After Domestic Abuse Is Proved, Who Should Have to Prove Contact Is Safe?

JSH LAW · COMPARATIVE FAMILY JUSTICE
By Jessica Susan Hill · Reviewed 26 October 2026 · England and Wales unless stated otherwise

The hardest question in a domestic-abuse contact case is often framed as: Should this child see this parent? That question is important, but it starts too late. The prior question is more revealing: once domestic abuse has been proved, who must establish that future arrangements are safe?

Arizona has now given a clear answer. Under the Alec and Lydia Act, the burden shifts to the parent found to have committed domestic violence. It does not abolish contact. It does not treat an allegation as a finding. It does not replace judicial assessment with an automatic ban. What it does is refuse to make the child absorb the uncertainty created by established abuse.

This matters far beyond Arizona. Comparative law is not binding authority in the Family Court of England and Wales, and American terminology cannot simply be imported into our system. But a different legal model can expose choices within our own. Here, the law recognises domestic abuse, coercive control and harm to children. It can restrict or end contact. Yet, after abuse is found, there is still no equivalent statutory rule requiring the abusive parent to demonstrate that their proposed involvement is safe.

Why it is called the Alec and Lydia Act

The legislation bears the names of Alec and Lydia, aged seven and six. According to reporting by Arizona's Family, their father killed them during court-ordered unsupervised parenting time in May 2024 and then killed himself.

Their mother, Hope Hooton, subsequently campaigned for reform. The same reporting records her account that the family-court process had included information about domestic violence, strangulation, coercive behaviour, psychiatric instability and a protective order, but had not brought those indicators together into an adequate assessment of danger to the children. Those details are reported accounts, not findings made in a trial against a living defendant. That distinction should be maintained even where the outcome is devastating.

The Arizona Legislature approved House Bill 2995 as an emergency measure. It was signed and filed on 22 June 2026 and became operative immediately. Its formal short title is the Alec and Lydia Act.

What the Arizona law actually changes

The Act does not merely add domestic violence to a welfare checklist. It changes the architecture of decision-making after domestic violence is established.

1. The court must decide whether domestic violence occurred

The party alleging domestic violence must establish the claim by a preponderance of the evidence—broadly comparable to our balance of probabilities: more likely than not. The judge must evaluate the evidence admitted in the proceedings and make specific written findings.

The Act also makes clear that a victim's evidence does not require a supporting exhibit or a second witness before it is capable of being accepted. That does not mean an allegation is automatically true. It means testimony is evidence, and the court must assess it rather than impose an extra-statutory corroboration rule.

2. A mandatory rebuttable presumption follows the finding

Once the court determines that a parent committed domestic violence, it must apply a rebuttable presumption that awarding that parent legal decision-making authority or parenting time is contrary to the child's best interests. In English terms, “legal decision-making” is closer to authority over major decisions; “parenting time” is closer to time or contact. The concepts are not exact equivalents.

This is a burden-shifting model. Before a finding, the person asserting abuse must prove it. After a finding, the person who committed the abuse must produce sufficiently strong evidence that the proposed arrangement is safe and consistent with the child's welfare.

3. The court cannot demand proof that the abuse was “bad enough”

The Act prevents the court from requiring the victim to prove that domestic violence was sufficiently severe or frequent to justify restrictions. Severity and frequency remain relevant to whether the presumption is rebutted, but they are not an additional gateway the victim must cross after domestic violence has been established.

It also addresses the dangerously flattening label of “mutual abuse”. Conduct is not to be treated as mutual unless both parents lacked justification to a similar degree and caused similar injury. Where the conduct and motivation are materially unequal, the presumption applies to the parent responsible for the more serious violence. That matters because resistance, self-protection and survival behaviour are not analytically identical to coercive abuse.

4. Coercive control includes conduct inside and outside the home

The statutory definition reaches far beyond physical assault. It includes patterns such as isolation, financial monitoring, stalking, degrading behaviour, threats involving immigration status, surveillance or tracking technology, property damage and threats to disclose sensitive information.

Notably, it also recognises abuse through legal process: false or frivolous civil litigation, or litigation used to restrict another person's lawful activity, can form part of coercive control. That is not a licence to call every unsuccessful application abusive. It is recognition that procedure itself can become an instrument of domination when the evidence establishes that pattern.

5. “The child did not see it” is not an answer

Three matters cannot be used to rebut the Arizona presumption:

• the child was not present when the abuse occurred;
• the child did not know about the abuse; or
• the child says they want contact with the abusive parent.

The precision here matters. The child's wishes are not declared irrelevant to the whole welfare analysis. They are declared insufficient to rebut the safety presumption. A child may love a parent, miss a parent and still be exposed to an unacceptable risk. Attachment is evidence of relationship; it is not proof of safety.

