Facing the Future: Substantive Equality Under the Spotlight

LEGAL RESEARCH PAPER SERIES Paper No 57/2010 July 2010 Facing the Future: Substantive Equality under the Spotlight SANDRA FREDMAN This paper can be downloaded without charge from the Social Science Research Network electronic library at: http://ssrn.com/abstract=1649991 An index to the working papers in the University of Oxford Legal Research Paper Series is located at: <http://www.ssrn.com/link/oxford-legal-studies.html> Electronic copy available at: http://ssrn.com/abstract=1649991 Facing the future: Substantive Equality under the Spotlight Sandra Fredman Oxford University Published in: EQUALITY IN THE WORKPLACE: REFLECTIONS FROM SOUTH AFRICA AND ELSEWHERE Edited by Ockert Dupper and Christoph Garbers, University of Stellenbosch, South Africa (2010) 1. Introduction Transformation in South Africa is intimately bound up with a vigorous development of the principle of equality. However, as the second decade of the new democracy draws to a close, the challenges facing the equality principle are more complex than ever. Equality before the law, although a sine qua non, is far from sufficient to bring about the far-reaching social and institutional changes necessary to achieve true equality. In addition, to ensure that the constitutional project of transformation retains its energy into the next decade, the promise of substantive equality must be fulfilled. Yet there is still much debate as to what that promise entails and the best means to achieve it. Nor are such challenges confined to South Africa. Although the South African experience has been condensed into a mere decade and a half, the evolution of substantive equality has been shared in many other societies, and the modern dilemmas are strikingly similar. The aim of this paper is to frame some of the key challenges facing the principle of substantive equality. Central to substantive equality is the recognition of the connection between status and disadvantage, where status refers to race, gender, disability or other prohibited ground, and disadvantage is concerned with socioeconomic disadvantage. Whereas formal equality addresses discrimination based on status, substantive equality highlights the fact that it is not status per se which is problematic, but the disadvantage which attaches to status. Thus the focus should be on women or blacks, rather than on gender or race per se. However, the power of this Electronic copy available at: http://ssrn.com/abstract=1649991 insight is matched by the challenges it raises. In addressing both status and socioeconomic disadvantage within the vehicle of substantive equality, how is each to be formulated, and how should they interact? This paper considers three specific problematic issues which arise from this interaction. The first is the role of dignity. An increasingly popular concept within substantive equality discourse, dignity has also at times had the effect of dislocating status from socio-economic disadvantage. The paper asks what role dignity should play. The second problematic issue at the intersection of status and socio-economic disadvantage is concerned with affirmative action. Affirmative action specifically attaches socio-economic benefits to those disadvantaged by status, and therefore facilitates the achievement of substantive equality. However, there is a risk that this link perpetuates status inequality by freezing individuals into existing status groups, or by reinforcing stereotypes. This paper considers this problem particularly in relation to pregnancy and parenthood. Furthermore, as the status group begins to benefit from the affirmative action provisions, a gap opens up between status and disadvantage. The paper asks the question of whether the beneficiary group therefore needs to be narrowed to include only those who suffer both status and socio-economic disadvantage. Indeed, if the aim is indeed to redress socio-economic disadvantage, why link affirmative action to status at all? It is argued here that affirmative action performs more functions than simply to redress socio-economic disadvantage, and therefore that the link with status should be maintained, even when accompanied by relative economic prosperity. This leads onto the third problematic issue at the intersection between status and socio-economic disadvantage, namely the link between status and poverty. It is argued here that just as status brings with it socioeconomic disadvantage, so poverty brings with it status harm. This means that poverty should be a ground of discrimination. Another way of formulating the issue is to use the concepts of recognition and redistribution, fashioned by Fraser, Honneth and others.1 Based on Hegel‟s notion of 1 N Fraser and A Honneth Redistribution or Recognition (Verso London, New York 2003);S Fredman ' „Redistribution And Recognition: Reconciling Inequalities‟ ' (2007) 23 South African Journal of Human Rights 214 ; S Liebenberg 'Needs, Rights and Transformation: Adjudicating Social Rights' Center For Human Rights And Global Justice Working Paper Economic And Social Rights Series Number 8, 2005 ; J Fudge ''The Canadian Charter of Rights: Recognition, Redistribution, and the Electronic copy available at: http://ssrn.com/abstract=1649991 identity as inter-subjective recognition, or inter-personal affirmation, recognition is the value attached to status. Status inequality arises out of mis-recognition, which in turn comes from denigrating or humiliating individuals. Distributive inequality arises out of unfair allocation of resources. Whereas distributive inequality is usually dealt with through the welfare state, and recognition inequality through discrimination law, the power of substantive equality is its ability to bring both together.2 However, it remains the case that substantive equality is better at addressing status wrongs than socio-economic disadvantage, unless it is constructively allied with socio-economic rights.3 The paper begins with a brief description of the meaning of substantive equality. It then moves on to consider the three challenges above: dignity, affirmative action, and poverty as a ground of discrimination. I do not presume to provide answers to these questions, but merely to elucidate the nature of the challenge, and show how various jurisdictions have responded. 2. Substantive equality Substantive equality is born out of disappointment and frustration at the limits of formal equality. It aims to take up the baton where formal equality leaves off. However, while it is by now relatively easy to reach consensus on the limits of formal equality, there remain many unanswered questions about the ways in which substantive equality should be shaped to transcend these limitations. This section briefly rehearses the limitations of formal equality, before turning to the challenges faced in defining substantive equality. Formal equality is based on the premise that individuals should be treated as individuals, on the basis of their own merit, rather than on attributions based on irrelevant characteristics such as race, colour, gender, caste or other analogous status. Imperialism of the Courts'' in T CampbellKD Ewing and A Tomkins (eds) Sceptical Essays on Human Rights (Oxford University Press Oxford 2001) 2 S Fredman ' „Redistribution And Recognition: Reconciling Inequalities‟ ' (2007) 23 South African Journal of Human Rights 214 3 S Fredman ''Providing Equality: Substantive Equality and the Positive Duty to Provide'' (2005) 21 SAJHR 163 It is usually summed up by the Aristotelian formula that likes should be treated alike. In legal contexts, it finds expression in the principle of direct discrimination or equal treatment, which makes it unlawful to treat a person less favourably on grounds of her gender, race or other status than a person of a different gender, race or other status. Its achievements should not be understated. It has taken many years of political struggle to reach the point of recognition that gender and race are irrelevant criteria for key political and civil rights. Political thinkers from Aristotle onwards regarded gender as a relevant criterion for access to citizenship rights, ascribing only to men the rationality required to qualify as a subject of rights4. It hardly needs to be stated that the key initial achievement of the new democracy in South Africa was to remove all racial references from the criteria for access to basic rights. However, experience has quickly demonstrated that inequalities persist despite the introduction of formal equality. This can be attributed to five main characteristics of this concept. The first is the assumption that the individual can be abstracted from her gender, race or other status and dealt with entirely on merit. However, merit is itself a function of previous advantage, rather than an objective characteristic. Treating status as an irrelevant ground merely ignores the ongoing disadvantage experienced by individuals who have been previously been unequal before the law or subject to social prejudice. 5 The result is to entrench disadvantage. Secondly, formal equality assumes that the aim is identical treatment. In fact, given antecedent inequality, it may be necessary to provide very different sets of resources to achieve genuine equality. As Sen has argued: „Equal consideration for all may demand very unequal treatment in favour of the disadvantaged. The demands of substantive equality can be particularly exacting and complex when there is a good deal of antecedent inequality to counter.‟6 Thirdly, formal equality is premised on an abstract and universal individual. However, the possibility of an individual abstracted from her status characteristics is illusory. In reality, the abstract individual is clothed with the characteristics of the dominant group, which are then asserted as if they were universal. Only those who can conform 4 S Fredman Women and the Law (Oxford Monographs in Labour Law Oxford University Press Oxford 1997) chap 1. 5 See further, S Fredman Discrimination Law (2002) chapters 1 and 5; C McCrudden „ “The Merit Principle”‟ (1998) 18 Oxford J of Legal Studies 543. 6 A Sen Inequality Re-examined (Oxford University Press Oxford 1992). to this norm are sufficiently „alike‟ to be entitled to „like treatment.‟ The result is that formal equality demands conformity as a price for equal treatment. Yet diverse individual identities may be enriching and desired. The problem is not the diversity of characteristics, but the detrimental treatment attached to it. Thus the aim should not be to eliminate difference, but to prohibit the detriment attached to such difference. Fourthly, formal equality is a relative concept, agnostic as to the substantive outcome. It can therefore be fulfilled by treating everyone equally badly, or by removing benefits from the better off in order to bring them in line with the worse off. This means that the achievement of formal equality can be a hollow victory, or even a defeat. Finally, formal equality is based on a negative conception of liberty, aiming to restrain the state from interfering with individual rights, rather than placing positive obligations on the State to promote equality. Substantive equality is fashioned to address each of these limitations. Firstly, substantive equality does not aim to abstract the individual from the social context. It takes into account existing power structures and the role of status or identity within them. Status is not regarded as irrelevant. Indeed, it may be highly relevant in addressing inequalities in society. Substantive equality therefore focuses on the disadvantage associated with a prohibited ground rather than the ground itself: on women rather than gender; on blackness rather than colour. Secondly, and as a result of this, substantive equality is sensitive to outcomes rather than just to treatment. This opens the way to requiring different treatment in order to achieve equalities of outcome. Thirdly, substantive equality recognises that identity can be a source of value. It does not therefore aim to treat all individuals identically, but to affirm and accommodate differences. In its transformative form, substantive equality requires social institutions to change, rather than expecting the individual to conform. Fourthly, substantive equality is not neutral as to the outcome. Equality cannot be achieved by treating all equally badly, or by removing benefits from the advantaged class. It is substantive in the sense that it advances individuals rather than formal in ensuring only consistency. Finally, substantive equality can go beyond a fault based model to one which includes positive duties to respect, protect, promote and fulfil. Framing substantive equality as a response to the limits of formal equality has achieved considerable consensus. However, the specific content of substantive equality remains contested. There is a temptation to reduce substantive equality to a single dimension, such as dignity, or socio-economic disadvantage. I argue instead that it has four different dimensions.7 First, it is an asymmetric principle. Rather than the abstract individual of formal equality, substantive equality focuses on disadvantage. It aims therefore is to break the cycle of disadvantage associated with status or out-groups. This is the redistributive dimension. Secondly, substantive equality promotes respect for the equal dignity and worth of all, thereby redressing stigma, stereotyping, humiliation, and violence because of membership of an outgroup. Sexual harassment, racist abuse, humiliation of old people, and homophobic bullying at schools are all examples of situations in which the primary claim is based on dignity. Thirdly, it does not exact conformity as a price of equality. Instead, it entails an accommodation and positive affirmation and celebration of identity within community. The second and third are dimensions of recognition. Finally, out-groups are under-represented in decision-making at all levels. Many are also socially excluded. Substantive equality should therefore facilitate full participation in society. This four-dimensional understanding of substantive equality leaves open the question of how the dimensions relate to each other. The rest of the paper attempts to throw some light on this interaction. 3. Dignity The desire to find a substantive core to prevent equality from being no more than an exercise in consistency has led many academics, practitioners, and judges to frame dignity as the central equality value. Thus the Canadian Supreme Court, rejecting a formal equality approach, located dignity at the centre of the equality principle. „Equality means that our society cannot tolerate legislative distinctions that treat certain people as second class citizens, that demean them, that treat them as less capable for no good reason, or that otherwise offend fundamental human dignity‟.8 Similarly, according to Chaskalson P : „Inequality is established not simply through 7 S Fredman The Future of Equality in Great Britain (Working Paper No.5 Equal Opportunities Commission Manchester 2002) 8 Law v Canada [1999] 1 SCR 497 (Canadian Supreme Court) para 51. group-based differential treatment, but through differentiation which perpetuates disadvantage and leads to the scarring of the sense of dignity and self-worth.