Skip to content
Patriarchy

The domestic violence myth

Domestic violence does not have a sex. A theory that assigns the man the role of perpetrator in advance breaks on the data.

The myth is simple: a man hits a woman because patriarchy taught him to establish power and control over her. Once that premise is accepted, sex supposedly tells us who the perpetrator is, who the victim is, and why the violence happened.

That mechanism is real in some cases. A person can build a system of intimidation around money, movement, social contact, threats, and physical force. But one form of coercive violence cannot explain all intimate partner violence. The research literature also contains female violence, bidirectional violence, situational conflict escalation, and multiple risk factors that do not reduce to a male desire to dominate women.

Symmetry in acts does not mean symmetry in harm. Men are physically stronger on average and are more likely to inflict injury. That matters. It still does not reveal the motive behind every violent act.

Our concern here is what disappears from a female-centered account: the man as victim, the woman as aggressor, prior provocation, expectations of male non-retaliation, and the barriers that keep men from seeking help. Female victims do not disappear when we restore these facts. Male victims should not be reduced to statistical noise merely because their experience fits the dominant political story badly.

How power and control became a universal explanation

The best-known version of this framework is the Duluth Model. Ellen Pence, Coral McDonnell, and Michael Paymar created the Power and Control Wheel in 1982 for a court-ordered program for men who batter. The recurring tactics on the wheel came from the experiences of women who had been subjected to sustained abuse.1

That origin matters. The source material was not a random sample of all couples and it was not an experiment establishing the cause of every assault. It was a practical model of battering: an ongoing pattern of violence, intimidation, and coercion. The Duluth Model itself still describes battering as an intentional pattern of control and domination, rather than treating every physical conflict between partners as the same phenomenon.2

The problem begins when that distinction disappears. A useful description of one pattern turns into a universal causal rule: if a man used violence against a woman, the motive is already known — power. Investigation is replaced by a motive assigned in advance.

Pence later acknowledged the problem

In 1999 Ellen Pence wrote directly about the philosophy behind the project. She acknowledged that treating the need or desire for power as the universal motivating force behind battering did not fit the lived experience of many of the men and women with whom the project worked. She also described her own confirmation error: when men did not articulate a desire for power, she tended to interpret that as denial until case experience forced the team to reconsider the framework.3

This was not a wholesale repudiation of the Duluth Model. Pence continued to connect violence to socially constructed entitlement and expectations. Claims that she simply “admitted the entire model was false” go too far.

The narrower point is stronger and better documented: one of the model’s architects acknowledged that a desire for power could not simply be assigned as the motive in every case.

Female aggression does not disappear from the data

John Archer’s 2000 meta-analysis found that women in the heterosexual samples studied were slightly more likely than men to report at least one act of physical aggression and to report such acts somewhat more frequently. Men, however, were more likely to inflict injury; women accounted for 62% of injured partners.4

That does not establish that “women are more violent than men.” Archer explicitly noted that the available database was heavily weighted toward young U.S. dating samples and that sample composition moderated the result.

It does establish something less dramatic and more important: female physical aggression against male partners is neither statistically absent nor theoretically exceptional. Any general theory of partner violence has to explain it rather than define it away.

“She hits only in self-defense”

This is one of the easiest ways to acknowledge female violence while preserving the assigned roles: the woman struck, but the man is still treated as the true source of the violence.

Self-defense genuinely matters. A systematic review of 23 studies on women’s motives for IPV found self-defense and retaliation among commonly reported reasons. In four of the 14 studies where motives could be ranked by frequency, self-defense was the leading motive. In other studies it was not. Across the review, anger, trying to get a partner’s attention, retaliation, and response to emotional hurt were recurring themes; control appeared as a motive but was not the leading motive in any included study.34

Even the category “self-defense” was unstable. Some women meant preventing imminent physical harm. Others used the term for violence after a partner had struck them or for actions framed as protecting their “emotional health.” In one study included in the review, 47% of women initially attributed their violence to self-defense, but that figure fell to 23% after they were told their account would be corroborated with other family members; the review authors correctly note that the reason for the change is unknown.34

