Julia Hörnle, Professor of Internet Law, CCLS, Queen Mary University of London[1]

It is now well known that internet users are widely tracked and profiled by a range of actors and the advancements in data science mean that such tracking and profiling is increasingly commercially profitable[2]. This raises difficult questions about how to balance the value of data with individual privacy. But since there is no point in having privacy (or data protection) rights if no redress can be found to vindicate them, it is even more important to investigate how internet users can obtain justice, if their privacy has been infringed. Given the power of Big Tech Companies, their enormous financial resources, cross-jurisdictional reach and their global impact on users’ privacy, there are two main litigation challenges for successfully bringing a privacy claim against Big Tech. One is the jurisdictional challenge of finding a competent court in the same jurisdiction as the individual users.[3] Secondly, the challenge is how to finance mass claims, involving millions of affected users. In privacy claims it is likely that there is significant user detriment, potentially with long-term and latent consequences, which are difficult to measure. This constellation provides a strong argument for facilitating collective redress, as otherwise individual users may not be able to obtain justice for privacy infringements before the courts. In privacy infringement claims these two challenges are intertwined and present a double-whammy for successful redress. Courts in a number of recent cases had to grapple with questions of jurisdiction in consumer collective redress cases in the face of existing provision on consumer jurisdiction and collective redress, which have not (yet) been fully adapted to deal with the privacy challenges stemming from Big Tech in the 21^st century.

In Case C-498/16 Max Schrems v Facebook Ireland[4] the Court of Justice of the EU in 2018 denied the privilege of EU law for consumers to sue in their local court[5] to a representative (ie Max Schrems) in a representative privacy litigation against Facebook under Austrian law. By contrast, courts in California and Canada have found a contractual jurisdiction and applicable law clause invalid as a matter of public policy in order to allow a class action privacy claim to proceed against Facebook.[6] In England, the dual challenge of jurisdiction and collective actions in a mass privacy infringement claim has presented itself before the English Courts, first in Vidal-Hall v Google before the Court of Appeal in 2015[7] and in the Supreme Court judgment of Google v Lloyd in November 2021[8]. Both cases concerned preliminary proceedings on the question of whether the English courts had jurisdiction to hear the action, ie whether the claimant was able to serve Google with proceedings in the USA and have illustrated the limitations of English law for the feasibility of bringing a collective action in mass-privacy infringement claims.

The factual background to Vidal- Hall and Lloyd is the so-called “Safari workaround” which allowed Google for some time in 2011-2012 to bypass Apple privacy settings by placing DoubleClick Ad cookies on unsuspecting users of Apple devices, even though Safari was trying to block such third party cookies, used for extensive data collection and advertising. The claimants alleged that this enabled Google to collect personal data, including sensitive data, such as users’ interests, political affiliations, race or ethnicity, social class, political and religious beliefs, health, sexual interests, age, gender, financial situation and location. Google additionally creates profiles from the aggregated information which it sells. The claim made was that Google as data controller had breached the following data protection principles set out in the Data Protection Act 1998 Schedules 1 and 2: 1^st (fair and lawful processing), 2^nd (processing only for specified and lawful purposes) and 7^th (technical and organizational security measures). In particular, it was alleged that Google had not notified Apple iPhone users of the purposes of processing in breach of Schedule 1, Part II, paragraph 2 and that the data was not processed fairly according to the conditions set out in Schedules 2 and 3.

Vidal-Hall^[9] concerned the first challenge of jurisdiction and in particular whether the court should allow the serving of proceedings on the defendant outside the jurisdiction under the Civil Procedure Rules[10]. For privacy infringement, previous actions had been brought under the cause of action of breach of confidence^[11], which is a claim in equity and, thus it was unclear whether for such actions jurisdiction lies at the place of where the damage occurs. The Court of Appeal held that misuse of private information and contravention of the statutory data protection requirements was a tort and therefore, if damage had been sustained within England, the English courts had jurisdiction and service to the USA (California) was allowed.

The second hurdle for allowing the case to proceed by serving outside the jurisdiction was the question of whether the claimant was limited to claiming financial loss or whether a claim for emotional distress could succeed. The Court of Appeal in Vidal-Hall decided that damages are available for distress, even in the absence of financial loss, to ensure the correct implementation of Article 23 of the (then) Data Protection Directive, and in order to comply with Articles 7 and 8 of the Charter of Fundamental Rights of the EU. The Court therefore found that there was a serious issue to be tried and allowed service abroad to proceed, at which point the case settled.