‟9 There is much that makes dignity an intuitively appealing concept. To heal the scars of apartheid requires, first and foremost, an affirmation of each person‟s humanity, prohibiting the dehumanisation and degradation of anyone. Dignity has also been used specifically to bring sexual harassment into the fold of equality. Whereas formal equality engages in fatuous attempts to find an appropriate comparator, substantive equality simply prohibits sexual harassment because it is inconsistent with respect for a woman‟s basic dignity and humanity. Thus under EU law, harassment is a species of discrimination, where harassment is defined as unwanted conduct with the purpose or effect of violating the dignity of a person and of creating an intimidating, hostile, degrading, humiliating or offensive environment.10 Dignity also plays a central role in preventing the principle of equality from becoming no more than a levelling down exercise. To implement equality by removing benefits, or treating everyone equally badly, would be inconsistent with the fundamental dignity of each individual. The importance of this can be seen by comparing the experience in the EU and South Africa over the question of equalising pension ages. The lower pension age of 60 for women, as against 65 for men, has been contested in both jurisdictions as discriminatory against men. Although older women remain among the poorest in both Europe and South Africa, increasing unemployment for male workers above 50 has meant that access to pension rights at an earlier age is of growing importance to men. In a case brought initially to the UK courts, and subsequently, under EU law, to the ECJ, this claim of discrimination was upheld.11 However, pension funds, claiming that the cost would be exorbitant, reacted by instituting policies to raise women‟s pension age to that of men over time. This strategy has been upheld in further litigation before the ECJ.12 The result is that poor women are worse off, and poor men are no better off.13 In South Africa, by contrast, 9 A Chaskalson „Human Dignity as a Constitutional Value‟ in D Kretzmer & E Klein (eds) The Concept of Human Dignity in Human Rights Discourse (2002) 140. 10 EP & Council Directive 2006/54/EC Article 2(1)(c) and Article 2(2)(a); for the similarly worded definition of racial harassment see Council Directive 2000/43/EC Article 2(3) 11 Barber v Guardian Royal Exchange Assurance Group [1990] IRLR 240 (ECJ) 12 C-408/92 Smith v Avdel Systems Ltd [1994] ECR I-4435 (ECJ) 13 S Fredman ''The Poverty of Equality: Pensions and the ECJ'' [1996] 25 ILJ 91 - 109. when litigation was contemplated on the same issue, the risk of levelling down could be averted by reference to the substantive content of equality, including dignity. Litigation was, in the event, rendered unnecessary the welcome intervention by political authorities in the form of legislative change. Even here, however, a substantive view was taken. In providing for age equalisation for state old age grants, there was no question of increasing women‟s pension age. Instead, the Social Assistance Amendment Bill opens the way for men aged 60 to 64 years to apply for the old age grant, potentially benefitting over 450,000 men. Dignity, however, carries risks. Most serious is the danger that it might operate to negate the link between socio-economic disadvantage and substantive equality. This manifests itself by requiring the claimant to prove, not just that she has been disadvantaged, but that this signifies lack of respect of her as a person. A measure which imposes socio-economic disadvantage on individuals on grounds of their status would not in itself be regarded as discriminatory unless the applicant could prove, in addition, that the measure assails her dignity. This danger has been floridly demonstrated in two recent cases in the Supreme Court of Canada, in which the Court agreed that the claimant suffered socio-economic disadvantage on grounds of her status (in this case, her age), but held that this did not signify that society regarded her of less value than others.14 Her discrimination claim therefore failed. This was particularly problematic in Gosselin,15 where the claimant challenged a scheme according to which full benefit was only available to welfare recipients over 30. Those under 30 received significantly less unless they participated in a designated work activity or education programme. In practice, there was a significant shortfall in places available, and those which were available were relatively short-term. As a result, many young people, including the claimant, experienced real poverty. Nevertheless, the majority of the Court held that „the provision of different initial amounts of monetary support to each of the two groups does not indicate that one group's dignity was prized above the other‟s.‟16 It therefore rejected her claim of breach of the equality guarantee in s.15(1) of the Charter. 14 Law v Canada [1999] 1 SCR 497 (Canadian Supreme Court); Gosselin v Quebec 2002 [SCC] 84 (Canadian Supreme Court) 15 Gosselin v Quebec 2002 [SCC] 84 (Canadian Supreme Court) 16 Gosselin v Quebec 2002 [SCC] 84 (Canadian Supreme Court) para 61 (McLachlin J). A more subtle manifestation of this effect of dignity is found in the Hugo case17, where, again, dignity was used to defeat an equality claim. The case concerned the pardon issued by President Mandela to all women prisoners who were mothers of young children. The pardon was challenged by a male prisoner, the sole carer of his young children, on the basis that it discriminated on grounds of gender. The Court rejected the case. According to Goldstone J, „The Presidential Act might have denied fathers an opportunity it afforded mothers, but it could not be said to have fundamentally impaired their rights of dignity or sense of equal worth.‟18 Yet the opportunity denied to fathers was one which goes to the heart of the gendered division of labour in our society, and therefore to the heart of the socio-economic disadvantage experienced by women. Women‟s role as primary child-carers has been a major source of inequality for women; to attempt to resolve the problem on the basis that men‟s dignity had not been impaired merely sidesteps the reality of gender inequality. Equally problematic is the assertion that the measure did not impair fathers‟ rights of dignity or sense of self worth. As Kriegler J stated in his dissenting opinion, it is the assumption that women are the primary child-carers which constitutes an assault on their dignity: „One of the ways in which one accords equal dignity and respect to persons is by seeking to protect the basic choices they make about their own identities. Reliance on the generalisation that women are the primary care givers is harmful in its tendency to cramp and stunt the efforts of both men and women to form their identities freely…‟19 Dignity can also be used in a way which abstracts the equality analysis from power relations in society. This can be seen in the early case of Walker20. The case concerned a charging system for electricity which favoured the predominantly black and poor residents of the former townships outside Pretoria over the predominantly white and affluent residents of Pretoria itself. The court held that although the charging system itself did not constitute an invasion of the dignity of the white residents, the selective enforcement of the charges constituted unfair discrimination against them. According to Langa DP: „No members of a racial group should be made to feel that they are not deserving of equal „concern, respect and consideration‟ and 17 Case CCT 11/96 President v Hugo Hugo para 47 19 Hugo para 80 20 City Council of Pretoria v Walker 1998 (2) SA 363 (CC) (South African Constitutional Court) 18 that the law is likely to be used against them more harshly than others who belong to other race groups‟.21 These difficulties are not, however, insurmountable. The Canadian Court itself has now recognised that dignity has been used in a problematic way. In R v Kapp,22 the Court acknowledged that „several difficulties have arisen from the attempt in Law to employ human dignity as a legal test. There can be no doubt that human dignity is an essential value underlying the s. 15 equality guarantee. In fact, the protection of all of the rights guaranteed by the Charter has as its lodestar the promotion of human dignity. …But as critics have pointed out, human dignity is an abstract and subjective notion that…has .. proven to be an additional burden on equality claimants, rather than the philosophical enhancement it was intended to be.‟23 More than this, however, dignity should be used in alliance with socio-economic rights to create a powerful means of upholding substantive equality. This can be seen in the judgement of Mokgoro J in Khosa,24 which concerned the exclusion of permanent residents from the right to child benefit and old-age pensions. Mokgoro J emphasised that the consequences of exclusion were not only socio-economic. In addition, the exclusion of permanent residents had a strong stigmatizing effect, creating the impression that they were inferior to citizens and less worthy of social assistance. Permanent residents were in effect „relegated to the margins of society and deprived of what may be essential to enable them to enjoy other rights vested in them under the Constitution‟.25 Thus dignity is the dimension of equality which speaks to our basic humanity. Equality attaches to all individuals, not because of their merit, or their rationality, or their citizenship or membership of any particular group, but because of their humanity. Individuals should not be humiliated or degraded through racism, sexism, violence or other status-based prejudice. But dignity is not a separate and additional 21 Walker para 81 R. v. Kapp 2008 SCC 41 (Supreme Court of Canada) 23 Kapp para 21, 22 24 Khosa and Mahlaule v Minister for Social Development 2004 (6) BCLR 569 (South African Constitutional Court) 25 Khosa para 77. 22 element to socio-economic disadvantage in an equality claim. Socio-economic disadvantage is itself an assault on an individual‟s basic humanity. The recognition and redistributive elements of equality should pull together rather than against each other. 3. Affirmative action (i) Affirmative action and substantive equality One of the main advantages of substantive equality over formal equality is its asymmetry. This means that it is not race or gender per se which is regarded as problematic, but the detriment and disadvantage associated with subordinated groups: black people, women, or other out-groups. This asymmetry means that equality is not necessarily breached by measures which specifically use race or gender as a means of distributing benefits and burdens. Indeed, provided that they aim to benefit the subordinated group, race or gender specific measures may be necessary to achieve substantive equality. Thus, whereas formal equality would regard affirmative action as a breach of equality, substantive equality sees such programs as a means to achieve equality. 26 This understanding of affirmative action has been endorsed by the South African Constitution, which makes express provision for affirmative action as a means of achieving substantive equality27. According to section 9(2): „Equality includes the full and equal enjoyment of all rights and freedoms. To promote the achievement of equality, legislative and other measures designed to protect or advance persons, or categories of persons, disadvantaged by unfair discrimination may be taken.’ The meaning of this provision was elaborated in Van Heerden,28 a claim of race discrimination brought by a white Afrikaner member aggrieved at a measure which enhanced the pension contributions of post apartheid members of Parliament but not pre-apartheid members. The High Court took an emphatically formal view of 26 S Fredman Women and the Law (Oxford Monographs in Labour Law Oxford University Press Oxford 1997); S Fredman Discrimination Law (Clarendon Law Series Oxford University Press Oxford 2002); S Fredman „Reversing Discrimination‟ [1997] 113 Law Quarterly Review 575–600. 27 South African Constitution, s.9(2). 28 Minister of Justice v Van Heerden 2004 (6) SA 121 (CC); 2004 (11) BCLR 1125 (South African Constitutional Court) equality, regarding the relatively advantaged position of the affected white members as irrelevant. It therefore struck down the programme as unfair discrimination. The Constitutional Court reversed the decision. As Moseneke J stressed, instead of being an exception to equality, restitutionary measures are an essential part of it. „What is clear is that our Constitution and in particular section 9 thereof, read as a whole, embraces for good reason a substantive conception of equality inclusive of measures to redress existing inequality. ..Such measures are not in themselves a deviation from, or invasive of, the right to equality guaranteed by the Constitution. They are not “reverse discrimination” or “positive discrimination” as argued by the claimant in this case. They are integral to the reach of our equality protection. In other words, the provisions of section 9(1) and section 9(2) are complementary; both contribute to the constitutional goal of achieving equality to ensure “full and equal enjoyment of all rights.”‟ 29 Similarly, the Canadian Charter makes it clear that measures whose object is to ameliorate the conditions of disadvantaged individuals or groups will not be a breach of the equality guarantee in s.15(1) of the Charter. Thus section 15(2) states: „Subsection (1) does not preclude any law, program or activity that has as its object the amelioration of conditions of disadvantaged individuals or groups including those that are disadvantaged because of race, national or ethnic origin, colour, religion, sex, age or mental or physical disability.‟30 While there was some suggestion in earlier cases that section 15(2) should be read as an exception to the equality guarantee, the Supreme Court of Canada in its 2008 decision in Kapp,31 emphatically held that „Sections 15(1) and 15(2) work together to promote the vision of substantive equality that underlies s. 15 as a whole. Section 15(1) is aimed at preventing discriminatory distinctions that impact adversely on members of groups identified by the grounds enumerated in s. 15 and analogous grounds. This is one way of combating discrimination. However, governments may also wish to combat discrimination by developing programs aimed at helping disadvantaged groups improve their situation. Through s. 15(2), the Charter preserves the right of governments to implement such programs, without fear of challenge under s. 15(1). This is made apparent by the 29 Minister of Justice v Van Heerden 2004 (6) SA 121 (CC); 2004 (11) BCLR 1125 (South African Constitutional Court) para 30. 30 Canadian Charger of Rights s. 15(2) 31 R. v. Kapp 2008 SCC 41 (Supreme Court of Canada) existence of s. 15(2). Thus s. 15(1) and s. 15(2) work together to confirm s. 15‟s purpose of furthering substantive equality.‟32 Moreover, „Section 15(2) supports a full expression of equality, rather than derogating from it.