Rethinking Gender-Based Asylum: A Look at the Advocate General’s Opinion on Women Fleeing the Taliban

 

Türkan Ertuna Lagrand (Assistant Professor, Utrecht University School of Law) and Salvo Nicolosi (Senior Assistant Professor, Utrecht University School of Law)

Photo credit: USAIDAfghanistan, via Wikimedia Commons

Can a woman be recognized as a refugee, merely because of her gender? This has been one of the most debated questions in international refugee law, which has recently reached the Court of Justice with two joined cases in AH and FN on which last 9 November 2023, Advocate General Richard de la Tour delivered his Opinion. These joined cases relate to a request for a preliminary ruling from the Austrian High Administrative Court. The referring judge was in doubt whether, in light of Article 9 (1) (a) and (b) of the Qualification Directive, it is sufficient that a woman who is affected, merely on the basis of her gender, by the accumulation of government-imposed or supported restrictive measures can be recognised as a refugee without the need to assess the woman’s individual situation.

While awaiting the ruling from the Court of Justice, it is worth underscoring the significance of the Advocate General’s Opinion in light of three complementary dimensions, each representing crucial elements within the Advocate General’s assessment. These dimensions refer to the nature of persecution arising from discriminatory measures; the issue of whether women can be recognized as constituting a distinct social group; the need to conduct an individual assessment as the foundational basis for granting refugee status.

By addressing these three dimensions, this short post aims to flag the potential for the Court of Justice to expand the scope of protection in Europe and contribute to the progressive development of international refugee law.

Systematic discrimination against women

The reference from the Austrian Court is rooted in the rise of the Taliban regime in Afghanistan in 2021, marked by the implementation of a set of discriminatory measures specifically aimed at women in the country. As the Austrian High Administrative Court as well as experts of the United Nations have underscored, the Taliban has put in place measures severely restricting civil and political rights. Such measures consist of preventing women from travelling without a male companion, obligating them to cover their bodies; denying participation in political office and political decision-making processes; denying women access to legal means to obtain protection from gender-based and domestic violence; and lack of protection against forced marriages. Additionally, these measures have significantly curtailed social, economic and cultural rights, such as women’s right to engage in gainful employment, and women’s access to health care, education and sports. In this regard, it is worth stressing that Afghanistan is the only country in the world where girls and young women are forbidden from attending secondary school and higher education institutions.

Admittedly, the accumulation of these measures has led to a situation that the Advocate General considered of 'severe, systematic and institutionalised discrimination' against women.

The concept of ‘discrimination’ is  of paramount importance to determine the existence of persecution for the purposes of seeking recognition as a refugee. However, not all discrimination amounts to persecution. On this point,  Advocate General’s Opinion is especially illustrative as it, in light of the UNHCR Handbook, clarifies that ‘a measure of discrimination will only amount to persecution if it leads to consequences of a substantially prejudicial nature for the person concerned, such as… access to available educational facilities’.

The Advocate General concluded that while some measures individually amount to persecution under Article 9 (1) (a) of the Qualifications Directive (which states that to be considered ‘persecution’, an act must be ‘be sufficiently serious by its nature or repetition as to constitute a severe violation of basic human rights, in particular the rights’ which States cannot derogate from under Article 15(2) ECHR), others cumulatively meet the threshold under Article 9 (1) (b), due to systematic violations of human rights which are ‘sufficiently severe as to affect an individual' in a similar manner as mentioned in Article 9 (1) (a).

 

Women as a particular social group

Interestingly, these joined cases gave the Advocate General de la Tour an opportunity to reiterate and finetune a line of argumentation that he followed earlier this year in the case of WS. On that occasion, the Advocate General argued that women can constitute a particular social group ‘solely on account of their condition as women.’ They in fact share an innate and immutable characteristic, because of which they are seen differently by society, according to their country of origin, by reason of the social, legal or religious norms of that country or the customs of the community to which they belong (para 72). In the same opinion, Advocate General de la Tour convincingly rejected the tendency,  often followed by senior courts, according to which to use women qua women cannot constitute a relevant social group owing to the size of the group. On the contrary, de la Tour concluded that ‘the concept of “distinct identity” of a group, in that it is perceived differently by the surrounding society, cannot be interpreted as entailing a quantitative assessment.’ From this perspective, the Advocate General echoed the position of the UNHCR  reminding that other grounds are not bound by the question of size and upheld the scholarly view that have unearthed the fallacy of such an approach.

Challenging the individual assessment as the foundational basis for refugee status

Borrowing Hathaway and Foster’s words ‘it is now widely understood that where a woman has a well-founded fear of being persecuted for reasons of her gender,… refugee status ought to be recognized.’ Despite the equivocal terms of the Qualification Directive (Article 10), recognizing only that ‘[g]ender related aspects, including gender identity, shall be given due consideration for the purposes of determining membership of a particular social group or identifying a characteristic of such a group,’ there is nonetheless widespread recognition in Europe that women constitute a social group for Convention purposes. As the situation of women and girls in Afghanistan has deteriorated rapidly, Member State authorities went even further to recognize women as refugees solely on the basis of their gender without assessing on a case-by-case basis whether there is an individual a risk of persecution.

While Austria was the first Member State to officially explore the possibility of accepting Afghan women as refugees without an individual assessment, a number of other States have recognized women from Afghanistan as refugees without further examination as to the individual situation. Sweden announced in December 2022 that any Afghan woman asking for refugee status will be granted this status. After firstly declaring to  continue with an individual examination of female asylum seekers from Afghanistan, Denmark later decided to hold a ‘relaxed assessment of evidence’ and as of 30 January 2023 all women and girls from Afghanistan have been granted asylum solely because of their gender. Similarly, the Finnish Immigration Service has announced in early 2023 that ‘all Afghan women and girls are granted refugee status.’

Such a domestic practice offered the Advocate General the opportunity to explain that this practice falls within the margin of appreciation that is left to the Member States by Article 3 of the Qualification Directive to introduce or retain more favourable standards for determining who qualifies as a refugee, in so far as those standards are compatible with the Directive. In this regard, the Advocate General also referred to the ruling in LW e, in relation to which the Court- decided that such standards may consist, inter alia, in relaxing the conditions for granting refugee status and should not prejudice to the general scheme and objectives of that Directive (paras 39-40). Indeed, concurring with the Advocate General, asylum applications submitted by women and girls from Afghanistan have specific characteristics that would allow the competent authorities to deviate from the individual assessment method, in principle required by Article 4 (3) of the Qualification Directive. The discriminatory measures to which Afghan women and girls are exposed are part of a regime of segregation and oppression imposed solely on account of the women’s presence on the territory, regardless of their identity or personal circumstances. Such a circumstance makes unnecessary to establish that the applicant is targeted because of distinctive characteristics other than her gender.

