Adoption Proportionality Requires Child-Specific Harm Consequences and Consideration of DAPO/Supervision Alternatives

1. Introduction

In F (Children: Placement Orders: Proportionality) [2026] EWCA Civ 990, the Court of Appeal (Peter Jackson LJ giving the lead judgment, with Dingemans LJ and Edis LJ agreeing) allowed the mother’s appeal against care and placement orders made by the Family Court at Cardiff.

The case concerned two sisters: D (aged 4½) and Z (aged 1½). The children had always lived with their mother and were said to be thriving in her supervised care within a residential assessment setting. The local authority sought (and obtained) a plan of adoption, despite there being no interim removal during the year-long proceedings and despite the children’s strong attachment to their mother.

Two features made the welfare evaluation unusually complex: (i) the parents’ precarious immigration status, and (ii) the reality that the children’s lived experience was of stable day-to-day care by the mother under constant professional oversight, while the risk case was substantially future-facing (risk of recurrence of domestic abuse and further injury, compounded by parental dishonesty and lack of insight).

The central appellate issue was proportionality: whether adoption was truly necessary—“where nothing else will do”—given the harm the children would certainly suffer by removal and the possible protective measures short of adoption.

2. Summary of the Judgment

  • The Court of Appeal rejected the complaint that the welfare evaluation was procedurally flawed in a way that caused injustice (the “linear” vs “holistic” point), though it expressed misgivings about the staged way the first instance reasoning was delivered.
  • The Court of Appeal allowed the appeal on substance: the welfare/proportionality evaluation was inadequate and/or wrong in two linked respects:
    1. insufficient focus on the realistically likely consequences for these particular children of the harm that might occur if they remained in maternal care; and
    2. insufficient recognition of the potential for state-supported risk management within the family, notably through Children Act 1989 orders and a Domestic Abuse Protection Order (DAPO) under the Domestic Abuse Act 2021.
  • The care and placement orders were set aside; the interim care orders revived; threshold findings were preserved; and the matter was remitted for a final welfare determination.
  • Dingemans LJ added guidance emphasising liaison between the Family Court and the FTT(IAC) where immigration outcomes may affect (and be affected by) family proceedings, referencing existing protocols and forms.

3. Analysis

A. Precedents Cited

1) Appellate restraint in welfare decisions

The Court anchored the appellate standard in Re J (A Child) [2005] UKHL 40 and the approach to review and deference explained in Piglowska v Piglowski [1999] UKHL 27. The Court also relied on Re H-W (Children) [2022] UKSC 17 (per Dame Siobhan Keegan) for the proposition that where the judge has adopted the correct approach, an appellate court intervenes only if the judge was “wrong”.

This framework mattered because the Court accepted the first instance judge’s careful fact-finding and the absence of significant challenge to core evaluations; the appeal therefore turned on whether the proportionality analysis was sufficiently rigorous and child-specific, not on re-trying disputed facts.

2) “Nothing else will do” and tolerance of imperfect parenting

The Court drew on two foundational statements:

  • Re B (A Child) (Care Proceedings: Threshold Criteria) [2013] UKSC 33 (Baroness Hale at [198]): severance of the parent-child relationship demands a very strict test—only where “nothing else will do”, and in many cases it will be necessary to explore and attempt alternatives.
  • Re L (Care: Threshold Criteria) [2007] 1 FLR 2050; [2006] EWCC 2 (Fam) (Hedley J at [50]): society must tolerate diverse and sometimes harmful parenting; the state’s role is not to eliminate all adverse consequences of imperfect parenting.

The Court treated these as active constraints in a case where the feared harm had not clearly materialised as actual harm to the children (beyond exposure concerns), yet the proposed order was the most drastic—adoption.

3) Structured proportionality and risk analysis in adoption/placement cases

The Court emphasised the structured questions mandated by Re H-W (Children) [2022] UKSC 17 at [56], and aligned them with a line of Court of Appeal authorities: Re F (A Child: Placement Order: Proportionality) [2018] EWCA Civ 2761, L-G (Children: Risk Assessment) [2025] EWCA Civ 60, Re N [2025] EWCA Civ 1541, and Re T (Children: Risk Assessment) [2025] EWCA Civ 93.

The Court used these authorities not to impose rigid “tick-box” compliance, but to identify what was missing: a genuine, evidenced evaluation of consequences and mitigation possibilities, integrated into the comparative welfare balance.

4) The Re B-S discipline

The history of the proceedings shows the trial judge herself identified a flaw in the professional evidence by reference to Re B-S (Children) [2013] EWCA Civ 1146; [2014] 1 WLR 563; [2014] 1 FLR 1035—the absence of a proper comparative analysis of options, particularly the advantages/disadvantages of placement with the mother versus adoption. Although that prompted an adjournment and further work, the Court of Appeal ultimately concluded the final proportionality analysis still fell short in critical respects.

5) Immigration/family court interaction

Dingemans LJ’s concurring remarks draw attention to the relevance of immigration outcomes to family welfare decisions (and vice versa), citing RS (immigration and family court proceedings: India) [2012] UKUT 00218 (IAC) and pointing to the 2013 protocols and Form EX660 arrangements. While not determinative of the appeal outcome, this is a significant operational reminder in cases involving precarious immigration status and parallel tribunal proceedings.

B. Legal Reasoning

1) The facts were largely accepted—so the appeal turned on proportionality

The Court treated the following as essentially fixed: serious domestic abuse findings; inflicted bruising to Z with inability to identify perpetrator; extensive parental dishonesty; lack of acceptance/insight; and the judge’s finding that the mother was likely to reunite with the father absent external pressure. The mother’s real challenge targeted the conclusion that no protective measures could reduce risk sufficiently to avoid adoption.

