EU lawyers had been awaiting 18 December with particular anticipation, which coincidentally or not, was also International Migrants Day. On that day, the Grand Chamber of the Court of Justice of the European Union delivered several significant judgments, including two concerning Frontex’s involvement in push-backs and unlawful returns of migrants, and one addressing the Danish housing law under the EU Race Equality Directive. The category of “non-Western immigrants and their descendants” was central to that latter case.

These cases emerged in a politically charged context regarding migration. Across Europe, migrants are increasingly portrayed as threats, both from “outside” and from “within”. Few governments have been willing to articulate a genuinely positive discourse on migration, despite its broadly beneficial effects. At the same time, racism (here and here) and xenophobia have intensified markedly. Importantly, this racism does not operate only at the level of individual prejudice. It is also reproduced through law and institutions, through the cultural process identified by sociologists as “racialisation” which entrenches ethno-racial hierarchies irrespective of the subjective intentions of policymakers.

For EU lawyers working on migration and anti-discrimination, these cases offer a striking illustration of the continuity of racialised governance within the Union. That continuity begins at the EU’s external borders – exemplified by Frontex – and extends inward into the most intimate sphere of social life: the home. Even individuals who have long settled and acquired citizenship may continue to be treated as migrants, as not fully belonging, particularly when classified as “non-Western”. The message is unmistakable: not quite “one of us”.

While the Frontex judgments have rightly been welcomed for challenging what had come to resemble a regime of near-impunity, the Danish housing law judgment leaves a more ambivalent impression. It is neither a decision to celebrate nor one to condemn outright. Rather, it exposes the Union’s persistent difficulty – indeed, its immaturity – in confronting one of Europe’s most entrenched pathologies: the systematic subordination of those perceived as failing to conform to white, middle-class, Christian norms.

Much has already been said about the judgment, including by Serde Atalay, Lucía López Zurita and Boris Belortaja as well as during the Oxford–European Law Review online event featuring Iyiola Solanke, Tamara Ćapeta, Nozizwe Dube, Hannah Eklund, and Bruno de Witte. What follows offers a focused reading, while reserving broader questions for a longer contribution. It first briefly unpacks the Court’s judgment, before turning to an analysis of three interrelated issues: the Court’s understanding of ethnic origin, its entrenchment of a two-tier system of justification based on nationality, and the racialization of poverty.

The Danish Housing Law as Direct Discrimination

The Danish Housing Law has already attracted extensive academic and public commentary. Its objective is the “gentrification” of neighbourhoods designated as socioeconomically disadvantaged, identified on the basis of four criteria: unemployment rates, educational attainment, criminal convictions and average gross income. Among the neighbourhoods that meet these thresholds, only those in which at least fifty percent of residents are classified as “immigrants and their descendants from non-Western countries” – including both Danish nationals and non-nationals – are designated as “transformation areas” (previously called “hard ghetto areas”) and are consequently the only neighbourhoods subjected to these “gentrification” measures, notably the termination of lease agreements and forced relocation.

Applicants from two designated areas – in Slagelse and in Copenhagen – challenged before the High Court of Eastern Denmark the termination of their leases and the ministerial decision approving the “gentrification” plan of their neighbourhoods. That court referred two questions to the CJEU: whether the categorisation of “non-Western immigrants and their descendants” constitutes a distinction based on ethnic origin under the Race Equality Directive, and whether the scheme gives rise to direct or indirect discrimination.

In a widely discussed landmark Opinion, Advocate General Ćapeta answered both questions in the affirmative, concluding that the categorisation is based on ethnic origin and that the resulting differential treatment amounts to direct discrimination. The Grand Chamber ultimately reached the same substantive conclusion, albeit through a different route.

What is striking, however, is that despite effectively characterising the legislation as involving direct discrimination, the Court nonetheless proceeded to conduct a full analysis of indirect discrimination. In doing so, it also acknowledged the existence of indirect discrimination, conceding that the measure places persons belonging to certain ethnic groups at a particular disadvantage (§175). To reach this conclusion, the Court held that, if the measure indeed placed persons of certain ethnic groups at a particular disadvantage, the objective of “integration” could not qualify as an overriding reason in the public interest capable of justifying it with regard to nationals of Member States (§151), and that, even if such an objective could be relied upon in respect of others, the measure failed to satisfy the requirements of proportionality: it was neither appropriate (§§162–163) nor necessary (§164), nor proportionate stricto sensu (§§165–174). This additional step is puzzling: once direct discrimination is established, the legal analysis should, in principle, have ended at that point, as Bruno de Witte has observed.

