter 8
Domestic Violence, Law Enforcement, and Court Responses to Domestic Violence

POLICE RESPONSE TO DOMESTIC VIOLENCE

The police are often an abuse victim's initial contact with the judicial system, making the police response particularly important. The manner in which the police handle a domestic violence complaint will likely color the way the victim views the entire judicial system. Not surprisingly, when police project the blame for intimate partner violence on victims, the victims may be reluctant to report further abuse.

According to the 2004 National Crime Victimization Survey, the police came to the aid of victims for about three-quarters (75.3%) of all violent crimes. In about half the incidents of rape and sexual assault, police came to victims (52.8%); 25.6% of victims went to police. Table 9.1 shows that police came to the aid of more victims of aggravated assault (76.6%), simple assault (73.8%), and robbery (85.9%) than victims of rape and sexual assault.

Family disturbance calls constitute most calls received by police departments throughout the country. Historically such calls were not taken seriously, a reflection of society's attitude about domestic violence at the time. For instance, in the mid-1960s Detroit police dispatchers were instructed to screen out family disturbance calls unless they suspected "excessive" violence. A 1975 police guide, The Function of the Police in Crisis Intervention and Conflict Management, taught officers to avoid arrest at all costs and to discourage the victim from pressing charges by emphasizing the consequences of testifying in court, the potential loss of income, and other detrimental aspects of prosecution.

Changes in society's tolerance for domestic violence mean that these approaches to domestic violence no longer enjoy official support (although they may still influence the actions of individual officers). By 2006 every state had moved to authorize probable cause arrests (arrest before the completion of the investigation of the alleged violation or crime) without a warrant in domestic violence cases. (A warrant is a written legal document authorizing a police officer to make a search, seizure, or arrest.) Many police departments have adopted pro-arrest or mandatory arrest policies. Pro-arrest strategies include a range of sanctions from issuing a warning, to mandated treatment, to prison time.

REPORTING DOMESTIC VIOLENCE TO THE POLICE

It is well established that a significant amount of intimate partner violence is unreported or underreported. In the past most women did not report incidents of abuse to the police. Several studies estimate that only about 10% of battering incidents are ever reported to authorities. The 1985 National Family Violence Resurvey finds that only 6.7% of all husband-to-wife assaults were reported to police. When the assaults are categorized by severity as measured on a Conflict Tactics Scale, only 3.2% of minor violence cases and 14.4% of severe violence cases were reported.

There is some evidence that over time, slightly higher percentages of female victims have reported instances of intimate partner violence to the police. In "The 'Drunken Bum' Theory of Wife Beating" (Social Problems, 1987), Glenda Kaufman Kantor and Murray A. Straus estimate that between 7% and 14% of intimate partner assaults were reported to police. A 1995 study using National Crime Victimization Survey data made the most optimistic projections, estimating that 56% of battering incidents were reported to the police.

The women who chose not to report their abuse cited a variety of reasons, including fear of retaliation, loss of income, or loss of their children to child protection authorities. When abuse is reported to law enforcement agencies, it is often by health care professionals from whom the woman has sought treatment for her injuries. In many states health professionals are mandated by law to report all instances of domestic violence to law enforcement authorities.

The National Crime Victimization Surveys (NCVS) are ongoing, nationwide surveys that gather data on criminal victimizations from a national sample of eighty thousand household respondents over age twelve. The surveys provide a biannual estimate of crimes experienced by the public, whether or not a law enforcement agency was contacted about the crime. In National Crime Victimization Survey: Criminal Victimization, 2004 (September 2005, http://www.ojp.usdoj.gov/bjs/pub/pdf/cv04.pdf), Shannan M. Catalano notes that 55.2% of female victims of violent crime in 2004 reported the crime to police. Only about one-third (35.8%) of women who were victims of rape or sexual assault reported the assaults to the police. African-American women (66.9%) and Hispanic women (65.1%) were more likely to report violent crimes to the police than were white women (52.1%).

Table 9.2 shows that women who were victims of violent crime at the hands of strangers were more likely to report the crime than were women who were victimized by nonstrangers (58.6% and 53%, respectively). However, recent data have not been broken down further to examine reporting behaviors of women assaulted by intimate partners. Using seven years of data from the NCVS, Callie Marie Rennison examines trends, including police notification, in Intimate Partner Violence and Age of Victim, 1993–99 (October 2001, http://www.ojp.usdoj.gov/bjs/pub/pdf/ipva99.pdf). She finds that nearly three-quarters of intimate partner violence against females ages twelve to fifteen is not reported to police and that more than half of females ages sixteen to nineteen and over age fifty do not make police reports.

The reasons people give for their decision to report victimizations to the police are shown in Table 9.3. For all personal crimes, high proportions of victims said they reported the incident to stop this incident (18.5%) or to prevent further crimes by the offender against them (17.7%). One out of five (20.4%) reported the incident simply because it was a crime. Other reasons included to stop the offender from victimizing anyone else (10.9%), to punish the offender (7.2%), to catch the offender (4.9%), and a feeling of duty to notify the police (4.8%).

Other victims chose not to report personal crimes to the police. Nearly one out of five people who chose not to report a personal crime said they did not do so because it was a private or personal matter (19.2%). Others were afraid of reprisal (4.6%), an issue particularly salient for battered women. (See Table 9.4.) When the crime was perpetrated by someone the victim knew rather than by a stranger, victims were even more likely to say they did not report because it was a private or personal matter (21.8% versus 16.3%) or because they feared reprisal (6.2% versus 2.9%). (See Table 9.5.)

In The Reporting of Domestic Violence and Sexual Assault by Nonstrangers to the Police (March 2005, http://www.ncjrs.gov/pdffiles1/nij/grants/209039.pdf), Richard B. Felson and Paul-Philippe Paré use data from the NCVS to examine the effects of the gender of the victim and the relationship between victim and offender on the victim's decision to report or not report a physical or sexual assault to the police. They find that victims are just as likely to report assaults by intimate partners as they are to report assaults by other people they know. Sexual assaults are less likely to be reported than are physical assaults.

Marsha E. Wolf et al., in "Barriers to Seeking Help for Intimate Partner Violence" (Journal of Family Violence, April 2003), interviewed forty-one battered women to find out what kept them from calling police. The factors cited included the idea that they must have physical proof that battering had occurred, the desire to avoid a humiliating physical examination in the case of rape or sexual abuse, cultural attitudes about domestic violence, poor self-esteem, being physically prevented from calling the police by the batterer, poor police response when battering was previously reported, and fears of possible retaliation by the batterer or removal of children from the home by child protective services. These women also came up with a "wish list" for how they wanted police to treat them when they called about domestic violence. This list, in the words of Wolf et al., "reflects the women's desires to have responsive police who treat victims with dignity, listen to them, and send appropriate messages to victims and batterers." (See Table 9.6.)

