Police responding to an allegation of domestic violence judge on the scene of the call who is the victim and who is the suspect. Sensational public claims about domestic violence against women and specific domestic violence laws and policies have biased that difficult police judgment toward arresting men. The U.S. criminal justice system operates with de facto gender profiling of men for arrest for domestic violence.

Domestic violence occurs in circumstances relatively favorable to low-level, mutual violence. Most domestic violence occurs between persons having close contact for an extended period. Close, long-term contact facilitates aggression in circumstances of drug or alcohol intoxication, aggression in circumstances of physical vulnerability (during sleep or when attention is diverted), mutual violence, indirect aggression, and retaliatory attacks. Amid the decades long, largely ignored scholarly controversy about the facts of domestic violence, social-scientific data indicates that about 40% of violence between marital partners is mutual violence.^ ^ Nonetheless, criminal justice systems forcefully impose the categories “victim” and “perpetrator.”

Police can arrest all parties involved in an incident of domestic violence. In Connecticut from 1992 to 2011, 35% of persons arrested for domestic violence were arrested in circumstances of arresting all parties involved in the incident. In Rhode Island from 1999 to 2002, the corresponding statistic was 4.3%. Almost all domestic-violence all-party arrests are dual arrests (two persons arrested). Wide dispersal in dual-arrest rates suggests that police have difficulty categorizing victim and perpetrator on the scene of a specific domestic violence incident.

Police data, while biased through reporting categorization, nonetheless indicates the importance of police discretionary judgements in determining who is the domestic violence victim and who is the suspect. Persons arrested in dual arrests for domestic violence assaults amounted to about 8% of all persons arrested for domestic violence assaults by Alaska State Troopers in 2004. Among persons that the Alaska State Troopers judged to be the victims of domestic violence assault, 29% were intoxicated at the time of the assault and 18% had attacked the suspect, according to the police report. The most frequent types of injuries to both victims and suspects were “bruising” and “laceration/bite marks.” Injuries to suspects accounted for at least 14% and 31% of such injuries from incidents of domestic violence assault. Although emergency law is routinely applied to domestic-violence incidents, most domestic violence incidents involve neither weapons nor serious injury. The broad scope of actions legally defined as domestic violence increases the likelihood that all parties involved in a domestic violence incident will have committed an act of domestic violence as defined under domestic violence law. The most important judgment with respect to a domestic violence incident is the judgment least subject to legal review: the police judgment of who to report as victim, and who to report as suspect.

Dual arrests have been highly disfavored in domestic-violence law and policy. According to the publicly controlling understanding of domestic violence, men batter women to uphold patriarchy. Couples don’t abuse each other because they have gotten intoxicated and lost impulse control. Couples don’t mutually abuse each other because they were raised in circumstances where vicious interpersonal behavior was normal. Like a drunken, jealous woman throwing a bottle at her boyfriend, police making dual arrests are thought to be making a mistake. They lack proper education in the cause of domestic violence, as that cause is pervasively asserted in authoritative domestic-violence literature.

Arrests of women for domestic violence have been an explicit concern among domestic-violence scholars and criminal justice agencies. About 25% of persons arrested for domestic violence in the U.S. are female. Whether significant sex differences exist in criminally culpable domestic violence is highly controversial among scholars. Whether significant sex differences exist in competitiveness, sociality, reasoning, and commitment to paid-work career success are also controversial among some persons. Within the amazingly broad legal definition of domestic violence, common sense suggests that large sex differences are unlikely. Nonetheless, domestic-violence literature and policy has pervasively gender-stereotyped domestic violence: women are victims of domestic violence, men are perpetrators. Dual arrest typically involves police reporting a man and a woman as each both suspect and victim of domestic violence. Dual arrests challenge pervasive gender stereotyping of domestic violence. Under actual arrest practices, dual arrests also tend to increase the share of women arrested for domestic violence. Reducing dual arrests serves to reduce the number of women arrested for domestic violence.

