The Primacy of EU Law in Bulgaria after CJEU’s Judgment in Case C‑56/25: A Thorny Path Ahead

 

Dr Radosveta Vassileva, Adjunct Senior Research Fellow, UCD Sutherland School of Law

Photo credit: Raggatt2000, via wikimedia commons

On 12 February 2026, the CJEU delivered its judgment in Case C-56/25 which concerns a preliminary reference from Bulgaria raising a question about the application of the principle of primacy of EU law in view of a provision of the Procedural Rules of Bulgaria’s Constitutional Court. The CJEU judgment was rendered without an Opinion by the Advocate General, indicating that, in the eyes of the CJEU, the question was neither new, nor difficult. Moreover, the CJEU quotes its settled case law on the primacy of EU law.

However, a closer look at the Bulgarian legislation, which gave rise to the question by the Bulgarian court, especially against the broader country-specific context, reveals that the judgment seems to foreshadow inevitable conflicts between the Bulgarian and the EU legal orders in the future. From a Bulgarian perspective, the place of EU law in the hierarchy of norms is far from being conclusively determined. In parallel, the judgment demonstrates some of the flaws of a controversial constitutional reform carried out in Bulgaria in 2023, whose grand innovation was the ‘individual constitutional complaint’, which led to the hierarchy of norms dilemma of the referring court.

‘The chicken-or-the egg’ problem of the Bulgarian court

The preliminary reference was made in 2025 by the Sofia City Court, acting as a first instance in criminal proceedings, which seemingly faced a ‘chicken-or-egg’ dilemma.

The Sofia City Court was concerned that a provision of national law relevant to the qualification of the criminal offence and, respectively, the punishment in the criminal proceedings before it violated both Bulgaria’s Constitution and EU law – namely, Article 4 of Council Framework Decision 2004/757/JHA of 25 October 2004 laying down minimum provisions on the constituent elements of criminal acts and penalties in the field of illicit drug trafficking.

The Sofia City Court deemed that it was more appropriate to ask the Constitutional Court to rule on the constitutionality of the contested provision of national criminal law before it made a preliminary reference to the CJEU about the compatibility of Bulgarian criminal law with the aforementioned council framework decision. However, in its eyes, there was a catch – Article 18(3) of the Procedural Rules of Bulgaria’s Constitutional Court requires that a request to it

…must contain a reasoned assessment of the applicable law, including of the consequences of the application of EU law where the contested provision or act comes within its scope.

In this context, the Sofia City Court asked the CJEU the following:

Are Article 267 TFEU, Article 94(b) of the Rules of Procedure of the Court … and the principle of the primacy of EU law … to be interpreted as meaning that, where a national court has doubts as to the compatibility of a provision of national law with EU law, and is at the same time convinced that that provision of law is [contrary to the national Constitution], it is obliged or entitled, before submitting its request for a preliminary ruling, to establish whether that provision of national law is indeed applicable in the main proceedings by making an application to the… Constitutional Court… for a declaration as to its unconstitutionality?(para 27 of judgment).

Contextual background: ‘the individual constitutional complaint’ as the grand innovation of the 2023 constitutional reform

The significance of the question by the Sofia City Court can be better appreciated against the broader context of the Bulgarian constitutional reform of 2023, which, regrettably, is neither explained in the text of the preliminary reference itself, nor in the CJEU judgment. This reform, however, made Article 18(3) of the Procedural Rules of Bulgaria’s Constitutional Court, which is at the heart of the hierarchy of norms dilemma of the Sofia City Court, a legal irritant from an EU law perspective.

Before the 2023 reform, the opportunities to request a review by the Constitutional Court were ‘significantly limited and depend[ed] on the political climate in the country’ (see Radosveta Vassileva, ‘A Perfect Storm: The Extraordinary Constitutional Attack against the Istanbul Convention in Bulgaria’ (2022) 1 Osteuropa Recht 84). According to the previous wording of Article 150(1) of the Bulgarian Constitution, the Constitutional Court could only be approached by one-fifth of the Members of Parliament, the President, the Council of Ministers, the Supreme Court of Cassation, the Supreme Administrative Court, or the General Prosecutor. In some limited cases, it could be approached by the Ombudsman or the Supreme Bar Council (see the previous wording of Articles 150(3) and 150(4) of the Constitution).

To this end, one of the grand ‘innovations’ of the 2023 constitutional reform was the introduction of the ‘individual constitutional complaint’. The name given by the Bulgarian legislator to this new procedure, nevertheless, seems to be a misnomer because a private citizen can neither directly ask the Constitutional Court for constitutional review nor appeal court decisions there on constitutional grounds. The new mechanism for constitutional review, in essence, mimics the preliminary reference procedure before the CJEU. In fact, as argued in the explanatory memorandum accompanying the reform bill of 2023, ‘the situation in Bulgaria [was] paradoxical because any court (judge) could [make a preliminary reference to the CJEU], but could not turn directly to the national constitutional jurisdiction’ (p 17 of the aforementioned explanatory memorandum).

Following the 2023 reform, Article 150(2) of the Constitution stipulates:

Any court, at the request of a party to the case or on its own initiative, may refer to the Constitutional Court a request to establish an incompatibility between a law applicable to the specific case and the Constitution. The proceedings in the case shall continue, and the court, whose decision is final, hands down its judgment after the proceedings pending before the Constitutional Court have been concluded.

It is important to note that the usage of ‘may’ in the wording implies that the judge or the judicial panel in question has discretion over whether to honour such request by a party in the proceedings. In practice, it is common for courts to choose not to honour requests for preliminary references to the CJEU and/or requests to approach the Constitutional Court under the new wording of Article 150(2) coming from the parties. In other words, there are cases in which judges turn a blind eye to the likely incompatibility between Bulgarian legislation, on the one hand, and EU law and the Constitution, on the other. There are also cases in which judges can be seen in a more activist role, making references when the parties concerned did not argue any such incompatibility themselves.

The ‘ticking procedural time bomb’ left by the Bulgarian legislator

Bulgaria’s Constitution, adopted in 1991, clearly specifies the hierarchy of norms. Its Article 5(1) explicitly states that it is the ‘supreme law’, while its Article 5(3) stipulates:

International treaties, ratified in accordance with the constitutional procedure, promulgated, and having entered into force for the Republic of Bulgaria, are part of the country’s domestic law. They take precedence over those provisions of domestic legislation that contradict them.

Hence, the Constitution takes precedence over any international treaty and Bulgarian law while international treaties take precedence over Bulgarian laws that contradict them. Bulgaria has, of course, ratified its EU Accession Treaty of 2005.

