February 07, 2023
In the recent case of Elmazova and Others v. North Macedonia, the European Court of Human Rights (ECtHR or the Court) received another opportunity to speak out against (Roma) segregation in education. It rose wonderfully to the occasion. In a unanimous and well-reasoned judgment, the Court condemns the existing segregation and clarifies that States have a positive obligation to implement desegregation measures that will ‘ensure the end of the segregation’ (§ 89).
In the past, commentators have compared the ECtHR case law on Roma segregation to that of the United States Supreme Court on racial segregation in education, having its origin in the seminal judgment of Brown v. Board of Education. Brown marked the beginning of an important shift in the US legal landscape and an incredible move towards desegregation. It has been argued that the ECtHR, in delivering its judgment in D.H. and Others v. the Czech Republic, was at a Brown v. Board of Education moment. However, the Court’s case law has never really risen up to the challenge of combatting segregation. Similarly, where the US Supreme Court’s judgments provided a very promising start, with later rulings the Supreme Court has undermined much of the realised progression by increasingly limiting avenues for desegregation. I argue that, with its most recent judgments on Roma segregation in education, the Court is surpassing Brown’s achievements and proving that it can take a clear stance in condemning segregation and requiring desegregation.
The case of Elmazova concerns the segregation of Roma pupils in two primary schools, G.S. and G.D. In the first school, G.S, 83.5% of pupils were Roma. In another school in the same catchment area, T.A., 95.1% of pupils were ethnic Macedonians. In the preceding few years, 58 requests for transfers to T.A. had been made – of which three were submitted by Roma – which had all been granted. In assessing the complaint of segregation made before it, the majority of the Constitutional Court held that, although the children attended two ethnically divided schools, the statistical information regarding the ethnic composition of classes in both schools was insufficient to establish segregation. The Court was not competent to assess the overall social context and the applicants had not presented any facts or evidence to support their alleged inability to enrol in T.A.
In the second school, G.D., 64-67% of all first-grade pupils were Roma. Of the three first-grade classes, two were mixed and one was Roma-only. The Constitutional Court established that two contextual factors contributed to the high presence of Roma pupils in the school: on the one hand, a large Roma community resided in G.D.’s catchment area, and on the other, many pupils of Macedonian ethnic origin enrolled in out-of-catchment-area schools. This situation prevented G.D. from ensuring a balanced placement of pupils in terms of their ethnicity, sex and social status. The school did try to take measures to ensure this balance, such as attempting to convince majority parents to enrol their children or asking the municipality for stricter implementation of decisions regarding school catchment areas, but this was to no avail. The Constitutional Court concluded that ‘there was no evidence that Roma pupils in the “segregated” classes had been treated differently in comparison with pupils in the mixed classes, namely that they had obtained an inferior education’ (§ 13).
After the date of application to the ECtHR, in both cases, the European Roma Rights Centre lodged a complaint of segregation before the Commission for Prevention and Protection against Discrimination. In both cases, the Commission issued an opinion condemning the ethnic segregation as indirect discrimination and pointing to the fact that this was a systemic problem in the educational process.
After reiterating the general principles applicable to cases of Roma segregation, the ECtHR begins by establishing that both schools offered a programme that was the same as, or at least similar to, that of other public primary schools (contrary to the situations in the earlier cases of D.H. and Others v. the Czech Republic (GC), Sampanis and Others v. Greece, Oršuš and Others v. Croatia (GC),Sampani and Others v. Greece and Horváth and Kiss v. Hungary).
Concerning G.S., the Court observed that this school was predominantly attended by Roma children (83.5%), while T.A. was almost exclusively attended by ethnic Macedonians (95.1%). It highlighted that there was a large disproportion between Roma and non-Roma placements in G.S. It moreover accepted that the main reason for this was the manner in which the regulatory framework on the admission and transfer of pupils was applied in practice, owing to the opposition of non-Roma parents to enrolling their children in G.S. (§ 72). Refuting the government’s arguments, the Court held that the residential ethnic composition of the catchment area could not, in the circumstances, be sufficient to objectively justify the segregation of Roma in G.S., as the ethnic composition of T.A. (located 600 metres from G.S.) was totally different (§ 73). Contrary to the Constitutional Court’s findings, the Court did not consider that the applicants’ failure to seek a transfer to other schools pointed to a lack of victim status. G.S. had been these pupils’ school of compulsory admission and the State had been required to
‘take positive effective measures to correct the applicants’ factual inequality and avoid the perpetuation of the discrimination that resulted from their over-representation in G.S., thereby breaking the circle of marginalisation and allowing them to live as equal citizens from the early stages of their life.’ (§ 74, with reference to X and Others v. Albania, § 84 and Horváth and Kiss, § 116)
It noted that the government had provided no explanation as to
‘why no measure had been taken to correct the imbalance in the ethnic composition of the two schools in question, or to redistrict the school district map, or why no other appropriate desegregating measure has been taken’ (§ 74, emphasis added).