Domestic-violence law and policy has reduced dual arrests by increasing police officers’ on-the-scene judicial responsibilities. States have enacted statutes requiring police, on the scene of an incident, to judge the case with respect to statutory definitions such as “primary aggressor,” “primary physical aggressor,” “dominant aggressor,” or “predominate aggressor.” For example, a California statute encouraging arrests for domestic violence differentiates between classes of domestic-violence offenders:

The dominant aggressor is the person determined to be the most significant, rather than the first, aggressor. In identifying the dominant aggressor, an officer shall consider the intent of the law to protect victims of domestic violence from continuing abuse, the threats creating fear of physical injury, the history of domestic violence between the persons involved, and whether either person acted in self-defense. (Cal. Penal Code §13701(b))

“Threats creating fear of physical injury” might be directly related to the specific incident to which the police responded. In contrast, “the history of domestic violence between the persons involved” points to a wide-ranging evaluation of facts not directly apparent in the incident. In a judicial proceeding in a court of law, “whether either person involved acted in self-defense” would be evaluated with respect to a well-developed body of self-defense law. Police officers’ judgments of self-defense aren’t made with professional expertise in that law. Finally, the plain meaning of “the intent of the law to protect victims of domestic violence from continuing abuse” seems not to have any relevance to determining the “dominant aggressor” in a particular incident. In the circumstances of sensationally exaggerated claims about domestic violence against women and pervasive gender-stereotyping of domestic violence, that statutory clause suggests legislative intent to encourage arrest of men for domestic violence.

Legislative intent to encourage arrest of men for domestic violence is implied through stereotypes of men as dangerous to women. For example, the National District Attorneys Association posted online a “DV 101” briefing from the American Prosecutors Research Institute. That briefing explained:

Keep in mind that the laws on arresting the primary or predominant aggressor are for the purpose of safety. Law enforcement officers frequently express frustration at being told to do such an analysis and arrest only one of the parties, when the other party also has some culpability. The domestic violence laws that have been passed in the last 10 to 20 years have come about for one main reason: to make battered women safer. This does not mean that women are never to be arrested, or that we can ignore laws of probable cause, etc. in arresting the batterers. This is simply a reminder that the purpose of the law is safety. Focusing on the safety aspect should help law enforcement officers to understand the analysis they should make in determining primary or predominant aggressor.^ ^

Deeply entrenched gender roles associate men with danger: men have the most dangerous jobs, men have long been expected to sacrifice their lives for women and children, and men suffer 40% more serious injuries from violence than do women. Men also predominate among the direct perpetrators of violent crimes. Domestic violence laws were passed in circumstances of sensational, false claims about domestic violence against women and willful blindness to domestic violence against men. In that deliberative context, focusing on safety means arresting men.

A major domestic-violence policy report indicates that New York’s primary aggressor law was intended to bias domestic-violence arrest decisions toward arresting men. This report was issued to the Governor and Legislature of New York State in 2001. It includes a prefatory “Dear Colleague” letter from the New York State Director of Criminal Justice and the Executive Director of New York’s Office for the Prevention of Domestic Violence. This letter highlighted an imperative to eliminate dual arrests for domestic violence:

Further monitoring is also necessary to ensure that dual arrests do not occur and that officers are making primary physical aggressor determinations.^

The report itself observed:

Recognizing that the implementation of mandatory arrest was causing an increase in situations where both parties were being arrested, termed Dual Arrests, the mandatory arrest statute was amended to add a a “Primary Physical Aggressor” (PPA) provision (N. Y. Crim. Proc. Law § 140.10 (4)(c) (McKinney Supp. 2001). New York State lawmakers created the PPA provision based on the assumption that any increase in the number of Dual Arrests was not reflective of increased rates of female to male violence, but of procedural problems in implementing the new {mandatory arrest} law. Thus, in 1997, the N.Y. Criminal Procedure Law § 140.10 (c) was expanded to provide that when there is reasonable cause to believe that more than one family or household member has committed a misdemeanor crime {specifically a family offense}, the officer is not required to arrest both parties. Police officers are directed to identify the PPA {“Primary Physical Aggressor”} by evaluating the following four factors:

• The comparative extent of any injuries inflicted by and between the parties;
• Whether either party is threatening or has threatened future harm against another household member;
• Whether either party has a prior history of domestic violence that can be reasonably ascertained;
• Whether either party acted defensively to protect himself or herself from injury.^