A community sample of 412 women who used IPV against male partners likewise produced five motive clusters rather than one defensive explanation: expression of negative emotions, self-defense, control, jealousy, and a “tough guise.” All five were related to more frequent aggression.35 In another study of young women, anger, retaliation for emotional hurt, getting a partner’s attention, jealousy, and stress were common motives, while few women identified self-defense as the reason for their abusive behavior.36

The male-centered conclusion is not that women never defend themselves. They do — and they also hit out of anger, retaliation, jealousy, attempts to make a partner listen, and other motives. Female aggression cannot automatically be rewritten as self-defense simply because the victim is male.

The male victim is not statistical noise

A newer synthesis makes the point even more clearly. A 2025 review included 246 studies and 517 estimates of physical IPV in industrialized English-speaking countries. Average physical victimization prevalence was 14.6% for women and 11.8% for men. For perpetration, the pooled rates were 17.5% for women and 10.6% for men. The authors explicitly conclude that prevention and intervention should support victims and perpetrators regardless of gender.27

That is not perfect symmetry. Female victimization remains higher on average. But male victimization is not a rare exception: the gap is measured in percentage points, not orders of magnitude.

The official Crime Survey for England and Wales shows the same problem in a different dataset. For the year ending March 2025, the ONS estimated 2.2 million female and 1.5 million male victims of domestic abuse — roughly two men for every three women among estimated victims. Male representation is much lower in police-recorded domestic-abuse crimes. The ONS explicitly warns that police-recorded crime is not a prevalence measure and should not be substituted for population survey estimates.28

For a male-centered analysis, this distinction matters. If you look only at shelters, police, and severe injury, men become less visible than they are in population surveys.

How male victims disappeared from the refuge movement

The history should not be rewritten in the opposite direction. The first Chiswick safe house opened in 1971 genuinely was a refuge for women and children, not a formally gender-neutral shelter. The UK National Archives documents that clearly.29

But the philosophy of its founder, Erin Pizzey, quickly diverged from the formula “man = source, woman = victim.” In the 1982 introduction to Prone to Violence, Pizzey described domestic violence as a human problem, wrote that violence occurs in both men and women, noted that Chiswick employed male workers, and said the organization also ran a special house for men from troubled families.30

Pizzey later repeatedly told the same story: after the women’s refuge she obtained a separate house in North London for battered men, but wealthy men who had funded women and children refused to fund the men’s house, and she had to give it up. This is Pizzey’s own testimony, not an independently reconstructed accounting record, and should be labeled as such.31

Another part of the story is independently documented. Sandra Horley became director of the Chiswick refuge in 1983, after Pizzey’s departure; the organization was later renamed Refuge.32 In 1992 The Spectator asked Horley about treating abuse of men as a comparable social problem. She said there were resource implications because women’s refuges were struggling, and that such a message “could seriously affect our funding.”33

This does not prove that the entire women’s shelter system was created “for the money.” It proves something narrower and harder to dismiss: women-only infrastructure had a direct financial incentive not to widen the public frame to men, and a leading director stated that concern publicly.

For an encyclopedia of the male position, this is a central historical case. Male victims were not “discovered recently.” They were visible near the beginning of the modern refuge movement. The dispute was whether they belonged inside the same social problem and whether resources should be shared with them.

Bidirectional violence is a major pattern

A 2024 systematic review covering 42 empirical studies found bidirectional violence to be the most common pattern in the included literature, with psychological violence the most frequently reported bidirectional form. Men and women reported bidirectional violence at similar rates.5

That finding does not mean every battered person “started it too,” and it does not mean severe coercive abuse is normally mutual. IPV is a broad category. A shove during an escalating argument and a years-long regime of terror are not the same event merely because both count as partner violence.

