‘Safe countries of origin’ in asylum law: the CJEU first interprets the concept

 

Professor Steve Peers, Royal Holloway University of London

Photo credit: viewsridge, via Wikimedia commons

 

Introduction

Along with judgments on Afghan women asylum seekers (discussed here) and the status of Turkey as a ‘safe third country’ (discussed here), the Court of Justice of the European Union recently gave its first judgment interpreting the substance of the concept of ‘safe countries of origin’ in EU law. The judgment addressed a number of important points, interpreting the law in ways that may have broad relevance to EU law on asylum procedures going forward.

EU legal framework

Previous rules

The concept of ‘safe countries of origin’ goes back, at EU-wide level, initially to ‘soft law’ adopted in the early 1990s (one of the ‘London Resolutions’ of 1992). Next, the principle first took on binding legal form at EU level in the first-phase 2005 asylum procedures Directive, which provided for an option for Member States to accelerate considering asylum applications (albeit in accordance with the usual procedural rules), inter alia where the applicant is from a ‘safe country of origin’. The latter concept was then further defined: it applied to countries (as described in Annex II to the Directive) where:  

on the basis of the legal situation, the application of the law within a democratic system and the general political circumstances, it can be shown that there is generally and consistently no persecution as defined in Article 9 of Directive 2004/83/EC [the first-phase qualification Directive], no torture or inhuman or degrading treatment or punishment and no threat by reason of indiscriminate violence in situations of international or internal armed conflict.

When ‘making this assessment’, Member States had to take account, ‘inter alia, of the extent to which protection is provided against persecution or mistreatment by’:

	- the relevant laws and regulations of the country and the manner in which they are applied;
	- observance of the rights and freedoms laid down in the European Convention for the Protection of Human Rights and Fundamental Freedoms and/or the International Covenant for Civil and Political Rights and/or the Convention against Torture, in particular the rights from which derogation cannot be made under Article 15(2) of the said European Convention;
	- respect of the non-refoulement principle according to the Geneva Convention;
	- provision for a system of effective remedies against violations of these rights and freedoms.

When applying these tests, Member States had to ‘have regard to the legal situation, the application of the law and the general political circumstances in the third country concerned’, and the ‘assessment of whether a country is a safe country of origin’ had to ‘be based on a range of sources of information, including in particular information from other Member States, the UNHCR, the Council of Europe and other relevant international organisations’.

Furthermore, the Directive expressly provided that Member States could designate ‘part of a country as safe where the conditions in Annex II are fulfilled in relation to that part’. It was also open to Member States to continue in place pre-existing rules that fell short of the standards in the Directive to some extent; these rules could also be applied to part of a country, or to designate a country (or part of it) ‘as safe for a specified group of persons’. There were safeguards for individuals: for the ‘safe country of origin’ concept to apply, they would have to be nationals of the country concerned (or stateless persons who were habitual residents of it), and must not have ‘submitted any serious grounds for considering the country not to be a safe country of origin in his/her particular circumstances and in terms of his/her qualification as a refugee in accordance with’ the qualification Directive. Also, Member States were obliged to ‘lay down in national legislation further rules and modalities for the application of the safe country of origin concept’ – although there was no indication of what those rules and modalities might concern.

Finally, there was a power for the EU to adopt common lists of ‘safe countries of origin’, by means of a special procedure. But this power was soon annulled by the CJEU, when the European Parliament successfully challenged the prospect of using such a special procedure.

Current rules

The current asylum procedures Directive, adopted in 2013, retains the ‘safe countries of origin’ concept as an option for Member States to apply an accelerated procedure to. There is no longer a reference to potentially treating only part of a country as safe, and the option for Member States to retain pre-existing lower standards on this issue (along with pre-existing rules on designating part of a country as ‘safe’, or as ‘safe’ for groups of people) was dropped. The notion of a common EU list was not revived in the Directive; although the Commission subsequently proposed such a list (consisting of Western Balkans states and Turkey) in 2015, that proposal did not receive enough support and was withdrawn in 2019. An express requirement for regular review of the supposedly ‘safe countries of origin’ by Member States has been added. Otherwise the rules on ‘safe country of origin’, including the core definition of the concept, were not amended.

However, the rules on appeals, elaborated compared to the previous Directive, now address the issue too. Although the 2013 Directive provides that in principle asylum-seekers can remain on the territory pending resolution of their appeal, there is an exception for (inter alia) most accelerated procedures, including the ‘safe country of origin’ rule. In such cases Member States, if they opt not to give an appeal suspensive effect, must at least allow a court to rule on whether or not the asylum-seeker can stay pending the appeal; and the asylum-seeker can at least stay until a court has ruled on this point.

Before the recent judgment, the CJEU has only ruled on these provisions once, and on a purely procedural point: according to the Court, Sweden could not apply a ‘safe country of origin’ rule until it legislated for one in national law. Sweden duly did so. In passing, the Court confirmed the usual understanding that the ‘safe country of origin’ rules set out a ‘rebuttable presumption’.  

Future rules

The 2024 asylum procedures Regulation has amended the ‘safe country of origin’ rules again. This entails: explaining the concept of ‘non-refoulement’; confirming that rebuttal of the presumption by an asylum-seeker must take place ‘in the framework of an individual assessment’; again expressly providing for exceptions for ‘specific parts’ of the non-EU country’s territory and (not only in the context of pre-existing law) for ‘clearly identifiable categories of persons’; and again allowing for the adoption of a common EU list (this time not via a special procedure, so not repeating the previous legal problem with the concept of such a list). There is no longer any reference to national rules providing for further ‘rules and modalities’ of the concept.

In the broader context, the ‘safe country of origin’ concept is still on the list of possible accelerated procedures, but the use of those procedures will now be mandatory for Member States. In practice the concept might cross over sometimes with the new ground of accelerated procedures where the country of origin has an international protection recognition rate below 20%. It will be expressly possible (although arguably it will only be optional) to apply this ground for accelerated procedures to unaccompanied minors. The implications of accelerated procedures are more explicitly harmonised: three months (instead of the usual six months) to decide on the application, and between five and ten days (instead of the usual two weeks to a month) to bring an appeal. The exception to the right to remain pending appeal (subject to a court ruling otherwise) in such cases becomes mandatory, with some harmonised rules added: a time limit of at least five days to request a court to remain on the territory, along with interpretation and legal aid. As before, accelerated procedures are linked to the rules on a border procedure, which have also been heavily amended.    

National practice

The EU asylum agency has produced a useful map of which countries have been designated as ‘safe countries of origin’ (and ‘safe third countries’) by Member States. Furthermore the agency reported on the national application of these concepts at the end of 2022. As regards ‘safe country of origin’, the report indicates that a large majority (although not quite all) Member States apply the concept. The exceptions are Poland, Latvia, Spain, Bulgaria, Lithuania, Portugal and Romania; Finland applies the concept, but without a list.

Several Member States apply geographic exceptions, namely for parts of Armenia, Bosnia and Herzegovina, Georgia, Moldova, the USA and India. Some Member States also apply exceptions for groups of people, in particular LGBT applicants, minorities, criminal cases, political activists, journalists, human rights defenders, victims of discrimination or forced marriage, and women and girls. There is some variation in which countries are listed as ‘safe’, with Western Balkans states being listed most frequently. The report details a number of judgments of national courts concerning both the listing of particular countries and their application to individual circumstances, with mixed results – some judgments uphold the listing or its application to an individual asylum-seeker, but some do not.  

The judgment

The recent judgment concerned the Czech designation of part of Moldova (except the Russian-occupied Transnistria) as ‘safe’. Faced with an appeal by a Moldovan asylum seeker who was subjected to a ‘safe country of origin’ rule, the national court asked three questions: a) could a country be designated a ‘safe country of origin’ even if it had triggered the derogation from (some) ECHR rights set out in Article 15 ECHR (which Moldova had done after the Russian invasion of neighbouring Ukraine)?; b) could only part of a country be designated as ‘safe’; and c) could a national court raise the legality of the designation of its own motion?

