There is an uncomfortable paradox at the heart of child protection in cases involving domestic abuse. The state is increasingly willing to recognise the devastating impact that domestic abuse has upon children, yet the machinery intended to protect those children can, at times, turn its attention away from the perpetrator and towards the parent who was abused.

The language is familiar: failure to protect. Lack of insight. Poor protective capacity. Failure to prioritise the children. Failure to recognise risk. Failure to engage. Each phrase may appear, in isolation, to be a legitimate assessment of parenting. But when they are applied to a non-abusive parent without properly accounting for coercive control, trauma, fear, dependency and the realities of leaving an abusive relationship, safeguarding can become indistinguishable from victim blaming. This is not merely a question of semantics. It is a question of who the child protection system considers responsible for the violence.

Of course children must be safeguarded from domestic abuse. A child who witnesses violence or lives within coercive control is a child who requires protection. The question is not whether intervention is necessary, but what that intervention should look like. If the mother is the non-abusive parent, why does intervention so readily become an assessment of her "protective capacity", rather than an assessment of what support is required to make her and her children safe? Why is the answer so often a list of courses, assessments and expectations, followed by criticism when the mother does not recover quickly enough, rather than meaningful investment in the consequences of the abuse?

The Psychology of the Traumatised

The psychological evidence is important here. Research consistently demonstrates that domestic abuse can have profound psychological consequences. In one study of 260 women accessing domestic-abuse services, more than three-quarters met the clinical threshold for PTSD and 76% met the threshold for significant psychological distress; the severity of psychological symptoms was associated with the severity of the abuse experienced. The point is not that every victim will develop PTSD, but that trauma is not an incidental consequence of abuse. It can fundamentally affect how a person thinks, feels, regulates emotion, makes decisions and relates to other people. Treating those consequences as evidence of deficient parenting without first understanding their cause risks confusing the symptoms of victimisation with the characteristics of the victim.

Substance misuse illustrates this particularly well. Ogden, Dichter and Bazzi's 2022 systematic review examined research concerning intimate-partner violence and subsequent substance-use outcomes among women. The authors found evidence that IPV was associated with subsequent substance use and substance-use disorders and identified pathways involving coping and self-medication. Their review also found evidence that IPV can affect substance-use treatment outcomes, including treatment completion and relapse. Importantly, the authors did not claim that IPV inevitably causes substance misuse; rather, the evidence supported several possible pathways through which victimisation may contribute to later substance use.

That distinction matters enormously in child protection. If a traumatised woman drinks or takes drugs to numb psychological distress, the substance use may be dangerous and may require intervention, but it is not necessarily the beginning and end of the story. But the mother is rarely given the proper tools to address her coping mechanisms. The social worker tells her to "just stop using" as if they are entirely unaware of the meaning of an addiction. They tell her to attend a substance misuse programme. They will fund these services, but not the ones that truly matter. Therapy. Treating the substance while ignoring the trauma can leave the underlying problem intact. A domestic-abuse course may educate a woman about coercive control; it does not, by itself, treat PTSD, addiction, emotional dysregulation, financial dependency or the psychological consequences of prolonged violence. The same problem emerges when the woman returns to her social worker having completed the prescribed domestic-abuse work. She can now explain coercive control. She can identify the signs of abuse. She can articulate the risks. Yet suddenly the language changes: disguised compliance. She is only telling professionals what they want to hear. The learning has not been embedded. The standard becomes almost impossible to satisfy. If she cannot articulate the learning, she lacks insight. If she can articulate it, she may simply be repeating what she has been taught. If she makes progress, it is not yet sufficiently embedded. If she struggles, it confirms the original concerns.

There is, however, nothing unreasonable about professionals wanting to know whether change is sustainable. But there is a danger when "embedded" becomes a moving target: the woman engages with therapy, understands the abuse, completes the programme, changes her behaviour and demonstrates insight, yet the conclusion remains that the learning has not been sufficiently embedded. At what point is the victim permitted to have recovered? Trauma recovery is not a school examination in which a woman passes once she can repeat the correct terminology about domestic abuse. Psychological recovery is often uneven. A woman can understand intellectually that a relationship was abusive while still experiencing fear, shame, anxiety or difficulties trusting her own judgment. Continued therapeutic need is not synonymous with parental incapacity. 

Peele [2009] and Miller [1992] found that a one-size-fits-all approach in substance misuse treatment is inadequate.Tick-box exercises are rarely enough for sustained sobriety. In fact, just 49% of patients enter a substance misuse programme actually complete it. NICE recommends evidence-based interventions including contingency management, motivational approaches, psychological treatment and, where appropriate, residential treatment. The question in child-protection proceedings therefore should not simply be whether a mother has a history of substance misuse. It should be whether she has been given appropriate treatment, whether that treatment has addressed the reasons for the substance use, what evidence exists of her response to treatment, and what support is required to maintain that progress. A referral to a service is not treatment. Attendance at a course is not recovery. And a history of addiction is not, by itself, a prognosis of permanent parental incapacity. But Local Authorities refuse to fund actual treatment, yet will fund a foster placement.

Why Does She Stay?

