Skip to content

Outside the Man Box

The education of domestic violence offenders

When A.J. asked Faith to marry him, he considered it a practical question. He had returned in 2012 to Camp Pendleton in California after a tour of duty in Afghanistan, the war leaving him a quieter person than he’d been before deployment. He was less interested in socializing, now vigilant to surprise, and could no longer tolerate crowds or loud noises. Meanwhile, his fellow Marines were marrying their high school sweethearts and moving out of the barracks to start homes, marriage offering eligibility for a higher salary. A.J. wanted to get married too. He was twenty years old, proud of his self-sufficiency, and far from where he was born and raised, in Harlem. But there was one friend he’d made at Martin Luther King Jr. High School with whom conversations had grown longer over the last few years. He believed she was a good person, and he liked that they’d watched each other grow up.

Soon after Faith crossed the country to be with A.J., they began sleeping together, and their affection for each other grew. A.J. considered it meaningful that Faith witnessed him at his lowest points, when his trauma peaked. Yet the couple also began to argue. A.J. had envisioned a mutually beneficial partnership, not a clutter of emotional inconveniences, such as jealousies prompted by his admittedly inappropriate flirtations online or disagreements about money after he was discharged from the Marine Corps in 2014. A.J. hadn’t been looking for a housewife; he’d been looking for a partner, he declared. Briefly, they broke up, before reuniting upon discovering that Faith was pregnant.

In August 2015, their daughter was born. To A.J., the labor was miraculous. He’d wanted a boy, but when A.J. held his daughter, he had never been so happy. She had jaundice, and she looked like a tiny, yellower version of him; when she was placed in a phototherapy bed, he thought it was just like a microwave bringing her to temperature. In their new life together as parents, A.J. observed that Faith was a wonderful mother, that she would do anything for their child.

Yet having a baby would neither reconcile their disagreements nor mitigate the financial strain as A.J. attempted to establish a post-military career back in New York. He tried construction. He tried rideshare driving. He tried security. None satisfied him, and as the list of jobs grew, A.J. convinced himself that his wife was perfectly comfortable allowing him to assume financial responsibility for the family. He began to imagine her a burden. Where once he’d encouraged her, now he enumerated her supposed shortcomings. He knew she was insecure about her body, so he insulted it. He thought up the most hurtful things he could say, then said them. When their daughter was two years old, Faith moved out to live in a shelter.

Though domestic violence constitutes one of the direst public health and criminal justice crises in the country, its gravity has been a belated and recent revelation in the American psyche.

One day in early 2018, when she returned to the apartment they’d once shared, they had a fight. A.J. pushed her, and, in his words, “physically restrained her,” holding her tightly enough that he left bruises on her bicep. That summer, they met again at a Harlem emergency room, where their daughter was being treated for a fever. The estranged couple began to argue, A.J. throwing anything at Faith within reach. The detective who investigated the incident wrote in a report that A.J. shoved his wife, kicked her, and struck her in the back of the head with the end of an umbrella. A.J. was charged with two counts of assault in the third degree, endangering the welfare of a child, aggravated harassment in the second degree, attempted assault in the third degree, and harassment in the second degree.

In September, the New York City Criminal Court ordered him to attend a psychoeducational group for perpetrators of domestic violence. Over the last few decades, it has become relatively common for such options to be offered as a condition for deferred and reduced sentencing or probation arrangements. Most of these programs follow the Duluth Model, which organizes its approach around the principle of offender accountability, framing abuse as volitional and therefore able to be changed by the actor.

On October 29, 2018, A.J.’s group was scheduled to begin. He did not want to go. He did not want to tell anyone what he was doing on a Monday night. He did not want to sit among people he estimated would be scumbags. But he anticipated that the class would improve his odds of a favorable custody agreement, and he loved his daughter. When he was with her, he liked bringing her to Chuck E. Cheese and taking her on walks through the park. They played karate and dolls and watched cartoons. Normally fastidious in his diet, he permitted himself to eat the sugary things she favored: cake, cookies, ice cream. He loved watching her grow up.

And so, that autumn evening, he arrived at a building in midtown Manhattan. He signed a clipboard to be granted entry to a room high in the building where abusive men went to remake themselves as men who had once been abusive. In the lobby hung paper flowers and illustrations inspired by Through the Looking-Glass. One card read, “Every adventure requires a first step.”

Reinventing the Power and Control Wheel

Though domestic violence constitutes one of the direst public health and criminal justice crises in the country, its gravity has been a belated and recent revelation in the American psyche, one many would still consider provisional. Around one in four American women will be harmed by a partner over the course of their lives, and while violent crime has declined in recent years, homicides due to domestic violence have not. Over half of the women killed in this country are killed by a loved one. Covid-19’s stay-at-home orders have left those suffering domestic violence more cut off from resources to protect them from abuse, including formal services like health care and shelters, as well as sites of informal social control like public playgrounds and churches, which have been shown to regulate the occurrence of abuse.

