The Right to Equal Treatment: Indirect Differential Treatment on the Grounds of Racial or Ethnic Origin
9 THE PREVIOUS CHAPTER established that the Dutch Act on Integration Abroad distinguishes between different groups of aliens on the basis of their nationality. This chapter investigates whether this distinction also amounts to differential treatment on the grounds of racial or ethnic origin. Clearly, the Act does not expressly mention racial or ethnic origin as a criterion to determine which aliens must pass the integration exam abroad. As asserted earlier however, the right to equal treatment also entails protection against indirect discrimination. Such discrimination may occur when a certain group of persons is disproportionately affected by a particular measure, even if the measure is not directly aimed at that group (section II.D. of chapter 8). An important objective of the legal provisions forbidding racial and ethnic discrimination is to protect persons against actions or practices inspired by racism or ethnic intolerance (section II.E). However, the concept of indirect racial or ethnic discrimination does not presuppose the existence of racist intentions or beliefs. Instead, it may be used to render visible the unintended effects that an apparently neutral measure – such as the Act on Integration Abroad – can have on particular ethnic or racial groups and that would have remained invisible if only forms of direct differential treatment were addressed. It is for this purpose that the concept of indirect discrimination is applied in this chapter. A general overview of the legal framework concerning the right to equal treatment is provided in chapter 8. The current chapter firstly examines the legal standards concerning differential treatment on grounds of race and ethnic origin and indirect differential treatment (section II). Particular attention is paid to the applicability and significance of these standards in relation to integration conditions for the admission of (certain groups of) aliens. Next, in section III, the legal framework established in section II is used to conduct an assessment of the Act on Integration Abroad. This section examines a number of legal norms that prohibit (indirect) discrimination on the grounds of race or ethnic origin. Various legal provisions and their applicability to the topic of this study are discussed in section II.A. After that, attention is paid to race and ethnic origin as differentiation grounds (section II.B), to the concept of indirect differential treatment (section II.C), the criteria for justification (section II.D) and the scope of review (section II.E). Lastly, in section II.F, some criteria are formulated concerning the prohibition of indirect racial and ethnic discrimination in relation to integration requirements for the admission of aliens. The equal treatment provisions of the ECHR, the ICCPR and the Dutch Constitution are discussed in some detail in chapter 8 (section II.B). It was established that each of these instruments contains open provisions on equal treatment that forbid discrimination based on any grounds or status, including race and ethnic origin.1 Discrimination on the grounds of race and ethnic origin is also prohibited under the Convention on the Elimination of All Forms of Racial Discrimination (hereafter CERD). The CERD prohibits racial discrimination, which is defined in Article 1(1) as: Any distinction, exclusion, restriction or preference based on race, colour, descent, or national or ethnic origin which has the purpose or effect of nullifying or impairing the recognition, enjoyment or exercise, on an equal footing, of human rights and fundamental freedoms in the political, economic, social, cultural or any other field of public life. It follows that the CERD does not cover all forms of differential treatment on the above grounds, but only those negatively affecting a person’s enjoyment of his or her human rights or fundamental freedoms. A list of such rights and freedoms is laid down in Article 5 CERD. Although this list is non-exhaustive, it may be observed that it does not include a general right to be admitted to a state of which one is not a national. Such a right is also not mentioned in several important human rights instruments, such as the Universal Declaration of Human Rights (UDHR), the ICCPR and the ECHR. It can therefore be questioned whether the CERD also applies to differential treatment in the field of immigration requirements. However, as shown in chapters 4 and 5, state measures concerning the admission of family migrants and religious servants may in certain cases come within the scope of national or international human rights obligations. It is therefore submitted that such measures are in principle covered by the CERD.2 This outcome is in line with ECtHR case law concerning Article 14 ECHR, in which the relationship between the prohibition of discrimination and the enjoyment of other rights and freedoms is also construed rather loosely and it is sufficient if the alleged discrimination comes ‘within the ambit’ of one of the other ECHR provisions (section II.B.i of chapter 8). Meanwhile several other limitations to the definition of Article 1(1) can be found in Articles 1(2), 1(3) and 1(4) CERD. According to these provisions, racial discrimination does not result from unequal treatment between citizens and non-citizens, from legal provisions concerning nationality, citizenship or naturalisation (provided they do not discriminate against a particular nationality) or from ‘positive discrimination’ measures. Discrimination on the grounds of race and ethnic origin is also addressed in various instruments of EU law, in particular Article 19 TFEU and Article 21(1) of the EU Charter of Fundamental Rights. Article 19(1) TFEU served as the legal