OUTCOME OF POLICE INTERVENTION

In the early 1970s it was legal for the police to make probable cause arrests without a warrant for felonies, but only fourteen states permitted it for misdemeanors. Because the crime of simple assault and battery is a misdemeanor in most states, family violence victims were forced to initiate their own criminal charges against a batterer. By 2006, however, all states authorized warrantless probable cause misdemeanor arrests in domestic violence cases. However, more than half of the states have added qualifiers, such as visible signs of injury or report of the violence within eight hours of the incident. Most state codes authorizing warrantless arrests require police to inform victims of their rights, which include the acquisition of protection orders and referral to emergency and shelter facilities and transportation.

In "Determining Police Response to Domestic Violence Victims" (American Behavioral Scientist, May 1993), a landmark study of four precincts of the Detroit Police Department and their responses to domestic violence in 1993, Eve S. Buzawa and Thomas Austin document several factors that affected police decisions to arrest offenders:

• The presence of bystanders or children during the abuse
• The presence of guns and sharp objects as weapons
• An injury resulting from the assault
• The offender and victim sharing the same residence whether they were married or not
• The victim's desire to have the offender arrested (of victims who expressed such a desire, arrests were made in 44% of the cases; when the victim did not want the offender arrested, arrests were made in only 21% of the cases)

TABLE 9.6

Battered women's "wish list" for police response

• Have quick police response
• Provide consistency in response and take time to listen to women—take victim statement
• Have more female officers
• Avoid questioning parents in front of children
• Get translators for non-English-speaking victims (do not use the batterer or children as translators)
• Send strong message to batterer that battering is wrong, he will be watched, caught, and prosecuted
• Tell women that battering is wrong, it can escalate, and it will not stop without help; take time to inform them of rights and resources
• Arrest appropriate person
• Arrest on felony charges when possible
• Enforce protection orders
• If needed, have advocate at scene who can help victim after police leave
• Provide follow up with victim

source: Marsha E. Wolf, Uyen Ly, Margaret A. Hobart, and Mary A. Kernic, "Table II. Battered Women's 'Wish List' for Police Response," in "Barriers to Seeking Police Help for Intimate Partner Violence," Journal of Family Violence, vol. 18, no. 2, April 2003. Reprinted with kind permission of Springer Science and Business Media.

Mandatory Arrests

Arrest gained popularity as a tactic after the publication of the first in a series of six studies funded by the National Institute of Justice known as the Spouse Assault Replication Program. All the studies were designed to explain how arrest in domestic violence cases could serve as a deterrent to future violence. Lawrence W. Sherman and Richard A. Berk, the authors of the influential first study in the series, "The Specific Deterrent Effects of Arrest for Domestic Assault" (American Sociological Review, April 1984), find that "the arrest intervention certainly did not make things worse and may well have made things better."

Although Sherman and Berk caution about generalizing from the results of a small study dealing with a single police department in which few police officers properly followed the test procedure, they conclude that in instances of domestic violence, an arrest is advisable except in cases where it would be clearly counterproductive. At the same time, Sherman and Berk recommend allowing police a certain amount of flexibility when making decisions about individual situations, on the premise that police officers must be permitted to rely on professional judgment based on experience.

Sherman and Berk's study had a tremendous impact on police practices, although five other Spouse Assault Replication Program studies found arrest had little or no effect on domestic violence recurrence (J. David Hirschel, Ira W. Hutchison, and Charles W. Dean, "The Failure to Deter Spouse Abuse," Journal of Research in Crime and Delinquency, 1992). A survey conducted two years after the Sherman and Berk report found more than a fourfold increase in the number of police departments reporting arrest as their preferred policy in domestic violence disputes. Some police departments have adopted a presumptive arrest policy. This policy means that an arrest should be made unless clear and compelling reasons exist not to arrest. Presumptive arrest provisions forbid officers from basing the decision to arrest on the victim's preference or on a perception of the victim's willingness to testify or participate in the proceedings. Proponents point out that arresting an offender gives the victim a respite from fear and an opportunity to look for help. Furthermore, they claim it prevents bias in arrests.

By the 1990s as many as one out of three police precincts had adopted a mandatory arrest policy in domestic abuse cases. In "The Influence of Mandatory Arrest Policies, Police Organizational Characteristics, and Situational Variables on the Probability of Arrest in Domestic Violence Cases" (Crime and Delinquency, October 2005), David Eitle finds that mandatory police arrest policies do result in more arrests when called to the scene of a domestic assault. In addition, he notes that these policies reduce somewhat the overrepresentation of African-Americans among those arrested. However, do these arrests reduce violence against women?

Some studies find that mandatory arrest policies have positive effects. For example, Jacquelyn C. Campbell et al. find in "Risk Factors for Femicide in Abusive Relationships: Results from a Multisite Case Control Study" (American Journal of Public Health, July 2003) that previous arrest for battering actually decreased women's risk of being subsequently killed by the batterer. Christopher D. Maxwell, Joel H. Garner, and Jeffrey A. Fagan argue in "The Effects of Arrest on Intimate Partner Violence: New Evidence from the Spouse Assault Replication Program" (July 2001, http://www.ncjrs.gov/pdffiles1/nij/188199.pdf) that arrest of batterers is consistently related to subsequent reduced aggression against their intimate partners. Figure 9.1 shows that during the follow-up period after batterers had either been arrested for domestic violence or had an alternative intervention (including mediation counseling, a citation to appear in court, a restraining order, or a warning), those who were arrested consistently "survived" (or did not batter) at a higher rate than those who were not arrested.

However, some battered women's advocates do not support mandatory arrest. They fear that poor and minority families are treated more harshly than middle-class families and that if the police arrive and both spouses are bloodied by the fight, both will be arrested, forcing the children into foster care.

Some studies, such as Ellen Pence and Martha McMahon's A Coordinated Community Response to Domestic Violence (January 1997, http://data.ipharos.com/praxis/documents/ccrdv.pdf), show that mandatory arrests may actually increase violence, especially if the batterer is unemployed or has a criminal record. Some observers suggest that mandatory arrest should be replaced with mandatory action, such as providing transportation to a shelter or granting the victim the option to have the offender arrested. Mandatory action would allow police officers to make decisions appropriate to each individual case.

Evan Stark, in "Mandatory Arrest of Batterers: A Reply to Its Critics" (Eve S. Buzawa and Carl G. Buzawa, eds., Do Arrests and Restraining Orders Work? 1996), offers reasons for a mandatory arrest policy other than deterrence. Stark asserts that mandatory arrest policies provide:

• A standard against which to judge variation in police responses
• Immediate protection from current violence and time for victims to consider their options
• Access to services and protection for victims that would not be available outside the criminal justice system

However, David Hirschel and Ira W. Hutchison, in "The Voices of Domestic Violence Victims: Predictors of Victim Preference for Arrest and the Relationship between Preference for Arrest and Revictimization" (Crime and Delinquency, April 2003), find support for a police policy of taking victim preferences into account in the decision to arrest. Victims based their preferences for arrest on the seriousness of the violence and the perpetrator's prior abusive behavior; in fact, victims proved to be good judges of the seriousness of the violence and the likelihood of it recurring. Hirschel and Hutchison find that victims who wanted their abusers arrested were more likely to suffer subsequent abuse than were victims who did not want their batterers arrested. "Based on these data," Hirschel and Hutchison state, "victim desire for arrest of the offender would appear to be a factor that police should take into account in determining subsequent action."