6. Reasons and rehabilitation must be evidence-based

Judges must explain in writing what evidence did—or did not—justify the domestic-violence finding, whether the presumption was rebutted, and why the safeguards ordered protect the child. The Act also tightens temporary-order reasoning where domestic violence is alleged.

A course-completion certificate is not enough, by itself, to prove rehabilitation. The court may examine whether the intervention was relevant and proportionate, whether records have been released, whether the parent understands the impact on the family, and whether denial, minimisation or rationalisation has genuinely reduced.

Available protections include safe handovers, professional or carefully safeguarded supervision, restrictions on overnight time, testing and treatment requirements, confidential addresses, payment of supervision costs by the abusive parent and suspension of access unless and until the order is varied. A victim cannot be compelled into joint counselling or assessment with the perpetrator. If the presumption is not overcome, it continues into later variation proceedings.

What the law already says in England and Wales

Our law is not silent about domestic abuse. The difficulty is the relationship between its principles, its presumptions and what happens in practice.

The child's welfare is paramount

Under section 1 of the Children Act 1989, the child's welfare is the court's paramount consideration when determining upbringing. The welfare checklist includes the child's wishes and feelings, needs, the likely effect of change, relevant characteristics, harm suffered or risked, and each relevant adult's capability.

The same section currently contains a rebuttable presumption that the involvement of each parent will further the child's welfare, unless the statutory risk condition is engaged. “Involvement” can be direct or indirect and does not prescribe equal time or any particular division of the child's life. It is therefore wrong to describe section 1(2A) as a legal entitlement to 50/50 care.

A child can be a victim without witnessing an assault

Section 3 of the Domestic Abuse Act 2021 recognises a child as a victim where the child sees, hears or experiences the effects of domestic abuse and is related to either party. The statutory language matters: harm is not confined to watching a violent incident.

Practice Direction 12J goes further in operational terms. It requires courts to identify domestic-abuse issues, determine disputed facts where necessary, consider patterns rather than isolated incidents, protect the child and resident parent, and ensure that any child-arrangements order does not expose either of them to an unmanageable risk of harm.

Where domestic abuse is proved and contact is ordered, the court must explain why the order will not expose the child to harm and why it is beneficial. The court must also consider whether the proceedings themselves are being used to continue abuse and whether the abusive parent appreciates the impact of past and possible future conduct.

In Re H-N and Others (Children) (Domestic Abuse: Finding of Fact Hearings) [2021] EWCA Civ 448, the Court of Appeal emphasised the importance of identifying the nature of the alleged abuse and its relevance to future parenting. Coercive and controlling behaviour often derives its significance from a pattern; atomising the case into disconnected incidents can conceal the very risk the court needs to understand.

Re CX: our courts can recognise indirect pathways of harm

A powerful recent example is Re CX (No 2) (Domestic Abuse: No Contact) [2026] EWFC 230, decided in August 2026.

The court had previously found that the father subjected the mother to extreme emotional and psychological abuse over a period of years. The child loved the father and had experienced approximately 200 hours of positive supervised contact. Taken in isolation, those observations might have supported continued contact.

But the court did not examine contact in isolation. Expert evidence established that continuing contact and litigation triggered the mother's complex post-traumatic stress disorder, prevented recovery and threatened her long-term capacity to care for the child. The risk to the child operated through the serious injury being perpetuated against the child's primary carer. After alternatives had been tried, the court ended direct and indirect contact and made a three-year section 91(14) order restricting further applications without permission.

This was an exceptional, evidence-rich decision—not a general rule. Its importance lies elsewhere. It demonstrates that:

• a child's enjoyment of supervised time does not answer every welfare question;
• harm to a protective parent can be a foreseeable pathway of harm to the child;
• the absence of poor behaviour during a contact session does not erase an established pattern outside it; and
• no-contact orders remain available where the evidence shows they are necessary, proportionate and the least harmful realistic option.

Re CX shows what the present law is capable of doing. It does not prove that every family has equal access to the expert evidence, continuity, time and judicial analysis that made that outcome possible.

Why implementation matters as much as the wording

The Ministry of Justice’s parental-involvement review and the Domestic Abuse Commissioner’s Everyday Business report, published in October 2025, examine the gap between formal safeguards and practice. Their findings support scrutiny of how domestic abuse, children’s voices and risk are considered. They are not proof that every disputed case has the same features or that every contact order is unsafe.

A finding that abuse occurred, the evidence about current risk, the available safeguards and the reasons for a proposed order all need to remain visible. Changing statutory language cannot, by itself, supply missing professional resources or improve the quality of every welfare assessment.

Arizona and England & Wales: the real comparison

Repealing the parental-involvement presumption is not the same reform

The Government has proposed removing the presumption of parental involvement. On the Parliament bill record checked on 26 October 2026, the Courts and Tribunals Bill had reached the House of Lords. A proposal before Parliament is not an enacted and commenced repeal. The current section 1 text must be distinguished from the proposed reform.

