This appeal concerns the Hague Convention on the Civil Aspects of International Child Abduction (25 October 1980) (“the Abduction Convention”). It raises general questions relating to:

the habitual residence of the children in the case; and

whether a wrongful retention in fact occurred, and if so when.

The Abduction Convention is in force for some 97 States. Its preamble records the desire of those States:

The general scheme of the Convention is to enable a left behind parent to make this application in the State to which a child has been taken, seeking return of the child. States are required to set up Central Authorities to transmit and receive such applications. Where the removal from the home State, or the retention in the destination State is wrongful, the courts of the recipient State are required by article 12 to order the return of the child “forthwith”. Apart from a saving provision in article 20 which permits refusal to return where such would amount to a breach of the requested State’s fundamental principles of human rights, that obligation to return is subject to very limited exceptions which, if present, enable (but do not require) return not to be ordered. Those exceptions are found in article 13 (rights of custody not being exercised; consent or acquiescence of the left-behind parent; grave risk that return would expose the child to physical or psychological harm or would place him/her in an intolerable situation; child’s objections), and in article 12 (child has been in the recipient State for one year from the wrongful removal or retention and is now settled there). Where prompt notice of wrongful removal or retention is received, the recipient State is required by article 16 to abstain from any decision on the merits of rights of custody, unless it is determined that return is not to be ordered. Moreover, States are required to act fast on any request. By article 11 an initial period of six weeks is stipulated, and the applicant or his Central Authority are entitled to an explanation from the recipient State if that period is exceeded. Thus the return is summary and its object is to enable merits decisions as to the child’s future to be made in the correct State, rather than in the State to which the child has been wrongfully taken, or in which he/she has been wrongfully retained. The general purposes and scheme of the Convention are expanded upon in an explanatory report by Professor Elisa Pérez-Vera on the work of the drafting conference, which report accompanied the original framing of the Convention; it is accordingly an aid to construction recognised in international law and in particular under article 32 of the Vienna Convention on the Law of Treaties (1969). In England and Wales the Convention is given domestic effect by the Child Abduction and Custody Act 1985, section 1(2).

Four key concepts underlie the Convention: wrongful removal, wrongful retention, rights of custody and return. The principal provisions which require attention in the present case, apart from the preamble and article 1, set out above, are articles 3, 4, 5, 12 and 16. So far as relevant, they say:

The mother, although born in Canada, was brought up in England and is originally of British nationality. The father is Australian. Mother went to live in Australia in 2008. There she met, and later married, the father. She took Australian citizenship in 2014 and so now has dual British/Australian nationality. Two children were born to them in 2012 and 2014. By the end of 2014 the marriage was in difficulties. Mother was on maternity leave from her job at the time. She told Father that she wanted to make a trip to England with the children before going back to work. Although initially reluctant, he agreed to an eight-week visit. Mother and the children came to England on 4 May 2015 with return tickets then scheduled for 24 June. They went to stay with the maternal grandmother, where they have since remained.

Discussions between Mother and Father then resulted in Father agreeing to an extension of the eight-week visit. Initially, Father agreed to a four-week extension. But then, on 28 June 2015 he sent Mother an email which said:

The children had entered England on six month visitor visas, so steps needed to be taken to regularise the longer stay now contemplated. What Mother then did loomed large at the hearing before the judge. Without telling Father she applied on 2 November 2015 for British citizenship for the children. She engaged solicitors to make the application. Those solicitors wrote on her behalf to the United Kingdom immigration authorities on 4 November 2015. In the course of a long letter they asserted that the marriage had irretrievably broken down, that Mother had been the object of repeated domestic abuse which had, moreover affected the elder child adversely, that she had been “effectively forced not to return to Australia in order to safeguard herself and the children” and that the children could not return to Australia because there was nowhere safe for them to go. The letter added:

Meanwhile in continuing correspondence between the parents, Father pressed Mother on her expected date of return. On 11 February 2016 she wrote saying that she did not know what her plans were but “Short term I will not be returning in May”. She added “I will not base my return to Australia at your demand.” Later, Father referred her to the Abduction Convention and instructed solicitors who wrote formally to ask Mother when she planned to return. She replied in June 2016:

In due course both parents gave oral evidence at the hearing before the judge of Father’s application under the Abduction Convention. By then it was accepted that Mother did not propose to return. The issue of when she had so decided was much in contention. The judge’s conclusions on the topic are considered below: [2016] EWHC 3535 (Fam). But Mother’s own case was that by April 2016 she had “felt that we wouldn’t be going back”. That meant that on any view there had been a decision not to return before the expiry of the agreed year of stay in England. That gives rise to the second general question in this case, namely: whether and when such a decision can make the retention in the destination country wrongful for the purposes of the Abduction Convention before the expiry of any agreed or sanctioned term of residence there.

The judge also had to make findings as to the place of the children’s habitual residence. The details of his conclusions are set out below, but he found that they were habitually resident in England and Wales by at the latest the end of June 2016, which was the last possible date for the expiry of the agreed year of stay. He added that in his view it was eminently arguable that they had acquired habitual residence significantly before that date. Those findings give rise to the first general question in this case, namely: what is the effect on an application under the Abduction Convention if the child has become habitually resident in the destination (requested) State before the act relied on as a wrongful removal or retention occurs.

In the simple paradigm case of wrongful removal, one parent will have taken the child from the State where s/he is habitually resident to a destination State. Similarly, in the simple paradigm case of wrongful retention, one parent will have travelled with the child from the State of habitual residence to the destination State, for example for an agreed fortnight’s holiday (and thus without the removal being wrongful), but will then wrongfully have refused to return. In each of those paradigm cases, the child will have remained habitually resident in the home State. An application under the Abduction Convention will be made in the destination (or “requested”) State for the return of the child to the State of habitual residence. The return will be a summary one, without investigation of the merits of any dispute between the parents as to custody, access or any other issue relating to the upbringing of the child (article 16). Such merits decisions are for the courts of the State of the child’s habitual residence.