In New York State about 1996-1997, before primary physical aggressor legislation went into effect, police-reported suspects of domestic violence were 25% female.^ While authorities avoided reporting shares of females arrested for domestic violence, the female arrest share was lower than 25% because the arrest rate for female domestic-violence suspects was lower than that for male suspects.^ Persons arrested in dual arrests for domestic violence amounted to about 9% of total persons arrested for domestic violence.^ Empirical study of domestic violence, the range in dual arrest statistics across other jurisdictions, and other police statistics on female domestic violence offenders don’t clearly indicate that the New York dual-arrest and female arrest shares were unreasonably high. They were too high relative to gender stereotyping of domestic violence. The intent of New York’s primary aggressor law was to reduce the number of females arrested for domestic violence by reducing the number of dual arrests. Justice system researchers in New York report that, in situations of mutual domestic violence, police now usually just arrest the man.^

Profiling men for domestic violence arrests has been quite direct. Guides for New York police went beyond the criteria in the New York primary physical aggressor statute and declared the criminal significance of “difference in the size or strength of the parties.”^ ^ One California police department’s domestic violence protocol includes these questions for determining the “dominant aggressor”:

• Was one party in actual fear of the other?
• Was one party physically larger and stronger than the other?
• Who has access to and control of resources?
• Which party will be in greater danger if nothing is done?^

In determining the “predominate aggressor,” Montana requires, by statute, police to consider “the relative sizes and apparent strength of each person” as well as “the apparent fear or lack of fear between the partners or family members” Mont. Code §46-6-311(b). Such factors set out in domestic-violence law and policy effectively profile men for arrest in cases of mutual domestic violence. Men typically are bigger, stronger, and less fearful than women are.^ ^ ^ Men historically have provided resources to women to whom they are intimately related. Danger to men historically has attracted less public concern than danger to women. These gender-group statistical differences have no relationship to criminal culpability of any sort. In Iowa, some evidence suggests that specially funded prosecutors are allowed to prosecute only male suspects.^ Particularly within the historical context of wildly exaggerated, widely distributed claims about domestic violence against women, scarcely any publicly expressed concern for domestic violence against men, and widespread gender stereotyping of domestic violence, gender-profiling men for arrest for domestic violence is likely to have real, invidious effects.

Gender-profiling men in domestic-violence law and policy draws upon gender biases found even in judgments of professional psychologists concerning psychological abuse. Matched, social-scientific surveys administered to professional psychologists compared identical actions from husband-to-wife and from wife-to-husband. Analysis of the surveys found:

psychologists, irrespective of demographics, rated the husband’s behavior as more likely to be psychologically abusive and more severe in nature than the wife’s use of the same actions.^

Gender bias was apparent even in the willingness to consider alternate scenarios of psychological abuse:

The response rate {for the survey of wife-to-husband psychological abuse} of 26% was substantially lower than that obtained in the previous study of husband-to-wife behaviors (45%) using identical methodological procedures. The lower response rate was in itself informative, as were several incomplete surveys returned with comments. These few participants were reluctant to consider whether behaviors enacted by a woman were abusive because they perceived that women lacked the requisite power to effectively enact psychologically abusive behaviors and they could not comprehend that females could coerce male partners into submission using only psychological tactics.^

Gender stereotypes of men as bigger, physically stronger, and less fearful than woman apparently are more powerful than the gender stereotype that women are more guileful than men. The deeper stereotype seems to be that men are more criminal than women. Lack of public concern about the highly disproportionate imprisonment of men is consistent with criminal gender stereotyping.

Empirical studies indicate that gender stereotypes influence discretionary police judgments about who is the suspect and who is the victim in domestic violence incidents. In Edmonton, Canada, in 2001, the share of women among those charged with intimate-partner violence was 13.0%, 20.4%, 27.9%, and 39.1% within no-injury, low-injury, medium-injury, and high-injury intimate partner violence incidence groups.^ Incidents with a lower degree of physical injury support greater charging discretion and better opportunity for gender stereotyping. Study of the underlying data, and additional independent data, found gender discrimination against men throughout the process of charging, arresting, prosecuting, and judging intimate partner violence.^