While “wife beating” has been illegal in every state since 1920, it was, until the 1970s, a crime often treated as a marginal misfortune. Its prohibition went largely unenforced by police and prosecutors. Psychologists tended to label survivors narcissistic, labile, and irresponsible. It took two to tango, the prevailing attitude held, and authorities were generally unwilling to step in to end the dance.

The battered women’s movement, as it was then known, emerged in the 1970s to advocate for the interests of those harmed. The first shelter in the United States opened in St. Paul, Minnesota, in 1974. Six years later, Ellen Pence, a feminist activist who would go on to earn her doctorate in sociology, co-founded the Domestic Abusive Intervention Project (DAIP). At once learned and funny, Pence bantered with judges, told irreverent jokes, and her warmth served her ambitious vision of building a “coordinated community response” to abuse, encompassing measures such as systematic arrest of offenders, court action, probation, resources for victim support, and auxiliary social services.

It soon became apparent, however, that at least one social service was failing: programs for anger management seemed unable to address how gendered beliefs could structure intimate partner violence—for example, that a man who believed a husband was entitled to his wife’s unqualified support might consider his abuse a warranted response to unmet expectations. Abusers in anger management tended to minimize or justify their actions: “I have a short fuse,” they might say, or “my temper got the best of me.” Counselors, in turn, colluded in assigning blame to victims: “You have to stop letting her push your buttons.” To hear it told in many of these dialogues, abusers were reacting, being triggered by their victims.

In 1980 and 1981, Pence collaborated with a Minneapolis City Council member named Michael Paymar, on what eventually came to be known as the Duluth Model of interagency domestic violence work. Pence and Paymar viewed abuse as both a choice and a learned behavior. They imagined a program that would not simply treat offenders as angry people but would teach them to take responsibility for what they’d done and, crucially, change how they behaved in the future. The duo consulted with survivors of domestic violence, survivor advocates, and domestic violence service providers, designing a curriculum that adapted and combined cognitive behavioral therapy strategies with an education in how gender and power contribute to tactics of abusive control. Modifying the activist-educator Paolo Freire’s critical pedagogy, Pence and Paymar settled on a group education process organized in facilitated discussion.

Central to the curriculum was the Power and Control Wheel, a visual representation of eight tactics typically used by batterers: coercion and threats; intimidation; financial abuse such as withholding access to family funds; exercising male privilege, as in the case of impelling domestic servility; using children, as in threatening custody withdrawal; minimizing, denying, blaming; isolation; and emotional abuse. The chart was paired with the Equality Wheel, which offered eight healthy relationship behaviors: negotiation and fairness, non-threatening behavior, respect, trust and support, honesty and accountability, responsible parenting, shared responsibility, and economic partnership.

In 1984, the U.S. Attorney General’s Task Force on Family Violence began recommending batterer intervention programs in cases not involving grave injury to the victim, and the Duluth Model became “the North Star” of abusive partner intervention work. It was revolutionary: a feminist education had been foregrounded as a solution to a crisis in crime and public health.

Watching Me, Watching You

On March 18, 2019, six men quietly entered a room at the Children’s Aid building in New York City, serious in their sweatshirts and tees. Among them was A.J., his hair running in neat braids down his scalp and his white T-shirt pristine. He took a seat at a conference table surrounded by walls lined with hand-written posters. One listed abusive behaviors such as gaslighting, justification, and deflection. Another defined accountability: “Taking full ownership/responsibility of the consequences that come with your thoughts, feelings, action. Also taking initiative to repair any harm done.” A third showed the Anger Volcano, in which a vertical line divided specific thoughts and actions and horizontal lines delineated their intensity from low to high. On the low end were thoughts like “YOU BUGGIN’”; on the high end, “I AM GOING TO KILL SOMETHING.”

The Family Wellness Program at Children’s Aid models its curriculum after Duluth and offers services specifically to fathers who have committed intimate partner violence. Receiving referrals from the Integrated Domestic Violence Court, Child Protective Services, Family Court, foster care, and attorneys, its ambitious goal is to end intergenerational cycles of abuse. According to Kerry Moles, the founding director of the Family Wellness Program, along with sexual abuse, family violence is “the grandmother” of nearly every social problem we face, a trauma begetting trauma. Family Wellness looks to end this replication of harm.

On this night, the topic was how exposure to domestic violence affects children. The group facilitator, Albery Abreu, had prepared a PowerPoint, “Helping Children Heal from Domestic Violence: A How-To Guide.” He offered the men a worksheet that asked them to check off symptoms of trauma their children may have shown. The point of the exercise was that the traumas of intimate partner violence live long and run through the family.

Over several weeks, men in the group had accrued a file folder full of resources: breathing techniques; family charts; lists of emotion words that might be used to denote their interior lives rather than expressing them through abusive behavior. Abreu was tasked with conveying what the textures of non-abusive behavior are, how the men could concretely manifest the abstract ideal. He often led the men in role-playing exercises designed to build empathy. And perhaps most important was his ability to offer supportive confrontation, holding the men to account and urging change with empathy.

