Evasive Manoeuvres: Strasbourg, the Hague Child Abduction Convention and the Absolute Prohibition on Ill-Treatment

Abstract

This article identifies a troubling omission on the part of the European Court of Human Rights, when it considers cases involving the Hague Convention on the Civil Aspects of International Child Abduction. Such cases often involve one parent, who has abducted the child, alleging that their actions were necessary to protect themselves or the child from violence by the other parent. The Court has avoided considering arguments based on Article 3 of the European Convention on Human Rights in such cases, despite the now established relevance of that Article to the risks of family violence. A change of course is needed, in order to articulate clear principles for the ECtHR’s own Hague cases, and reinforce the absoluteness of Article 3 protection elsewhere.

1. INTRODUCTION

To those seeking safety in a new country, international law problems of ‘norm interaction’ have real and dramatic consequences. International human rights law (‘IHRL’) offers protections against removal from the state of refuge, but treaties pursuing other objectives may conflict with or undermine them. EU law measures which distribute responsibility for processing asylum claims among member states - sometimes involving the transfer of asylum seekers to states which will offer them inadequate protection - are a prominent example. In that context the European Court of Human Rights (‘ECtHR’) has taken significant steps to reassert the primacy of the protective duty under the European Convention on Human Rights.¹ In another context, on which this article will focus, the ECtHR’s message has been less clear, with troubling implications for fundamental ECHR protections. That context is a specific point of interaction between the ECHR and the Hague Convention on International Child Abduction (the ‘Hague Convention’).² The norms at issue are as follows:

• (i) the requirement under the Hague Convention to promptly return a child who has been removed (most often by one of their parents, from the other) from their state of habitual residence, and taken to another state; and

• (ii) the ECHR ‘non-refoulement’ obligation, which arises by implication from Article 3 of the ECHR, enshrining the right to freedom from torture and inhuman or degrading treatment or punishment. That obligation prevents states from returning individuals to another state where they face a ‘real risk’ of such ill-treatment.

Hence, if a child is taken to an ECHR state, and the return of the child from that state is then sought under the Hague Convention, the state must consider both norms. The tension between them arises where allegations are made by the ‘taking’ parent, who removed the child, that the ‘left-behind’ parent’s violence (against the taking parent or child) necessitated the child’s removal.

While the Hague Convention provides exceptions from its general requirement of prompt return, they are narrow and must be assessed quickly. The ECHR non-refoulement obligation, on the other hand, normally requires a thorough assessment of any such alleged risks before a person can be returned to their state of origin. It accords a high priority to the ‘absolute’ right to freedom from ill-treatment under Article 3 of the ECHR. While exceptions to the return obligation built into the Hague Convention can accommodate non-refoulement in many cases, there is significant potential for conflict in other cases between the expeditious return objective of the Hague Convention and the protection objective of Article 3.

The resolution of that tension is of course of greatest importance to the divided families at the heart of any Hague Convention case, if the child is within the jurisdiction of an ECHR state. But the issue has broader significance. Firstly, other bodies of IHRL establish similar ‘non-refoulement’ obligations prohibiting the return of individuals to face danger in another country: these include the Refugee Convention,³ and customary law.⁴ Accordingly the potential for tension with non-refoulement obligations can arise for any party to the Hague Convention,⁵ and in the United Kingdom there is evidence of an increasing number of cases where the two overlap.⁶ More broadly still, the interaction here is evidently relevant, at least by analogy, to other areas in which states may seek to establish arrangements under international law to ‘contract out’, to other states, their protective responsibilities under IHRL (in particular towards those seeking asylum).⁷

Given the lack of an international tribunal to consider individual claims under the customary or Refugee Convention non-refoulement obligations, the ECtHR might be expected to play an especially important role in addressing the Hague Convention/non-refoulement tension. In fact, although the ECtHR has considered many cases connected to Hague return proceedings, it has been silent on the specific issue of non-refoulement in those cases. Indeed it has been almost silent on the applicability of Article 3 in any form. Similarly, while the relationship between non-refoulement and the Hague Convention has received only limited scholarly attention,⁸ in respect of the ECtHR’s jurisprudence it has received almost none.⁹ In the scholarship that considers the ECtHR’s jurisprudence on the Hague Convention in other respects, ¹⁰ non-refoulement is absent, presumably reflecting its absence in the jurisprudence itself.

The ECtHR, and commentators, instead primarily consider such cases by reference to the Article 8 rights to family life enjoyed by the taking parent, the child, and also the left-behind parent. The flexibility associated with that right more readily accommodates the expeditious return objective of the Hague Convention,¹¹ but leaves a troubling silence on the crucial question of how the ECHR’s more specific obligations to protect against ill-treatment operate in the Hague Convention context. That question, as was noted above, assumes additional importance as states increasingly explore options to ‘outsource’ their compliance with IHRL’s protective obligations to those whose flight from danger takes them across international boundaries.

While Article 8 rights may yield in the face of other international rights, obligations and objectives of the relevant states, ‘absolute’ Article 3 ones should not.¹² Such ‘absolute rights’, as Mavronicola has argued in her recent in-depth study of Article 3, ‘amount to entitlements that are non-displaceable’.¹³ Although the state’s obligations to secure Article 3 are not unlimited, the permissible limits relate to the state’s capacity to protect the individual in question, and cannot be altered to allow for balancing against ‘extraneous considerations’ such as national security.¹⁴

The integrity of Article 3, and the absolute status it is accorded within the ECtHR’s jurisprudence, is the central reason for changing course. It may also be the reason that its application has so far been avoided by the ECtHR. John Finnis has argued that ‘logic requires that a norm’s absoluteness narrow, not broaden its scope as an absolute’, and (reflecting that logic) in his view the Article 3 non-refoulement obligation is too broad to be appropriately treated as an absolute norm.¹⁵ Since the ECtHR nonetheless does consider the non-refoulement obligation to be absolute, it should recognise, and address, the difficult interaction of the Hague Convention’s objectives with the jurisprudence, which the ECtHR has developed on absolute Article 3 protections, rather than succumb to the temptation to avoid considering the Article 3 question entirely. It should expressly apply Article 3, and uphold its absolute status, when considering Hague cases. Such an approach is desirable, important and realistic for four reasons.

Firstly, and most importantly, it affirms the broader principle that Article 3’s protections, for refugees and others, cannot be displaced by other international law rules, which contemplate moving protective responsibility (together with the affected individuals themselves) to another state. Omitting to address Article 3 at all, by contrast, undermines that principle and the absolute status of that Article generally. Secondly, the present silence leaves uncertainty. As well as affecting litigants before the Court, that uncertainty clouds questions, which municipal courts in ECHR states must address, in respect of the interaction between Hague proceedings and non-refoulement obligations. There are two particular areas—the procedural rigour with which risks are assessed, and the protection of the taking parent—in which there is potential for conflict, and a need for clarity should such conflict arise. If it were clearer, the ECtHR’s approach could, by way of analogy, exert influence in non-ECHR states, which are party to the Hague Convention; and indeed in proceedings outside the Hague regime where states have nonetheless opted to follow equivalent rules.¹⁶

Thirdly, in many cases, recognising the applicability of the non-refoulement obligation will not result in conflicts with, nor even undermine, the Hague Convention and its objectives; that will primarily be the case because the Hague Convention incorporates exceptions from the obligation to return.¹⁷ Such recognition will nonetheless be helpful in those cases: it will confirm the correct approach, and support Article 3’s absolute status more generally, at no cost to the Hague regime. Even in the more difficult areas mentioned in the above paragraph, conflict could often be avoided provided measures are taken by states to ensure that rigorous risk assessments are undertaken expeditiously; again, recognising Article 3’s applicability will support clarity as to what is required, and exert pressure on states to deliver it.