On the first point, the Court ruled that invoking Article 15 ECHR did not automatically prevent or cancel the designation of a country as a ‘safe country of origin’. This was because the ECHR set out guarantees for use of the derogation in Article 15, limiting its use to cases where it was necessary, requiring no conflict with other international law, allowing derogations only from certain rights and remaining subject to review by the European Court of Human Rights. (On the relevant case law, see that Court’s guide to Article 15). Also, it could not necessarily be assumed that invoking Article 15 meant that rights were actually derogated from; nor did use of the derogation as such determine the nature and extent of the derogation.

However, the CJEU pointed out that triggering an Article 15 ECHR derogation must mean that Member States had to review the ‘safe country of origin’ listing, on the basis of the obligation to review the designations in light of developments set out in the Directive. That review should be triggered by significant events likely to affect whether a non-EU country still meets the ‘safe country of origin’ criteria, which include the invocation of a derogation from the ECHR – even though the derogation would affect rights other than Article 3 ECHR (the ban on torture or other inhuman or degrading treatment), because Article 15 ECHR does not allow derogation from Article 3.

The Court did not address the arguments in the Advocate-General’s opinion, concerning the protocol on asylum for EU citizens by analogy. That protocol provides that EU Member States are ‘safe countries of origin’ for each other’s citizens, meaning (in the context of the protocol) that asylum applications cannot be considered at all – unless a Member State either: invokes a derogation under Article 15 ECHR; or the Article 7 TEU process of either sanctioning a Member State for a ‘serious and persistent breach’ of EU values, or warning a Member State that there is a ‘clear risk of a serious breach’ of such values, has been applied (or triggered, as regards a warning; on this point, see the Hungary v European Parliament judgment); or a Member State decides to do so unilaterally for an individual applicant. In the Advocate-General’s view, the protocol could not apply by analogy due to its different context: an expression of mutual trust applicable to EU citizens only.

On the second question – whether a non-EU country could be designated as a ‘safe country or origin’ only in part – the Court first looked at the wording of the Directive, which used the word ‘country’ without indicating that it may be referring to only part of the territory. Secondly, the context of the Directive included the definition of ‘safe country of origin’ in the Annex; that definition did not suggest either that only part of a country could be designated, and the words ‘generally and uniformly’ suggested the opposite.

Next, the ‘safe country of origin’ rule was an ‘exception’ to the normal process of considering asylum applications, and such derogations should have a ‘strict interpretation’ (referring to the case law on inadmissible applications), notably where an interpretation had no support from the text. Also, the Court looked at the historical context, noting that: the 2005 Directive expressly provided for parts of countries to be listed; the 2013 Directive does not; and the proposal for what became the 2013 Directive explicitly stated an intention to remove the possibility. Finally, the Court referred to the objectives of the asylum procedures Directive, taking the view that the EU legislature was trying to balance the objectives of a speedy versus a thorough assessment of asylum applications. This was a discretionary choice that was reversed by the 2024 Regulation; a change of mind was the prerogative of the EU legislature, provided that it complied with the Refugee Convention and the Charter. And the wording of the 2024 Regulation reinforced the point that designation of only part of a country was not provided for in the 2013 Directive.  

Again the Court’s judgment does not discuss a point analysed by the Advocate-General – as to whether the ‘effective territorial control’ principle, which limits the liability of countries (including Moldova, in this context) for breaches of the ECHR on their territory controlled by another State, applies in the context of EU asylum law to allow for the designation of only part of their territory as ‘safe’. In the Advocate-General’s view, the issue of state liability for breaches of international law is distinct from the issue of the listing of ‘safe country of origin’ under EU asylum law; and the EU legislature decided to provide for listing of whole countries only, even being aware of the ‘effective territorial control’ issue.

On the third point – consideration of the designation of ‘safe country of origin’ by a court on appeal, of its own motion – the CJEU ruled that, taking account of its prior case law giving a strong role to courts reviewing asylum decisions, including on procedural issues, such designations have to be reviewed by courts of their own motion even if the asylum applicant does not raise the point. In particular, the courts have to review the invocation of derogations from the ECHR and the territorial scope of the ‘safe country of origin’ designation. Again, the Court did not address an issue raised by the Advocate-General – the analysis, in the alternative, of how the EU law principle of equal and effective remedies would apply in a case like this one.

Comments

The Court’s judgment raises a number of points that could be relevant to EU asylum procedures law more generally. On the first point (the ECHR derogation), the ruling at first sight may only apply to ECHR parties; but logically it applies by analogy also to any derogations from other human rights treaties (including the ICCPR, which is expressly mentioned in the Directive), and/or from national human rights protection – taking into account any relevant differences from the ECHR (the prospects of effective judicial review of the derogation, and the particular rights being derogated from, which were key features of the Court’s judgment). After all, human rights abuses (triggering asylum claims) often take place during supposed states of emergency. 

The judgment also makes clear that reviews of ‘safety’ must be triggered by significant events – not only, as Member States might prefer, taking place at regular prescheduled intervals. It follows that a failure to conduct a review in light of a significant event is a procedural flaw that can be challenged. Moreover, this judgment should logically apply by analogy to the reviews of the EU common list provided for in the asylum pact Regulation – as well as to national and EU reviews of ‘safe third countries’.  

As for the dog that didn’t bark – the comparison with the protocol on asylum for EU citizens – the better argument than in the Advocate-General’s opinion is surely that in addition to the obvious different in wording, the context of the Directive is hugely different too. Although the protocol uses the phrase ‘safe countries of origin’, unlike the Directive it does not create a rebuttable presumption, but rather a complete ban on considering applications (in principle) from EU citizens; so it makes sense that the conditions for applying the protocol are more stringent, since the key safeguard of the possibility of the individual right to attempt to rebut the presumption is absent.

On the second point – the inability to label part of a country as ‘safe’ – as we can see from the practice above, this would mean changes in a number of Member States. Of course, this is only short term, because the asylum pact legislation will soon allow such designations again. But the Court’s express reference to the new Regulation may give us a broader indication of how it will interpret the pact. It suggests that while the EU legislature has a discretion to choose between prioritising speedy or thorough consideration of applications, that choice must comply with the Charter and the Refugee Convention. Given the reduction in standards in the Regulation – and the further reductions in standards that the Commission and many Member States seem to be gagging to make in the near future – the Court’s assessment  of whether EU asylum law indeed complies with the Charter and the Refugee Convention is likely to be of increasing importance in the future.

This part of the judgment also confirms that the concept of ‘safe country of origin’ is an exception which must be interpreted strictly, by analogy with the case law on inadmissibility which often finds that Member States’ interpretations of the grounds for inadmissibility have been too restrictive. This logically by analogy to all grounds for accelerated procedures, a fortiori after they become mandatory under the 2024 asylum pact.

Finally, the third point – judicial review of the ‘safe country of origin’ designation, on its own initiative – logically applies by analogy too, to other aspects of the principle besides those at issue in this case, and to listings of ‘safe third countries’ too. Although strengthening judicial review may have limited effect in such cases, given the absence of suspensive effect of appeals, confirming the wider power of courts to examine the validity of designating countries in the list on their own motion may make it easier to convince them to grant such suspensive effect.

Latest Updates on The Legitimate Interest Ground for Processing Personal Data (Article 6(1)(f) of GDPR): the latest CJEU Case and EDPB New Guidelines

 

 

 

Aolan Li*

*The author is a third-year PhD candidate in Law at Queen Mary University of London. Her ongoing doctoral thesis research delves into the application of Article 6(1)(f) of GDPR from a comparative perspective. Email: aolan.li@qmul.ac.uk

Photo credit: TheDigitalArtist, via Wikipedia Commons

 

A positive spirit has spread among business-side stakeholders across the EU since the Court of Justice of the European Union (CJEU) published its preliminary ruling in the Koninklijke Nederlandse Lawn Tennisbond case (C-621/22) on 4 October 2024, where the court confirms a purely commercial interest could constitute a legitimate interest for processing personal data under Article 6(1)(f) of GDPR. Commentators go as far as to say - “what this means is that under the GDPR, your data can be used without your consent solely for a company’s commercial interests.”