“I want you to step inside your children’s lives a little bit,” Abreu said. “What questions do you think your children are going to have for you if they have a chance to talk to you about what happened between mommy and daddy?”

A.J. raised his hand. He volunteered that his daughter had asked if he loved her mother. He had not been allowed to see her for four or five months. He recognized that his absence from his daughter’s life might relate to the question, but why, he wondered, would she ask such a thing if she hadn’t witnessed anything bad happen between her parents for a year? She was only a toddler.

“I’m curious,” Abreu responded. “You asked that question, where she got it from. Do you know where she got it from?”

“She asked me, ‘Do you love me, do you like me?’ I said, ‘I love you, I like you. You my best friend.’ And then the next question is, ‘Do you like my mommy? Do you love my mommy?’”

“So let’s not minimize it, because the absence is one thing. The question is, even when you were in front of mommy, could you say you displayed love to mommy?”

The Duluth Model was revolutionary: a feminist education had been foregrounded as a solution to a crisis in crime and public health.

The group was familiar with minimizing. The term was on a poster. But Duluth was designed around Pence’s application of Freire’s critical pedagogy theory. As she understood it, “if people who had been taught to lead their lives in uncritical or unreflective ways are to begin to think critically and reflectively, then the educational process must start in the real moments of their lives, not in the abstract.” And so, those in A.J.’s group needed to notice minimizing as it happened, to identify minimizing behaviors in practice. Fathers in the program often admit they know that it is wrong to “lay hands on a woman.” But, if one is to believe them, as they often understand it, the denotation of the phrase is slim—hitting or punching. The curriculum asks them to contemplate not only other forms of physical aggression but verbal, psychological, emotional, financial, nutritional, and sexual abuse too. Program leaders hope to instill that abuse is much more capacious than its rendering in the popular imagination, bigger even than the United States Department of Justice has understood it since April 2018, when it erased from the definition of domestic violence the phrase “non-physical abuse.”

A.J. pitched over the table to answer Abreu. “We haven’t been in the same room but one time besides court. There’s one time at my daughter’s performance. We was on opposite sides of the auditorium.”

“So she hasn’t seen what a two, three-year-old brain would consider love. It’s not a question that’s coming out of nowhere.”

Abreu’s voice was soft. He folded his hands on the table then hinged them out, fingers splayed. Much of his job was to prod the men toward empathetic thinking for those they had harmed. He wanted A.J. to consider what his daughter might be feeling. He wanted the group to do so too. Duluth’s group design was meant to circumvent some offender defensiveness so that men would examine abuse external to themselves and in the process acquire “new eyes” on their own acts of violence.

“They watching you literally like we was watching when we was coming up,” J.L., a lean man sitting diagonally across from A.J., added. “When we close the door to talk, me and wifey, my daughter, she opens the door. She wants to know what’s going on: ‘Why’s the door closed?’” J.L.’s voice was becoming louder, mimicking the distress of his daughter, or maybe his own. “‘Why’s the door closed. Why’s the door closed? Why’s the door closed? Mommy, everything okay?’” J.L. stood up. His hoodie was pulled over his hat, and his boots fell heavily on the carpet as he paced. He described his four-year-old daughter running between them during arguments, saying, “What’s the matter, momma? I don’t want you guys to fight.”

J.L. walked to the wall, pivoted. His head bobbed, as though affirming that he’d heard himself. “Maybe this is something that I brought upon myself with my kids and I gotta find out how to deal with it.”

“I apologized ten times,” A.J. said.

“Yeah, sometimes it takes that,” J.L. said. He was weeping.

A.J. leaned back. “Ten years later you gotta deal with that shit?”

“Sometimes it takes that long because of what we brought up in the past,” J.L. said.

In the phrase facilitators favor, it can be “difficult to sit with” having committed domestic violence, and the men’s apparent acceptance of their own accountability fluctuated over the course of this session. At one point, J.L.’s frustration mounted as he related how his oldest son had rejected him when he was released from prison. His son seemed to hate him, and he was tired of trying to win over a full-grown man.

E.T., sitting close to the head of the table, offered another view. “Your twenty-one-year-old son feels the same as his nine-year-old,” he said, gesturing toward another man, whose child had expressed feelings of neglect. “But because he’s from the street, he’s in that man box. That’s how he expresses it. He doesn’t have that nine-year-old mentality where he can just tell you, ‘I feel abandoned.’”

At the end of the class, after the PowerPoint and worksheet, Abreu reminded them that what they were doing in the group was practice. They had not yet perfected non-abusive behavior. But he believed they were getting closer.

Duluth Be Told

A.J. told me that he knew what he did to Faith was wrong but also that it was “normalized” in the environment of his upbringing. By “normalized,” he meant that violence between couples was somewhat routine, an observation borne out by the fact that over twelve million men and women experience domestic violence in the United States every year. After an instance of domestic violence, life went on, and usually, nothing happened to the perpetrator.