DO MANDATORY ARREST POLICIES DISEMPOWER THE VICTIM?

In "Battered Women Add Their Voices to the Debate about the Merits of Mandatory Arrest" (Women's Studies Quarterly, 2004), Paula C. Barata and Frank Schneider look at the debate about mandatory arrest policies from a different angle: from the point of view of the victim. They find that a large proportion of battered women actually support a mandatory arrest policy, although they "were more likely to see the benefits of mandatory arrest for other victims/survivors than for themselves." Barata and Schneider hypothesize that the victims' support for mandatory arrest policies in their own situations were influenced by such factors as love for their battering partner, their fears about retaliatory abuse, or their worries about money.

Battered women whose partners had been arrested generally described a pattern of initial decreased violence when the batterer was removed from the home, but an increase in violence when the abuser, who was potentially angry about being arrested, returned home. Only about a third of the women interviewed (38.5%) believed the threat of another arrest would decrease the violence in the long run.

However, despite their disbelief in deterrence, a majority of women still thought that domestic violence was a crime, not a family problem, and several participants liked mandatory arrest policies because they proved to the abuser that assault is wrong and will be punished by the legal system. In fact, most of the battered women interviewed by Barata and Schneider did not view mandatory arrest policies as disempowering. Barata and Schneider argue that this is because of two reasons: first, victims do not want the responsibility for deciding if their battering partner will be arrested and were relieved when the decision was taken out of their hands, and second, victims of battering do not believe they have much influence over the decision to arrest, even without a mandatory arrest policy in place. Also, the vast majority of women believed that mandatory arrest policies made police take their abuse more seriously.

JoAnn Miller, in "An Arresting Experiment: Domestic Violence Victim Experiences and Perceptions" (Journal of Interpersonal Violence, July 2003), also studied victims' perceptions of empowerment or disempowerment as a result of arrest. She focused on the Spouse Assault Replication Program victims' perceptions of police interventions and arrest. She examined two concepts of power: personal power (control of economic and social resources) and legal power (perceived empowerment in response to police intervention).

Miller finds that women did not use their sense of personal power (derived from an independent income) to end domestic violence. However, she notes that victims' perception of legal power (derived from their satisfaction with the police action taken) could be used to feel safer and to control interactions with violent partners in the future. She concludes that "the most reasonable criminal justice and social service responses to domestic violence are those that consider the victim's needs by taking into account her subjective experiences, her cultural and social resources, and her personal and legal resources." In other words, in Miller's view, mandatory arrest policies tend to undermine victims' personal power because they do not take into account victims' needs.

Victims' Attitudes toward Police Response

Interviewing 110 victims as part of their study of the Detroit Police Department's response to domestic violence mentioned previously, Buzawa and Austin find that 85% of victims were satisfied with the police response. Not surprisingly, the victims were particularly satisfied when the police responded to their preferences for arresting or not arresting the offenders.

In "Perceptions of the Police by Female Victims of Domestic Partner Violence" (Violence against Women, November 2003), Robert Apsler, Michele R. Cummins, and Steven Carl investigate "what female victims of domestic violence wanted from the police, the extent to which they perceived they obtained what they wanted, and how helpful they found the actions of the police." They find that women in the study were satisfied with the police response to their call. The women believed the police had been helpful, and more than 80% of them said they would definitely call the police for help in the future. Apsler, Cummins, and Carl emphasize that the particular police department involved in the study had recently instituted policies specifically designed to help battered women.

NOT ALL VICTIMS WHO SEEK POLICE ATTENTION ARE THE SAME

Apsler, Cummins, and Carl speculate that the female victims who seek police aid are likely the most frightened women and that they are seeking assistance to prevent future instances of abuse. In "Fear and Expectations: Differences among Female Victims of Domestic Violence Who Come to the Attention of Police" (Violence and Victims, August 2002), Apsler, Cummins, and Carl review police officers' interviews of ninety-five consecutive victims who came to the attention of a police department in a Boston suburb.

The study participants had either contacted the police department via telephone or by personal appearance at the police station to request intervention in a violent intimate partner dispute. Police officers administered a standardized questionnaire that included questions about the severity of the abuse victims had experienced, their level of fear, and their expectations about the future. About half the victims were living with their abusers at the time the incident occurred.

Apsler, Cummins, and Carl find that just one-quarter of respondents said they were afraid of their abusers. Another 6% were fairly afraid, 12% said they were slightly afraid, and 36% claimed they were not at all afraid of their abusers. Taken together, these latter two groups accounted for nearly half of participants reporting little or no fear of their abusers.

The results were similar in terms of participants' expectations of future abuse. Apsler, Cummins, and Carl find that just 21% of victims thought future abuse was likely, and well over half of women surveyed said future abuse was not at all likely or only slightly likely. These findings challenge long-standing beliefs that the victims who tend to come to the attention of the police are those who most fear future abuse.

Interestingly, there were no statistically significant relationships found between victims' expectations of future violence and whether they lived with their abusers, had children under eighteen years old, or were able to support themselves financially. A less surprising result was that victims' expectations of future abuse strongly influenced their desired future relationships with their offenders. A strong majority of the participants (90%) who thought future abuse was fairly likely or likely wanted to permanently separate from the offenders. In contrast, only about half of the women who thought further abuse was not at all likely wanted permanent separations.

Apsler, Cummins, and Carl conclude that the differences between victims of domestic violence and their varied expectations when seeking police attention point to a need for law enforcement agencies to offer a variety of police responses tailored to victims' needs. For example, they suggest that mandatory arrest of victims' aggressors might not help as a universally applicable strategy for all victims, especially women who do not fear further abuse. By contrast, fearful victims might be reassured and experience greater security if police maintained regular, ongoing contact with them following the incident. Apsler, Cummins, and Carl add that police follow-up might also send a powerful message to perpetrators—that they are under surveillance and that future violations will not be tolerated.

PROTECTION ORDERS

What Are They?

An abuse victim in any state may go to court to obtain a protection order. Also referred to as "restraining orders" or "injunctions," civil orders of protection are legally binding court orders that prohibit an individual who has committed an act of domestic violence from further abusing the victim. Although the terms are often used interchangeably, restraining orders usually refer to short-term or temporary sanctions, whereas protection orders have longer duration and may be permanent. These orders generally prohibit harassment, contact, communication, and physical proximity to the victim. Although protection orders are common and readily obtained, they are not always effective.

All states and the District of Columbia have laws that allow an abused adult to petition the court for an order of protection. States also have laws to permit people variously related to the abuser to file for protection orders. Relatives of the victim, children of either partner, couples in dating relationships, same-sex couples, and former spouses are among those who can file for a protection order in a majority of the states, the District of Columbia, and Puerto Rico. In Hawaii and Illinois, those who shelter an abused person can also obtain a protective order against the abuser.