‟33 The same approach, albeit within stricter parameters, can be found in EU law, where several provisions state that, with a view to ensuring full equality in practice, the principle of equal treatment should not prevent a member state from maintaining or adopting measures to prevent or compensate for disadvantages linked to gender34, race or ethnic origin,35 or religion, disability, age and sexual orientation.36 At international level, affirmative action is again endorsed. Thus the Convention on the Elimination of Discrimination Against Women (CEDAW) states that adoption by States Parties of temporary special measures aimed at accelerating de facto equality between men and women „shall not be considered discrimination‟; and Article 2(2) of International Convention on the Elimination of Race Discrimination (CERD), which states: „States Parties shall, when the circumstances so warrant, take, in the social, economic, cultural and other fields, special and concrete measures to ensure the adequate development and protection of certain racial groups or individuals belonging to them, for the purpose of guaranteeing them the full and equal enjoyment of human rights and fundamental freedoms.37 However, there remain important challenges for affirmative action measures. At one level, affirmative action constitutes a valuable synthesis between status and disadvantage, specifically attaching socio-economic benefits to those disadvantaged by status. However, there is a risk that this link perpetuates status inequality. It is not necessarily the case that a programme singling out one group for special protection will in fact advance substantive equality. It may freeze individuals into the very status identity which substantive equality aims to eliminate. It may perpetuate stereotypes. This is particularly true for special measures in favour of women in their child-caring role. Secondly, as the status group begins to benefit from the affirmative action provisions, a gap opens up between status and disadvantage; and the class of 32 Ibid para 16 Ibid para 37 34 Article 141(4) EU 35 Directive 2000/43 EC Article 5 36 Directive 2000/78/EC Article 7(1) 37 CERD Article 2(2) 33 beneficiaries identified by the measure might become over- or under-inclusive. In particular, as the association between race and disadvantage loosens, it may become necessary to include a further criterion, such as that known as the „creamy layer‟ in India, which excludes status members who are no longer socio-economically disadvantaged. Both these questions in turn require a deeper understanding of the relationship between status and socio-economic disadvantage. Behind these issues is the third challenge, namely the extent to which affirmative action can be genuinely transformative, in that it brings about structural change, rather than simply changing the colour or gender composition of classes within the existing structure. These challenges in turn raise questions as to the role of courts, and in particular, the standard of scrutiny which courts should apply in order to determine whether affirmative action measures breach a constitutional equality guarantee. Since these challenges are often fought out in the judicial arena, it is necessary to consider the standard of scrutiny before turning to the challenges themselves. (ii) Standard of Scrutiny Although substantive equality is not breached in principle by affirmative action measures, there remains an important role for judicial supervision. The controlling mechanism used by most courts is that of proportionality: are the aims legitimate, and do the means „fit‟ the aims. However, courts vary widely in the intensity of scrutiny. This is true both for the definition of a „legitimate purpose‟ and for the tightness of the fit between the affirmative action measure and that purpose. This variation is accounted for both by the court‟s own understanding of equality and the extent to which they consider it appropriate to defer to legislative or executive decisionmakers‟ understanding of equality. As is well known, the US courts have in recent years adopted a standard of „strict scrutiny‟.38 Based on an underlying hostility to the substantive equality values behind affirmative action, US courts have framed the proportionality analysis in the narrowest terms. Thus instead of simply requiring a legitimate State interest, the State must demonstrate a „pressing social need‟. Similarly, it is not sufficient for the means 38 Adarand Constructors v Pena 515 U.S. 200 (1995) (United States Supreme Court) to be reasonably related to these ends. Instead, the measure must be „narrowly tailored‟ to that end. Nevertheless, this standard does not revert to a purely formal approach. As O‟Connor J has emphasised, „strict scrutiny‟ does not amount to „fatal in fact‟.39 There are two interests which have been recognised as compelling under the strict scrutiny test. The first is the compelling interest of remedying the effects of past intentional discrimination.40 The second is the interest in diversity.41 Building on the well-known suggestion of Powell J in the Bakke42 case that diversity could be a compelling interest, the Court in Grutter upheld a race conscious policy by the University of Michigan on the grounds that it furthered the legitimate aim of diversity in higher education. Grutter also elaborated on the standards to be applied for an affirmative action or race conscious programme to be „narrowly tailored‟. It is not legitimate to impose inflexible quotas; instead, to fulfil the test of strict scrutiny, a programme must be flexible enough to ensure that each applicant is evaluated as an individual and not in a way that makes race or ethnicity the defining feature of the application. Thus race or ethnicity may be considered, but only as a “ „plus‟ in a particular applicant's file.‟43 At the same time, the Court rejected the argument that the Law School should have used race-neutral means to obtain the educational benefits of student body diversity. „Narrow tailoring does not require exhaustion of every conceivable race-neutral alternative or mandate that a university choose between maintaining a reputation for excellence or fulfilling a commitment to provide educational opportunities to members of all racial groups.‟44 The standard of strict scrutiny is nevertheless clearly inconsistent with a genuine adherence to the principle of substantive as against formal equality. Does this mean, however, that Courts should entirely abdicate their supervisory function? The Canadian Court in Kapp took the view that review should be primarily directed to the genuineness of the government‟s stated ameliorative purpose. As will be recalled, section 15(2) states that a programme does not breach the equality guarantee if it has 39 Adarand Constructors v Pena 515 U.S. 200 (1995) (United States Supreme Court) at 237 Parents Involved in Community Schools v. Seattle School Dist. No. 1 127 S.Ct. 2738 (2007) (US Supreme Court) at 2752. 41 Grutter v. Bollinger 539 U.S. 306; 123 S.Ct. 2325 (2003) (US Supreme Court) at 328. 42 Regents of the University of California v Bakke 438 U.S. 265 (1978) (United States Supreme Court) 314-315. 43 Grutter 539 US at 317. 1 44 Grutter 539 US at 309. 40 „as its object the amelioration of conditions of disadvantaged individuals or groups‟. The Court rejected the argument that judicial review should focus on the outcome of a measure, since results might be unpredictable or difficult to evaluate. Instead, as required by the text of the provision, the focus should be on the purpose. Nor did amelioration need to be the sole purpose, as long as it constituted a genuine purpose. So far as the fit between means and ends was concerned, the Court required no more than a correlation between the program and the purpose. „Analysing the means employed by the government can easily turn into assessing the effect of the program. As a result, to preserve an intent-based analysis, courts could be encouraged to frame the analysis as follows: Was it rational for the state to conclude that the means chosen to reach its ameliorative goal would contribute to that purpose? For the distinction to be rational, there must be a correlation between the program and the disadvantage suffered by the target group. Such a standard permits significant deference to the legislature but allows judicial review where a program nominally seeks to serve the disadvantaged but in practice serves other, non-remedial objectives.‟ 45 There is force in the Canadian courts‟ concern with the purpose of a programme rather than its effects. The Kapp case concerned the federal government‟s strategy to enhance aboriginal involvement in commercial fishing. As part of the strategy, the government had issued a communal fishing licence to three aboriginal bands permitting only fishers designated by the bands to fish for salmon in the mouth of the Fraser River for a period of 24 hours and to sell their catch. The appellants, commercial fishers, mainly non-aboriginal, were excluded from the fishery during this 24-hour period. They argued that the communal fishing licence discriminated against them on the basis of race. The trial judge found upheld their claim. In the Supreme Court of Canada , Mr Kapp argued that the program did not offer a benefit that effectively tackled the problems faced by these bands. The Court was rightly reluctant to adjudicate on the effectiveness of the programme in this sense. As it stated, „If the sincere purpose is to promote equality by ameliorating the conditions of a disadvantaged group, the government should be given some leeway to adopt innovative programs, even though some may ultimately prove to be unsuccessful. The 45 Kapp para 49 government may learn from such failures and revise equality enhancing programs to make them more effective.‟46 Nevertheless, it was alive to the fact that courts have regarded a programme as ameliorative in surprising circumstances. Particularly worrying are measures which place restrictions on individuals which the Government claims to be in their best interests.47 Thus the Court drew a careful line between deference to the judgement of the executive, and appropriate judicial supervision. „The meaning of “amelioration” deserves careful attention in evaluating programs under s. 15(2). We would suggest that laws designed to restrict or punish behaviour would not qualify for s. 15(2) protection. Nor, as already discussed, should the focus be on the effect of the law. This said, the fact that a law has no plausible or predictable ameliorative effect may render suspect the state‟s ameliorative purpose. Governments, as discussed above, are not permitted to protect discriminatory programs on colourable pretexts.‟48 The South African Constitutional Court, while rejecting the strictness of „narrow tailoring‟, has nevertheless retained for itself a role which goes beyond checking only for genuine ameliorative purpose. Instead, it requires a measure to meet the standard of reasonableness. Although the constitutional text also talks of measures „designed to protect and advance‟, in van Heerden, Moseneke J stated that the remedial measures must be „reasonably capable of attaining the desired outcome‟, namely to protect or advance individuals or categories of persons who have been disadvantaged by unfair discrimination. This excludes measures which are arbitrary, capricious or display naked preference, or are not reasonably likely to achieve the end of advancing or benefiting the interests of those who have been disadvantaged by unfair discrimination. 49 Importantly, he emphasised that the Constitution did not „postulate a standard of necessity between the legislative choice and the governmental objective. The text requires only that the means should be designed to protect or advance. It is sufficient if the measure carries a reasonable likelihood of meeting the end.‟50 46 Kapp para 47 Kapp para 53 48 Kapp para 54 49 Minister of Justice v Van Heerden 2004 (6) SA 121 (CC); 2004 (11) BCLR 1125 (South African Constitutional Court) para 38 - 40 50 Van Heerden para 41 47 This can be contrasted with the standard of scrutiny applied by the European Court of Justice, which continues to regard affirmative action as a derogation from the principle of equality, and therefore to be construed strictly. According to the ECJ: „In determining the scope of any derogation from an individual right such as the equal treatment of men and women laid down by the Directive, due regard must be had to the principle of proportionality, which requires that derogations must remain within the limits of what is appropriate and necessary in order to achieve the aim in view and that the principle of equal treatment be reconciled as far as possible with the requirements of the aim thus pursued.‟51 Ultimately, the role of the court should be to further substantive equality, and therefore to support government measures which aim to achieve substantive equality. There is no equivalence between classification which perpetuates disadvantage or causes detriment, and measures which use status in order to achieve substantive equality. However, judicial support for affirmative action which furthers substantive equality should not be confused with deference to governmental decisions. Courts should still require governments to demonstrate that the aim of the measure is indeed to further substantive equality and to justify their choice of means. Justification need not amount to proof that there were no suitable alternatives, nor should the State have to demonstrate that the programme is effective. This preserves the space for innovation, as emphasised by the Canadian Court. But justification does require that the State demonstrate that its measures are not based on assumptions, generalisations or stereotypes. In both Hugo and Gosselin, the South African and Canadian Courts were prepared simply to assume facts in favour of the State. Thus in Hugo, although the President had relied exclusively on his desire to protect the interests of young children, the Court was particularly influenced by what it regarded as the impossibility of extending the remission to fathers of young children. „Male prisoners outnumber female prisoners almost fiftyfold. A release of all fathers would have meant that a very large number of men prisoners would have gained their release. As many fathers play only a secondary role in child rearing, the 51 Briheche 24, C-476/99 Lommers v Minister Van Landbouw, Natuurbeheer En Visserij [2004] 2 C.M.L.R. 49 (ECJ) at [39], E-1/02 EFTA Surveillance Authority v Norway [2003] 1 C.M.L.R. 23 (EFTA Court) at [37] –[43] release of male prisoners would not have contributed as significantly to the achievement of the President‟s purpose as the release of mothers. In addition, the release of a large number of male prisoners in the current circumstances where crime has reached alarming levels would almost certainly have led to considerable public outcry. In the circumstances it must be accepted that it would have been very difficult, if not impossible, for the President to have released fathers on the same basis as mothers. Were he obliged to release fathers on the same terms as mothers, the result may have been that no parents would have been released at all.