Michael Johnson’s typology makes that distinction explicit. He separates situational couple violence, intimate terrorism, violent resistance, and mutual violent control. In his account, situational couple violence dominates general population surveys, while intimate terrorism and violent resistance are more prominent in agency samples.6

Johnson’s change in terminology is especially revealing. He had previously used the term “patriarchal terrorism,” then abandoned it because it prejudged both men’s and women’s involvement and implied that all such controlling violence was rooted in patriarchal structures. He still argued that most intimate terrorism in heterosexual relationships is perpetrated by men and that much of it is connected to patriarchal traditions. What he stopped doing was building that conclusion into the definition of the phenomenon itself.6

That is the proper order: identify the violence and its control context first; explain its cause second.

Counting acts is not the same as counting harm

Much of the symmetry argument collapses because different studies count different outcomes. One survey may count acts — pushing, hitting, grabbing. Another may count fear, injury, sexual violence, stalking, coercion, and danger to life. Those are not interchangeable measures.

Archer’s meta-analysis already showed the distinction: close-to-symmetric results for physical aggression could coexist with an injury burden tilted toward women.4 Current CDC surveillance likewise shows a heavier combined burden for women when physical violence, contact sexual violence, stalking, and related impact are considered together.7

Two claims can therefore be true at once:

  • women also use physical aggression, and violence is bidirectional in many samples;
  • severe consequences and lethal risk are not distributed evenly between the sexes.

The second fact does not erase the first. The first does not erase the second.

Real risk factors do not fit into one word

The CDC states explicitly that intimate partner violence is not caused by a single factor. Risk exists across individual, relationship, community, and societal levels. Its list includes heavy alcohol and drug use, anger and hostility, impulsivity, poor nonviolent problem-solving skills, economic stress, jealousy, relationship conflict and separation, childhood abuse, witnessing parental violence, dominance and control, and attitudes that justify violence.8

Meta-analyses support parts of that broader model. Adverse childhood experiences are associated with both later IPV perpetration and victimization. Emotion-regulation problems are associated with IPV perpetration; in a 2023 meta-analysis, the strength of that association did not differ by gender.9

Alcohol is another real factor, but inflated statistics weaken the argument. We do not repeat the claim that alcohol is present in “60–80% of domestic conflicts,” because there is no single defensible denominator for that figure. A 2026 meta-analysis instead found harmful alcohol use associated with increased odds of both man-to-woman and woman-to-man IPV perpetration.10

Control and sexist attitudes belong on the list. They are not the whole list.

A bad diagnosis produces bad intervention

If every violent person is told in advance that the single cause of the violence is a patriarchal drive for power, some actual mechanisms will never be treated. Substance misuse does not disappear after a lecture about power. Impulsivity does not disappear. Childhood trauma does not disappear. Conflict-regulation skills do not appear by themselves.

That is not an argument for indiscriminate couples therapy. A victim should not be sent to negotiate with someone who is using sustained coercive control and high-risk violence. Those cases require safety, risk management, and accountability.

Perpetrator programs still have to prove that they reduce violence. An updated 2024 meta-analysis of 59 controlled outcome studies found small overall effects. Newer approaches, including Acceptance and Commitment Therapy and Circles of Peace, produced the largest effects in direct comparisons with Duluth-curriculum groups.11

The useful cause is not the ideologically correct cause. It is the cause that actually explains the case and changes the behavior.

Where does aggression start: with the hit or with the provocation?

This question exposes a measurement problem. Research usually distinguishes provocation or instigation from the aggressive response. A provocation may be a disagreement, rejection, insult, threat, humiliation, jealous accusation, or another hostile act. Some provocations are not aggression at all. Others — threats, deliberate humiliation, verbal abuse — are already psychological aggression.

Experimental evidence leaves little doubt that provocation changes aggressive behavior. Bettencourt and Miller’s meta-analysis found that men were more aggressive than women when unprovoked, but that provocation sharply reduced that sex difference. Under sufficient provocation, reactive aggression becomes much less male-specific.12

Longitudinal couple research shows why counting only the first physical act can miss the beginning of an aggressive sequence. In 393 engaged couples followed into marriage, a person’s own psychological aggression and the partner’s psychological aggression both predicted first incidents of physical aggression; the pattern was consistent across sexes. A later study of 634 newlywed couples likewise found prior verbal aggression by both partners to be an important predictor of later physical aggression.13

So if one partner spends an argument threatening, humiliating or verbally attacking the other and the other partner then delivers the first slap, a dataset that measures only physical acts will correctly identify who initiated physical aggression. It cannot tell us who initiated the aggressive episode. Those are different questions.