The preceding saying is a total misunderstanding. The positive spirit should have been dampened as, on 9 October 2024, the European Data Protection Board (EDPB) published its new guidelines on Article 6(1)(f) for public consultation (hereafter as the new EDPB guidelines).

Bearing in mind the optimistic bubbles in the market, this writing articulates the EDPB’s stringent stance on the application of Article 6(1)(f) of GDPR, focusing on what has changed compared to the Article 29 Data Protection Working Party’s opinion on the legitimate interest ground under Directive 95/46/EC (hereafter as the WP29 Opinion).

General remark

The newly published EDPB guidelines align with the WP29 Opinion in some basic stances. First and foremost, the recognition of a legitimate interest is not itself sufficient to rely on Article 6(1)(f) of GDPR as a legal basis (this is why the saying is misleading) as there are three cumulative conditions for its application. Secondly, Article 6(1)(f) of GDPR should not be used “by default” nor as a “last resort”. The open-ended nature of Article 6(1)(f) has a unique role in the EU data protection law.

Not surprisingly, the new EDPB guidelines also substantially update the WP29 Opinion.

The update is partially attributed to judgments of CJEU issued after the adoption of the WP29 Opinion, including Rīgas (Case C-13/16), Fashion ID (C-40/17), TK (C-708/18), MICM (C-597/19), Meta v Bundeskartellamt (C-252/21), SCHUFA Holding (Joined Cases C-26/22 and C-64/22), and the latest Koninklijke Nederlandse Lawn Tennisbond (C-621/22). Many practical examples in the new guidelines mirror scenarios disputed in the abovementioned cases. For example, example 4 is analogous to Rīgas.

Building upon more detailed case law, the new EDPB guidelines are more logical and clearly articulated. Unlike the WP29 Opinion, the new guidelines make effects to draw a clearer line between the six grounds for legitimising data processing under Article 6(1) of GDPR. Also, the new guidelines follow the now well-accepted three-step approach to applying Article 6(1)(f), which was established by the CJEU in its judgment in Rīgas.

The update also corresponds to the evolvement of the law itself (GDPR vs Data Protection Directive). GDPR has strengthened data subject rights. It is worth noting the improvement of the right to object - a specific right for the processing based on Article 6(1)(e) and (f) of GDPR - as the burden of proof has been reversed on the controller. Also, GDPR and CJEU case law have escalated the reasonable expectation of data subjects to a more significant position in determining the application of Article 6(1)(f) of GDPR. Therefore, the new guidelines are observed to enhance the position of data subjects accordingly.

Besides being consistent with legislative developments and the CJEU’s case law, the EDPB is observed to add its unique understanding to narrow down the scope of Article 6(1)(f) of GDPR; here’s why I said the EDPB takes a stringent stance. The next part provides more discussions.

Overall, the new guidelines have been compiled from rich and up-to-date sources and provide much more nuanced interpretations of Article 6(1)(f) of GDPR. However, one might lament that Part IV of the new guidelines hesitates to touch on the application of Article 6(1)(f) of GDPR in more complicated and controversial contexts. For example, its application in the credit scoring industry seems like a real-world need, as demonstrated in SCHUFA Holding. Let alone its silence on applying Article 6(1)(f) of GDPR in AI-related scenarios.

The writing below touches on the substantial content of Article 6(1)(f) of GDPR. However, it does not intend to sketch the 37-page guidelines reductively. Instead, it aims to highlight the stringent stance of the new guidelines, read together with the Koninklijke Nederlandse Lawn Tennisbond case.

The Three Steps Approach

As mentioned above, three cumulative conditions must be fulfilled to rely on Article 6(1)(f) of GDPR as a legal basis, called the three steps approach, which are 1) the pursuit of a legitimate interest by the controller or by a third party; 2) the need to process personal data for the purposes of the legitimate interest(s) pursued; 3) the interests or fundamental freedoms and rights of the concerned data subjects do not take precedence over the legitimate interest(s) of the controller or of a third party (the new EDPB guidelines, p 2).

For the first step, the new guidelines narrow down the scope of interests with respect to the controller’s own interests and disentangle the third party’s interests from wider public interests.

As the information circulated, the qualifier “legitimate” is interpreted broadly, covering any interests that are not contrary to the law (Koninklijke Nederlandse Lawn Tennisbond, para 49).

However, sourced to the CJEU judgment in Meta v. Bundeskartellamt, the new guidelines confine that “as a general rule, the interest pursued by the controller should be related to the actual activities of the controller.” (the new EDPB guidelines, para 19) It means that, within the meaning of Article 6(1)(f) of GDPR, a controller whose activity is economic and commercial in nature is only allowed to pursue economic and commercial-related interests.

Other legitimate but non-economic/commercial interests might fall within the scope of interest(s) pursued by a third party. The new guidelines clarify that the controller needs to demonstrate the legitimate interest(s) are pursued by one or more specific third parties (para 20-25) and should not be confused with broader public interests despite the fact they can overlap, as seen in SCHUFA Holding.

Remarkably, the new EDPB guidelines indicate that relying on the interest(s) pursued by a third party in the first step is generally more challenging to pass the latter two steps (the necessity and balance test) than relying on the controller's own interests. (para 30)

For the second step, the processing involved should be necessary for the purposes of that interest identified in the first step, called the necessity test. The concept of necessity has its own free-standing meaning in EU law. The controller must demonstrate that there are no other reasonable, just as effective, but less intrusive alternatives to achieve the pursued legitimate interests.

Despite no given example in the new EDPB guidelines, the CJEU judgment in Koninklijke Nederlandse Lawn Tennisbond provides a least intrusive scenario in the direct marketing context. To be brief, without asking for consent, a Netherlands sports federation (KNLTB) sold its members’ personal data to its sponsors for the latter’s marketing purposes. The court considers it possible for KNLTB “to inform its members beforehand and to ask them whether they want their data to transmitted to those third parties for advertising or marketing purposes.” (para 51) The court deems a procedure as such may involve the least intrusion of data subjects’ rights and compliance with data minimisation principles. As will be explained below, the proposed approach resonates with the right to object and controllers’ notification obligations.

For the third step, the balance test entails a balancing of the controller side's rights and interests against those of the data subject side. The controller needs to ascertain, on a case-by-case basis, that the processing at issue would not disproportionately impact the data subject’s rights and interests.

One can observe the improved position of data subjects directly from the structures of exercising the balance test in the new EDPB guidelines and the WP29 Opinion. (See table below)

Methodology for the balancing test under new EDPB guidelines	Methodology for the balancing test under the WP29 opinion
The data subjects’ interests, fundamental rights and freedoms.	Assessing the controller’s legitimate interest. -      Exercise of a fundamental right; -          Public interests/the interests of the wider community; -          Other legitimate interests; -          Legal and cultural/societal recognition of the legitimacy of the interests.
The impact of the processing on data subjects, including -          The nature of the data to be processed; -          The context of the processing; -          Any further consequences of the processing.	Impact on the data subjects -          Assessment of impact; -          Nature of the data; -          The way data are being processed; -          Reasonable expectations of the data subject; -          Status of the data controller and data subject.
The reasonable expectations of the data subject.	Provisional balance.
The final balancing of opposing rights and interests, including the possibility of further mitigating measures.	Additional safeguards applied by the controller to prevent any undue impact on the data subjects.

Despite most of the content continuing to work, some remarkable points exist.

Firstly, the reasonable expectation of the data subject has been escalated to an independent element. It goes beyond the controller’s notification obligation and highlights the data subject’s genuine understanding; as the new EDPB guidelines put it, more than the mere fulfilment of Articles 12, 13, and 14 is needed to consider that the data subject can reasonably expect the said processing. (para 53)

Secondly, the mitigating measures, be it technical and organisational, within the meaning of Article 6(1)(f) of GDPR must go beyond existing principles and obligations set out in the GDPR. In this sense, the new EDPB guidelines encourage controllers who intend to rely on Article 6(1)(f) of GDPR to pursue a higher level of personal data protection than legal obligations.

