The Interplay between the 1980 Hague Convention on the Civil Aspects of International Child Abduction and Domestic Violence

Abstract

When a mother commits an international child abduction, even if she is fleeing domestic violence perpetrated by the left-behind father, she is bound to face complicated return proceedings under the 1980 Hague Child Abduction Convention. Such mothers are particularly vulnerable; apart from the costly, cross-border proceedings they face, if the court issues a return order, they risk returning to the abusive setting they fled from. This article explores avenues for safeguarding the protection of abducting mothers in return proceedings. The authors provide a range of potential avenues for improving the standing of the abducting mother fleeing domestic violence, including judicial and legislative interventions. The article delves deeper by considering the interplay between international child abduction law and international refugee law in cases involving domestic violence allegations. Particular emphasis is given to Article 20 and the growing instances of mothers defending return orders on asylum grounds pursuant to Article 20 and the flowing human rights implications. The authors point out a niche area for further research: the interplay between domestic violence and asylum claims.

1. Introduction

This article addresses the problem of domestic violence in the context of the 1980 Hague Convention of 25 October 1980 on the Civil Aspects of International Child Abduction (‘the 1980 (Hague) Convention’). It is concerned specifically with the position of mothers who have abducted their child(-ren) across international borders and are involved in return proceedings under the 1980 Hague Child Abduction Convention in circumstances where the child abduction was motivated by domestic violence by the left-behind father. In doing so, the article also touches on the interplay between international child abduction law and international refugee law in cases involving allegations of domestic violence (e.g., Hegar and Greif 1994; Norris 2010; Hayman 2018; Al-Shargabi 2022). The role of the ‘best interests of the child’ principle in child abductions committed against the background of domestic violence is explored before a range of possible judicial and legislative interventions to secure the protection of abducting mothers in return proceedings under the 1980 Hague Convention is examined.

2. The Problem of Domestic Violence in the Context of the 1980 Convention

The latest statistical analysis carried out by the Hague Conference on the application of the 1980 Convention showed that as of 2015, 73% of the taking persons were mothers, noting an increase of 4% compared to the earliest set of data of 1999 (See Permanent Bureau of the Hague Conference on Private International Law (2017b, pp. 3 and 7)). Further, the number of abducting fathers, which had been reported to be 30% in 1999, had reduced to only 24% in 2015.1 With hindsight, the 1980 Convention effectively established a global scheme for the return of abducted children under which fathers regularly request return and find themselves successful in approximately half of cases (Permanent Bureau of the Hague Conference on Private International Law 2017b, p. 11). Today, abducting mothers regularly report the existence of domestic violence directed against them, the child or both by the left-behind parent in return proceedings under the 1980 Convention (e.g., Weiner 2000; Brown Williams 2011; Hale 2017; Trimmings and Momoh 2021; Masterton et al. 2022). Although there are no comprehensive statistics, it is suspected that domestic violence is a present issue in as many as 70% of the total parental child abduction cases (Trimmings and Momoh 2021, p. 5; Pérez-Vera 1982, p. 34). Abducting mothers who flee domestic violence and face return proceedings find themselves in a particularly vulnerable position; they are faced with complex and costly cross-border proceedings, and if they do not wish to return, as is the case most times, they must prove to the court that the domestic violence they experienced at the hands of the left-behind father presented a ‘grave risk of harm’ for their child or otherwise placed him or her in an ‘intolerable situation.’2 If the court mandates the return of the child, the now returning mother is likely to return to the unsafe situation she fled from, become financially dependent on the left-behind father or in more extreme cases, face homelessness (Masterton et al. 2022, pp. 376–81).

In some cases, the abducting mother escaping domestic violence may apply for a refugee status under the 1951 Convention Relating to the Status of Refugees (‘the 1951 Refugee Convention’) (Zimmermann 2011)3 in conjunction with its 1967 Optional Protocol Relating to the Status of Refugees (‘the 1967 Optional Protocol’). At the European Union level, additional legal instruments may come into play.4 Additionally, soft law instruments such as the principles and guidelines on the human rights protection of migrants in vulnerable situations, which focus on the human rights situation of migrants who may not qualify as refugees but who are nevertheless in vulnerable situations and therefore in need of protection by international human rights law, can also be invoked (Office of the High Commissioner for Human Rights 2018b). Return proceedings under the 1980 Hague Convention may be initiated by the left-behind father either during the asylum proceedings, leading to complex interaction between international family law and refugee law, or after the conclusion of the asylum proceedings, potentially raising the question of whether the principle of non-refoulement5 can be undermined by a return order granted in favour of the left-behind father under the 1980 Hague Convention.