The more recent English Supreme Court judgment in Lloyd concerned the second challenge, collective redress. As pointed out by Lord Leggatt in the judgment, English procedural law provides for three different types of actions: Group Litigation Orders (CPR 19.11), common law representative actions, and statutory collective proceedings under the Competition Act 1998. Their differences are significant for the purposes of litigation financing in two respects: first the requirement to identify and “sign-up” claimants and secondly, the requirement for individualized assessment of damages. Since both these requirements are expensive, they make collective redress in mass privacy infringement cases with large numbers of claimants impractical.

Group actions require all claimants to be identified and entered in a group register (“opt-in”) and are therefore expensive to administer, which renders them commercially unviable if each individual claim is small and if the aim is to spread the cost of litigation across a large number of claimants.

English statutory law in the shape of the Competition Act 1998 provides for collective proceedings before the Competition Appeal Tribunal in competition law cases only.[12] Since the reforms by the Consumer Rights Act in 2015, they can be brought under an “opt-in” or “opt-out” mechanism. Opt-out means that a class can be established without the need for affirmative action by each and every member of the class individually. The significance of this is that it is notoriously difficult (and expensive) to motivate a large number of consumers to join a collective redress scheme. Human inertia frequently prevents a representative claimant from joining more than a tiny fraction of those affected. For example, 130 people (out of 1.2-1.5 million) opted into the price-fixing case against JJB Sports concerning replica football shirts.[13] Likewise, barely 10,000 out of about 100,000 of Morrison’s employees joined the group action against the supermarket chain for unlawful disclosure of private data on the internet by another employee.[14] Furthermore, s.47C (2) of the Competition Act obviates the need for individual assessment of damages, but limits the requirement to prove damages to the class as a whole, as an aggregate award of damages, as held by Lord Briggs in Merricks v Mastercard[15]. However no such advanced scheme of collective redress has yet been enacted in relation to mass privacy infringement claims.

While the Supreme Court held that Mr Lloyd’s individual claim had real prospect of success, the same could not necessarily be said of everyone in the class he represented. This case was brought as a representative action where Mr Lloyd represented the interests of everyone in England and Wales who used an iPhone at the relevant time and who had third party cookies placed by Google on their device. One of the interesting features of representative actions is that they can proceed on an opt-out basis, like the collective actions under the Competition Law Act. Common law representative actions have been established for hundreds of years and have now been codified in CPR Rule 19.6: “Where more than one person has the same interest in a claim by or against one or more of the persons who have the same interest as representatives of any other person who have that interest”. Thus representative actions are based on the commonality of interest between claimants. The pivotal issue in Lloyd was the degree of commonality of that interest and in particular, whether this commonality must extend to the losses, which claimants have suffered, and proof of damages.

Lord Leggatt in Lloyd emphasized the spirit of flexibility of representative actions. Previous caselaw in the Court of Appeal had held that it was possible for claimants to obtain a declaration by representative action, which declares that they have rights which are common to all of them, even though the loss and amount of damages may vary between them.[16] He held that a bifurcated approach was permissible: a representative action can be brought seeking a declaration about the common interests of all claimants, which can then form the basis for individual claims for redress. Lord Leggatt held that, depending on the circumstances, a representative action could even be brought in respect of a claim for damages, if the total amount of damages could be determined for the class as a whole, even if the amount for each individual claimant varied, as this was a matter which could be settled between the claimants in a second step. He held that, therefore, a representative action can proceed even if a claim for damages was an element of the representative action, as in Lloyd.

Lord Leggatt found that the interpretation of what amounts to the “same interest” was key and that there needed to be (a) common issue(s) so that the “representative can be relied on to conduct the litigation in a way which will effectively promote and protect the interests of all the members of the represented class.”[17] The problem in Lloyd was that the total damage done to privacy by the Safari workaround was unknown.

Lord Leggatt saw no reason why a representative action for a declaration that Google was in breach of the Data Protection Act 1998, and that each member was entitled to compensation for the damage suffered as a consequence of the breach, should fail. However, commercial litigation funding in practice cannot fund actions seeking a mere declaration, but need to be built on the recovery of damages, in order to finance costs. In order to avoid the need for individualised damages, the claim for damages was formulated as a claim for uniform per capita damages. The problem on the facts of this case was clearly that the Safari workaround did not affect all Apple users in the same manner, as their internet usage, the nature and amount of data collected, as well as the effect of the data processing varied, all of which required individualised assessment of damages.

For this reason, the claimant argued that an infringement of the Data Protection Act 1998 leads to automatic entitlement to compensation without the need to show specific financial loss or emotional distress. This argument proved to be ultimately unsuccessful and therefore the claim failed. The Court examined Section 13 of the Data Protection Act 1998, entitling the defendant to compensation for damage, but the court held that each claimant had to prove such damage. The level of distress varied between different members of the represented class, meaning that individual assessment was necessary.