Interestingly, despite these readily apparent, objective circumstances in the country of origin, the Advocate General did not approach the cases on the basis of prima facie recognition of refugee status, which as confirmed by the UNCHR, constitutes an exception to the principle of individual assessment. As argued by Zieck, prima facie recognition is in essence ‘a collective form of status determination that presumes that each individual member of a particular group qualifies for refugee status based on objective information on the circumstances causing flight.’ While particularly suited to situation of large scale arrivals of refugees, prima facie recognition may also be appropriate in relation to groups of similarly situated individuals whose arrival is not on a large scale. However, as the legal foundations and contours of this practice remains still opaque, the Advocate General might have chosen to confine his argumentation within the specific features of EU asylum law, notably Article 4 of the Qualification Directive. This was a reasonable but also particularly relevant choice as it clarifies to what extent EU asylum law allows a departure from an individual assessment, thereby offering a valuable standpoint to the Court to proceed in the same direction.

Concluding remarks

To conclude, the Joined Cases in AH and FN bear significant relevance, particularly in the near term for Member States that have already adjusted their asylum policies concerning Afghan women, and, prospectively, for those Member States poised to emulate such modifications, following an expected favourable ruling by the Court. If the Court were to adopt the analytical framework proposed by Advocate General de la Tour, this would, therefore, contribute to aligning the EU's stance with that of the UNHCR, the Human Rights Council, thereby substantially contributing to the advancement of international refugee law.

 

Amazon v DSA: insights from interim proceedings

 

 

 

Laureline Lemoine, Senior Associate, AWO Agency

 

Photo credit: KarleHorn, via Wikimedia Commons

 

On September 27, 2023, the President of the General Court, Marc van der Woude, issued an order in the case of T‑367/23 R Amazon Services Europe v Commission, one of the first legal challenges against the Digital Services Act (DSA). The order sheds light on Amazon's stance and the arguments that can be anticipated in the main proceedings, and has potential implications for other providers of very large online platforms (VLOP). 

 

Amazon was designated as a Very Large Online Platform (VLOP) by the European Commission on 25 April 2023. This designation meant that Amazon, along with other VLOPs, had until 25 August 2023 to comply with DSA obligations, while other entities (platforms, intermediaries) have until 17 February 2024. In response, Amazon initiated legal action to annul the Commission's VLOP designation, asserting that it infringed upon principles of equal treatment and constituted a disproportionate restriction on its rights. However, as EU acts are presumed lawful, Amazon was required to comply with DSA obligations throughout the proceedings, leading them to request interim measures to suspend specific obligations stemming from their VLOP designation.

 

Examining Amazon's arguments 

 

One of Amazon's primary concerns was Article 38 of the DSA, which mandates an opt-out for recommender systems based on user profiling. Amazon argued that this requirement would adversely impact their business, customers, and third-party sellers. However, President Marc van der Woude noted that Article 38 does not prohibit the use of such systems but merely offers users an opt-out option, and suggested that Amazon could inform customers effectively about the benefits and risks of such an option. The President suggested that platforms could employ “precise and effective measures” to inform customers of the “benefits of the recommender systems and the risks that will ensue from opting out”. In this context, Amazon could only claim financial harm, which, based on the evidence, was deemed insufficient to imperil its financial viability before the final judgement. Consequently, no interim measures were granted concerning Article 38 of the DSA.

 

The suggestion from the Court could lead to platforms employing pop-ups and persuasive language to prompt users to continue using profiling-based recommender systems, similar to how they request users to approve personalised advertising via cookie banners. 

 

Amazon's objections were more substantial concerning Article 39, where it argued that the obligation to publish an advertisement repository would expose confidential information, causing harm to their advertising activities and partners and leading to the loss of consumers. 

For the purpose of interim proceedings, the President had to assume the confidentiality of the information and therefore agreed that the repository revealed sensitive information, which could potentially be exploited by competitors. The Commission tried to counter-argue that the main novelty of the DSA was to consolidate information, explaining that Amazon was already obligated to disclose most of the required information under existing EU legal acts. However, the President highlighted that certain elements of the DSA, particularly pertaining to the duration of advertisements (Article 39(2)(d) DSA) and the total of recipients reached (Article 39(2)(g)), appeared to be genuinely new and not covered by previous legislation.

 

One of the key issues of the main proceedings will therefore centre around the question whether the information Amazon is supposed to publish under Article 39 is genuinely confidential. To prevent jeopardising the ongoing main proceedings, the General Court President chose to suspend Amazon’s obligation to make the advertisement repository public, but Amazon is still required to create and compile the repository pending the outcome of the main proceedings. 

 

Implications for VLOPs

 

The outcome of this interim proceeding has broader implications, especially for other VLOPs. 

The absence of a clear case and the failure to demonstrate real harm in these interim proceedings regarding Article 38 could deter other VLOPs from pursuing similar challenges. Since Article 38 offers an opt-out mechanism closely aligning with GDPR principles, it also makes it less likely to be contested.

 

Conversely, other VLOPs could argue that Article 39 directly impacts their core business models as well. The fact that Amazon obtained a suspension may inspire them to explore similar avenues, given their shared concerns and arguments, which could increase their chances of securing a similar exemption.

Conclusion

 

The interim proceedings in the case of T‑367/23 R Amazon Services Europe v Commission provides interesting insights into how Amazon is navigating their obligations under the DSA. The legal processes and arguments emerging from this case offer a glimpse of what is to come and as this case unfolds, it will continue to be a focal point in the broader discussion surrounding DSA enforcement and its implementation.

Europol’s Joint and Several Liability Regime: Revolutionizing EU Fundamental Rights Responsibility?

 

Dr Joyce de Coninck, University of Ghent

Photo credit: Oseveno 

Introduction

 

The Europol Regulation introduces a system of joint and several EU liability for unlawful data processing in violation of Article 7 and 8 of the Charter of Fundamental Rights. This nascent EU liability regime features at the heart of the dispute in the Marián Kočner v Europol saga, and much like the recent WS and others v Frontex case before the General Court, highlights the urgency for clarification on joint responsibility for human rights violations as a result of shared conduct between the EU’s operational agencies and the EU Member States.

 

One of the drivers prompting this need for clarification, relates to the increased cooperation between the EU’s operational agencies on the one hand, with EU Member States on the other hand, in achieving common objectives. While Frontex is increasingly endowed with (executive) powers in the EU’s Integrated Border Management (see here, here and here), Europol is endowed with increased powers regarding the processing of large datasets, the screening of foreign direct investment in security-related cases and the acquisition of data from private companies in dealing with terrorist or child abuse material. These enhanced powers result in a multiplicity of public and private actors working together in achieving common goals, where previously such tasks fell within the exclusive purview of the Member States.  

 

The ‘crowding of the operational field’, referred to by Gkliati and McAdam as the ‘many hands’ problem, reveals a significant disconnect between the EU’s contemporary liability regime on the one hand, and the application of this liability regime in practice to situations of joint conduct that give rise to human rights harms on the other hand. In other words, the EU’s liability regime was not legally designed to accommodate questions of joint responsibility for human rights harms flowing from concerted conduct by the EU institutions, bodies, offices and agencies and the EU Member States. The incompatibility – or rather, unsuitability – of the EU’s human rights regime in dealing with joint conduct, features on two distinct levels, and on both levels, a driving force behind the unsuitability is one of legal design.