Furthermore, the status and place of CJEU case law in the hierarchy of norms is subject to debate from a national perspective. CJEU’s case law is not defined as binding in Bulgarian legislation, unlike the case law of the Constitutional Court or the decrees and decisions on interpretation by the country’s supreme courts. However, the Code of Civil Procedure allows cassation on the grounds of violation of the case law of the CJEU (see Article 280(1), point 2).

It is notable that throughout the years, Bulgaria’s Constitutional Court has tried to avoid direct confrontation with the CJEU. In its own case law, however, it has argued that it is up to each entity requesting constitutional review to determine the applicable law, including EU law, to the proceedings before it. It has also stressed that the determination of the applicable law should always precede requests for constitutional review (see, for instance, the detailed reasoning in Ruling 2 of 24 February 2022 by the Constitutional Court).

In this light, the ‘individual constitutional complaint’ can be seen as a ‘ticking procedural time bomb’ since the reform of 2023 left Article 5 of the Constitution untouched, while expanding the jurisdiction of the Constitutional Court itself, which can now receive requests directly from judges. In practice, this ‘individual complaint’ means a chance for more frequent encounters and, respectively, potential clashes between Bulgarian constitutional law and EU law in ordinary Bulgarian courts.

The legislative choice not to tamper with Article 5, in turn, was informed by the restrictions on amendments imposed by the Constitution itself – any amendments to Article 5 require the convocation of a ‘grand national assembly’ as opposed to a ‘regular national assembly’ (on the difference between them, see here; see also Article 158 of the Constitution). Grand national assemblies are notoriously difficult to convene. The Constitution of 1991 was adopted by such an assembly.

CJEU’s judgment as an exercise in judicial diplomacy

In the judgment in Case C-56/25, the CJEU held that Article 267 TFEU, the principle of the primacy of EU law and Article 94(b) of the Rules of Procedure of the Court of Justice do not preclude procedural rules, such as Article 18(3) of the Procedural Rules of Bulgaria’s Constitutional Court (para 61 of judgment). Leading authorities on the rule of law like Professor Laurent Pech have already called the judgment ‘unusually generous’, especially in view of challenges to CJEU’s jurisdiction from other EU member states, such as Poland.

Yet, when one reads the reasoning of the judgment in light of the particularities of the Bulgarian constitutional order, it seems that the ‘generosity’ came with a few price tags – namely, demands for a voluntary surrender of jurisdiction by the Bulgarian Constitutional Court. These requests, nevertheless, may not bear the fruit intended by the CJEU, precisely because they seem oblivious of context, including national legislation.

One of the reasons why the CJEU was not troubled by the requirements of Article 18(3) of the Procedural Rules of Bulgaria’s Constitutional Court was the Constitutional Court’s alleged obligation to make preliminary references itself. Namely, the CJEU stated: ‘…if a constitutional court is seised of a request for a review of the constitutionality of a provision of national law coming within the scope of EU law, that court is in principle obliged to make a reference to the Court of Justice for a preliminary ruling, in accordance with the third paragraph of Article 267 TFEU…’ (para 57 of judgment). Bulgaria’s Constitutional Court, however, has no record of making preliminary references to the CJEU and it is highly unlikely that it will endeavour to make such references in the future if an amendment to Article 5 of the Constitution is not made. That is because the Constitutional Court sees its own role solely as the guardian of the Bulgarian Constitution. First, the Constitutional Court has the monopoly on authoritative, binding interpretation of the Constitution (Article 149(1), point 1 of the Constitution). Second, in case of conflict between the Constitution and EU law, the Constitution prevails according to its own text. Third, as highlighted above, the Constitutional Court has stated in its own settled case law that the establishment of the applicable EU law is in the prerogative of the court examining the dispute on the merits. On the one hand, this approach protects the Constitutional Court’s jurisdiction. On the other hand, it prevents direct confrontation with the CJEU – it is up to the ordinary courts to identify and raise any questions about the compatibility between EU law and constitutional law and to consider how such conflict affects the outcome of the case.

It is also interesting that the CJEU reminds that a ‘national court is required, in order to ensure the full effectiveness of the rules of EU law, to disregard, in the dispute before it, the rulings of a national constitutional court which refuses to give effect to a judgment given by way of a preliminary ruling by the Court of Justice…’ (para 59 of judgment). From an EU perspective, this approach has numerous merits, such as empowering judges in lower courts to disregard controversial decisions by constitutional courts, especially when they are unlawfully composed. Nevertheless, in the Bulgarian context, beyond the invitation to disregard key express provisions of the Bulgarian Constitution, this conclusion ignores important practical aspects. For example, pursuant to Article 280(1), point 2 of the Bulgarian Code of Civil Procedure, non-compliance with case law by the Constitutional Court serves as grounds for cassation. While the Code of Criminal Procedure does not contain such explicit reference to constitutional case law, its Article 348 allows cassation for ‘violations of the law’ broadly conceived.

A thorny path ahead for the primacy of EU law

Overall, the path ahead for the primacy of EU law in Bulgaria appears rather thorny. CJEU’s judgment in Case C-56/25 may be seen as a precursor to a series of legal and political challenges. It foreshadows inevitable clashes between EU law and Bulgarian constitutional law in ordinary courts, which are facilitated by the shortcomings of the 2023 constitutional reform.

It is also doubtful to what extent the approach, which the CJEU endorses in the judgment, will contribute to affirming the primacy of EU law in Bulgaria. Ensuring such primacy on paper requires important constitutional amendments by a grand national assembly as well as a large-scale legislative reform. It also necessitates a change of mentalité in practice on behalf of both Bulgarian judges and the CJEU itself.

First, CJEU case law remains terra incognita for many jurists because of their educational and professional background. Moreover, for more than a decade following Bulgaria’s EU accession in 2007, preliminary references from the country came from a very small circle of judges (see Aleksander Kornezov, ‘Ten years of preliminary references – a critical review and appraisal’(2017) Evropeiski praven pregled).

Second, between 2018-2022, a vast number of preliminary references were made by Bulgaria’s specialised criminal court, which had all features of an extraordinary repressive tribunal, and used this EU law mechanism primarily as a tool to affirm its legitimacy in light of criticism of its abusive practices. In 2022, this court was dissolved for good for undermining the rule of law upon the proposal by a short-lived opposition government and after years of public discontent. As a result, doubts about the status of its case law and the EU judgments resulting from its references, which are intertwined with it, are on the rise. So is the bitter feeling that the CJEU empowered ‘non-judges’ the way it did vis-à-vis Poland (see here).