Many abusers in the Family Wellness program witnessed intimate partner violence growing up. One man recalled pretending to sleep while his father beat his mother. Another remembered hiding in a bush with his mother while his father pursued her with a machete. Programs like Family Wellness adopt a trauma-informed approach, meaning they acknowledge that abusers themselves have often experienced trauma, that trauma is ubiquitous—and that inflicting it is, regardless of circumstance, insupportable.

Though only 2 percent of North American men agree that it is “OK to hit your wife/girlfriend to keep her in line,” half the abusers in a National Institute of Justice study reported that they saw “battering as acceptable in certain situations.” These statistics are not irreconcilable. An Abusive Partner Intervention Program (APIP) aims to show offenders that the “certain situations” in which they believe battering to be acceptable are not, and that the belief in their legitimacy is grounded in an unacceptable desire to keep intimate partners in line, a desire that is structured by patriarchy.

Family violence is “the grandmother” of nearly every social problem we face, a trauma begetting trauma.

In 1998, however, the psychologist Donald Dutton proposed a theory that discarded the social critique underpinning the Duluth model. In The Abusive Personality, Dutton argued that cyclical abusers present with what he called borderline personality structure. It was immaterial whether the abuser was male or female: early in life, abusers formed a dysfunctional attachment style in which intimacy was entwined with pain and rage. The individual then relied on intimate partners to cohere a precarious sense of self and ameliorate their own anxiety. Unable to communicate their fear of abandonment or other emotional needs, they viewed their intimate partner as failing the task of holding together their threatened selfhood. This pattern explained their “transient psychotic states.”

In Dutton’s telling, the abusive personality, plotted from early wound to eventual tragic misstep, was nearly archetypal. His theory was operatic and narrative. It was also distinctly dismissive of the feminist and sociological dimensions that characterized the thinking behind batterer intervention, and it established Dutton as one of the most prominent and persistent critics of APIPs. “Male ‘sexist beliefs’ are deemed to be at the heart of IPV [Intimate Partner Violence], which is seen as exclusively male perpetrated,” Dutton wrote of the Duluth Model. “The facilitator confronts these beliefs. Clients who do mention anger, impulse problems, their partner’s abuse, or the abuse they experienced in their family of origin are told by the facilitator that they are in denial. Needless to say, a therapeutic bond cannot form.” The abusive behavior remains, he claimed, “largely because of policy that is based on a misconceptualization of the causes of IPV as due to gender, rather than to psychological, factors.”

Soon after the publication of The Abusive Personality, Edward Gondolf, a sociologist with a background in psychiatric epidemiology, published the results of a four-city study. Gondolf saw little evidence of Dutton’s “abusive personality” and wrote that domestic violence is committed largely by men who don’t demonstrate psychopathology.

Over the course of roughly two decades, Gondolf and Dutton volleyed arguments in the pages of academic journals. Gondolf was irked that Dutton ignored cognitive behavioral therapy’s centrality to the Duluth Model. While Gondolf acknowledged a small subset of “particularly problematic men” who would not respond to Duluth techniques, he emphasized that patients with severe psychiatric disorders were often unresponsive to any treatment type. He pointed out several obstacles to producing the empirical evidence Dutton called for to justify the Duluth model’s strategies, as well as weaknesses in the data that did exist: studies received low response rates from survivors and program attendees, generally failed to assess facilitator competence prior to the experiments, and did not administer the minimum “treatment dose.” Some men in a National Institute of Justice study, for example, had received only eight weeks of group work, while others received twenty-six. It was little wonder that studies so far had not proved the programs’ efficacy.

Dutton, for his part, insisted that Gondolf, whose interest in APIPs originated with his work at women’s shelters, had forsaken data in favor of feminist ideology. He and his colleague Kenneth Corvo impugned Duluth as “data-impervious” and “coercive,” arguing that Duluth’s focus on accountability was an “anti-therapeutic” form of “gender-shaming.” In the portrait they drew up, Duluth did not adequately address mental health and substance abuse problems. Positing the superiority of “evidence-based” psychological practice, they criticized Duluth as too political, even as they failed to acknowledge Dutton’s own ideological affiliations as the founding member of a legislative advocacy organization who counted among his supporters Father’s Rights groups.

Many of Dutton and Corvo’s criticisms mischaracterized Duluth. The model acknowledges psychological factors in abusive behavior but does not claim it is caused by pathology alone. Though Duluth-based programs don’t provide substance-abuse services in their nonviolence programming, many do refer clients to substance-abuse treatment. And Duluth recognizes that women commit abuse but does not understand it to be “symmetrical” to male abuse, considering it primarily “secondary offender” violence—defensive and less destructive.

Material Harms

Developing accountability through abusive partner intervention requires commitment. Around half of all participants in such groups drop out, but A.J. had shown himself to be motivated in other aspects of his life. After gaining weight during a previous period of depression, he had taken up bicycling forty miles every Sunday. He had done bootcamp, of course. And because he believed it aided emotional regulation, he walked his South African Mastiff, Nala, for two hours each day.