 

On the one hand, historical accounts of the constitutionalization of fundamental rights in the EU, giving rise to the Charter of Fundamental Rights in particular, explain that this process was by and large the result of constitutional concerns over EU fundamental rights protection by domestic courts. In other words, this exercise of constitutionalization came about in reaction to constitutional objections by Member States regarding the level of protection of fundamental rights provided under the EU’s chapeau. An unintended consequence of this development appears to be that the drafters of the Charter did not necessarily consider joint and inseparable operational conduct by EU entities and the EU Member States. In turn, and as predicted by Weiler, it did not bring the added clarity to how the state-centric Charter rights – many of which were inspired by and textually almost identical to state-centric international human rights treaties – would translate into enforceable negative and positive human rights obligations that give flesh to the bones of these human rights commitments. In other words, the mere fact that EU entities are bound by fundamental rights in the Charter, does not relay much on how the EU must conduct itself in order to comply with these rights, as I have discussed at length elsewhere (here, here and here).

 

On the other hand, the EU’s liability regime also was not legally designed to respond to questions of responsibility-allocation flowing from unlawful joint conduct giving rise to human rights harms. This is textually and historically supported, as the EU’s action for damages falls within the exclusive purview of the CJEU (Article 268 in juncto 340 TFEU) and case law has set out rules proclaiming that national courts shall be seized where damages are the result of the incorrect or correct implementation by Member States of EU legislative acts (for a general discussion, see here). In other words, the EU’s action for damages was not developed to consider joint non-contractual responsibility and the conditions for liability subsequently developed through the CJEU’s case law were also not developed with such liability in mind.

 

However, the increased reliance on inseparable and operational cooperation between EU entities and its Members giving rise to fundamental rights harms, brings to the fore a new dimension of liability that was not foreseen in either the normative human rights developments giving rise to the Charter, nor the liability regime that currently exists within the EU’s framework. Yet it is precisely this question of joint liability that sits at the heart of the case of Marián Kočner v Europol currently pending before the CJEU and the accompanying opinion by Advocate General Rantos as developed and discussed in what follows.

 

The Case

 

In 2018 Marián Kočner was being investigated by the Slovak criminal authorities within the context of a murder investigation. The investigation resulted in the domestic authorities taking possessing of two mobile phones and a USB drive belonging to the Applicant, which were subsequently handed over to Europol at the request of the domestic authorities in October 2018. Several months later, Europol returned the mobile phones and the USB-drive along with relevant scientific reports concerning its contents, as well as a hard-drive with encrypted data derived from the mobile phones to the Slovak authorities. The contents of the mobile phones and USB drive – transcripts of intimate conversations involving the applicant and his girlfriend, as well as the inclusion of his name on the ‘mafia lists’ – were subsequently leaked in large quantities and made public by the press. On the basis of these leaks the Applicant claimed compensation from Europol for non-material damage stemming from unlawful data processing, underscoring that the leaks by the press violated his right to a private and family life as protected under Article 7 CFR.

 

In the subsequent action for damages on the basis of Article 268 and Article 340 TFEU, the General Court dismissed the Applicant’s claims (Kočner v Europol T-528/20) holding that no causal link could be established between Europol’s conduct and the purported damages stemming from the data made public from the mobile phones, and that the Applicant had not provided any evidence demonstrating that the ‘mafia lists’ had been drawn up by Europol.

 

In his appeal, the Applicant asks the Court of Justice to set aside the General Court’s ruling on the basis of six points of law. For the purpose of the current contribution however, the focus will be on the argument raised by the Applicant concerning the nature of the EU’s liability. Specifically, the Applicant argues that the General Court erred in law for having disregarded Europol’s liability in light of recital 57 of the Europol Regulation related to joint and several liability. In other words, this claim by the Applicant juxtaposes the concept of ‘joint and several liability’ with the notion of joint responsibility more generally, contending that the implications of these different approaches to responsibility may have yielded a different outcome in the case. According to the Applicant, the fact that the General Court did not consider Europol’s liability through the standard of ‘joint and several liability’ constitutes an error depriving recital 57 of the Europol Regulation of any significance.

 

The arguments advanced by the Applicant provide the Court of Justice with the first-ever opportunity to rule on the scope and implications of the concept of joint and several liability of Europol, which – given the marginal case law on joint responsibility for human rights harms more generally – could prove very instructive in clarifying the conditions of joint responsibility and the manner in which such responsibility should be allocated between the EU and the Member States.

 

The Opinion

 

After dismissing an admissibility objection by Europol, Advocate General Rantos identifies six grounds of appeal, of which four relate to the question of whether unlawful data processing occurred by Europol. The remaining two points of appeal concern the nature of Europol’s liability and the concept of ‘joint and several liability’ specifically.

 

The question of the nature of Europol’s responsibility essentially revolves around recital 57 and Article 50 of the Europol Regulation. As aforementioned, recital 57 introduces the concept of joint and several liability where it may “…be unclear for the individual concerned whether damage suffered as a result of unlawful data processing is a consequence of action by Europol or by a Member State”. This provision covers only liability issues relating to unlawful data processing and only insofar it is unclear to which party the (unlawful) data processing should be attributed, whereas the preceding recital 56 recalls that for all other questions of non-contractual liability, the EU’s general liability rules – as articulated in the CJEU’s Bergaderm ruling – apply.

 

Chapter 7 of the Europol Regulation covers remedies and liability and Article 50 specifically, addresses liability stemming from unlawful data processing. This provision holds in its first paragraph that anyone having suffered damage from unlawful data processing will be entitled to receive compensation from either Europol in line with the general liability rules of article 340 TFEU, or from the Member State in which the unlawful data processing occurred in accordance with its domestic law. The second paragraph (Article 50(2)) holds that where a dispute arises concerning the ultimate responsibility for compensation, the Management Board of Europol shall decide by a two-thirds majority who bears the burden of ultimate responsibility for compensation. Grosso modo the relevant recitals appear to refer to modalities of responsibility allocation between Europol and the implicated Member States, whereas Article 50 is concerned with the ensuing obligation of compensation insofar responsibility has effectively been established.  

 

AG Rantos begins his opinion on the nature of the EU’s liability by pointing out that while the relevant recitals do introduce a solidarity-based responsibility mechanism, this is not mentioned explicitly in its operative counterpart. In fact, the absence of any explicit reference to joint and several liability in Article 50 led the General Court to the conclusion that liability in accordance with the general rules on liability embedded in Article 340 TFEU, could not be causally established.

 

After recalling the conditions to establish EU liability generally (para 34 – 35), AG Rantos addresses the question of the nature of Europol’s liability in a threefold manner, recalling that a provision of EU law must be interpreted mindful of its wording (1), the context in which it was drafted (2), and its objective and purpose (3), which may be inferred from its legislative history and through comparative interpretation.

 

Contrary to Europol, AG Rantos concedes that the wording of the relevant recitals (which appear to introduce new modalities of joint responsibility under EU law), and the wording of the Article 50 (which neglects any reference to joint and several liability and refers only to compensation) is not unambiguous. To this end, he underscores that the reference to joint and several liability in recital 57 suggests concurrent liability for Europol and the Member States, whereas Article 50 literally suggests responsibility for compensation as being a responsibility of either the Member State or Europol. Similarly, the generic reference to non-contractual EU liability in Article 340 TFEU, which is to be considered in line with the general principles common in the laws of the Member States, leaves room for interpretation.