Petitioners may file for protection orders in circumstances other than violent physical abuse, including sexual assault, marital rape, harassment, emotional abuse, and stalking. Protection orders are valid for varying lengths of time depending on the state. In thirty states the orders are in force for six months to a year. In Illinois and Wisconsin the orders last two years, and in California and Hawaii they are in effect for three years. Furthermore, some states have extended the time during which a general or incident-specific protective order is effective. For example, a no-contact order issued against a stalker convicted in California remains in effect for ten years. In Iowa five-year protection orders are issued and additional five-year extensions may be obtained. New Jersey offers permanent protective orders, and a conviction for stalking serves as an application for a permanent restraining order. Judges in Connecticut may issue standing criminal restraining orders that remain in effect until they are altered or revoked by the court.

Protection orders give victims an option other than filing a criminal complaint. Issued quickly, usually within twenty-four hours, they provide safety for the victim by barring or evicting the abuser from the household. Statutes in most states make violating a protection order a matter of criminal contempt, a misdemeanor, or even a felony. However, this judicial protection has little meaning if the police do not maintain records and follow through with arrest should the abuser violate the order.

The "full faith and credit" provision of the Violence against Women Act was passed to establish nationwide enforcement of protection orders in courts throughout the country. States, territories, and tribal lands were ordered to honor protection orders issued in other jurisdictions—although the act did not mandate how these orders were to be enforced. According to Christina DeJong and Amanda Burgess-Proctor in "A Summary of Personal Protection Order Statutes in the United States" (Violence against Women, January 2006), most states have amended their state domestic violence codes or statutes to reflect the new requirement, although the states vary widely on how easy it is for battered women to get their protection orders enforced. Courts and law enforcement agencies in most states have access to electronic registries of protection orders, both to verify the existence of an order and to assess whether violations have occurred.

Effects of Protection Orders

In "Protection Orders and Intimate Partner Violence: An 18-Month Study of 150 Black, Hispanic, and White Women" (American Journal of Public Health, April 2004), Judith McFarlane et al. report on their study of the effects on intimate partner violence of the application for and receipt of a two-year protection order against abusers. McFarlane et al. interviewed 150 women over an eighteen-month period to determine whether protection orders diminished violence. Although almost half (44%) of the women reported at least one violation of the order, McFarlane et al. find significant reductions in physical assaults, stalking, and threats of assault over time among all women who applied for a protection order, even if they had not been granted the order. McFarlane et al. hypothesize that it was not the protection order itself that led to the diminished violence but the contact with the criminal justice system that exposed the battering to public view.

Victoria L. Holt et al., in "Do Protection Orders Affect the Likelihood of Future Partner Violence and Injury?" (American Journal of Preventive Medicine, January 2003), came to a similar conclusion in their prospective study of 448 female victims of intimate partner violence. They measured the number of unwelcome calls or visits, threats, threats with weapons, psychological, sexual, or physical abuse, and abuse-related medical care among women who had obtained a civil protection order and those who had not. Holt et al. find that women who obtained a protection order following an abusive incident had a significantly decreased risk of contact by the abuser, threats by weapons, injury, and abuse-related medical care.

Holt et al. find in "Civil Protection Orders and Risk of Subsequent Police-Reported Violence" (Journal of the American Medical Association, August 7, 2002) that having a permanent protection order in effect was associated with an 80% reduction in police-reported physical violence in the twelve months following an incident of intimate partner violence. They also report that women who had obtained temporary protection orders, in effect for two weeks only, were more likely than victims with no protection orders to be psychologically abused in the six months after the reported incident of intimate partner violence. Holt et al. speculate that temporary protection orders may have restrained abusers from inflicting physical violence, producing a commensurate increase in psychological abuse. They observe that while temporary orders were linked to increased psychological abuse, the orders did not generate the increased physical violence that many victims and service providers fear will follow. Holt et al. conclude that concern about increased physical violence after obtaining temporary protection orders may be unfounded and that permanent protection orders may be more powerful deterrents to prevent violence recurrence than previously believed.

Enforcement of Orders of Protection

Robert J. Kane examines patterns of arrest of batterers who violate restraining orders in "Police Responses to Restraining Orders in Domestic Violence Incidents" (Criminal Justice and Behavior, October 2000). Although all violators in his study were required by Massachusetts state law to be immediately arrested, in reality only between 20% and 40% of violators of restraining orders were taken into custody. Kane finds that restraining orders had no significant effect on arrest rate; instead, police perception of imminent danger to the victim was the strongest predictor of arrest. He also finds that as the number of domestic violence calls from one victim to police increased, the rate of arrest decreased, regardless of whether a restraining order was in place. Kane suggests that further studies should be done into variations in arrest rates that include personal characteristics of police officers and the social contexts of the couples involved in domestic violence incidents.

George Rigakos, in "Constructing the Symbolic Complainant: Police Subculture and the Nonenforcement of Protective Orders from Battered Women" (Violence and Victims, Fall 1995), studies how police officers' attitudes influenced the nonenforcement of protective orders in a suburb of Vancouver, British Columbia. He finds that a traditional masculine culture contributed to negative stereotypes of women as liars, manipulators, and unreliable witnesses.

According to Rigakos, interviews with police officers and justice officials revealed four major themes. First, justice officials and police felt they were doing all they could for women but contended that other criminal justice institutions impeded their work. Second, the police in this study held conservative attitudes about marriage that resulted in their excusing men's abusive behavior. Third, because traditional beliefs influenced police attitudes toward the victims, many male officers seemed determined to make the women's behavior appear "unreliable." Some felt that the women were using restraining orders to manipulate their husbands to give themselves advantages in custody battles and divorces. Fourth, the officers made generalizations that were supported by their beliefs about the women but were not substantiated by official court records. The officers believed that after they had spent tremendous amounts of time and effort to prepare a case, the women frequently did not pursue the charges. These negative perceptions of the women produced "selective memory" that magnified every instance of a battered woman failing to appear to testify against a partner and diminished the number of times when the women followed through with legal action.

The U.S. Department of Justice, in the report Enforcement of Protective Orders (January 2002, http://www.ojp.usdoj.gov/ovc/publications/bulletins/legalseries/bulletin4/ncj189190.pdf), observes that while all states have passed some form of legislation to benefit victims of domestic violence, and thirty-two states have integrated these rights at the constitutional level, the scope and enforcement of these rights varies. The Department of Justice calls for law enforcement agencies, prosecutors, and judges to be completely informed about the existence and specific terms and requirements of orders and to act to enforce them. Furthermore, it asserts that "unequivocal standardized enforcement of court orders is imperative if protective orders are to be taken seriously by the offenders they attempt to restrain."