3. The Role of the Best Interests of the Child

One of the obvious challenges in the operation of the Hague Convention is ensuring that at the heart of the decision-making process the interests of children prevail, especially in cases involving allegations of domestic violence. Where intra-EU cases are concerned, it was observed that the Brussels IIa Regulation6 carried particular emphasis on the best interests of children (Kruger and Samyn 2016, p. 155).7 That being said, Kruger and Samyn went further and suggested that the Regulation would be clearer, stronger and more credible if it referred explicitly to the United Nations Convention on the Rights of the Child (UNCRC).8 The Brussels IIa Recast Regulation9 now in force appears to go further, emphasising in its preamble that matters of parental responsibility shall be ‘shaped in the light of the best interests of the child’. This is significant, because over a decade after the adoption of the 1980 Hague Convention, Dyer made a comparison between the number of ratifications of the United Nations Convention on the Rights of the Child and the Hague Convention (Dyer 1993, p. 273).10 At that stage, there were 30 Contracting States to the Hague Convention which was ‘less than one fourth’ of the UNCRC ratifications.11 Dyer suggested that ‘an obvious point’ was that the obligations under the Hague Convention ‘are more precise and constraining than the obligations described in an “umbrella” Convention like the UNCRC’ (Dyer 1993, p. 273). The argument follows that the provisions of the Hague Convention place a heavier burden on Contracting States, being both stricter and more specific in its objectives. Therefore, the ‘execution of this obligation requires discipline on the part of the courts’ and an ‘acceptance of new points of view by both judges and populations’ (Dyer 1993, p. 274). In contrast, the UNCRC upholds principles on the rights of children that one would expect to be universally acknowledged, and therefore, there is an ease for States to commit themselves to meeting those obligations. We are now at a stage where these treaties cannot be seen in isolation, a preference for one over the other is uncompelling.

In cases under the 1980 Hague Convention, it is also important to distinguish between the concepts of the best interests of the child generally and the interests of a child involved in return proceedings.12 In terms of the terminology, the preamble to the Hague Convention refers to the ‘interests of children, and the Convention’s core philosophy13 is that it is not in the interests of children to be wrongfully removed or retained across international borders; it is generally thus not in the ‘best interests’ of children to be abducted.14 The Explanatory Report reiterates that ‘the right not to be removed or retained’ demonstrates one of the objective examples of the interests of a child and that the two objectives under Article 1 of the Hague Convention embody the best interest of the child (Pérez-Vera 1982, paras. 24–25; Chamberland 2012). The Explanatory Report further states that the ‘true interests’ of a child are ‘inspired’ by a desire to protect children against the harmful effects of international child abduction (Pérez-Vera 1982, para. 24). However, in cases involving domestic and family violence, a departure from this principle may not only be justified but necessary.15 Article 13(1)(b) provides an exception to the return of a child where there is a grave risk of physical harm, psychological harm or an otherwise intolerable situation. Accordingly, in return proceedings involving allegations of domestic violence under Article 13(1)(b), it has often been argued that to return a child in such circumstances would be contrary to their best interests (Chamberland 2012, pp. 27–30).16 In recent times, the publication of the HCCH Guide to Good Practice on Article 13(1) (b) has sought to address head on the issues around the proper and consistent application of the grave risk of harm exception, though it would be premature to assess its utility and impact. Having said that, it has not abated concerns that the grave risk of harm exception is not doing enough to protect victims fleeing from domestic violence and their children. Some may argue that the concerns raised by Vesneski, Lindhorst and Edleson in their research remains true today (Pretelli 2021). Vesneski, Lindhorst and Edleson have argued that court decisions under Article 13(1)(b) were frequently ‘against the interests of even battered women and their children’17 and that ‘abused women arguing grave risk face a more difficult path’ (Vesneski et al. 2011, p. 17).18 A stronger view advanced by Weiner is that the Hague Convention has become a ‘substantial barrier to some women’s ability to escape domestic violence’ (Weiner 2003, p. 799). Weiner takes the view that it is rarely in a child’s best interest to return in the face of serious allegations of domestic violence, arguing that the Hague Convention ‘offers too little hope for the domestic violence victim who flees with her children’ (Weiner 2003, p. 703).19 Norris also asserts that in cases involving the grave risk of harm exception the courts should ‘apply the best interests of the child as its guiding criterion, rather than the need for prompt return’ and to ensure that a decision to return a child is not harmful (Norris 2010, pp. 185–86 and 194–95). In recent times, Pretelli weighs in on the concern, arguing that current legal framework ‘places women in an impossible situation, in a double bind’ (Pretelli 2021, p. 376). One of Pretelli’s concluding remarks reflects on the pursuit of achieving universal principles such as the best interests of children including in child abduction cases (Pretelli 2021, p. 393). On the other hand, an opposite view held by Browne is that the best interest enquiry should be avoided as it ‘threatens to invite the type of gender stereotype prevalent in custody disputes’ (Browne 2011, p. 1222). Nevertheless, any consideration afforded to the best interests of the child principle in Hague Convention proceedings is by no means intended to invoke a detailed examination of welfare issues or a merits exercise of the custody dispute (Beaumont and Walker 2013; Silberman 2011; Pérez Manrique 2012).20

4. Safeguarding the Protection of Abducting Mothers in Return Proceedings

When determining whether an exception to return under the 1980 Convention applies, ‘it is the situation of the child which is the prime focus of the inquiry’;21 the Convention has no explicit regard to the safety of the abducting mother upon the return. Although it is not mandatory for the abducting mother to return together with the child, the mother (in particular if she is the primary carer),) will typically accompany the child back to the State of habitual residence, even if it means that she has to compromise her own safety. The lack of consideration for the abducting mother’s safety in return proceedings involving allegations of domestic abuse is concerning.22 It highlights the pitfalls of applying the 1980 Convention in isolation from international human rights law—an approach which is contrary to the wider trend towards a more pronounced confluence of private international law and public international law (e.g., Mills 2009). In this context, it has been rightly remarked that as both the public and the private international systems coordinate human behaviour, the values that inform both systems should be the same (Maier 1982). In addressing the problem, this section analyses relevant case-law, whilst making suggestions for judicial interpretations and legislative interventions that have the potential to assist in securing the protection of abducting mothers in return proceedings in child abduction cases committed against the background of domestic violence (See also Trimmings et al. 2023).