Leaving an abusive relationship can be the most dangerous point in the relationship. Government statutory guidance expressly recognises that women are at greatest risk of homicide at separation or after leaving, because the post-separation period can involve escalation, stalking, harassment and attempts by perpetrators to regain control. Campbell et al.'s landmark 2003 multisite case-control study examined 220 intimate-partner femicide victims and 343 abused control women and found that separation after living together was associated with a substantially increased risk of femicide, with an adjusted odds ratio of 3.64. Where the perpetrator was highly controlling and the couple had separated, the risk was higher still.

The UK evidence is equally uncomfortable. Analysis published by the Office for National Statistics found that, among 68 women killed by an ex-partner or ex-spouse where the timing of separation was known, 49% were killed within one month of separation, 79% within six months and 90% within one year. Read that once more. In the U.S, women are murdered by partners more than by any other perpetrator.  The most recent ONS figures continue to demonstrate the intensely gendered nature of domestic homicide: in England and Wales in the year ending March 2025,  and all but four of women were killed by a male suspect. Of adult femicide victims overall, the most common suspect was a partner or ex-partner. Of domestic homicides, around 70-80% were female victims. This pattern emerged year after year after year.

These statistics fundamentally complicate the criticism that a woman should simply have left. A woman who remains with an abusive partner is not necessarily demonstrating that she does not understand the risk. She may understand it exceptionally well. She may know what happens when she attempts to leave. She may fear retaliation against herself or her children. She may have been financially controlled, isolated from family, threatened with losing her children or conditioned over years to believe that resistance will make the situation worse. Government guidance expressly recognises that victims may perceive remaining in the relationship as safer and that fear of escalation can be a significant barrier to leaving.

Walker originally developed her theory partly to challenge the assumption that women remain in violent relationships because of some inherent psychological defect or because they want the relationship. Her theory proposed that repeated violence can produce psychological responses associated with learned helplessness, while her cycle-of-violence theory described recurring phases of tension-building, acute violence and a period of reconciliation or relative calm. This is why the concept of "protective capacity" needs to be used carefully. A mother who returns to an abusive partner has not necessarily demonstrated an absence of maternal concern. She may have demonstrated the extraordinary difficulty of attempting to escape a coercive relationship while simultaneously trying to protect her children. That does not mean that professionals should ignore the risk. It means that the intervention should address the reason the risk remains. And yet, too often, the practical response appears to be remarkably narrow: attend a domestic-abuse course, engage with substance-misuse services, complete parenting work, demonstrate insight and show that the learning has become embedded. But the catch is - if you don't do it fast enough, your children will be put into care, or even adopted. There is nothing inherently wrong with any of these interventions. The problem is what happens when they are treated as sufficient. A woman experiencing PTSD may need psychological treatment. A woman using substances to cope with trauma may need integrated addiction and trauma services. A woman attempting to leave a controlling partner may need safe accommodation, financial support, specialist advocacy and an effective safety plan. A woman who has been repeatedly victimised may need long-term therapeutic and social support. The existence of a referral is not evidence that meaningful support has been provided. The point is not that victims require endless sympathy without accountability. Walker's work itself recognised the need for therapeutic intervention. The question is what kind of intervention is actually capable of addressing the psychological consequences of abuse. If a woman has developed PTSD, substance misuse, impaired self-efficacy or difficulties with interpersonal functioning as a consequence of prolonged victimisation, a short domestic-abuse awareness course cannot reasonably be treated as equivalent to psychological treatment addressing those difficulties.

The Domestic Abuse Commissioner's work has identified precisely this problem. Its research into children affected by domestic abuse found that non-abusive parents can be blamed for failing to protect children from perpetrators and that survivors reported a lack of empathy and increasing demands from social workers. It recommends moving away from the language of "failure to protect" towards a strengths-based approach which recognises the strategies victims use to manage, resist and survive abuse. That is a significant distinction: the question becomes what the victim has done to survive and what support she needs, rather than simply what she has failed to do.This also exposes a fundamental contradiction in the purpose of child protection. If the objective is, wherever safely possible, to enable children to remain within their families, then supporting the non-abusive parent should be central to safeguarding. It cannot be enough to identify that a mother is traumatised, addicted, frightened or psychologically affected by years of violence and then conclude that these characteristics make her unsuitable to care for her children. If those difficulties are capable of treatment, why is the emphasis not on providing that treatment?

Of course there are cases where a child cannot safely remain with a parent. There are mothers whose own conduct creates an unacceptable risk, and domestic abuse cannot be used as a universal explanation for every safeguarding failure. But there is a crucial distinction between removing a child because the parent is the source of the danger and removing a child because the parent has not, without adequate support, overcome the consequences of somebody else's violence. The latter should make us profoundly uncomfortable. We must ask has the Local Authority done their duty to exhaust their resources and support for the victimised parent before cutting family ties and potentially traumatising children even further by placing them into care?