LANDMARK LEGAL DECISIONS

Before the 1962 landmark case Self v. Self, when the California Supreme Court ruled that "one spouse may maintain an action against the other for battering," women had no legal recourse against abusive partners. The judicial system had tended to view wife abuse as a matter to be resolved within the family. Maintaining that "a man's home is his castle," the federal government traditionally had been reluctant to violate the sanctity of the home. Furthermore, many legal authorities persisted in "blaming the victim," maintaining that the wife was, to some degree, responsible for her own beating by somehow inciting her husband to lose his temper. Yet even after the landmark California case, turning to the judicial system for help was still unlikely to bring assistance to or result in justice for victims of spousal abuse. Jurisdictions throughout the United States continued to ignore the complaints of battered women until the late 1970s.

Many victims of domestic violence have sought legal protection from their abusive partners. This section summarizes the outcomes of several landmark cases that not only helped to define judicial responsibility but also shaped the policies and practices aimed at protecting victimized women.

Baker v. The City of New York

Sandra Baker was estranged from her husband. In 1955 the local domestic relations court issued a protective order directing her husband, who had a history of serious mental illness, "not to strike, molest, threaten, or annoy" his wife. Baker called the police when her husband created a disturbance at the family home. When a police officer arrived, she showed him the court order. The officer told her it was "no good" and "only a piece of paper" and refused to take any action.

Baker went to the domestic relations court and told her story to a probation officer. While making a phone call, she saw her husband in the corridor. She went to the probation officer and told him her husband was in the corridor. She asked if she could wait in his office because she was "afraid to stand in the room with him." The probation officer told her to go to the waiting room. Minutes later, her husband shot and wounded her.

Baker sued the city of New York, claiming that the city owed her more protection than she was given. The New York State Supreme Court Appellate Division, in Baker v. The City of New York (1966), agreed that the city of New York failed to fulfill its obligation. The court found that she was "a person recognized by order of protection as one to whom a special duty was owed … and peace officers had a duty to supply protection to her." Neither the police officer nor the probation officer had fulfilled this duty, and both were found guilty of negligence. Because the officers were representatives of the city of New York, Baker had the right to sue the city.

Equal Protection

Another option desperate women have used in response to unchecked violence and abuse is to sue the police for failing to offer protection, alleging that the police violated their constitutional rights to liberty and equal protection under the law.

The Equal Protection Clause of the Fourteenth Amendment provides that no state shall "deny to any person within its jurisdiction the equal protection of the laws." This clause prohibits states from arbitrarily classifying individuals by group membership. If a woman can prove that a police department has a gender-based policy of refusing to arrest men who abuse their wives, she can claim that the policy is based on gender stereotypes and therefore violates the equal protection laws.

THURMAN V. CITY OF TORRINGTON

Between October 1982 and June 1983 Tracey Thurman repeatedly called the Torrington, Connecticut, police to report that her estranged husband was threatening her life and that of her child. The police ignored her requests for help no matter how often she called or how serious the situation became. At one point her husband attacked her in view of the police, and was arrested. Thurman obtained protection and restraining orders against him after this incident. But when her husband later came to her home and threatened her again, in violation of his probation and the court orders, police refused to intervene.

On June 10, 1983, Thurman's husband came to her home. She called the police. He then stabbed her repeatedly around the chest, neck, and throat. A police officer arrived twenty-five minutes later but did not arrest her husband, despite the attack. Three more police officers arrived. The husband went into the house and brought out their child and threw him down on his bleeding mother. The officers still did not arrest him. While his wife was on the stretcher waiting to be placed in the ambulance, he came at her again. Only at that point did police take him into custody. Thurman later sued the city of Torrington, claiming she was denied equal protection under the law.

In Thurman v. City of Torrington (1984), the U.S. District Court for Downstate Connecticut agreed, stating:

City officials and police officers are under an affirmative duty to preserve law and order, and to protect the personal safety of people in the community. This duty applies equally to women whose personal safety is threatened by individuals with whom they have or have had a domestic relationship as well as to all other people whose personal safety is threatened, including women not involved in domestic relationships. If officials have notice of the possibility of attacks on women in domestic relationships or other people, they are under an affirmative duty to take reasonable measures to protect the personal safety of such people in the community.

… A police officer may not knowingly refrain from interference in such violence, and may not automatically decline to make an arrest simply because the assailant and his victim are married to each other. Such inaction on the part of the officer is a denial of the equal protection of the laws.

There could be no question, the court concluded, that the city of Torrington, through its police department, had "condoned a pattern or practice of affording inadequate protection or no protection at all, to women who complained of having been abused by their husbands or others with whom they have had close relations." Therefore, the police had failed in their duty to protect Tracey Thurman and deserved to be sued.

The federal court jury awarded Thurman $2.3 million in compensatory damages. Almost immediately the state of Connecticut changed the law, calling for the arrest of assaultive spouses. In the twelve months following the new law, arrests for domestic assault almost doubled, from 12,400 to 23,830.

Due Process

The Due Process Clause of the Fourteenth Amendment provides that no state can "deprive any person of life, liberty, or property, without due process of law; nor deny to any person within its jurisdiction the equal protection of the laws." It does not, however, obligate the state to protect the public from harm or provide services that would protect them. Rather, a state may create special conditions in which that state has constitutional obligations to particular citizens because of a "special relationship between the state and the individual." Abused women have used this argument to claim that being under a protection order puts them in a "special relationship."

MACIAS V. HIDE

During the eighteen months before her estranged husband Avelino Macias murdered her at her place of work, Maria Teresa Macias had filed twenty-two police complaints. In the months before her death, Avelino Macias sexually abused his wife, broke into her home, terrorized her, and stalked her. The victim's family filed a wrongful death lawsuit against the Sonoma County Sheriff's Department in California, accusing the department of failing to provide Macias equal protection under the law and of discriminating against her as a Hispanic and a woman.

The U.S. District Court for the Northern District of California dismissed the case because Judge D. Lowell Jensen said there was no connection between Macias's murder and how the sheriff's department had responded to her complaints. On July 20, 2000, the U.S. Court of Appeals for the Ninth Circuit reversed the earlier decision and ruled that the lawsuit could proceed with the discovery phase and pretrial motions. Judge Arthur L. Alarcon of the U.S. Court of Appeals for the Ninth Circuit conveyed the unanimous opinion of the court when he wrote, "It is well established that there is no constitutional right to be protected by the state against being murdered by criminals or madmen. There is a constitutional right, however, to have police services administered in a nondiscriminatory manner—a right that is violated when a state actor denies such protection to disfavored persons."

After this decision the case proceeded to trial. On June 18, 2002, Sonoma County agreed to pay $1 million to the Macias family to settle the case. The settlement agreement did not include an admission of any wrongdoing by the county. Nevertheless, domestic violence activists lauded the result. According to Katherine P. Califa in "Domestic Violence Civil Suit Settles for $1 Million" (July 5, 2002, http://www.now.org/issues/violence/070502macias.html), Kim Gandy, the president of the National Organization for Women (NOW), stated, "This settlement shows that law enforcement cannot get away with denying equal protection under the law to victims of domestic violence."