 

As concerns the context of the contested provisions, the AG notes that while recitals have no legally binding force as such, they nevertheless function as an indicator of the intent of the legislator. In casu, the intent of the legislator was to favor the aggrieved parties and eliminate any questions of attribution. The AG concludes that this is not in conflict with Article 50, following which the latter must be interpreted in light of recital 57 and the concept of joint and several liability.

 

Finally, the objectives of recital 57 of the Europol Regulation may be discerned through its legislative history and a comparative interpretation of its meaning in light of general principles common to the Member States. Here, the AG recalls that the concept of ‘joint and several liability’ had been introduced in the very first Commission proposal and had been included among others to limit the difficulties encountered by aggrieved parties in attributing unlawful processing to either the Member States or the EU. Furthermore, a comparative analysis of this concept reveals that Member States make use of this mode of liability in cases where attribution of unlawful conduct may be hard to establish. The Advocate General concludes that suspending the procedure before EU courts while the concomitant domestic procedure against the Member State is pending – as typically occurs for questions of joint responsibility – would deprive Article 50 interpreted through recital 57 of any significance. It flows from this that concurrent proceedings would thus be possible. 

Analysis

 

The case deals with a situation of ‘many hands’ cooperation involving a Member State which gives rise to a question of unlawful data processing, arguably falling within the ambit of Article 7 (respect for private and family life) and 8 (protection of personal data) of the Charter. Flowing from this, the Applicant argues that Europol should be held responsible under the rules of joint and several liability, whereas Europol contends that this should be assessed under the standard rules of joint responsibility which are derived from the Bergaderm ruling. In essence, this is a question of whether the lex generalis applies or instead, whether a lex specialis applies. As aforementioned, the Advocate General recommends that the case be re-examined by the General Court, in light of the (underdeveloped) rules on joint and several liability, whereby he concurs with the Applicant that it is unclear to which party the conduct should be attributed.

 

The Francovich and Brasserie du Pêcheur judgments, spell out the conditions for Member State liability under EU law, whereas the Bergaderm judgment spells out the conditions for non-contractual responsibility of the EU institutions. These conditions require that for responsibility to arise, there must be a (sufficiently serious) breach of EU law, that causally gives rise to damage. In certain cases, the CJEU will also demand that the conduct must be attributable to the EU actor under scrutiny.

 

These rules apply to responsibility and joint responsibility between the EU and its Member States generally, but importantly do not prejudice more tailored, specific or alternative rules on (joint) liability. An alternative, bifurcated approach to liability exists in the realm of EU data processing. On the one hand, there are the data-processing specific rules for Member State liability embedded in the GDPR. On the other hand, there are specific liability rules for data processing applicable to EU institutions, bodies, offices and agencies as embedded in the Data Protection Law Enforcement Directive, as well as the Data Processing by the EU Institutions and Bodies Regulation. These data processing-specific rules apply, unless there are more specific rules that have been developed, which is the case for processing of operational data by Europol (Article 2(3) Data Processing by the EU Institutions and Bodies Regulation). In other words, more specific rules have been developed for situations involving processing of data for Europol. Accordingly, when it is clear to which actor (the Member State or Europol) unlawful data processing should be attributed, the regular rules on liability apply, in accordance with the domestic regime for Member State liability and in accordance with the action for damages concerning Europol’s liability (Article 50(1) Europol Regulation). However, when attribution is not clear, joint and several liability applies (recital 57 in juncto Article 50(2) Europol Regulation), leaving it to the Management Board to decide in case of conflict who bears the ultimate responsibility to provide compensation for the inflicted harm (Article 50(2) Europol Regulation).

 

Juxtaposing Joint Liability and Joint and Several Liability

 

This approach appears to give rise to procedural efficiency from the perspective of the Applicant and appears to relax the Bergaderm conditions for EU responsibility to arise. 

 

Choosing the Judicial Forum

 

The objective of the joint and several liability mechanism is to ensure that the Applicant’s rights are safeguarded. This means that unlike the system of joint EU-Member State responsibility, the domestic court will not necessarily be the primary forum to establish responsibility and the ensuing burden of reparations. Instead, the aggrieved individual could go through either the domestic legal system or the EU’s action for damages to have responsibility established. Upon conclusion of the legal procedures and once the Applicant has been awarded damages, these actors could subsequently settle any dispute on the duty to provide reparations in a subsequent procedure within the Management Board of Europol, the decision of which could also be subject to legal scrutiny under the annulment procedure. Under this mechanism, the Applicant enjoys a much lesser of a burden in choosing the appropriate judicial venue and is not constrained by which actor will be able to provide reparations. Instead, reparations (in case of responsibility) will be the default from the perspective of the Applicant.

 

Attribution and Causation Revisited

 

The system of joint and several liability suggests that as soon as a situation implicates both Europol and a Member State, and the questionable conduct cannot be definitely attributed to either entity, the requirement of attribution becomes obsolete, as the conduct will be considered attributable to both in full. Interestingly, by relaxing the requirement to establish attribution, the condition of causation will arguably also be relaxed. It is important to recall that while attribution links a particular line of conduct to an actor, causality links that actor to the damage. Relaxing the rules of attribution under the joint and several liability regime and doing away with the requirement to definitively attribute conduct to one or the other, ipso facto entails that the requirement of causality as it currently is being applied, can never be met. Causation under general EU liability law demands that there is an uninterrupted relationship between the unlawful conduct by a certain actor, giving rise to damage. Yet, in the absence of an obligation to attribute to either the Member State or the EU, the unlawful data processing will be considered attributable to both. If the unlawful conduct is considered attributable to both, it is then unclear how this impacts the causality requirement, which demands that the chain of causation linking the damage to the unlawful conduct by a particular actor, be uninterrupted by intervening acts.

 

Lingering Questions for the EU Courts

 

In light of the limited case law on EU (joint) responsibility generally, a number of questions remain unaddressed including by Advocate General Rantos either.

 

Attribution

 

A first small but pervasive question that demands further clarification concerns when Article 50 read in light of recital 57 of the Europol Regulation is triggered. The presumption appears to be that it is straightforward to distinguish between scenarios in which attribution can be definitively established, and situations in which it is unclear to which entity the unlawful data processing should be attributed. Yet, to date no clear standard of attribution can be definitely discerned under the general system of EU liability. In fact, practice by the EU institutions internally, in international relations, and across different EU policy fields, suggests that the rule of attribution differs significantly in a rather haphazard manner. This is complicated by the absence of a common legal forum to settle responsibility questions implicating the EU and Member States in unlawful data processing. The applied attribution rules under domestic regimes may very well differ from attribution rules under the EU’s liability regime for example, and to date, it is not clear which attribution rules should prevail, much less how this impacts whether Europol’s joint and several liability mechanism is triggered. Arguably, the absence of a coherent and clarified approach to attribution under EU law means that it will be easier for Applicants to trigger joint and several liability under the Europol Regulation. However, this remains to be seen, and is as always, dependent on the applicable burden, standard and method of proof required to show that it’s unclear to which actor the unlawful data processing should be attributed.