The claimant sought to apply the cases on the tort of misuse of private information by analogy. In this jurisprudence the courts have allowed for an award of damages for wrongful intrusion of privacy as such, without proof of distress in order to compensate for the “loss of control” over formerly private information.[18] Lord Leggatt pointed out that English common law now recognized the right to control access to one’s private affairs and infringement of this right itself was a harm for which compensation is available.

However in this particular case the claim had not been framed as the tort of misuse of private information or privacy intrusion, but as a breach of statutory duty and Lord Leggatt held that the same principle, namely the availability of damages for “loss of control” did not apply to the statutory scheme. He pointed out that it may be difficult to frame a representative action for misuse of private information, as it may be difficult to prove reasonable expectations of privacy for the class as a whole. This may well be the reason that the claim in this case was based on breach of statutory duty in relation to the Data Protection Act. Essentially the argument that “damages” in Section 13 (1) included “loss of control” over private data was unsuccessful. Both Article 23 of the Data Protection Directive and Article 13 made a distinction between the unlawful act (breach of data protection requirements) and the damage resulting, and did not conceive the unlawful act itself as the damage. Furthermore, it was not intended by the Directive or the Act that each and every contravention led to an entitlement to damages. He held that “loss of control” of personal data was not the concept underlying the data protection regime, as processing can be justified by consent, but also other factors which made processing lawful, so the control over personal data is not absolute.

Furthermore, it did not follow from the fact that both the tort of misuse of private information and the data protection legislation shared the same purposes of protecting the right to privacy under Article 8 of the European Convention of Human Rights that the same rule in respect of damages should apply in respect of both. There was no reason “why the basis on which damages are awarded for an English domestic tort should be regarded as relevant to the proper interpretation of the term “damage” in a statutory provision intended to implement a European directive”.[19] He concluded that a claim for damages under Section 13 required the proof of material damage or distress. He held that the claim had no real prospect of success and that therefore no permission should be given to serve proceedings outside the jurisdiction (on Google in the US).

This outcome of Lloyd raises the question in the title of this article, namely whether the cross-border battle on collective actions in mass privacy infringement cases against Big Tech has been lost, or whether on the contrary, it has just begun. One could argue that it has just began for the reason that the facts underlying this case occurred in 2011-2012, and therefore the judgment limited itself to the Data Protection Act 1998 (and the then Data Protection Directive 1995/46/EC). Since then the UK has left the EU, but has retained the General Data Protection Regulation[20] (“the UK GDPR”) and implemented further provisions in the form of the Data Protection Act 2018, both of which contain express provisions on collective redress. The GDPR provides for opt-in collective redress performed by a not-for-profit body in the field of data protection established for public interest purposes.[21] This is narrow collective redress as far removed from commercial litigations funders as possible. Because of the challenge of financing cross-border mass-privacy infringements claims, this is unlikely to be a practical option. The GDPR makes it optional for Member States to provide that such public interest bodies are empowered to bring opt-out collective actions for compensation before the courts.[22] These provisions unfortunately do not add anything to common law representative actions or group actions under English law. As has been illustrated above, representative actions can be brought on an “opt-out” basis, but have a narrow ambit in that all parties must have the same interest in the claim and Lloyd has demonstrated that in the case of distress this communality of interest may well defeat a claim. For group actions the bar of communality is lower, as it may encompass “claims which give rise to common or related issues of fact or law”[23]. But clearly the downside of group actions is that they are opt-in. Therefore, while English law recognizes collective redress, there are limitations to its effectiveness.

The Data Protection Act 2018 imposes an obligation on the Secretary of State to review the provision on collective redress, and in particular, consider the need for opt-out collective redress, and lay a report before Parliament. This may lead to Regulations setting out a statutory opt-out collective redress scheme for data protection in the future.[24] This Review is due in 2023.

Thus, the GDPR and the Data Protection Act 2018 have not yet added anything to the existing collective redress. It can only be hoped that the Secretary of State reviews the collective redress mechanisms in relation to data protection law and the review leads to a new statutory collective redress scheme, similar to that enacted in respect of Competition Law in 2015, thereby addressing the challenge of holding Big Tech to account for privacy infringement.[25]