Finally, recognition of the non-refoulement obligation’s applicability is not only logically required by the ECtHR’s jurisprudence on that obligation, but it is consistent with the substance of a significant (and more recent) part of the ECtHR’s jurisprudence on the Hague Convention. While the ECtHR has not explicitly applied Article 3, its decisions have moved away from giving priority to the Hague Convention at the expense of ECHR rights, and of proper examination of alleged violence. Recognition of Article 3’s role would help to confirm the correct path, as between the contradictory routes offered by the existing ECtHR Hague Convention jurisprudence.

For all of these reasons, it is important that the ECtHR should now move to openly articulate the correct approach. The argument will proceed as follows. Section 2 will introduce the ECtHR’s approach to Hague cases, highlighting the Court’s refusal to consider Article 3 in the cases where it has been raised by an applicant. Section 3 then explains why Article 3, and specifically the non-refoulement obligation, should properly be regarded as applicable to Hague cases. In Section 4, the likely impact of the non-refoulement obligation on the operation of the Hague Convention will be discussed, observing both the areas in which the two regimes are compatible and the possible areas of tension. It will argue that in both the areas of harmony, and of tension, Article 3’s applicability should be recognised. Focusing more specifically on the areas of tension, Section 5 then identifies the trend in more recent ECtHR Hague jurisprudence towards an approach more amenable to accommodating Article 3, while Section 6 identifies the arguments based on which that shift (and the more explicit approach advocated here) can be justified, and outlines the Conclusion.

2. ECHR HAGUE CASES AND THE AVOIDANCE OF ARTICLE 3

The factual circumstances in which a Hague case arises inevitably implicate ECHR rights. The left-behind parent will have sought the return of the child, abducted from their state of habitual residence (against the laws of that state), from the authorities of the state to which the child has been taken. The taking parent is likely to have invoked exceptions under the Hague Convention from the general requirement of return, based on the interests of the child. Against that factual background, two ECHR rights are likely to have particular relevance: Article 8, which will always be relevant, and Article 3, which will be relevant if the exceptions invoked are based on apprehended Article 3 harm.

A. The Article 8 Jurisprudence

Both of the parents involved in the Hague dispute, and the child, have a right to respect for family life under Article 8 (the right ‘to respect for private and family life’). The proceedings, determining whether the child should be returned or not, evidently have a significant impact on that right, for all concerned.¹⁸ The ECtHR has no jurisdiction to determine whether state conduct is or is not compliant with international law rules external to the ECHR, such as those imposed by the Hague Convention, nor to interpret treaties other than the ECHR.¹⁹ However, to perform its function, in child abduction cases where return is sought from a state, which is party to both the ECHR and the Hague Convention, any decision by the ECtHR necessarily has relevance to the Hague Convention. If the Hague Convention were ignored, and the ECtHR found that the return of the child was not permitted under the ECHR, the state could be confronted with a conflict in its obligations, if, on the facts, the Hague Convention requires return. It would be reasonable to expect the ECtHR to make some effort to avoid such an outcome (although not to do so at all costs).²⁰

Article 31(3)(c) of the Vienna Convention on the Law of Treaties,²¹ which the ECtHR applies to its interpretation of the ECHR,²² offers a vehicle for addressing potential conflicts. It requires that a treaty be interpreted taking into account ‘relevant rules of international law applicable in the relations between the parties’. It thus requires at least some incidental interpretation of other instruments, in order to interpret the ECHR,²³ providing a possible basis for rejecting interpretations of the ECHR which would require the state to violate another rule of international law.²⁴ Treaty rules ratified by all ECHR states unambiguously fall within Article 31(3)(c) of the VCLT, and so must be taken into account in interpreting the ECHR. The ECtHR has frequently affirmed this requirement, and expressed the need where possible to harmonise diverging obligations.²⁵ Indeed it has also done so when considering treaties not binding on all ECHR states,²⁶ and it is on this basis that consideration of the Hague Convention, to which all but two ECHR states are party, apparently occurs.²⁷ Accordingly the Hague Convention has been extensively discussed by the ECtHR, and has significantly affected its jurisprudence.

The ECtHR’s earliest jurisprudence regarding the Hague Convention arose in situations where left-behind parents alleged failures by ECHR states to expeditiously return children under the Hague Convention. The ECtHR held that the Article 8 rights to family life of those left-behind parents imposed a positive obligation on states to, in effect, comply with the Hague Convention, by ensuring the prompt return of the child.²⁸ A positive obligation existed under Article 8 to reunite the left-behind parent and child, and had to be interpreted ‘in light of’ the Hague Convention.²⁹ Indeed even where the state is not party to the Hague Convention, it must provide some alternative framework to discharge the relevant positive obligation.³⁰

The importance of the obligation to return was also evident when the ECtHR came to consider applications by taking parents, contesting ECHR states’ decisions to effect Hague returns.³¹ In those cases, the ECtHR has also focused on Article 8, considering whether ordering return would interfere disproportionately with Article 8 rights of the taking parent and the child (weighed against Article 8 rights of the left-behind parent, and potentially of the child to a reunion with them). However, sometimes, taking parent applicants have sought to invoke Article 3 in addition to Article 8 rights.

B. The Article 3 Question

Article 3’s prohibition on torture and ‘inhuman or degrading treatment or punishment’ can be relevant if the actions taken by the returning state themselves amount to such ill-treatment. In Maumousseau and Washington v France (‘Maumousseau’),³² the applicant alleged that the trauma caused by the French authorities’ enforcement of a Hague return order by way of police intervention at the child’s nursery school violated Article 3. The ECtHR considered that argument under Article 8 (on which basis it failed).³³ It is not these cases, however, but those in which the possibility is of ill-treatment by other parties—particularly the left-behind parent—on which this article will focus.³⁴ The point arises because the behaviour of actors other than the state can amount to ill-treatment prohibited by Article 3. Accordingly, Article 3 does not only impose a ‘negative obligation’ on the state to refrain from inflicting such treatment itself. In its Article 3 ‘positive obligations’ jurisprudence, the ECtHR has outlined various state obligations to take positive actions to prevent ill-treatment by private actors.³⁵ These require action to investigate and punish ill-treatment where it has occurred, reasonable measures to prevent it where a ‘real and immediate risk’ of it occurring is (or should be) identified, and an appropriate legal and regulatory framework in place to protect against such treatment.³⁶

The treatment covered by Article 3 has evolved, increasing its relevance to the allegations often made against left-behind parents in Hague cases. The ECtHR has observed that ‘the prohibition of ill-treatment under Article 3 covers all forms of domestic violence without exception’.³⁷ Regarding domestic corporal punishment of children, which remains legal in many states (including a significant number of parties to the ECHR),³⁸ long-standing jurisprudence established the need for effective prohibition of more severe forms of corporal punishment.³⁹ While previous jurisprudence also suggested that less severe forms of corporal punishment would not trigger Article 3,⁴⁰ the ECtHR now appears to consider that domestic corporal punishment is ‘invariably degrading’,⁴¹ and thus contrary to Article 3, although its stance still appears ‘somewhat hesitant’ as to the extent of state obligations to comprehensively prohibit it.⁴²