4.1. Case-Law Analysis and Suggestions for Appropriate Judicial Interpretations

4.1.1. The ‘Grave Risk of Harm’ Exception (Article 13(1)(b)) (See (Permanent Bureau of the Hague Conference on Private International Law 2020, paras 57–59))

Article 13(1)(b) contains the ‘grave risk of harm’ defence, which, at its core, will exempt the abducting parent from returning the child to the State of his/her habitual residence if there is a grave risk that on return the child would be exposed to a ‘physical or psychological harm’ or be otherwise placed in ‘an intolerable situation.’ It is typical for abducting mothers who have fled domestic violence to seek to rely on Article 13(1)(b) to resist a return application by the left-behind father. In 2015, the ‘gave risk of harm’ defense was the ‘most frequently relied upon ground for refusal’ and was amongst the reasons for judicial refusal in 25% of applications (Permanent Bureau of the Hague Conference on Private International Law 2017b, p. 16). Despite being frequently invoked, Article 13(1)(b) contains integral key terms such as ‘grave risk’ and ‘intolerable situation’ which are undefined by the Convention, thus relying on domestic courts for interpretation (Brown Williams 2011, p. 62).23 Through years of application and with knowledge of the drafters’ intention that Article 13(1)(b) should have restricted application, courts have discerned the a number of principles pertaining to the interpretation and application of this defence (Pérez-Vera 1982, paras. 7 and 34; Trimmings and Momoh 2021).

The Correlation between Domestic Violence Directed towards the Mother and a Grave Risk of Harm to the Child (See (Permanent Bureau of the Hague Conference on Private International Law 2020, paras 57–59))

Most courts have adopted a ‘literal interpretation’ and only in the past decade did the UK Supreme Court draw a connection between domestic violence directed towards the mother and a grave risk of harm to the child (Hale 2017, p. 7; Quillen 2014, p. 632; Brown Williams 2011, p. 62).24 Internationally, not all courts acknowledge that the former is directly related to the latter (Zashin 2021, p. 585).25 Accordingly, the existence of domestic violence alone is insufficient to satisfy the grave risk of harm defence (Trimmings and Momoh 2021, p. 6). Instead, ‘the key question is whether the effect of domestic violence on the child upon his/her return’ will have such an impact as to place him/her in grave risk of harm (Trimmings and Momoh 2021, p. 6). The Article’s limited reach is particularly manifested in the word ‘intolerable’ which has been interpreted to denote ‘a situation which this particular child in these particular circumstances should not be expected to tolerate’,26 and which is beyond the ‘tumble, discomfort, and distress’ that is acceptable for a child to tolerate.27 An ‘intolerable situation’ could be one where the child is harmed by exposure to domestic violence in the form of physical or psychological abuse towards a parent (Permanent Bureau of the Hague Conference on Private International Law 2020, para. 58).28 The Hague Conference Permanent Bureau of the Hague Conference on Private International Law 2012 on Article 13(1)(b) further includes the potential risk of harm upon the return of the child and circumstances where the grave risk of harm manifests itself in the form of ‘significantly impairing the ability of the taking parent to care for the child’ (Trimmings and Momoh 2021, p. 5; See also Permanent Bureau of the Hague Conference on Private International Law (2020, para. 57)).

Additionally, the UK Supreme Court has held that it is irrelevant whether the risk is the result of objective reality or of the abducting mother’s subjective perception of reality.29 Accordingly, anxieties of an abducting mother about a return with the child which are not based on objective risk to her but are nevertheless of such intensity as to be likely, if returned, to affect her mental health so as to destabilise her parenting of the child to a point where the child’s situation would become intolerable, can found the grave risk of harm defence under Article 13(1)(b).30 It is not important whether the mother’s anxieties are reasonable or unreasonable.31 This means that if the court concludes that there is a grave risk of harm to the child, the source of the risk is irrelevant. Therefore, the grave risk of harm defence may successfully be established, for example, “where a mother’s subjective perception of events leads to a mental illness which could have intolerable consequences for the child.”32 The court shall, however, examine an assertion of intense anxieties not based upon objective risk very critically, and shall consider whether it can be dispelled through protective measures.33

Assessment of Allegations of Domestic Violence in Return Proceedings (See (Permanent Bureau of the Hague Conference on Private International Law 2020, paras 43–48 and 50–54))

When assessing the granting of a (non)-return order under Article 13(1)(b), courts in the UK34 and internationally35 have mainly followed the so-called ‘protective measures approach’ or variants of it. The said approach is a two-step assessment and at the first instance, involves the court considering the following question: ‘If [the domestic violence allegations] are true, would there be a grave risk that the child would be exposed to physical or psychological harm or otherwise placed in an intolerable situation?’36 If the court answers in the affirmative, the court ‘will consider whether protective measures to mitigate the harm are available in the requesting State’ (Trimmings and Momoh 2021, p. 6). The court will grant a non-return order ‘only if the protective measures cannot ameliorate the risk’; in all other circumstances, the court will entrust the resolution of the merits of the issues to the courts of the requesting State, assuming that they are best-suited to deal with the substantive questions (Trimmings and Momoh 2021, p. 6).