basis for the adoption of the EU Racial Equality Directive (RED), which affords protection against racial and ethnic discrimination in various fields such as employment and education.3 Although the directive itself states that it does not apply to differences in treatment that are based on nationality, it may be argued that this exception does not concern differences amounting to indirect differential treatment on grounds of racial or ethnic origin.4 More important, for the purposes of this chapter, is that the directive also does not apply to provisions and conditions concerning the entry and residence of third-country nationals and stateless persons in the territory of Member States.5 As EU nationals in any case do not have to comply with integration requirements (chapter 6, section II), it must be understood that the RED is of little relevance for the examination conducted in this chapter. On the other hand, Article 21(1) of the Charter of Fundamental Rights is not subject to such limitations. This provision prohibits ‘any discrimination’, including on the grounds of race and ethnic origin.6 The prohibition is addressed to the institutions of the European Union, as well as to the Member States when they are implementing EU law.7 It follows that, within the context of this study, Article 21(1) of the Charter is of relevance with regard to integration requirements for third-country nationals who apply for family reunification under the Family Reunification Directive (FRD) or who move to a second Member State in accordance with the Long-term Residents Directive (LRD) (section V of chapter 4 and section VI of chapter 6). Race and ethnic origin are difficult concepts, which are closely related to social and cultural identity. No definition of these concepts is provided in any of the legal instruments mentioned above. However, in Timishev, the ECtHR stated that: Ethnicity and race are related and overlapping concepts. Whereas the notion of race is rooted in the idea of biological classification of human beings into subspecies according to morphological features such as skin colour or facial characteristics, ethnicity has its origin in the idea of societal groups marked by a common nationality, tribal affiliation, religious faith, shared language, or cultural and traditional origins and backgrounds.8 It can be derived from this definition, as well as from academic literature on the subject, that both race and ethnic origin are socially constructed concepts which are based on the presumption that people can be divided into distinguishable groups or communities.9 This presumption may be held by outsiders, as well as by the members of the group themselves. As the ECtHR noted in Timishev, racial and ethnic categorisations are commonly linked to particular features or characteristics (Busstra uses the term ‘proxies’), whereby the term ‘race’ is generally used for categorisations on the basis of physiological features (notably skin colour) and ‘ethnicity’ refers to categorisations based on cultural or sociological characteristics (such as language or religion).10 However, it must be observed that such features and characteristics cannot be equated to race or ethnic origin, but only form an indication thereof.11 Given that race and ethnicity refer to perceptions of belonging, rather than to actual behaviour or physical features, one particular difficulty in the legal context is how a person’s racial or ethnic origin is to be established.12 Probably the most reliable way of doing this is by means of self-identification.13 Alternatively, a person’s racial or ethnic origin can be determined by someone else on the basis of the above proxies. Importantly however, racial and ethnic classifications are not static, but subject to change and contestation.14 It follows that the proxies by which racial and ethnic groups are designated cannot be established once and for all, and that regard must be had to the particular social context. In this connection Busstra pointed out that international human rights law has come to recognise an increasing number of characteristics as proxies for race or ethnicity, including geographical affiliation and even social origins and class.15 The use of proxies can also be a relevant instrument for establishing whether a particular measure leads to a difference in treatment between persons on account of their race or ethnic origin. This may be the case, in particular, where the applicability of a measure is determined by a criterion that is also a proxy for race or ethnicity. Much will depend, however, on the nature of the criterion, as well as on the context in which the measure is taken. For example where a distinction is based on skin colour, there will most likely be differential treatment on the grounds of racial origin as skin colour has historically functioned as a proxy for race and is not likely to be relevant in any other way. On the other hand, the situation often becomes more complicated if other criteria are used as grounds for differentiation. In the context of the Netherlands, for example, someone may be refused a job because he or she speaks Arabic or because he or she does not speak Dutch. In both cases the difference in treatment is based on language. However, the fact of speaking Arabic can be taken as a proxy for belonging to a particular ethnic group (in other words, Muslims or Moroccans). With regard to the fact of not speaking Dutch, such a connection is much more difficult to make. Yet it may be that this criterion, even if not directly indicative of racial or ethnic origin, nevertheless entails indirect differential treatment on these grounds. The issue of indirect differential treatment is discussed in more detail in section II.C below. Several of the legal provisions referred to in section II.A mention