It is therefore unsurprising that some taking parents have invoked Article 3, in arguing that Hague proceedings should not require the return of the child to an allegedly abusive left-behind parent. In MR and LR v Estonia,⁴³ the taking parent argued that she, and the child, would face a risk of violence from the father if returned from Estonia to Italy. However when she invoked Article 3 (as well as Articles 6, 8 and 14) the ECtHR, as ‘master of the characterisation to be given in law to the facts of the case’, chose to consider the case only under Article 8.⁴⁴ In OCI and others v Romania,⁴⁵ in addition to pleading Article 8, the taking parent argued that the return of her children to their father in Italy, ordered by the Romanian courts, violated their Article 3 rights, relying primarily on his record of using corporal punishment against the children.⁴⁶ Again, the ECtHR declined to rule on the Article 3 claim, although as will be seen, with important differences in reasoning and outcome.⁴⁷

The unwillingness to address Article 3 arguments can also be seen in two other recent cases. In Nedelcu v Romania, the applicants argued that a Romanian court had ordered return without proper consideration of the facts, including allegations of violence by the left-behind parent, and Article 3 was among the grounds for the application.⁴⁸ As it had in MR and LR, the Court simply asserted that it could choose to consider the case under Article 8 alone, and did so.⁴⁹ Finally, in YS and OS v Russia, the ECtHR considered Article 3 arguments primarily based on the armed conflict-related risks, which the child would face if returned by the Russian authorities to conflict-affected areas of Ukraine.⁵⁰ The ECtHR opted to consider those risks (very different to those generally alleged in Hague cases) under Article 8, and declined to rule under Article 3.⁵¹

The ECtHR’s reticence to engage with Article 3 arguments in these cases is understandable but not, it will be argued here, justified. It is necessary to explore further the operation of Article 3 to understand the significance of its omission from these decisions, and to identify the concerns, which may have caused the ECtHR to avoid open engagement with the issue.

3. ARTICLE 3 AND THE NON-REFOULEMENT OBLIGATION IN ECHR HAGUE CASES

To begin, an important distinction must be drawn between Article 3 and the other rights potentially at issue, in terms of their flexibility. Article 8 rights are ‘qualified’, allowing restrictions, which are adequately prescribed by law and are proportionate to a legitimate aim.⁵² Compliance with Hague obligations might itself be considered a ‘legitimate aim’ justifying limitation, and, more substantively, the Hague Convention’s aims of deterring abductions,⁵³ and the protection of the Article 8 rights of the left-behind parent, could also be. The ECtHR has accordingly had little difficulty reconciling Article 8 with the Hague Convention, invoking Article 31(3)(c) of the VCLT in support of its approach.⁵⁴

Article 3, on the other hand, is ‘absolute’ in its prohibition of ill-treatment which meets the Article 3 threshold: it does not admit restrictions in pursuit of legitimate aims. That said, while Article 3 absolutely precludes resort by the state to such treatment, there is flexibility in respect of the ‘positive’ obligations to prevent ill-treatment by third parties; the ECtHR recognises that the state’s capacity to identify and avert such treatment is not unlimited.⁵⁵ For example, as was noted in the previous section, the central positive obligation to address specific threats from others is framed as a duty to take ‘reasonable’ measures to prevent a ‘real and immediate’ risk of threats of ill-treatment, which the state is or should be aware of.⁵⁶ If that positive obligation were the correct framework for approaching Hague cases, then its limitation to requiring measures which are ‘reasonable’ could provide a basis for reconciliation with the Hague Convention. It might be argued that refusing or delaying transfer goes beyond what is ‘reasonable’, since it would breach another international obligation (the Hague Convention). Whether, and to what extent, the Article 3 positive obligations might, effectively, yield to other international law obligations in this way is an important question with broader ramifications.⁵⁷ However when the applicants in OCI invoked Article 3 positive obligations in their argument,⁵⁸ the ECtHR declined to decide the point. Instead, it drew indirectly upon its Article 3 positive obligations jurisprudence, on the unacceptability of corporal punishment, to inform its Article 8 decision that the return of the child was unlawful.⁵⁹ It then declined to decide the Article 3 question, on the basis that it had addressed the relevant issues under Article 8.⁶⁰

In fact, had the ECtHR decided the Article 3 question, it should have recognised that the (potentially) flexible positive obligation is not the appropriate vehicle for assessing Hague returns in any event, since Hague cases involve the transfer of an individual (the child) to another state. The removal of a person from one state to another normally calls for assessment under Article 3’s implied obligation of ‘non-refoulement’, rather than the positive obligations outlined above. Under that obligation, the removal will be prohibited if there are substantial grounds for believing that the relevant person would face a real risk of treatment contrary to Article 3.

The non-refoulement obligation was first articulated in Soering v the United Kingdom (Soering),⁶¹ a case which considered, and prohibited, the applicant’s intended extradition by the UK to the United States, where he would have faced a risk of ill-treatment within the criminal justice system of Virginia. Since then, it has been established that the obligation applies not only to extraditions, but also to deportations,⁶² and not only to ill-treatment by the authorities but also by non-state actors.⁶³ Non-refoulement has not been applied, nor apparently argued,⁶⁴ in Hague return cases before the ECtHR. However, the ECtHR’s framing of the rule, and insistence upon its comprehensive protection, do not favour treating Hague returns differently from other removals:

whenever substantial grounds have been shown for believing that an individual would face a real risk of being subjected to treatment contrary to Article 3… if removed to another State, the responsibility of the Contracting State to safeguard him or her against such treatment is engaged in the event of expulsion.⁶⁵

A closely analogous question has been expressly considered by the UK Supreme Court, in respect of the non-refoulement obligation enshrined in the Refugee Convention. In that context, in 2021, the Supreme Court dismissed an intervener’s arguments ‘that a return order was not a form of “refoulement” but rather the court was determining whether the child wished to remain in the United Kingdom’.⁶⁶ The court rightly observed that ‘the substantive effect of a return order... is that the child is being returned to the country from which they seek refuge’, and as such the non-refoulement obligation was applicable.⁶⁷

Although the ECHR non-refoulement obligation was not at issue in that case, the same logic is applicable; in both scenarios, the non-refoulement obligation offers protection against ‘serious and irreparable’ harm,⁶⁸ risked by removal from the state to a location in which it can no longer offer effective protection. Previous obiter comments by the UK Supreme Court, and a ruling of the High Court, also support the conclusion that the ECHR non-refoulement obligation applies in principle to Hague returns.⁶⁹ That outcome is logical for an ‘absolute’ right. As Mavronicola has observed:

An approach which varies the meaning of Article 3’s terms simply on the basis of where the treatment is administered, or which (partly) displaces the prohibition on torture or inhuman or degrading treatment found in Article 3 on the basis of extraneous considerations (such as the desirability of extradition), contradicts the absoluteness starting point.⁷⁰

Accordingly, it appears in principle correct that the non-refoulement obligation should apply to Hague transfers. If that is the case, it is necessary to consider the impact of the obligation’s applicability, and the extent of departure that it might require from the Hague Convention and from the ECtHR’s existing approach. The following sections will undertake that analysis.

4. THE QUESTION OF IMPACT

As was noted above, qualified Article 8 rights can readily accommodate limitation in favour of the Hague objective of prompt return. Mavronicola’s comment above, in excluding the possibility of giving weight to ‘the desirability of extradition’, emphasises that upholding the consistent applicability of the non-refoulement obligation is more likely to come at a cost. In contrast to Article 8, under the absolute Article 3, if a ‘real risk’ exists the transfer is prohibited regardless of any competing interests.⁷¹ States’ powers to deport non-nationals, as well as their ability to co-operate on criminal justice through extradition and other transfers of suspected criminals, are necessarily limited as a result.⁷² Applying non-refoulement to Hague transfers has the potential, likewise, to compromise the Hague Convention’s objective of expeditious return.