‟ Yet as Kriegler J stated, the President had not sought to justify his decision by reference to the numbers of male prisoners who would have gained their release and no evidence was led to that effect. Nor was there evidence to the effect that such numbers would have produced public disquiet, nor did the Court know anything about the administrative burden of weighing the family circumstances of individual prisoners. The Court simply assumed these issues by way of justification for the differential treatment. This technique, namely a reliance on judicial notice or assumptions of unproved facts, is a familiar one for a Court to resort to when it is unwilling to deflect what it regards as a as a legitimate government policy to further substantive equality. In the Gosselin case in Canada, for example, the Court accepted without substantiating evidence the State‟s assertion that it was easier for young people to find jobs than older people. Judicial deference in this context operates as an obstacle to transformation because it protects the State from having to explain its decision to exclude a particular group from a particular benefit. Respect for the democratic process requires greater, not less, attention to the duty to account and explain. Providing an explanation for a decision enhances the accountability of decision-makers, exposing them to public scrutiny and debate. Requiring the state to provide a reasoned explanation does not require the judges to substitute their own opinions for those of policy makers on resource allocation questions. The judicial role in applying substantive equality remains a supervisory one, to guard against stereotypical assumptions and unwarranted generalisations which can cause or perpetuate disadvantage. This role is not easy to define. However, by protecting the state from having to explain an allocation in the first place, courts abdicate their rightful role in developing an appropriate judicial standard of scrutiny52. (iii) Achieving equality or reinforcing stereotypes: pregnancy and parenthood The dangers of avoiding the formulation of an appropriate level of scrutiny are highlighted in relation to pregnancy and maternity, where a deep sensitivity to the structural nature of gender discrimination is necessary to check whether a measure is likely to bring about change or simply reinforce stereotypes. Whereas special measures for women are necessary for child-bearing, the question is more complex in relation to child-care and family work. Genuinely transformative change can only occur when both parents are equally responsible for child-care53. Special measures for women, however well-intentioned, run the risk of reinforcing their primary role as child-carers, and therefore perpetuating their disadvantage. This does not mean that there should be no special provision for parents. That would be to revert to formal equality, which contemplates levelling-down as an appropriate response. Instead, such measures will only achieve real change if they refer to both parents. Yet, in a string of cases in various jurisdictions, courts have upheld measures which excluded men from rights related to parenting on the ground that they achieved equality for women by recognising their position as mothers. We have already seen this in the Hugo case, where mothers of minor children pardoned, but not fathers. As Kriegler J stated in his dissent, the stereotype that fathers were not responsible for children and that mothers were had a more invidious effect than the benefit to the few hundred women released under the scheme: „The benefits in this case are to a small group of women - the 440 released from prison and the detriment is to all South African women who must continue to labour under the social view that their place is in the home. In addition, men must continue to accept that they can have only a secondary/surrogate role in the care of their children. …There is also more diffuse disadvantage when society imposes roles on men and 52 See further S Fredman ''Providing Equality: Substantive Equality and the Positive Duty to Provide'' (2005) 21 SAJHR 163 53 See further S Fredman Women and the Law women, not by virtue of their individual characteristics, qualities or choices, but on the basis of predetermined, albeit time-honoured, gender scripts..‟54 A similar result can be seen in the European Union case-law. The European Court of Justice has generally been strict in its scrutiny of affirmative action measures, regarding them as a derogation from the equality principle. But in a recent case, it held that to provide child-care facilities for working mothers and not working fathers was a measure which fell within the exception in EU law for measures which, „although discriminatory in appearance, are in fact intended to eliminate or reduce actual instances of inequality which may exist in the reality of social life‟.55 The case concerned a programme according to which a work-place nursery provided places for children of female but not male employees. The measure was defended by the employer, the Dutch Ministry of Agriculture, on the basis that the distinction drawn on grounds of sex reflected its determination to tackle the under-representation of women in the Ministry. According to Dutch law, the programme was legitimate because it was a “ distinction intended to place women in a privileged position in order to eliminate or reduce de facto inequalities” and was reasonable in relation to that aim.56 The European Court of Justice endorsed this approach. The Court pointed out that there was a significant under-representation of women in the Ministry of Agriculture and particularly in the higher grades, and that there was similarly a proven insufficiency of suitable and affordable nursery facilities, which, it regarded as likely to induce women employees to give up their jobs. The Court was not oblivious to the danger that „a measure such as that at issue …, whose purported aim is to abolish a de facto inequality, might nevertheless also help to perpetuate a traditional division of roles between men and women.‟57 However, it was not prepared to hold that the measure was disproportionate for failing to include working fathers, particularly in light of the very limited number of nursery places available, and the fact the employer could grant requests from male officials in cases of emergency, to be determined by the employer. The Court did stress that a measure 54 Hugo para 83 C-476/99 Lommers v Minister Van Landbouw, Natuurbeheer En Visserij [2004] 2 C.M.L.R. 49 (ECJ) para 32; see Equal Treatment Directive, Article 2(4). 56 Art.5 of the Netherlands Law on Equal Treatment of Men and Women. 57 C-476/99 Lommers v Minister Van Landbouw, Natuurbeheer En Visserij [2004] 2 C.M.L.R. 49 (ECJ) para 41 55 which excluded male officials who take care of their children by themselves from access to a nursery scheme subsidised by their employer would go beyond the permissible derogation. But it seemed reassured on this score by the indication given by the Ministry of Agriculture that male officials who bring up their children by themselves should, on that basis, have access to the nursery scheme at issue. Valuable guidance as to how to distinguish between discrimination and affirmative action in this context is given by the CEDAW Committee. In its Recommendation 25 on special temporary measures, it states that „women‟s biologically determined permanent needs and experiences should be distinguished from other needs that may be the result of past and present discrimination against women by individual actors, the dominant gender ideology, or by manifestations of such discrimination in social and cultural structures and institutions.‟ Moreover, „as steps are being taken to eliminate discrimination against women, women‟s needs may change or disappear, or become the needs of both women and men. Thus, continuous monitoring of laws, programmes and practices directed at the achievement of women‟s de facto or substantive equality is needed so as to avoid a perpetuation of non-identical treatment that may no longer be warranted.‟58 (iv) Demarcating beneficiaries The second challenge for courts in their supervisory role concerns the question of whether the class of beneficiaries is appropriately delineated, or whether it is over- or under-inclusive. A key strength of substantive equality is its move away from the individualism of formal equality, which requires proof that an individual has suffered discrimination at the hands of an identified perpetrator before an individual remedy can be granted. Affirmative action, understood as an element of substantive equality, does not require proof of individual detriment in this sense. Therefore the class of beneficiaries need not consist only of proved victims. Nevertheless, there needs to be some correlation between the beneficiaries and the previous disadvantage. This raises 58 General recommendation No. 25, on article 4, paragraph 1, of the Convention on the Elimination of All Forms of Discrimination against Women, on temporary special measures, para 11. the important question of the relationship between socio-economic disadvantage and status groups. If the aim of the measure is to address socio-economic disadvantage, then a group demarcated by status, such as race or gender, might be over-inclusive in including wealthier blacks or women, and under-inclusive, in excluding poor white men. If the aim is to address mis-recognition or status inequalities, then the demarcation might be appropriate. It is here that the controlling function of proportionality comes in: are the means appropriately correlated with the ends? But it is here too that the problem is overlaid by the need to take account of relative institutional competence. Should judges be leaving this decision to law-makers and the executive, or should they be exercising some supervision? Both the Canadian and South African courts have addressed this question. In Kapp, the question arose as to whether the three Indian bands were an „identifiable disadvantaged group‟ for the purposes of section 15(2), which it will be recalled, permits ameliorative measures whose object is „the amelioration of conditions of disadvantaged individuals or groups including those that are disadvantaged because of race, national or ethnic origin, colour, religion, sex, age or mental or physical disability‟. It is notable that this formulation focuses on status as a means of identifying the group, but does not distinguish openly between attendant socioeconomic disadvantage and status disadvantage. The Court in Kapp regarded them as intertwined. Thus, having identified the distinction as being based on race, it went on to establish whether, as required by s15(2), the program targeted a „disadvantaged group identified by‟ race. For this, it referred to both status disadvantage - “the legacy of stereotyping and prejudice against Aboriginal peoples” - and socio-economic disadvantage: “the evidence shows in this case that the bands granted the benefit were in fact disadvantaged in terms of income, education and a host of other measures.” However, it did not require the classification to be solely concerned with such disadvantage. Thus it held that not all members need be disadvantaged. „The fact that some individual members of the bands may not experience personal disadvantage does not negate the group disadvantage suffered by band members.‟59 59 Kapp para 59 In Van Heerden, a similar question was faced. Although an overwhelming majority of the new members of Parliament were excluded from parliamentary participation by past apartheid laws, not all new parliamentarians of 1994 belonged to the class of persons prejudiced by past disadvantage and unfair exclusion. Moseneke J recognised that it would often be „difficult, impractical or undesirable to devise a legislative scheme with “pure” differentiation demarcating precisely the affected classes. Within each class, favoured or otherwise, there may indeed be exceptional or “hard cases” or windfall beneficiaries. That however is not sufficient to undermine the legal efficacy of the scheme. The distinction must be measured against the majority and not the exceptional and difficult minority of people to which it applies…‟60 He thus held that the measure should be judged by „[whether an overwhelming majority of members of the favoured class are persons designated as disadvantaged by unfair exclusion.‟ 61 Like the Canadian Court, he regarded the validity of the remedial measures as unaffected by the existence of a tiny minority of members of Parliament who were not unfairly discriminated against, but who benefited from the measure. This can be contrasted with the decision of Mogkoro J, who held that a more exact fit was necessary. If the class of beneficiaries could not be precisely delineated, then the measure should have to reach the high standards of fairness required in relation to an ordinary discrimination claim. She regarded the measure as too loosely related to a proscribed group to fall within section 9(2), which relieves the state of the burden of proving unfairness. Instead, it needed to pass muster under section 9(3), which, on the facts, it did. The difference between Moseneke and Mokgoro JJ reflects an underlying principle about the role of judicial scrutiny. Requiring a higher level of scrutiny does not in itself mean that affirmative action programmes will be struck down as a breach of equality. Judges could exercise supervisory control without reverting to a formal equality approach. But this means that more attention needs to be given to the criteria of review which would ensure that an affirmative action measure is consistent with substantive equality. Most important in the context of the delineation of the beneficiary group is the relationship between redistribution and recognition, or socioeconomic disadvantage and status inequality. This relationship need not achieve much 60 61 Van Heerden para 39 Van Heerden para 40 attention when there is a substantial overlap, as in the case of the First Nation tribes in the Kapp case. However, as affirmative action measures become more effective, and some members of the status group begin to prosper, questions arise as to the legitimacy of an over-inclusive group. One response is to overlay status with socio-economic disadvantage in demarcating the group of beneficiaries. To qualify for the benefit, the individual must show socioeconomic disadvantage as well as membership of the status group. This can be seen in both India and the US. The Indian Constitution permits special provision to be made for two categories of disadvantaged groups. The first are known as „Scheduled Castes and Scheduled Tribes‟ and are specified by the President.62 The second are referred to as „socially and educationally backward classes of citizens,‟63 or other „backward classes‟.64 The latter are specified in a list drawn up by the National Commission for Backward Classes.65 The equality guarantee in the Indian Constitution provides that special provision may be made for the advancement of any these categories.66 Particularly controversial has been the question of whether the reservations should be available to the „creamy layer‟, or those members of the certified groups which are in fact no longer socially or educationally disadvantaged. In two of the foremost cases on reservations, the Indian Supreme Court has made it clear that exclusion of the creamy layer is not an addition principle, but one that goes to the very purpose of the reservation. As Justice Jeevan Reddy stated in Indra Sawhney's case, „In our opinion, it is not a question of permissibility or desirability of such test but one of proper and more appropriate identification of a class as a backward class.