 

Joint and Several Liability Beyond Data Processing

 

The question of human rights liability for violations occurring at the hands of operational EU agencies has gained much traction in recent years. The current pending actions for damages against Frontex prompt the question whether a – CJEU clarified – system of joint and several liability may be a way forward. Anyone who has attended a conference or workshop involving Frontex representatives, has undoubtedly been confronted with the scripted answer to questions of human rights responsibility: ‘Frontex is not responsible for such actions – Frontex merely coordinates Member State actions’. Leaving aside the veracity of this response, it is undisputed the current regime of liability allocation has resulted in much blame shifting at the expense of individual rights. Conversely, the system of joint and several liability introduced by the Europol Regulation may very well be a way to circumvent this type of blame-shifting, safeguard the rights of the individual while ensuring that the burden of reparation is not circumvented by one at the expense of the other. A well-developed system of joint and several liability could thus fulfill both a remedial function – namely to protect the Applicants’ fundamental rights, as well as a deterrence function. By increasing the likelihood of legal responsibility through more relaxed rules on attribution and causation, EU institutions, bodies, offices and agencies may be disincentivized to resort to ‘many hands’ to circumvent responsibility claims in implementing their policies, or at least be incentivized to clarify their own rules on (human rights) responsibility allocation. Of course, I write this knowing full well that it is precisely these institutions that prefer to continue operating in the ‘many hands’ murkiness and that clarified rules on responsibility will receive political push-back and may disincentivize operational agencies from providing support in tackling transnational issues. Yet, once every so often, a unicorn-like development surfaces in the field of EU human rights responsibility, as evidenced by the joint and several liability mechanism in this case. Who knows – maybe this same unicorn will resurface in the EU’s responsibility acquis more generally? In any event, I await the CJEU’s perspective on this matter eagerly.  

Analyzing the legality of the EU Commission’s proposed withdrawal of preferential tariffs for third countries when cooperation on migrant returns ‘fails’--an unholy alliance of trade and migration?

 

 

Marion Panizzon, Senior Research Fellow, World Trade Institute, University of Bern*

 

*Privat-Docent, Dr. iur., LL.M., Senior Research Fellow, World Trade Institute, University of Bern and Legal Consultant, World Trade Advisors, Ltd. Geneva. I thank Dr. Alan Desmond, Leicester University for his thoughtful comments on several earlier drafts. I’m grateful to Dr. Christian Häberli, World Trade Institute, for his critical read of an earlier draft in light of current WTO law and practice.

 

Photo credit: NOAA, via Wikimedia commons

 

 

As a strategy to rebalance uneven negotiating positions, the ‘comprehensive approach’ defined in paragraph 11 of the Global Compact for Safe, Regular and Orderly Migration (GCM) sets the stage for rendering more attractive EU trade and EU external migration policy to third countries. However, the comprehensive approach, considered a component of ‘shared responsibility’ under the GCM is often confounded with conditionality, because both might, according to Hocquét 2023, expand the radius of migration policies, to areas outside its immediate realm. There are marked differences though, since conditionality establishes a  co-dependency between measures the EU wishes to implement, with areas of interest to the third country, being education, energy, natural resources, climate adjustment, public health (Peers 2016). Conversely, the comprehensive approach while based on mutuality, rather than reciprocity, ideally strives to create the win-win-win situations, in most cases, breaks down to integrating safe pathways with border management (Vitiello 2022)

 

In trade, ‘rights-based conditionality’ for obtaining trade preferences, has been criticized by academics (Hafner-Burton et al. 2019) and UNCTAD alike (Irish 2007), and materializes when preferential import duties are leveraged for securing the developing or least-developed country’s cooperation to fulfill an EU public good, including combatting narcotics trading, child labor and worker exploitation,  as well as the smuggling and trafficking. At the outset, the trade and development chapter inserted in the 1960s into the General Agreement on Tariffs and Trade (GATT) had empowered developing and least-developed countries to catch-up. One such avenue came in the shape of GSP developed under the auspices of the UNCTAD in 1968 and anchored by several temporary waivers into GATT, to protect infant industries thru non-reciprocal treatment (Michalopoulos 2020). Today, the expectation on trade preferences is that they fulfill non-trade policy objectives (NTPO), which is an attribution that distorts the original idea behind the GATT Art. XXIV and the subsequent Enabling Clause, permanently waiving the most-favored nation treatment otherwise due if a WTO Member lowers a tariff(ECDPM 2020).

 

Initiated in 2021, the EU Commission’s reform of Regulation EU 978/2012 Generalized System of Preferences (GSP) for developing and least-developed countries, proposes to expand by the area of migrant readmission, the cooperation the EU requires from beneficiary countries for exports from those countries to benefit from a lower or zero import duty on two-thirds of tariff lines under standard GSP, a zero duty on the same tariff lines, conditioned on the ratification of 27 conventions (GSP+) or a zero import duty on all products from LDCs except for arms and ammunition (Everything-but-Arms, EBA)  into the EU. Adding to the EU’s long list of incentives to buy origin countries’ approval for sending back their citizens in irregular stays in the EU, the Commission’s proposal, critically viewed by the EU Parliament, NGOs, and academics alike, would have complemented that listing by adding preferential trade initiatives to it. Inversely, the GSP Regulation, equally boasting an ever increasing to-do list of criteria countries need to fulfill in order to enjoy the preferences, has now been topped by the criteria of readmitting (irregular) migrants.

 

The legality of both the EU external migration system with the new addition of trade and the EU GSP regulation with its expansion to include migration policy, poses challenges of legality and practicability under WTO law, as academics and practitioners have analysed and this post discusses.

 

In its reform proposal of EU GSP Regulation 978/2012, the EU Commission suggest for the very first time, to interlace EU external migration policy, notably EU readmission agreements and cooperation on assisted and voluntary return to the EU’s GSP for the period of 2024-34. In particular the proposal foresees to up the ante of EU Regulation 978/2012 withdrawal of tariff preferences procedure by adding migrant readmissions as one benchmark to measure good governance, the former which is, alongside sustainable development and human rights one of the areas of cooperation which can conditionality rewards either positively or which is sanctioned off negatively, by the withdrawal or suspension of preference, ranging from visa relaxation, development cooperation or lowered import tariffs into the EU. As Grundler and Guild 2023 have observed, the negative conditionality (Sabourin and Jones 2023) which the Commission would like to see, is insofar not surprising, as EU member states have traditionally taken to constructs double binds, pitting legal pathways against migration control (Garcia-Andrade 2020:260) in bilateral migration agreements, with questionable outcome.

.

Whereas the EU has inserted Art. 25(a), a readmission clause in its 2019 revision of the  Schengen Visa Code (Bisong 2019; Grundler and Guild 2023), the Commission’s activating the Union’s common commercial policy Art. 207 TFEU, to secure cooperation on readmissions, uses a new tool for the same purpose, yet without specifying the periodicity of review of third countries’ cooperation or lack thereof, as Art. 25a does. Several human rights organizations have remarked in response to the Commission’s proposal and the Council’s endorsement in 2022, that the insertion of migration, is shaky on more than one legal ground (Human Rights Watch, FIDH, ECRE).