That said, the Hague Convention does not require return of the child ‘at all costs’,⁷³ and makes its own provision for the interest—averting risks of ill-treatment—which the non-refoulement obligation serves. Accordingly, to identify the impact of non-refoulement in this area, it is necessary to consider how far the applicability of non-refoulement either contradicts or supports the relevant Hague Convention provisions. The enquiry necessitates consideration of both the priorities of the Hague Convention as a whole and of the specific exceptions it incorporates. As will be seen, Article 3 has considerable potential to operate harmoniously with, and in support of, the Hague Convention, in some cases; however there is also clear potential for conflict in others, depending on the interpretations of the Hague Convention, which states adopt. Either way, on the argument made here and in the following sections, its applicability should be recognised.

A. The Hague Convention’s Objectives and Exceptions

The potential conflict between the Hague Convention and Article 3 arises because of the objective of the Hague Convention procedure. It aims to ‘secure the prompt return’ of the child to the ‘home’ jurisdiction, where any family dispute and issues regarding the conduct of either parent should then be resolved according to that jurisdiction’s laws. It also aims to deter unilateral removals of children contrary to those laws,⁷⁴ since it applies to ‘wrongful’ removals, which are in breach of custody rights under the laws of the state of habitual residence.⁷⁵ Thus, consistent with a long-standing approach in international family law instruments,⁷⁶ it seeks to identify which state should have responsibility for determining the correct outcome, rather than prescribing correct outcomes itself. That state should be the state of the child’s ‘habitual residence’.⁷⁷ In this framework, the priority is for the state to which the child has been taken to return them as quickly as possible.⁷⁸ Prompt return should allow for a quicker end to the disruption caused by the abduction (including to the relationship between the child and the left-behind parent), and also generally reduce any incentive to abduct arising from the prospect of more favourable treatment for a potential taking parent in another state.⁷⁹ The status quo ante should be swiftly restored.⁸⁰ However if the taking parent alleges that that status quo was characterised by violence on the part of the left-behind parent, the Article 3 obligation of the state to which the child was taken, to examine the risk fully and (if it is real) allow them to remain, could conflict with its Hague obligation to return the child promptly and without extensive court proceedings.⁸¹

The Hague Convention incorporates two provisions, which considerably reduce the extent of this potential conflict. One is Article 13(1)(b), which provides grounds to refuse return where ‘there is a grave risk that his or her return would expose the child to physical or psychological harm or otherwise place the child in an intolerable situation.’ Accordingly, in any case where a ‘real risk’ of Article 3 ill-treatment also amounts to a ‘grave risk’ in Hague terms, the Article 13(1)(b) exception offers a basis to uphold the non-refoulement obligation without any conflict with the Hague obligation.⁸² The need for a ‘grave’ risk under Article 13(1)(b) appears to establish a higher threshold than that which gives rise to non-refoulement obligations.⁸³ In theory, at least, that suggests the possibility that a child facing a risk, which meets the non-refoulement threshold, but is not ‘grave’, may be required to return, in breach of non-refoulement. However insofar as Article 13(1)(b) could leave a gap, resort can be made to Article 20. That Article allows refusal if return ‘would not be permitted by the fundamental principles of the requested State relating to the protection of human rights and fundamental freedoms’. While the rarity of successful invocations of Article 20 have prompted its characterisation as a ‘ghost’,⁸⁴ its potential in this area has been recognised by commentators.⁸⁵

Considering the cumulative effect of Articles 13(1)(b) and 20, the ECtHR’s omission to address Article 3 arguments appears somewhat surprising; seemingly, it need not lead to a significantly different result to the Hague Convention anyway. However it is troubling that Article 3’s applicability has not been affirmed, for two reasons, which will be explored in the following section.

B. The Impact of Omitting Article 3

One reason for concern is that, as Article 3’s ‘absolute’ status, especially in respect of expulsions, faces challenge,⁸⁶ the failure to affirm its consistent and comprehensive applicability is at best a missed opportunity. At worst, it risks undermining the absoluteness of protection associated with Article 3 generally, by (tacitly) developing an exception to the rule of non-refoulement. While two recent cases appear to imply that Article 3 is in principle applicable to Hague cases,⁸⁷ express affirmation of the applicability of the non-refoulement obligation remains lacking. If the essence of Article 3’s absoluteness is its non-displaceability,⁸⁸ then its tacit displacement in some cases is an important challenge, meriting more open remedy than its vague reinstatement by implication.

The other concern, of more practical significance for the outcome of Hague cases, is the uncertain legal position, which is left by the ECtHR’s failure to consider Article 3. The Hague Convention’s emphasis on expeditious returns may be given undue weight in determining cases, unless it is recognised that Article 3’s absolute protection is also at stake. Without that guidance, while some domestic courts may reach results, which mirror non-refoulement protection by applying the Hague Convention exceptions and Article 8, others may not.

The approach of the Romanian courts in OCI offers an example: the Bucharest Court of Appeal had not (in the ECtHR’s view) recognised that corporal punishment ‘cannot be tolerated’, and was contrary to the prohibition of domestic abuse.⁸⁹ The example is informative, since the danger of such inadequate consideration of corporal punishment risks, in the absence of reliance on Article 3, may be particularly relevant in Hague cases. An assumption that only more extreme or ‘excessive’ forms of corporal punishment should be considered as an obstacle to return has sometimes been a feature of Hague Convention proceedings at the domestic level in non-ECHR states.⁹⁰ Cultural relativism (in the form of deference to the allegedly normal standards of child raising in the state to which return is sought) may play a role.⁹¹ Drawing Article 3 and the non-refoulement obligation to the attention of national courts and litigants should help to clarify matters; in OCI, it was to Article 3 jurisprudence that the ECtHR turned to explain the correct position, albeit within its Article 8 decision.⁹²

Further, while OCI appears to resolve the cultural relativism question concerning corporal punishment, other questions remain. More specifically, the uncertainty regarding Article 3 leaves two important questions about the interaction of the Hague Convention and the non-refoulement obligation unanswered, giving neither litigants before the ECtHR, nor domestic courts, clarity as to how those questions would or should be addressed if they arise. They relate firstly to the rigour with which any contested allegations about the risk to the child may be considered, in light of the need for expedition in Hague cases, and secondly to the relevance of risks faced by the taking parent but not the child.

(i) The rigorous assessment question

The first question could be labelled as a procedural, rather than a substantive one, but it is fundamental. In the non-refoulement context, Article 3 requires an assessment of the risk of ill-treatment, which is ‘rigorous’, ‘adequate and sufficiently supported’.⁹³ Once ‘evidence capable of proving… substantial grounds for believing’ that there is a real risk of ill-treatment has been adduced by the applicant, the burden of proof shifts to the state to ‘dispel any doubts’.⁹⁴ Non-refoulement thus requires a precautionary approach to the assessment of risk. However, as was noted above, consistent with the Hague Convention’s broader priorities, the review of alleged risks permitted before a Hague return is ordered is quite different from the non-refoulement assessment. The exceptions to return at Articles 13(1)(b) and 20 of the Hague Convention must be interpreted in light of the Hague Convention’s object and purpose,⁹⁵ to ensure a rapid process and to have the merits of the underlying dispute resolved under the law of the state from which the child was removed. Accordingly the Article 20 exception is to be interpreted narrowly,⁹⁶ and the burden of proving a grave risk under Article 13(1)(b) is on the person resisting return,⁹⁷ in either case without a full review of the family’s circumstances by the courts considering the request.⁹⁸ Full decisions on the merits of disputes regarding family arrangements should generally be made in the state of habitual residence following the return.⁹⁹

In the context of considering the interaction of the Refugee Convention non-refoulement obligation with the Hague Convention, the English courts have noted the potential for the latter to be significantly undermined by the need to wait until asylum claims have been resolved, in some situations.¹⁰⁰ These assessments suggest that a significant gap can exist between the non-refoulement obligation (including its no less rigorous ECHR manifestation) and the Hague Convention, in terms of the required procedures for assessing risk. As will be seen in section (iii), however, it is in relation to the rigour of assessment regarding past violence against the taking parent (and the future risks to them or the child it may indicate) that the practice of the ECtHR gives greatest cause for concern. Before engaging in that discussion it is necessary to further explain the issue regarding the position of ‘taking parents’ in general.

