Domestic Abuse in the Family Court: What the 87% and 4% Findings Really Mean

1. What the researchers actually examined

The academic paper draws upon the Domestic Abuse Commissioner’s Domestic Abuse Commissioner’s Everyday Business report, published in October 2025. The research examined three court sites in England and Wales operating under the conventional Child Arrangements Programme (CAP). It was designed to establish a detailed baseline and to test a method for monitoring how child arrangements proceedings respond to domestic abuse.

• 298 case files, representing proceedings concluded during 2023 and reviewed across three different courts.
• 95 observed hearings in 2024, including different procedural stages, contested hearings and judicial levels.
• Survivor participation: six focus groups involving 35 female survivors, plus one interview with a male survivor.
• 22 professional interviews with judges, magistrates and Cafcass/Cafcass Cymru family court advisers.

This matters because the researchers could compare court files with observed hearings and survivors’ experiences. In an ordinary single hearing, abuse may barely be mentioned even when it appears repeatedly in the history. A case file may record a professional assessment but fail to show how the court dealt with it. Comparing sources gives a fuller picture of systemic practice.

These were not all courts in England and Wales, and this was not a national prevalence survey. The files largely preceded the January 2026 amendments to PD12J and the latest Child Focused Courts rollout. The findings should therefore be treated as serious research evidence about the studied process, not as unqualified statistics about every family court today.

2. The 87%, 4% and 62% findings: what they prove, and what they do not

The authors also report that in 36% of cases in which domestic abuse was raised, the initial safeguarding letter either did not mention it or regarded it as irrelevant. In observed hearings in which domestic abuse was identified, special measures were in place in 53%. The statistics describe different samples and stages; they should not be casually combined into a single number of children harmed.

It is tempting to present 87% and 4% as a direct proof of widespread unlawful refusal to find facts. That would overstate the research and the law. The more defensible argument is that a system handling abuse-related issues this frequently needs transparent, rigorous and child-focused reasons for how it determines which disputes require further investigation.

3. The hidden assumption behind the adversarial contest

Adversarialism assumes that each party can put forward their case, obtain and present evidence, challenge the other account and persuade an independent decision-maker. Family proceedings already include many departures from that model: judges actively manage cases, Cafcass contributes to welfare analysis, and the child is a central person whose interests cannot be reduced to either parent’s position.

Yet the underlying contest remains powerful. One parent may have financial resources, specialist legal assistance and control of relevant records. Another may be unrepresented, traumatised, disabled, frightened of the other party or unfamiliar with basic litigation vocabulary. Equal speaking time does not automatically produce effective equality of participation.

The study identifies difficulties both where neither party has representation and where one party is represented but the other is not. Even where both have lawyers, unequal resources, differing advocacy quality and a strong professional culture of encouraging settlement can shape what is raised and what is quietly left untested.

A litigant in person may not understand the distinction between a position statement and evidence; between a document existing and the court having lawful access to it; between a police outcome and a judicial finding; or between saying a welfare report is inaccurate and showing the exact material error. Those are not matters of intelligence or good faith. They are technical features of court procedure.

The concern is not that judges should believe one party because they describe themselves as a survivor. Procedural fairness remains essential for both parents. The concern is that a test of evidence cannot be fair if the system assumes both parties have the resources and ability to participate on equal terms when the facts demonstrate otherwise.

4. Agreement is a resolution method. It is not, by itself, a risk assessment

The Family Court is right to encourage appropriate agreement. Unnecessary contested litigation can cause delay, expense and harm to children. But an agreement reached in the shadow of an intimidating or unaffordable hearing is not necessarily evidence that the underlying safeguarding concerns have disappeared.

Burton and Hunter describe how an insistence on examining allegations may be interpreted as refusing to move on from parental conflict. A person who has experienced controlling behaviour may feel compelled to choose between a painful contested hearing and an arrangement that does not fully address risk. The dynamic becomes particularly concerning where the alleged behaviour continues through communication, handovers, financial pressure, surveillance or repeated applications after separation.

A parent’s apparent consent should not be treated as proof that abuse has occurred. Neither should it be treated as proof that abuse did not occur. The court must still consider what is known, what remains disputed, the practical implications for the child and whether interim or final arrangements can be made safely.

Care is also needed with the term “high conflict”. Both parents may disagree strongly about arrangements, but coercive control is not the same thing as mutual poor communication. Equally, a contested allegation is not automatically established because it is described as coercive control. The court must examine the pattern alleged, the evidence, the other party’s response and the child’s welfare.