Third, the CJEU has acted as a Pontius Pilate, washing its hands of responsibility, on key matters pertaining to Bulgaria’s rule of law backsliding and human rights abuses. It has a record of prioritising formalism and refusing to address the issues at their core, thus discouraging judges from raising politically sensitive questions (on judicial independence, see here; on standards of proof in pretrial detention, see here).

To what use is then primacy for primacy’s sake? This is a question which many judges from Bulgaria’s ordinary courts may ask in the aftermath of CJEU’s judgment in Case C-56/25.

 

The author would like to thank Prof Laurent Pech for his helpful comments and suggestions on an earlier draft.

The Chios Incident: Echoes of Pylos Humanitarian Disaster and Greece's Criminalization of Solidarity

 

Georgios Athanasiou, PhD Researcher, University of Antwerp

Photo credit: Julian Lupyan, via Wikimedia Commons

The Chios migrant boat shipwreck of 3 February 2026 exemplifies the acute tensions between Greece’s increasingly securitized border management and its obligations under EU law, the ECHR, and international maritime conventions such as the 1979 Search and Rescue (SAR) Convention. More specifically, off the coast of Chios island, a Hellenic Coast Guard patrol vessel collided with an inflatable boat carrying approximately 39 Afghan nationals, resulting in 15 deaths and 24 injuries, including 11 minors, and cases of miscarriage. All of the deaths were attributed to severe head trauma rather than drowning, per up-to-date autopsy reports, with survivors claim that the coast guard did not offer any prior warning or communication before ramming the migrant boat, contrary to official claims of the migrants’ speedboat initiating contact. Interestingly, the on boat cameras of the patrol vessel had been deactivated.   

This event parallels the 14 June 2023 Pylos shipwreck, Europe’s deadliest maritime migration tragedy, where over 500 lives (mainly Syrian, Pakistani, and Egyptian) were lost after the overcrowded trawler Adriana capsized, allegedly due to Coast Guard towing maneuvers following delayed rescue operation, despite prior distress alerts. The ongoing criminal proceedings in Greece have charged 17 Coast Guard personnel, including the rescue vessel captain, with felony offenses such as endangering lives and contributing to the shipwreck,. This development appears to be part of a systematic attempt to portray the eastern Mediterranean migration route as inherently life-threatening for asylum seekers, thus reflecting a pervasive securitization narrative guiding border policies of the Greek government that overshadows State accountability.

Legal Parallels and Accountability Gaps

From a legal aspect, both incidents implicate Greece’s positive obligations under Article 2 ECHR (right to life), requiring States to safeguard lives within their jurisdiction, including during maritime interceptions, and conduct effective, independent investigations into fatalities. The ECtHR has repeatedly held Greece accountable in analogous cases: in Safi and Others v. Greece, for inadequate protection and probing of a sunk migrant boat; Alkhatib and Others v. Greece, for excessive lethal force lacking “absolute necessity” and deficient regulatory frameworks for Coast Guard firearms use. Such repeated failures in border management operations seem to formulate a consistent pattern of action of the Greek authorities in handling migrant routes, in an attempt to not allow migrants to enter Greek territory/territorial waters. In this sense, although Article 3 ECHR (prohibition of inhuman/degrading treatment) further prohibits collective expulsions or pushbacks, this practice has been deemed systematic by Greek authorities in A.R.E. v. Greece (also see, here).

Under EU law, the Qualification and Asylum Procedures Directives respectively mandate the upholding of the principle of non-refoulement and individual assessments of asylum applications, while Article 4 of Protocol 4 of the ECHR and Article 19 CFR bar collective expulsions. Meanwhile, the SOLAS and SAR Conventions impose duties to render assistance “without delay” to persons in distress, disembarking them to a place of safety, irrespective of nationality or the existence of a right to enter the country. Hence, any form of interception framed as SAR mission cannot justify pushbacks or endangering the lives of migrants.

Greece’s Restrictive Policies and Criminalization of Solidarity

Domestically, this incident aligns with broader migration policy tendencies, as Greece has instrumentalized criminal law in an attempt to restrict migration, rendering irregular entry, stay and exit of the country a felony punishable with up to 5 years of imprisonment coupled with a minimum fine of €5,000 (Law 5226/2025 Government Gazette Α' 154/8.9.2025). Similarly, rejected asylum seekers face administrative fines up to €10,000, as well as up to five-year sentences or electronic ankle monitoring. Hence, the 2025 deportation law, hailed as Europe’s most stringent, essentially attempts to streamline expulsions of “economic migrants,” given that long-term regularization after 7 years of stay in the country is equally abolished.

This framework cannot be dissociated from Greece’s post-2019 migration hardening: escalated border fortifications (Evros 35 km wall), freezing of asylum applications, and systematic pushbacks exceeding 540 incidents between 2020-2022 (also see, here). Hence, high-seas shipwrecks, such as the Chios and Pylos lethal incidents, epitomize how this apparatus practically overrides positive obligations under the ECHR, as well as international humanitarian and maritime law, subordinating the protection of life at sea to national security imperatives.

The Greek Government defends its approach as prevention of illegal entry, invoking safe third country safeguards, especially for migrants arriving from Turkey, yet these yield no derogation from non-refoulement or collective expulsion bans. Hence, the Government’s approach in migration policies embodies a “fortress mentality,” which, coupled with its recent attempts to criminalize solidarity, further sets in danger the lives of migrants attempting to cross the Eastern Mediterranean route. A prominent example of this criminalization tendency include the recent Lesbos case against 24 rescuers, who were acquitted after years on charges, like espionage and smuggling, that carried up to 20-year sentences. Similarly, Norwegian activist Tommy Olsen faced an arrest warrant in February 2026 for documenting pushbacks via Aegean Boat Report, accused of criminal organization. Finally, a February 2026 migration law amendment (Law 5275/2026, Government Gazette Α’ 17/06-02-2026) makes NGO membership an aggravating factor, escalating misdemeanors (e.g., facilitation of stay) to felonies with fines exceeding €100,000, constituting the largest criminalization of solidarity in the EU.

Analysis

It is apparent that the Chios tragedy, when assessed alongside the Pylos shipwreck, does not constitute an isolated operational failure but rather indicative of a structural recalibration of border governance in the Eastern Mediterranean. The shift from the enforcement of search-and-rescue obligations to human rights violations at the EU’s external borders under SAR cover reveals a normative inversion: life-saving obligations are operationalized through a security prism that treats irregular entry primarily as a threat vector rather than a protection trigger.