 

Firstly, withdrawing the preferential tariff treatment, which LDC and developing countries obtain under one of the three pillars of the GSP, GSP+ and Everything-but-Arms (EBA) treatment might be unlawful under the WTO Appellate Body jurisprudence, as shown by DG External Policies’ Report on the proposed GSP reform (2022). In that WTO decision of 2004, the benefit of lower tariffs must be granted on a non-discriminatory basis to ‘similarly situated beneficiary countries’ and a clear link between the benefit granted and the ‘development objective’ be made. Hence, the idea of threatening a developing or least developed country with the withdrawal of a tariff preference, is not new, as Peers (2016) points out (534-537), but the EU Parliament in its criticism of the EU Commission’s proposed new EU GSP Regulation  had legitimate concern, that the EU would be creating the similar situation that had led the WTO Appellate Body ruling in EC-Tariff Preferences, to condemn it and which had put a stop to any selective imposition of trade preferences (Bartels 2003).

 

Up until this day, the Commission rewards countries of origin for cooperating on forced returns, border screening, information campaign, via visa relaxations for diplomatic staff, journalists, by facilitating remittances transfers and lowering costs, by a promise of better integration of third country nationals (Peers 2016). In so doing, the EU has treated different migrant origin countries, differently, yet, so far, without risking WTO incompatibility, since if visa, remittances or integration remain outside the scope of the WTO. The situation is different, given the recently suggested negative trade conditionality, contingent on withdrawing tariff preferences, when a readmission cannot take place. It implies that the Commission must define ‘objective’ benchmarks around readmission, for when that tariff treatment is to be removed in addition to treating similarily situated countries, identically.

 

Secondly, the EP during the 2022 inter-institutional trilogues between Council, Commission, ascertained that tying preferences to readmission of one’s own nationals, might be too far removed a conditionality. Recently, the EP’s international trade (INTA) committee on 19 September 2023 voted in favour of extending the current EU GSP scheme to 31 December 2027, which the Commission proposed to avoid the current GSP lapsing in light of the lack of agreement on the main proposal, until the Parliament and Council can agree on how to update that Regulation. Meanwhile, INTA has not further entered into discussions about circumscribing the exact legal scope triggering a potential tariff withdrawal, whether a non-implementation of an EU readmission agreement or of a bilateral readmission agreement must be shown, or whether the refusal to negotiate such an agreement in the first place is sufficient to trigger the clause, or, as the DG External Relations suggested, the non-compliance with international obligations under a EURA. Nonetheless, the Commission’s proposal currently stands at the brink of extinction. Yet, it seems timely to analyse its legality with WTO and international obligations, as a final vote, after EP elections, could overturn the INTA’s refusal to follow the Commission’s proposal.

 

In this blogpost, I discuss legality of the EU’s reform proposal under two WTO instruments, the Enabling Clause 1971, setting the legal basis for the Special and Differential Treatment of trade in goods from developing and least-developed WTO Members and the LDC Services Waiver 2011, to understand which out of the two takes origin country concerns seriously. In so doing, I draw on the discussion started by Vidigal (2023) and Tans (2023) about why the  Commission proposal conflicts with WTO rules. In so doing, I touch upon the number of preferential trade agreements (PTAs) which, similarily, have used a trade component as the quid-pro-quo for obtaining a partner’s cooperation on irregular migration. Since PTAs pit trade in services, and thus a form of legal pathways (as opposed to trade in goods) to return migration, within their chapters on the temporary movement of natural persons, the conditionality is more closely or directly contingent on people-on-the-move.

 

Consequently, the EU Commission were better advised to negotiate such openings of its services markets for service supplying natural persons, from countries of origin, within one of its deep and comprehensive free trade agreements (DCFTA) than to go freestyle by attempting to match migrant readmission with trade-in-goods. Not only are countries of origin deeply in need of docking onto the global services markets, but that linkage to readmission, at least in theory, appears to be an immediate one, since services is the only trade flow, hinging, for face-to-face delivery, directly on human factor mobility. Finally, there is in WTO law, a legal basis for enabling such one-way flows of natural persons from developing and least-developing countries (LDCs), without this asymmetric trade posing problem under the WTO GATS most-favored nation clause (MFN), as discussed below.

 

Aggregated conditionality as catalyst of informalizing EU migration cooperation

 

As Frasca (2023) and Desmond (2023) point out, soft law, in the EU external migration context, re-asserts EU sovereignty in instances, where a legally binding obligation on shaky grounds. The key catalyst to jumpstart the process turns out to be conditionality, whitewashed as the comprehensive approach, because it allows to create the traction that soft law lacks. Consequently, conditionality substitutes for a norm failing to deploy a legally binding effect, whether a country refuses to embrace the internationally binding quality of the duty to take back one’s own nationals or considers not being concerned by the duty to take back one’s own nationals. However, the role of conditionality when the EU deploys its armada of informalized migration arrangements, whether partnerships, technical readmission arrangements, standard operation procedures, dialogues still needs more research.

 

Under the New Pact on Migration and Asylum of 23 September 2020, conditionality was up for a supposedly ‘fresh start’ with the  Communication, Attracting Skills and Talents to the EU (27 April 2022), spearheading the Talent partnerships complementing EU mobility partnerships as a multidimensional response to the 18 EU readmission agreements (2023) and 6 arrangements, which regularly fail for non-reciprocally engaging with the sending country ‘s interests and needs (Moraru, Cornelisse and de Bruycker 2022).  Whereas the Commission was not yet breaking with positive conditionality driving much of the EU external migration policy, conditionality’s new focus on vocational and professional upskilling of trainees thru circular Talent Partnerships bears evidence that the Commission is on the lookout for new anchors by which to reinvigorate conditionality, and its EU external migration policy (Tsourdi, Zardo and Sayed 2023).

 

Whereas negative conditionality, which retributes a third country for its lack of cooperation on implementing EU migration policy, has prompted Ethiopia and Afghanistan to sign non-binding readmission declarations (SWP Berlin 2020), the threatened suspension of development aid, has never materialized. Speaking against negative conditionality, as the one the Commission proposes for trade preferences is that raising visa processing charges has not encouraged cooperation (Grundler and Guild 2023), nor is cutting development aid the appropriate penalty for a country such as Nigeria, where remittances are high and thus, installing of vocational training makes more of a difference (Nigeria-Switzerland Migration Partnership of 2011).

 

In 2021, against the background of arrivals by at-sea crossings and over the central route to Europe (ICMPD 2023) tripling, the Commission introduced a new feature to its palette of negative conditionality for non-cooperation over irregular migration. Under its Proposal for a revised GSP Regulation, COM(2021) 579 final for 2024-2034, the EU would now withdraw preferential trade benefits, either under the GSP+ (Generalized System of Preferences) granted to the seven EU beneficiaries (Bolivia, Cape Verde, Kyrgyzstan, Mongolia, Pakistan, Philippines, Sri Lanka), or for least-developed countries (LDCs) under its Everything-but-Arms (EBA), if ‘beneficiary countries on migration and the readmission of rejected asylum seekers’ refuse to cooperate (Guild 2023 in this blogpost).