5. Why coercive control cannot be understood as five disconnected incidents

Coercive and controlling behaviour may be cumulative. A single message, complaint, financial restriction or disputed handover can appear minor when extracted from its context. A persistent pattern of intimidation, monitoring, isolation, economic restriction or interference with family relationships may have a different significance when events are considered together.

The Court of Appeal in Re H-N [2021] EWCA Civ 448 recognised the need to assess patterns of behaviour and the limitations of traditional “Scott Schedules” for that purpose. The approach was reinforced through the January 2026 amendments to Practice Direction 12J, which removed the older term in the relevant provisions and now refers to a schedule or table capable of showing specific incidents and examples of alleged patterns, enabling holistic consideration.

Changing the template label is not enough. A schedule can help identify precise issues for trial and allow a fair response, but it cannot carry the whole context of a complex alleged pattern by itself. Depending on the court’s directions, a focused narrative statement, accurate chronology, original messages, professional records and contextual exhibits may be necessary.

This is not permission to file unlimited material or relitigate every unhappy episode of a relationship. The court is determining what is relevant to the child and what requires proof. Precision makes patterns easier to understand: identify the period, the recurring behaviour, its progression, the child’s exposure or impact, and why it informs future arrangements.

The correct objective is not maximum volume. It is evidential clarity. A small number of properly evidenced, material examples may reveal a pattern more effectively than an unwieldy collection of screenshots without dates, context or an identified welfare purpose.

6. Special measures may change the room without changing the questions

Family Procedure Rules Part 3A and Practice Direction 3AA require consideration of vulnerability and participation. Depending on the circumstances, measures may include screens, remote evidence, structured questioning, managed arrivals or departures, breaks and other arrangements to enable a person to take part and give their best evidence.

The study observed improvements in access to special measures, but also their limitations. A screen may reduce direct eye contact while leaving a survivor exposed to intimidating questioning, isolation from their own representative or unsafe proximity in the waiting areas. A procedural arrangement can look adequate on paper and feel entirely different to the person required to use it.

Burton and Hunter argue that adjustments to the mode of giving evidence do not necessarily reform the substance of the questions. Repetitive challenges focused on trivial differences, stereotyped assumptions or hostile framing may prevent a coherent narrative from emerging. That is not an argument against testing contested evidence; it is an argument for relevant, proportionate, well-managed testing that does not confuse distress or minor inconsistency with an automatic credibility finding.

The Family Court also has statutory restrictions on direct cross-examination by a party in specified circumstances. A Qualified Legal Representative (QLR) can be appointed for questioning where the legal conditions are met. But a QLR does not become that party’s solicitor throughout the proceedings, prepare their whole case, or guarantee effective overall participation.

Practical consequence: participation requests should be raised early, identify the actual barrier to taking part and propose workable measures. Neither requesting special measures nor declining them proves the underlying disputed allegations.

7. The law as at November 2026: welfare, relevance, necessity and fairness

The welfare principle remains paramount

Section 1(1) of the Children Act 1989 makes the child’s welfare the court’s paramount consideration in decisions about upbringing. The welfare checklist includes the child’s wishes and feelings, physical and emotional needs, the effects of change and harm suffered or at risk of being suffered. The question is not which parent deserves contact; it is which arrangements best serve this particular child.

As at 4 November 2026, the qualified statutory presumption of parental involvement in section 1(2A) remains in force. It does not require a particular division of time or justify exposure to harm. The Courts and Tribunals Bill proposes repeal; a proposed change in a Bill must not be presented as enacted law.

Domestic abuse does not automatically require a separate hearing

Practice Direction 12J governs the approach where domestic abuse is raised in children proceedings. The court must identify whether allegations are relevant to welfare and whether factual determination is necessary. In Re K [2022] EWCA Civ 468, the Court of Appeal stressed relevance, the sufficiency of other evidence, necessity and proportionality. A fact-finding hearing is not a general inquiry into all past grievances.

But where the court decides no fact-finding hearing is necessary, paragraph 18 of PD12J requires its order to record the reasons. That matters in a system where a later welfare decision may otherwise rely on assumptions or an incomplete account of what was disputed.

An unproved allegation is not a finding that nothing happened

A disputed allegation not proved on the balance of probabilities cannot be silently converted into an established fact for the welfare decision. Equally, “not proved”, “not investigated”, “withdrawn” and “found not to have happened” are not identical descriptions. It is important to identify precisely what the judge determined and what was left unresolved.

Other established facts, admissions, existing orders and current independently identified risk factors may still need assessment. Police no-further-action decisions do not in themselves prove that abuse did not occur; equally, a recorded report to police does not establish the allegation as true. A Cafcass or section 7 report must be evaluated against what was actually available to its author and the findings, if any, that bind the welfare analysis.