At the doctrinal level, Article 2 ECHR imposes both substantive and procedural duties on States. More specifically, from a substantive aspect, States shall refrain from unlawful deprivation of life and adopt preventive operational measures where authorities knew or ought to have known of a real and immediate risk. Meanwhile, procedural obligations mandate to conduct prompt, effective, and independent investigations capable of leading to accountability. In both Chios and Pylos, the central legal question is whether Greek authorities fulfilled the due diligence threshold required during maritime interception. The reported deactivation of onboard cameras in Chios and the delayed rescue response in Pylos indicate the State’s unwillingness to comply with these operational standards.

In a similar vein, the 2026 legislative reform represents an internal consolidation of the securitization paradigm. By reclassifying irregular entry and facilitation-related conduct as felonies and elevating NGO affiliation to an aggravating factor, the Greek legal framework operationalizes criminal law as a migration-management instrument, fully adopting a “Crimmigration” approach, that is the convergence of criminal and immigration enforcement logics, as border management framework.

In this sense, the prosecution of humanitarian actors in the Lesbos case and proceedings against figures associated with monitoring networks reinforce a chilling effect on civil society oversight. When accountability mechanisms (NGO monitoring, documentation of pushbacks) are suppressed, the evidentiary architecture for fundamental rights protection is simultaneously weakened. In practical terms, criminalization of solidarity indirectly facilitates impunity.

Greece’s approach cannot be decoupled from the broader EU externalization strategy. Financial and operational support through Frontex, coupled with political endorsement of deterrence metrics (reduced arrivals as “success indicators”), generates structural incentives that privilege interdiction over protection. In this context, it appears that the Eastern Mediterranean has become a testing ground for this hybrid governance model of the EU’s external borders. This primarily includes operational opacity (restricted access, disabled recording systems), normative elasticity (expansive security justifications), and penal reinforcement (domestic felony frameworks). In other words, the legal tension at stake is not merely compliance with international human rights law but the hierarchy of values underpinning EU border management. If border integrity consistently supersedes the core values of life and human dignity, the doctrinal architecture of human rights law is functionally subordinated to security rationales.

Accordingly, the Chios incident should be analyzed not only as a maritime tragedy but as a constitutional stress test for the EU human rights regime. The decisive issue is whether accountability mechanisms, domestic courts, the ECtHR, EU oversight bodies, will be able to effectively recalibrate operational practice toward a life-preserving baseline or tacitly normalize deterrence-driven fundamental rights erosion.

Conclusion

Greece exemplifies a broader European paradigm: a so‑called “success story” for deterrence‑based migration control, yet in reality a humanitarian catastrophe for those seeking protection. The country’s migration policies mirror a wider EU strategy that prioritizes border fortification over human life. Hence, a rights‑first recalibration is urgently required. This entails independent and transparent investigations into all reported maritime incidents, such as the full public release of the Chios and Pylos footage, and unhindered support for NGOs engaged in SAR operations, paired with the domestic decriminalization of humanitarian assistance to migrants. In the absence of these measures, the prevailing doctrine of “prevention at all costs” will perpetuate watery graves, turning the Mediterranean into an open cemetery and rendering the protection of migrants’ fundamental rights mere eulogies in default.

 

The Aleb judgment on ‘safe third countries’ in asylum law: the CJEU’s answer to EU legislative amendments?

 

 

Tamta Gventsadze, PhD candidate in law, UNITUS

Photo credit: Mstyslav Chernov, via Wikimedia Commons

 

Introduction

On February 5, 2026, the Court of Justice delivered its judgment in Case C-718/24, Aleb concerning the interpretation of Articles 33(2)(c), 38 and 46 of Directive 2013/32 in the context of the ‘safe third country’ concept and the right to an effective remedy, Article 47 of the Charter of Fundamental Rights of the European Union.

This analysis examines the Aleb judgment in light of the ‘safe third country’ concept under Directive 2013/32 (Asylum Procedures Directive, APD) and its replacement by Regulation (EU) 2024/1348 (Asylum Procedures Regulation, APR) from 12 June this year. After mapping the factual background and the Court’s clarification of the cumulative safeguards governing the presumption of safety, it then considers the impact of the 2026 amendments and assesses their implications for judicial control and fundamental rights protection.

I. Facts of the Case

The applicant, NP, is a Syrian national and an unaccompanied minor who lodged an application for international protection in Bulgaria on 2 November 2023. During the interview conducted on 1 December 2023, he stated that he had lived in Aleppo (Syria) and had left two to three months earlier with his brothers because of the war. Before “illegally entering the Bulgarian territory” (para. 19), NP stayed in Türkiye for approximately one month, where his brothers remained with three of his sisters who already lived there.

By decision of 18 June 2024, the Chairperson of the Bulgarian National Refugee Agency rejected the application, refusing to grant him both refugee and humanitarian status. The authority accepted that Syria was affected by internal armed conflict and indiscriminate violence and acknowledged that the applicant was exposed to a real threat to his life or person. Nevertheless, it refused to grant protection on the ground that Türkiye constituted a ‘safe third country’ in which the applicant could seek protection. The decision relied, inter alia, on the fact that the applicant had already lived in Türkiye for about a month without suffering harm, had close family members there, and that Syrian nationals in Türkiye benefited from temporary protection and protection against forced return; finally, “their basic needs [were] satisfied” (para. 20).

The referring court expressed doubts as to the compatibility of this approach with Directive 2013/32, especially considering the absence of both a defined methodology for applying the ‘safe third country’ concept and the lack of explicit procedural guarantees under Bulgarian law allowing the applicant to challenge the existence of a sufficient connection with Türkiye (paras 24-26).

II. The Presumption of Safety under Article 38 of the Directive

The judgment deals with the legal nature and limits of the presumption underlying the ‘safe third country’ concept. The Court begins by recalling that the application of Article 33(2)(c) of the APD (ie, providing that asylum applications are inadmissible where the ‘safe third country’ principle is applied) is conditional upon strict compliance with the requirements of Article 38 thereof (ie the definition of ‘safe third country’ and the conditions related to it). In para. 48 of the judgment, it is expressly stated that the conditions laid down in Article 38 are cumulative, with the result that the inadmissibility ground cannot be applied where any one of those conditions is not satisfied.