Data subject rights

A comprehensive review of the enhanced data subject rights under the GDPR goes beyond the subject matter of this writing. Calling back to the least intrusive approach proposed in Koninklijke Nederlandse Lawn Tennisbond, this part of the writing articulates the significance of controllers’ notification obligations and data subjects’ right to object in the context of Article 6(1)(f) of GDPR.

The court considers that KNLTB can inform its members beforehand. KNLTB’s notification obligations are set out in Articles 13 and 14 of GDPR. Its members (data subjects) should be informed about, among other things, the legal basis of processing, the specific legitimate interests pursued by KNLTB or its sponsors, and data subject rights. According to Article 13(3), KNLTB should inform its member concerned prior to further processing.

The court also considers it good practice for KNLTB to ask members concerned whether they want their data transmitted to third parties for advertising or marketing purposes. One might feel at odds with the reintroduction of “consent” in assessing Article 6(1)(f) of GDPR. Actually, it is better to understand the “ask” as informing its members concerned about their right to object under Article 21 of GDPR.

The objection to direct marketing based on Article 6(1)(f) of GDPR is absolute. In other cases, however, the controller might have compelling legitimate grounds to disapprove the right. Here, it involves another balancing test to determine whether the controller has a compelling legitimate ground. Unlike Directive 95/46, the burden of proof is on the controller.

The new EDPB guidelines promote the idea that the controller’s compelling legitimate grounds can only be recognised in exceptional cases. The controller cannot circumvent the right to object by merely showing that the processing would be beneficial to the controller. Rather, the concept of compelling is understood as essential to the controller.

From the preceding standpoint, the right to object has been improved in favour of the data subject, and it is not much inferior to the right to withdraw consent.

Concluding remarks

In conclusion, while the CJEU’s preliminary ruling in the Koninklijke Nederlandse Lawn Tennisbond case initially sparked optimism among business stakeholders by holding that purely commercial interests could qualify as legitimate under Article 6(1)(f) of the GDPR, this enthusiasm is misplaced. The ongoing EDPB’s new guidelines underscore a more restrictive interpretation of the legitimate interest ground than the earlier WP29 Opinion, reinforcing the need for careful application and a balanced approach to personal data protection. This writing calls for a self-reassessment of GDPR compliance, in particular for controllers relying on legitimate interest as a main legal basis.

Pyrrhic victory for the Greek government: the CJEU rules on Turkey as a “safe third country”

Professor Steve Peers, Royal Holloway University of London

Photo credit: Mstyslav Chernov, via Wikimedia Commons

Hamster idiom credit: Adam Sharp

(Thanks to Zoe Gardner for discussion of the case; the following analysis is my own)

 

Introduction

Remember the ‘refugee crisis’ of 2015-16? One part of the EU’s response to it was to negotiate a ‘joint statement’ with Turkey (see comments and criticism here, here and here) that, among other things, led to Turkey agreeing to take back asylum-seekers from the Greek islands. Greece facilitated the application of this process by determining that Turkey was a ‘safe third country’ (ie that asylum-seekers could and should have applied there instead, given that this other country is ‘safe’) under EU asylum law, initially for Syrian asylum-seekers in the eastern Greek islands. Faced with a legal challenge to this determination, the Greek Council of State ruled in 2017 that the determination was legal, deciding by a 13-12 vote not to ask the CJEU to interpret the issue of EU law at stake. For its part the EU General Court (upheld on appeal by the CJEU) rejected legal challenges to the joint statement, on the frankly bizarre ground that the European Council had nothing to do with the EU/Turkey joint statement (which referred to EU law issues such as visa waivers).

Subsequently in 2020, Turkey decided to refuse all readmission of asylum-seekers from Greece, stating that (as quoted in note 10 of the Advocate-General’s opinion):

 Responding to repeated requests from the Greek authorities and the European Commission regarding the resumption of return operations, [the Republic of Türkiye] has stated that no return operation would take place unless the alleged pushbacks along the Turkish-Greek border stop and [the Hellenic Republic] revokes its decision to consider [the Republic of Türkiye] a Safe Third Country.

Greece nevertheless decided to list Turkey as a ‘safe third country’ more broadly in 2021, for asylum-seekers from Syria, Afghanistan, Pakistan, Bangladesh or Somalia; it added Albania and North Macedonia to its ‘safe third country’ list soon afterwards. In light of the EU law rules on ‘safe third country’ (discussed further below), two NGOs (the Greek Council for Refugees and Refugee Support in the Aegean) decided to challenge the Greek government decision to list Turkey, resulting this time in the view of the Council of State in February 2023 (by an 18-4 vote) that the listing of Turkey was invalid, on the grounds that EU law precluded designating a country which refused to readmit asylum-seekers as a ‘safe third country’. (Other EU law arguments against the designation of Turkey were rejected by the national court and not referred to the CJEU: see the analysis by Minos Mouzourakis).  But unlike in 2017, there was great deference to the view of the (much smaller) minority (two judges apiece argued that the issue of readmission was relevant instead when ruling applications inadmissible, or when enforcing that decision). So the national court decided to ask the CJEU questions about the interpretation of EU law on the readmission point before giving its final ruling. The CJEU delivered its judgment last Friday.

This is not only the first time that the CJEU has been asked about an aspect of the EU/Turkey asylum relationship. Remarkably, despite the huge number of asylum seekers that have come to Greece over the last decade and more, this is also the first time that the CJEU has been asked any questions about EU asylum law by any Greek court.

 

EU legal framework

Legislation

The definition of ‘safe third country’ for the EU is currently set out in Article 38 of the EU’s asylum procedures Directive. (This is about the ‘safety’ of non-EU countries, ie it is distinct from the EU Member States regarding each other as ‘safe’ countries, under the Dublin rules) First, Article 38(1) provides that the (optional) ‘safe third country’ concept can ‘only’ be applied if Member States are satisfied that certain ‘principles’ are ensured for asylum-seekers in that country: (a) ‘life and liberty are not threatened on account of race, religion, nationality, membership of a particular social group or political opinion’; (b) ‘there is no risk of serious harm as defined in’ the EU qualification Directive; (c) ‘the principle of non-refoulement in accordance with the Geneva [Refugee] Convention is respected’; (d)  ‘prohibition of removal, in violation of the right to freedom from torture and cruel, inhuman or degrading treatment as laid down in international law, is respected’; and (e) ‘the possibility exists to request refugee status and, if found to be a refugee, to receive protection in accordance with the Geneva Convention’.

Secondly, Article 38(2) requires Member States to subject the application of the ‘safe third country’ rule to national law rules, including: (a) ‘requiring a connection between the applicant and the third country concerned on the basis of which it would be reasonable for that person to go to that country’; (b) ‘rules on the methodology by which the competent authorities satisfy themselves that the safe third country concept may be applied to a particular country or to a particular applicant. Such methodology shall include case-by-case consideration of the safety of the country for a particular applicant and/or national designation of countries considered to be generally safe’; and (c) ‘rules in accordance with international law, allowing an individual examination of whether the third country concerned is safe for a particular applicant which, as a minimum, shall permit the applicant to challenge the application of the safe third country concept on the grounds that the third country is not safe in his or her particular circumstances’. Also, an asylum-seeker must ‘be allowed to challenge the existence of a connection between him or her and the third country’, as referred to in point (a).

Next, Article 38(3) requires Member States to inform the applicant that the principle is being applied to them, and also to give the asylum-seeker a document informing the supposedly ‘safe’ third country that the merits of their asylum claim have not been examined.

Crucially for this case, Article 38(4) then provides that:

Where the third country does not permit the applicant to enter its territory, Member States shall ensure that access to a procedure is given in accordance with the basic principles and guarantees described in Chapter II. [ie an ‘ordinary’ examination of the merits of an asylum claim]

Finally, Article 38(5) obliges Member States to inform the Commission of their use of the concept.