Protection is not permission to bypass fairness

The other party must have a fair opportunity to know the material case against them and respond. The court has to distinguish evidence from submissions and ensure proportionate directions. The objective is a reliable evidential foundation for safeguarding decisions, not an assumption that one party is credible based solely on status or presentation.

8. Child Focused Courts: better investigation at the beginning, but what at the hearing?

The Family Court’s Child Focused Court model, formerly called Pathfinder, seeks to move away from the conventional adversarial structure towards early information-gathering, multi-agency working and attention to the child’s lived experience. The Child Focused Courts Target Operating Model and delivery update, published on 28 October 2026, describe phased implementation with intended nationwide coverage by March 2029. The model is not yet universal.

In Re C [2026] EWCA Civ 1216, the Court of Appeal recognised the importance of the new investigative, child-focused procedure while confirming that the core legal obligations remain unchanged. The child’s welfare stays paramount; PD12J remains applicable; and a Child Impact Report or enhanced early information gathering does not remove the need for proper decisions on relevant facts and risk.

Burton and Hunter identify a potential fault line: a more investigative front end may produce a richer picture of the family, but contested hearings can still be conducted through traditional adversarial questioning and binary findings. A genuinely investigative model would need to carry child-focused inquiry into the actual hearing, while preserving judicial independence, fair testing of contested evidence and the opportunity for both parties to be heard.

The Government’s June 2026 response to Everyday Business accepted eight recommendations fully and five partly, while six were not taken forward. It committed £500,000 across two years to further independent monitoring, focused on the Child Focused Court model. These are identifiable reform steps, not proof that the underlying structural difficulties have been resolved.

9. A practical approach for litigants in person

It is not the responsibility of an individual parent to repair the justice system. It is, however, useful to present information in a way that makes the legal and safeguarding questions clear. Start with the latest sealed order and the actual question the judge must decide, not with a general account of why the proceedings feel unfair.

1. Identify the decision. Is the next hearing about fact-finding, interim contact, a section 7 report, enforcement, variation or a final child arrangements order? Your documents should serve that decision.
2. Preserve original sources. Keep original messages, files, dates, attachments and context. Work from copies; do not manipulate evidence, ask a child to investigate or share confidential family-court materials inappropriately.
3. Separate evidence statuses. Label allegations, admissions, earlier judicial findings, professional opinions, police outcomes and factual matters still genuinely unknown. Do not turn a source into something it does not prove.
4. Explain the alleged pattern. If coercive control is raised, record the period, repeated conduct, escalation, post-separation features and material examples, rather than only a table of disconnected episodes.
5. Show the welfare link. For each issue, explain why it may matter to the child’s safety, emotional wellbeing, expressed wishes, contact, parental functioning or practical arrangements.
6. Analyse reports precisely. If Cafcass or a local authority treats abuse as irrelevant, identify the exact passage, the missing evidence, the contradictory record and the possible welfare consequence. Precision is stronger than an unsupported accusation of bias.
7. Seek proportionate directions. Depending on the case, consider whether lawful third-party disclosure, a clarification, written reasons, a focused factual determination or safe interim arrangements are necessary. Ask for a specific and workable action.
8. Protect participation. Raise any need for safety arrangements, adaptations or restrictions on questioning promptly. Do not wait until the witness box to explain that you cannot participate fairly.
9. Check representation options. Where domestic abuse or serious child-safety concerns arise, check legal-aid eligibility and seek regulated legal advice, particularly for urgent protective measures, complex hearings or appeals.

A simple issue-to-evidence table

Illustrative framework, not a prescribed court form and not a recommendation that the directions suit every case.

A measured way to put the issue to the court

“I ask the court to consider whether the matters identified in paragraphs [X–Y] are material to the child’s welfare and risk assessment. The allegations remain disputed, and I have identified the supporting and contrary sources. I ask the court to determine whether further fact-finding is necessary and proportionate, or to record its reasons for concluding otherwise. I also request that appropriate safeguarding and participation arrangements be considered.”

Any proposed wording must be adapted to the actual order, evidence, hearing and legal advice. It is not a substitute for an application where the rules require one, and it does not suspend existing deadlines or child arrangements.