(ii) The taking parent question

Regarding the position of the taking parent, formally speaking the non-refoulement obligation’s direct relevance is limited, since the state will not be ordering their return under the Hague process, only the child’s; the non-refoulement context can be distinguished in this respect. However looking to the ECtHR’s general emphasis on the need for ECHR rights to be secured in a manner, which is ‘practical and effective’,¹⁰¹ the possibility that requiring the child’s return is, de facto, forcing the return of the taking parent to a real risk of ill-treatment should be considered. That possibility is most clearly relevant if, for the purposes of assessing any risks to the child (whether or not related to Article 3), it is being assumed that the taking parent will return with the child. Such an assumption may remove one of the potential grounds for resisting return. In particular, if the taking parent is the main care-giver, returning the child without them could plausibly involve a grave risk of psychological harm or an ‘intolerable situation’ for the child under Article 13(1)(b) of the Hague Convention, and thus preclude their return entirely.¹⁰² If the parent’s return is assumed then the case for ordering the child’s return may thus be significantly strengthened.

Accordingly, to avoid a gap in non-refoulement protection, at least where the acceptability of ordering the child’s return depends on the taking parent also returning, a real risk of Article 3 harm to the taking parent should, logically, preclude a return order.¹⁰³ Otherwise, the Article 13(1)(b) assessment of the child’s position would be proceeding on an assumption that the taking parent will expose themselves to a real risk of ill-treatment—something, which the authorities cannot (by virtue of non-refoulement) force them to do. The non-refoulement obligation does not prevent the taking parent from choosing to return, but to order the return of the child—on the assumption that the taking parent will also return, to avoid an ‘intolerable situation’ for their child—appears to place heavy pressure on the parent to accompany the child and accept the risk to themselves. While allowing the taking parent to ‘veto’ the child’s return, by refusing to return themselves, would present obvious problems for the Hague Convention regime, a ‘real risk’ of Article 3 ill-treatment should be considered a compelling justification for such a refusal.

The Hague Convention offers limited protection in respect of taking parent risks. While recent guidance relating to the Hague Convention acknowledges the potential for evidence of domestic violence against the taking parent to be considered in assessing the risk to the child, it also emphasises that the focus is on whether such violence constitutes a ‘grave risk’ to the child: ‘[e]vidence of the existence of a situation of domestic violence, in and of itself, is therefore not sufficient to establish the existence of a grave risk to the child.’¹⁰⁴ It further emphasises that the Article 13(1)(b) exception is ‘only rarely upheld’ on the basis that ‘the taking parent cannot or will not return’.¹⁰⁵

(iii) The questions in practice

Although in principle the rigorous assessment and taking parent questions reflect distinct issues, in the ECtHR’s practice it is at the point of convergence between those questions that the risks to Article 3 protection appear greatest. If Article 3 can prevent return by reference to the risks faced by the taking parent, rigorous assessment of a credibly alleged risk to the parent would also be necessary. Such an assessment is likely to require investigation of alleged previous violence against that parent. Further, the alleged previous violence against the taking parent may often indicate a risk to the child as well; either a direct risk of being subjected to violence, or an indirect risk through witnessing it.¹⁰⁶ However requiring allegations of past domestic violence by the left-behind parent to be sufficiently examined to meet the requirements of Article 3 may delay return under the Hague Convention. The ECtHR has at times appeared willing to give priority to the Hague Convention’s expedition objective, by accepting domestic court decisions to curtail risk assessment, instead of examining allegations of past violence by the left-behind parent, which was not directed against the child, as indicators of future risk.

In Maumousseau,¹⁰⁷ the ECtHR clearly aligned its approach to the priorities of the Hague Convention, when considering taking parent arguments against return (which were based on the Hague Convention exceptions). Allegations of domestic violence by the left-behind parent (not inflicted upon the child) were among the arguments made by the applicant against the return of her child by the French authorities to the father in New York.¹⁰⁸ However the ECtHR upheld the objectives of the Hague Convention, and endorsed their efficacy for protecting the best interests of the child, emphasising the need for a strict interpretation of the Article 13 and 20 exceptions as a result.¹⁰⁹ Referring to Article 31(3)(c) of the VCLT,¹¹⁰ the ECtHR considered the domestic decision to return the child (based on a finding that the exceptions did not apply) a proportionate interference in the mother’s and daughter’s Article 8 rights to their own shared family life.¹¹¹

As to the potential Article 3 dimension, the allegations of past domestic violence by the father received little attention in the ECtHR, although a vigorous dissent by two of the judges considered that those allegations had not been adequately taken into account by the French courts.¹¹² The French courts, holding that the Hague Convention Article 13 exception was not applicable, had simply observed that the alleged violence had not been directed against the child.¹¹³ Without any discussion on that specific point, including whether alleged domestic violence by the father might also indicate risk to the child if she were returned to him, the ECtHR accepted their general conclusion.¹¹⁴ In doing so, it gave effect to the general principles it had elaborated regarding the Hague Convention’s priorities, instead of applying the kind of assessment associated with the non-refoulement obligation. It also implicitly accepted the readiness of the French authorities to ignore alleged violence by the left-behind father, which was not inflicted on the child. Two features of the context for the decision deserve mention in understanding its significance for Article 3. Firstly, Article 3 was not argued by the applicants, except in relation to the enforcement actions of the French authorities themselves.¹¹⁵ Secondly, as was discussed in Section 2, Article 3’s relevance to family violence has significantly increased over time. Accordingly the Maumousseau approach could perhaps be considered consistent with the Article 3 jurisprudence of its time, and thus in need of revisiting as that jurisprudence evolved, as and when Article 3 might be invoked.

One opportunity for revisiting arose in the MR and LR case, mentioned earlier,¹¹⁶ which offers a more recent example of the prioritisation of the Hague Convention, in circumstances which might have been expected to attract scrutiny on the basis of Article 3. It will be recalled that the applicants invoked Article 3 in that case, but the ECtHR chose to characterise the complaint under Article 8.¹¹⁷ The applicants alleged that they faced a risk of violence at the hands of the child’s father, if returned from Estonia to Italy. The taking parent submitted ‘written statements from a number of persons (her mother and other relatives, friends, colleagues and a teacher) describing the relations between the parties and [the taking parent’s] abusive behaviour’.¹¹⁸

In that case the ECtHR upheld the presumption, in favour of return without a broader determination of the merits of custody issues, and again accepted the need for expedition and a narrow interpretation of the exceptions to return, reflecting the priorities of the Hague Convention.¹¹⁹ It ‘attache[d] particular importance in this context to the need to conduct the proceedings in question swiftly and to the fact that these proceedings were not meant to determine the merits of the custody issue’.¹²⁰ Accordingly, the ECtHR considered that the Estonian court’s refusal to allow a further hearing, witness examination, and psychiatric evaluation of the father, before ordering return, was acceptable.¹²¹