The practical relevance of applying the ‘safe third country’ notion is (among other things) that it is a ground of (optional) inadmissibility under Article 33 of the Directive. And if the case is inadmissible, the asylum application is not further considered on the merits and has failed in that Member State (subject to an appeal of the inadmissibility decision, or the limited prospect of a ‘repeat application’, discussed further below). The failed asylum-seeker can then deprived of support as an asylum seeker under the reception conditions Directive and detained and removed under the Returns Directive. Of course, the impact of this is in theory ameliorated in ‘safe third country’ cases because the supposedly ‘safe’ country will readmit the person concerned and consider their asylum application properly. In practice, as we have seen, Turkey was explicitly and unambiguously refusing to readmit asylum-seekers from Greece in the context of the ‘safe third country’ concept.

Case law

The CJEU had interpreted the ‘safe third country’ rule in some prior case law, all concerning Hungary. First of all, in Case C‑564/18 (paras 29-51 of the judgment), the CJEU confirmed that the grounds of inadmissibility listed in the Directive were exhaustive (‘only’), and ruled that a national law applying a ‘safe third country’ principle to non-EU States where the asylum-seeker ‘as not exposed to persecution or a risk of serious harm, or in which a sufficient degree of protection is guaranteed’ was not consistent with the requirements of Article 38(1) of the Directive, in particular because it omitted the ‘non-refoulement’ requirement. Moreover, mere transit through the non-EU country was not enough for there to be a ‘connection’ with that country as required by Article 38(2).

Next, the Court reiterated these points in its judgment in Joined Cases C-924/19 and C-925/19 PPU (paras 148-165); but questions about what happens when the non-EU State refuses to readmit the asylum-seekers – the issue since addressed in last week’s judgment – were ruled inadmissible (paras 166-174). However, the Advocate-General’s opinion in that case discusses the issue (paras 107-127), taking the view that the refusal of readmission is only relevant at the point of enforcement of a decision based on the principle.  

Finally, in Commission v Hungary (Case C‑821/19), along with ruling on the criminalisation of assistance to asylum seekers, the Court again reiterated that the Hungarian law breached Article 38(1) and (2) of the Directive. The existence of a (then) pending proposal to amend the Directive on this point (since adopted as part of the 2024 asylum pact – but without some of the changes originally proposed) was irrelevant in determining whether a Member State was in breach of EU law at the time of the alleged breach.

 

Greek practice

As noted above, Greece had listed Turkey as a ‘safe third country’ despite Turkey’s refusal to readmit anyone in that context. While the CJEU judgment says nothing about the practice of the authorities, and the Advocate-General’s opinion says little (paras 65-68 of the opinion), one can find rather more information at the bottom of the ivory tower, in the recent Pro-Asyl report on the ‘safe third country’ principle in Greece, and in the annual AIDA report on Greece (‘safe third country’ section).

The reports state that over 10,000 asylum claims have been found inadmissible on the basis of the Greek listing of Turkey as a ‘safe third country’, although around 12,000 asylum claims have been found admissible (in some cases because the asylum-seekers argued that Turkey had not allowed them to make an asylum claim, or due to past exposure to refoulement, ie return to an unsafe country, or because they lacked a sufficient ‘connection’ with Turkey, on the basis of the national law implementing Article 38(2) of the Directive). Also, the reports are highly critical of the conclusion that Turkey is ‘safe’ – although the recent CJEU judgment does not address this issue, because the national court did not ask the CJEU about it.

Like the Advocate-General’s opinion, the Pro-Asyl study reports that Greek authorities apply the readmission test at the point of enforcement of a ‘safe third country’ decision. When Turkey then refuses readmission, the Greek bodies refuse to withdraw their prior decisions, leaving asylum-seekers to make a ‘repeat application’. But there are delays in registering these applications, and the authorities often reject them because they do not believe that the refusal of readmission is a ‘new element or finding’ (as required for a repeat application to be admissible under the EU Directive). But at least some courts rule against detention in such cases, given that there is no reasonable prospect of return (as required by the Returns Directive).

 

The judgment

The CJEU’s judgment began by answering the first question referred by the national court, rephrasing it as asking ‘in essence, whether’ the ‘safe third country’ rule in the Directive, ‘read in the light of Article 18 of the Charter’ on the right to asylum, ‘must be interpreted as precluding legislation of a Member State classifying a third country as generally safe for certain categories of applicants for international protection where, despite the legal obligation to which it is subject, that third country has generally suspended the admission or readmission of those applicants to its territory and there is no foreseeable prospect of a change in that position’.

First, the Court reiterated its case law that the grounds for inadmissibility of asylum applications in the procedures Directive (including ‘safe third country’ as defined in the Directive) are ‘exhaustive’, ie the Member States cannot add more grounds of their own (such as a divergent national definition of ‘safe third country’). It pointed out that the EU law definition ‘is subject to compliance with the cumulative conditions laid down’ in Article 38 of the Directive, as summarised above. But as the national court had already ruled that Greek law complied with the basic principles in Article 38(1), the CJEU did not examine that point. But it summarised the requirements for national law set out in Article 38(2) and the information obligations in Article 38(3), then recalled the Article 38(4) rule that ‘where the third country concerned does not permit the applicant for international protection to enter its territory, the Member States are to ensure that access to a procedure is given in accordance with the basic principles and guarantees described in Chapter II of that directive’.

In the CJEU’s view, it followed from this first from the wording of the Directive that a Member State can designate a country as ‘generally safe’ by ‘an act of general application’. The references to national law in Article 38(2) referred to such a designation, as did the preamble to the Directive. Also, the Directive did not make the validity of such a general act ‘subject to the condition that it be proved that the applicants for international protection concerned will actually be admitted or readmitted to the territory of that third country’. This was not among the national law rules in Article 38(2); and the wording of Article 38(4) implied that the designation of the third country as ‘safe’ preceded the consideration of the issue of whether that country actually readmitted asylum seekers.

The CJEU also believed that this followed from the context of the Directive, in particular by comparison with the wording of the rules on ‘first country of asylum’ (ie where the asylum seeker already had protection, as distinct from where they could have sought it) on the readmission issue.

Finally, in the CJEU’s view, this interpretation also did not conflict with the objective of the Directive, as set out in the preamble, that in the interests of asylum-seekers in particular, asylum applications ‘should be made as soon as possible, without prejudice to an adequate and complete examination being carried out’. This objective was addressed by Article 38(4), requiring that a Member State ‘ensure’ access to an asylum decision on the merits, in accordance with the usual rules, if the ‘safe third country’ refused readmission. And then a key point:

54      It follows that, where it is established that the third country designated as generally safe by a Member State does not in fact admit or readmit the applicants for international protection concerned, that Member State cannot reject their applications for international protection as inadmissible on the basis of Article 33(2)(c) of Directive 2013/32. Furthermore, that Member State may not unjustifiably postpone the examination of those applications and must, inter alia, ensure that that examination is conducted on an individual basis, in accordance with Article 10(3)(a) of that directive and in compliance with the time limits set out in Article 31 thereof.

This interpretation, in the Court’s view, ‘is likewise not such as to deprive of any practical effect the right of an applicant for international protection, as enshrined in Article 18 of the Charter and given specific expression by that directive, to obtain the status of beneficiary of international protection, provided that the conditions required by EU law are met.

Having ruled, in effect, that the designation of Turkey was valid, the CJEU felt it was unnecessary to answer the second and third questions referred by the national court, on whether the refusal of readmission was a factor to be considered at the admissibility or the enforcement stage. But as we have seen, the Court had already de facto answered the second question: if it is ‘established’ that the ‘safe’ third country ‘does not in fact’ readmit the asylum seekers concerned, a Member State ‘cannot’ find their applications admissible, and must consider those applications on the merits (moreover, without ‘unjustifiable’ postponement, and subject to the deadlines in the Directive). And that renders the third question moot.