Other studies have found similar results in research contexts showing little concern for gender bias against men. Under the heading “Has mandatory arrest {for domestic violence} disproportionately increased arrests of particular sub-populations such as women and minorities?” a 2001 New York study found that, at all sites where the relevant data were available and controlling for relevant incident differences, the odds of a man being arrested was 2.0 to 3.2 times higher than the odds of a woman being arrested. The report went on to rationalize and obscure that gender disparity.^ A study of intimate partner violence in 1994 in Boulder County, Colorado, found that police-identified male victims “were more than three times as likely to be part of a dual-arrest couple than were those individuals identified as female victims.” This study, which termed “battering calls” all calls to police concerning intimate partner violence, apparently sought to “improve the likelihood of strong police responses to intimate partner battering.” This study also showed little concern for extralegal, anti-men sex bias in that police response.^ A study of the police response to domestic violence in three Massachusetts towns in the late 1990s found:

even when a man was a victim, the woman was 5 times less likely to be arrested than a man in a similar situation. In addition, current organizational and political pressures may discourage officers from dual arrests when the incident could in fact be defined as mutually violent. For example, if a woman initiated violence by throwing an object at her partner, resulting in a bruise or cut, and the man retaliated violently, causing similar bruising, officers were found to make no arrest or simply to arrest the man.^

That observation of gender bias was buried in a section entitled, “Is Dual Arrest a Likely Outcome of the Push for Arrest?” Another study produced a logistics regression model showing that a female domestic-violence suspect had 40% lower odds of arrest relative to a male domestic-violence suspect. This study concluded:

there may be valid reasons for not treating male and female offenders alike. … one can question the appropriateness of using the same factors in decisions to arrest for men and women. Acknowledging that women’s experiences and realities can be different may make a more rational and just system.^

Given the demographics of U.S. mass incarceration and pervasive domestic-violence gender stereotyping, that reasoning is as enlightening as arguments for racial segregation in the Old South.

Evidence of anti-men gender bias in domestic violence arrest is pervasive and largely ignored. Many studies and much data indicate anti-men gender bias in arrests for domestic violence.^ ^ Yet that central issue of justice is largely ignored. A study of arrest and prosecution for domestic violence offenses excluded women perpetrators of intimate partner violence against men on the basis that “reforms, at both state and federal levels, were unambigiously directed toward female victims.”^ Studies soberly ponder whether they can reclassify females arrested for domestic violence as victims of domestic violence, but don’t provide an equal scholarly re0judgment service for men arrested for domestic violence.^ An anti-men, gender-biased study of dual arrests obtusely observed in its conclusion: “mandatory arrest policies may have negative, unanticipated consequences for the women victims they were intended to protect.”^ Another tendentious analysis lamented:

policies intended to directly protect women from partner violence by encouraging police to arrest domestic batterers—which is generally construed by policy makers and victim advocates to mean proarrest of male batterers— have resulted in increased arrests of women for domestic violence and other assaults.^

The criminal justice system’s response to domestic violence has been fundamentally and unjustly oriented toward arresting men and protecting women. In practice, implementing that gender bias perfectly has been difficult. Most domestic-violence scholarly literature on arrest for domestic violence ponders how to implement more effectively anti-men gender bias in arrests.

About the year 2001, the Domestic Violence Action Team in Lethbridge, Canada, produced a 106-page manual entitled Our Community Response to Domestic Violence. A scholar without institutional affiliation made a thorough analysis of domestic violence incident reports and police statistics from nearby Edmonton. He also reviewed Our Community Response to Domestic Violence. After several pages of analysis, he gave up:

It is beyond the scope of this report to critique the Domestic Violence Action Team’s manual exhaustively. Suffice it to say that it continues in the same naive, ideological, self-defeating, and blatantly sexist manner for another 90 pages. One should not lose sight of the fact that this document was produced with the support of two separate divisions of Alberta Justice, the City of Lethbridge, Lethbridge Police Services, Lethbridge Family Services, and the Chinook Health Region. These governmental agencies should be embarrassed and ashamed.^

A large share of domestic-violence scholarly and policy reports are of similar quality. Anti-men gender bias is so deliberately and so deeply rooted in public discussion of domestic violence that interested parties are scarcely able to show concern for it without risking great harm to their professional careers. Profiling men for domestic violence arrest has been a central feature of domestic violence literature, law, and policy for decades. Public discussions of arrest for domestic violence typically express concern that profiling men for arrest for domestic violence hasn’t been sufficiently successful.