TOWN OF CASTLE ROCK V. GONZALES

The U.S. Supreme Court reaffirmed in 2005 that governments are not required to offer citizens protection from violent abusers. In the case of Town of Castle Rock v. Gonzales, Jessica Gonzales's three children were taken from her home by her estranged husband in violation of an order of protection. When she alerted the police and repeatedly asked them for assistance, they made no effort to find the children or enforce the state's mandatory arrest law. The children were killed later that night. Gonzales sued the town, arguing that by failing to respond to her calls the police had violated her right to due process of law, and the case eventually reached the U.S. Supreme Court. In its review the Court found on June 27, 2005, that "Colorado law has not created a personal entitlement to enforcement of restraining orders" and that the state's law did not really require arrest in all cases, but actually left considerable discretion in the hands of the police. So while the Castle Rock police made a tragically incorrect decision in this case, they did not do so in violation of Gonzales's due process rights. The American Civil Liberties Union subsequently filed the case with the International Human Rights Organization, arguing that domestic violence victims have the right to be protected by the state from their abusers.

KEY DOMESTIC VIOLENCE LEGISLATION

While appealing to the judicial system for help will not solve all the problems an abused woman faces, the reception a battered woman can expect from the system—police, prosecutors, and courts—improved markedly in the late twentieth century. The Violence against Women Act, signed into law by President Bill Clinton in September 1994, did much to help. The act simultaneously strengthened prevention and prosecution of violent crimes against women and provided law enforcement officials with the tools they needed to prosecute batterers. Although the system is far from perfect, legal authorities are far more likely to view abuse complaints as legitimate and serious than they had in the past.

Violence against Women Act

A key provision in the Violence against Women Act, the civil rights provisions of Title III, declared that violent crimes against women motivated by gender violate victims' federal civil rights—giving victims access to federal courts for redress for the first time. In testifying in favor of the passage of the act, Sally Goldfarb (http://www.binghamton.edu/womhist/vawa/doc14b.htm), an attorney for NOW, told Senate members, "The enactment of civil rights legislation would convey a powerful message: that violence motivated by gender is not merely an individual crime or a personal injury, but is a form of discrimination, an assault on a publicly-shared ideal of equality. When half of our citizens are not safe at home or on the streets because of their sex, our entire society is diminished." When the Violence against Women Act was passed into law, the text of the civil rights provision was:

Federal civil rights action as specified in this section is necessary to guarantee equal protection of the laws and to reduce the substantial adverse effects on interstate commerce caused by crimes of violence motivated by gender; and the victims of crimes of violence motivated by gender have a right to equal protection of the laws, including a system of justice that is unaffected by bias or discrimination and that, at every relevant stage, treats such crimes as seriously as other violent crimes.

In 1999 the civil rights section of the act was tested in the U.S. Supreme Court. Christy Brzonkala, an eighteen-year-old freshman at Virginia Polytechnic Institute, was violently attacked and raped by two men, Antonio Morrison and James Crawford, on September 21, 1994. Brzonkala did not immediately report the rape, and no physical evidence of the rape was preserved. Two months later, she filed a complaint with the school; after learning that the college took limited action against the two men, she withdrew from the school and sued her assailants for damages in federal court.

Brzonkala's case reached the Supreme Court in 1999. Briefs in favor of giving victims of gender-based violence access to federal courts were filed by dozens of groups—the American Medical Women's Association, the National Association of Human Rights Workers, the National Coalition against Domestic Violence, and the National Women's Health Network among them—as well as the briefs filed by law scholars and human rights experts. However, five of the Supreme Court justices decided in Brzonkala v. Virginia Polytechnic Institute (2000) that Congress could not enact a law giving victims of gender-motivated violence access to federal civil rights remedies. The majority opinion emphasized that "the Constitution requires a distinction between what is truly national and what is truly local"—and it ruled that the violent assault of Christy Brzonkala was local.

In October 2000 Congress responded to the Supreme Court decision by passing new legislation, the Victims of Trafficking and Violence Protection Act of 2000. The new statute included these titles: Strengthening Law Enforcement to Reduce Violence against Women, Strengthening Services to Victims of Violence, Limiting the Effects of Violence on Children, and Strengthening Education and Training to Combat Violence against Women. The act allocated $3.3 billion over five years to fund traditional support services along with prevention and education about dating violence, rape, and stalking via the Internet, as well as new programs for transitional housing and expanded protection for immigrant women. The new act did not mention women's civil rights. The Violence against Women Act was reauthorized in 2005, providing funding to programs through 2009.

Other Federal Laws and Public Policies

The following are among other federal laws and public policies aimed at addressing violent crimes against women:

• The Hillory J. Farias and Samantha Reid Date Rape Drug Prohibition Act of 2000 made it illegal to manufacture, distribute, or dispense gamma hydroxybutyric acid (GHB, also known as "liquid ecstasy" and the "date rape drug") and created a special unit to evaluate abuse and trafficking of GHB and other drugs associated with instances of sexual assault.
• The Social Security Administration in 1998 changed its rules governing new social security numbers. Previously, a woman needed to provide evidence that an abuser was using her social security number to locate her or that the violence was life threatening. With the new law women need only to provide written corroboration of domestic violence from a third party, such as a doctor, lawyer, clergy member, or even a family member or friend, to obtain a new number.
• A Postal Service Release of Information Final Rule was published on January 25, 2000, and became effective February 24, 2000. If an individual postal customer presents the U.S. Postal Service with a court order of protection, then the postal service may not disclose identifying information such as address, location, or post office box, unless ordered by the court.
• A Final Rule on Documentation of Immigrants and Nonimmigrants—Visa Classification Symbols was published on June 18, 2001. This rule amended the Violence against Women Act to create new nonimmigrant categories for victims of trafficking for illicit sexual purposes and slavery and those who have suffered abuse, such as battering, and other forms of violence.

In addition, on January 26, 2001, the U.S. Sentencing Commission published the Sentencing Guidelines for United States Courts, which increased the base sentencing levels for offenses and required stricter sentences for stalking, domestic violence, and cases involving the use of GHB.

The International Marriage Broker Regulation Act of 2005 was geared to stop serial abusers of women who come to the United States as "mail-order brides." The new law requires all men seeking visas for a bride-to-be to disclose any convictions for domestic violence, sexual assault, or child abuse. In addition, men seeking a new visa for a second fiancée within two years will be denied, and no petitions for a third fiancée visa will be granted.

DOMESTIC VIOLENCE GUN BAN

Federal law includes batterers convicted of domestic violence crimes or those with domestic violence protection orders filed against them among the people who are prohibited from owning or carrying guns. Since 1994, most people attempting to purchase guns in the United States have had to pass a background check first. According to the report Background Checks for Firearm Transfers, 2004 (October 2005, http://www.ojp.usdoj.gov/bjs/pub/pdf/bcft04.pdf), Michael Bowling et al. note that from March 1, 1994, through December 31, 2003, 61.6 million applications for firearm permits or transfers were subjected to background investigations by the Federal Bureau of Investigation (FBI) and state and local agencies. Of these applications, about 1.2 million were rejected. While the rejection rate is low (Table 9.7 shows that the FBI rejection rate for selected states was only 1.2% in 2004), a substantial proportion of those rejections are because of prior domestic violence convictions. In 2004, about 16% of those rejected were rejected because the FBI or state or local police agencies found that applicants had either been convicted of a domestic violence misdemeanor or had a restraining order issued against them. (See Table 9.8.) After prior felony convictions, domestic violence was the second-leading reason for rejecting applicants' gun permit requests.