 

Comments

Greece can have a dog; but it will be wagged by its Turkish tail. That is the effect of the Court’s judgment: in more technical legal language, Turkey can remain on Greece’s ‘safe third country’ list, but the practical impact of this is zero as long as Turkey refuses to readmit asylum-seekers from Greece, because asylum applications cannot be found inadmissible on this ground. The wheel of designation of Turkey as a ‘safe third country’ keeps spinning, but the hamster of inadmissibility of asylum applications is dead. Let’s unpack this further, examining first the reasoning of the judgment and then its potential consequences.  

Judicial reasoning

The CJEU begins by, with respect, answering a ‘straw man’ question that it had not been asked.  The national court did not ask the CJEU if it was possible to designate ‘safe third countries’ at all; rather it asked if those countries could be designated if they failed to readmit asylum-seekers. The references to common principles on ‘safe third countries’ in the preamble to the Directive do not answer the question as to what those common principles are; a rule that failure to readmit precludes listing a country would be a ‘common principle’ too. However, the Court’s points that Article 38(2) on national law and (especially) Article 38(4) on readmission presume that the country concerned is already on a list are more convincing.

Most importantly, the Court’s finding that applications cannot be inadmissible on ‘safe third country’ grounds where the country concerned refuses to readmit, requiring a Member State to consider the merits of such applications, is particularly convincing.  Given that between Greece and Turkey the readmission will not take place, and in light of the evidence that repeat applications are difficult to make and are often rejected, this is the best way to ensure that the obligations set out in Article 38(4), coupled with the principle of considering applications as soon as possible, are met.

Unlike some judgments, this is not so much a ‘liberal’ or ‘activist’ interpretation of EU asylum law by the CJEU, but rather the overdue termination of conservative activism by the Greek authorities and lower courts. In light of the clear obligation in Article 38(4) to consider the merits of an asylum application if the asylum seeker is not readmitted to a ‘safe third country’, in the context of the absolute refusal of Turkey to readmit, the failure of the authorities to consider the merits of many of those applications, and of the courts to require them to do so, is unjustifiable and frankly shocking.

Consequences in practice

First and foremost, what happens to the 10,000 plus asylum seekers whose applications were already rejected as inadmissible? As noted above, some had previously tried to make repeat applications, but faced the difficulty that they had to invoke ‘new elements or findings’ in order to do so, which was often difficult to do. But their ability to invoke the ‘repeat applications’ clause in the Directive has now, as a consequence purely of the Court of Justice’s recent judgment, undergone a Copernican revolution. This is because the CJEU has previously ruled (see Joined Cases C-924/19 and C-925/19 PPU, paras 175-203, confirmed in Case C-216/22) that its own judgments on EU asylum law are a new element or finding in filing a repeat application, where those judgments are relevant and add significantly to the likelihood of an asylum application being successful. For those asylum-seekers whose application was found inadmissible on ‘safe third country’ grounds despite Turkey’s refusal to readmit them, last week’s judgment is obviously highly relevant; and overturning the finding that their applications were inadmissible equally obviously significantly increases the odds that their application will be successful. (Moreover, the judgment in Joined Cases C-924/19 and C-925/19 PPU is especially relevant by analogy here, since the prior national judgments which conflicted with a later CJEU judgment in that case also concerned an unlawful application of the ‘safe third country’ rules).

Therefore, if the rejected asylum-seekers apply again, the Greek asylum system will now have to decide on the merits of thousands of asylum applications that it wrongly ruled were inadmissible in the first place. A pending CJEU case referred from Ireland may be relevant by analogy in establishing whether Greece has to pay damages too. Finally, if any Afghan women asylum-seekers were among those whose applications were considered inadmissible, they can also rely on another CJEU judgment from the same day (discussed here) as a ‘new element or finding’, supporting the merits and fast-tracking of their gender-based persecution claims.

As for current or future applicants, their applications cannot be rejected as inadmissible on the grounds that Turkey is a 'safe third country' at all. Nor can they be rejected on the grounds that Turkey is a ‘first country of asylum’ for them: without even looking at the substance of that concept, the CJEU judgment explicitly stated that this principle cannot even get off the ground if the country concerned will not readmit the asylum seeker.

What about applicants after the asylum pact takes effect? The Court of Justice judgment from last week only mentions the new asylum procedures Regulation (discussed here) in passing, but as the Advocate-General’s opinion pointed out (in para 64), the interpretation of the Directive ultimately followed by the Court is explicitly set out in the Regulation, now being built into the definition of inadmissibility – also as regards the ‘first country of asylum’ principle (these grounds of inadmissibility can be invoked for either principle ‘unless it is clear that the applicant will not be admitted or readmitted to that country’). (See also recital 53 in the preamble to the Regulation). Whether the planned review of the ‘safe third country’ rules in the Regulation in 2025 affects these issues remains to be seen.

Finally, the Court’s judgment understandably only dealt with situations where the refusal of readmission is certain at the time of considering the possible application of the ‘safe third country’ rule. It did not deal with the scenario where refusal was merely a possibility – as that was hypothetical on the facts. How should such cases be dealt with?

The starting point is that the current Directive (recital 44 in the preamble) implies the possibility of an inadmissibility decision in that scenario: Member States do not have to examine the merits of an application where ‘there are grounds for considering that the applicant will be admitted or readmitted to that [‘safe’ third] country’. The inadmissibility rule in the new Regulation points in the same direction: it can’t be applied if ‘it is clear that’ readmission is not going to happen (see also recital 43 in the preamble to the Regulation). But this implies that the inadmissibility rule can be applied if it is not clear that readmission is ruled out. On the other hand, the recent judgment points to the importance of deciding on applications as soon as possible, without unjustifiable postponements, while ensuring the application of the Charter right to asylum. How best to reconcile the possibility of the adoption of inadmissibility decisions with the need to take decisions speedily and afford an effective prospect of having an asylum application decided on the merits, if there is uncertainty about whether readmission will actually take place?

The 2024 procedures Regulation already expressly provides (in recital 53 in the preamble) that access to the Member State’s asylum system must be offered again if readmission does not take place:

…if the applicant is eventually not admitted or readmitted to the third country after the application has been rejected as inadmissible, the applicant should again have access to the procedure for international protection in accordance with this Regulation

How should that principle apply in concrete terms? The best approach is to infer an obligation upon the Member State to act with due diligence to attempt to secure the readmission, starting as soon as it takes the inadmissibility decision (after any appeals, or the prospect of appeals, against that decision are spent). From that point, the Member State should have a reasonable time frame to request readmission from the non-EU country concerned. By analogy with the procedures Regulation time limit for deciding on inadmissibility, two months should be sufficient. If the non-EU country refuses readmission, or does not reply within the time limit set out in the readmission arrangements in force, or accepts readmission in principle but does not carry it out in practice within the relevant time limit, then the Member State’s obligation to consider the application on the merits where readmission does not take place must be triggered. Procedurally either the inadmissibility decision must be considered as withdrawn from that point (it makes sense to consider it as only provisional until readmission is actually carried out), or a repeat application must be considered – the requirement of new developments being satisfied by the inability, subsequent to the inadmissibility decision, to ensure readmission in practice within a reasonable time frame. The inadmissibility decision was based on the presumption that readmission would be secured within a reasonable period; that presumption has since been rebutted.

The final word on the impact of last week’s judgment should go to the asylum NGOs who brought the case:

…the CJEU’s judgment is of major importance because it overturns an arbitrary and abusive practice of the Greek authorities which has been in force for many years, of wholesale rejecting asylum applications as inadmissible under the safe third country principle, and puts an end to the denial of the rights of thousands of applicants, by prohibiting their rejection when there is no possibility of their readmission to Türkiye and dictating their individual examination in accordance with the Directive.

Is this the end of football’s transfer system? An immediate reaction to the Court’s ruling in Diarra (C-650/22).