national origin as a potential differentiation ground, sometimes in addition to race and/or ethnic origin. In the previous chapter it is argued that this ground refers primarily to nationality as an ethnographical category.16 In this sense, the term ‘national origin’ does not differ to any relevant extent from ‘ethnic origin’ and therefore does not require separate discussion.17 The concept of indirect differential treatment was briefly introduced in the previous chapter (section II.V). In short, indirect differential treatment occurs if a measure formulated in neutral terms nevertheless disproportionately affects a particular group of persons falling under a relevant differentiation ground. As mentioned earlier, this effect may be caused intentionally, for example where a measure distinguishes between groups on account of their nationality in order to circumvent the prohibition of discrimination on the ground of race or ethnicity. In the context of the legal provisions discussed in this chapter, however, no evidence of discriminatory intent is required: it is the effect rather than the aim of the measure that is relevant.18 By using the concept of indirect indiscrimination in this way it becomes possible to pursue a more substantive concept of equality (see again section II.V of chapter 8) and to address structural inequalities including, but not limited to those resulting from a history of intentional discrimination.19 In cases of alleged indirect discrimination, it is normally up to the applicant to demonstrate that a particular measure has a disproportionate effect on the group to which he or she belongs. If this succeeds, a presumption of indirect discrimination is thereby established. The burden of proof then shifts to the respondent state, which has to show that the difference in treatment is based on a reasonable and objective justification (section II.D).20 The determination of the existence of a disproportionate effect is nonetheless a complex issue, which raises many questions. While an in-depth examination of these questions goes beyond the scope of this study, some of the main difficulties involved are highlighted below. A primary issue relating to the establishment of a presumption of indirect discrimination is of course the standard of proof that is required. In EU law, where the concept of indirect discrimination has been well developed, different criteria have been devised in this regard. In the field of nationality discrimination the CoJ has accepted that indirect differential treatment occurs when a particular measure is ‘intrinsically liable’ to affect migrant workers more than national workers, for example because it differentiates on the basis of geographic elements.21 This is a rather light standard, which does not require the applicant to demonstrate that the measure has actually produced an adverse effect, but merely that it is, by its nature, likely to do so. By contrast, in cases of sex discrimination, the Court has asked applicants to show that the contested measure has in fact resulted in a disadvantage for a substantially higher proportion of the members of one sex than the other.22 This criterion requires much stronger evidence of a disproportionate effect, which will often involve statistics (see below).23 The above standards for establishing a presumption of indirect discrimination differ primarily with regard to the type of evidence to be adduced.27 If it is enough for a particular group to risk being disproportionately affected, the applicant may rely on qualitative evidence to prove such a risk. Such evidence may consist of facts of general knowledge.28 For example, it may be submitted that the socio-economic position of Turkish and Moroccan communities in the Netherlands is generally less favourable than that of the majority population. Where necessary, these facts may be complemented or backed up by other materials such as scientific or other reports, policy evaluations or judicial decisions.29 In some cases the risk of disproportionate effects may also be directly related to the criterion used. For instance, a rule that applies only to persons needing a residence permit automatically entails indirect differential treatment on the grounds of alienage (section III.A.i of chapter 8).30 Meanwhile, proof of actual disproportionate effects on a particular group has to be demonstrated by means of quantitative (statistical) evidence. One practical problem relating to this type of proof is that statistical data will not always be available. In addition, the evidence must be significant, and this may be problematic if the disputed measure concerns only a small group.31 The latter problem is not likely to play a role with regard to generally applicable legislation. Nevertheless a presumption of indirect discrimination is clearly more difficult to demonstrate where statistical evidence is required. Related to the type of proof is the level of disparity that is demanded to establish a case of indirect differential treatment. In other words, it must be determined when a certain effect can be considered ‘disproportionate’. This aspect is particularly relevant where statistical evidence is used or required. As discussed above, the RED demands that persons belonging to a racial or ethnic group would be put ‘at a particular disadvantage’. This term is not further explained. In cases involving differential treatment on the grounds of sex, the CoJ has previously required the percentage of women affected to be ‘considerably’ higher than the percentage of affected men. Additionally, the Court also accepted that there could be indirect differential treatment in the case of a ‘lesser but persistent and relatively constant’ level of disparity over a long period of time.32 While these criteria provide some indications, they are