The ‘protective measures approach’ suffers from pitfalls that typically jeopardise domestic violence victims. By design, the approach is paradoxical in that it ‘relies on the availability of adequate and effective measures as a substitute for determining facts’ (Trimmings and Momoh 2021, p. 7). An assessment of ‘grave risk’ and available protective measures cannot reasonably come before exploring whether domestic violence exists and if it does, what risks it encompasses (Trimmings and Momoh 2021, p. 9).

An alternative approach, which is considered more appropriate, has been termed as the ‘assessment of allegations approach’(Trimmings and Momoh 2021, p. 7).37 Under this approach, the court will first seek to determine, to the extent possible within the confines of the summary nature of the return proceedings, the merits of the disputed allegations of domestic violence. Once the assessment of allegations has been carried out, the court will determine whether a grave risk of harm exists. Only afterwards, as part of the exercise of discretion,38 the court will assess availability of protective measures. This approach is based on the premise that it is necessary to assess the disputed allegations in order to evaluate the risk. Admittedly, this approach may raise concerns over the length of the proceedings; however, speed should not take priority over the proper assessment of risk and consideration of the safety of the child and the abducting mother. Indeed, the emphasis on speed may encourage judges to minimise or ignore allegations of domestic violence rather than determining them, leaving thus an unassessed risk of harm. Importantly, this approach seems to be supported by the jurisprudence of the European Court of Human Rights, specifically the case of X v Latvia39 where the Grand Chamber introduced the concept of ‘effective examination’(Beaumont et al. 2015; Momoh 2019, pp. 650–56). As Judge Albuquerque explained in his concurring opinion, ‘effective examination’ means a ‘thorough, limited and expeditious’ examination. Accordingly, it is recommended here that a ‘thorough, limited and expeditious’ examination of disputed allegations of domestic violence be carried out by the judge in return proceedings.40

Protective Measures (See Permanent Bureau of the Hague Conference on Private International Law 2020, paras 43–48)

There is a further note to be made regarding the scope of ‘protective measures’ available, as their robustness is what the domestic violence victim and the child will rely on upon their return. In the UK, the Practice Guidance on case management of child abduction cases ‘distinguishes between protective measures that “are available” and protective measures that “could be put in place,” making clear the potential extensive scope of the exercise’ (Munby 2018; Trimmings and Momoh 2021, p. 11). In England and Wales, when assessing the availability and effectiveness of ‘protective measures’ the courts have included ‘general features’ of the requesting State’s legal system such as ‘access to courts and other legal services, state assistance and support, including financial assistance, housing assistance, health services, women’s shelters and other means of support to victims of domestic violence’ (Trimmings and Momoh 2021, p. 11).41 The expansive understanding of ‘protective measures’ may mean that the court’s assessment is not focused on the measures that can facilitate the protection to a returning domestic violence victim such as ‘decisions of courts and/or other competent authorities (as appropriate)’.42

In the extent of ‘protective measures’ available, common law courts will further include the so-called ‘undertakings’, which may be defined as ‘promises’ sometimes granted by the left-behind parent that aspire to address the reasons behind the taking parent resisting return (Trimmings and Momoh 2021, p. 12; Brown Williams 2011, p. 66; Zashin 2021, p. 577). ‘The concept of undertakings is a judicial creation and is not included or defined in the 1980 Convention’ (Zashin 2021, p. 577). In reality, undertakings are not effective because they are regularly breached by their grantors43 and suffer from limited enforceability since they are not recognised in civil law jurisdictions (Trimmings and Momoh 2021, p. 13). Therefore, in deciding what weight should be given to protective measures, the judge must take into account the extent to which they will be enforceable in the State of habitual residence. In intra-EU child abduction cases recognition and enforcement of protective measures can be facilitated by either the Brussels IIa Recast Regulation and/or the Protection Measures Regulation.44 Outside of the EU, in cases where the State of habitual residence and the State of refuge are both contracting parties to the 1996 Hague Convention,45 this Convention should be utilised to facilitate cross-border recognition and enforcement of protective measures in return proceedings. However, where the State of habitual residence is not a party to the 1996 Convention, extreme caution should be exercised by the judge when protective measures are sought.

Even where a legal mechanism for cross-border circulation of protective measures exists, judges should be guarded when considering making a return order conditioned on such measures. In particular, they should be wary of the fact that protection orders are often breached, and that satisfactory follow-up measures by relevant authorities in the State of habitual residence may be lacking. In any case, employment of protective measures with a view to making a return order should never be considered in cases where it has been established that there is a future risk of severe violence.