This formulation makes clear that the presumption of safety is neither automatic nor self-standing, it is legally constructed and constrained by a series of substantive and procedural safeguards (para. 46). More precisely, Article 38(2) requires Member States to regulate the safe third country concept through national law and that the national rules ensure: (i) there is a sufficient connection between the applicant and the third country so that return there is reasonable; (ii) must define a methodology for applying the concept, which includes either a case-by-case safety assessment or the designation of generally safe countries; (iii) must guarantee an individual examination and allow the applicant to challenge both the safety of the third country in their specific circumstances and the existence of the required connection.

Therefore, the key requirement in the safe third country designation is the existence of a “connection” between the applicant and the third country. The Court emphasizes that Article 38(2)(a) obliges Member States to define in national law criteria enabling authorities to determine whether such a connection exists and whether return to that country is reasonable (paras 51–52). Since the Directive does not define “connection,” Member States retain discretion to specify the criteria, but within EU limits.

Importantly, the Court further reiterates its prior case law that mere transit through a third country cannot, on its own, justify the conclusion that return there is reasonable (para. 54). This statement substantially narrows the presumption. It prevents Member States from relying on minimal factual links and requires a qualitative assessment of the relationship between the applicant and the third country, considering factors such as duration and circumstances of stay and family ties. It is evident that the presumption cannot be based merely on the “transit” criterion. Even where national law relies on the notion of “stay,” national courts must assess, in light of all circumstances, whether that stay genuinely establishes a sufficient connection (para. 55).

The Court also addresses national lists of safe third countries. Member States may, in principle, designate safe third countries by general act. However, such designation does not dispense with the obligation to conduct an individual assessment. National law must provide a methodology for a case-by-case evaluation of both the country’s safety for the applicant and the existence of a sufficient connection (para. 65). The presumption must remain rebuttable, and the applicant must be able to challenge the existence of that connection.

The judgment firmly situates the safe third country concept within the framework of effective judicial protection. Article 38(2)(c) must be read in conjunction with the ‘effective remedy’ rights in both Article 46 of the Directive and Article 47 of the Charter (paras 69-74). Even if national law does not explicitly confer such power, a court hearing an appeal must verify whether a sufficient connection exists.

Accordingly, the Court recalls that Article 46(1) of the Directive guarantees a right to an effective remedy in asylum cases and that Article 46(3) requires a full and ex nunc examination of both facts and law in asylum appeals. This standard applies even in inadmissibility cases and does not necessarily require a substantive assessment of protection needs, but it does require full judicial scrutiny of admissibility conditions. Furthermore, Article 47 of the Charter enshrines the principle of effective judicial protection and is directly applicable, thereby requiring national courts to conduct comprehensive review consistent with EU fundamental rights standards.

Therefore, when reviewing a decision declaring an application inadmissible on ‘safe third country’ grounds, national courts must conduct a full and up-to-date examination of whether the third country is safe for the applicant and whether all cumulative conditions, including the connection requirement, are fulfilled (para. 75). The Court thus subjects the presumption of safety to meaningful judicial scrutiny grounded in Article 47 of the Charter.

Finally, the Court confirms that an application may be declared inadmissible on ‘safe third country’ grounds even where the applicant faces a real risk of serious harm in the country of origin (ie, as distinct from the ‘safe third country’). This confirms that the ‘safe third country’ concept functions as a jurisdiction-allocating mechanism rather than as a substantive denial of risk in the country of origin. Precisely because it allows displacement of protection despite such risk, strict compliance with Article 38 safeguards is imperative.

III. Relevant changes of legislative framework

It is further necessary to underline that while the Court has provided meaningful clarifications regarding safe third country concept and judicial protection in light of corresponding provisions under Directive 2013/32 in a few months’ time this instrument will be replaced by another secondary EU legislation in the form of Regulation (EU) 2024/1348 of the European Parliament and of the Council of 14 May 2024 establishing a common procedure for international protection in the Union and repealing Directive 2013/32/EU (already discussed in great detail here).

Upon a comparative assessment of these two contrasting instruments, several illustrative differences emerge between the provisions concerning safety presumptions, and those of appeals.

Article 33 of the APD addresses inadmissible applications, framing them as an exception to the obligation to examine the substance of an application, meaning that Member States are not required to assess the merits of qualification for international protection where one of the exhaustively listed grounds for inadmissibility applies. The list is limited: Art. 33(2) includes protection granted by another Member State, first country of asylum, ‘safe third country’, subsequent applications without any new elements, and applications lodged by dependents who had consented to be included in another application.

Article 38 of the APR restructures this framework, by separating the decision on admissibility from the decision on the merits. Article 39(3) explicitly governs rejection as unfounded following substantive examination pursuant to the ‘Qualifications’ Regulation (EU) 2024/1347. This structural clarification seems to strengthen the distinction between inadmissibility and unfoundedness, which had already been emphasized by the Court.

In addition, relevant to the current analysis, a notable development concerns the first country of asylum and ‘safe third country’ grounds of inadmissibility. In the APD, Article 33(2)(b) and (c) refer to Articles 35 and 38 without expressly conditioning inadmissibility on the likelihood of admission or readmission. The Regulation introduces an explicit safeguard, with Article 38(1)(a) and (b) essentially stating that inadmissibility may apply unless it is clear that the applicant will not be admitted or readmitted to the third country. Therefore, changes regarding admissibility grounds combine expansion with procedural tightening and partial codification of judicial safeguards.

To further detail the elements regarding the concept of ‘safe third country’, it seems clear that the evolution of the rules on this principle from Article 38 of the APD to Article 59 of the 2024 Regulation, especially as amended in 2026, reflects an apparent shift in structure and scope, even if several aspects remain unchanged.

First off, at the level of safety criteria, both instruments require absence of threats to life or liberty on Convention grounds, absence of serious harm, respect for non-refoulement, and protection against removal contrary to international law. The Regulation clarifies the content of “effective protection” through reference to Article 57 and to Regulation (EU) 2024/1347. As defined by Article 57, effective protection can be attained in a country that has ratified and respects the Geneva Convention on Refugee Status, within any permitted reservations or limitations. However, where a geographical limitation applies, or where the Convention has not been ratified, protection must be assessed against minimum criteria of permission to remain on the territory, access to sufficient means of subsistence to ensure an adequate standard of living, access to healthcare and essential treatment, access to education under general national conditions, and the availability of protection until a durable solution is found. Interestingly, these minimum criteria resemble those of subsidiary protection guarantees, but the context refers to international protection.

Secondly, additional significant changes concern Union and national designation mechanisms introduced by the Regulation. Article 59(2) and (3) of the Regulation, which have no previous equivalent in the Directive, allow partial designation for specific territorial parts or identifiable categories of persons and require reliance on a broad range of sources. Importantly, Art. 59(5)(b) explicitly maintained, that the ‘safe third country’ concept may only be applied where “there is a connection between the applicant and the third country in question on the basis of which it would be reasonable for him or her to go to that country” (the same standard of ‘reasonable’ connection as employed by the Court in Aleb).