The “ball”, so to speak, is now back in the Danish courts. That said, the answer appears clear: the Danish housing legislation gives rise to direct discrimination, requiring the Danish courts to determine the appropriate consequences. Consistent with the logic underlying the principle ex injuria jus non oritur, measures adopted on the basis of an unlawful framework cannot, in principle, be regarded as lawful.

Ethnic Origin: Contortions Between Objective and Social Understandings

A central issue raised by the judgment is whether the category of “non-Western immigrants and their descendants” falls within the notion of ethnic origin or race under the Race Equality Directive. The importance of this question can hardly be overstated. Once race or ethnic origin is accepted as the relevant ground, the conclusion that the Danish regime entails direct discrimination follows almost inevitably.

The Court of Justice has long had an uneasy relationship with the concepts of race and ethnic origin. Its case law has favoured an ostensibly objective and restrained understanding of ethnic origin, distancing legal analysis from social constructions, exemplified by Jyske Finans or Land Oberösterreich v KV. Such formalism has been widely criticised for weakening the protection of EU anti-discrimination law (here and here).

Against this backdrop, one of the most welcome – if overdue – aspects of the present judgment lies in the Court’s clarification that ethnic origin – understood as encompassing “societal groups marked, in particular, by common nationality, religious faith, language, cultural and traditional origins and backgrounds” (§72) – need not be determined by reference to a specific ethnic group, nor to a homogeneous one (§104; §§135–140). Bringing Feryn back to life (§103), the Court clarified that a broad understanding covering different (ethnic) groups may include the notion of ethnic origin (§§101–103; §143). To reach this conclusion, the Court relied on a “complex combination of criteria” (§§100 and 109) – nationality and place of birth, each of which may be plural – thereby implicitly endorsing a more contextual and socially grounded mode of reasoning. In this respect, the Court’s willingness to broaden its understanding of ethnic origin constitutes a meaningful development.

At the same time, this openness seems to remain circumscribed. The Court confines its reasoning to criteria of nationality and place of birth while leaving unclear where, precisely, it draws the boundaries of this expanded understanding and crucially, why certain characteristics are included while others are excluded in casu. Notably absent is any engagement with socioeconomic status or religion, despite their central role in the Danish housing regime and their obvious interaction with ethnicised or racialised classifications in practice.

This limited openness contrasts with scholarly proposals to conceptualise such distinctions through the lens of racialization, as advanced by Nozizwe Dube (here and here), building on Jennifer Nash’s account of race as produced through everyday practices and state-based policies. This racialization approach is particularly illuminating, as it explicitly captures the social construction of race and ethnic origin by a process of othering through the accumulation and interaction of multiple criteria, including language, legal status, national origin, religion, class, etc. While the Court does not fully embrace this framework, it advances only a semi-intersectional approach, acknowledging the interaction between multiple factors under the concept of ethnic origin while failing to identify all relevant factors and to articulate the nature of their interaction. As a result, the judgment falls short of engaging with the broader dynamics of racialisation at stake.

The judgment also raises troubling questions and ambiguities – most notably regarding the relationship between ethnic origin, on the one hand, and nationality or country of birth, on the other. It is precisely at this juncture that the Court appears to retreat to its long-standing, static understanding of ethnic origin, grounded in allegedly “objective” characteristics. As in Jyske Finans, the Court again declines to subject nationality or place of birth to the level of scrutiny they warrant, reasoning that, taken in isolation, these criteria do not give rise to a presumption of belonging to a particular ethnic group (§73) because “a person’s nationality cannot, in itself, give rise to a presumption of belonging to a particular ethnic group” (§81) and “it cannot be presumed that each sovereign State has one, and only one, ethnic origin” (§85).

This reasoning not only contradicts the Court’s own semi-intersectional approach to interpreting the ground of ethnic origin, but it also fundamentally misses the point. In practice, markers such as nationality, national origin, place of birth, religion, or legal status are, as Lilla Farkas has emphasized, among the most common – often self-standing – vehicles through which processes of racialisation are produced and reproduced – particularly when, following the Court’s own formulation, “the context surrounding the adoption of a general criterion” is taken into account (§105). Otherisation does not hinge on the empirical claim that a State has a single ethnic origin, nor on the formal link between nationality and ethnicity. It operates, rather, through the attribution of social meaning to individuals via markers such as nationality, religion or place of birth, which in practice serve as vehicles for racialised differentiation.

Although discrimination on grounds of nationality is formally excluded from the scope of the Race Equality Directive, this does not preclude nationality from operating in certain contexts as a proxy for race and ethnic origin, amounting to racial and ethnic discrimination. This is precisely what the Court fails to acknowledge. By excluding in principle the possibility that such criteria may function as proxies on their own for race and ethnic origin, the Court risks opening a significant loophole in EU anti-discrimination law – one that Member States may readily exploit by structuring exclusionary measures around place of birth or nationality, while insulating them from scrutiny under the prohibition of racial and ethnic discrimination.