‟67 Similarly, the Chief Justice in Ashoka Kumar stated: „To fulfil the conditions and to find out truly what is socially and educationally backward class, the exclusion of "creamy layer" is essential.‟68 In other words, the definition of the beneficiary class must correspond with the purpose of the provision, namely to advance those who are disadvantaged. However, the requirement that a group display 62 Constitution of India http://lawmin.nic.in/coi/coiason29july08.pdf Articles341(1), 342(2). See Constitution of India s15(4) (http://lawmin.nic.in/coi/coiason29july08.pdf) 64 See Constitution of India Article 16(4) 65 See Constitution of India Article 340(1); the list can be found at http://ncbc.nic.in/backwardclasses/index.html 66 Article 15(4) 67 Indra Sawhney p.724. 68 Ashoka Kumar para 149. 63 both misrecognition and socio-economic disadvantage is only imperfectly executed. On the one hand, there is no creamy layer exclusion for the Scheduled Tribes or Scheduled Castes. On the other hand, there is no provision for reservation in favour of disadvantaged Muslims. A similar attempt to create a closer fit between status and socio-economic disadvantage is found in US affirmative action programmes, particularly in respect of set-aside programmes authorizing preferential treatment in the award of government contracts. In order to comply with the strict scrutiny test, Federal legislation mandating set asides for status groups now also includes a requirement of evidence of socio-economic disadvantage. This can be seen in the major vehicle for present-day set-asides, section 8(a) of the Small Business Act (SBA). 69 To participate in what has become known as a section 8(a) program, a business must be 51% owned by individuals who qualify as “socially and economically disadvantaged.” 70 The Act defines “socially disadvantaged individuals” as “those who have been subjected to racial or ethnic prejudice or cultural bias because of their identity as a member of a group without regard to their individual qualities.”71 It defines “economically disadvantaged individuals” as “those socially disadvantaged individuals whose ability to compete in the free enterprise system has been impaired due to diminished capital and credit opportunities as compared to others in the same business area who are not socially disadvantaged.”72 There is a rebuttable presumption that members of certain groups are socially disadvantaged, including: Black Americans; Hispanic Americans; Native Americans; Asian Pacific Americans; subcontinent Asian Americans and members of other groups designated from time to time by SBA.73 The presumption of social disadvantage may be overcome with „credible evidence to the contrary‟, which can be submitted by third parties74. An individual who is not a member of a listed group can also be included if she or he can „establish individual social disadvantage by a preponderance of the evidence.75‟ The presumption does not apply 69 15 U.S.C. § 631 et seq. As defined in 13 C. F. R. § 124.105 71 15 U.S.C. § 637(a)(5) 72 15 U.S.C. § 637(a)(6)(A). 73 13 C.F.R § 124.103 (b) 74 13 C.F.R § 124.103 (b)(3) 75 13 C.F.R § 124.103(c) 70 to economic disadvantage: each participant to must prove such disadvantage.”76 The result is that businesses which, regardless of race, are not in fact socially or economically disadvantaged are excluded, a similar effect to that achieved by the Indian „creamy layer‟ provisions above. Amendments adopted in 1994 specifically included “small business concerns owned and controlled by women” in addition to “socially and economically disadvantaged individuals.” Also of importance has been the Transportation Equity Act for the 21st Century (known as TEA-21),77 which authorizes the use of race- and sex-based preferences in federally funded transportation contracts. As in the SBA regulations, the TEA-21 regulations presume that Black Americans, Hispanic Americans, Native Americans, Asian-Pacific Americans, Subcontinent Asian Americans, and women are socially and economically disadvantaged,78 a presumption which is rebutted where the individual has a personal net worth of more than $750,000 or a preponderance of the evidence demonstrates that the individual is not in fact socially and economically disadvantaged.79 Firms owned and controlled by someone who is not presumed to be disadvantaged (i.e., a white male) can qualify for DBE status if the individual can demonstrate that he is in fact socially and economically disadvantaged.80 This raises deep questions as to the function of affirmative action and the relationship between status and socio-economic disadvantage. By requiring individuals to manifest both socio-economic disadvantage and membership of an out-group, it is assumed that the function of affirmative action is only in relation to the socio-economic dimension of status wrongs. Pure socio-economic disadvantage without status wrongs and pure status wrong without socio-economic disadvantage both fall outside the affirmative action purview. It could be argued that affirmative action is an inappropriate remedy in cases in which status wrongs need to be addressed but socio-economic disadvantage is no longer an issue. On this view, status-only wrongs are better addressed through other measures, such as prohibitions on discrimination and 76 13 C. F. R. § 124.104 Transportation Equity Act for the 21st Century (TEA-21), extended through FY2009 by P.L. 109-59, signed into law during the 109th Congress. 78 .§ 26.67(a). 79 .§ 26.67(b). 80 .§ 26.67(d). 77 harassment. Indeed, to continue to use affirmative action for purely status wrongs could, on this view, simply reinforce stereotypes. However, this unnecessarily narrows the role of affirmative action. Its aim should not be regarded as merely redistributive. Affirmative action also functions in three further ways.81 Firstly, hidden and structural barriers to advancement can be more easily overcome by affirmative action than by individual claims for indirect discrimination. Despite apparently objective eligibility standards and ostensible equal opportunity policies, there remain many hidden obstacles to the advancement of women, historically disadvantaged South Africans, minorities and others. On the assumption that, in the absence of barriers, there would be a random spread of men and women, blacks and whites, and members of different ethnic groups across the labour force and government, the very fact that a group is seriously under-represented in a sphere or activity is evidence of the subtle operation of often invisible barriers. Yet could this not be dealt with by the familiar principle of indirect discrimination? Indirect discrimination expressly aims to remove apparently neutral barriers which in fact function to exclude more women than men or more blacks than whites unless they can be justified.82 However, it has proved to be too clumsy a tool to achieve its aims. Applicants seeking to prove indirect discrimination have initiate court proceedings and show disparate impact, often on the basis of complex statistics. Finally, even if she can surmount all these barriers, she may find that an employer successfully shows that the criteria, despite being exclusionary, are justifiable by reference to the needs of the business. Affirmative action resolves many of these difficulties. Instead of relying on litigation by individual victims, the employer takes the initiative. Nor is it necessary to prove that an exclusionary rule has had a disproportionate impact. Instead, it is sufficient to demonstrate a clear pattern of under-representation of women in particular grades or occupations. The complex questions above are unnecessary. Moreover, discriminatory selection criteria are unequivocally removed: by creating a presumption in favour of women in conditions of equal merit, it makes it impossible 81 The remainder of this section is taken from S Fredman Discrimination Law (Clarendon Law Series Oxford University Press Oxford 2002), chap 5. 82 Sex Discrimination Act 1975, s. 1(1)(b)(i). for such criteria to be reintroduced surreptitiously through subjective decisionmaking. Phrased in this way, affirmative action can be legitimated as an effective means of overcoming hidden barriers. At the same time, this formulation reveals its very limited impact. Most importantly, while preference policies may change the gender or racial composition of some higher paid occupations, they do not challenge the underlying structural and institutional forces leading to the discrimination. As Young argues83, because affirmative action diagnoses the problem as one of maldistribution of privileged positions, its objective is limited to the redistribution of such positions among under-represented groups. However, this narrow distributive definition of racial and gender justice leaves out the equally important issues of institutional organisation and decision-making power. The under-representation of women and historically disadvantaged South Africans in higher positions in the employment ladder, both public and private, is only partially solved by inserting some women or black South Africans into those positions. While some women „make it to the top‟, the vast majority will remain in poorly paid, low status jobs. It is not surprising that, in practice, reverse discrimination is often found to do no more than favour middle class women or blacks who are already relatively privileged in society.84 For fundamental change to occur, the structural and institutional causes of exclusion need to be changed, including the division of labour in the home, the interaction between work in the family and work in the paid labour force, education and others. A second, more dynamic way of justifying the use of affirmative action policies is to argue that the very presence of women or blacks in higher status positions will lead to structural changes. This is to argue, as Young does, that the particular life experience of the decision-maker is reflected in his or her view. Since gender and race remain such strong determinants of a person‟s life experience, the overwhelming 83 I. Young, Justice and the Politics of Difference (Princeton University Press, 1990), p. 193. 84 W. F. Menski „The Indian Experience and its Lessons for Britain‟ in B. Hepple and E. Szyszcak (eds.), Discrimination and the Limits of the Law (Mansell, 1992), 300 at p. 330. predominance of one gender or race in decision-making fora make it unlikely that the experience and perspectives of the excluded group will be articulated85. Indeed, a recent study in Britain demonstrated that the biggest barrier to advancement for ethnic minorities, women and disabled people within the senior Civil Service is believed to be a deeply embedded culture which acts to exclude those who are different from traditional Civil Service employees, who are generally middle class middle-aged white men86. On this view, it is possible to characterise the presence of women and historically disadvantaged groups as functioning to open up new perspectives on decision-making, to cast light on assumptions that the dominant group perceives as universal, and to enhance the store of „social knowledge.‟ This approach makes sense of the notion that women or minorities may have distinct perspectives, which the very process of exclusion negates, and therefore which need to be guaranteed a place in deliberative decision-making. In addition, it demonstrates the necessity for a critical mass both to reflect a diversity within the social group in question and to make the common interests more audible. At the same time, it is important to move beyond essentialising a status group, or regarding all members as sharing the same interests or perspectives. Young argues that groups are better understood, not as fixed categories with impermeable boundaries, but as a set of relationships between different people. Such a relational understanding moves beyond the notion that a group consists of members who all share the same fixed attributes and have nothing in common with members of other groups. Instead, a group is characterised as a social process of interaction in which some people have an affinity with each other. Assertion of affinity with a group may change with social context and with life changes; and members may have interests which differ from other members of the group but are similar to members of other groups87. A third and related justification for reverse discrimination is that it provides diversity in an educational institution or workplace; and that it facilitates the provision of role 85 Phillips, above n. 97, p. 52. 86 (1999) 87 EOR 4. 87 Young, above n. 93, pp. 171-2 models. In Bakke88, Powell J justified affirmative action in university admissions thus: „An otherwise qualified medical student with a particular background - whether it be ethnic, geographic, culturally advantaged or disadvantage - may bring to a professional school of medicine, experiences, outlook, and ideas that enrich the training of its student body and better equip its graduates to render with understanding their vital service to humanity.‟89 In other words, where a group has been excluded from a particular setting, be it a workforce or an educational institution, the likelihood is that the perspectives and experiences of members of the excluded group, particularly those relating to its exclusion, will be undervalued, misunderstood or ignored by the dominant group, making it impossible for the excluded group to change its disadvantaged position. While diversity operates to change the perspectives of the dominant group, the provision of role models operates on the self-perception of excluded groups, piercing stereotypes and giving them the self confidence to move into non-traditional positions. This too suggests the importance of critical mass. This is well illustrated by Canadian Supreme Court in Action Travail des Femmes v Canadian National Railway Co .90As Chief Justice Dickson put it, the aim of an employment equity programme (in this case setting a quota of one woman in four new hirings until a goal of 13 per cent women in certain blue collar occupations was reached) is not to compensate past victims; but "an attempt to ensure that future applicants and workers from the affected group will not face the same insidious barriers that blocked their forebears".91 He identified at least two ways in which such a programme is likely to be more effective than one which simply relies on equal opportunities or the proscription of intentional prejudice. Firstly, the insistence that women be placed in non-traditional jobs allows them to prove that they really can do the job, thereby dispelling stereotypes about women's abilities. This was particularly evident in the case at hand, in which the quotas ordered by the tribunal concerned traditionally male 88 Regents of the University of California v Bakke 438 U.S. 265 (1978) (United States Supreme Court) ibid., at 314, 2760. Upheld in Grutter v. Bollinger 539 U.S. 306; 123 S.Ct. 2325 (2003) (US Supreme Court) 89 90 Action Travail des Femmes v Canadian National Railway Co [1987] 1 SCR 1114, 40 DLR (4th) 193. 