Recent dyadic experimental work points in the same direction. In a 2024 study using real couples, negative emotion expressed by both partners predicted reactive aggression. Couples tended to retaliate in response to their partner’s aggression; the researchers found retaliation rather than a simple steadily escalating spiral. Forced breaks of only 5–15 seconds reduced aggression by interrupting the provoked state.14

Who starts it more often?

There is no strong basis for saying that one sex universally “provokes” more. The category is too broad, and studies operationalize the beginning of conflict differently. What the evidence does show is that male initiation cannot be treated as the default.

Capaldi, Kim and Shortt directly observed conflict discussions in a longitudinal sample of young, at-risk couples. At ages 18, 21 and 24, women were observed to initiate nonplayful physical aggression more often than men: 46% versus 10%, 28% versus 10%, and 20% versus 7% of sessions respectively. By age 26 the difference was no longer significant: 11% versus 8%. Across all four waves, men’s and women’s reciprocation rates were almost identical, about 24–26%. Among sessions with mutual physical aggression, 71.4% involved male reciprocation after female initiation and 28.6% the reverse.15

Those numbers are striking, but their limits matter. This was an at-risk sample, observed aggression was generally low-level, the number of mutual-aggression events was small, and some reciprocity and injury comparisons were descriptive rather than formally tested. It is evidence against a universal male-initiator assumption, not a population estimate saying women always start violence.

A 2018 study of 403 adults using hypothetical intimate-conflict scenarios found women more inclined than men to escalate the early stages of conflict after an aggressive verbal demand from a partner. That result concerns response to provocation, not proof that women initiate the original provocation, and the convenience sample and hypothetical design limit generalization.16

Ordinary couple-conflict research adds another warning. The familiar wife-demand/husband-withdraw pattern appears often, but it changes with who wants change and whose issue is being discussed. Both sexes become more demanding when discussing a change they want and more withdrawing when the partner wants the change. Recent work still finds men more likely to withdraw during the woman’s issue, while attachment and issue structure explain much of the remaining variation.17 A demand for change is not automatically a provocation and certainly not automatically abuse.

Population data are equally hostile to a simple sex monopoly. A six-country European study found high psychological-aggression victimization among both men and women and reported that participants predominantly experienced IPV as both victims and perpetrators, with few significant sex differences within cities.18

The cleanest way to study an incident is therefore to code the whole sequence: disagreement, insult, threat, coercion, psychological aggression, first physical act, retaliation, repeated physical acts, injury, fear and control. “Who hit first?” is one useful variable. It is not the same variable as “Who started the aggression?”

The invisible first half of the conflict

From a male point of view, there is another asymmetry that statistics on serious injury often miss. A man is on average physically more dangerous to a woman, so his retaliation is more likely to leave a visible injury. The same fact helps produce a much stronger social prohibition against male-on-female violence. Experimental studies judge violence against men as less serious and more justified, assign more blame to male victims, and prefer harsher punishment when the perpetrator is male.19

In practical terms, men know the rule: you do not hit a woman back, even if she hits first. There is direct, though old and limited, evidence that some female aggressors count on that asymmetry. In a study of 978 college women, 29% reported physical aggression against a male partner over five years. Among the reasons reported was an attempt to gain the partner’s attention, and assaultive respondents did not expect to seriously injure the man or expect him to retaliate.20 This is not a population estimate and it does not establish the motives of women as a class. It does establish that female aggression under an expectation of male non-retaliation is a real pattern, not a male fantasy.

Low-level aggression also does not begin and end with a slap. Standard IPV instruments treat threats, insults and breaking or destroying a partner’s belongings as forms of psychological aggression.21 From the male victim’s perspective, a slap, a thrown object, a smashed phone and deliberate destruction of property may belong to one continuous sequence even when no single event produces a medical injury.