However the new data protection law has improved the provision of recoverable heads of damage. This improvement raises the question, if the issues in Lloyd had been raised under the current law, whether the outcome would have been different. The Data Protection Act 2018 now explicitly clarifies that the right to compensation covers both material and non-material damage and that non-material damage includes distress.[26] Since non-material damage is now included in the Act, the question arises whether this new wording could be interpreted by a future court as including the privacy infringement itself (loss of control over one’s data). Some of the arguments made by Lord Leggatt in Lloyd continue to be relevant under the new legislation, for example that the tort of statutory breach is different from the tort of misuse of private information and that not each and every (minor) infringement of a statute should give raise to an entitlement for damages. Nevertheless it is clear from the new Act that non-material damage is included and that non-material damage includes distress, but is wider than distress. This means that claimants should be able to obtain compensation for other heads of non-material damage, which may include the latent consequences of misuse of personal information and digital surveillance. There is much scope for arguing that some of the damage caused by profiling and tracking are the same for all claimants. A future representative action in an equivalent scenario may well be successful. Therefore, the battle for collective action against Big Tech companies’ in privacy infringement cases may just have begun.

[1] J.hornle@qmul.ac.uk

[2] Shoshana Zuboff The Age of Surveillance Capitalism (2018)

[3] See further J. Hörnle, Internet Jurisdiction Law & Practice (OUP 2021)

[4] ECLI:EU:C:2018:37; discussed further in J. Hörnle fn 1 Chapter 8

[5] Ie the courts of the consumer’s domicile, if the business directed their activities to that state, Art 17 and Art 18 (1) Brussels Regulation on Jurisdiction (EU) 1215/2021

[6] In Re Facebook Biometric Information Privacy Litigation 185 F.Supp.3d 1155 (US District Court N.D. California 2016) and Douez v Facebook [2017] SCC 33; discussed further in J. Hörnle fn 1 Chapter 8

[7] [2016] QB 1003

[8] [2021] 3 WLR 1268

[9] [2015] 3 WLR 409 (CA)

[10] CPR PD 6B para.3.1(9)

[11] Campbell v MGN Ltd [2004] 2 WLR 1232 (HL)

[12] Section 47B

[13] The Consumers Association v JJB Sports Plc [2009] CAT 2

[14] Various Claimants v WM Morrisons Supermarkets Plc [2017] EWHC 3113 (QB)

[15] [2021] Bus LR 25, para 76

[16] David Jones v Cory Bros & Co Ltd (1921) 56 LJ 302; 152 LT Jo 70

[17] Paras 71-74

[18] by Mann J affirmed in the Court of Appeal Gulati v MGN Ltd [2017] QB 149

[19] Para 124

[20] Regulation (EU) 2016/679, L119, 4 May 2016, p. 1–88

[21] Art 80 (1)

[22] Arts 79 and 80 (2) in relation to effective judicial remedies

[23] CPR Part 19- Group Litigation Orders, Rules 19.10, 19.11

[24] ss. 189-190

[25] The current government, however seems to march in the opposite direction, see Consultation on reform of data protection law https://www.gov.uk/government/consultations/data-a-new-direction

[26] S. 168 (1)

Guest post by Willem Theus, PhD Researcher (KULeuven, cotutelle with UCLouvain)

The history of private international law (or ‘conflict of laws’) is incomplete. Private international law textbooks have always referred to the essentials of the history of our discipline.[1] However, these essentials are often solely based on the history of conflict of laws in the West and on the works of western authors such as Huber, Von Savigny and Story. It is undoubtedly true that these authors played an important role and that the  “modern” conflict of laws finds it origin in 19^thcentury Europe, when the split between private and public international law occurred.[2] This is however only one part of history.

Conflict of laws systems have been around much longer and are definitely not uniquely western. They were already present in the very first civilizations, with some rules of that ancient history still resembling our present-day rules.[3]Conflict of laws is “the body of law that aims to resolve claims involving foreign elements”.[4] A state or international border is therefore not required to have a conflict of laws system,[5] only different jurisdictions and laws (i.e. legal pluralism[6]) are. A distinction could therefore be made between “external” (i.e. crossing an international State border) conflict of laws or private international law and “internal” conflict of laws (i.e. within one State).[7] Both the historical research and the contemporary study of our field should arguably reflect much more on precolonial and/or non-western conflict of laws systems and on the unique linkage between the national (or “internal”) and international (or “external”) spheres. This is especially so given that “external” conflict of laws rules seem to sometimes guide “internal” conflict of laws cases.[8] I offer one historical example to highlight the new perspectives that such a widening of scope could offer.

In a not so distant and colonial past, there were multiple “internationalized” or mixed courts in various regions and nations. The last such mixed court only closed its doors in 1980.[9] In general, mixed courts were local courts that employed a mixed (read mostly Western) bench, bar and legal system to deal with legal conflicts that had a mixed or “foreign” element, i.e. conflicts not exclusively related to one local or foreign resident population.[10] Those exclusively local or intra-foreigner  -of the same nationality-  legal conflicts were often dealt with by various local or consular courts. The mixed or “foreign” element was however often widely interpreted and therefore quickly kicked in, leading to overlapping jurisdictions in many instances and therefore to a conflict of laws system.