One night, A.J. sat down to complete a homework assignment from the program: write a letter to his child about the violence he had committed. It wasn’t until after Afghanistan that he had realized he was uncomfortable with the thought that he might die never having repaired the relationship with his own father, who hadn’t been around when he was a child. Even five years after their reconciliation, A.J. had been too overwhelmed by the unfamiliarity of his father calling him to wish him a happy birthday to pick up the phone. Now, A.J. found himself inhabiting the same role of a father who had caused pain to his child.

In the first draft, A.J. wrote just the facts, in language a child would understand. Then he crumpled it up; it had not gone far enough. He decided to write a letter to his daughter not as she was now, a toddler, but as a teenager to whom he could speak in more complex terms. In this version, he apologized to her for depriving her of a traditional family, with two parents who modeled healthy love and stayed together their entire lives. He admitted to hurting her mother but wrote that he would never hurt her. He wrote that he loved her, loved her, loved her. He wrote that despite what he’d done, his daughter could count on him going forward. He promised to be better.

Whether and how much the men who go through programs actually do get better is, of course, highly contested. In the absence of federal regulation, APIPs vary from state to state. They can range in length from weeks to a year. Some facilitators hold doctorates in psychology or masters’ degrees in social work. Others, pejoratively referred to as “paraprofessionals” by the evidence-based school, do not. And despite Duluth’s primacy, another competing philosophy, the New York Model, has disputed that abusive partner intervention works at all. Developed by the New York State Office for the Prevention of Domestic Violence, the New York State Coalition Against Domestic Violence, and Phyllis Frank of the mental health nonprofit Volunteer Counseling Services, it recognizes that individual men can change, but it does not understand batterer programs to be an effective vehicle for that change. The New York Model’s main purpose, Phyllis Frank has stated, is not rehabilitative. It is to demonstrate to abusers that “the court means business.”

Jennifer Krapf, a survivor of domestic violence whose former partner attended a batterer program in Rhode Island during the 1990s, told me that physical abuse declined after the intervention classes, but verbal, psychological, and sexual abuse increased in severity. She was shocked and disappointed. Sometimes, she thought she’d rather he beat her. It seemed to her that in the absence of visible physical scars, he had friends and family fooled; he presented the false image of a reformed man. Indeed, after serving time in prison and attending a group, after Krapf left him and moved out of state, he committed felony-level domestic assault on multiple occasions with a new partner.

Whether and how much the men who go through programs actually do get better is, of course, highly contested.

Still, Krapf remains conflicted about abusive partner intervention as a whole, both recognizing that APIPs have evolved since her former partner attended one and skeptical that their project is viable. “I believe everybody deserves a chance. . . . You’re not born an abuser,” she said. “But I don’t know that some people can be cured. I don’t know. Is there even an answer for that?”

In APIP circles, when I asked about evidence of program efficacy, invariably the question was returned: How do you measure success? The responses were not meant to be opaque. No one in the APIP community claims the programs work for all offenders, but they consistently point to the lack of adequate research. Most clinical studies use re-arrest as the marker of recividism, in part because re-arrest is the most accessible measurement, available in criminal justice databases. However, these databases do not include non-physical forms of abuse, which survivors often name as the most insidious form of abuse they’ve suffered. What’s more, some studies that measure re-arrest may not accurately capture the frequency or intensity of abuse, an important measure of harmful effects. And, of course, not all domestic violence offenders are arrested in the first place.

In the spring of 2019, I went to Portland, Oregon, to discuss these results with Chris Huffine, who runs a program called Allies in Change. Huffine is a friendly, burly man, a doctor of psychology with a current of Chicago gruff in his voice. One graduate of Allies who said he had abused women for decades had told me the program changed his life. The man idolized Huffine, quoting him like the Tao Te Ching. And across the country, intervention folk will hear his name and touch their hearts, speaking his name in emphatic tones of admiration: “Chris.” When I met with him in Portland, the trees were flowering. We sat in the Allies office while Huffine described the productive discomfort when he confronted partners who have been abusive in “the hot seat.” But how, I asked him, did he know the programs worked? “Women tell us,” he said. “Survivors thank us. They’re the people we’re really, truly serving.”

Huffine’s answer is supported by the Project Mirabal study, conducted by Durham University and London Metropolitan University between 2009 and 2015; it attempted to honor survivor testimony as the most reliable and significant evidence in domestic violence cases. Surveying survivors three months before abusive partner intervention and one year after the program began, representing a fifteen-month period, the Mirabal researchers assessed program efficacy by looking at six measures of success, including an improved relationship underpinned by respect and effective communication; expanded “space for action” for women that restored their ability to make choices and improved their well-being; and safety and freedom from violence and abuse for women and children.

No one in the APIP community claims that the programs work for all offenders, but they consistently point to the lack of adequate research.