 

Stephen Weatherill, Somerville College and Faculty of Law, Oxford University

Photo credit: Addesolen, via Wikimedia Commons

 

Introduction

 

‘Is this the end of football’s transfer system?’ So shrieked the media in December 1995 when the Court of Justice decided the Bosman case (C-415/93). And the same question has now hit the headlines after the Court’s ruling on 4 October 2024 in Diarra (C-650/22). The answer was ‘no’ back in 1995. And the answer is still ‘no’ today.

 

The transfer system which enmeshed Diarra and led to the litigation is exposed by the Court’s judgment on 4 October as grossly deficient and incompatible with EU law. It will need to be changed. But a system of sorts, specific to football, can survive this judgment. And it is unlikely anything will happen quickly. It took several years after the Bosman ruling for the system to be revised, and Bosman was clearer on what needed to be changed – the exclusion of out-of-contract players - than Diarra is. But there will need to be change. Most of all, the system will need to be made less restrictive and the consequences of unilateral breach of contract will need to be made more predictable.

 

The litigation

 

Diarra played for Lokomotiv Moscow. A dispute arose. The club terminated his contract for reasons related to his conduct and brought a claim for compensation before the Dispute Resolution Chamber under FIFA’s Regulations on the Status and Transfer of Players. The player counterclaimed. In the meantime he tried to find a new club. Diarra’s claim which reached the Court in Luxembourg via a reference from the cour d’appel de Mons in Belgium concerns his failure to find a new club. Under the FIFA Regulations it was possible that the dispute with his old club might, once finally resolved, have resulted in any new club being liable alongside Diarra to pay compensation to Lokomotiv Moscow under the FIFA Regulations. So, Diarra claimed, the transfer system obstructed his chances of finding new employment. This, he also claimed, was not hypothetical. His evidence included a concrete (and remarkably convenient) offer from Charleroi which was stated to be subject to that club not being liable under the FIFA Regulations – a proviso which FIFA refused to accept. So Diarra’s case was that he had been treated by FIFA in breach of EU law – that the transfer system operated as an obstacle to his free movement in violation of Article 45 TFEU and as an anti-competitive practice incompatible with Article 101 TFEU.

 

 

The structure of the ruling

 

Much of the Diarra ruling is entirely familiar and unsurprising to an EU internal market lawyer.

 

An obstacle to inter-State trade within the meaning of Article 45 TFEU (on the free movement of workers) was found. FIFA’s rules, the Court concluded, were liable to obstruct football players resident or working in their Member State of origin who wish to work instead for a new club established in the territory of another Member State by unilaterally terminating their employment contract.

 

Similarly the presence of the key ingredients of Article 101 TFEU (on competition law cartels) were quickly ticked off – FIFA falls within the concept of an undertaking or association of undertakings, its rules were a 'decision by an association of undertakings', and the matter affected trade between Member States.

 

The sports-specific elements in the ruling are also no surprise. The Court follows the model of its December 2023 ruling in the European Superleague case (Case C-333/21), which is cited on 35 different occasions, and this is plainly now the established model for the several cases concerning the regulation of sport which are pending before it. So the Court repeats its (silly and plain wrong) observation that certain specific rules such as those relating to the exclusion of foreign players from the composition of teams participating in competitions between national teams or to the setting of the ranking criteria used to select athletes participating in competitions which were adopted exclusively for reasons of a non-economic nature and relate to matters relating solely to sport as such must be regarded as being unrelated to any economic activity, and so untouched by EU law. But no harm is done because in Diarra the Court quickly dismisses the notion that the transfer system is ‘unrelated to any economic activity’, just as in the Superleague case it had correctly refused to accept that UEFA’s rules on prior approval of new competitions were extraneous to economic activity – and just as it should also appreciate that rules on the composition of national teams and ranking criteria have direct economic impact. EU law applies – the key issue is not using some spurious ‘non-economic’ label to exclude EU law but rather how EU law applies to practices with both sporting and economic motivations and effects.

 

Paragraphs 124-133 address ‘The concept of conduct having as its “object” or “effect” the harm to competition’. Nothing here will surprise anyone familiar with Superleague. If conduct is found to have an anti-competitive object, it is not necessary to examine its effect on competition; the concept of anticompetitive 'object' must be interpreted strictly; it catches only practices which reveal a sufficient degree of harm to competition for it to be possible to consider that an examination of their effects is not necessary (and some such practices are itemised); subjective intention is not decisive; one must examine the content, the economic and legal context and the aims. Familiar fare from Superleague. Moreover, and confirming the revolutionary turn taken by the Court in Superleague, a practice may be excluded from the scope of Article 101(1) where justified by the pursuit of, and necessary to achieve, legitimate objectives in the public interest – but only where the practice exerts an anti-competitive effect, not where it pursues an anti-competitive object. In the latter case only Article 101(3) may save the practice (paras 149-152). The decision in Meca-Medina (Case C-519/04P), which envisaged (in short) a public interest/ legitimate objective exclusion from the scope of Article 101(1) irrespective of whether the challenged practice was a restriction on competition by object or by effect, is banished – so much so that the case is not even cited in Diarra.

 

 

The incompatibility of the challenged transfer system with EU law

 

The flaws in the transfer system, on paper and in practice, are brutally listed by the Court. This leads it to the finding that they violate both Article 45 and Article 101.

 

The Court does not deny that FIFA has a legitimate role a regulator. In interpreting Article 45 it notes that the rules may be appropriate as a means to achieve the objective of ensuring the regularity of club football competitions and by contributing to maintaining a certain degree of stability in the membership of the football clubs likely to participate in those competitions (para 103). In similar vein, in examining Article 101, it refers to the conditions under which professional football clubs may compose the teams participating in such competitions and those in which the players themselves may take part in them (para 143). It may be legitimate for FIFA to seek to ensure the stability of the composition of the squads of players during a given season, for example by prohibiting – as the FIFA Regulations do - the unilateral termination of employment contracts during the season (para 144).

 

But the Court’s detailed assessment is damning.

 

Paragraphs 103-113 deal with compliance with the principle of proportionality in the interpretation of Article 45. The Court does not mince its words. The rules appear to go in several respects beyond, and in some cases far beyond, what is necessary to attain their objectives. They apply to players who have a relatively short career and therefore the impact with typically be great. Compensation is payable by the player in the event of unilateral termination of the employment contract 'without just cause', but that expression is not precisely defined in the Regulations. Some criteria are, as the Court (rather gleefully, I suspect) notes is admitted even in the official FIFA commentary, almost never applied in practice, whereas others (such as the specificity of sport) have no precise definition. This leads to discretionary and unpredictable implementation. Absence of the necessary legal certainty is a central problem. Other criteria, though more objective and verifiable, seem to go far beyond what is necessary. This is especially so in relation to taking into account remuneration and costs in calculating compensation due. The criteria governing compensation – the Court notes drily – seem more attuned to preserving the financial interests of clubs than to ensure the effective organisation of sporting competitions. The ‘sporting sanction’ imposed on the new club, which operates on the basis of a presumption of incitement, appears far removed from the demands of proportionality. The same is true of the impact on the player.

 

It is a sorry list of inadequacies. And it leads the Court to conclude that the proper conduct of sporting competitions cannot provide a justification. The current transfer system violates Article 45.

 

The examination conducted pursuant to Article 101 has much in common with that presented in connection with Article 45. The analysis begins at paragraph 134. The Court notes – in line with its earlier remarks in the light of Article 45 - that the rules are general and imprecise and that they are subject to discretionary implementation which is unpredictable and difficult to control. So too they appear to allow compensation to be set at a very high and dissuasive level.

 

Paragraph 138 draws on the Opinion of Advocate General Szpunar, which itself was a savage indictment of FIFA’s rules. It declares that the transfer system acts as a general and severe restriction on competition between professional football clubs in the market for recruitment of players. The system amounts to a ‘no-poaching’ agreement between clubs which results in the artificial partitioning of national and local markets, which – crucially - is to the benefit of football clubs (para 145). This acts as a general, absolute and permanent prohibition on the unilateral recruitment of players who are already engaged, which the Court treats as a clear restriction on competition between clubs (para 146).

