No “Permanence” Requirement and Mediation-Delay Counts: The Court of Appeal Recasts the Article 12(2) Settlement Defence

Introduction

This Court of Appeal decision in B (Child Abduction: Settlement) (Rev1) [2025] EWCA Civ 1382 addresses the interpretation and application of Articles 12 and 13(b) of the 1980 Hague Child Abduction Convention, as enacted by the Child Abduction and Custody Act 1980. The case primarily concerns the Article 12(2) “settlement” defence where proceedings begin more than a year after a wrongful removal/retention. A secondary issue was the Article 13(b) “grave risk” exception and the use of undertakings as protective measures.

The child, B (now nearly 4), was born in Portugal to a Portuguese father (F) and British mother (M). In October 2022, B was wrongfully retained in England. Over the next two and a half years, the parents engaged in extensive negotiations and mediation, during which F frequently visited B in England and B regularly visited Portugal, maintaining strong relationships in both countries. After M ultimately withdrew from an agreed plan to relocate B back to Portugal, F issued Hague proceedings in March 2025. The High Court ordered a return with undertakings; M appealed.

The appeal raised three central questions:

• Article 12(2): Was the judge wrong to find that B was not “now settled in [her] new environment” and, if settlement had been established, was the judge’s prospective exercise of discretion to return rational?
• Discretion: How should delay and Convention policy weigh where proceedings are brought long after the wrongful retention, especially when delay is linked to mediation or non-court dispute resolution?
• Article 13(b): Did the judge err in rejecting the grave risk defence and in relying on protective undertakings not mirrored abroad?

Summary of the Judgment

The Court of Appeal (Baker, Birss and Asplin LJJ) allowed the appeal on Article 12 and discretion, but dismissed the appeal on Article 13(b):

• Settlement under Article 12(2):
  • “Settlement” requires arrangements that are long-term rather than transient, but does not require proof of “permanence.” The Court expressly rejected a permanence threshold (para 67–68).
  • The child’s “new environment” includes ongoing relationships and routines, including regular contact with the left-behind parent and cross-border visits. Such contact can support, not undermine, settlement (para 69–73).
  • On the facts, B was settled by March 2025. The High Court’s contrary conclusion was wrong (para 75).
• Discretion after settlement:
  • Applying Re M (Zimbabwe), in non–“hot pursuit” cases, Convention policy carries reduced weight; ordering a summary return without a full welfare analysis was wrong in principle given the uncertainties (notably M’s immigration situation) and the limited evidential foundation for F’s assertions about future travel (para 76–77).
  • Delay is relevant to discretion even if caused by mediation; the High Court’s “policy point” insulating mediation-related delay was rejected as inconsistent with Convention principles (para 78).
• Article 13(b) grave risk:
  • The High Court’s rejection of the grave risk defence stood; allegations, even at their highest and cumulatively, did not cross Article 13(b)’s high threshold (para 83).
  • Protective undertakings were appropriately relied upon; as measures under the 1996 Hague Convention, they are recognisable by operation of law in Portugal, and mirror orders were not a precondition to return on these facts (para 83).

Result: The return order was set aside on Article 12 grounds; the Article 13(b) challenge was dismissed.

Analysis

Precedents Cited and Their Influence

The Court revisited and refined a number of established authorities:

• Perez-Vera Explanatory Report:
  • The Report anchors Convention aims in children’s best interests, emphasising prompt return to restore the status quo and deter abductions, while recognising limited exceptions (paras 16–26, 107, 109). The Court used this to restate that, post–one year, if a child is settled, the ability to order summary return without a full welfare inquiry becomes “extremely limited” (para 65).
• Re N (Minors) (Abduction) [1991] 1 FLR 413 (Bracewell J):
  • Two elements of settlement—physical and emotional/psychological—remain orthodox (para 12). However, the Court of Appeal rejected the later gloss that settlement requires “permanence,” aligning with academic criticism and first-instance practice (para 14, 67–68).
  • The concept of “new environment” is broad (home, school, people, friends, activities, opportunities) and does not exclude the continuing relationship with the abducting parent; the Court extended this logic by holding that the relationship with the left-behind parent and cross-border contact can also be part of the “new environment” (para 69–73).
• Cannon v Cannon [2004] EWCA Civ 1330:
  • Concealment undermines settlement; emotional/psychological integration matters (para 13). Here, the opposite—open living and sustained contact with F—supported settlement (para 72–73).
• Re B (A Child) [2018] EWHC 1643 (Fam) (Williams J):
  • Holistic, child-focused evaluation; physical and emotional components; settlement sits on a spectrum; absence of a rigid two-stage test; overlap with habitual residence but lesser requirement than full integration (paras 17; 42; 57(3)(c)). The Court expressly endorsed this framework (para 66) and applied it to recognise cross-border parental contact as part of the new environment (para 69–73).
• AX v CY (Article 12 Settlement) [2020] EWHC 1599 (Fam) and Re G and B (Children) [2025] EWHC 795 (Fam):
  • Support the holistic approach; “long-term” vs “transient” focus. The Court of Appeal reaffirmed the “long-term not transient” threshold while repudiating any permanence requirement (para 18, 67–68).
• Re M (Zimbabwe) [2007] UKHL 55 (Baroness Hale):
  • Even where settlement is proved, discretion is at large but Convention policy diminishes with time; in non–“hot pursuit” cases summary return is seldom appropriate absent strong countervailing factors (paras 19–20). The Court relied on this to criticise reliance on assertions about future feasibility of contact and to emphasise the need for a full welfare analysis where settlement exists and time has elapsed (para 76–77).
• Re IG [2021] EWCA Civ 1123 and E v D (Return Order) [2022] EWHC 1216 (Fam):
  • Confirm the high threshold for Article 13(b), the use and scrutiny of undertakings, and the need to analyse protective measures in concrete terms (paras 21–22). The Court upheld the High Court’s approach on Article 13(b) and undertakings, consistent with these cases (para 83).

Legal Reasoning

The Court’s reasoning unfolds in three linked strands:

1. Re-stating the settlement test:
   • Settlement requires “long-term” (not transient) arrangements; “permanence” is not required (para 67–68). Imposing a permanence test wrongly elevates the threshold beyond the Convention’s language and purpose and is inconsistent with first-instance authority (Re C; Re E) which found settlement notwithstanding immigration uncertainty.
   • The child’s “new environment” is assessed holistically and includes continuing and structured contact with the left-behind parent and cross-border routines. The High Court erred in treating B’s repeated trips to Portugal and her ties there as undermining settlement in England; they were part of the stable, predictable pattern of her life and thus supported settlement (para 69–73).
   • Open living and cooperation (as opposed to concealment) is settlement-enhancing. F’s knowledge of M’s and B’s whereabouts, frequent visits, and acquiescent conduct at the end of visits reinforced the finding of settlement (para 72–73).
   • Excessive focus on parental communications and the lack of a “permanent” statement of intention misdirected the settlement inquiry away from the child’s actual lived environment (para 68, 74–75).
2. Discretion after settlement:
   • Once settlement within Article 12(2) is established in a non–“hot pursuit” scenario, the Convention’s policy weight declines (Re M (Zimbabwe)). A summary return should not be ordered on thin evidence where there are real uncertainties (such as immigration status, right to work/benefits, and mental health impacts) without a full welfare assessment (para 76–77).
   • Delay matters even if parties tried mediation: the Court rejected the High Court’s “policy point” that non-court dispute resolution should not be held against the applicant. Any significant delay is relevant to discretion, and its reasons may be considered, but mediation does not immunise the delay from its effects under the Convention (para 78).
3. Article 13(b) and protective measures:
   • Article 13(b)’s high threshold remained unmet on the mother’s case, even cumulatively (para 83). The High Court did not err in evaluating disputed allegations within the summary process or in concluding that proposed undertakings (including non-molestation-style terms, funding, and logistical supports) provided adequate protection.
   • Undertakings as measures: The Court endorsed the High Court’s reliance on undertakings as “measures” recognised under the 1996 Hague Convention, not requiring a mirror order as a precondition to return on these facts (para 83), while acknowledging the general principles that efficacy—not merely formal enforceability—must be assessed.