Why men do not call the police

The male victim is frequently absent not only from feminist narratives but from formal help-seeking data. A systematic review of male victims identified fear of disclosure, shame, challenges to masculinity, fear of not being believed, and fear of being labelled the perpetrator. Men also described domestic-violence services as being visibly organized around female victims.22

In a separate Canadian study, male victims explicitly described avoiding the police because they expected ridicule, disbelief, or being arrested themselves.23 That does not justify the literal claim that women are “absolutely protected by law”: women are arrested and prosecuted for domestic violence too. But behavior depends on the expected outcome of seeking help, not on an abstract statement that the statute is sex-neutral.

That expectation is not baseless. One U.S. study of arrest decisions following domestic disputes found male suspects more likely to be arrested than female suspects after accounting for other factors.24 In a different longitudinal sample of young couples, official IPV arrests were predominantly of men even though, outside the official incidents, women showed higher levels of physical and psychological aggression and severe physical aggression did not differ significantly by sex.25

This is not a universal law of policing. Arrest patterns vary with jurisdiction, severity, injuries, weapons, mandatory-arrest rules and sampling. But the male strategy — “do not call the police; you may end up treated as the offender” — has an empirical basis, and that strategy itself makes female aggression less visible.

Restraint threshold and destructive retaliation

This produces a male version of the incident that can disappear when researchers count only the final injury. A woman may repeatedly use relatively low-danger physical or psychological aggression while expecting that the man will not retaliate. The man may tolerate it, withdraw, and avoid formal help. Provocation increases the likelihood of reactive aggression, while anger and hostility are associated with more severe IPV.1226

When restraint finally fails, physical asymmetry now works against the woman: one male strike can do more damage than several earlier slaps or shoves. Medical records, injury photographs and police reports capture that final part of the sequence far better than the first.

We are not claiming that research has already proved this chain as the universal mechanism of domestic violence. There is no longitudinal study that lets us honestly write: “usually the woman provokes for a long time, the man restrains himself, then snaps.” It is a male causal model assembled from separately supported links: women also initiate physical aggression; some women explicitly expect no retaliation; male victims avoid help-seeking; violence against men is judged more leniently; provocation increases reactive aggression; and male physical strength magnifies the consequences of retaliation.

That is enough to make one point non-negotiable: the severity of the last blow cannot be allowed to rewrite the history of the whole conflict. If a woman has slapped, humiliated, broken belongings or otherwise attacked a man repeatedly, and he later retaliates once and causes a serious injury, the injury matters. It does not erase the preceding acts or retroactively turn them into nonviolence.

Test the myth

“Domestic violence is a male instrument of patriarchal control over women” is too broad to survive as a general theory.

The evidence supports a narrower claim: coercive control and male domination explain some serious partner violence, but they are not a universal mechanism of IPV. Women also perpetrate violence. Violence is bidirectional in many studies. Male victims form a large rather than marginal share of observed domestic violence. Female aggression is not reducible to self-defense. Situational escalation differs from sustained terror. Severity differs from frequency. Alcohol, childhood adversity, emotion regulation, conflict, and economic stress also matter. Provocation and psychological aggression can precede physical violence, and a male victim may remain invisible until the moment he retaliates.

A male-centered account therefore has to reconstruct the entire sequence, not photograph only its most damaging frame: who began the insults, threats, property destruction and physical acts; who counted on non-retaliation; who attempted to disengage; who sought help; who used physical force first; who retaliated; and who was injured. Consequences matter. Consequences do not retroactively determine who initiated the violence.