 

 

Along a well-trodden path—forerunners to identifying trade as leverage for incentivizing migrant returns

 

For the past decade or longer, EU Member States have sought to level the playing field over migration policy among North and West African countries of origin and transit, by proposing one-size-fits-all bilateral migration management, on the basis of Art. 79:5 TFEU in the shape of agreements linking legal pathways to cooperation on returns. France’s agreements on the joint management of migration flows and solidary development for example, stepped up labor admission quotas, or created the same new categories of admission, for all of the seven African countries willing to sign on to a readmission clause. The suggested EU GSP link to cooperation on migration, would run counter to precisely those efforts, that remove treating certain origin countries better than others, thus risking to re-install post-colonial privileges (Robertson 2017).

 

Another forerunner to the prospective trade and (return) migration linkage, is the EU Compact with Jordan which reduces tariffs to duty-free, quota-free exports (DFQF) for products manufactured in Jordan with 10% (first 2 years) and later on, 15% ‘refugee content’. For becoming eligible for this Everything-but-Arms (EBA) privilege, Jordan had to temporarily accept a least-developed country (LDC) status. Whereas Jordan was compensated for employing refugees, in an afterthought, the DFQF occurred with a view to reducing secondary onward movement of refugees to Europe (Lenner 2020). If the EU’s Jordan Compact targeted refugees and not migrants, it was critically received by scholars (Gordon 2021) and advocates of fair and ethical recruitment under ILO standards.

 

Tariff Reduction for Return Migrations? Criticism of the proposed EU GSP 2023-34

 

Trade preferences can be critical for the survival of a developing country on the global market, and their withdrawal carries ethical consequences, as discussed by Tans 2023. Questions about the legality of the EU GSP scheme under WTO law (ODI Report 2023) also arise. Firstly, for Tans, proposing to retract tariffs if migrant returns seem low (to the EU), is not immediate enough a link to human mobility (EP in-depth analysis of the Commission’s proposal of January 2022). In this line of thinking, the Commission would first need to open legal pathways for migrants under the temporary movement of persons, the so-called Mode 4, under an economic partnership agreement (EPA) before it could retract trade preferences in goods.

 

If we recall how Mode 4 GATS stands as the only format of international human mobility that is liberalised internationally, under the multilateral WTO/GATS (Chetail 2014), Mode 4 presents the very connection between trade and temporary migration that is missing from the EU’s proposal. Labelled ‘mode 4’ of Art. 1:2(d) GATS this temporary mobility is a sub-form of international migration. However, it is limited under an excruciatingly narrow definition, to a) temporary stay abroad, b) not entering the labor market (only the services) of the host country c) opened only under commitments by member states d) categories of persons are narrowly defined, often clustering in the highly-skilled segments (Trachtman 2009;). As (Tans in this blogpost) suggests, the EU conditionality working through Everything-but-Arms (EBA) and GSP+ should only apply to those nationals who move under the EU’s GATS mode 4 commitments, if they fail to return voluntarily at the end of their legal temporary stay.

 

Looking for Alternatives 1: Cotonou Convention’s Cooperation on Migrant Returns

 

To this day, the temporary, cross-border movement of natural persons, the so-called Mode 4 has been missing out of EU economic partnership agreements (EPA) with North African countries (Cottier and Shinghal 2021), such that threatening to withdraw tariff preferences under the EU GSP, or even suspending the cross-border mobility so as to penalize countries in North Africa for refusing to take back their own citizens in irregular stays abroad remains illusory. In addition, speaking against penalizing countries of origin for refusing to take back citizens, and thus, against applying the EU’s revised GSP, is Art. 74 on ‘return and readmission’ of the 15 April 2021 negotiated agreement text initialed by the EU and the ACPs chief negotiators -- a follow-up to the Cotonou agreement  -- restates Art. 5 Cotonou agreement with the exception of a return clause which is free from any conditionality—neither is there a negative consequence for failing to take back one’s nationals, nor are typical migrant host countries required to open their labor markets to potential migrant workers.

 

Looking for Alternatives 2: Preferential Trade Agreements and ‘Embedded’ Returns

 

Several economic partnership agreements (EPA) have been consolidating an emerging opinio juris of obliging the origin country to take back their own nationals, once these have terminated their temporary stay to supply a service abroad:  Japan’s EPAs with the Philippines (2008), Indonesia and Vietnam (2009) codify a return clause, which is linked to a services trade commitment. It spells out a requirement for the Filipino, Vietnamese nurses and caregivers to return home, who have failed Japan’s national board examination (NBE). Because this return clause applies solely to the closed-circuit of the categories of persons whose movement the EPA facilitates (Efendi at al. 2013; Naiki 2015), I label it ‘embedded return’. Japan’s EPA of 2019 with the EU, Annex 17 imposes a duty of cooperation on worker’s returns, even if it remains generic when compared to Japan’s EPA with the countries mentioned above. Unlike for what the EU Commission envisages with reforming its GSP regulation, Japan’s EPA carry no negative consequences if either Vietnam, Indonesia or the Philippines fail to cooperate on returns.

 

A Definitive ‘No’? Uncertain Legality of a Trade - Return Migration Linkage under WTO Law

 

As Carzaniga and Sharma 2022 note, WTO Members’ right to regulate emerges from Art. VI GATS. As such, there is a discretionary space under Art. VI, but not an unlimited one, which would permit biometric border surveillance and data collection at the border, under the condition that certain criteria, including transparency are met. However, such broad reading contrasts with the GATS Annex on the Temporary Movement of Natural Persons which stipulates that measures that regulate entry and stay remain under the sovereign right of WTO Members and fall outside the scope of the WTO. Under this narrow interpretation, any PTA linking return duties conditionally to the temporary movement of workers would, in theory, be in breach of the GATS Annex. To summarize, the EU Commission might be infringing WTO/GATS by proposing a GSP reform since the multilateral trade rules of the WTO/GATS Annex preclude a legal connection being made between migrants’ return and trade in services. Beyond the uncertain legality of the Commission’s proposed reform of Regulation EU 1083/2013, there are political economic reasons why arguing in favor of the reform would be bad judgment, discussed further below.

 

People-on-the-move and the ‘new’ EU GSP 2024-34: Moving to the GATS Services Waiver instead?

 

A key consideration speaking against using the revision of the EU GSP to manage the EU’s external migration policy, are the uncertain consequences for countries non-complying with the GSP+ or EBA? In general, countries subject to the GSP+, need to ratify the 27 international law conventions on good governance, labor and human rights, as well as environmental protection, for becoming eligible for the preferential tariff treatment. If the EU deems there is a failure to fully implement provisions of these 27 Conventions, it will normally suspend the preferential tariff and the country’s exports move back to the higher regular tariff (Cambodia in 2019). Now, if the EU considers a ‘failure to cooperate on return migration’ in the same rationale as the 27 Conventions, it conjures a questionable linearity between irregular migration and a non-existent international convention about protecting migrants’ rights. Naturally, the ILO Migrant Workers’ Convention could embody the 28th international convention countries of origin of migrants would be asked to sign to receive the tariff privilege, so as to create a closer lineage.