91 ibid. at 213. jobs such as "brakeman" or signaller at Canadian National Railways. Secondly, an employment equity programme helps to create a "critical mass" of women in the work-place. Once a significant number of women are represented in a particular type of work, "there is a significant chance for the continuing self-correction of the system."92 The critical mass overcomes the problem of tokenism, which would leave a few women isolated and vulnerable to sexual harassment or accusations of being imposters. It would also generate further employment of women, partly by means of the informal recruitment network and partly by reducing the stigma and anxiety associated with strange and unconventional work. Finally, a critical mass of women forces management to give women's concerns their due weight and compels personnel offices to take female applications seriously. As the Chief Justice concluded: "It is readily apparent that, in attempting to combat systemic discrimination, it is essential to look to the past patterns of discrimination and to destroy those patterns in order to prevent the same type of discrimination in the future."93 At the same time, it is important to stress the limitations of affirmative action as a strategy. The introduction of new perspectives, while an important goal, can only have a limited impact: entrenched structures are often resilient and indeed have powerful conformist pressures. Women, blacks or minorities may find themselves forced to hide their views and ignore their own needs and interests in order to ensure that their continued participation is viable. Even if they do articulate their perspectives, the process of recognition and affirmation is halting and erratic. Thus affirmative action needs to be only one part of a broad based and radical strategy, which does more than redistribute privileged positions but refashions the institutions which continue to perpetuate exclusion. 92 ibid. at 214. 93 ibid. at 215. 4. Should poverty be a ground for discrimination? It has been argued thus far that a major strength of substantive equality is its ability to achieve a synthesis between status inequality and socio-economic disadvantage. Poverty, however, is generally associated only with socio-economic disadvantage. It is often thought not to correspond to any status group, and similarly not to give rise to recognition inequality. For this reason, only measures dealing with socio-economic disadvantage are thought to be appropriate, and poverty is not regarded as being a separate ground within discrimination law. These assumptions are vividly demonstrated by the United States Supreme Court, which has refused to shape the Constitutional equality guarantee in the US Constitution to cover groups defined only by their poverty. In a case in which the claimants argued that a school system funded by local taxes discriminated against those who lived in poor areas, the Court drew a bright line between claims based on poverty and those traditionally identified as „suspect‟ under the Equal Protection Clause. The difference between status and poverty is highlighted in the judgement of Powell J: „The system of alleged discrimination and the class it defines have none of the traditional indicia of suspectness: the class is not saddled with such disabilities, or subjected to such a history of purposeful unequal treatment, or relegated to such a position of political powerlessness as to command extraordinary protection from the majoritarian political process.‟94 Questions of this sort, it held, were inappropriate for strict scrutiny since they involved „the most delicate and difficult questions of local taxation, fiscal planning, educational policy, and federalism, considerations counselling a more restrained form of review‟.95 The Court found it particularly difficult to accept that there could be justiciable criteria for what amounted to poverty: „The Texas system of school financing might be regarded as discriminating (1) against „poor‟ persons whose incomes fall below some identifiable level of poverty or who might be characterized as functionally „indigent,‟ or (2) against those who are relatively poorer than others, 94 San Antonio Independent School District v. Rodriguez 411 U.S. 959, 93 S.Ct. 1919. (U.S.Supreme Court)1294. 95 Ibid 1302. or (3) against all those who, irrespective of their personal incomes, happen to reside in relatively poorer school districts. Yet appellees have not defined the term „poor‟ with reference to any absolute or functional level of impecunity.‟96 Even if it had been able to define poverty, the Court was unwilling to view redistribution in egalitarian terms. Instead, it was „“a sufficient answer to appellees” argument that, at least where wealth is involved, the Equal Protection Clause does not require absolute equality or precisely equal advantages‟.97 However, this approach entirely ignores the fact that poverty is not only about socioeconomic disadvantage. It is also about recognition inequality. Welfare claimants are frequently stigmatised as scroungers, lazy, or irresponsible, and authorities often devote disproportionate resources to „welfare fraud‟, as compared to fraud committed by powerful politicians or entrepreneurs. Bruce Porter, describing Canadian welfare fraud campaigns, describes the consequences vividly in terms of recognition inequality: „It is difficult to appreciate just how profound is the social exclusion that results from this kind of government-endorsed promotion of discrimination and scapegoating. It transforms social assistance from an entitlement of citizenship linked with the right to security and dignity, into a source of shame, guilt, and insecurity.‟98 Should legislation therefore prohibit discrimination on grounds of poverty? To answer this question, it is worth revisiting the four dimensions of equality set out in the first part of this paper. These were to, first, to break the cycle of disadvantage associated with status or out-groups; secondly, to promote respect for the equal dignity and worth of all, thereby redressing stigma, stereotyping, humiliation, and violence; thirdly to accommodate and positively affirm different identities; and finally to redress under-representation in decision-making and social exclusions. It is clear that anti-discrimination law cannot address the primarily economic manifestations of poverty: the welfare state or socio-economic rights generally aim to redress socioeconomic disadvantage. Nevertheless, it remains paradoxical that disadvantaged or socially excluded individuals who are not members of historically stigmatised status groups gain no protection, and might even be called upon to make way for members 96 Ibid 1289. Ibid 1291. 98 Porter „Claiming Adjudicative Space: Social Rights, Equality and Citizenship‟. 97 of status groups in respect of affirmative action programmes. Instead, including poverty or social exclusion as a ground of discrimination would add a „recognition‟ dimension to this primary emphasis on redressing economic disadvantage within the welfare state. In this context, the role of anti-discrimination law would be to prohibit stigma and hostility, to affirm individual dignity and worth, and to facilitate social inclusion and participation in decision-making. South Africa has tentatively included „socio-economic status‟ as a potential ground of unfair discrimination, although at present, this ground is only a „directive principle‟ and awaits further activation.99 It is also possible to argue that poverty, or socioeconomic status, is an unenumerated ground under s.9 of the Constitution. This leaves open the question of how to define the status-ground. The South African Promotion of Equality and Prevention of Unfair Discrimination Act 4 of 2000 defines socioeconomic status to include the „social or economic condition or perceived condition of a person who is disadvantaged by poverty, low employment status, or lack of or lowlevel educational qualifications‟.100 A complementary suggestion could be to focus on social exclusion, a concept which has come to replace poverty as the defining criterion for state intervention. The focus would then be on those who are systematically excluded from participation in society. This means systematic exclusion from participation in economically or socially valuable activities, from political engagement and from social interaction as well as lacking the capacity to purchase goods and services.101 Social exclusion in this form arises from a variety of causes, which are partly material, but also relate to other issues, such as living in a deprived area, suffering partnership breakdown, being a elderly or disabled or a member of a historically disadvantaged group. Social exclusion is not just a temporary phase of poverty; it is systemic, often passed from generation to generation and can be self-perpetuating. The result of including socio-economic status defined as social exclusion would be that less favourable treatment on grounds of indicators of social exclusion, such as living in an informal settlements. Similarly, if a training scheme operated in such a way as to disproportionately exclude those who are socially excluded, for example, by being located a distance away from a council 99 Section 34(1). Section 1(xxvi). 101 T Burchardt, J Le Grand & D Piachaud „Degrees of Exclusion‟ in J Hills, J Le Grand & D Piachaud (eds) Understanding Social Exclusion (2001) 30–32. 100 estate, or requiring participants to have had relevant job experience, a claim of indirect discrimination could be pursued unless it was demonstrated that the requirement was necessary. However, it is unlikely that incorporating a status ground defined according to social exclusion would go far enough to address the dichotomy between the two spheres. It simply tacks an expressly redistributive claim onto a framework geared for recognition claims, which in turn is based on individualised, retrospective and monetary remedies. Status-based claims as presently constituted, as argued above, cannot command the resources necessary genuinely to address social exclusion. It is only as a way of challenging stigmatic sequels of social exclusion that it might have some marginal effect. Conclusion As the concept of substantive equality receives growing acknowledgement, both internationally and in domestic jurisdictions, more attention needs to be given to its specific meaning and impact. This paper has attempted to frame some of the complex questions which substantive equality raises. These primarily concern the relationship between status and socio-economic disadvantage in relation to three major issues: dignity, affirmative action and poverty. Cutting across these issues concerns the question of the appropriate role of the judiciary. The key to the their resolution lies in achieving a synthesis between both dimensions, rather than regarding one as trumping the other. But to achieve this, requires a greater sensitivity to structural inequality is required. At the same time, more attention is needed to the standard of judicial review. Courts should not cut across government measures to further substantive equality; indeed, courts should support such measures. However, courts should not regard this as a reason for deference. Respect for the democratic process requires greater, not less, attention to the duty to account and explain. The judicial role in applying substantive equality remains a supervisory one, to guard against stereotypical assumptions and unwarranted generalisations which can cause or perpetuate disadvantage. Facing the future: Substantive Equality under the Spotlight Sandra Fredman Oxford University Published in: EQUALITY IN THE WORKPLACE: REFLECTIONS FROM SOUTH AFRICA AND ELSEWHERE Edited by Ockert Dupper and Christoph Garbers, University of Stellenbosch, South Africa (2010) 1. Introduction Transformation in South Africa is intimately bound up with a vigorous development of the principle of equality. However, as the second decade of the new democracy draws to a close, the challenges facing the equality principle are more complex than ever. Equality before the law, although a sine qua non, is far from sufficient to bring about the far-reaching social and institutional changes necessary to achieve true equality. In addition, to ensure that the constitutional project of transformation retains its energy into the next decade, the promise of substantive equality must be fulfilled. Yet there is still much debate as to what that promise entails and the best means to achieve it. Nor are such challenges confined to South Africa. Although the South African experience has been condensed into a mere decade and a half, the evolution of substantive equality has been shared in many other societies, and the modern dilemmas are strikingly similar. The aim of this paper is to frame some of the key challenges facing the principle of substantive equality. Central to substantive equality is the recognition of the connection between status and disadvantage, where status refers to race, gender, disability or other prohibited ground, and disadvantage is concerned with socioeconomic disadvantage. Whereas formal equality addresses discrimination based on status, substantive equality highlights the fact that it is not status per se which is problematic, but the disadvantage which attaches to status. Thus the focus should be on women or blacks, rather than on gender or race per se. However, the power of this insight is matched by the challenges it raises. In addressing both status and socioeconomic disadvantage within the vehicle of substantive equality, how is each to be formulated, and how should they interact? This paper considers three specific problematic issues which arise from this interaction. The first is the role of dignity. An increasingly popular concept within substantive equality discourse, dignity has also at times had the effect of dislocating status from socio-economic disadvantage. The paper asks what role dignity should play. The second problematic issue at the intersection of status and socio-economic disadvantage is concerned with affirmative action. Affirmative action specifically attaches socio-economic benefits to those disadvantaged by status, and therefore facilitates the achievement of substantive equality. However, there is a risk that this link perpetuates status inequality by freezing individuals into existing status groups, or by reinforcing stereotypes. This paper considers this problem particularly in relation to pregnancy and parenthood. Furthermore, as the status group begins to benefit from the affirmative action provisions, a gap opens up between status and disadvantage. The paper asks the question of whether the beneficiary group therefore needs to be narrowed to include only those who suffer both status and socio-economic disadvantage. Indeed, if the aim is indeed to redress socio-economic disadvantage, why link affirmative action to status at all? It is argued here that affirmative action performs more functions than simply to redress socio-economic disadvantage, and therefore that the link with status should be maintained, even when accompanied by relative economic prosperity. This leads onto the third problematic issue at the intersection between status and socio-economic disadvantage, namely the link between status and poverty. It is argued here that just as status brings with it socioeconomic disadvantage, so poverty brings with it status harm. This means that poverty should be a ground of discrimination. Another way of formulating the issue is to use the concepts of recognition and redistribution, fashioned by Fraser, Honneth and others.1 Based on Hegel‟s notion of 1 N Fraser and A Honneth Redistribution or Recognition (Verso London, New York 2003);S Fredman ' „Redistribution And Recognition: Reconciling Inequalities‟ ' (2007) 23 South African Journal of Human Rights 214 ; S Liebenberg 'Needs, Rights and Transformation: Adjudicating Social Rights' Center For Human Rights And Global Justice Working Paper Economic And Social Rights Series Number 8, 2005 ; J Fudge ''The Canadian Charter of Rights: Recognition, Redistribution, and the identity as inter-subjective recognition, or inter-personal affirmation, recognition is the value attached to status. Status inequality arises out of mis-recognition, which in turn comes from denigrating or humiliating individuals. Distributive inequality arises out of unfair allocation of resources. Whereas distributive inequality is usually dealt with through the welfare state, and recognition inequality through discrimination law, the power of substantive equality is its ability to bring both together.2 However, it remains the case that substantive equality is better at addressing status wrongs than socio-economic disadvantage, unless it is constructively allied with socio-economic rights.3 The paper begins with a brief description of the meaning of substantive equality. It then moves on to consider the three challenges above: dignity, affirmative action, and poverty as a ground of discrimination. I do not presume to provide answers to these questions, but merely to elucidate the nature of the challenge, and show how various jurisdictions have responded. 