In some cases, however, it is possible that by the time of the act relied upon as a wrongful removal or retention, the child may have acquired habitual residence in the destination State. It is perhaps improbable in the case of removal, but it is not in the case of retention. It may particularly happen if the stay in the destination State is more than just a holiday and lasts long enough for the child to become integrated into the destination State. It is the more likely to happen if the travelling parent determines, however improperly, to stay, and takes steps to integrate the child in the destination State. Even in the case of wrongful removal it may be possible to imagine such a situation if, for example, there had been successive periods of residence in the destination State, followed by a removal from the State of origin which infringed the rights of custody of the left-behind parent.

In England and Wales at least, this possibility did not in practice arise in the past, since it was regarded as axiomatic that one parent could not by unilateral action alter the habitual residence of the child. This proposition dated from a dictum of Lord Donaldson MR in In re J (A Minor) (Abduction: Custody Rights) [1990] 2 AC 562, 572, and the decision of Wall J in In re S (Minors) (Child Abduction: Wrongful Retention) [1994] Fam 70, which was approved by the Court of Appeal in In re M (Abduction: Habitual Residence) [1996] 1 FLR 887, 892, and, as Baroness Hale explained in A v A (Children: Habitual Residence) [2014] AC 1 at para 39, it was thereafter taken for granted. Such a proposition is, however, not generally adopted in other countries, including the United States, sits uneasily with the equally axiomatic principle that habitual residence is a question of fact, not law, and is difficult to accommodate within the European approach which requires an examination of integration, as exemplified in Proceedings brought by A (Case C-523/07) [2010] Fam 42 and Mercredi v Chaffe (Case C-497/10PPU) [2012] Fam 22, and which is binding on this country via Council Regulation (EC) No 2201/2003 (“Brussels II Revised”): see the analysis of Baroness Hale in A v A. It was recognised in In re H (Children) [2015] 1 WLR 863 that such a rule could not survive.

If the habitual residence of the child may have changed to the destination State by the time of the wrongful act of removal or retention relied upon, then it becomes necessary to know whether the summary procedure of the Abduction Convention remains available in such a case or does not. Hence the first general question. If the answer is that it is not available, because the Abduction Convention pre-supposes an application made in a destination State which is not the State of habitual residence, then the second general question becomes of importance. It becomes important because deliberate acts aimed at integrating the child in the destination State may well be undertaken by the travelling parent once he has decided not to honour his obligation to return to the State of origin. It will then matter whether such acts, or other manifestations of his decision, can themselves amount to wrongful retention. If they can, then wrongful retention may occur before any change of habitual residence has been achieved and whilst the child is still habitually resident in the State of origin. If they cannot, and wrongful retention cannot occur until the day of agreed return arrives, it may be too late for any application under the Abduction Convention, because the same acts which derive from and accompany the decision not to return may themselves have resulted in the child becoming habitually resident in the destination State.

The first question is accordingly this: if by the time of the act relied on as wrongful removal or retention the child is habitually resident in the State where the application for return is made, is summary return under the Abduction Convention still available or not?

This question did not arise in either of the courts below, where everyone proceeded on the assumption that the answer was “no”. It arose in the course of argument in this court, and we have had the benefit not only of some immediate oral submissions, but of considered post-hearing written submissions from both parties and from the International Centre for Family Law, Policy and Practice as intervener.

The argument that summary return under the Abduction Convention remains available runs as follows:

Accordingly, it is said, on facts such as those of the present case, if the child’s habitual residence is in England by the time of the act relied on as wrongful retention, that simply means that it becomes English law which decides whether the retention was wrongful. If it is decided that it is wrongful, there can still be a return to Australia.

This may be a possible construction if one has regard simply to the wording of articles 3 and 12. It is, however, not a persuasive construction if one takes into account the general scheme of the Convention. Nor is it the way that the Convention has been operated over the nearly 40 years of its life. Nor is this construction consistent with the way in which the Convention has been treated by subsequent multi-lateral instruments in the general field of the conflict of laws in relation to disputes about the upbringing of children.

By the time of the Abduction Convention, habitual residence was already established as the principal internationally-recognised basis for according jurisdiction relating to the upbringing of children. At any rate by the time of the 1961 Hague Convention on the Protection of Infants, habitual residence was, together with in some respects the law of the child’s nationality, the principal basis for jurisdiction (see article 1). By the time of the Abduction Convention, Professor Pérez-Vera’s report was saying (in para 19) that the Convention:

The entire scheme of the Abduction Convention is to provide a summary remedy which negates the pre-emptive force of wrongful removal or retention. The aim was also to defeat forum-shopping. This is made clear by Professor Pérez-Vera’s report, especially at paras 14-15.

This underlying rationale of the scheme of the Abduction Convention was recognised by this court in In re E (Children) (Abduction: Custody Appeal) [2011] UKSC 27; [2012] 1 AC 144. Baroness Hale and Lord Wilson observed at para 8:

The whole point of the scheme adopted was to leave the merits to be decided by the courts of the place of the child’s habitual residence. The preamble makes this clear in almost the first words of the Convention. If, however, the child has by the time of the act relied on as wrongful become habitually resident in the requested State, then that State will be the appropriate place for the merits of any custody dispute to be resolved. If the requested State is the habitual residence of the child, there can be no place for a summary return to somewhere else, without a merits-based decision, still less for such to be mandatory. That would be so whether or not the removal or retention was, judged by the law of the requested State, as the State of habitual residence, wrongful, for even if it were, it would remain open to either party to ask the courts of that State to review the future plans for the upbringing of the child.