An example of such a set-up is the Tangier International Zone (1923-1956), a treaty-based multinational run zone, which remained under the Sovereignty of the Sultan of Morocco. It had various multinational institutions with local involvement. In the Zone, five different legal systems co-existed, each with their own courts. These were the American Consular Court, the Special Tribunal of the State Bank of Morocco, the Moroccan Sharia courts, the Moroccan Rabbinical courts and the Mixed Court. The latter dealt with all cases that had a “foreign” element (except American as they went to the aforementioned American Consular Court).[11] Both “internal” and “external” conflict of law systems in fact overlap here. Indeed the Mixed Court and the two Moroccan courts were “local” courts with the judges being formally appointed by the Sultan, whereas the American Consular Court was in essence an ad hoc American court in Tangier. The Special Tribunal was some sort of early investment protection court with very limited jurisdiction.

Naturally, in such a set-up conflict of laws cases were frequent, as illustrated by the Toledano-case which came before the Mixed Court. In 1949 a dispute between the heirs of the large inheritance of a Tangerine Jew, Isaac Toledano, broke out. The key question concerned the nationality of Isaac – and as such the questions of jurisdiction and applicable law. During his lifetime Isaac had become a Spanish citizen by naturalization, yet he had seemingly always lived in Morocco. Had he somehow lost his Moroccan citizenship? If so, the mixed courts would have jurisdiction and Spanish law would apply, leading his inheritance to be divided under all his children, including his married daughters. If not, the rabbinical courts of Tangier and rabbinical law would apply, leading to his inheritance to only go to his sons and unmarried daughters. On appeal the court overturned the judgment of first instance that held that he had retained his Moroccan nationality. He was deemed to be Spanish and therefore Spanish law was to be applied.[12]

Such jurisdictional caselaw is only a part of this conflict of laws treasure trove. The caselaw of the mixed courts seemingly encompasses all types of conflict of laws questions and many other legal questions. I have to say seemingly, as the caselaw of the mixed courts has in recent times barely been studied and their archives (if known at all) are scattered throughout the globe. A closer look could undoubtedly open up new perspectives to conflict of laws, and some of these mixed courts’ experiences and case-law could perhaps help to guide ever-recurring questions of personal status matters regarding foreigners. The Emirate of Abu Dhabi has for example reintroduced special personal status provisions for non-Muslim foreigners as reported on conflictoflaws recently. The courts also offer new perspectives for public international law as certain mixed courts acted as “true” international courts when interpreting their treaties. An example is the Court of Appeal of Mixed Court of Tangier going against the International Court of Justice in 1954 when it held that it alone had the authority to provide authoritative interpretations of the Zone’s constitutive treaties.[13] The Mixed Courts could even open new perspectives to EU-law as many early key EU lawyers and judges have ties to certain Mixed Courts.[14] Much work is therefore still to be done. This piece is a call to arms for just that.

[1] Hatzimihail, N.E. (2021) Preclassical Conflict of Laws. Cambridge University Press 51-52.

[2] For an overview of this period see: Banu, R.  (2018). Nineteenth Century Perspectives on Private International Law. Oxford University Press

[3] Yntema, Hessel E. (1953). The Historic Bases of Private International Law. The American Journal of Comparative Law, vol. 2, no. 3, 301.  Yntema refers to the following text found in a Fayoum Papyri: “Contracts between Greeks-who had established colonies in Egypt (red.)-and Egyptians, if in Greek form, should be tried before the chrematists, the Greek courts; if in Egyptian form, before the laocrites, the native courts, in accordance with the laws of the country.”

[4] Okoli, C.S.A. (2020). Private International Law in Nigeria. Hart, 3.

[5] Okoli, Op.cit., 3-7; Yntema. Op.cit., 299

[6] For a good overview of the different meanings of this term see: Benda-Beckmann, B. & Turner, B. (2018). Legal Pluralism, Social Theory, and the State. Journal of Legal Pluralism and Unofficial Law, 50(3), 255–274

[7] This distinction is not new and is used in legislation. See for example: Non-application of This Regulation to Internal Conflicts of Laws. (2016). In A. Calvo Caravaca, A. Davì, & H. Mansel (Eds.), The EU Succession Regulation: A Commentary (pp. 521-529). Cambridge University Press.

[8] Okoli Op.cit, 3.