At intake, 54 percent of survivors responded that their domestic violence perpetrator had “punched, kicked, burnt, or beaten” them. One year later, the number was down to 2 percent. Thirty percent said they’d been made to do something sexual they didn’t want to do. A year later, none had. Perhaps the most interesting findings were those regarding survivors’ “space for action,” the ability to act freely as agents of their own lives. Fifty-six percent of survivors initially agreed their abusive partner or ex-partner tried “to prevent me from participating in activities or groups outside the home”; that number was down to 12 percent at the time of the final survey. Project Mirabal captured data that the “evidence-based” studies had largely ignored: even if eventual re-arrest occurred, APIPs could dramatically improve the quality of everyday life for survivors and their children.

Critics of abusive partner intervention often point to a 2004 meta-analysis led by the psychologist Julia Babcock, which measured efficacy by police and partner reports of any physical re-assault. Babcock’s team found that APIPs produced only small improvements over arrest alone. Yet the treatment results were similar to those reported by substance abuse programs, and as Babcock pointed out, while the treatment effect might be small, it was greater than that of an early study on the use of aspirin for heart attacks; aspirin still saves many lives. Her conservative measure would still mean approximately forty-two thousand fewer women being battered per year in the United States alone.

Duluth-style administrators ask: If abusive partner intervention programs save, at a minimum, forty-two thousand American women a year from domestic violence and at best, as the Mirabal study reports, reduce severe physical violence by 96 percent, isn’t that sufficient reason to support abusive partner programs and further research? What is the alternative to reeducating offenders?

It’s become common to view arrest as a deterrent since 1984 when the Minneapolis Domestic Violence Experiment associated arrest with lower recidivism. However, subsequent studies have found the results of arrest are ambiguous. For example, offenders who were employed were less likely to lapse into abuse after arrest, while those who were unemployed were more likely to commit assault again if arrested. Though abusive partner intervention has been criticized for its putatively statistically insignificant results, a major new meta-analysis has found that arrest leads to reduced recidivism in less than 5 percent of cases.

In New York City, the average cost to incarcerate an inmate for a year is $168,000. Before the Office for the Prevention of Domestic Violence decided to cut funding for APIPs under agency budgetary restraints in 2009, it provided around $200,000 across the state, and the Mayor’s Office of Criminal Justice from FY 2013 onward has provided anywhere from $593,082 to $949,457 annually to abusive partner classes, less than the cost of incarcerating six individuals.

Abusive partner intervention programs and carceral solutions are not mutually exclusive—some APIP attendees are locked up, though most domestic violence cases are misdemeanors without prison sentences—and, in fact, the interagency Duluth Model, which includes the Duluth Curriculum classes, is committed to systematic arrest. Yet it is difficult to ignore the gap in funding between punitive and rehabilitative work, and one 2015 study found that the probability of recidivism for those who served jail time alone was two to four times greater than those assigned to one-year batterer intervention programs.

“We have consistently invested in a crisis response frame,” Purvi Shah, a consultant on the Interagency Working Group on NYC’s Blueprint for Abusive Partner Intervention, wrote, “while short-shrifting the long-term work of transformation.” This transformation could operate as an arm of larger abolition projects. Though the original coordinated community response envisioned by Pence and Paymar centered mandatory arrest, the future of intervention work might divest from the carceral paradigm, its commitments to the Power and Control curriculum and community-based care taking precedence. The emotional and behavioral education of Duluth is not incompatible with police abolition or defunding, and indeed, some of the most promising programming today raises youth consciousness on intimate partner and gender-based violence.

Abusive partner intervention is also relatively inexpensive. The Family Wellness Program spends around $12,500 per individual treated. Meanwhile, by the time a child who has been exposed to domestic violence reaches the age of sixty-four, the abuse has cost the national economy $50,000, according to one Case Western Reserve University Study. Allies in Change spends less at $2,184 per client per year but manages to survive only because some employees are volunteers. Huffine regrets that he is unable to offer health insurance or training opportunities to his staff, and subsidies from the county cover only approximately half the cost of the program.

But given the interdisciplinarity of APIPs, establishing who—or which public agency—is responsible for payment becomes complicated. Under austerity, there is little impetus for agencies with already strained budgets to pick up the bill for social services, education, and care-based work that might be assumed by others. In the criminal justice-oriented New York Model ethos, perpetrators ought to shoulder financial responsibility as part of their punishment, and many programs have adopted this practice.

Those who have experienced domestic violence are, in the conventional heroic narrative, supposed to gather a bit of dignity, call the cops, and leave.

Yet this, too, carries a cost. Family violence occurs across class lines, and it is frequently stated that statistics correlating poverty and abuse may be misleading, since more privileged people can elect to leave a relationship, rather than reporting abuse to the authorities. Nevertheless, billing offenders can exclude underprivileged communities from access to rehabilitative justice; economically disadvantaged men are less than half as likely to complete intervention programs. Intervention payment requirements can also leech financial resources that might normally be spent on family expenses, bringing material harm to those in already underserved communities.