Evidence-Related Matters (See (Permanent Bureau of the Hague Conference on Private International Law 2020, paras 50–54))

Abducting mothers pleading ‘grave risk of harm’ also grapple with more practical issues. The burden of proof that ‘grave risk of harm’ exists, rests with the party resisting return. However, there is no internationally agreed standard required for the purposes of Article 13(1)(b), and many times evidence of harm caused is unavailable or uncorroborated. This is an issue exacerbated by the fact that in many jurisdictions domestic violence directed to the mother is a bifurcated issue from harm caused (directly) to the child (Brown Williams 2011, p. 65). Even if evidence is recoverable, the policy of immediate return under the Convention contravenes the need of the court to assess the evidence, a procedure that would require time (Brown Williams 2011, p. 66). Nevertheless, the POAM project Best Practice Guide on the protection of abducting mothers in return proceedings sets out detailed guidance for courts and other authorities on matters related to evidence as they arise in return proceedings involving allegations of domestic violence, including an ‘evidence roadmap’ separately for documentary evidence, oral evidence and on navigating the evidence types.46

4.1.2. The ‘Child Objections’ Exception (Article 13(2))47

Article 13(2) states: ‘The judicial or administrative authority may also refuse to order the return of the child if it finds that the child objects to being returned and has attained an age and degree of maturity at which it is appropriate to take account of its views’.

In cases involving allegations of domestic violence, the ‘grave risk of harm’ defence is often invoked, and in some cases successfully made out, in conjunction with the ‘child’s objections’ defence under Article 13(2) of the Convention (Trimmings et al. 2020, p. 85; Honorati 2020, p. 3). The defence of child objections can of course be made out also independently of the ‘grave risk of harm’ defence.

Judges in all contracting states should be open to listening to children in return proceedings more frequently48 and, when reaching a decision on the return application, should attach importance to the child’s account of the incidents of domestic violence that occurred prior to the abduction and the impact of these incidents on him/her and/or the abducting mother. For example, in the UK, children as young as seven and half are routinely given the opportunity to be heard in return proceedings. This approach can be traced back to a 2006 House of Lords decision in the case of Re D (Abduction: Rights of Custody)49 and is recommended here as a model to follow by judges in other contracting states.

4.1.3. Human Rights Considerations (Article 20)

The product of a “laudable attempt” to compromise and resolve opposing views by the Convention’s drafters, Article 20 is no mere public policy clause (Pérez-Vera 1982, para. 33; Trimmings and Beaumont 2014; Weiner 2004). It transcends academic arguments on the rule of law and erosion of comity, developing a unique evaluation of factual circumstances when international human rights agreements may disrupt the Convention’s objectives. The exception under Article 20 of the 1980 Hague Convention provides that ‘the return of the child under the provisions of Article 12 may be refused if this would not be permitted by the fundamental principles of the requested State relating to the protection of human rights and fundamental freedoms.’ Even so, decades since the operation of the 1980 Hague Convention, Article 20 is seldomly utilised and, as it turns out, rarely successfully.50

Article 20 confirms that a refusal to return on human rights grounds is based on the internal laws of the requested state; that is to say that the source and foundation of ‘the fundamental principles of the requested state’ is to be found in national laws. But in reality, the national laws on human rights of Contracting States to the 1980 Hague Convention are influenced by, if not completely founded on, international treaties. There is a level of certainty and uniformity in human rights standards. The most obvious being the body of international human rights treaties51 created under the auspices of the United Nations as well as the European Convention on Human Rights52 (‘ECHR’). Often interconnected with the laws on immigration, these provisions may be invoked on the basis that the protection of human rights and fundamental freedoms are threatened by war zones, persecution on the basis of race, religion, political stance, nationality or membership of a particular group. There is also a scope to engage other international human rights treaties, such as the 1951 Refugee Convention and the 1967 Optional Protocol and the Council of Europe Convention on preventing and combating violence against women and domestic violence (‘the Istanbul Convention’).53 In summary, human rights principles applicable in a Contracting State are more likely than not to be a mirror of international agreements.

The majority of available cases54 that engage Article 20 show that the provision is often an anchor to core arguments based on the grave risk of harm or a child’s objections. This is because grounds based on domestic violence (on the basis of the abducting parent being a female) and the gravity and impact on the child are usually pleaded under Article 13(1)b).55 Similarly, grounds based on unsettled political environments for example, may be pleaded under the grave risk of harm. Indeed, the Guide to Good Practice on Article 13(1)b) highlights that risks associated with circumstances in the State of habitual residence such as political, economic or security situations may fall under asserted grave risks of harm (Guide to Good Practice on Article 13(1)(b) 2020, para. 61). As such, reliance on Article 20 is generally sparse in comparison to the other exceptions.

Of course, human rights grounds should not exclude invoking an argument based on Article 6 ECHR (right to fair hearing)56 or Article 8 ECHR (right to private and family life), but this is not what is envisaged under Article 20. These overarching arguments have been considered in cases with courts finding that Convention objectives do take into account and allow for consideration of ECHR values.57 What is also true is that Article 13(1)(b) in particular, when applied correctly, ensures that the court is not acting in a manner that is incompatible with human rights treaties such as the ECHR. Likewise, the 1989 Convention on the Rights of the Child (‘UNCRC’) has an integral role to play in upholding a child’s fundamental human rights and freedoms in return proceedings, and this appears to be naturally engaged.