This understanding of the scheme of the Abduction Convention is reflected in the provisions of both the 1996 Convention and Brussels II Revised. A large number of nations are party to these two multinational instruments, but not nearly so many as are party to the Abduction Convention. These two instruments are concerned, unlike the Abduction Convention, with recognition and enforcement. But they are scrupulous to ensure that wherever possible they are consistent with the Abduction Convention, whose scheme they very plainly seek to preserve.

The 1996 Convention adopts, by article 7(2) a definition of wrongful removal and retention in the same words as article 3 of the Abduction Convention. Substantively, article 7(1) provides for cases of wrongful removal and retention a limited exception to the ordinary rule in article 5 that jurisdiction moves with the habitual residence of the child. In effect, the State of habitual residence immediately before the wrongful removal or retention keeps jurisdiction until not only habitual residence has shifted but also there has been an opportunity for the summary return provided for by the Abduction Convention. The effect, plainly intended, is to preserve the regime of the Abduction Convention, and in particular the mandatory summary return. But if, at the time of the wrongful act, the requested State had become the State of habitual residence, the extension by article 7(1) to the jurisdiction of the previous State of habitual residence would have no application and the requested State would have sole jurisdiction; in such an event, there could be no question of a mandatory summary return without consideration of the merits.

Brussels II Revised adopts a similar structure to the 1996 Convention. article 2(11) provides a definition of wrongful removal and retention which, although not in identical words to article 3 of the Abduction Convention, achieves the same result, and in particular makes the test for wrongfulness the law of the State of habitual residence immediately before the act relied upon. Article 10 prolongs the jurisdiction of that State in the event of a wrongful removal or retention in much the same terms as does article 7 of the 1996 Convention. As with the 1996 Convention, the intention is plainly to preserve the regime of the Abduction Convention, and article 11 goes on to make supplemental provision for the handling of applications under it. It is revealing that it does so after introduction in the following terms:

There are other examples of legislative provisions making explicit the principle that return under the Abduction Convention presupposes return from a state other than that of habitual residence at the time of the wrongful act. In New Zealand, the Convention is given effect by the Care of Children Act 2004. In that Act, “removal” includes “retention”, in each case as defined in article 3 of the Convention. Section 103 provides:

In re H (Minors) (Abduction: Custody Rights), In re S (Minors) (Abduction: Custody Rights) [1991] 2 AC 476 the House of Lords addressed the question whether wrongful removal and wrongful retention were mutually exclusive concepts; the issue arose in the context of the commencement date for the 1985 Act as between the two States involved. The House held that for the purposes of the Abduction Convention the two concepts were mutually exclusive, and that because article 12 required it to be possible to calculate the 12-month period from a wrongful retention, as well as from a wrongful removal, the former could not be regarded as simply continuing, but had to have an identified date, in effect its beginning. Giving the sole speech, Lord Brandon explained, at 498G:

The researches of counsel, for which we are very grateful, have disclosed that a similar approach has been adopted in Scotland, France, Israel, Switzerland, Canada, Australia, New Zealand and various United States courts whether federal or state. Whilst those surveys cannot by their nature be exhaustive of every decision in every jurisdiction, what is significant is that none of them, including those conducted on behalf of those arguing against the currently assumed analysis (Father and the Intervener), has unearthed any decision to the contrary.

In C v M (2014) (Case C-376/14PPU); [2015] Fam 116 the Court of Justice of the European Union (“CJEU”) adopted the same analysis. The French Father had made application to the Irish Court for the return of children who had been taken to Ireland by Mother. The background was an initial decision of the French court permitting relocation to Ireland, which had been appealed promptly. Mother had moved notwithstanding the pending appeal, a stay having been refused to Father, and subsequently the French decision had been reversed by the appeal court. The Irish court was minded to find that the child had become at some stage habitually resident in Ireland. It referred a number of questions to the CJEU. The CJEU decided (1) that the initial removal to Ireland had not been wrongful, because of the then extant first instance decision permitting the move (para 44), (2) that the subsequent retention there after the French appellate decision might justify an order for return but (3) this would depend on whether by then the child was habitually resident in Ireland (paras 45-49 and 63). If habitual residence had by then been established in Ireland, there could be no order for return. At para 48 the court said:

It is certainly true that this paragraph proceeds from the words of article 11(1) of Brussels II Revised. But the application which the father had made was under the Abduction Convention. He had referred also to Brussels II Revised, but this Regulation does not contain the duty to return a child; what it does is to recognise that the Abduction Convention does contain such a duty, and by article 11 it provides supplementary rules for how this duty is to be performed. En route to its conclusion, the CJEU emphasised, first, that the Regulation and the Abduction Convention were to be “uniform”, that is to say consistent (para 58), and secondly that a decision to return under the Abduction Convention is not a decision on the merits and thus there can be no occasion for a conflict of jurisdiction between the requesting and requested State (paras 37 and 40-42). It left to the Irish court the decision of fact whether and when habitual residence had been established in Ireland. It may be that its proposition that for a return order under the Abduction Convention to be made it was essential that the child was habitually resident at the time of the wrongful act in the State of origin, as distinct from some State other than the requested State, might be wider than necessary, for it may not have considered the possibility of habitual residence in an intermediate State, which did not arise for debate. But what is abundantly clear is that it is only under the Abduction Convention that a summary order for return is provided for, and that such an order could not be made if the child was, by the time of the wrongful act relied upon, habitually resident in the requested State. There is no hint in the court’s decision that Brussels II Revised has in any sense modified the fundamentals of the scheme of the Abduction Convention for EU members; quite the contrary.