The treatment of the allegations that the left-behind parent had abused the child was, with one exception, not obviously problematic; the ECtHR outlined the reasonably extensive consideration that the allegations had received in the Estonian courts.¹²² The exception was the ECtHR’s problematic invocation of the presumption of innocence in criminal proceedings, in support of its determination that the Estonian courts had been entitled to find that the allegations had not been adequately proved.¹²³ The Court remarked that a criminal standard of proof ‘need not’ be applied to the assessment (for Hague return purposes) of whether or not the father had abused the child.¹²⁴ However there is no apparent justification for the implication that the necessarily high threshold for imposing criminal responsibility could be relevant when assessing risk under the ECHR, or indeed the Hague Convention. That comment suggests willingness to countenance an unduly high bar for establishing a ‘grave risk’ to the child (based on alleged past abuse, which might resume if they were returned). The concern of greater significance relates to the convergence of the taking parent and rigorous assessment questions. The ECtHR was dismissive of the mother’s argument that, based on an alleged risk of violence against her from the father (and her lack of connections in Italy) she could not return, and that the child’s return without her would result in serious harm to the child, engaging the Hague Convention exceptions.¹²⁵ The ECtHR did not actually address the risk of violence to the mother, or consider the extent of evidence on the point before the Estonian courts, and (as noted above) accepted that those courts were entitled to refuse the further procedural steps sought by the mother.

The Court’s acceptance of the curtailed consideration of risks to the mother, and of the assumption that she could return, is particularly troubling in light of its consideration of the practicalities of the return. The ECtHR suggested that mother and child could rely on the father’s offer of ‘lodging and subsistence’, declining to consider whether that offer would be binding in municipal law.¹²⁶ If the allegations of violence inflicted on the mother were true, however, the suggestion that she should rely on the perpetrator for accommodation would appear to involve significant risks of further domestic violence, exacerbated by the likely power imbalance if he were under no obligation to abide by the commitment. Indeed under the Hague Convention itself, the guidance more recently issued by the Hague Conference on Private International Law recommends caution in reliance on such unenforceable undertakings.¹²⁷ Despite these seemingly significant challenges, in the ECtHR’s ruling, the child’s return was not precluded by the possibility of separation. The mother’s return was assumed without any meaningful assessment by the ECtHR of the Estonian courts’ decisions to reject the mother’s allegations that she had been abused.

Both MR and LR and Maumousseau exemplify an approach, which enthusiastically supports the Hague obligation to return, at the expense of effective Article 3 protection for taking parents and, potentially, their children. Commentators have attached particular significance to Maumousseau within the ECtHR’s Hague jurisprudence.¹²⁸ If the priorities established in Maumousseau are accepted (including the focus on expedition, and restrictive reading of the exemptions from the obligation to return, which those priorities imply), then there is little space for the rigorous precautionary approach to returns, which the non-refoulement obligation requires. MR and LR is also important, since it was decided against the background of increasing reference to Article 3 in (non-Hague) domestic violence cases, which might have been expected to prompt a change, especially since Article 3 was actually invoked by the applicant in that case.

The impact of the Hague Convention was so significant, in some ECtHR cases, as to result in what has been termed an ‘unrestricted prioritization of return’.¹²⁹ On that approach, in the two overlapping areas of potential conflict, non-refoulement assessment is likely to be truncated, and its protection denied to taking parents in their own right. However the trend in other, more recent, cases has been towards a more cautious approach to return orders, requiring more extensive consideration of the exceptions to return. That jurisprudence offers a pathway for establishing Article 3’s full applicability, as will now be seen.

5. THE ECTHR'S SHIFTING PRACTICE

The 2010 case of Neulinger and Shuruk v Switzerland (‘Neulinger’)¹³⁰ attracted significant attention as an apparently major departure from Maumousseau’s prioritisation of return, and acceptance of a tightly circumscribed role for the courts of the returning state in considering the best interests of the child.¹³¹ It appeared to suggest that the returning state courts were required to undertake an ‘in-depth’ review of their own as to whether return would be in the child’s best interests, contrary to the provisions and intention of the Hague Convention.¹³² The principles it articulated, while capable of encompassing the Article 3 issues discussed here, were much broader, seemingly covering all matters of relevance to the child’s best interests.¹³³

Accordingly, while Neulinger may have signalled an important willingness to depart from Maumousseau, the approach it adopted has had variable purchase in subsequent cases.¹³⁴ It will thus not be relied on here as the basis for incorporating full non-refoulement protection into the ECtHR’s Hague jurisprudence. The ECtHR’s limited mandate to interpret the Hague Convention, by virtue of Article 31(3)(c) of the VCLT, is not a mandate to re-write it or to generally downgrade its objectives, especially since the Hague Convention has been ratified by many states outside the ECHR, who have not given any mandate to the ECtHR at all. Giving due effect to Article 3 can, and should, be achieved with less violence to the Hague Convention. Subsequent jurisprudence indicates a more appropriate route.

A. X v Latvia: A Constrained Recalibration

It is the case of X v Latvia (X) in 2013,¹³⁵ seen as something of a retreat from Neulinger,¹³⁶ which marks the significant turning point for present purposes. The ruling was not factually directly relevant to the question of abuse.¹³⁷ Instead it was primarily the anticipated psychological impact on the child of potential separation from her mother, if returned, which, in the ECtHR’s view, had received insufficient attention in the reasoning of the Latvian court, which ordered the return.¹³⁸ However, of relevance to the Article 3 discussion here, the reasoning was specific to Article 13(1)(b); the ECtHR held that the domestic courts were required to provide adequate review of any alleged ‘grave risk’.¹³⁹ In doing so they seemingly offered a means of addressing the rigorous assessment question. As McEleavy observes, criticising the decision, the ECtHR’s approach suggests that once the taking parent has made a prima facie case, the authorities are then required to establish whether or not the risk exists. That implies the kind of extended review that the Hague Convention seeks to avoid the returning state undertaking.¹⁴⁰ Such review is however consistent with the rigorous requirements for assessing the evidence of a risk of Article 3 ill-treatment described above.¹⁴¹ On the facts the ruling perhaps went rather beyond that approach, insofar as the focus was primarily on risks unlikely to meet the Article 3 threshold;¹⁴² however in doing so it opened space for Article 3 risks to receive the same scrutiny.

The fact that the risk was of separation-induced trauma is also significant, as regards the taking parent question. The ECtHR emphasised the need for the Latvian courts to consider and determine whether or not the mother could accompany the child and maintain contact in Australia, if a return order were granted.¹⁴³ In that respect, Beaumont et al. criticise the X ruling, arguing that the ECtHR should have followed MR and LR, in assuming that the return of the mother was permitted unless she could ‘prove’ (with ‘concrete evidence’) ‘sufficient reason for her not to return’.¹⁴⁴ However in MR and LR the mother had seemingly been deprived of the opportunity to prove the sufficient reasons (including abuse directed against her) which she alleged.¹⁴⁵  X would better allow for the Article 3 rights of the taking parent, by ensuring that if they make a prima facie case then the risks they face are adequately assessed, instead of forcing them to choose between separation from their child and a return to face a possible real risk of ill-treatment (in the absence of a more extended assessment).