The position in case law on the interplay between Article 20 and the protection of human rights pursuant to the Istanbul Convention is underdeveloped with decisions at times reiterating that Article 13(1)(b) is ample to plead domestic violence. In essence, where the overarching defence is based on domestic violence and a compelling public law element cannot be made out, relying on Article 13(1)(b) should suffice. For example, in the case of G (A Child: Child Abduction)58, the English Court of Appeal was concerned about unduly extending the scope of Article 20 when it was raised,59 and this was in the context of a principal claim relating to allegations of domestic violence. The UK Supreme Court60 in the same case reiterated that the provision should not be used ‘as a way around the rigours of the other exceptions to the return of the child’.61

Nevertheless, aside from an Article 13(1)(b) case on the grounds of domestic violence, a possible subsidiary argument is that domestic violence is a form of persecution pursuant to the 1951 Refugee Convention (Momoh 2023, p. 230). Further to this, women, as a particular social group within the meaning of Article 1A of the Refugee Convention are entitled to seek refuge and rely on the principle of non-refoulment62 where they have fled a country that is unable to protect them or other country where their life would be threatened (UN High Commissioner for Refugees 1990). Indeed, establishing a well-founded fear requires a subjective and an objective element.63 In the context of return proceedings under the 1980 Hague Convention, the fear of persecution may be the domestic violence perpetrated on the abducting mother, her gender being a protected characteristic, and the lack of adequate protection by the State of habitual residence being an objectively justifiable basis for human rights violations that give rise to an Article 20 case. Inadequate protection by the State of habitual residence has been demonstrated in cases such as Walsh v Walsh64 when the First Circuit refused a return order because the father’s perpetual disobedience of orders meant that any protective measures would be ineffective, or in State Central Authority, Secretary to the Department of Human Services v Mander,65 where consideration was given to the left-behind father’s behaviour, including a history of disobeying orders and violating undertakings in the home country. In Friedrich v Friedrich66 it was opined that there may be circumstances where ‘the court in the country of habitual residence, for whatever reason, may be incapable or unwilling to give the child adequate protection’.67

As noted, Article 20 may also be engaged where an abducting parent asserts that theirs (and the child’s) human rights and fundamental freedoms are in jeopardy in the State of habitual residence. The argument follows that there is a need for protection in the country of refuge, thus claiming asylum.68 One of the core arguments that have arisen in English jurisprudence is whether such application, successful or pending, places a bar to a return order being made under the 1980 Hague Convention.69 English court precedent is, arguably, significant in shaping a position that makes clear that a return order would break the principle of non-refoulement. In essence, that it would be impossible to make orders of a procedural nature (return orders) which would be in direct conflict with the substantive nature of the relief that is granted under the 1951 Refugee Convention and, indeed, the 1998 Human Rights Act.70 This position was reiterated in the Court of Appeal71 and subsequent Supreme Court decision of G v G (international child abduction).72 The case of G v G concerned the applicant father’s application for the return of the parties’ daughter (‘G’) to South Africa. The respondent mother opposed the return relying on Article 13(1)(b) (grave risk of harm) and Article 13(2) (child objections). Although not formally pleaded in legal arguments, Article 20 was raised on her behalf. The mother relied on facts that included allegations of domestic violence including sexual and racial abuse and aggressive and controlling behaviour, compounded by a vulnerability as a result of her mental health. The mother was also found to be HIV positive, the source of which was a matter of dispute. During the return proceedings, the mother revealed that she had feelings for women but had been brought up to believe that homosexuality was a sin. The mother applied for asylum in England, including the child as her dependant. At trial level, the order of Lieven J stayed the father’s return application pending the determination of the asylum application by the Secretary of State for the Home Department. On appeal, it was found that children who have been granted refugee status or have pending asylum applications are protected by the principle of non-refoulment; however, it was determined that because G did not have an independent asylum application, a return order could be made. This was overturned by the Supreme Court, which held that a child named as a dependant on a parent’s asylum application is also protected from refoulement. This meant that even if a court made a return order, the principle of non-refoulement applied so as to prevent the implementation of such an order. The Supreme Court in G v G also considered practical and desirable steps to take in future cases where the two Conventions apply. This included acknowledging that the Secretary of State has sole responsibility for both examining and determining claims for international protection. As a result of the decision in G v G, the Secretary of State has set up a Specialist Asylum Team to expedite such cases (Home Office 2021).

Despite the suggested uniformity across jurisdictions on the basis that internal laws have drawn inspiration from similar international treaties, a level of discord had previously arisen. Distinguishable from the English jurisdiction were decisions in the US and Canadian courts, where effectively the Hague court got another bite at the cherry. For example, in the Canadian decision of Court of Appeal (Ontario) in AMRI v KER73, it was found that even though refugee status had been granted to the mother and daughter, the Hague court may revisit and make a return order, bearing in mind what was considered a mere ‘rebuttable presumption’ as opposed to a bar to return. In the US Court of Appeal (first circuit) in the decision of Sanchez v RGL74, it was also found that the grant of asylum was not determinative in return proceedings.75 More recently, however, the Ontario Court of Appeal’s decision in the case of M.A.A v D.E.M.E 76 found that family courts cannot issue return orders for children if their applications for asylum are still pending. This is an encouraging progress across the Atlantic that aligns with the developments in the English courts and, it is hoped, will become the norm. Evidently, Article 20, like all the other exceptions to return, ought to be interpreted in a restrictive fashion. Having said that, arguably, where friction arises between parental child abduction law and international refugee law, it should be approached from a humanitarian perspective, where due regard is had to the substantive (rather than procedural) nature of the relief sought. It would not open the floodgates as indeed the anchoring of Article 13(1)(b) and Article 20 reminds courts that protection from a well-founded fear of prosecution amply qualifies as a grave risk of harm.