Village Logic

In December of 2018, around a month after he had begun the program, A.J. was arrested for second-degree harassment, to which he pled guilty. He admits that he became frustrated by Faith’s commentary in a Facebook group and lashed out. He had not realized that online harassment was a punishable offense. After the Facebook incident, Faith was granted an order of protection from A.J., and while enrolled in the Family Wellness Program, he was not allowed to make contact with her, relying instead on his mother to make arrangements to see his daughter. Others at Family Wellness now have new partners. Some, even after arrest, have sustained their relationships.

Though the goal of abusive partner intervention is not to preserve the couple as a unit, critics of the programs have advanced a troubling hypothesis: the safety of survivors is compromised by a false sense of security that allows to them to justify remaining in the abusive relationship. The charge follows from the popular conviction that the correct response to domestic violence is leaving the offender. Those who have experienced domestic violence are, in the conventional heroic narrative, supposed to gather a bit of dignity, call the cops, and leave.

Yet that is often unrealistic or undesirable for the families involved. As Danny Salim, the anti-violence program director at the Arab-American Family Support Center, points out, “In some cases, survivors want to keep their families, want the abuse to stop, and the family to reunite. In other words, many survivors express that they want to hold abusers accountable but give them another chance.”

Both Purvi Shah and the Attorney General’s Task Force on Family Violence found that some survivors do not see intimate partner violence as the worst harm they face. Undocumented immigrants know involving the criminal justice system likely will end in their own deportation or that of a loved one, who may be one of very few people with whom they share a language and culture. Leaving a partner may present a conflict with religious faith. The abused may not have the financial means to leave, nor job prospects that would allow them to support themselves or their children. And of course there may also be a resistance to soliciting police involvement at all, particularly in disproportionately policed Black and brown neighborhoods, where law enforcement is sometimes viewed as a more insidious threat than intimate partner violence. Mental health professionals have documented trauma symptoms in victims following negative experiences with the criminal justice system. Police—40 percent of whom self-report as domestic violence perpetrators—may aggravate psychological and emotional damage.

The idea that APIPs cause survivors to remain with abusive partners is, further, a view that infantilizes the predominantly female survivors, presuming that they cannot be trusted to decide what is best for themselves. Those shaming the abused for the choice to stay “almost contribute to the abuse,” one survivor told me, by faulting those harmed for “allowing” themselves to be hurt.

And against the presiding notion that those harmed by intimate partners ought to buckle up and break up stands a stark body count. The period immediately after a woman leaves an abusive partner is the most lethal; on average, in the year after ending the relationship, women are 3.64 times more likely to be killed by their partner than abused women who stay. Judicial protections may not help. One-third of intimate partner homicides occur within the month after a restraining order is issued.

An APIP may be one step in a survivor’s safety plan. And, of course, a breakup is not necessarily an absolute severance of association, particularly for those with shared custody. The programs are, then, not intervention alone. They are also prevention. Behavior continues over a lifetime. APIPs may assist in reforming it, whether the participant remains in the relationship or finds a new partner.

Nevertheless, the programs’ reputations as services only for perpetrators that divert funding from resources for survivors has limited the financial support they receive from governmental agencies, even as judicial referrals continue. Program administrators often struggle to afford the labor-hours necessary to complete paperwork required by the courts—and to pursue the private donations and grant money that must fill the gap left by the state. As such, many facilitators are part-time employees. A.J.’s group was meant to be led by two facilitators, one male and one female, to model an egalitarian relationship across gender. But it needed to make do with a single group leader after one was no longer able to accommodate her part-time work at the Family Wellness Program when the schedule changed at her primary job. Many practitioners cite one year of APIP classes as the minimum required to instigate behavioral change. Yet shoestring budgets force programs to offer shorter terms. More limited staff training, less follow-up with clients, and fewer meetings with supervisors all mean less successful outcomes.

And as Edward Gondolf elaborated in his debate with Donald Dutton, these programming discrepancies also make it difficult for researchers to compare efficacy. The meagerness of the research means legislation and oversight vary wildly. Some states, like New York, have found it impossible to define best practices, never adopting regulations around APIPs at all. While the lack of regulation has encouraged innovation, it has also permitted lackluster programs to swing the empirical evidence that does exist against them. This only makes it more difficult to obtain funding. APIPs find themselves in a vicious cycle: the belief that they don’t help survivors limits their access to funding, that lack of funding in turn damages their performance, and their scattered performance weakens their ability to prove they help survivors and thereby fund additional research and development.

Chris Huffine thinks of the phenomenon as self-perpetuating, not unlike the problems facing under-resourced schools. What happens when excellent teachers work in schools with broken desks? When he talks about it, he’ll often paraphrase a line from the movie Spotlight—“If it takes a village to raise a child, it takes a village to abuse one”—the idea being that the systematic dismissal of abuse intervention sustains intimate partner violence too.