Loopholes in state and federal laws allow batterers to purchase guns despite the federal ban. In many states private gun owners can sell their firearms without background checks. In addition, many states keep incomplete records of domestic violence offenders and orders of protection. Still other evidence suggests that some gun dealers knowingly allow people who are not legally eligible to purchase firearms to buy them through a third party.

DIFFICULTIES IN THE COURT SYSTEM

An appeal to the U.S. judicial system should be an effective method of obtaining justice. For battered women, however, this has not always been the case. In the past ignorance, social prejudices, and uneven attention from the criminal justice system all tended to underestimate the severity and importance of battering crimes against women. Although society has become significantly less tolerant of domestic violence, and laws in many states criminalize behavior previously considered acceptable, old attitudes and biases continue to plague intimate partner violence and spouse abuse cases in the courts.

Intimate partner abuse cases are often complicated by evidence problems, because domestic violence usually takes place behind closed doors. The volatile and unpredictable emotions and motivations influencing the behavior of both the abuser and victim may not always fit neatly into the organized and systematic framework of legal case presentation. Finally, the varying training mandates to ensure that prosecutors and judges are better informed about the social and personal costs of domestic violence, along with society's changing attitudes toward abuse, influence the responses of the judicial system.

The victim-offender relationship is an important factor in determining how the offender is treated by the criminal justice system. On the one hand, strangers are treated more harshly because stranger offenses are considered more heinous and the true targets of the justice system. As a result, criminal law is strictly enforced against them. On the other hand, the justice system has traditionally perceived nonstranger offenses as a victim's misuse of the legal system to deal with strained interpersonal relationships.

Several studies confirm that while intimate partners are frequently charged with and convicted of more serious offenses, stranger offenders generally receive longer sentences. In addition, most intimate partner crimes never go to sentencing, because victims drop charges or settle out of court. Those cases that do proceed through the judicial system are likely to be the more serious crimes.

Willingness of Victims to Prosecute

One of the most formidable problems in prosecuting abusers is the victim's reluctance to cooperate. Although many abused women have the courage to initiate legal proceedings against their batterers, some are later reluctant to cooperate with the prosecution because of their emotional attachment to their abusers. Other reasons for their reluctance are fear of retaliation, mistrust or lack of information about the criminal justice system, or fear of the demands of court appearances. These reasons are among the findings by Lisa Goodman, Lauren Bennett, and Mary Ann Dutton in "Obstacles to Victims' Cooperation with the Criminal Prosecution of Their Abusers: The Role of Social Support" (Violence and Victims, Winter 1999) and by JoAnn Miller in "An Arresting Experiment: Domestic Violence Victim Experiences and Perceptions" (Journal of Interpersonal Violence, July 2003). A victim's fear and ambivalence about testifying, and the importance of her behavior as a witness, can undoubtedly discourage some prosecutors from taking action.

However, a victim might choose not to move forward with the prosecution because the violence ceases temporarily following the arrest while the batterer is in custody. In most cases, a woman does not want her husband to go to jail with the attendant loss of income and community standing. She simply wants her husband to stop beating her.

Religious convictions, economic dependency, and family influence to drop the charges place great pressure on victimized women. Consequently, many prosecutors, some of whom believe abuse is a purely personal problem and others who believe winning the case is unlikely, test the victim's resolve to make sure she will not back out. This additional pressure drives many women to drop the charges because after being controlled by their husbands, they feel that the judicial system is repeating the pattern by abusing its power. Hence, the prosecutors' fears contribute to the problem, creating a self-perpetuating cycle.

Goodman, Bennett, and Dutton explore the reasons many domestic violence victims refuse to cooperate in the prosecution of their abusers. Surprisingly, they find that the relationship between emotional support and cooperation with prosecutors was not significant. Similarly, institutional support, whether from police or victim advocates, was also unrelated to cooperation. Neither level of depression nor degree of emotional attachment to the abuser had an effect. These findings refute the common perception that the battered woman is too depressed, helpless, or attached to the abuser to cooperate in his prosecution. Instead, Goodman, Bennett, and Dutton's findings show that many domestic violence victims persevere in the face of depression and the sometimes complex emotional attachment to their partner.

Consistent with findings from earlier studies, Goodman, Bennett, and Dutton find that the more severe the violence, the more likely the abused women were to cooperate with prosecutors. Participants rearing children with the abuser were also more likely to cooperate, perhaps because these women hoped that the criminal justice system would force the abuser into treatment. In contrast, women with substance abuse problems were less than half as likely as other women to cooperate with the prosecution. Goodman, Bennett, and Dutton conclude that the women who used alcohol or drugs believed that the abuse was partly their fault or that a judge would not take them seriously. Some also feared that their substance abuse might negatively affect the court proceedings and possibly even lead to criminal charges or the loss of their children.

Factors Associated with Prosecutors' Charging Decisions in Domestic Violence Cases

In "Modeling Prosecutors' Charging Decisions in Domestic Violence Cases" (Crime and Delinquency, July 2006), John L. Worrall, Jay W. Ross, and Eric S. McCord investigate what factors influenced the decisions of prosecutors to charge a batterer and what factors influenced the decisions of prosecutors to pursue a misdemeanor or a felony charge. They collected data on 245 domestic violence cases filed by police officers, examined the impact of characteristics of the victim, offender, and the offense, and determined how these characteristics influenced the prosecutors' charging decisions.

Worrall, Ross, and McCord find that prosecutors were more likely to charge offenders if they had been arrested, or if they had inflicted serious injuries on the victim. They find that criminal charges were more likely to be filed against a male batterer than against a female batterer. They also find that if the victim supported prosecution, felony charges were more likely to be filed against the batterer.

TREATMENT FOR MALE BATTERERS

Rather than serving a prison term, many convicted batterers enter treatment programs. As a requirement of probation, most courts will order a batterer into an intervention program. Regardless of an intervention program's philosophy or methods, program directors and criminal justice professionals generally monitor the offenders' behavior closely. Most batterers enter intervention programs after having been charged by the police with a specific incident of abuse.