4.2. Legislative Interventions

Legislative interventions can be contemplated at the global level or the domestic level.

4.2.1. Global Level

Amending the 1980 Hague Convention

At the global level, the most extreme but, admittedly, least practicable solution would be for the Hague Conference on Private International Law as the global law-making body in the area of private international law to amend the wording of the 1980 Convention to take account of the concerns over the safety of abducting mothers in return proceedings. This could take, for example, the form of a separate exception to return on the grounds of domestic violence or a wholly separate ‘pathway’ for applications involving allegations of domestic violence, including provisions related to evidentiary matters; legal aid; the availability of alternative dispute resolutions methods; channels for direct judicial communication; and the availability of psychological and other support services to the abducting mother during the return proceedings. However, as alluded to above, this solution lacks feasibility as the process of amending an international convention is complex in itself and becomes even more challenging where a large number of contracting parties is involved as is the case of the 1980 Hague Convention.77

Amending an international convention refers to the formal modification of the convention provisions affecting all the contracting parties. Such alterations must be effected with the same formalities that attended the original formation of the treaty. Where the convention does not lay down specific requirements to be satisfied for amendments to be adopted (as is the case with the 1980 Hague Convention), amendments require the consent of all the parties.78 The ‘stone tablet quality’ of international conventions makes it extremely unlikely that the contracting parties to the Convention would come down in favour of a revision of the instrument (Thorpe 2006, p. 10).

Adopting a Protocol to the 1980 Hague Convention

An alternative option would be the adoption of a Protocol to the Convention.79 This form of legislative intervention is more pragmatic than amending the Convention; however, it has other shortcomings. Most importantly, the fact that contracting parties to the Convention are not bound to participate in a Protocol initiative would mean that the safety of abducting mothers would be guaranteed at a restricted scope only. Unfortunately, this would significantly lessen the value of the Protocol. Nevertheless, one can agree with Thorpe LJ that the Protocol would ‘at least enable like-minded States to strengthen the Convention inter se’ and that ‘a Protocol with a limited range of operation would be better than no Protocol at all’ (Thorpe 2006, p. 10).

4.2.2. Domestic Level

At the national level, contracting parties could adopt new or amend relevant domestic legislation to clarify that allegations of domestic violence including the safety of the abducting mother should be considered before a return order is made for the child under the 1980 Hague Convention. A recent example of such legislative intervention is an Australian piece of legislation,80 which provides safeguards to mothers and children fleeing domestic violence when Australian courts consider cases brought under the 1980 Hague Convention (‘the 2022 Regulations’). The 2022 Regulations make clear inter alia that domestic violence is a consideration under the ‘grave risk of harm’ exception to return and a court does not need to be satisfied that such violence has occurred or will occur before it is taken into account (The Hon Mark Dreyfus KC MP 2022). It is recommended that domestic legislation includes also supplementary provisions to strengthen the position of abducting mothers who had fled domestic violence and are involved in return proceedings. Such provisions could pertain to matters such as legal aid, availability of ADR channels, and a legal basis for the use of and the functioning of direct judicial communication (see below ‘Alternative avenue: ADR/mediation’).

4.3. Alternative Avenue: ADR/Mediation

The use of alternative methods of dispute resolution (‘ADR’), and specifically mediation,81 for the resolution of domestic family disputes is an alternative avenue to court proceedings. The popularity of ADR, including mediation, has grown significantly over the past decades.

When domestic violence is involved or even suspected as the reason behind an international child abduction, mediation becomes a questionable option.82 Experts point out that mediation can do more harm than good in disputes involving abusive relationships. The concerns are threefold. From the victim’s perspective, participating in mediation (or any ADR mechanism) will result in delayed access to the court83 and therefore court orders to protect the victim.84 For victims that have already distanced themselves from their abuser, mediation can result in a risk of physical or mental harm or even re-traumatisation (González Martín 2014, p. 343; see also Permanent Bureau of the Hague Conference on Private International Law 2012, p. 73). Another group of concerns is related to the integrity of the mediation process. The existence of domestic violence often comes hand in hand with broken-down communication, toxic dynamics and severe power imbalance between the abuser and the victim. Accordingly, during the mediation, the victim might be unable to voice concerns equally to the abuser, leading to a potentially disadvantageous or coerced result. In the context of international child abduction, there are additional dimensions that might make the mediation process more challenging; these primarily relate to the cross-border element and, in particular, the cultural diversity, potential language barriers and the need for close cooperation between the Central Authorities of the states involved (González Martín 2014, p. 343). The final type of concern is that of policy. Mediation is founded on the objective of reaching a mutually agreeable solution to a private dispute; therefore, by definition, it arguably becomes inappropriate when domestic violence is present because reaching a private agreement results in ‘no-punishment’ and even normalisation of domestic violence (González Martín 2014, p. 343; Permanent Bureau of the Hague Conference on Private International Law 2012, p. 73).