Sources and notes

  1. Domestic Abuse Intervention Programs, history of the Power and Control Wheel: the wheel was created in 1982 by Ellen Pence, Coral McDonnell, and Michael Paymar for a court-ordered program for men who batter.
  2. Domestic Abuse Intervention Programs, Power and Control Wheel FAQ: DAIP distinguishes battering as an intentional pattern of control and domination within the broader category of IPV.
  3. Ellen L. Pence, “Some Thoughts on Philosophy,” in Melanie F. Shepard and Ellen L. Pence (eds.), Coordinating Community Responses to Domestic Violence: Lessons from Duluth and Beyond, SAGE, 1999, pp. 25–40, especially pp. 29–30.
  4. John Archer, Sex differences in aggression between heterosexual partners: a meta-analytic review, Psychological Bulletin, 2000, 126(5):651–680, DOI 10.1037/0033-2909.126.5.651.
  5. Andreia Machado, Catarina Sousa, Olga Cunha, Bidirectional Violence in Intimate Relationships: A Systematic Review, Trauma, Violence, & Abuse, 2024, 25(2):1680–1694, DOI 10.1177/15248380231193440.
  6. Michael P. Johnson, Conflict and Control: Gender Symmetry and Asymmetry in Domestic Violence, Violence Against Women, 2006, 12(11):1003–1018.
  7. U.S. Centers for Disease Control and Prevention, About Intimate Partner Violence, current NISVS summary of victimization and related impact.
  8. U.S. Centers for Disease Control and Prevention, Risk and Protective Factors for Intimate Partner Violence, 2024.
  9. Jenney Zhu et al., Adverse childhood experiences and intimate partner violence: A meta-analysis, Development and Psychopathology, 2023, DOI 10.1017/S0954579423000196; Molly A. Maloney, Christopher I. Eckhardt, Daniel W. Oesterle, Emotion regulation and intimate partner violence perpetration: A meta-analysis, Clinical Psychology Review, 2023, 100:102238, DOI 10.1016/j.cpr.2022.102238.
  10. Amany Tanyos et al., Perpetrator alcohol use and intimate partner violence in high-income countries: a systematic review and meta-analysis, Addiction, 2026, DOI 10.1111/add.70540.
  11. Julia C. Babcock et al., Which battering interventions work? An updated meta-analytic review of intimate partner violence treatment outcome research, Clinical Psychology Review, 2024, 111:102437, DOI 10.1016/j.cpr.2024.102437.
  12. B. Ann Bettencourt, Norman Miller, Gender differences in aggression as a function of provocation: a meta-analysis, Psychological Bulletin, 1996, 119(3):422–447, DOI 10.1037/0033-2909.119.3.422.
  13. Christopher M. Murphy, K. Daniel O’Leary, Psychological aggression predicts physical aggression in early marriage, Journal of Consulting and Clinical Psychology, 1989, 57(5):579–582; Julie A. Schumacher, Kenneth E. Leonard, Husbands’ and wives’ marital adjustment, verbal aggression, and physical aggression as longitudinal predictors of physical aggression in early marriage, 2005, 73(1):28–37.
  14. Annah G. McCurry, Robert C. May, David I. Donaldson, Both partners’ negative emotion drives aggression during couples’ conflict, Communications Psychology, 2024, 2:73.
  15. Deborah M. Capaldi, Hyoun K. Kim, Joann Wu Shortt, Observed Initiation and Reciprocity of Physical Aggression in Young, At-Risk Couples, Journal of Family Violence, 2007, 22(2):101–111, DOI 10.1007/s10896-007-9067-1.
  16. Zeev Winstok, Ronit Smadar-Dror, Michael Weinberg, Gender differences in intimate-conflict initiation and escalation tendencies, Aggressive Behavior, 2018, 44(3):276–284, DOI 10.1002/ab.21750.
  17. Andrew Christensen, Christopher L. Heavey, Gender and social structure in the demand/withdraw pattern of marital conflict, Journal of Personality and Social Psychology, 1990, 59(1):73–81; Gender, attachment, and demand/withdraw patterns in the context of moderate couple conflict in cisgender, heterosexual relationships, Journal of Marital and Family Therapy, 2024.
  18. Diogo Costa et al., Intimate partner violence: a study in men and women from six European countries, International Journal of Public Health, 2015, 60(4):467–478, DOI 10.1007/s00038-015-0663-1.