[9] Pacific Manuscripts Bureau,  Collection MS 1145: Judgements of the Joint Court of the New Hebrides. Retrieved from <https://asiapacific.anu.edu.au/pambu/catalogue/index.php/judgements-of-joint-court-of-new-hebrides> accessed 13 December 2021. It was known as a ‘Joint’ Court and not ‘Mixed’ as there were only two powers involved: France and the UK. Although in French it was still referred to as a Tribunal Mixte. Mixed Courts mostly existed in countries that were not-directly colonized, yet still under heavy Western influence such as Siam, China and Egypt. They were mostly founded due to western distrust for the local legal systems and build forth on the principle of personal jurisdiction (and the connected later principle of extraterritoriality and the connected Capitulations and Unequal Treaties).

[10] Erpelding, M. (2020). Mixed Courts of the Colonial Era. In Hélène Ruiz Fabri (ed), Max Planck Encyclopedia of International Procedural Law. Oxford University Press.

[11] Erpelding, M & Rherrousse, F. (2019) The Mixed Court of Tangier. In Héne Ruiz Fabri (ed), Max Planck Encyclopedia of International Procedural Law. Oxford University Press, paras 22-24.

[12] de Radigues de Chenneviere, C. (5 April 1949). ‘Procès Toledano’. Tangier, P 452/717, AF-12-A-3 (Diplomatic Archives of the Kingdom of Belgium)

[13] Grawitz, M. (1955). Arrêt du 13 août 1954. Annuaire français de droit international, 1(1), 324–328

[14] Erpelding, M. (2020). International law and the European Court of Justice: the Politics of Avoiding History, Journal of the History of International Law, 22(2-3), 446-471.

This post was contributed by Dr. Vito Bumbaca, who is Assistant Lecturer at the University of Geneva

In a ruling of 2 August 2021 (A v. B, C-262/21 PPU), the Court of Justice of the European Union (CJEU) clarified that a child who is allegedly wrongfully removed, meaning without consent of the other parent, should not return to his/ her habitual residence if such a removal took place as a consequence of the ordered transfer determining international responsibility based on the Dublin III Regulation. The judgment is not available in English and is the first ever emanating from this Court concerning the Brussels IIA-Dublin III interplay.

The Council Regulation (EC) No 2201/2003 of 27 November 2003 concerning jurisdiction and the recognition and enforcement of judgments in matrimonial matters and the matters of parental responsibility, repealing Regulation (EC) No 1347/2000 (Brussels IIA Regulation) complements the Hague Convention of 25 October 1980 on the Civil Aspects of International Child Abduction, and is applicable to 26 EU Member States, including Finland and Sweden. The Regulation (EU) No 604/2013 of 26 June 2013 establishing the criteria and mechanisms for determining the Member State responsible for examining an application for international protection lodged in one of the Member States by a third-country national or a stateless person (recast) (Dublin III), is pertinent for asylum seekers’ applications commenced at least in one of the 31 Dublin Member States (EU/EFTA), comprising Finland and Sweden, bound by this Regulation.

Questions for a CJEU urgent preliminary ruling:

The CJEU was referred five questions, but only addressed the first two.

‘(1) Must Article 2(11) of [Regulation No 2201/2003], relating to the wrongful removal of a child, be interpreted as meaning that a situation in which one of the parents, without the other parent’s consent, removes the child from his or her place of residence to another Member State, which is the Member State responsible under a transfer decision taken by an authority in application of Regulation [No 604/2013], must be classified as wrongful removal?

(2) If the answer to the first question is in the negative, must Article 2(11) [of Regulation No 2201/2003], relating to wrongful retention, be interpreted as meaning that a situation in which a court of the child’s State of residence has annulled the decision taken by an authority to transfer examination of the file, and to take no further action since the mother and child have left the State of residence, but in which the child whose return is ordered, no longer has a currently valid residence document in his or her State of residence, or the right to enter or to remain in the State in question, must be classified as wrongful retention?’

Contents of the CJEU judgment:

In 2019, a married couple, third-State nationals (Iran), both with regard to Brussels IIA and Dublin III respective Member States, moved from Finland to settle in Sweden. Since 2016, the couple had lived in Finland for around three years. In 2019, a child was born in Sweden. The couple was exercising joint custody over the child in conformity with Swedish law. The mother was holding a family residency permit, in both Finland and Sweden, through the father’s employment rights. The approved duration of the mother’s residency right in Finland was around one year longer than in Sweden.

Two months after the child’s birth, the latter and the mother were placed under Swedish residential care (hostel). Essentially, the Swedish administrative decision to uphold this care protective measure was the result of the father’s violence against the mother, so to protect the child from the risks against his development and health, as well as to prevent his wrongful removal to Iran possibly envisaged by his father. Limited contact rights were granted to the father. A residency permit was requested, individually, by the father and the mother based on the family lien – request respectively filed on 21 November and 4 December 2019.