In the later case of OL v PQ (2017) (Case C-111/17PPU), a different chamber of the CJEU reached a similar conclusion. The court held that a child born in Greece was habitually resident there, despite the originally Italian home of her parents, and that in consequence an order under the Abduction Convention for return from Greece to Italy could not be made by the Greek court. At para 38 the court said:

The nearest case proffered as any indication to the contrary is In re G (A Minor) (Enforcement of Access Abroad) [1993] Fam 216. There, the Court of Appeal held that a Canadian-resident father could use the Abduction Convention (article 21) to enforce his Canadian-given rights of access in relation to a child who was habitually resident in England by the time the mother declined to comply with them. But that was not a case involving any question of return. The provisions of the Convention in relation to access are notably more fluid and flexible. They simply require the central authorities to facilitate co-operation with a view to preserving access rights. They make no demands of the courts of the requested State and to the extent that they contemplate that an application may be made there, they appear to assume that those courts will conduct a merits hearing. They provide no guide to the scheme of the Convention in relation to applications for orders for summary return.

These various examples of the practice as to the application of the Abduction Convention thus all point in the same direction. The Convention cannot be invoked if by the time of the alleged wrongful act, whether removal or retention, the child is habitually resident in the State where the request for return is lodged. In such a case, that State has primary jurisdiction to make a decision on the merits, based on the habitual residence of the child and there is no room for a mandatory summary return elsewhere without such a decision. It may of course be that in making a merits decision, the court of the requested State might determine that it is in the best interests of the child to be returned to his previous home State, and indeed might do so without detailed examination of all possible evidence, as the English courts may do (see In re J (A Child) (Custody Rights: Jurisdiction) [2005] UKHL 40; [2006] 1 AC 80). But so to do is very different from making a summary order for return without consideration of the merits under the Abduction Convention.

The submissions made to this court addressed also the separate question of whether a return under the Abduction Convention, if made, must always and only be made to the State of habitual residence immediately before the wrongful act. It is to be noted that article 12 does not contain any such restriction, and that Professor Pérez-Vera’s report at para 110 makes clear that the decision not to do so was deliberate. The reason given is that whilst ordinarily that State will be the obvious State to which return should be made, there may be circumstances in which it would be against the interests of the child for that to be the destination of return. The example given is of the applicant custodial parent who has, in the meantime, moved to a different State. The propriety, in such circumstances, of an order returning the child to the new home state of the custodial parent is not in issue in this case. For the reasons given above, the silence of article 12 on the destination of a return order is of no help on the issue which does arise, namely whether an order for return can be made if at the time of the wrongful act the child was habitually resident in the requested State. It is however to be observed in passing that the unusual circumstances envisaged in para 110 of the Pérez-Vera report were held at first instance to have arisen in O v O (Child Abduction: Return to Third Country) [2013] EWHC 2970 (Fam); [2014] Fam 87 and there did result in an order for return to the new home State.

This was the question of principle on which leave to appeal to this court was given. If the child has been removed from the home State by agreement with the left-behind parent for a limited period (and thus the removal is not wrongful), can there be a wrongful retention before the agreed period of absence expires? The classic example of the possibility is where the travelling parent announces, half way through the agreed period (say of a sabbatical year of study for the parent) that he will not under any circumstances return the child in accordance with the agreement he made. He might do more. He might effectively make it impracticable to return, by, for example, selling his house in the home State, abandoning his job there, and obtaining residency in the new State for himself and the child on the basis of an undertaking that they will both remain there indefinitely. No doubt other examples could be postulated. The question is whether, if such a thing occurs, there is then and there a wrongful retention, or whether his retention of the child cannot in law be wrongful until the date agreed for return arrives and, as it was graphically put in the American case of Falk v Sinclair (2009) 692 F Supp 2d 147, the aeroplane lands and the child is not among those who disembark.

There is some difficulty in devising a suitable shorthand for the possibility of wrongful retention in advance of the due date for return. One which has been used is “anticipatory retention”. This is certainly convenient but it may lead to misconceptions. If early wrongful retention is a legal possibility, it is not because there is an anticipation of retention. On the contrary, the child is retained in the destination State from the moment of arrival, just as he is removed from the home State at the moment of departure. If the departure and arrival are permitted by agreement with the left-behind parent, or sanctioned by the court of the home State, they are still respectively removal and retention, but they are not wrongful. So what is under consideration is a retention which becomes wrongful before the due date for return.

The key to the concept of early wrongful retention, if it exists in law, must be that the travelling parent is thereafter denying, or repudiating, the rights of custody of the left-behind parent and, instead of honouring them, is insisting on unilaterally deciding where the child will live. In the absence of a better expression, the term which will be used here will, for that reason, be “repudiatory retention”. That is not to import contractual principles lock stock and barrel into the concept, for the analogy with a contract is only partial. It is simply to attempt a shorthand description.

The expert and thorough analysis of the known cases in several different jurisdictions which was undertaken in this case by Black LJ, as she then was, cannot be improved upon. It is to be found at paras 28-97 of her judgment [2017] EWCA Civ 980; [2017] 3 FCR 719. On this part of her judgment the Court of Appeal was unanimous. It shows that a concept of repudiatory retention has been recognised in some jurisdictions, and for many years now: early examples included Wall J’s decision in In re S (Minors) (Abduction: Wrongful Retention) [1994] Fam 70 and the Canadian case of Snetzko v Snetzko (1996) CanLII 11326. Other cases have rejected the concept, for example in Australia. There are cases going either way in the United States. It follows that there is no generally accepted international practice on the point, nor is there clear authority either way in this jurisdiction. In those circumstances it is necessary for this court to address the principle of the suggested concept.

The Court of Appeal concluded unanimously that there was a concept of repudiatory retention known to the law. It divided, however, as to whether it could exist only when the repudiation was communicated to the left-behind parent (or at least manifested by action), as Black LJ held, or whether such communication was not necessary in law, as Sharp and Thirlwall LJJ concluded. In considering the existence of the concept, it is necessary also to address how repudiatory retention, if it exists at all, may occur.

The helpful submissions made to this court identified six suggested reasons why such a concept is inconsistent with the Abduction Convention and not known to the law.