Turning to the second school, G.D., the Court noted that Roma pupils represented 64 to 67% of all first-grade pupils and that in the 2018/2019 academic year, of the three first-grade classes, two were mixed and one was Roma-only. From this information, it concluded that there was no ‘general policy to automatically place Roma pupils in separate classes in the school at issue’ (§ 75). At the same time, in the 2017-2018 academic year, the Roma-only class had comprised of only 18 pupils (compared to 31 and 32 in the mixed classes), which was below the minimum threshold of class size set by law (§ 76). The school acknowledged the existence of segregation and took certain measures to tackle the problem. These measures failed mainly because of opposition by the parents of non-Roma children (§ 77).
The Court unanimously concluded that the segregation of Roma children in G.S (overrepresentation). and G.D. (Roma-only class) had amounted to discrimination, due to the state’s failure to implement the required positive (desegregation) measures. Under Article 46 of the ECHR, it held that ‘the coexistence of members of society free from racial segregation is a fundamental value of democratic societies and that inclusive education is the most appropriate means of guaranteeing the fundamental principles of universality and non-discrimination in the exercise of the right to education’ (§ 89, with reference to Çam v. Turkey § 64). Therefore, measures need to be taken in order to ensure the end of the segregation of Roma pupils in G.S. and G.D., as recommended by the European Commission against Racism and Intolerance, the National Commission for Prevention and Protection against Discrimination, and the Ombudsman (§ 89).
The current judgment lives up to the expectations set by the previous judgment of X and Others v. Albania, which marked a promising evolution in the Court’s case law on Roma segregation in education. In my analysis of that case, I reminded readers of the previous Roma segregation cases and put that judgment into the context of these cases. Together, X and Others and Elmazova mark a turn in the Court’s assessment of segregation in education, both in terms of what is required in order to find a situation of segregation discriminatory and in terms of its focus on desegregation measures. In the current commentary, I will take a positive approach, by focusing on all the laudable elements of the judgment – leaving the (limited opportunities for) criticism for elsewhere.
In the current judgment, the Court held that the overrepresentation of 83.5% Roma pupils in G.S. amounted to segregation, in violation of the prohibition of discrimination. The mere fact of overrepresentation was sufficient to find a violation of Article 14. This finding crystallises the shift towards lowering the threshold that needs to be reached in order to prove that a situation of segregation is discriminatory under the Convention.
In addition, the Court can be applauded for its realism in assessing the existence of segregation: the fact that there is a clear overrepresentation of pupils of a certain background indeed points to segregation, even where there is no ‘full’ segregation. This reasoning could also be extended to the composition of different classes within a school. This is an approach which the Court has not (yet) taken, as the findings on the second school of G.D. show. In the cases on class composition, the Court holds on to its reasoning developed in Oršuš and Others v. Croatia, where it investigated whether a general policy existed to automatically place Roma pupils in separate classes in the school at issue (Elmazova, § 75, with reference to Oršuš, § 152). Instead of also considering class composition in terms of overrepresentation, the Court applies this criterion for deciding whether classes in the school in general were segregated. If this criterion is not met, the Court looks at individual classes, where segregation may still exist – as was the case in both Oršuš and Elmazova.
In its assessment of whether segregation existed, the Court can also be cheered upon for making an explicit comparison to the ethnic composition of the majority school, T.A., in assessing whether G.S. was segregated. It is indeed extremely relevant that there is another segregated school in the same catchment area, one which is ‘almost exclusively’ attended by ethnic Macedonians (§ 71). In the past, attention has mostly been geared towards the disadvantaged schools and how segregation should be solved in these schools. However, segregation can and should not be solved by focusing only on these schools. This approach can be compared to solving a problem (e.g. bullying or sexual harassment) by focusing on changing the behaviour of the victim instead of that of the perpetrator. Concerning segregation, changes need to be made in both minority- and majority-segregated schools in order to solve the issue of inequality. In the current example, the acceptance policies of T.A. are part of the reason why G.S. is segregated. Solving the issue of segregation in G.S. will require other schools in the area to take on additional Roma pupils and will require a coordinated approach among them, one that includes engagement with the majority parents who may be opposed to the desegregation out of a fear that it would lower the educational quality. By explicitly shining its light also on T.A., the Court seems to embrace this.