In August 2020, the mother submitted an asylum request, for the child and herself, before the Swedish authorities. The same month, the Finnish authorities declared themselves internationally responsible over the mother’s and child’s asylum request by virtue of article 12(3) of Dublin III – based on the longer duration of the residency permit previously delivered according to Finnish law. In October 2020, the Swedish authorities dismissed the father’s and rejected the mother’s respective residency and asylum requests, and ordered the transfer of the child and his mother to Finland. Taking into account the father’s presence as a threat against the child, the limited contacts established between them, and the father’s residency right in Finland, the Swedish authorities concluded that the child’s separation from his father was not against his best interests and that the transfer was not an obstacle to the exercise of the father’s visitation right in Finland. In November 2020, the mother and the child moved to Finland pursuant to article 29(1) of Dublin III. In December 2020, the father filed an appeal against the Swedish court’s decisions, which was upheld by the Swedish Immigration Tribunal (‘Migrationsdomstolen i Stockholm’), although it resulted later to be dismissed by the Swedish Immigration Authorities, and then rejected by the Immigration Tribunal, due to the child’s relocation to Finland (CJEU ruling, § 23-24).

In January 2021, the father filed a new request before the Swedish authorities for family residency permit on behalf of the child, which was still ongoing at the time of this judgment (CJEU ruling, § 25). During the same month, the mother deposited an asylum application before the Finnish authorities, which was still ongoing at the time of this judgment – the mother’s and child’s residency permits were withdrawn by the Finnish authorities (CJEU ruling, § 26). In April 2021, the Swedish Court (‘Västmanlands tingsrätt’), notwithstanding the mother’s objection to their jurisdiction, granted divorce, sole custody to the mother and refused visitation right to the father – upheld in appeal (‘Svea hovrätt’). Prior to it, the father filed an application for child return before the Helsinki Court of Appeal (‘Helsingin hovioikeus’), arguing that the mother had wrongfully removed the child to Finland, on the grounds of the 1980 Hague Convention. The return application was rejected. On the father’s appeal, the Finnish authorities stayed proceedings and requested an urgent preliminary ruling from the CJEU, in line with article 107 of the Luxembourg Court’s rules of procedure.

CJEU reasoning:

The Court reiterated that a removal or retention shall be wrongful when a child holds his habitual residence in the requesting State and that a custody right is attributed to, and effectively exercised by, the left-behind parent consistently with the law of that State (§ 45). The primary objectives of the Brussels IIA Regulation, particularly within its common judicial space aimed to ensure mutual recognition of judgments, and the 1980 Hague Convention are strictly related for abduction prevention and immediate obtainment of effective child return orders (§ 46).

The Court stated that, pursuant to articles 2 § 11 and 11 of the Brussels IIA Regulation, the child removal to a Member State other than the child’s habitual residence, essentially performed by virtue of the mother’s right of custody and effective care while executing a transfer decision based on article 29 § 1 of the Dublin III Regulation, should not be contemplated as wrongful (§ 48). In addition, the absence of ‘take charge’ request following the annulment of a transfer decision, namely for the purposes of article 29 § 3 of Dublin III, which was not implemented by the Swedish authorities, would lead the retention not to being regarded as unlawful (§ 50). Consequently, as maintained by the Court, the child’s relocation was just a consequence of his administrative situation in Sweden (§ 51). A conclusion opposing the Court reasoning would be to the detriment of the Dublin III Regulation objectives.

Some insights from national precedents:

In the case ATF 5A_121/2018, involving a similar scenario (cf. FamPra.ch 1/2019), the Swiss Federal Court maintained that a child born in Greece, who had lived for more than a year with his mother in Switzerland, had to be returned to Greece (place of the left-behind parent’s residence) based on the established child’s habitual residence prior to the wrongful removal to Switzerland, notwithstanding his pending asylum application in the latter State. Indeed, the Greek authorities had been internationally responsible over the child’s asylum request on the basis of his father’s residence document. However also in that case it was alleged that the father had been violent against the mother and that a judgment ordering the child’s return to Greece, alone or without his mother (§ 5.3), would not have caused harm to the child under the 1980 Hague Convention, art. 13.

In the case G v. G [2021] UKSC 9, involving a slightly different scenario in that no multiple asylum requests were submitted, the UKSC judged that a child, of eight years old born in South Africa, should not be returned – stay of proceedings – until an asylum decision, based on an asylum application filed in England, had been taken by the UK authorities. The UKSC considered that, although an asylum claim might be tactically submitted to frustrate child return to his/ her country of habitual residence prior to wrongful removal or retention, it is vital that an asylum claim over an applicant child, accompanied or not by his/ her primary carer, is brought forward while awaiting a final decision – in conformity with the ‘non-refoulement’ principle pursuant to article 33 of the 1951 Geneva Convention relating to the Status of Refugees.