The crux of the issue lies in the first two contentions, which are different ways of expressing the same point. If there is no breach of the rights of custody of the left-behind parent, then it is clear that the Convention cannot bite; such a breach is essential to activating it, via articles 3 and 12. It is clearly true that if the two parents agree that the child is to travel abroad for a period, or for that matter if the court of the home State permits such travel by order, the travelling parent first removes, and then retains the child abroad. It is equally true that both removal and retention are, at that stage, sanctioned and not wrongful. But to say that there is sanctioned retention is to ask, rather than to answer, the question when such retention may become unsanctioned and wrongful.

When the left-behind parent agrees to the child travelling abroad, he is exercising, not abandoning, his rights of custody. Those rights of custody include the right to be party to any arrangement as to which country the child is to live in. It is not accurate to say that he gives up a right to veto the child’s movements abroad; he exercises that right by permitting such movement on terms. He has agreed to the travel only on terms that the stay is to be temporary and the child will be returned as agreed. So long as the travelling parent honours the temporary nature of the stay abroad, he is not infringing the left-behind parent’s rights of custody. But once he repudiates the agreement, and keeps the child without the intention to return, and denying the temporary nature of the stay, his retention is no longer on the terms agreed. It amounts to a claim to unilateral decision where the child shall live. It repudiates the rights of custody of the left-behind parent, and becomes wrongful.

The plain purpose of the Abduction Convention is to prevent the travelling parent from pre-empting the left-behind parent. The travelling parent who repudiates the temporary nature of the stay and sets about making it indefinite, often putting down the child’s roots in the destination State with a view to making it impossible to move him home, is engaging in precisely such an act of pre-emption.

It is possible that there might also be other cases of pre-emptive denial of the rights of custody of the left-behind parent, outside simple refusal to recognise the duty to return on the due date. It is not, however, necessary in the present case to attempt to foresee such eventualities, or to consider whether fundamental failures to observe conditions as to the care or upbringing of the child might amount to such pre-emptive denial. It is enough to say that if there is a pre-emptive denial it would be inconsistent with the aim of the Abduction Convention to provide a swift, prompt and summary remedy designed to restore the status quo ante to insist that the left-behind parent wait until the aeroplane lands on the due date, without the child disembarking, before any complaint can be made about such infringement.

It is no doubt true that a travelling parent might change his mind after an act of repudiation. But so he might after a failure to return on the due date, and commonly does when faced by notice of the provisions of the Abduction Convention, or by an application under it. So also he might, after making an unsanctioned move to an unagreed country, or after embarking on an unsanctioned programme of religious conversion. The possibility of a change of heart is no reason not to recognise that the heart needs changing if rights of custody in the left-behind parent are to be respected. On the contrary, the desirability of inducing a prompt change of mind is an argument for recognising a repudiatory retention when and if it occurs. Proof that it has occurred is a matter of evidence, and what manifestation of it must be demonstrated is considered below.

If a concept of repudiatory retention exists, it would indeed follow that once such an act occurs, the article 12 12-month clock would begin to run at that point. If the left-behind parent knows of the repudiation, there is every reason why it should run. If he does not, the possibility exists that the 12-month period partly, or sometimes wholly, may pass before he finds out and can make an application under the Abduction Convention. But it is a mistake to think of the 12-month period as a limitation period, of the kind designed in Limitation Acts to protect a wrongdoer from claims which are too old to be pursued. It is not a protection for the wrongdoer. Rather, it is a provision designed in the interests of the child. It operates to limit the mandatory summary procedure of the Convention to cases where the child has not been too long in the destination State since the wrongful act relied on. Where it applies, it does not prevent a summary return; it merely makes it discretionary. In the event that an act of wrongful repudiatory retention had been concealed, that concealment might well be one factor in the decision whether to order return or not. In other cases, the settlement of the child might be so well established that notwithstanding the wrong done by the travelling parent, it is too late to disturb it. Such decisions are fact-sensitive ones which are properly left to the court of the requested State. The risk of the 12-month period running without the knowledge of the left-behind parent is in any event distinctly less fatal to his interests than the risk of the child’s habitual residence being changed without his knowledge, or indeed with his knowledge but without him being able to invoke the Convention because the due date for return has not yet arrived. The latter risk creates a complete bar to return under the Convention; the former a discretionary one.

The concern that Abduction Convention applications may become longer and more complicated is a point well made. It was convincingly voiced in the Court of Appeal by Black LJ. It is of the essence of such cases that the remedy is a swift and summary one. Oral evidence should be the exception, not the rule. But some limited disputes of fact are bound to arise. In the kind of case where retention is in question, it will often be critical to establish what the terms were of any arrangement under which the child travelled. That may be as necessary to establish the date of due return (and thus conventional wrongful retention) as to establish an earlier repudiatory retention. The Family Division judges who hear these cases are well used to managing them actively and to controlling any tendency to spill outside the issues necessary to determine them. If the correct rule is that repudiatory retention must be demonstrated by overt act or statement (see below) the danger of speculative applications being made, or of hearings degenerating into speculative cross-examination as to the internal and undisclosed thinking of the travelling parent ought not to arise.