Despite the above legitimate concerns, ‘mediation has particular advantages over litigation in international child abduction cases,’ and ‘inherent benefits […] regardless of the outcome’ (González Martín 2014, p. 322; Vigers 2011, p. 71). From the victim’s perspective, elective mediation offers a strategic route to an acceptable arrangement. Particularly in cases where the abducting mother returns with the child to the state of habitual residence, mediation can be a significantly better option to litigation. Empirical research into abducting mothers post-Hague proceedings has shown that victims of domestic violence that have fled and subsequently returned (following a return order) face a wreath of issues from returning to the abusive context they fled from, homelessness and domestic litigation on the custody and related issues regarding the child, often resulting in mother–child separation (Masterton et al. 2022, pp. 376–81; Quillen 2014, p. 641). Sometimes, returning mothers might even face ‘criminal prosecution, extradition and incarceration’ (Alanen 2008, p. 52). The findings demonstrate that victims experience multifaceted and severe consequences for child abduction despite their actions being driven by domestic violence. It is argued that mediation can help ease some of these consequences upon the return of the mother and child and make the experience of return less traumatizing. Mediation is a flexible process that allows the parties to broaden its scope beyond the child’s return to consider a broader range of issues such as custody, visitation and living arrangements. The victim can utilize the context of the mediation, and the tools made available there, to reach an agreement that, in hindsight, might be more favourable than a court order.85 Further, ‘prosecutors (…) might drop criminal charges once the child is returned to the custodial parent or the parents have stipulated to a valid, enforceable parenting agreement’ (Alanen 2008, p. 52; Quillen 2014, p. 641; González Martín 2014, p. 337).

Apart from the more controlled outcome, elective mediation can have the opposite effect from what is feared by specific experts; instead of silencing the victim, it can empower her to make reasonable requests that will improve the entire family’s quality of life. It can be forgotten that abducting mothers are victims that have found the courage to flee, and fleeing is the first step in their journey of empowerment. Accordingly, stripping the victims of choice to mediate is counterintuitive.86 When administered by a domestic violence-informed mediator(s), the mediation process can be tailored in multiple ways to allow space for the victim and avoid any further traumatization (González Martín 2014, p. 342). For instance, mediation may not necessarily be delivered face-to-face (Vigers 2011, p. 23). It can instead be delivered entirely online so that the victim feels physically safe. In ‘shuttle’ mediation, an experienced mediator will make use of techniques such as ‘face-saving’, whereby the mediator has private meetings with either party and puts forward the parties’ requests to each other in a controlled and strategic manner to diffuse high-emotion and promote a mutually agreeable outcome (Whatling 2012, pp. 49 and 157). The entire process is carried out without the parties coming in contact. However, even in the case of face-to-face mediation, the process can be built and tailored in a manner suitable to the specific circumstances, with as many private and joint sessions as necessary to work through the issues.87 In meetings where the victim and the abuser are in the same space, an experienced mediator will be mindful of power imbalance and act as an equaliser, ensuring both sides are heard. In this context, mediation becomes a safe space for ‘the victim to have a voice, to not fear repercussions’ (Kucinski 2010, p. 318). Reported accounts of past mediations administered under the auspices of Reunite International note that ‘the victim often becomes empowered and finds a voice, and grows during the [mediation] process, more so than in a courtroom’ (Kucinski 2012, p. 84).88

Mediation cannot occur in a vacuum, and of course, not all cases will be suitable to mediate. It is pertinent that the selection process is performed by an expert who is trained to identify signs of domestic violence and is able to adapt the process accordingly. Further, it is essential that the victim receives independent legal advice on what the mediation process entails and her specific circumstances to aid in deciding. Family mediators are supportive of a case-by-case assessment of suitability; instead of a pre-determined approach, ‘informed consent and thorough assessment’ can maximise the positive impact of mediation on the lives of victims and their children.89

5. Conclusions

Over the past nearly fifteen years, the interplay between international child abduction and domestic violence has generated attention and divided positions amongst academic commentators and judges in the Contracting Parties to the 1980 Hague Convention. The change in the profile of a typical parental child abductor, combined with better understanding of the seriousness and impact of domestic violence on the victims and, by extension, their children, has led to increased awareness of the need to safeguard protection of abducting mothers in child abduction cases committed against the background of domestic violence. This article has proposed several measures that could help achieve this objective, ranging from possible legislative interventions at the global level (e.g., a Protocol to the 1980 Convention) to judicial interventions to be employed on a case-by-case basis when applying the exceptions to return available under the 1980 Convention, in particular Article 13(1)(b), Article 13(2) and Article 20. Additionally, the role of Article 20 has been explored in the context of an interplay between domestic violence and asylum claims. This niche area of law would benefit from concentrated and comprehensive research. Nevertheless, it ought to be said that defending a return order on asylum grounds pursuant to Article 20 has earned a standing of its own right. At the very least, it engages the 1951 Refugee Convention, specifically Article 33. Further, it aligns with the observations of Professor Peréz-Vera that to invoke Article 20 is to address the contradiction between the 1980 Hague Convention and domestic human right laws, as well as establishing how such a return would breach the protective principles of human rights (Pérez-Vera 1982, para. 3).