Comment:

The CJEU ruling is momentous dictum in that it holds the not any longer uncommon intersection of private international law and vulnerable migration, especially with regard to children in need of international protection in accordance with both Brussels IIA and Dublin III Regulations (cf. Brussels IIA, § 9, and Dublin III, article 2 lit. b). The Luxembourg Court clarifies that a child who is allegedly wrongfully removed, meaning without consent of the other parent, should not return to his/ her habitual residence if such a removal took place as a consequence of the ordered transfer determining international responsibility based on the Dublin III Regulation. It is emphasised that, contrary to the Swiss judgment, the child in the instant case did not have any personal attachments with Finland at the time of the relocation – neither by birth nor by entourage – country of destination for the purposes of the Dublin III transfer. Moreover, the ‘transfer of responsibility’ for the purposes of Dublin III should be contemplated as an administrative decision only, regardless of the child’s habitual residence.

It is observed as a preamble that, according to a well-known CJEU practice, a child should not be regarded as to establish a habitual residence in a Member State in which he or she has never been physically present (CJEU, OL v. PQ, 8 June 2017, C-111/17 PPU; CJEU, UD v. XB, 17 October 2018, C-393/18 PPU). Hence, it appears procedurally just that the Swedish courts retained international jurisdiction over custody, perhaps with the aim of Brussels IIA, article 8 – the child’s habitual residence at the time of the seisin, which occurred prior to the transfer to Finland. On that procedural departure, the Swedish courts custody judgment is substantially fair in that the father’s abuse against the mother is indeed an element that should be retained for parental responsibility, including abduction, merits (CJEU ruling, § 48; UKSC judgment, § 62).

However, it is argued here that, particularly given that at the relevant time Sweden was the child’s place of birth where he lived for around 14 months with his primary carer, the Swedish and the Finnish authorities might have ‘concentrated’ jurisdiction and responsibility in one Member State, namely Sweden, ultimately to avoid further length and costs related to the asylum procedures in line with the same Dublin III objectives evoked by the CJEU – namely “guarantee effective access to the procedures for granting international protection and not to compromise the objective of the rapid processing of applications for international protection” (§ 5, Dublin III). Conversely, provided that the child’s relocation was not wrongful as indicated by the Finnish authorities, and confirmed by the CJEU ruling, the Swedish authorities may have opted for the ‘transfer of jurisdiction’ towards the Finnish authorities on the basis of Brussels IIA, article 15(1) lit. b, indicating the child’s new habitual residence (cf. Advocate General’s opinion, § 41) following the lawful relocation (cf. article 15.3., lit. a).

Importantly, concentration of jurisdiction-responsibility over a child seeking international protection in one Member State, in light of the Brussels IIA-Dublin III interplay, would essentially determine a coordinated interpretation of the child’s best interests (cf. Brussels II, § 12, and Dublin III, § 13), avoiding two parallel administrative-judicial proceedings in two Member States whose authorities may not always come to similar views, as opposed to the present case, over such interests (AG’s opinion, § 48). This is particularly true, if the child (non-)return to his/ her habitual residence might likely be influenced, as stated in the CJEU ruling, by his/ her administrative situation, which would potentially have an impact on the international custody jurisdiction determination. An example of controversial outcome, dealing with child abduction-asylum proceedings, is the profoundly divergent opinion arising from the UK and Swiss respective rulings, to the extent of child return in a situation where the mother, primary carer, is or could be subject to domestic violence in the requesting State.

Similarly, the UKSC guidance, in ‘G v. G’, affirmed: “Due to the time taken by the in-country appeal process this bar is likely to have a devastating impact on 1980 Hague Convention proceedings. I would suggest that this impact should urgently be addressed by consideration being given as to a legislative solution […] However, whilst the court does not determine the request for international protection it does determine the 1980 Hague Convention proceedings so that where issues overlap the court can come to factual conclusions on the overlapping issues so long as the prohibition on determining the claim for international protection is not infringed […] First, as soon as it is appreciated that there are related 1980 Hague Convention proceedings and asylum proceedings it will generally be desirable that the Secretary of State be requested to intervene in the 1980 Hague Convention proceedings” (UKSC judgment, § 152-157). Clearly, the legislative solution on a more efficient coordination of child abduction-asylum proceedings, invoked by the UK courts, may also be raised with the EU [and Swiss] legislator, considering their effects on related custody orders.

– Cross posted at the EAPIL blog.