It may be that in many cases which would be covered by the concept of repudiatory retention the left-behind parent may have remedies alternative to an application under the Abduction Convention. We were pressed with the contention that ordinarily he will be able to seek an order for return in the home State, and then enforce it in the destination State. This may indeed sometimes be possible. It will be possible if both States are party to the 1996 Convention and if at the time of the application to the court of the home State the child is still habitually resident there. In that event, the home State has jurisdiction (article 5) and the destination State must enforce its decision (article 23). Article 7 of the 1996 Convention prolongs the jurisdiction of the home State if there has been a wrongful retention, but if the habitual residence of the child has been changed to the destination State by the time of the act relied upon, there will be no wrongful retention and article 7 will not apply. Nevertheless, the necessity for habitual residence in the home State presents no greater hurdle to the left-behind parent under the 1996 Convention than under the Abduction Convention, because if the habitual residence of the child has shifted to the destination State by the time of the act relied on, neither form of machinery will work. Likewise, if both States are members of the EU and governed by Brussels II Revised. All that said, the critical fact is that by no means all States which are party to the Abduction Convention are party to the 1996 Convention; at the time of the hearing in this court there were some 49 States which are not. Even fewer are members of the EU. The Abduction Convention has its own self-contained scheme and should function as such. The recognition and enforcement provisions in the 1996 Convention are, as explained above, meant to preserve that scheme and not to substitute for it. Moreover, such an application to the home State would have to trigger a merits hearing, in which the home State has to adjudicate upon where the best interests of the child now lie, and upon whether habitual residence has shifted, all depending on facts occurring perhaps some thousands of miles away. That is not at all the same as the mandatory summary remedy provided by the Abduction Convention. Even in jurisdictions, such as England and Wales, which retain the practice of sometimes returning children without a full investigation of the facts (In re J, para 34 above), the remedy is not, for the left-behind parent, the equivalent of the Abduction Convention’s mandatory summary return.

For all these reasons, the principled answer to the question whether repudiatory retention is possible in law is that it is. The objections to it are insubstantial whereas the arguments against requiring the left-behind parent to do nothing when it is clear that the child will not be returned are convincing and conform to the scheme of the Abduction Convention. The remaining question is what is needed to constitute such repudiatory retention.

As with any matter of proof or evidence, it would be unwise to attempt any exhaustive definition. The question is whether the travelling parent has manifested a denial, or repudiation, of the rights of the left-behind parent. Some markers can, however, be put in place.

The judge ([2016] EWHC 3535 (Fam)) held that there was no concept of repudiatory retention known to the law. But he helpfully addressed the facts on the hypothesis that he was wrong about that. He held that the application to the immigration authorities made on 4 November 2015 did not amount to such a repudiatory retention, because although it was concealed from Father, something had to be done to regularise the stay of the children once it was to last more than their six month visas permitted. Father, he held, could not properly have objected to such regularisation, even if Mother feared that he might have tried.

There can be no doubt that the judge significantly misdirected himself here. It was not the application for permission to stay which was potentially significant. It was what was said, in support of it, about Mother’s intentions. Of course it was said by her solicitors, but if it showed that by that date she had determined that “the children’s centre of life is, and will be, in the UK” indefinitely, then it would be capable of being an objectively identifiable manifestation, made to an official third party, of her repudiation of Father’s rights of custody, and of the fact that thereafter her retention of the children in the United Kingdom was not in accordance with the arrangement she had made with him, but in defiance of it.

However, the question which matters is not whether the judge made this error, but whether it affected his conclusion that Mother had not, before the expiry of the agreed year (which he determined was at the end of June 2016) made any act of repudiatory retention.

The judge went on to examine Mother’s state of mind. He found that she vacillated in what she meant to do. He had seen her examined and cross-examined, and it is clear that he believed her when she said that as at both November 2015 and February 2016, she had not yet made up her mind. In February she had told Father only that she would not be returning in May (when the year would not, on the judge’s findings, have expired). He attributed her uncertainty in part to anticipation of “harassment” from Father. He then directed himself that even though she gave evidence that by the end of April 2016 she had resolved not to return, that could not be a date for repudiatory retention because it was too imprecise and thus inconsistent with the In re H; In re S rule that retention must be a definite occurrence rather than a continuing process. To the extent that he relied on imprecision he was, again, clearly wrong. There is, as explained above, nothing in In re H; In re S which prevents a court from saying that retention had occurred not later than the end of April. But what does prevent there from being a repudiatory retention in April is that Mother’s internal thinking could not by itself amount to such. If she had had such an intention in November, the application to the immigration authorities would have been capable of amounting to an objective manifestation of her repudiation, but the judge believed her when she said that she did not. It was open to him to believe her or not to believe her about this. He saw her and this court has only a transcript. It does not provide nearly sufficient basis for overturning his decision. His error about the potential significance of what was said to the immigration authorities in November is not inconsistent with his yet believing the witness whom he saw when she said that she had not then (or until April) made up her mind to stay.

These findings need to be considered alongside the judge’s decision as to the habitual residence of the children. He reviewed a body of evidence from Mother, relatives, neighbours and the playschool manager, to the effect that the children were, by the Summer of 2016, firmly integrated into the social and family environment of the part of England in which they had lived for a year, and, in the case of the younger child, for somewhat longer than he had lived in Australia. By reference to the decision of Hayden J in In re B (A Child) (Custody Rights: Habitual Residence) [2016] EWHC 2174 (Fam); [2016] 4 WLR 156, he directed himself correctly as to the test of habitual residence and the factors relevant to the integration necessary to establish it. He found that the children were, by the time of their otherwise wrongful retention at the end of June 2016, already habitually resident in the United Kingdom, so that the Abduction Convention could not apply to call for a mandatory summary return. He expressed the view that they had probably become habitually resident in England “much earlier” than June 2016.

There is no basis in law for criticising the judge’s decision as to habitual residence. His remark that it was “arguable” that the children had established habitual residence by the time of the November application to the immigration authorities may well be going too far, for at that stage they had been in the United Kingdom only since May, a period of about six months, but that remark does not alter the propriety of his decision as to June 2016, by which time more than a year’s residence had passed, during which the children had clearly become integrated parts of English life.

For my part, I recognise the force of the contention that the judge’s error about the potential significance of what was said at the time of the November application to the immigration authorities infected his decision that there was no combination of intention not to return and outward manifestation of that decision until the following summer. But for the reasons given above I conclude that that infection did not in fact take place. It follows that by the time the children were retained in the United Kingdom inconsistently with Father’s rights of custody they had become habitually resident here. That being so, the application under the Abduction Convention cannot succeed. The consequence is that Mother’s appeal against the order of the Court of Appeal must succeed, whilst Father’s cross-appeal in relation to the finding as to habitual residence must be dismissed.