Graduation, or Commencement

Abusive partner intervention participants tend to say they haven’t always understood the mindfulness that being non-abusive entails. One man told me that though he’d previously gone to individual therapy, he had never been taught to delineate thoughts, feelings, and action until joining a program. Once, during what he called “petty season,” he texted his former partner, “I’ll piss on your grave.” Now he understood that his language had been a threatening action, not a natural expression of the feeling word he learned in group, “anxiety,” which he might have simply named in the moment. Another graduate of a program said that after decades of being an abusive man, he still needed to be attentive to practicing these metacognitive skills. Still another said the first abusive partner intervention program he’d attended had not helped. It was only when he was sent to a second, different program that his mind began to change.

In April 2019, I returned to the Family Wellness Program to see a group as it was intended to run, with two facilitators: Albery Abreu and Cheryl Hurst. That night, the group was supposed to reflect on their violence. One man had threatened his partner with a knife. Another had attempted to control the food his wife ate. J.F. mentioned becoming angry in the car with his family. He’d begin shouting, and soon his partner would be offering suggestions to him about road rage, and he was raising his voice, and people in other cars couldn’t even hear him. Cheryl wanted to know more about his reaction to his partner’s advice.

“Sometimes you even getting more upset,” J.T. said.

“Who’s getting upset?” Albery said.

“Me,” J.T. admitted.

He was a quiet man but spoke quickly. The desks were arranged in a horseshoe that night, and J.F. turned his head to look at his peers occasionally as he drew up the choreography of when he would, enraged, “blow out.”

“I’m wondering how gender plays into this,” said Cheryl. “What does your son think about women? What does your daughter think about women if you are being dismissive to their mother?”

That same month, A.J. finished his classes at the center. When we spoke, he was eager to catalog the “idiotic impulses” he’d left behind. He said he was no longer interested in acting like “a monarch or tyrant.” At other moments, he criticized Faith and expressed incredulity that online harassment had landed him with a misdemeanor, though he had in fact pled guilty. Moments later, as if catching himself, he’d speak again of his own stupidity.

“I was raised by women,” A.J. told me. “I would never want women to feel insecure or in danger while in my presence.” He paused. “I would never want my daughter to think that’s acceptable behavior.”

Once, I saw Cheryl Hurst ask a man what would happen if he extended the listening skills he declared he knew he must assimilate into his parenting to his children’s mother. A.J., it seemed, was still working through something similar: how to apply the empathy he offered his daughter—and knew she deserved—to intimate relationships. He said he was not yet ready to date again. He probably would not be for some time.

The last time we spoke, A.J. was somber. He had hoped to become a firefighter, yet his history of domestic violence disqualified him. In New York City, misdemeanors can make individuals ineligible for subsidized Section 8 housing, and he was worried about paying his rent. He feared that if he didn’t find a job soon, he wouldn’t be able to find a shelter that would permit him to bring Nala. A.J.’s abusive behavior had been unacceptable regardless of the circumstances, but the fallout from his entrance into the criminal justice system also situates him in a risk pool for reoffense. Men who are unemployed are more likely to be arrested for reassault than their employed peers, and eviction has been correlated with severe physical intimate partner violence perpetration. Had this punishment facilitated the safety of him, his family, or his community? I wondered.

Once, A.J. had told me that the emotion word he’d learned in class that lit him with self-recognition was “melancholy,” and he did seem sad that night in April when we spoke. I felt for A.J. too, but, I told him, I also had to ask again about that day in the hospital during the summer of 2018.

At first, A.J. denied that he had struck his estranged wife with an umbrella. Then he reversed his position. He had, he said, but not intentionally. In the midst of the altercation, he had yanked an umbrella from a hook and, somewhere in the confusion of it all, the umbrella made contact.

Days after our conversation, A.J. was scheduled to receive the final decision in the hospital case. But he failed to appear for his court date. A.J. had devoted twenty-one weeks to abusive partner intervention to satisfy the terms of a favorable sentence. Now, the judge had issued a bench warrant for him.

I was uncertain about how to reconcile this outcome with the moments in which he had demonstrated accountability. A.J., too, faced many uncertainties: how he’d manage his relationship with his estranged wife, how they would co-parent, how he would sustain the changes he’d made and make more. There were also uncertainties for the Family Wellness Program, which has only survived by cobbling together a budget made up of private donations and limited government grant money. At the time of this writing, program director Nazy Kaffashan was applying to renew grant funding for the Family Wellness Program. If denied, she would need to cut some abusive partner intervention programming.

At the end of every class cycle, the Family Wellness Program records exit interviews with the graduates. It is a chance to meditate on what they’ve learned and what they will need to do in the future. The men are consistently remorseful for how their own acts of domestic violence have affected their children, though some need to be reminded that their progress in class does not entitle them to forgiveness and that they must continue to work on their empathy for the women they’ve harmed. It is not uncommon for fathers to relate that they enjoyed being “challenged” to be better men, better partners, to incorporate new relational idioms. Often, they express a wish to stay in class for longer, to learn more about healthy relationships and good parenting.

“I need more information,” one program graduate said. “I do. It’s a hunger for it. Where am I going to get this information? I know I need to continue.”