Lastly, in accepting that the mere fact of overrepresentation gives rise to segregation, the Court clearly distances itself from the limiting case law of the US Supreme Court, in which a ‘racial imbalance’ (i.e. overrepresentation) cannot lead to the adoption of desegregation measures – be they voluntary or otherwise. In the Supreme Court’s view, this racial isolation is the result of societal discrimination, economic characteristics and/or innocent private choices, with which the Court has no business. This approach has been widely criticised as unfit to tackle the issue of segregation as a social reality. The ECtHR is thus moving to surpass what has ultimately been achieved in the US in the wake of Brown by engaging with social reality. It accepts, contradicting also the Macedonian Constitutional Court (§ 8), that the ‘overall social context’, including the practical application of the regulatory framework, the opposition of non-Roma parents and residential segregation, is legally relevant.
Aside from the Court’s assessment of the existence of segregation, its focus on desegregation measures is also significant. In its assessment under article 46 of the Convention, concerning the measures necessary for the implementation of the judgment, the Court holds that ‘measures to be taken in the context of the present case must ensure the end of the segregation of the Roma pupils in G.S. and G.D.’ (§ 89, emphasis added). This is very clear and explicit guidance on what is needed for compliance with the judgment – guidance that is usually not present in the Court’s judgments. The Court hereby sends a strong message to the state in question – and, one may imagine, all the other states that still have high Roma segregation rates in education, often despite earlier condemnations by the Court – that actual, positive measures will need to be taken. The message reminds of the requirement in X and Others v. Albania that desegregation measures need to be swift and effective, which doubtless will also be applicable to the current situation.
In my analysis of X and Others v. Albania, I criticised the Court for suggesting two desegregation measures which the Government had proposed, although it was quite unclear whether the suggestions were evidence-based or would be effective. Moreover, I questioned whether the Court would be able to assess the appropriateness of the measures suggested. Of course, where the aim is to actually solve the problem of segregation in education, it is necessary to reflect on these issues. In the current judgment, the Court cleverly refers to recommendations by specialised international and national institutions for the question of how to ensure the end of the existing segregation. The European Commission against Racism and Intolerance (ECRI), the National Commission for Prevention and Protection against Discrimination and the Ombudsman all have certain expertise on how to combat segregation and/or insight into the complexity of the problem at the national level. Combining these institutions’ expert knowledge will ensure the quality and effectiveness of future desegregation measures by the state.
Both X and Others v. Albania and the current judgment in Elmazova and Others v. North Macedonia show a shift by the Court from a focus on the assessment of whether situations of segregation in education are discriminatory under the Convention to a focus on positive obligations stemming from segregation and desegregation measures. The question that now remains, is, when will the goal of putting an end to segregation be achieved? This question has also been put before the US Supreme Court for assessment, which has elaborate case law on when a school district can be considered ‘unitary’ and when desegregation measures should thus be lifted. Through years and years of implementation of desegregation and declarations of unitary status, it has become clear what the benefits and the pitfalls of the Supreme Court’s approach are. In many instances the applied measures have clearly led to decreasing segregation rates, although the declaration of unitary status and the ensuing obligation to lift all desegregation measures have resulted in a new rise of segregation. In the future, the ECtHR may be required to give further guidance on the question of when segregation has been ‘ended’. It might be wise to look at the US Supreme Court for inspiration on what approach towards this question to adopt – and on what approach to evade.
The unanimous judgment in Elmazova and Others v. North Macedonia marks another tremendous step forward in prohibiting segregation and guaranteeing equality of all pupils in education. Together with the recent case of X and Others v. Albania, it shows a shift in the ECtHR’s approach from a focus on an assessment of whether situations of segregation in education are discriminatory under the Convention to a focus on positive obligations stemming from segregation and desegregation measures. These judgments have made proving segregation easier, by accepting that merely showing overrepresentation is sufficient for shifting the burden of proof. Elmazova itself can moreover be applauded for its realism in assessing the existence of segregation, contrasting the overrepresentation of Roma pupils in one school to the very homogeneous ethnic composition of the majority school in the same catchment area. By engaging with social reality, in this ruling, the ECtHR is moving to surpass what has ultimately been achieved in the US in the wake of Brown. Lastly, the judgment clearly orders the government to ensure the end of the existing segregation while pointing the member state to specialised international and national institutions for guidance on how to achieve this result. This judgment shows that the Court has genuinely understood the harms of segregation and will take a strong stance in combatting it, while guiding the member states struggling with this complex issue. It has the potential to create (more) equal opportunities for countless children.