Again, on the facts, X possibly courted criticism by going further than was needed—given the seemingly complete lack of evidence that the mother had been ill-treated in X, ¹⁴⁶ the case for further investigation was rather less compelling from an Article 3 perspective than it had been in MR and LR.¹⁴⁷ Nonetheless, X provided a basis for the taking parent question also to be addressed, in cases where there is some evidence of risk.¹⁴⁸

B. The Recent Jurisprudence

Although X created space for the ECtHR to determine cases in a manner that takes fuller account of the requirements of Article 3, as was noted above those requirements had little direct relevance to the case. Accordingly, the practical possibilities for such an approach have only started to emerge subsequently, and they have emerged without explicit acknowledgment of the change. In particular, two Hague Convention decisions in 2019 went further than merely making theoretical allowance for Article 3, and reached conclusions that were of practical importance for the Article 3 rights at issue. The first is the admissibility decision in BS v Poland, a case brought by a left-behind parent.¹⁴⁹ In that case, the Polish authorities took over nine months to decide the applicant’s Hague case seeking return of his son to Germany (which was ultimately declined).¹⁵⁰ The ECtHR considered that his Article 8 rights were not violated by the delay, since allegations that he had sexually abused the child had been made and were under criminal investigation. Ultimately the allegations were not substantiated by the criminal investigation and were not the basis for the decision against return.¹⁵¹ However the ECtHR suggested that although such allegations are (it asserted) ‘frequently’ made falsely in Hague proceedings, they were ‘not to be treated lightly, and they ought to be investigated by domestic authorities before those authorities can decide whether the Hague Convention exception applies in the circumstances of the case.’¹⁵² In that light, and faced with various complicating factors, the Polish authorities had progressed the matter as expeditiously as possible.¹⁵³ The case thus demonstrates how the ECtHR’s requirements of review of the Article 13(1)(b) exceptions, established by X, can be relied on to address the rigorous assessment question, at the expense of the Hague regime’s expedition requirement.

The other 2019 case, OCI and others v Romania, has been mentioned earlier.¹⁵⁴ The (taking parent) applicant argued that the return of her children to their father in Italy, ordered by the Romanian courts, violated their and her Article 8 rights, relying primarily on his record of using corporal punishment against the children.¹⁵⁵ The Article 8 discussion extensively cited the Article 3 positive obligations case, DMD,¹⁵⁶ which had established the ECtHR’s view that corporal punishment is inherently degrading.¹⁵⁷ On the basis of inadequate consideration of ‘the potential risk of ill-treatment’, the return order breached the ECHR. Having identified the breach under Article 8, by reference to that risk, the ECtHR held that there was ‘no separate issue’ under Article 3.¹⁵⁸ Thus although the ECtHR declined to rule on the Article 3 claim, which had also been pleaded, it did so on the explicit basis that the Article 8 analysis in the decision had addressed the risk of ill-treatment.¹⁵⁹ In contrast to decisions where the ECtHR has simply opted to characterise the case in terms of Article 8 even where Article 3 allegations have been made,¹⁶⁰ the wording implicitly acknowledged that the Article 3 aspect of the case needed to be addressed.

Some reinforcement of the shift can be identified in subsequent cases. In YS and OS v Russia, like OCI, the ECtHR addressed the applicants’ argument that Article 3 (as well as Article 2) was relevant by remarking that it had ‘already examined the principal arguments raised under Articles 2 and 3’ when it considered Article 8.¹⁶¹ Its Article 8 findings, that the Russian courts had not sufficiently considered the armed conflict-related risks the child faced in ordering her return to Eastern Ukraine, obviated the need to separately examine Articles 2 and 3 in relation to the return of the child.¹⁶² Again the ECtHR implicitly acknowledged the applicability of Article 3, as well as challenging the adequacy of the domestic courts’ assessment of risks of violence clearly relevant to Article 3 (albeit very different from the domestic violence risks more commonly relied on in Hague proceedings).

The cases above relate primarily to the rigorous assessment issue. However in two other cases, particularly relevant to the taking parent question, brought by left-behind parents, the ECtHR endorsed domestic decisions to refuse return based on risks of violence or abuse. Moga v Poland is most interesting, since the ECtHR raised no objection in principle to the domestic court’s decision to refuse return, based primarily on the risk that the child would witness domestic violence against the taking parent.¹⁶³ Although it considered that on the facts that basis was unsustainable, since the couple had separated (and so relied on other concerns as establishing an adequate basis for refusal), the implicit assumption that such grounds could have justified refusal points to a vehicle for some recognition of risks to the taking parent, within the assessment of risk for the child.

Finally, PD v Russia deserves mention because, notwithstanding the Hague Convention’s specific focus on risks to the child, it was past abuse by a friend of the (left-behind parent) applicant against the child’s half-brother that formed the basis for identifying a risk, to the child, if returned.¹⁶⁴ Hence the alleged abuse demonstrating a ‘grave risk’ had not been committed by the applicant himself, nor against the child whose return was sought. In rejecting the applicant’s claim that this basis ‘did not involve a genuine and objective evaluation of the alleged risk to the applicant’s daughter’ (whose return was sought), the ECtHR allowed sufficient flexibility to take into account broader evidence of risk. Comparison to the ECtHR’s lack of objection to the French authorities’ rejection of domestic violence allegations in Maumousseau, on the basis that the child had not been the alleged victim, illustrates the shift. The flexibility shown in PD v Russia could facilitate increased responsiveness to Article 3 risks, in such circumstances.

Accordingly there is some evidence for a change in the ECtHR’s approach, which can allow greater scope for Article 3 protection in Hague cases, and a greater willingness to recognise (although only implicitly) that such protection is indeed relevant. It was observed in Section 4 that such protection has the potential to impose a cost on the Hague Convention regime and its priorities, in some cases. The following section considers whether imposing that cost can be justified.

6. ACCOMMODATING THE NON-REFOULEMENT `COACH AND HORSES'

Even if the apparent change in practice in Neulinger is ruled out, as too sweeping a departure from the Hague Convention regime’s intentions, the more limited changes outlined above, from the subsequent jurisprudence, also require examination with that concern is mind. The argument for the explicit recognition of Article 3 protection in Hague cases raises that issue as well. While earlier sections have emphasised that the ‘absolute’ nature of Article 3 justifies the ECtHR in departing from Hague requirements where necessary, this section revisits the question of how far departures are in fact necessary.

There are legitimate arguments for interpreting the Hague Convention to make allowance for Article 3 protection, notwithstanding the costs it may sometimes impose on fulfilment of the Hague Convention’s object and purpose. Further, with appropriate care and recognition of the task at hand, those costs can be significantly limited, achieving an optimal reconciliation between Article 3 and the Hague Convention, in which neither is unacceptably compromised. Those considerations will be explored in turn in this section, before the Conclusion summarises the outcome of the analysis conducted in this piece as a whole.

A. The Interpretation of the Hague Convention

By maintaining its silence on Article 3, the ECtHR avoids being seen either to trespass on matters beyond its jurisdiction, or to admit to the existence of an area in which ECHR states face a conflict in their international obligations. Doing so, however, neglects the possibilities for the ECtHR to contribute to the evolution in the Hague Convention’s interpretation in light of evolving IHRL. The ECHR’s Article 3 evolution has various parallels beyond its specific regional context, in wider IHRL. Hence on the question of corporal punishment, the ECtHR’s more recent restrictive approach accords with (and indeed appears to be inspired by) the position of the UN Committee on the Rights of the Child, in its interpretation of the 1989 Convention on the Rights of the Child.¹⁶⁵ The expanded coverage of Article 3 in this respect is not therefore special to the ECHR’s regional context.