There is much in Lord Hughes’ judgment with which I agree. Like Lord Wilson (with whose proposed disposal I fully agree) I would have dismissed the appeal. There is perhaps a slight difference in emphasis between us, however, on the reasons that the appeal should be dismissed and, on that account, I add this short judgment.

When dealing with the effect of wrongful retention of a child by what has been described as a travelling parent, one can recognise that various factors are in play. One starts with the proposition that, in general, it should not be possible for a child to acquire or for a parent to bestow habitual residence after the time that wrongful retention begins. A strong imperative exists for discouraging travelling parents from the view that they can avoid the consequences of the Abduction Convention by concealing an intention to retain the child in the country to which they have travelled, on the pretext, for instance, of a holiday of fixed or limited duration. To insist that wrongful retention can only occur at the end of an agreed period of absence could lead to absurd results; would encourage dissimulation on the part of the travelling parent; and would permit habitual residence to be acquired by the perpetration of deception on the left-behind parent.

As against that, it is often difficult retrospectively to decide when wrongful retention began. It may be the outcome of a gradual change of attitude on the part of the travelling parent. Retention in the country travelled to may be acquiesced in by the left-behind parent, even if she or he suspects that the travelling parent may be in the process of forming an intention not to return the child to the country where she or he was habitually resident. If the child has formed relationships in the travelled to country and is well settled there (albeit as a result of the travelling parent’s covertly formed intention not to return him or her) do the best interests of the child obtrude on the question of where her or his habitual residence should be found to be?

No final answers to these potentially difficult questions need be given in the present appeal. I raise them solely to illustrate the extremely trying problems that can arise in this fraught area.

How is the fact (and the time of onset) of wrongful retention to be established? Clearly the intention of the travelling parent wrongfully to retain is needed. Must this be accompanied by some overt act or event by which the intention becomes manifest? Not without misgivings, I am prepared to accept that this is required. The reason for my misgivings can be explained by taking a simple but not, I suggest, fanciful, example.

Suppose a husband persuades his wife to allow him to take their children to his parents’ native country on the promise that he will return within a stipulated period. Days after leaving, he conceives a firm determination that the children and he will never return. He does not communicate this to anyone. Some months later, he takes action which clearly demonstrates that he has no intention of returning the children. Evidence emerges that this was his plan from the outset. Is the period between his first determining not to return the children and the later “event” reckonable in the assessment as to whether they have acquired habitual residence in the country of their paternal grandparents? If we say that the retention only becomes wrongful when the intention of the retaining parent becomes manifest, how is the claim by the father in my example that the children have become habitually resident in his parents’ country to be resisted?

Again, however, this conundrum does not require to be solved in the present appeal and, having expressed my misgivings about the notion that some manifestation of the wrongful retention is required, I say no more about it.

For the reasons given by Lord Hughes and Lord Wilson, the judge ([2016] EWHC 3535 (Fam)) was wrong to hold that the law did not recognise repudiatory retention. His examination of when such a wrongful retention might have occurred (if, contrary to his view, the concept exists in law) appears to have been coloured by that primary finding, for he concluded that it had not arisen in this case at all. That finding simply cannot be reconciled with his statement in para 80 that “as the months went by, the mother gradually came to the conclusion that she and the children should remain in England. She had reached that conclusion by around April though it was not communicated to the father”. And this, notwithstanding that he had earlier said, at para 62, that a “finding that there was a wrongful retention on some unspecified date in April 2016 … is too imprecise.”

The opportunity for a firm finding as to the precise timing that an intention was formed is, in the nature of things, unlikely to be always possible. Intentions are formed over days, weeks or even years. Because it is not possible to make a positive finding of the date on which it had been formed is not a reason for not making a finding as to the time by which it had been formed. And indeed Judge Bellamy appears to have done precisely that when he said in para 80 that the mother had decided by “around April” that the children should remain in England.

The judge, having made that finding, was obliged to consider whether the children’s habitual residence had been established in England by April 2016. He did not do that. On that account alone, his decision cannot be allowed to stand, in my opinion. It is impossible to say that, if he had recognised the true implication of his statement that the mother had, by April 2016, formed the intention not to return the children, he would nevertheless have decided that habitual residence in England had by then already been established. In the absence of a finding to that effect, or alternatively the inevitability of such a conclusion, it is quite impossible to conclude that the habitual residence of the children had changed at a time which would displace the father’s rights under the Abduction Convention.

There is a more fundamental problem with the judge’s judgment. This concerns the communications to the Home Office in November 2015. In the letter from the mother’s solicitor, it was asserted that she had been advised not to return to Australia; that it was necessary that she remain in England “to safeguard herself and her children”; and that there was no doubt that “the children’s centre of life is and will be in the UK”. At paras 53 and 59 of his judgment, the judge dealt with the application for British citizenship in the following terse passages:

From these passages, two reasons for the judge’s conclusions can be discerned. First, the circumstance that the Home Office did not have to exercise a discretion meant that any misleading or inaccurate information in the letter should be discounted or ignored. Secondly, the fact that the father knew that something would have to be done to allow the children to remain in the United Kingdom after 5 November 2015 eliminated any possibility of the mother having wrongfully retained the children from that date.

Neither reason is sustainable. More importantly, the conclusions that he reached on those matters deflected the judge from recognising and considering the significance of the evidence provided by the November 2015 correspondence as to the mother’s intention at that time. The failure to give proper consideration to that evidence fatally undermines the conclusion reached by the judge as to the time at which the mother had conceived the intention to retain the children in England. In turn, this extinguishes the basis for his decision that the wrongful retention did not begin until June 2016 and that, by that time, the habitual residence of the children was England.