clearly not very specific. As a reference, it could be assumed that a disproportionate effect exists when the percentage of persons belonging to a particular racial or ethnic group that is disadvantaged by a particular measure is at least 1.5 times as high as the percentage of disadvantaged persons not belonging to that group.33 As argued below, however, other comparisons are possible. Alternatively, especially in cases where significant figures are indeterminate or not available, the statistical information may be looked at in combination with qualitative evidence.34 The case law and comments of the other monitoring bodies (the ECtHR, the HRC and the CERD Committee) do not display a clear choice for any of the above standards, let alone a further specification of them. In earlier cases the ECtHR set a very high standard for evidence of indirect discrimination, which made a presumption thereof almost impossible to prove.35 This approach was significantly mitigated, however, in later judgments. In DH and others v the Czech Republic, the ECtHR stated that in order to ‘guarantee those concerned the effective protection of their rights, less strict evidential rules should apply in cases of alleged indirect discrimination’.36 In this case it was explicitly accepted that a presumption of indirect discrimination could be derived from statistical evidence, provided the statistics presented were ‘reliable and significant’.37 As to the level of disparity, the Court considered in DH and others that a presumption of indirect discrimination had been established, considering that the disputed education measures had had ‘considerably more impact’ on children of Roma origin.38 In the same judgments, however, the ECtHR declared that ‘there are no procedural barriers to the admissibility of evidence or pre-determined formulae for its assessment’ and that ‘the level of persuasion necessary for reaching a particular conclusion [is] intrinsically linked to the specificity of the facts, the nature of the allegation made and the Convention right at stake’.39 It follows that the type of proof that can be accepted as evidence of indirect differential treatment and the standard of evidence will vary from case to case. It follows that, in EU law as under Article 14 (and 1 Twelfth Protocol) ECHR, there is scope to assume a presumption of indirect racial or ethnic discrimination in the absence of ‘hard’ statistical evidence proving actual adverse effects on persons of a particular racial or ethnic origin. However, no very specific evidential standards have been formulated, and what is considered sufficient proof may therefore differ from one situation to another. In the view of the ECtHR, such differentiations are also required because of the varying circumstances of each case. With regard to the Racial Equality Directive, Busstra proposes applying lenient standards when determining the existence of a presumption of indirect discrimination so as to maximise the number of situations that may be scrutinised under the directive. She argues that this inclusive interpretation is balanced by the objective justification test, which is also prescribed by the directive and allows for the consideration of interests that may validate the difference in treatment.40 This is a useful approach, which may also be applied in relation to other non-discrimination norms, provided the existence of a possible justification is assessed. The justification test and the scope of review to be applied are discussed in sections II.D and II.E below. In order to establish whether a racial or ethnic group is disproportionately affected by a particular measure, the effects of the measure on that group must be compared to the effects on other groups. This raises the question of which comparison must be made and which groups are to be compared.41 Again, this question primarily arises in cases involving statistical evidence. As far as the identification of relevant groups is concerned, it has been argued that the comparison should be made among all persons addressed by the disputed measure.42 With regard to immigration requirements, however, this criterion does not resolve all the questions that may arise. If the requirement applies to all immigrants seeking admission, it could be assumed that the comparison must be made within this group, by comparing those able to meet the requirement with those unable to do so. Yet, as the Dutch Act on Integration Abroad shows, it may also be the case that immigration requirements apply only to certain categories of immigrants and not to others. If the comparison in such cases is limited to those having to meet the requirement, it will be unable to take account of any indirect differential treatment resulting from the criteria used to define the target group.43 In this type of situation, therefore, the comparison should include all applicants for admission rather than just those who have to meet the requirement.44 Even then, however, the comparison will still not identify the effects on those who may have wished to apply for admission, but who refrained from doing so because they knew or believed that the requirement would stand in their way. To address this omission, the comparison needs also to include all potential immigrants, which in turn can make it more difficult to obtain reliable data.45 Table 3. Establishing a presumption of indirect discrimination
The Right to Equal Treatment:
Indirect Differential Treatment on the
Grounds of Racial or Ethnic Origin
I. INTRODUCTION
II. LEGAL STANDARDS REGARDING INDIRECT DIFFERENTIAL
TREATMENT ON GROUNDS OF RACIAL OR ETHNIC ORIGIN
A. Relevant Legal Provisions
i. (Inter)national Human Rights Provisions
ii. EU Law
B. Discrimination Grounds
C. Indirect Differential Treatment
i. Standard of Proof
ii. Which Comparison to Make?