Impact

This judgment provides important practical and doctrinal guidance for Hague abduction cases issued after one year:

• Resetting the settlement threshold: Courts should not require “permanence.” The enquiry is whether the child’s circumstances are long-term, not transient, assessed holistically from the child’s perspective. Practitioners should avoid over-technical arguments focusing on rigid declarations of parental intention.
• Cross-border contact as part of settlement: Regular, cooperative contact with the left-behind parent—even involving periodic returns to the requesting State—can support settlement in the requested State. Attempts to argue that ongoing ties to the requesting State negate settlement will carry less force where the child’s overall life has predictability and stability.
• Delay and mediation: Left-behind parents should not assume that mediating for many months (or years) insulates them from the consequences of delay. If settlement arises, discretionary return becomes inherently harder to justify. Early issue of Hague proceedings—possibly stayed to allow mediation—may be advisable to preserve Convention objectives.
• Discretion demands more in settled cases: Where settlement is established, courts should be slow to order summary return on assertions; immigration, welfare, and practical uncertainties will often require a full welfare analysis in the forum now seized of long-term arrangements.
• Protective measures and undertakings: The Court reaffirms that undertakings can, in appropriate cases, suffice as protective measures, especially where both States are 1996 Hague Convention parties. Nevertheless, their efficacy must be scrutinised in concrete terms and proportional to the weight placed on them.

Complex Concepts Simplified

• “Settlement” (Article 12(2)):
  • Meaning: The child is “now settled in [their] new environment.”
  • What it covers: Both physical integration (home, schooling/nursery, community, routines) and emotional/psychological stability.
  • Threshold: Long-term, not transient stability; “permanence” is not required.
  • Scope of “new environment”: Includes relationships with both parents, extended family, and structured cross-border routines.
• “Discretion” after settlement:
  • Even if a child is settled, a court retains discretion to order return—but the weight of Convention policy declines with time.
  • Where serious uncertainties exist (immigration, welfare impact), ordering return without a full welfare inquiry risks error.
• “Hot pursuit”:
  • Cases issued promptly (within a year, or very soon after removal/retention), where the Convention’s objective of swift return can still be realised. Outside hot pursuit, policy considerations weaken.
• Article 13(b) “grave risk”:
  • A high threshold: the return must expose the child to a grave risk of physical/psychological harm or to an intolerable situation.
  • Courts assess whether protective measures (e.g., undertakings) can mitigate risk, bearing in mind enforceability and practical efficacy.
• Undertakings, mirror orders, and the 1996 Hague Convention:
  • Undertakings are promises to the court by a parent (e.g., not to harass, to provide housing/funding, to facilitate contact).
  • Where both States are parties to the 1996 Convention, undertakings approved as “measures” can be recognised abroad by operation of law, although courts must still evaluate their practical effectiveness.
  • Mirror orders (obtaining a corresponding order in the requesting State) may be prudent in some cases, but are not invariably required.

Conclusion

B (Child Abduction: Settlement) significantly clarifies the Article 12(2) landscape. The Court of Appeal definitively removes any “permanence” requirement from settlement, insists on a holistic, child-centred assessment that recognises cross-border contact as part of the child’s new environment, and emphasises that open, cooperative parenting can facilitate—rather than undermine—settlement. In the discretionary stage, the Court reaffirms Re M (Zimbabwe): after prolonged delay, Convention policy arguments carry less weight; ordering summary return on limited evidence—particularly amidst immigration and welfare uncertainties—is inappropriate without a full welfare analysis. The Court also confirms continued adherence to the high Article 13(b) threshold and the legitimate role of undertakings as protective measures under the 1996 Hague Convention when shown to be effective.

Practically, the decision counsels left-behind parents to issue proceedings promptly, even when exploring mediation, and guides first-instance judges away from over-technical settlement analyses or heavy reliance on parental intention statements. It promotes a pragmatic, welfare-sensitive application of the Convention in long-delay cases, ensuring that the settlement defence functions as intended: to avoid compounding earlier harm by summarily uprooting a child who has, in fact, become established in a new environment.