In general terms, a range of states, international organisations, and bodies responsible for IHRL treaties, regionally and globally, have moved to emphasise the incompatibility of domestic violence with fundamental human rights in international law, including established prohibitions on torture and ill-treatment.¹⁶⁶ Specifically regarding the rigorous assessment question, the evolution in the interpretation of the Refugee Convention, recognising that its own non-refoulement obligation can also be triggered by the threat of domestic violence,¹⁶⁷ necessarily implies that states should not remove a child before the risk of such violence has been rigorously assessed and found not to meet the applicable threshold.¹⁶⁸ Under Article 31(3)(c) of the VCLT, these developments should be taken into account in interpreting the Hague Convention. Although accommodating full non-refoulement protection compromises the Hague Convention’s objectives, Article 20 of the Hague Convention at least provides a plausible vehicle in the treaty text for such an accommodation. Indeed that is not a new suggestion; Weiner’s argument that Article 20’s role should be ‘strengthened’ in judicial practice, taking into account its text and its negotiating history, specifically endorsed its applicability to pending asylum claims (analogous to the ‘rigorous assessment’ issue) and to domestic violence threats faced by the taking parent.¹⁶⁹

As will be recalled, Article 20 provides that return is not required if it would breach ‘the fundamental principles of the requested State relating to the protection of human rights and fundamental freedoms.’ Regarding the taking parent question, if the non-refoulement obligation were interpreted to restrict returns by reference to risks to the taking parent (as this piece has argued it should),¹⁷⁰ it appears reasonably clear that Article 20 should then allow for such an approach.¹⁷¹ There is little reason to question that the substantive prohibitions arising from Article 3 constitute prohibitions expressing ‘fundamental principles’ referred to in Article 20.¹⁷²

The position is arguably more complicated for the rigorous assessment question and the procedural requirements it imposes. While Weiner considers that Article 20 would afford protection in order to allow full consideration of protection claims (in the context of Refugee Convention asylum proceedings),¹⁷³ Trimmings and Beaumont have argued that:

Article 20 of the Hague Abduction Convention only applies to human rights that constitute a reason for refusing to return a child. It does not apply to create procedural rights as to how the courts should go about the process of determining whether or not a return of the child would be a breach of human rights or would be contrary to one of the Article 13 exceptions in the 1980 Convention.¹⁷⁴

That view is of course consistent with the object and purpose of the Hague Convention, aiming to avoid a lengthy process in the state to which the child has been abducted.¹⁷⁵ However it is not apparent from the terms of Article 20 itself; if human rights principles prohibit expulsion unless a proper procedure is followed (as they do), then the only issue is whether the procedural principles are somehow not ‘fundamental’.

On that point, as Trapp and Mills observe, ‘“procedural” rules may go to the heart of substantive justice, in facilitating or denying a remedy to a claimant’.¹⁷⁶ In this context, where the issue is protection from future risk (as opposed to remedies for past breaches), the procedural matter of risk assessment is integral to the substantive protection in question.¹⁷⁷ That was the reason for the ECtHR’s original departure, in Soering, from its general procedure of refusing to consider ‘potential violations’ before the fact: the departure was ‘necessary, in view of the serious and irreparable nature of the alleged suffering risked, in order to ensure the effectiveness of the safeguard provided by [Article 3]’.¹⁷⁸ Accordingly, a strong argument can be made for interpreting Article 20 to allow for the operation of full, procedural, non-refoulement protection.

B. Limiting the Cost

Finally, accommodating non-refoulement in this respect does not require the evisceration of the Hague Convention. Article 3 will not always be at stake, and, where it is, measures should be taken to ensure that Article 3 protection is provided without compromising the Hague Convention’s objectives any more than is truly necessary. The ECtHR jurisprudence highlights a range of scenarios in which Hague Convention requirements play a dispositive role, unaffected by any Article 3 allegations. A survey of the ECtHR’s Hague Convention jurisprudence in 2022 shows plentiful examples of states being found to have breached the obligation to return, in circumstances where Article 3 had no role to play, or would not have altered the outcome; in particular where there are unjustified delays in proceedings (in contrast to the delays in BS v Poland).¹⁷⁹ Even where the ECtHR upheld the domestic court’s refusal of return in Moga v Poland, it held that the unnecessary delay in reaching that decision violated the left-behind parent applicant’s Article 8 rights.¹⁸⁰ The Hague Convention, and the ECtHR’s support for its enforcement, would provide a basis for identifying and condemning such failures even if Article 3’s applicability is fully recognised—it can only justify the delays, which are genuinely necessary.

Proactive changes to Hague proceedings at the domestic level could further mitigate the conflict between expedition and rigour. Momoh argues that, in fact, a rigorous but expedited assessment of relevant factual allegations, capable of sufficiently addressing the Hague Convention’s requirements and the need for proper investigation of the risk, is in principle achievable.¹⁸¹ While that is not necessarily to say that such an assessment would always offer the degree of rigour and protection expected under the non-refoulement obligation, it suggests that it should be possible to significantly reduce the number of cases where the two regimes are at odds. In the UK context the Supreme Court has noted, and endorsed, measures of institutional co-operation between UK judicial and administrative authorities, to ensure expedition and accuracy where assessments of non-refoulement risk and Hague return requests overlap.¹⁸²

Further reduction, and perhaps complete reconciliation, could be achieved by focusing on the measures for protection and family dispute resolution that the state to which the child would be returned makes available. For example, if the risk of violence against the taking parent can be removed by prohibiting the left-behind parent from contacting them, and the risk to the child removed by requiring that their contact with the left-behind parent be supervised, then the non-refoulement objection can be removed.¹⁸³ It is therefore necessary that the returning state authorities have sufficient awareness of the situation in the receiving state to consider those measures appropriately. To that end, active ongoing contact between states, which the Hague Convention requires,¹⁸⁴ would support the achievement of the Hague Convention’s objectives. If there is adequate evidence in the sending state proceedings that the speed, accessibility, effectiveness and enforceability of the destination state’s measures will be sufficient then the return can proceed without breaching non-refoulement. If there is not, then returning states should be expected to interpret the Hague Convention not to require return, based either on Article 13(1)(b) or Article 20.¹⁸⁵

7. CONCLUSION

Once the opportunities for proactive measures to harmonise the operation of the two regimes are exploited, the number of cases in which a conflict between the Hague regime’s objectives and non-refoulement protection may arise should be considerably reduced, and perhaps eliminated. Recognition by the ECtHR of the significance of Article 3 and the non-refoulement obligation in Hague cases would provide the impetus for ECHR states to adopt such measures. It remains possible to imagine cases in which non-refoulement protection precludes a return that would ordinarily be required by the Hague Convention. However that hypothetical possibility should not prevent the ECtHR from upholding the absolute nature of Article 3, even if it requires states to invoke Article 20 of the Hague Convention in circumstances where they would otherwise not have, and thereby somewhat undermine the priority of expeditious return.

The ECtHR’s refusal to openly address Article 3 in Hague cases is problematic. It leaves a lack of clarity for applicants, for states and for ECtHR judges themselves as to the correct approach, and in particular the potential gaps in protection relating to the rigorous assessment and taking parent questions. An explicit approach would also help to fortify the absolute status of Article 3 protection against removal, at a time when it is under pressure in the context of refugee rights.¹⁸⁶ Doing so would, in addition, support the approach taken by domestic courts, such as those in the UK, which have held that the Hague Convention cannot cut across the assessment requirements of the non-refoulement obligation in the Refugee Convention.¹⁸⁷

The early signs of a shift in the ECtHR’s approach to Hague cases, allowing a more active role for the protection which Article 3 requires against domestic violence, are welcome. The challenge that remains is ending the silence, in order to articulate clear principles for ECHR Hague cases, and reinforce the priority of human rights protection elsewhere.

ACKNOWLEDGMENTS

I am very grateful to anonymous reviewers for their extremely valuable comments, to Dr Nikolaos Pavlopoulos for his kindness in reading an early draft, and to Professor Kimberley Trapp for her insightful support on the research from which this project ultimately sprung; any errors are my own.

Footnotes

Author notes