The state knows that trauma can affect behaviour. It knows that substance use can be connected to attempts to cope with abuse. It knows that repeated victimisation is associated with previous trauma and other vulnerabilities. It knows that leaving can substantially increase the risk of serious violence and homicide. It knows that perpetrators can use children, finances, threats and the fear of social-services involvement as mechanisms of control. Yet the non-abusive mother can still be placed in the impossible position of having to prove that she has recovered sufficiently, quickly enough, to retain her children.That is not accountability. It is an expectation of invulnerability. The state should not require a traumatised woman to demonstrate that she will never again make a poor decision, never again trust the wrong person, never again relapse, never again experience psychological distress and never again become vulnerable to coercive control. No such standard is realistically achievable.

There is a fundamental contradiction at the heart of the child protection system's response to domestic abuse. The state increasingly recognises that domestic abuse is a form of harm to children, yet the response to that harm can still place the burden of remedying it upon the person who did not cause it. The mother becomes responsible for managing the perpetrator's violence, mitigating his behaviour, overcoming the psychological consequences of his abuse, addressing any resulting substance misuse, demonstrating insight into the risks and, ultimately, proving that she is sufficiently recovered to deserve to keep her children.

That is an extraordinary burden to place upon a victim.The issue is not whether mothers should be held accountable where their own conduct places children at risk. They should. Nor is it an argument against intervention, assessment or professional scrutiny. Children have a right to safety, and that right must remain paramount. But accountability must be directed towards the source of the risk. Where a mother's difficulties are themselves consequences of domestic abuse, the response cannot simply be to catalogue those difficulties as evidence of deficient parenting. Otherwise, the system risks transforming victimisation into culpability. The language of "failure to protect" is particularly revealing. It can imply that the mother possessed the power to prevent the abuse in circumstances where the perpetrator deliberately constructed a relationship in which she did not have that power. Coercive control is, by its nature, an erosion of autonomy. It operates through fear, dependency, isolation, intimidation and the systematic restriction of another person's choices. To subsequently assess that person as though she had been making free and unconstrained choices risks reproducing, within the child protection system, the very dynamics of control that characterised the abusive relationship.

This is also where the rhetoric of the child's "best interests" requires greater interrogation. Of course the child is paramount. But the child's interests are not necessarily served by treating the removal of a non-abusive mother as the default solution to the consequences of a perpetrator's abuse. Sometimes removal is necessary. Sometimes there is no safe alternative. But where a mother is capable of safely caring for her child with appropriate intervention, the question should be whether the state has done everything reasonably available to make that intervention possible before concluding that separation is necessary.

A domestic-abuse course cannot treat PTSD. A substance-misuse referral cannot, without more, address the trauma which may have contributed to that substance use. A woman completing a programme does not necessarily become psychologically invulnerable simply because she can now describe coercive control in the language professionals expect. And a relapse, a moment of fear or a failure to leave immediately cannot automatically erase every piece of progress that came before it.The distinction is therefore between expecting accountability and demanding invulnerability. The former is legitimate. The latter is neither realistic nor humane.

A genuinely child-centred approach would ask not merely what the mother has failed to do, but what the system has failed to provide. Has she received appropriate trauma-informed psychological treatment? Has her substance misuse been treated alongside its underlying causes? Has she been provided with meaningful assistance to leave safely? Has the perpetrator's behaviour been properly addressed? Has she been given the practical, financial and therapeutic resources necessary to protect herself and her children? Have less interventionist options genuinely been exhausted before family separation is contemplated? These are not questions designed to excuse parental shortcomings. They are questions designed to determine whether removal is actually necessary.

There is, ultimately, something deeply uncomfortable about a system that will spend substantial public resources undertaking the extensive machinery of care proceedings, removing a child from a traumatised mother, placing that child into foster care while being unwilling or unable to provide the mother with the intensive therapeutic and practical intervention that might have made separation unnecessary. If the state is serious when it says that children should, wherever safely possible, remain within their families, then supporting the non-abusive parent cannot be an optional extra. It must form part of the safeguarding response itself.

The feminist question is therefore not whether mothers should be above scrutiny. It is who is being scrutinised, for what, and why. When the perpetrator's violence becomes the backdrop against which a mother's psychological distress, addiction, fear or imperfect decision-making is assessed, responsibility has already begun to shift. When those consequences are then relied upon to justify separating her from her children, the system risks imposing a second injustice upon the original victimisation.

Children should be protected from domestic abuse. But protecting children does not necessarily mean protecting them from their mothers. Sometimes protecting a child means protecting the mother from the perpetrator. Sometimes it means treating the trauma rather than punishing its symptoms. Sometimes it means providing the resources necessary for a family to become safe rather than removing the child because the family was not made safe quickly enough.

The question should therefore not simply be: "Has she done enough?". It should also be: "Have we done enough?"

Because if the state has identified the violence, identified the trauma, identified the addiction, identified the fear and identified the risks, but has failed to provide the intervention capable of addressing them, it cannot then regard the consequences of those failures as evidence that the mother is incapable of parenting. A child-centred system must be more than a system capable of removing children from unsafe circumstances. It must also be a system capable of making families safe.

And if the state genuinely believes that the child is paramount, then it must be willing to invest in keeping children safely with the parent who was trying to protect them, rather than making that parent pay the price for the violence of someone else.