However, the 2026 amendments substantially reshape the connection criterion, by removing it as a mandatory condition and introducing two additional bases. The concept may now apply where the applicant transited through the third country (explicitly opposing to what the Court prohibits in Aleb). It may also apply where an agreement or arrangement exists requiring the third country to examine protection claims. As already wittily named elsewhere, this ‘Rwanda clause’ permits transfer without prior connection or transit. Moreover, the removal of automatic suspensive effect in safe third country appeals in the 2026 amendments seem to further intensify a restrictive shift, although there is still a possibility for requesting a judicial suspension.

The most problematic aspect seems this possibility of transferring an individual to a state where they might have never even been to, but the secondary law would permit it solely due to the existence of an agreement (often in non-legally binding form of MoUs) between a EU Member State or the Union and possibly any third country. It is true that an applicant will maintain a possibility to appeal this decision, however, they might have to first ask for a suspensive effective of the appeal, which further complicates an already arduous procedure. In theory, a person might end up being transferred to such a supposedly ‘safe’ third country before a decision is made upon their request to remain, which could create a rather unclear legal consequence; would a person potentially have to be brought back, shall their appeal succeed? It is true that this particular scenario might be extremely rare in practice, but does this rarity allow for its legality?

Nevertheless, despite this expansion in scope, the APR maintains several core safeguards, explicitly stating that admission or readmission must be ensured and that individual assessment remains required (which could become more restricted in practice). Special guarantees also apply to unaccompanied minors, including best interests and prior assurances of protection; and the ‘Rwanda’ clause cannot apply to unaccompanied minors at all.

As confirmed by the Alace judgment, designations of ‘safe countries of origin’ must remain subject to judicial review under Article 47 of the Charter. The Court held that national courts must be able to examine compliance with material designation criteria and to rely on independent sources of information. This reasoning applies by analogy to ‘safe third country’ designations and considering that the Charter has the same legal value as primary EU law, legislative attempts to narrow judicial review cannot override it.

In this context, the Aleb judgment constitutes an additional message. It reaffirms that application of the ‘safe third country’ concept is subject to cumulative conditions and full judicial scrutiny. Even as the legislature broadens the concept and limits suspensive effect, the Court insists on effective judicial protection and strict assessment of safety criteria.

IV. Concluding Assessment

The Aleb judgment does not abolish the presumption of safety inherent in Article 38 of the Directive. However, it subjects the presumption to cumulative substantive conditions, mandatory individualized assessment, and full judicial review.

In more detail, the ECJ allows that the determining authority may apply the concept of ‘safe third country’ on the basis of information from publicly available sources and rely on a national list of safe third countries, if such exists, but this is only provided that national law also defines the methodology applicable for assessing, on a case-by-case basis, according to the particular circumstances of the applicant for international protection, whether the third country concerned satisfies the conditions for being regarded as safe for that applicant and the possibility for that applicant to challenge the existence of a connection, within the meaning of Article 38(2)(a).

Moreover, the Court’s insistence, particularly in paragraphs 48, 54 and 65 of the judgment, on the cumulative nature of the safeguards, the insufficiency of mere transit, and the necessity of a defined methodology, significantly limits the discretionary space of national authorities. The ECJ reinforces the procedural containment of the ‘safe third country’ mechanism, by linking these cumulative requirements to Article 46 of the Directive and Article 47 of the Charter, and by requiring national courts to verify the existence of a connection even where national law is silent in this regard.

Therefore, the Court ties the existence of the ‘safe third country’ presumption to two co-existing elements: clearly defined methodology underlining individual assessment (which includes the existence of “reasonable” connection between the applicant and the safe third country) and a possibility for judicial review of the connection requirement. It seems evident from this judgment, that the safety presumption, be it national or supranational level, would otherwise be invalid. In a way, Aleb strengthens the doctrinal link between inadmissibility decisions and effective judicial protection. The presumption of safety is permitted, per se, but only as a structured, reviewable and rebuttable legal construction embedded within the broader guarantees of EU fundamental rights law.

The combined effect of Alace and Aleb indicates that the Court of Justice does not seem to be prepared for relaxing the standards governing the designation and review of ‘safe third countries’. Nevertheless, it remains to be seen whether and how the Court will respond to the legislative changes, considering that they aim to abolish the mandatory connection element and restrict safeguards when challenging it.

Race, Housing, and the Limits of EU Anti-Discrimination Law: A Commentary on the CJEU’s Judgment in the Danish ‘ghetto law’ case

 

 

Karin de Vries (professor of fundamental rights law, Utrecht University) and Sarah Ganty (JSD candidate, Yale Law School; FNRS Post doc fellow, UCLouvain)

 

Photo credit: Kristoffer Trolle, via Wikimedia Commons

 

The so-called Danish ‘ghetto law’ case has been one of the most closely watched judgments of 2025 before the Court of Justice of the European Union (CJEU), long awaited by lawyers working in EU anti-discrimination law and beyond. Its resonance has extended well beyond legal circles, attracting sustained attention from the media and NGOs.

Much has already been written on the judgment, from different perspectives (for example, here, here, here and here). The focus of this contribution is on how the judgment shapes the definition of racial discrimination in EU law, including the emerging tension between objective and social understandings of ethnic origin, the Court’s application of the concepts of direct and indirect discrimination and its reticence to address the structural racism at the roots of the Danish housing policy. Before proceeding to our analysis we first offer a brief outline of the case and the Court’s reasoning.

Background: the Danish ‘ghetto law’

At the heart of the case lies a practice of state-driven gentrification as part of the Danish housing policy. The Danish Law on public housing seeks to reduce the proportion of public housing in certain designated neighbourhoods, officially labelled “transformation areas”. Under the scheme, authorities may sell buildings to private developers, demolish existing housing, or convert family dwellings into accommodation for young people. These measures may entail the unilateral termination of tenants’ leases, resulting in their forced displacement.