2) The missing “consequences” analysis

Applying the Re H-W structure, the Court accepted the judge was entitled to find a substantial possibility of future significant harm. But it found the judgments inadequately addressed a distinct and decisive question: what were the realistically likely consequences for these particular children if the feared harm occurred?

The Court noted that references to worst-case outcomes (including “death”) lacked evidential foundation on the facts, and that general statements about domestic abuse (while true in general) were not sufficiently connected to the children’s actual circumstances and evidential history. In effect, the feared harm was described at a high level of generality, while the harm of removal was concrete, immediate, and accepted as significant.

3) The missing mitigation analysis—especially the role of DAPO and Children Act orders

The Court’s second linked criticism was that the judgments insufficiently grappled with the range of protective mechanisms short of adoption. The lead judgment explicitly identified a combination that should have been squarely evaluated:

  • s.31 Children Act 1989 measures (including a supervision order and/or care order mechanisms as appropriate);
  • a Domestic Abuse Protection Order (DAPO) under the Domestic Abuse Act 2021 (not limited in time; breach is a criminal offence); and
  • voluntary and multi-agency supports and monitoring, potentially adding “another layer of vigilance”.

Importantly, the Court did not hold that these measures would necessarily succeed; rather, it held that the first instance evaluation treated protective possibilities as essentially futile (“worthless”) without sufficiently analysing (i) the mother’s proven capacity to seek help when abuse occurred in the UK, (ii) the real-world detectability of risk escalation, and (iii) whether structured legal controls could meaningfully reduce risk below the “nothing else will do” threshold.

4) Proportionality required an explanation of why certain and severe removal harm was outweighed

The Court highlighted the children’s certain future harm if adopted: separation from their primary attachment figure, placement with strangers, probable further move(s), and uncertainty about sibling placement and cultural/identity needs. Against that, the feared harm was future and probabilistic. The key deficiency was the absence of a convincing, evidence-led explanation of why adoption was necessary notwithstanding that asymmetry.

5) Process point: staged judgments and the risk of “surrogate” decision-making

On Ground 2 the Court offered a caution rather than a finding of unfairness: when a case is determined in stages, there is a risk that intermediate issues (acceptance, insight, risk) become narrower surrogates for the overall welfare decision. However, the Court ultimately found no serious irregularity causing injustice in the process adopted here.

6) Disposal

Because the proportionality assessment was ineffective and/or the conclusion was wrong, the Court set aside the care and placement orders, revived interim care orders, preserved the threshold findings, and remitted the matter for a final welfare determination.

C. Impact

  • Sharper insistence on child-specific consequences: even where future harm is found “high” risk, courts must articulate the realistically likely impact on the child, not rely on generic statements or unsupported catastrophic outcomes.
  • DAPO becomes central in public law proportionality: the judgment signals that where domestic abuse risk is central and adoption is proposed, the court should expressly evaluate whether a DAPO under the Domestic Abuse Act 2021, alongside Children Act orders and multi-agency oversight, could reduce risk sufficiently.
  • “Nothing else will do” in “intermediate seriousness” cases: where there are strong protective factors (secure attachment, no addiction/mental health/neglect markers, children thriving under supervised care), the court must give a particularly cogent account of why adoption is necessary despite inevitable removal harm.
  • Operational reminder on immigration interface: Dingemans LJ’s remarks strengthen expectations of effective liaison and non-assumptive sequencing between family proceedings and the FTT(IAC), especially where immigration outcomes may materially alter future parenting feasibility or risk.

4. Complex Concepts Simplified

Care order vs supervision order (Children Act 1989)
A care order gives the local authority parental responsibility and greater control over the child’s care. A supervision order keeps the child with the parent but requires the local authority to “advise, assist and befriend” and provides a structured oversight framework (typically time-limited but extendable), potentially relevant where risk can be managed without removal.
Placement order and adoption
A placement order authorises a local authority to place a child for adoption (often a decisive step towards permanent severance of legal ties). Because adoption is irreversible in practical terms, the law demands stringent justification.
Proportionality and Article 8
Article 8 protects family life. Removing children permanently must be necessary and proportionate: the least interventionist measure that can safely meet the child’s welfare.
“Nothing else will do”
A shorthand for the strict test that adoption (or equivalent permanent severance) is justified only where no other realistic option can meet the child’s welfare needs safely.
DAPO (Domestic Abuse Act 2021)
A Domestic Abuse Protection Order can impose requirements/prohibitions to protect a person from domestic abuse, can last without time limit, and breach is a criminal offence—making it a potentially significant tool in safeguarding analysis where domestic abuse risk is central.
Risk assessment is not “zero risk”
The Court reaffirmed that safeguarding is not about eliminating all risk; the state must tolerate some degree of imperfect parenting, and the proportionality question asks whether risks can be managed to an acceptable level without the most drastic intervention.

5. Conclusion

F (Children: Placement Orders: Proportionality) [2026] EWCA Civ 990 reinforces that adoption decisions must be driven by a disciplined proportionality analysis: courts must identify not only the likelihood and type of future harm, but the realistically likely consequences for the child, and must squarely evaluate state-supported risk management alternatives—expressly including a DAPO where domestic abuse risk is in play.

The decision also highlights the practical importance of coordination between family and immigration jurisdictions where precarious immigration status may shape both risk and feasibility of long-term family arrangements.