2. Substantive equality Substantive equality is born out of disappointment and frustration at the limits of formal equality. It aims to take up the baton where formal equality leaves off. However, while it is by now relatively easy to reach consensus on the limits of formal equality, there remain many unanswered questions about the ways in which substantive equality should be shaped to transcend these limitations. This section briefly rehearses the limitations of formal equality, before turning to the challenges faced in defining substantive equality. Formal equality is based on the premise that individuals should be treated as individuals, on the basis of their own merit, rather than on attributions based on irrelevant characteristics such as race, colour, gender, caste or other analogous status. Imperialism of the Courts'' in T CampbellKD Ewing and A Tomkins (eds) Sceptical Essays on Human Rights (Oxford University Press Oxford 2001) 2 S Fredman ' „Redistribution And Recognition: Reconciling Inequalities‟ ' (2007) 23 South African Journal of Human Rights 214 3 S Fredman ''Providing Equality: Substantive Equality and the Positive Duty to Provide'' (2005) 21 SAJHR 163 It is usually summed up by the Aristotelian formula that likes should be treated alike. In legal contexts, it finds expression in the principle of direct discrimination or equal treatment, which makes it unlawful to treat a person less favourably on grounds of her gender, race or other status than a person of a different gender, race or other status. Its achievements should not be understated. It has taken many years of political struggle to reach the point of recognition that gender and race are irrelevant criteria for key political and civil rights. Political thinkers from Aristotle onwards regarded gender as a relevant criterion for access to citizenship rights, ascribing only to men the rationality required to qualify as a subject of rights4. It hardly needs to be stated that the key initial achievement of the new democracy in South Africa was to remove all racial references from the criteria for access to basic rights. However, experience has quickly demonstrated that inequalities persist despite the introduction of formal equality. This can be attributed to five main characteristics of this concept. The first is the assumption that the individual can be abstracted from her gender, race or other status and dealt with entirely on merit. However, merit is itself a function of previous advantage, rather than an objective characteristic. Treating status as an irrelevant ground merely ignores the ongoing disadvantage experienced by individuals who have been previously been unequal before the law or subject to social prejudice. 5 The result is to entrench disadvantage. Secondly, formal equality assumes that the aim is identical treatment. In fact, given antecedent inequality, it may be necessary to provide very different sets of resources to achieve genuine equality. As Sen has argued: „Equal consideration for all may demand very unequal treatment in favour of the disadvantaged. The demands of substantive equality can be particularly exacting and complex when there is a good deal of antecedent inequality to counter.‟6 Thirdly, formal equality is premised on an abstract and universal individual. However, the possibility of an individual abstracted from her status characteristics is illusory. In reality, the abstract individual is clothed with the characteristics of the dominant group, which are then asserted as if they were universal. Only those who can conform 4 S Fredman Women and the Law (Oxford Monographs in Labour Law Oxford University Press Oxford 1997) chap 1. 5 See further, S Fredman Discrimination Law (2002) chapters 1 and 5; C McCrudden „ “The Merit Principle”‟ (1998) 18 Oxford J of Legal Studies 543. 6 A Sen Inequality Re-examined (Oxford University Press Oxford 1992). to this norm are sufficiently „alike‟ to be entitled to „like treatment.‟ The result is that formal equality demands conformity as a price for equal treatment. Yet diverse individual identities may be enriching and desired. The problem is not the diversity of characteristics, but the detrimental treatment attached to it. Thus the aim should not be to eliminate difference, but to prohibit the detriment attached to such difference. Fourthly, formal equality is a relative concept, agnostic as to the substantive outcome. It can therefore be fulfilled by treating everyone equally badly, or by removing benefits from the better off in order to bring them in line with the worse off. This means that the achievement of formal equality can be a hollow victory, or even a defeat. Finally, formal equality is based on a negative conception of liberty, aiming to restrain the state from interfering with individual rights, rather than placing positive obligations on the State to promote equality. Substantive equality is fashioned to address each of these limitations. Firstly, substantive equality does not aim to abstract the individual from the social context. It takes into account existing power structures and the role of status or identity within them. Status is not regarded as irrelevant. Indeed, it may be highly relevant in addressing inequalities in society. Substantive equality therefore focuses on the disadvantage associated with a prohibited ground rather than the ground itself: on women rather than gender; on blackness rather than colour. Secondly, and as a result of this, substantive equality is sensitive to outcomes rather than just to treatment. This opens the way to requiring different treatment in order to achieve equalities of outcome. Thirdly, substantive equality recognises that identity can be a source of value. It does not therefore aim to treat all individuals identically, but to affirm and accommodate differences. In its transformative form, substantive equality requires social institutions to change, rather than expecting the individual to conform. Fourthly, substantive equality is not neutral as to the outcome. Equality cannot be achieved by treating all equally badly, or by removing benefits from the advantaged class. It is substantive in the sense that it advances individuals rather than formal in ensuring only consistency. Finally, substantive equality can go beyond a fault based model to one which includes positive duties to respect, protect, promote and fulfil. Framing substantive equality as a response to the limits of formal equality has achieved considerable consensus. However, the specific content of substantive equality remains contested. There is a temptation to reduce substantive equality to a single dimension, such as dignity, or socio-economic disadvantage. I argue instead that it has four different dimensions.7 First, it is an asymmetric principle. Rather than the abstract individual of formal equality, substantive equality focuses on disadvantage. It aims therefore is to break the cycle of disadvantage associated with status or out-groups. This is the redistributive dimension. Secondly, substantive equality promotes respect for the equal dignity and worth of all, thereby redressing stigma, stereotyping, humiliation, and violence because of membership of an outgroup. Sexual harassment, racist abuse, humiliation of old people, and homophobic bullying at schools are all examples of situations in which the primary claim is based on dignity. Thirdly, it does not exact conformity as a price of equality. Instead, it entails an accommodation and positive affirmation and celebration of identity within community. The second and third are dimensions of recognition. Finally, out-groups are under-represented in decision-making at all levels. Many are also socially excluded. Substantive equality should therefore facilitate full participation in society. This four-dimensional understanding of substantive equality leaves open the question of how the dimensions relate to each other. The rest of the paper attempts to throw some light on this interaction. 3. Dignity The desire to find a substantive core to prevent equality from being no more than an exercise in consistency has led many academics, practitioners, and judges to frame dignity as the central equality value. Thus the Canadian Supreme Court, rejecting a formal equality approach, located dignity at the centre of the equality principle. „Equality means that our society cannot tolerate legislative distinctions that treat certain people as second class citizens, that demean them, that treat them as less capable for no good reason, or that otherwise offend fundamental human dignity‟.8 Similarly, according to Chaskalson P : „Inequality is established not simply through 7 S Fredman The Future of Equality in Great Britain (Working Paper No.5 Equal Opportunities Commission Manchester 2002) 8 Law v Canada [1999] 1 SCR 497 (Canadian Supreme Court) para 51. group-based differential treatment, but through differentiation which perpetuates disadvantage and leads to the scarring of the sense of dignity and self-worth.‟9 There is much that makes dignity an intuitively appealing concept. To heal the scars of apartheid requires, first and foremost, an affirmation of each person‟s humanity, prohibiting the dehumanisation and degradation of anyone. Dignity has also been used specifically to bring sexual harassment into the fold of equality. Whereas formal equality engages in fatuous attempts to find an appropriate comparator, substantive equality simply prohibits sexual harassment because it is inconsistent with respect for a woman‟s basic dignity and humanity. Thus under EU law, harassment is a species of discrimination, where harassment is defined as unwanted conduct with the purpose or effect of violating the dignity of a person and of creating an intimidating, hostile, degrading, humiliating or offensive environment.10 Dignity also plays a central role in preventing the principle of equality from becoming no more than a levelling down exercise. To implement equality by removing benefits, or treating everyone equally badly, would be inconsistent with the fundamental dignity of each individual. The importance of this can be seen by comparing the experience in the EU and South Africa over the question of equalising pension ages. The lower pension age of 60 for women, as against 65 for men, has been contested in both jurisdictions as discriminatory against men. Although older women remain among the poorest in both Europe and South Africa, increasing unemployment for male workers above 50 has meant that access to pension rights at an earlier age is of growing importance to men. In a case brought initially to the UK courts, and subsequently, under EU law, to the ECJ, this claim of discrimination was upheld.11 However, pension funds, claiming that the cost would be exorbitant, reacted by instituting policies to raise women‟s pension age to that of men over time. This strategy has been upheld in further litigation before the ECJ.12 The result is that poor women are worse off, and poor men are no better off.13 In South Africa, by contrast, 9 A Chaskalson „Human Dignity as a Constitutional Value‟ in D Kretzmer & E Klein (eds) The Concept of Human Dignity in Human Rights Discourse (2002) 140. 10 EP & Council Directive 2006/54/EC Article 2(1)(c) and Article 2(2)(a); for the similarly worded definition of racial harassment see Council Directive 2000/43/EC Article 2(3) 11 Barber v Guardian Royal Exchange Assurance Group [1990] IRLR 240 (ECJ) 12 C-408/92 Smith v Avdel Systems Ltd [1994] ECR I-4435 (ECJ) 13 S Fredman ''The Poverty of Equality: Pensions and the ECJ'' [1996] 25 ILJ 91 - 109. when litigation was contemplated on the same issue, the risk of levelling down could be averted by reference to the substantive content of equality, including dignity. Litigation was, in the event, rendered unnecessary the welcome intervention by political authorities in the form of legislative change. Even here, however, a substantive view was taken. In providing for age equalisation for state old age grants, there was no question of increasing women‟s pension age. Instead, the Social Assistance Amendment Bill opens the way for men aged 60 to 64 years to apply for the old age grant, potentially benefitting over 450,000 men. Dignity, however, carries risks. Most serious is the danger that it might operate to negate the link between socio-economic disadvantage and substantive equality. This manifests itself by requiring the claimant to prove, not just that she has been disadvantaged, but that this signifies lack of respect of her as a person. A measure which imposes socio-economic disadvantage on individuals on grounds of their status would not in itself be regarded as discriminatory unless the applicant could prove, in addition, that the measure assails her dignity. This danger has been floridly demonstrated in two recent cases in the Supreme Court of Canada, in which the Court agreed that the claimant suffered socio-economic disadvantage on grounds of her status (in this case, her age), but held that this did not signify that society regarded her of less value than others.14 Her discrimination claim therefore failed. This was particularly problematic in Gosselin,15 where the claimant challenged a scheme according to which full benefit was only available to welfare recipients over 30. Those under 30 received significantly less unless they participated in a designated work activity or education programme. In practice, there was a significant shortfall in places available, and those which were available were relatively short-term. As a result, many young people, including the claimant, experienced real poverty. Nevertheless, the majority of the Court held that „the provision of different initial amounts of monetary support to each of the two groups does not indicate that one group's dignity was prized above the other‟s.