 

This, then, is as ruthless as it is damning. As already in its treatment of Article 45 the Court portrays the transfer system in its current iteration as a means to improve the economic position of clubs at the expense of workers. Given the negligible input into its shaping allowed to workers’ representatives, this comes as little surprise. 

 

The conclusion is therefore that the system counts as a restriction of competition by object. As such, as explained at paragraphs 149-152 and as decided in Superleague, it cannot be saved by reference to the general public interest/ legitimate objective test applied to sport in Meca-Medina, although the Court cites only Superleague and Em akaunt BG (C-438/22), not Meca-Medina. That in turn means that only Article 101(3) can save FIFA’s rules. But the Court’s treatment of Article 101(3) at paragraphs 153-157 is as brusque as it is brisk. The ultimate decision belongs with the national court, but the Court gives a strong steer that the discretionary and/or disproportionate nature of the rules and their severe restriction on cross-border competition between clubs means they cannot be regarded as indispensable or necessary to achieve economic benefit, even were any such benefit shown.

 

 

The legitimate role of FIFA

 

The Court’s ruling treats the transfer system as rotten. But there is plenty in the Court’s judgment for FIFA to be pleased about.

 

The Court’s interpretation of both Article 45 and Article 101 allows for recognition of a legitimate regulatory role performed by FIFA in adopting common rules to regulate sport, and provide some detail on how far this may reach. Paragraphs 100-103, dealing with Article 45 and paragraphs 143-144, dealing with Article 101, contain a receptivity to FIFA’s regulatory aspirations on which FIFA (and other governing bodies) are likely to rely in framing defence of their practices in future.

 

The objective of ensuring the regularity of sporting competitions constitutes a legitimate objective in the public interest which may be pursued by a governing body. Making the organisation and conduct of international competitions subject to common rules intended to guarantee the homogeneity and coordination of those competitions within an overall annual or seasonal calendar is recognised as legitimate. So too the protection of the essential role played by equal opportunities and sporting merit in the conduct of competitions organised at both European and national level; and ensuring teams compete against each other under homogeneous regulatory and technical conditions.

 

These paragraphs of the judgment also reflect specifically on how these common rules shall affect players. The Court accepts that rules are needed to regulate the composition of teams participating in competitions; that there may be rules relating to the time limits for transfers of players during the competition; that there may be rules intended to ensure the maintenance of a certain degree of stability in the squads of clubs, which also entails concern for the continuity of related contracts; rules setting deadlines for player transfers in order to avoid transfers at a late stage of the season which would damage the overall integrity of the competition. Paragraph 144 seems open to prohibiting - as the current Regulations do -the unilateral termination of employment contracts during the season or even in a given year. These interventions, the Court accepts, serve as means to contribute to the pursuit of the legitimate objective of ensuring the regularity of club football competitions.

 

So this is to recognise the need for regulation of the sport generally and regulation of the place of players within it in particular. It is to recognise the place of a governing body in doing so. Presumably these activities, then, fall within the scope of the legitimate regulatory role performed by a governing body, and they are not a practice which has the object of restricting competition. This echoes Superleague. In that ruling UEFA’s detailed procedures on prior approval were condemned as non-transparent, lacking objective criteria and discriminatory – much as FIFA’s transfer system fell apart under scrutiny in Diarra. But in Superleague the Court did not direct that the market for supply of sporting competitions shall become a free-for-all. Quite the reverse. It accepted that a prior authorisation system may be used to refuse a competition which is not based on sporting merit. It seems that the object of requiring that new competitions be open and based on sporting merit is not to restrict competition but rather ‘the pursuit of legitimate objectives, such as ensuring observance of the principles, values and rules of the game underpinning professional football’ (ESL para 176, and see analyses here and here). Superleague provides opportunities for UEFA to re-define its regulatory purposes in defence of the values of the game – provided it meets the required standards of transparency, objectivity and non-discrimination. Similarly Diarra challenges FIFA to pin down with more care how and why rules governing the consequences of unilateral termination of contract are necessary to protect the integrity of sporting competition, and to devise rules that genuinely do so.

 

 

Revising the rules

 

It seems clear that FIFA’s rules on transfers must be made less restrictive and more predictable than they are now. The difference between the status of a player out of contract and a player still in contract will need to be reduced, but I do not think it will need to be eliminated.  The Court leaves room for FIFA to adopt common rules which are necessary to sustain the integrity and regularity of sporting competition. It seems clear from paragraphs 100-103, dealing with Article 45, and paragraphs 143-144, dealing with Article 101, that the Court is not condemning the very idea that sport needs common rules, operating independently of local contract and labour law, which are designed to protect the sport’s functioning and which may have an impact on a player who commits a unilateral breach of contract. The transfer window will survive, even though it plainly exerts some deterrent effect on clubs’ willingness to act in the market for players. I think it remains open to FIFA to devise a system that will maintain some degree of control over the eligibility of the player who commits a unilateral breach of contract. I think FIFA could provide that a player will not be able to quit one club and insist on being available to play for a new club immediately, even if local law permitted that. How long would the player have to wait? – well, that remains to be decided as the FIFA rules come to be revised. In this sense footballers will still not be treated in exactly the same way as plumbers, sausage-makers and University teachers – they will be subject to special rules applied within their industry in addition to applicable local contract and labour law. In some parts of the judgment, especially paragraph 145, the Court seems to assume that clubs can be adequately protected from the harm flowing from a breach of contract through the application of contract and employment law, and that may be so, but there is also the wider interest in the integrity of the sporting competition to take into account. It would be damaging to sporting integrity if the local law entitled the player immediately to take up employment with a new club. It would be especially damaging in a transnational competition if different laws applied in different states, as they doubtless would. So there must be room also for FIFA, as the transnational regulator of the sport, to address the phenomenon of unilateral breach of contract. But – how? There is a tension – when does action taken against a player who has committed a unilateral breach of contract cross the line from a (lawful) scheme designed to protect the integrity of sporting competition to an (unlawful) attempt to use regulatory power to extract advantage to the clubs at the expense of the player? FIFA’s job will be to show when and why such rules are needed given the special demands of sporting competition. None of this is easy, and it will take time, but paragraphs 100-103 and 143-144 of Diarra are a goldmine for FIFA.

 

A major objection which runs through the ruling is directed at the discretionary case-by-case evaluation of the consequences of unilateral breach of contract. The Court in Diarra notes that a system, if shown to be justified and proportionate, may be supported by sanctions but only on condition that they are set according to transparent, objective, non-discriminatory and proportionate criteria and also they shall be subject to review (para 111), which is a requirement which has close thematic links with the Court’s treatment of UEFA’s rules on prior approval in Superleague. FIFA’s rules on transfers need to be made clearer. Probably it is required that it be possible to predict in advance what would be the consequence of a unilateral breach of contract. That, though, will be tricky if the individual circumstances of each case need to be taken into account, as is suggested by the Court in paragraphs 110, 111, 112, and 137. I am not suggesting FIFA’s re-design of its rules will be easy.

 

The Diarra ruling does not require that a renegotiation of the transfer system shall involve a formal role for player unions, but a strong theme in the ruling is that FIFA has presided over a system which is of huge advantage to some stakeholders and operates to the detriment of others. The Court is not shy of commenting on how the system works very well for clubs as employers (paras 107, 145). Re-negotiation of the transfer system through a process which gives thorough and sincere respect to the voice of the workers would increase the likelihood that a revised version will be compatible with EU law. More generally the Court will have performed a great service if this ruling triggers a deeper reform of governance in sport so that all affected interests – players, even fans - enjoy a louder voice in decision-making than occurs right now. Perhaps FIFA will do this; perhaps, if not, the EU will be tempted to adopt legislation mandating improvement in the standard of governance in sport in general or football in particular. I am aware how naïve that may sound.

 

Diarra summarised? Sport is special. Just look at paragraphs 100-103 and 143-144. But sport is not as special as special as the governing body claims, and the particular practices at stake have been found to violate EU law, and will require reform in the shadow of EU law. In that Diarra is in perfect alignment with Bosman and Superleague.