To determine which neighbourhoods are subject to restructuring, an earlier version of the Law on public housing introduced the labels of “ghettos” and “hard ghettos”. These labels have since been replaced by the less openly yet still stigmatising terms “parallel societies and “transformation areas”, without however incurring any substantive changes to the scheme. A neighbourhood classifies as a parallel society when it satisfies at least two out of four socio-economic criteria—relating to unemployment, education levels, criminal convictions, and average gross income—and at least 50 per cent of its residents are “immigrants and their descendants from non-Western countries”. Where such a classification persists for five consecutive years, the area is classified as a “transformation area” and becomes subject to the abovementioned far-reaching restructuring measures. The Law on Public Housing also identifies “vulnerable areas”, neighbourhoods that meet the same socio-economic criteria as parallel societies but do not have a majority of “non-Western” residents. These areas are not eligible for restructuring. Hence, the possibility of restructuring turns explicitly on the “Western”/“non-Western” distinction, making the risk of forced relocation contingent upon the demographic composition of a neighbourhood.

The applicants are residents of designated “transformation areas” in the municipalities of Slagelse and Copenhagen. The applicants from Slagelse challenged the termination of their leases, whereas the applicants from Copenhagen sought invalidation of the ministerial decision approving the development plan for their neighbourhood. Both cases reached the High Court of Eastern Denmark, which referred two preliminary questions to the Court of Justice First, whether the criterion “non-Western immigrants and their descendants” constitutes a distinction based on ethnic origin within the meaning of the Race Equality Directive (RED); and second, whether the Danish legislation gives rise to direct or indirect discrimination.

 

The Judgment

 

The Opinion of the Advocate General (Ćapeta) answered both questions in the affirmative, concluding that the distinction is based on ethnic origin and that the difference in treatment amounts to direct discrimination, although it could also be qualified as indirect discrimination. While the Grand Chamber does not fully adopt the AG’s Opinion, much of its reasoning is reflected in the judgment. The CJEU, first of all, concurs with the AG that a public housing scheme such as that existing in Denmark is capable of falling within the material scope of the RED (§ 66). It then proceeds to examine whether the Danish legislation constitutes direct ethnic discrimination. While leaving it to the referring court to ultimately decide this issue, the Court offers detailed guidelines which allow very little room for a negative answer. Still, in case the referring court should conclude that there has been no direct discrimination, the CJEU also provides guidelines to determine whether there has been indirect discrimination. Here the Court interprets the requirement of Article 2(2)(b) RED that persons of a racial or ethnic origin must be put at a ‘particular disadvantage’, as well as the criteria for objective justification. The Court leaves to the referring court to decide, notably, if the Danish legislation puts persons belonging to certain ethnic groups at a particular disadvantage (§ 143) and, if that is the case, if the restructuring measures are appropriate, necessary and proportionate in relation to the public interest of promoting social cohesion and integration (§§ 164-165).

 

On ethnic origin: a step forward but not quite there yet

A central issue in the judgment is whether the category of “non-Western immigrants and their descendants” falls within the prohibited ground of racial or ethnic origin under the RED. The Court of Justice has long displayed an uneasy relationship with the concepts of “race” and ethnic origin — most notably with the former, which it has traditionally been reluctant to invoke expressly. In CHEZ, the Grand Chamber held that ethnic origin ‘has its origin in the idea of societal groups marked in particular by common nationality, religious faith, language, cultural and traditional origins and backgrounds’ (§ 46). While this definition in itself is not unnecessarily restrictive, its application in subsequent cases demonstrated a very restrained understanding of the concept. In Jyske Finans as well as Land Oberösterreich v KV the Court’s focus was on establishing the presence of objective characteristics rather than identifying social processes of racialisation (here) and othering through which such characteristics gain social meaning and become markers of racialized social hierarchy. Moreover, it was suggested that the requirement of a ‘particular’ disadvantage in Article 2(2)(b) RED meant that only ethnically homogeneous groups could seek protection under the Directive, thus excluding differential treatment of heterogeneously composed groups such as ‘third-country nationals’. This approach has been widely criticised in the literature (for example here, here,and here). Commentators have argued that the Court’s understanding of ethnic origin risked hollowing out the protective scope of EU anti-discrimination law and failed to account for the lived realities of racialised exclusion suffered by migrants in particular.

Despite AG Ćapeta’s invitation (Opinion, § 69) the Court does not expressly recognize ‘ethnic origin’ as a socially constructed category. It confirms, however, that the term ‘racial or ethnic origin’ in the RED must be understood in light of the International Convention on the Elimination of All Forms of Racial Discrimination (ICERD) and of Article 14 European Convention of Human Rights (ECHR) and the case law of the European Court of Human Rights (ECtHR). This confirms the Court’s growing tendency to engage with external sources, especially its Strasbourg counterpart, even as EU accession to the European Convention on Human Rights remains pending. The reference to ECtHR case law is moreover significant as that court has previously espoused a less rigid understanding of ethnicity, notably in the Grand Chamber judgment in Biao v. Denmark where it held that ‘Danes of foreign origin’ were subject to ethnic discrimination. The CJEU also recalls the ECtHR’s consistent qualification of racial discrimination as a ‘particularly invidious form of discrimination which, in view of its perilous consequences, requires […] special vigilance and a vigorous reaction’ (§ 79). It thus makes clear that the issue at stake is one of racial discrimination and that no relevant distinction exists in this regard between the concepts of “race” and ethnic origin. Still, it fails to draw any strong doctrinal consequences from this acknowledgement in the form of a more explicitly social conception of racial discrimination.

A very welcome aspect of the case is the Court’s clarification that ethnic origin—as previously defined in CHEZ—need not be determined by reference to a single or homogeneous ethnic group (§§101-104; §§134-140). Returning to its first judgment on the RED, Feryn (§103), the Court clarifies that the notion of ethnic origin may apply to broadly formulated criteria such as ‘allochtones’, ‘foreigners’ or, indeed, ‘non-Western immigrants’. Still, it continues to insist, as it did in Jyske Finans, that a single characteristic such as nationality or country of birth, can never be indicative of ethnic origin (§ 86). Instead, ethnic origin must always be determined based on a combination of factors. As we argued earlier, this insistence on multiple characteristics denies the reality of racial discrimination in which a single feature (such as someone’s nationality, surname or skin colour, for that matter) may be the reason for differential treatment if that feature functions, in the given context, as a marker of “race” or ethnicity. The Court thus keeps open a significant loophole in EU anti-discrimination law—one that Member States may exploit by designing exclusionary measures around formally non-protected criteria, while shielding them from scrutiny under the prohibition of racial and ethnic discrimination.

In the present case, however, the single characteristic requirement does not stand in the way of a finding of ethnic discrimination. The Court is satisfied that the criterion of ‘immigrants from non-Western countries and their descendants’ is based on a ‘complex combination of criteria’, including country of birth, the nationality and country of birth of the parents and whether those countries of birth and/or nationality are considered, under Danish law, as ‘non-Western’ countries (§ 100).