Why was the judge wrong to decide that, because the Home Office did not have to exercise a discretion, any misleading or inaccurate information in the letter should be discounted or ignored? Because this was nothing to the point. The significance of the letter in the context of these proceedings was its potential to provide an insight into what the mother’s intention was at the time that it was written. The purpose of the letter, the result that it sought to achieve, was entirely incidental to that critical consideration. The importance of the letter bore on the question of what the mother’s sentiments about the retention of her children in England were at the time of its dispatch. What it sought to persuade the Home Office of was entirely irrelevant to that question. But the judge dismissed the letter as a potential source of evidence on that central question. Until that question is addressed, the conclusion that the mother had not formed any intention wrongfully to retain the children in England in November 2015 is simply insupportable.

Likewise, the fact that the father knew that something would have to be done in November 2015 to ensure that the legal entitlement of the children to remain living in England was preserved, has no direct bearing on the question whether the letter from the mother’s solicitor showed that, as early as that date, the mother had decided that she would not return the children to Australia. The contents of the letter certainly suggested that that was the case. As already observed (in para 11 above), it had said that she had been advised not to return to Australia; that it was necessary that she remain in England “to safeguard herself and her children”; and that there was no doubt that “the children’s centre of life is and will be in the UK”. What the judge should have asked himself was, “is it conceivable that such a letter would be sent if the mother had not already decided that she and the children would not return to Australia?”. Instead, he elided that question by concentrating on the circumstance that the husband must have known that the mother would have to do something to regularise the children’s continued stay in England.

The important question was why the letter was couched in the terms that it was, if it did not reflect the mother’s settled intention to remain here. That question was never asked by the judge and it has not been possible to address it since. It needs to be asked and satisfactorily answered before any conclusion as to the mother’s intention in November 2015 about returning her children to Australia can be reached. That is why, in my opinion, remittal of the case for a proper hearing is unavoidable.

I respectfully agree with the exposition of law in the judgment of Lord Hughes. I disagree with him only when, from para 52 onwards under the heading “This case: the judge’s decision”, he reaches the conclusion that the mother’s appeal should be allowed.

I consider that this court should have dismissed the mother’s appeal.

The trial judge (“the judge”) held that the law did not recognise a repudiatory retention and that the mother’s retention of the children in the UK became wrongful only on 28 June 2016, which he found to have been the agreed date for their return to Australia.

The judge added, however, that, even if the law did recognise a repudiatory retention, he did not consider that it had arisen in the present case, whether in November 2015 or in April 2016 or at all.

As Lord Hughes has explained, the Court of Appeal was right to hold that the law does indeed recognise a repudiatory retention. The majority (Sharp and Thirlwall LJJ) proceeded to hold that the judge’s conclusion that in any event it had not arisen in the present case had been flawed; and they ordered that the case be remitted for further inquiry in that regard, particularly in relation to circumstances in November 2015.

In my view the majority were right to order that the possibility of a repudiatory retention, particularly in November 2015, required further to be explored. It required further to be explored by reference in particular to the mother’s intention; to the need for some objectively identifiable act of repudiation; and to whether, immediately before any repudiatory retention, the children had already acquired their habitual residence in the UK.

Although, like the majority in the Court of Appeal, I will focus principally on circumstances in November 2015, I wish briefly to address the possibility of a repudiatory retention of the children on the part of the mother in April 2016.

The judge found:

So why was there no repudiatory retention in April 2016? In para 55 above Lord Hughes explains

More importantly, however, the majority in the Court of Appeal were in my view right to set aside the judge’s finding that the mother’s intention to retain the children beyond 28 June 2016 arose only by around April 2016. For he did not grapple with evidence which seemed clearly to point to her having developed that intention by November 2015.

This evidence was the letter dated 4 November 2015 from the mother’s solicitor to the Home Office, which accompanied her applications on behalf of the children to be registered as UK citizens.

The context was that the children had entered the UK on 5 May 2015 pursuant to visitors’ visas due to expire on 5 November 2015. In the light of the father’s agreement that they could remain with the mother in the UK until 28 June 2016, it was necessary for their visas to be extended for almost eight months. But the regularisation of their stay in the UK for that extended period could have been achieved without their becoming UK citizens. So the mother’s applications for them to be registered as UK citizens called for an explanation. Her failure to notify the father in advance that she was making the applications also called for an explanation and, in cross-examination, it received one: she explained that she believed that he would have obstructed them.

To her statement in answer to the father’s application, the mother exhibited her solicitor’s letter dated 4 November 2015. In the letter the solicitor said:

The terms of the letter therefore appear to be entirely inconsistent with an intention on the part of the mother to return with the children to Australia in June 2016 or at all.

In the body of her statement the mother said that her decision not to return the children to Australia in June 2016 had developed over time and had not arisen long before that date. But she made no comment upon the content of her solicitor’s letter to the Home Office. She did not say that any part of it had been written without her approval or was untrue. On the contrary she said that her solicitor had been “utterly clear that there was nothing wrong or deceptive” in the applications for citizenship, being an assertion with which she seems there to have associated herself.

In cross-examination the mother was taxed, albeit perhaps in terms too general, about the content of the solicitor’s letter. She agreed that it did not indicate that she and the children would return to Australia in June 2016. She denied that, as at the date of the letter, she had formed an intention to stay with them indefinitely in the UK but, whether in re-examination or otherwise, she offered no explanation for what her solicitors had said.

In my view the content of the solicitor’s letter dated 4 November 2015, in support of applications for the children to acquire UK citizenship, represented a major obstacle to any finding that the mother had not by then intended to keep the children indefinitely in the UK. Before making any such finding, the judge was obliged to weigh that evidence and, on some basis or another, to explain it away. But, apart from an early reference to “any misleading or inaccurate information set out in that letter”, he did not address its content in any way. He said simply:

With respect to the judge, he was there missing the main point and was indeed making an unconvincing subsidiary point. His crucial finding about the mother’s intention in November 2015, not even expressly made but to be inferred from his reference to “around April” 2016, was flawed; and the majority in the Court of Appeal were correct to order that inquiry into it should be conducted again.