Secondary education diploma | No secondary education diploma | |
Ethnic group A | 500 | 1,000 |
Ethnic group B | 3500 | 5000 |
On the basis of the above information, it can be established that the admission rate among persons belonging to ethnic group A is one-third (33.3 per cent) whereas two-thirds (66.6 per cent) will be denied admission. On the other hand, 41.2 per cent of the persons belonging to ethnic group B will be admitted and admission will be denied to 58.8 per cent. The next question is whether the difference between the groups is significant (which is not the case if the 1.5-criterion is used, see section II.C.i).
Alternatively, it is possible to compare the percentage of persons belonging to ethnic group A in the overall pool with their share of the group that is admitted. This shows that, although persons belonging to ethnic group A constitute 15 per cent of the overall pool (1500 out of 10000), they only constitute 12.5 per cent of the group that is admitted. By contrast, persons belonging to ethnic group B constitute 85 per cent of the overall pool and 87.5 per cent of the group that is admitted.46 Hence this form of comparison also shows that persons belonging to ethnic group A are relatively more affected by the immigration requirement than those belonging to ethnic group B, thus leaving the question of when a difference is sufficiently significant to prove a presumption of indirect discrimination.
The above comparison is made between only two groups. In reality, however, the ethnic diversity among immigrants seeking admission to a particular state will normally be much greater, which means that group A will not only have to be compared to group B, but also to groups C, D, E and so on. If the latter form of comparison is used, it may turn out that persons belonging to ethnic groups A, B, C and D all constitute approximately 25 per cent of the overall pool, while 5 per cent of the group admitted belongs to group A, 7 per cent to group B, 25 per cent to group C and 63 per cent to group D. In my view, the admission rate of each group in this situation needs to be considered separately in relation to the share they represent of the overall pool. In the example provided here, this would lead to the conclusion that a presumption of indirect discrimination can be established with regard to ethnic groups A and B, but not with regard to ethnic groups C and D.