The criminal justice system categorizes offenders based on their potential danger, history of substance abuse, psychological problems, and risk of dropout and rearrest. Ideally, interventions focus on the specific type of batterer and the approach that will most effectively produce results, such as linking a substance abuse treatment program with a batterer intervention program. Other program approaches focus on specific sociocultural characteristics, such as poverty, race, ethnicity, and age. In "Analyzing the Studies" (Batterer Intervention Programs: Where Do We Go from Here? June 2003, http://www.ncjrs.gov/pdffiles1/nij/195079.pdf), Shelly Jackson argues that the effectiveness of batterer intervention programs might improve if the programs were seen "as part of a broader criminal justice and community response to domestic violence that includes arrest, restraining orders, intensive monitoring of batterers, and changes to social norms that may inadvertently tolerate partner violence."

Several states require that the victim be notified at various points of the intervention and that programs with a strong advocacy policy contact victims every two or three months. Victims may be asked for additional information about the relationship, given information about the program's goals and methods, and helped with safety planning. In addition, the batterer's counselor will inform the victim if further abuse appears imminent.

Batterers leave the program either because of successful completion or because they are asked to leave. Reasons for termination include failure to cooperate, nonpayment of fees, revocation of parole or probation, failure to attend group sessions regularly, or violation of program rules. Successful completion of a program means that the offender has attended the required sessions and accomplished the program's objectives. With court-mandated clients a final report is also made to probation officials. To be successful, batterer intervention programs must have the support of the criminal justice system, which includes coordinated efforts between police, prosecutors, judges, victim advocates, and probation officers.

Program Dropout Rates

Dropout rates in battering programs are high, even though courts have ordered most clients to attend. Several studies, such as that by Jennifer Rooney and R. Karl Hanson in "Predicting Attrition from Treatment Programs for Abusive Men" (Journal of Family Violence, June 2001), record varying dropout rates, some finding that as many as 90% of the men who begin short-term treatment programs do not complete them.

High dropout rates in batterer intervention programs make it difficult to evaluate their success. Evaluations based on men who complete these programs focus on a select group of highly motivated men who likely do not reflect the composition of the group when it began. Because a follow-up is not conducted with program dropouts—the men most likely to continue their violence—research generally fails to accurately indicate the success or failure of a given treatment program.

Certain characteristics are generally related to dropout rates. Bruce Dalton, in "Batterer Characteristics and Treatment Completion" (Journal of Interpersonal Violence, December 2001), finds that the level of threat that the batterer perceived from the referral source (for example, the court) was, surprisingly, not related to program completion. Unemployment was the one characteristic most consistently related to dropping out of treatment. Dalton theorizes that these men have both trouble paying for the treatment and a lower investment in the "official social order."

Other researchers find that factors influencing completion rates of batterer intervention programs include youth, not being legally married, low income and little education, unstable work histories, criminal backgrounds, and excessive drinking or drug abuse. Voluntary clients, especially those with college educations, remain in treatment longer. Some researchers find better attendance among college-educated men, regardless of whether their enrollment in a program is court ordered or voluntary (see Rooney and Hanson, above; and Edward W. Gondolf, "A Comparison of Four Batterer Intervention Systems: Do Court Referral, Program Length, and Services Matter?" Journal of Interpersonal Violence, January 1999).

Nearly all professionals involved in domestic violence prevention and treatment programs concur that batterer intervention programs must address the issue of dropouts. Reducing or eliminating intake sessions and immediately engaging batterers in useful interventions may help to promote attendance and participation. Counselors should provide more information about the purpose of the program in the preprogram orientation sessions. Other suggested retention measures include courtroom assistance, mentors, and harsher and quicker punishment for dropouts. One study finds that home visits after a batterer misses a meeting also help decrease dropout rates. Researchers on this subject include Bruce Dalton in "Batterer Characteristics and Treatment Completion" (Journal of Interpersonal Violence, December 2001) and A. DeMaris in "Attrition in Batterers' Counseling: The Role of Social and Demographic Factors" (Social Service Review, 1989).

Recidivism Rates

Recidivism (the tendency to relapse to old ingrained patterns of behavior) is a well-documented problem among people in intimate partner violence treatment programs. In "Predictors of Criminal Recidivism among Male Batterers" (Psychology Crime and Law, December 2004), R. Karl Hanson and Suzanne Wallace-Capretta examine risk factors associated with recidivism of 320 male batterers within a five-year follow-up period. Of those men, 25.6% recidivated with a battering offense. Risk factors included being young, having an unstable lifestyle, being a substance abuser, and having a criminal history. Batterers were not deterred by expectations of social or legal negative consequences. However, maintaining positive relationships with community treatment providers was associated with deterrence of future battering.

Julia C. Babcock and Ramalina Steiner report some cautiously optimistic findings in "The Relationship between Treatment, Incarceration, and Recidivism of Battering: A Program Evaluation of Seattle's Coordinated Community Response to Domestic Violence" (Journal of Family Psychology, March 1999). Their research measured recidivism of domestic violence after arrest and completion or noncompletion of a mandatory, coordinated program of treatment involving the courts, probation officers, and treatment providers.

Babcock and Steiner followed 387 people arrested for misdemeanor domestic violence offenses, thirty-one of whom were women. More than three-quarters of participants had no prior domestic violence convictions and 69% had no prior criminal history. Participants were referred to one of eleven certified domestic violence treatment programs. The majority attended programs that use the Duluth model (based on the feminist idea that patriarchal ideology, or men's desire to oppress women, causes domestic violence), whereas the remainder participated in feminist, psychoeducational, and cognitive-behavioral men's groups. About 31% completed at least twenty-four sessions of treatment, and those batterers considered to have completed treatment attended an average of thirty-two sessions. In contrast, batterers who did not complete treatment attended an average of just 5.8 sessions. Treatment completers were generally first-time offenders, better educated, employed, and had less prior criminal involvement. Of the noncompleters, 58% did not attend any sessions, and the majority were not legally punished, despite their failure to attend court-ordered treatment.

Program completion was related to lower rates of recidivism—treatment completers had significantly fewer domestic violence arrests at follow-up than noncompleters, and this difference remained even when Babcock and Steiner controlled for differences in prior criminal record and history. Batterers who had been court ordered to attend treatment and failed to complete it were more likely to commit further offenses than treatment completers. Babcock and Steiner conclude that their findings support the premise that completing treatment is directly related to reduced rates of domestic violence. They caution, however, that participants who completed treatment were probably not representative of the entire population of batterers—they likely had more to lose as a result of failure to complete treatment than the treatment dropouts.

In "The Effects of Domestic Violence Batterer Treatment on Domestic Violence Recidivism" (Criminal Justice and Behavior, February 2003), Jill A. Gordon and Laura J. Moriarty study the effect of batterer treatment on recidivism. They find that attending treatment had no impact on recidivism when comparing the treatment group as a whole with the experimental group. Christopher I. Eckhardt et al., in "Intervention Programs for Perpetrators of Intimate Partner Violence: Conclusions from a Clinical Research Perspective" (Public Health Reports, July-August 2006), concur in their review of the literature concerning batterer intervention programs and recidivism rates. However, Gordon and Moriarty also find that among the treatment group, the more sessions a batterer completed, the less likely he was to batter again. Batterers who completed all sessions were less likely to be rearrested for domestic violence than were batterers who had not completed all sessions.