On direct and indirect discrimination

A distinction based on ethnic origin constitutes direct discrimination if it involves a person being treated less favourable than another in a comparable situation (Art. 2(2)(a) RED). In line with the AG’s Opinion, the Court identifies two types of less favourable treatment that could result from the Danish legislation. The first is that tenants of public family housing in “transformation areas” face a higher chance of having their lease terminated compared to tenants with a similar lease in “vulnerable residential areas”, which are comparable to “transformation areas” in socioeconomic terms but where “non-Western immigrants and their descendants” do not form more than half of the population. It follows from the judgment that the risk of a lease being terminated early is, in itself, sufficient to constitute less favourable treatment, there is no requirement that this risk must have materialised. The Court moreover confirms what it had already decided in CHEZ, namely that there is direct discrimination when less favourable treatment results from a distinction based on ethnic origin, even if among those affected there are people who do not belong to the targeted ethnic group (§ 107). After all, tenants who are not themselves “non-Western” immigrants face the same risk of losing their homes as their “non-Western” neighbours. What is at stake here is a form of discrimination by association, although this concept is not mentioned explicitly by the Court.

Besides the risk of having their leases terminated, the Court notes a second form of harm that the applicants may have suffered, which is stigmatisation. Here the Court shows itself much more deferential to the Danish courts: it is up to the referring court to investigate if ‘the very name “transformation area”, which for the residents of areas classified as such gives rise to an increased risk of early termination of their lease, and which replaced the name ‘hard ghetto area’ is, at national level, offensive and stigmatising’ (§ 126). It is interesting that, at this point, the Court does show consciousness of the fact that the meaning of certain terms is socially constructed and can therefore differ over time and place, in this case leaving the Danish courts in a better position to determine the stigmatising connotations of the notion of “transformation areas”.

After having thus applied a broad definition of direct discrimination, which acknowledges both redistributive and recognitional harm, the Court moves on to examine the possibility of indirect discrimination. According to the Court itself, this analysis is necessary in case ‘the referring court concludes that the national legislation at issue […] does not constitute direct discrimination’ (§ 130). While this may at first sight seem a logical step to take, the Court’s consideration of indirect discrimination becomes less obvious if it is taken into account that the previous part of the judgment leaves very little to no room for the referring court to find that there has not been any direct discrimination. Indeed, once direct discrimination is established, the analysis can end there: discrimination is either direct or indirect, but not both, at least not on the same ground.

This move is troubling. The scheme at issue is plainly not one of indirect discrimination, and introducing this layer of analysis risks generating confusion rather than clarity for national courts. For one, if the referring court concludes that there is no direct discrimination because, after all, the criterion concerning “non-Western immigrants and their descendants” does not result in differential treatment based on ethnic origin, this would raise the question of what would then be the ethnic group (or groups) that could be put at a particular disadvantage, as required by Article 2(2)(b) RED. This is a question on which the judgment remains silent. The risk of confusion is further exacerbated by the fact that the Court’s own conceptualization of indirect discrimination is fraught with difficulties, as illustrated most clearly by its case law on religious symbols and headscarves (see here and here).

On the upside, the judgment does provide valuable clarification as to how the case should be assessed if viewed through the lens of indirect discrimination. This may prove practically significant, as experience shows that even where direct discrimination is formally removed, discriminatory effects often re-emerge—consciously or unconsciously—through ostensibly neutral criteria. One can easily imagine, for instance, that the Danish legislator might in the future abandon the explicit reference to “non-Western immigrants and their descendants”, while continuing to target specific neighbourhoods through socio-economic indicators that would, in practice, place that very group at a particular disadvantage. Some key takeaways from The Court’s analysis are: 1) that the aim of “ensuring successful integration of third-country nationals (TCNs)” is legitimate in principle; 2) the potential of the Danish scheme to actually promote social cohesion and integration is viewed as doubtful, not least because it does not apply to “vulnerable residential areas” that are socioeconomically comparable to “transformation areas” but without a majority “non-Western” population (§§ 162-163); 3) in addition to the requirements of appropriateness and necessity, which are expressly mentioned in Article 2(2)(b) RED, the Court examines the measures’ proportionality stricto sensu, something that has not commonly done before (see here and here); 4) regarding this proportionality stricto sensu, it must be taken into account that the right to respect for the home is a fundamental right protected by the EU Fundamental Rights Charter (Article 7) and that, according to ECtHR case law, the loss of one’s home constitutes “a most extreme form of interference” (§ 170).

Structural discrimination

In sum, there is a clear suggestion that the Danish scheme would also have to be considered as indirectly discriminatory – at least if a particular disadvantage to one or more ethnic groups can be established. The judgment can thus be readily qualified as an important blow to Denmark’s “ghetto” policy. Still, from a perspective of structural equality several problematic aspects remain – we round off this blog by mentioning two of them. The first, already raised by one of us elsewhere, is that the judgment entrenches the existing distinction between EU citizens and third-country nationals whereby only the latter are deemed in need of integration. This becomes evident in the indirect discrimination analysis, where it is pointed out that “integration” is a legitimate public interest only where third-country nationals are concerned (§ 151) – the subtext being that EU citizens are by definition already integrated.

Second, the Court seems to leave open the possibility that, under certain circumstances, “gentrification” measures such as those imposed by Denmark might be proportionate. This suggestion is troubling, given that those measures reflect a discriminatory logic that runs even deeper than the use of stereotypes in legislative documents or the harm suffered by individual tenants who are forced to leave their homes. At the basis of Denmark’s (or perhaps any) integration policy lies the rationality of Goldberg’s ‘racial state’ – the notion that at its core the modern state serves to protect public order by excluding the racialised other. It follows that any proportionality analysis cannot meaningfully abstract from the fact that the policy is premised on the stigmatization of racialised and impoverished communities. Yet is seems that the Court failed to recognise this, unlike in CHEZ where it expressly acknowledged the offensive and stigmatising nature of the practice at issue (§§84 and 108).

In short, there are several positive points to be noted about the judgment. These are, in particular, the Court’s acknowledgement that race discrimination does not turn on the presence of a homogeneous ethnic group, its recognition of the stigmatisation inherent in the so-called ‘ghetto law’ and its engagement with international instruments and proportionality in the indirect discrimination analysis. Still, the Court stops short of addressing the central issue at stake: the systemic and structural character of racialised exclusion. By avoiding a direct confrontation with structural racism, the Court ultimately does not get to the bottom of what makes the Danish ‘ghetto law’ so deeply problematic.