 

However, since no EU Member State has signed onto to it, this option falls out of question, even if thematically it would address the linkage the EU desires to build. If not the ICMWR, would the 7 GSP+ beneficiaries of the EU, which are Sri Lanka, Cap Verde, Pakistan, the Philippines, Kyrgysztan, in addition to ratifying the 27 UN Conventions be required to sign onto EU readmission agreements, or EU mobility partnerships or the UN anti-smuggling/trafficking protocols as the benchmark for obtaining the lower tariff? Would cooperating with a single EU Member State thru a bilateral readmission agreement be sufficient to hold off higher tariffs on cotton T’shirts or cocoa products or coffee? By the very act of withdrawing trade preferences, if the EU perceives efforts of reducing irregular migration as waning, becomes comparable to suspending tariff preferences from a beneficiary country where corruption starts to spread, narcotics are being produced or trafficked, labor standards are neglected, human trafficking and smuggling take place. In this logic, irregular migration becomes an act that is to be penalized and sanctioned, in the same order as narcotics trading, corrupting business practices or human trafficking and smuggling, which ethically and legally is a questionable nexus to make.

 

There are better ways to incentivize countries of origin to take back their citizens in unlawful stays abroad than to withdraw tariff preferences. One is to use the LDC Waiver of WTO/GATS whereby a host country increases sectoral labor market openings in services for nationals of a country of origin. This scheme has the advantage of closely matching with the EU Talent Partnerships, the former which are sending potential migrants for a training and upskilling to Europe. In concreto, the mechanism is the following: if the LDC cooperates with the EU on irregular migration, it shall obtain additional market access on mode 4 or mode 3 for its natural persons involved in service provision. Under this paradigm, unlike with for the GSP+/EBA, migration is neither treated as a criminal activity that needs to be contained. In fact, the EU is already testing an LDC Waiver type of model in its Skills Partnerships, which ‘buy professionals from an LDC to deliver services in Europe’ (WTO Council for Trade in Services, Webinar on LDC waiver 2-3 June 2021).

 

In preferring the LDC Waiver alternative to vamping up the GSP, the EU would make a strong statement that cross-border mobility, in the first instance, occurs for improving migrants’ wellbeing that it is ‘cooperative rather than punitive’ (FIDH 2023).

 

In addition, the EU GSP+/EBA reform, as proposed by the Commission, discredits its pilot projects on labor migration and skills partnerships, which already have incorporated a return obligation for all the trainees sent to an EU Member State for upskilling. Even if the contentious term of ‘return’, is replaced by the expression of a so-called ‘soft landing’ back home (Garcia Andrade 2020), these EU Talent Partnerships are anything but free from return obligations. Hence, to now enlist the GSP+/EBA scheme for securing even more returns, puts developing countries and LDCs at risk of an additional sanctioning mechanism (Bisong 2022).  

 

Exploring the ‘LDC Waiver’: Securing Cooperation on Returns by Providing Legal Pathways on Mode 4

 

Special and differential treatment (SDT) for the Global South under Art XIX GATS (OECD 2016) calls on  WTO Members to ‘give special priority’ when opening services markets to exports of LDCs.  On the basis of Art XIX, WTO Members took a Ministerial Conference decision on 17 December 2011 to install a LDC services waiver. Since WTO Members were not using it, LDCs were encouraged to make a collective request under the lead of Uganda, to indicate in which sectors of their services industries a waiver of the MFN and a removal of discriminatory barriers  to national treatment (including quotas, licensing requirements, authorization procedures, labor market tests or professional qualifications) could prove development-friendly. By 2015, 50 WTO Members had made offers, and at the Nairobi Ministerial Conference, a decision was adopted to prolong duration of the waiver until 2030.

 

Under the LDC waiver, the EU, US, Canada, Singapore offer ‘best Free Trade Agreement (FTA)-level’ or in 25% of cases above best FTA level, which is the only way for LDC service suppliers to enjoy a competitive advantage on the global services markets. The LDC waiver provides predictability to service traders who otherwise operate under high volatility, so that a sustainable services industry can grow in LDCs and is more in line with day-to-day reality of LDC economies, many of which are no longer export-based in terms of goods. Australia’s opening of warehouse services beyond the categories it has liberalized in the WTO is promising, while Switzerland opened insurance services to LDC providers with lower qualification/notification requirements.

 

Special and Differential Treatment (SDT) under the early years of GATT was limited to developing countries granted preferences amongst each other (1971) and later on, industrialized countries followed suit, when the GATT Enabling Clause L/4903 0f 1979 transformed the 10-year waivers  of the 1970s into permanent ones. In contrast, the LDC Services waiver, while based on the idea of SDT, requires countries, like the EU, to offer non-reciprocal market access in sectors or modes of services delivery of interest to the services exports of developing and least developed Members. Moreover, it is temporary without a clear scope for discretion over how much reciprocity the grant-giving country is prepared to offer. Such facts, including that SDT is generally considered more difficult to implement than trade facilitation (Elsig 2010), render the Commission proposal difficult to digest and even more treacherous to implement. On the upside, the causality of openings in all four modes of services supply or Mode 4 only, would make the case to embed international people-on-the-move and their forced returns within the Services Waiver more plausible, because the fourth mode of service delivery, the so-called Mode 4 of GATS is the only WTO entry point for the temporary movement of natural persons, which can involve migrants, including in irregular stays, depending on national immigration and labor legislation.

 

Conclusions

 

In this blogpost I put forward the case for subsuming a compensation mechanism for countries of origin taking back migrants in irregular stays under the LDC Services Waiver of the GATS, rather than under the 1971 GATT Enabling Clause legitimizing the sequence of EU GSP Regulations.

 

The EU Commission’s heralded ‘positive outcome’ for states cooperating on migrants’ return, appears to be a negative conditionality of withdrawing tariff duties. As such it is less attractive than pledging market access under the LDC Waiver, in more than one way. Firstly, sending countries are often serving as regional services hubs in construction, logistics, and production-related services or becoming global players in tourism and healthcare, such that benefitting from the LDC Waiver, if cooperating on return migration with the EU, resonates with the Global South’s evolution from export-based manufacturing to service economy.

 

Secondly, source countries are more likely to embrace a pledge by the EU to open a services sector, in exchange for ensuring a functioning readmission procedure, than they will actually benefit from already low tariffs. Therefore, to co-opt the LDC Waiver for migration management leads to a fuller integration of countries of origin into global value-chains.

 

Thirdly, the LDC Waiver works without attaching conditionalities. This is key because having too many conditionalities can increase the compliance costs on developing and least-developed countries and backfire, as a disproportionate amount of resources is invested into meeting criteria, rather than on the ground. (US Congress, GSP 2022).

 

In sum, the LDC Waiver not only responds to the WTO’s call for special and differential treatment of developing countries, but it offers more credibility to the revised EU GSP 2024-34 than the GSP, because of its co-ownership by countries of origin. The LDC Waiver opens up valuable export markets in exchange for a duty of taking back one’s own citizens. It would certainly be more in line with the WTO Marrakech Agreement’s Preamble which stresses ‘the need for positive efforts designed to ensure that developing countries …secure a share in the growth in international trade’ over negative retribution as a way to elevate nations out of poverty.