  19. Victim Gender, Rater Attitudes, and Rater Violence History Influence Perceptions of Intimate Partner Violence, Journal of Interpersonal Violence: violence against male victims was rated less serious and more justified, and male victims were assigned more blame; Jenna M. Wilson, Kimberly Smirles, College Students’ Perceptions of Intimate Partner Violence: The Effects of Type of Abuse and Perpetrator Gender, 2022.
  20. College women who initiate assaults on their male partners and the reasons offered for such behavior, Psychological Reports, 1997, 80(2):583–590, DOI 10.2466/pr0.1997.80.2.583.
  21. Ryan C. Shorey et al., Arrest History and Intimate Partner Violence Perpetration in a Sample of Men and Women Arrested for Domestic Violence, 2012: the study’s definition of psychological aggression includes threats and breaking or destroying a partner’s personal belongings.
  22. Alyson L. Huntley et al., Help-seeking by male victims of domestic violence and abuse: a systematic review and qualitative evidence synthesis, BMJ Open, 2019, 9:e021960, DOI 10.1136/bmjopen-2018-021960.
  23. Emmanuel E. Dim, Valeriya Lysova, Male Victims’ Experiences With and Perceptions of the Criminal Justice Response to Intimate Partner Abuse, Journal of Interpersonal Violence, 2022.
  24. Melissa Hamilton, Meredith G. F. Worthen, Sex disparities in arrest outcomes for domestic violence, Journal of Interpersonal Violence, 2011, 26(8):1559–1578, DOI 10.1177/0886260510370592.
  25. Deborah M. Capaldi et al., Official incidents of domestic violence: types, injury, and associations with nonofficial couple aggression, Violence and Victims, 2009, 24(4):502–519, DOI 10.1891/0886-6708.24.4.502.
  26. Christopher I. Eckhardt et al., Anger, Hostility, Internalizing Negative Emotions, and Intimate Partner Violence Perpetration: A Meta-Analytic Review, Clinical Psychology Review, 2015, 35:23–33.
  27. Alexandra Lysova, Karl Hanson, James Mackay, Rates of male and female physical victimization and perpetration in intimate relationships: a comprehensive review, Partner Abuse, 2025, DOI 10.1891/PA-2024-0038: 246 studies, 517 estimates; physical victimization 14.6% for women and 11.8% for men, perpetration 17.5% for women and 10.6% for men.
  28. Office for National Statistics, Domestic abuse victim characteristics, England and Wales: year ending March 2025: 2.2 million female and 1.5 million male victims; the ONS explicitly warns that police-recorded crime is not a prevalence measure for domestic abuse.
  29. The National Archives, How the first women’s refuge enacted change in the UK: Chiswick Women’s Aid opened the first safe house for women and children in 1971.
  30. Erin Pizzey, Prone to Violence, Introduction and Preface, 1982: Pizzey describes a human rather than anti-male philosophy, writes that violence occurs in both men and women, notes male workers at the refuge, and describes a separate house for men from troubled families.
  31. Erin Pizzey, interview on The Respondent, Episode 34: Pizzey says she obtained a separate North London house for battered men but could not secure donor funding and had to give it up. This is her own testimony.
  32. Refuge, Sandra Horley’s retirement message: Horley states that she became director of the Chiswick refuge in 1983 and that the organization was renamed Refuge in 1993.
  33. Domestic Violence: The Other Side, The Spectator, 28 November 1992: Sandra Horley said that presenting male abuse as comparable in scale could seriously affect women’s-refuge funding.
  34. Megan H. Bair-Merritt et al., Why Do Women Use Intimate Partner Violence? A Systematic Review of Women’s Motivations, Trauma, Violence, & Abuse, 2010, 11(4):178–189, DOI 10.1177/1524838010379003.
  35. Jennifer E. Caldwell et al., Why I Hit Him: Women’s Reasons for Intimate Partner Violence: community sample of 412 women who used IPV; motives clustered into negative-emotion expression, self-defense, control, jealousy, and tough guise.
  36. Physical and emotional abuse in romantic relationships: motivation for perpetration among college women: anger, retaliation for emotional hurt, attention, jealousy, and stress were common reasons; few women identified self-defense as their motive.