D. Justification of Indirect Differential Treatment
It follows from the above that establishing a presumption of indirect racial or ethnic discrimination can be a complex undertaking. Once such a presumption has been established, the next question is whether the difference in treatment is nevertheless based on an objective and reasonable justification. This test requires the differential treatment to pursue a legitimate aim and also to be in a reasonable relationship of proportionality to that aim (section II.C of chapter 8). With regard to indirect differential treatment, the requirement for a reasonable and objective justification moreover entails that the measure must not be intended to distinguish between groups of different racial or ethnic origin, even where this is not explicitly put forward as a differentiation ground.47
In cases of indirect differential treatment, the nature of the justification provided may also be different than in situations involving direct differential treatment. The fact that a particular group is disproportionately affected by a certain measure may be due to social circumstances that cannot easily be changed by the legislator (for example, the socio-economic position of some ethnic groups may be worse than that of others, which means they will be hit harder by a measure that imposes a financial burden). In such situations, the question of justification will not primarily concern the aim of the measure or its suitability, but rather whether another measure could have been adopted with less adverse effects on the group concerned and whether the advantage gained through the measure is of sufficient importance to outweigh these effects.48
With regard to the prohibition of racial discrimination in the CERD, it is observed that a justification test is not foreseen in the Convention. Article 1(1) CERD constitutes a ‘closed provision’ (section II.C of chapter 8), which implies that differences in treatment falling under its definition of racial discrimination are prohibited unless they are covered by one of the exceptions of Article 1(2)–(4). Nevertheless, the CERD Committee has accepted that differential treatment does not amount to racial discrimination if it is justified in relation to the objectives and purposes of the Convention.49
Although the Committee’s interpretation goes against the wording of Article 1(1) CERD, its approach is not without its merits. In the absence of a possibility for justification, the definition of racial discrimination can be expected to be interpreted in a restrictive way so as to preclude reasonable and legitimate state actions from falling foul of the prohibition laid down in the Convention. Such a restrictive interpretation would, however, necessarily limit the scope of the Convention, with the result that protection against racial discrimination may not always be available even if the action concerned is not based on a reasonable and objective justification. The introduction of a justification test thus allows for a more refined examination in cases where a difference in treatment is based on grounds of racial or ethnic origin.50 It is therefore submitted that the CERD Committee’s approach is defensible in view of the purpose of the Convention, which is to eliminate racial discrimination in all its forms and manifestations.51 It may also be noted that the acceptance of the possibility of justification is in line with the approach taken by the ECtHR with regard to racial discrimination under Article 14 ECHR and with the definition in the Racial Equality Directive.52
Despite accepting a justification test, the CERD Committee has not yet provided detailed criteria for determining whether, in a particular case, a difference in treatment is justified in relation to the objectives and purposes of the Convention. However, the Committee has on occasions stated that the criteria for differentiation must be applied pursuant to a legitimate aim and that they must be proportional to the achievement of that aim.53
E. Scope of Review
The issue of the scope of review in equal treatment cases was already addressed in a general manner in chapter 8 (section II.D), when it was established that the scope of review is determined by various factors, specifically including the differentiation grounds at stake. Differences in treatment that are based on grounds of racial or ethnic origin are generally held to require strict scrutiny.54 This view finds confirmation in the case law of the ECtHR, which has held on various occasions that such differentiations constitute a violation of Article 14 ECHR, unless they are based on a very strong justification. In several cases, the Court stated that where differences in treatment are based on race, colour or ethnic origin, ‘the notion of objective and reasonable justification must be interpreted as strictly as possible’.55 Another formulation used by the Court is that ‘no difference in treatment which is based exclusively or to a decisive extent on a person’s ethnic origin is capable of being objectively justified in a contemporary democratic society built on the principles of pluralism and respect for different cultures’.56 Although the latter criterion appears even more stringent than the former, the Court itself does not seem to make a clear distinction as, in several judgments, both criteria are used alongside each other.57 Nevertheless it is clear that the Court considers differentiations based on race or ethnic origin to be a particularly serious form of discrimination, which can even amount to inhuman or degrading treatment.58
The importance of the prohibition of racial discrimination is also stressed by Article 4(1) ICCPR, which states that this prohibition is one of the norms from which no derogation is possible in times of emergency. Arguably, this supports the view that differentiations on the grounds of race or ethnic origin should in principle be subject to intense scrutiny. The HRC has also stated that its review will be stricter when a differentiation is based on one of the grounds mentioned in Article 26 ICCPR, which include race, colour and national origin (section II.D of chapter 8).
Nevertheless, it is submitted that the scope of review should also be determined in relation to the nature of the differential treatment and that alleged indirect discrimination on grounds of race or ethnic origin will not always require an equally strong justification as when the difference in treatment is directly based on these grounds.59 One important reason underlying the prohibition of racial or ethnic discrimination is the wish to combat racism or racist ideologies and the acts or practices resulting from them.60 As Howard explains, the concept of racism is ‘based on the belief that some races are superior to others’.61 Busstra similarly relates the term racism to the belief that persons belonging to a certain group are ‘unreliable’ or ‘less intelligent’ than others.62 Racist ideologies and practices are therefore inherently contrary to the principles of equality and human dignity that lie at the basis of human rights instruments in general and the right to equal treatment in particular.63