Worsening the “Apartheid Européen”

In its unnecessary – and in many respects problematic – examination of indirect discrimination, the Court proceeds to assess whether the contested measures may be justified by the objective of integration (§§ 147–151). What is striking is the asymmetrical way in which this aim is treated.

The Court holds that integration cannot justify such measures where they affect the nationals of Member States (§151), while considering that the same objective could justify them when they affect non-nationals. In the present case – where the majority of the applicants were Danish nationals – this renders the measure largely unlawful under indirect discrimination, a welcome outcome for the applicants and beyond. Yet this should not be mistaken for a broader advance in anti-discrimination protection.

On the contrary, the Court entrenches a two-tier system of justification, affording stronger protection to Member State nationals and weaker protection to third-country nationals. This move is troubling. While it is well established that EU citizens enjoy preferential treatment in many areas of EU law, protection against racial discrimination is not one of them; racial discrimination does not become less objectionable because its targets are not EU citizens. Indeed, the Race Equality Directive was designed with immigrants in mind and has long constituted one of the few domains of EU law in which protection has not been explicitly stratified according to citizenship status – unlike, for example, free movement law. In other words, the statement of the Court regarding integration might have been doctrinally intelligible (and even welcome) in a free-movement context, but it is misplaced in a case concerning racial and ethnic discrimination. By reintroducing nationality at the level of justification in the interpretation of the Race Equality Directive, the Court fragments the uniform protection against racial and ethnic discrimination that the Directive is meant to secure. More broadly, this reasoning reinforces the structural divide between EU citizens and third-country nationals within the EU legal order – what Étienne Balibar famously described as an “apartheid européen”.

An unresolved ambiguity in the Court’s reasoning concerns the meaning of “nationals of Member States”. It remains unclear whether this category refers exclusively to Danish nationals residing in Denmark or whether it also includes other EU citizens living there, as the reference to “including the Kingdom of Denmark” would suggest. The latter case would sit uneasily with the much criticised strand of the Court’s free movement case law, where “integration” and “real links” have been used to restrict EU citizens’ rights in host Member States. If the Court instead meant nationals in their own Member State only, the judgment reinforces nationality as the primary axis of belonging for EU citizens, entrenching a deeply nationalised conception of Union citizenship that runs counter to the logic of European integration.

The core problem, however, is not that the Court rejected integration as an overriding reason in the public interest with respect to Member State nationals. It is that it failed to extend this reasoning to all persons affected by such measures, irrespective of citizenship, as Advocate General Ćapeta proposed (§170). Integration is not a neutral or a benign objective; it is an inherently malleable and increasingly instrumentalised concept, frequently deployed to justify exclusion and otherisation – a trend that the Court’s reasoning appears more apt to endorse than to constrain.

Racialisation of Poverty

To conclude, the Danish housing law is an instructive example of what has long been described as the racialisation of poverty. Racialised groups are disproportionately represented among those living in poverty in Europe (here and here), due to historical and structural inequalities, while poverty itself becomes associated with these groups and framed as an individual or cultural failing rather than systemic injustice (see here and here).

While extensively analysed in sociology and critical race theory, the racialisation of poverty remains strikingly undertheorised in law. Poverty is treated as a neutral socioeconomic condition, rarely recognised as a site of discrimination or as a racialised mode of governance. As Vandita Khanna has shown, racialised poverty remains marginally treated in law in Europe, including before the European Court of Human Rights, allowing policymakers to justify the exclusion, paternalisation and punishment of “racialised others”.

The Danish housing law fits squarely within this logic. This legal regime conveys a powerful message to racialised residents: that there is something inherently wrong with them, to the point that their displacement from their homes can be justified in the name of “gentrification”, with strong implicit logics of profit optimisation and market rationality. Racialisation and capitalism thus operate here in tandem. Crucially, the problem does not lie solely in the law’s explicit language – such as the use of the term “ghettos” – although, as the Court correctly observed, this terminology is itself stereotypical (§127). This is only the tip of the iceberg. Rather, the core issue lies in the structural incorporation of stereotyping within the legal regime, which the Court fails to confront, thus obscuring, in the words of Serde Atalay, “the dynamics behind the instrumentalisation of law to govern racialised groups”. Even in the absence of overtly stigmatising language, the law actively produces and legitimises racialised hierarchies. It is precisely this deeper, structural form of stereotyping that the Court ultimately failed to grasp.

I am immensely grateful to Vandita Khanna, Karin de Vries and Harry Panagopulos for their generous comments and suggestions. All remaining errors are, of course, my own.