10. A meaningful test of reform: six questions for the system

At JSH Law, we consider that a genuinely child-focused process should be capable of demonstrating the following, not merely asserting it:

1. Recognition: Did the court and safeguarding agencies identify abuse-related concerns early, including alleged patterns, post-separation conduct and economic or digital coercion?
2. Evidence continuity: Were relevant existing findings, police, local authority, health and school records identified and obtained through proper lawful routes, without conflating different evidential statuses?
3. Effective participation: Could each parent understand the case, respond fairly and participate without avoidable intimidation? Were differences in legal resources and vulnerabilities recognised?
4. The child’s voice: Was the child’s perspective obtained sensitively and assessed as part of welfare, rather than mechanically used to validate an adult’s preferred narrative?
5. Reasoned decisions: Are the reasons for treating abuse as relevant or not relevant, and for directing or declining fact-finding, identifiable in orders and judgments?
6. Real-world safety: Is there a meaningful way to evaluate whether orders remain safe, workable and protective from continuing abuse?

An investigative court is not automatically a fairer court. Too much untested power in one set of hands creates its own risks. Fairness requires reliable evidence, disclosure of material points, opportunities to challenge contested accounts and independent reasoning. Child-focused safeguards and procedural fairness should reinforce one another.

The essential issue is this: the Family Court makes decisions about a child’s future. It cannot do so responsibly if the process itself prevents it from understanding the relevant history, the present circumstances and the risks that may follow from its order.

11. Frequently asked questions

Does the research say domestic abuse was proved in 87% of child arrangements cases?

No. The 87% figure refers to domestic abuse being raised in the sampled files. It is not a national statistic of judicial findings. Nevertheless, it demonstrates that abuse-related issues were common in the courts studied and should not be treated as exceptional administrative distractions.

If the court did not hold a fact-finding hearing, has it breached PD12J?

Not necessarily. A hearing depends on what facts are material, whether they genuinely need resolution, what other evidence exists and proportionality. The court must nevertheless apply PD12J and record reasons where it considers fact-finding unnecessary. The particular order and evidence must be reviewed before any legal error can be asserted.

Can domestic abuse matter if a child did not witness an incident?

Yes. Harm can be indirect as well as direct, and abuse may affect the child’s emotional welfare, the safety of the caregiving parent and future arrangements. But relevance and the evidential basis must be assessed in the individual case rather than presumed.

Are Scott Schedules now banned?

No. The current language of PD12J has changed, and the court should consider how a schedule or table can reflect a pattern of coercive control. A court may still direct a table to identify the issues. Read the actual directions before filing any alternative document.

Does the Child Focused Court model remove the need for PD12J?

No. The Court of Appeal made that clear in Re C [2026] EWCA Civ 1216. The new process changes how information is gathered and cases are managed; it does not abolish the statutory welfare requirement or the safeguards relating to domestic abuse.

What if I cannot afford representation?

Check whether you qualify for legal aid and whether urgent regulated advice is needed. Defined-scope document preparation and McKenzie Friend support may assist you to organise evidence and understand procedure, but it is not a substitute for authorised legal representation in matters requiring reserved legal activity.

12. Original sources, linked guidance and research credit

Academic research: Mandy Burton and Rosemary Hunter, original open-access Modern Law Review paper, The Modern Law Review (first published 21 October 2026), DOI: 10.1111/1468-2230.70076. This is the original open-access paper discussed here. The underpinning research was commissioned by and carried out with the Office of the Domestic Abuse Commissioner.

Intelligence acknowledgment: Hana Sakkal drew attention to the research through her LinkedIn commentary. We credit her for identifying and sharing its significance. The statistics and substantive academic analysis are credited to Burton and Hunter and the Domestic Abuse Commissioner’s research programme, not attributed to Hana as original data.

• Domestic Abuse Commissioner’s Everyday Business report
• Ministry of Justice response to Everyday Business, 10 June 2026
• Children Act 1989, section 1
• Practice Direction 12J and Practice Direction 3AA
• Re H-N [2021] EWCA Civ 448; Re K [2022] EWCA Civ 468; Re C [2026] EWCA Civ 1216
• Child Focused Courts Target Operating Model (28 October 2026) and delivery update
• Courts and Tribunals Bill: parliamentary progress

Related JSH Law guides

• Domestic Abuse Evidence & PD12J Support
• Schedules of Allegations and Patterns of Coercive Control
• Re C: The Obligations Are Unchanged
• Child Focused Courts: What the New Process Means

General-information and regulatory notice: This article is research commentary and general legal information for England and Wales, not individual legal advice. JSH Law Ltd is not an SRA-authorised solicitors’ firm. It provides non-reserved litigation support, legal information, document-preparation assistance and McKenzie Friend support. It does not undertake reserved conduct of litigation or exercise advocacy rights as of right. Court outcomes depend on facts, evidence, directions and judicial discretion. For urgent danger or safeguarding concerns, seek appropriate specialist or emergency assistance.