Repeal would be significant. It would remove a statutory instruction favouring some form of parental involvement and return the inquiry to an open application of welfare and risk. But it would not reproduce the Arizona model.

There is a decisive difference between:

1. removing a presumption in favour of involvement, which creates a neutral starting point; and
2. creating a presumption against unsafe involvement after proved abuse, which reallocates the burden created by a finding.

Neutrality may be better than a pro-involvement default. It does not, on its own, answer who must demonstrate safety once abuse has been judicially established.

What a principled reform could look like here

A serious proposal for England and Wales should not copy and paste an American statute. Our evidential law, terminology, human-rights framework and family procedure are different.

In particular, Arizona's “clear and convincing evidence” test does not map neatly onto English family law. Re B (Children) [2008] UKHL 35 confirms that the standard for findings in Children Act cases is the balance of probabilities; seriousness does not create a floating intermediate standard. Any domestic reform must be drafted with that principle in mind.

But the underlying architecture deserves proper debate. A proportionate model could include:

1. A judicial-finding gateway. No burden shift on allegation alone. The parties must have notice of the case, a fair opportunity to challenge it and a reasoned decision distinguishing admissions, evidence, professional opinion and findings.
2. A rebuttable safety presumption. Following relevant findings of domestic abuse, the starting point could be against shared decision-making and unsupervised direct contact unless the court is satisfied that the arrangement is safe and advances welfare.
3. Individualised proportionality. A finding should not produce an automatic, permanent no-contact order. The court should consider the nature, pattern, recency and parenting relevance of the abuse, together with the child's needs and all realistic safeguards.
4. Evidence of insight and change. Attendance is not transformation. The relevant questions are whether responsibility is accepted, risk mechanisms are understood, learning is applied and sustained behaviour supports the claim of change.
5. Mandatory written reasoning. The judgment should connect the facts found to the pathway of harm, the proposed arrangement, the safeguards, the child's wishes and the court's conclusion. “Contact is the norm” is not a welfare analysis.
6. Recognition of asymmetry. Courts must distinguish coercive domination from resistance, self-defence or situational conflict. Labels such as “high conflict” and “mutual hostility” should never replace findings where the evidence permits a more precise analysis.
7. Protection from procedural abuse. Repeat applications, unnecessary disclosure demands, oppressive communications and litigation used as surveillance or punishment should be identified as possible continuations of the established pattern.
8. A properly resourced route to safe involvement. Where supervised, supported or indirect contact may genuinely benefit the child, the necessary services must exist. A legal safeguard that families cannot access is not a functioning safeguard.
9. Transparent monitoring. Outcomes should be audited by reference to findings, risk factors, orders and reasons—not by crude measures of whether “contact happened”.

Any interference with family life must remain lawful, necessary and proportionate. That is precisely why a rebuttable, evidence-based rule is more defensible than an automatic ban. Fairness belongs to every party. But procedural fairness does not require the court to pretend that the risk landscape is unchanged after abuse has been proved.

For litigants in person: turning abuse evidence into a welfare case

A family court cannot safely analyse what has not been clearly identified. Volume is not the same as proof, and a history of abuse should not be reduced to a collection of unexplained screenshots. The task is to show, with precision, what is alleged, what is evidenced, what has already been found and why it matters to this child's future.

This is also why language matters. “He abused the mother, not the child” is not a complete legal analysis. Neither is “the child loves him”, “contact went well” or “she completed the course”. Each may be relevant evidence. None resolves the central question without examining the pattern, the pathway of harm and the safeguards required.

The question our system must now answer

The Alec and Lydia Act is not perfect simply because it is new, and it is not a template to be imported without scrutiny. Courts will still have to interpret its thresholds, apply its safeguards consistently and protect due process. Evidence of its operation over time will matter.

But its moral and legal premise is difficult to dismiss: after domestic violence is established, the legal system should not behave as though the parties return to a neutral evidential landscape. A finding changes what is known. The law should change who must answer the remaining uncertainty.

England and Wales already has the vocabulary of welfare, coercive control, indirect harm and procedural abuse. It has judges capable of rigorous decisions, as Re CX demonstrates. What it lacks is the same clear statutory allocation of risk after findings are made.

Repealing the presumption of parental involvement would remove one structural obstacle. The next question is harder and more important: when abuse has been proved, should the child and protective parent continue to prove danger—or should the parent responsible be required to demonstrate safety?

Arizona has answered. We should at least be brave enough to ask.

General information and commentary, not advice on an individual case. JSH Law provides non-reserved litigation support, document-preparation assistance and McKenzie Friend support, subject to the court’s permission where required. It is not an SRA-authorised solicitors’ firm and does not conduct litigation or exercise reserved rights of audience on a client’s behalf. Obtain appropriately regulated legal advice where needed. Comparative foreign law is not binding in England and Wales.