‟16 It therefore rejected her claim of breach of the equality guarantee in s.15(1) of the Charter. 14 Law v Canada [1999] 1 SCR 497 (Canadian Supreme Court); Gosselin v Quebec 2002 [SCC] 84 (Canadian Supreme Court) 15 Gosselin v Quebec 2002 [SCC] 84 (Canadian Supreme Court) 16 Gosselin v Quebec 2002 [SCC] 84 (Canadian Supreme Court) para 61 (McLachlin J). A more subtle manifestation of this effect of dignity is found in the Hugo case17, where, again, dignity was used to defeat an equality claim. The case concerned the pardon issued by President Mandela to all women prisoners who were mothers of young children. The pardon was challenged by a male prisoner, the sole carer of his young children, on the basis that it discriminated on grounds of gender. The Court rejected the case. According to Goldstone J, „The Presidential Act might have denied fathers an opportunity it afforded mothers, but it could not be said to have fundamentally impaired their rights of dignity or sense of equal worth.‟18 Yet the opportunity denied to fathers was one which goes to the heart of the gendered division of labour in our society, and therefore to the heart of the socio-economic disadvantage experienced by women. Women‟s role as primary child-carers has been a major source of inequality for women; to attempt to resolve the problem on the basis that men‟s dignity had not been impaired merely sidesteps the reality of gender inequality. Equally problematic is the assertion that the measure did not impair fathers‟ rights of dignity or sense of self worth. As Kriegler J stated in his dissenting opinion, it is the assumption that women are the primary child-carers which constitutes an assault on their dignity: „One of the ways in which one accords equal dignity and respect to persons is by seeking to protect the basic choices they make about their own identities. Reliance on the generalisation that women are the primary care givers is harmful in its tendency to cramp and stunt the efforts of both men and women to form their identities freely…‟19 Dignity can also be used in a way which abstracts the equality analysis from power relations in society. This can be seen in the early case of Walker20. The case concerned a charging system for electricity which favoured the predominantly black and poor residents of the former townships outside Pretoria over the predominantly white and affluent residents of Pretoria itself. The court held that although the charging system itself did not constitute an invasion of the dignity of the white residents, the selective enforcement of the charges constituted unfair discrimination against them. According to Langa DP: „No members of a racial group should be made to feel that they are not deserving of equal „concern, respect and consideration‟ and 17 Case CCT 11/96 President v Hugo Hugo para 47 19 Hugo para 80 20 City Council of Pretoria v Walker 1998 (2) SA 363 (CC) (South African Constitutional Court) 18 that the law is likely to be used against them more harshly than others who belong to other race groups‟.21 These difficulties are not, however, insurmountable. The Canadian Court itself has now recognised that dignity has been used in a problematic way. In R v Kapp,22 the Court acknowledged that „several difficulties have arisen from the attempt in Law to employ human dignity as a legal test. There can be no doubt that human dignity is an essential value underlying the s. 15 equality guarantee. In fact, the protection of all of the rights guaranteed by the Charter has as its lodestar the promotion of human dignity. …But as critics have pointed out, human dignity is an abstract and subjective notion that…has .. proven to be an additional burden on equality claimants, rather than the philosophical enhancement it was intended to be.‟23 More than this, however, dignity should be used in alliance with socio-economic rights to create a powerful means of upholding substantive equality. This can be seen in the judgement of Mokgoro J in Khosa,24 which concerned the exclusion of permanent residents from the right to child benefit and old-age pensions. Mokgoro J emphasised that the consequences of exclusion were not only socio-economic. In addition, the exclusion of permanent residents had a strong stigmatizing effect, creating the impression that they were inferior to citizens and less worthy of social assistance. Permanent residents were in effect „relegated to the margins of society and deprived of what may be essential to enable them to enjoy other rights vested in them under the Constitution‟.25 Thus dignity is the dimension of equality which speaks to our basic humanity. Equality attaches to all individuals, not because of their merit, or their rationality, or their citizenship or membership of any particular group, but because of their humanity. Individuals should not be humiliated or degraded through racism, sexism, violence or other status-based prejudice. But dignity is not a separate and additional 21 Walker para 81 R. v. Kapp 2008 SCC 41 (Supreme Court of Canada) 23 Kapp para 21, 22 24 Khosa and Mahlaule v Minister for Social Development 2004 (6) BCLR 569 (South African Constitutional Court) 25 Khosa para 77. 22 element to socio-economic disadvantage in an equality claim. Socio-economic disadvantage is itself an assault on an individual‟s basic humanity. The recognition and redistributive elements of equality should pull together rather than against each other. 3. Affirmative action (i) Affirmative action and substantive equality One of the main advantages of substantive equality over formal equality is its asymmetry. This means that it is not race or gender per se which is regarded as problematic, but the detriment and disadvantage associated with subordinated groups: black people, women, or other out-groups. This asymmetry means that equality is not necessarily breached by measures which specifically use race or gender as a means of distributing benefits and burdens. Indeed, provided that they aim to benefit the subordinated group, race or gender specific measures may be necessary to achieve substantive equality. Thus, whereas formal equality would regard affirmative action as a breach of equality, substantive equality sees such programs as a means to achieve equality. 26 This understanding of affirmative action has been endorsed by the South African Constitution, which makes express provision for affirmative action as a means of achieving substantive equality27. According to section 9(2): „Equality includes the full and equal enjoyment of all rights and freedoms. To promote the achievement of equality, legislative and other measures designed to protect or advance persons, or categories of persons, disadvantaged by unfair discrimination may be taken.’ The meaning of this provision was elaborated in Van Heerden,28 a claim of race discrimination brought by a white Afrikaner member aggrieved at a measure which enhanced the pension contributions of post apartheid members of Parliament but not pre-apartheid members. The High Court took an emphatically formal view of 26 S Fredman Women and the Law (Oxford Monographs in Labour Law Oxford University Press Oxford 1997); S Fredman Discrimination Law (Clarendon Law Series Oxford University Press Oxford 2002); S Fredman „Reversing Discrimination‟ [1997] 113 Law Quarterly Review 575–600. 27 South African Constitution, s.9(2). 28 Minister of Justice v Van Heerden 2004 (6) SA 121 (CC); 2004 (11) BCLR 1125 (South African Constitutional Court) equality, regarding the relatively advantaged position of the affected white members as irrelevant. It therefore struck down the programme as unfair discrimination. The Constitutional Court reversed the decision. As Moseneke J stressed, instead of being an exception to equality, restitutionary measures are an essential part of it. „What is clear is that our Constitution and in particular section 9 thereof, read as a whole, embraces for good reason a substantive conception of equality inclusive of measures to redress existing inequality. ..Such measures are not in themselves a deviation from, or invasive of, the right to equality guaranteed by the Constitution. They are not “reverse discrimination” or “positive discrimination” as argued by the claimant in this case. They are integral to the reach of our equality protection. In other words, the provisions of section 9(1) and section 9(2) are complementary; both contribute to the constitutional goal of achieving equality to ensure “full and equal enjoyment of all rights.”‟ 29 Similarly, the Canadian Charter makes it clear that measures whose object is to ameliorate the conditions of disadvantaged individuals or groups will not be a breach of the equality guarantee in s.15(1) of the Charter. Thus section 15(2) states: „Subsection (1) does not preclude any law, program or activity that has as its object the amelioration of conditions of disadvantaged individuals or groups including those that are disadvantaged because of race, national or ethnic origin, colour, religion, sex, age or mental or physical disability.‟30 While there was some suggestion in earlier cases that section 15(2) should be read as an exception to the equality guarantee, the Supreme Court of Canada in its 2008 decision in Kapp,31 emphatically held that „Sections 15(1) and 15(2) work together to promote the vision of substantive equality that underlies s. 15 as a whole. Section 15(1) is aimed at preventing discriminatory distinctions that impact adversely on members of groups identified by the grounds enumerated in s. 15 and analogous grounds. This is one way of combating discrimination. However, governments may also wish to combat discrimination by developing programs aimed at helping disadvantaged groups improve their situation. Through s. 15(2), the Charter preserves the right of governments to implement such programs, without fear of challenge under s. 15(1). This is made apparent by the 29 Minister of Justice v Van Heerden 2004 (6) SA 121 (CC); 2004 (11) BCLR 1125 (South African Constitutional Court) para 30. 30 Canadian Charger of Rights s. 15(2) 31 R. v. Kapp 2008 SCC 41 (Supreme Court of Canada) existence of s. 15(2). Thus s. 15(1) and s. 15(2) work together to confirm s. 15‟s purpose of furthering substantive equality.‟32 Moreover, „Section 15(2) supports a full expression of equality, rather than derogating from it.‟33 The same approach, albeit within stricter parameters, can be found in EU law, where several provisions state that, with a view to ensuring full equality in practice, the principle of equal treatment should not prevent a member state from maintaining or adopting measures to prevent or compensate for disadvantages linked to gender34, race or ethnic origin,35 or religion, disability, age and sexual orientation.36 At international level, affirmative action is again endorsed. Thus the Convention on the Elimination of Discrimination Against Women (CEDAW) states that adoption by States Parties of temporary special measures aimed at accelerating de facto equality between men and women „shall not be considered discrimination‟; and Article 2(2) of International Convention on the Elimination of Race Discrimination (CERD), which states: „States Parties shall, when the circumstances so warrant, take, in the social, economic, cultural and other fields, special and concrete measures to ensure the adequate development and protection of certain racial groups or individuals belonging to them, for the purpose of guaranteeing them the full and equal enjoyment of human rights and fundamental freedoms.37 However, there remain important challenges for affirmative action measures. At one level, affirmative action constitutes a valuable synthesis between status and disadvantage, specifically attaching socio-economic benefits to those disadvantaged by status. However, there is a risk that this link perpetuates status inequality. It is not necessarily the case that a programme singling out one group for special protection will in fact advance substantive equality. It may freeze individuals into the very status identity which substantive equality aims to eliminate. It may perpetuate stereotypes. This is particularly true for special measures in favour of women in their child-caring role. Secondly, as the status group begins to benefit from the affirmative action provisions, a gap opens up between status and disadvantage; and the class of 32 Ibid para 16 Ibid para 37 34 Article 141(4) EU 35 Directive 2000/43 EC Article 5 36 Directive 2000/78/EC Article 7(1) 37 CERD Article 2(2) 33 beneficiaries identified by the measure might become over- or under-inclusive. In particular, as the association between race and disadvantage loosens, it may become necessary to include a further criterion, such as that known as the „creamy layer‟ in India, which excludes status members who are no longer socio-economically disadvantaged. Both these questions in turn require a deeper understanding of the relationship between status and socio-economic disadvantage. Behind these issues is the third challenge, namely the extent to which affirmative action can be genuinely transformative, in that it brings about structural change, rather than simply changing the colour or gender composition of classes within the existing structure. These challenges in turn raise questions as to the role of courts, and in particular, the standard of scrutiny which courts should apply in order to determine whether affirmative action measures breach a constitutional equality guarantee. Since these challenges are often fought out in the judicial arena, it is necessary to consider the standard of scrutiny before turning to the challenges themselves. (ii) Standard of Scrutiny Although substantive equality is not breached in principle by affirmative action measures, there remains an important role for judicial supervision. The controlling mechanism used by most courts is that of proportionality: are the aims legitimate, and do the means „fit‟ the aims. However, courts vary widely in the intensity of scrutiny. This is true both for the definition of a „legitimate purpose‟ and for the tightness of the fit between the affirmative action measure and that purpose. This variation is accounted for both by the court‟s own understanding of equality and the extent to which they consider it appropriate to defer to legislative or executive decisionmakers‟ understanding of equality. As is well known, the US courts have in recent years adopted a standard of „strict scrutiny‟.38 Based on an underlying hostility to the substantive equality values behind affirmative action, US courts have framed the proportionality analysis in the narrowest terms. Thus instead of simply requiring a legitimate State interest, the State must demonstrate a „pressing social need‟. Similarly, it is not sufficient for the means 38 Adarand Constructors v Pena 515 U.S. 200 (1995) (United States Supreme Court) to be reasonably related to these ends. Instead, the measure must be „narrowly tailored‟ to that end. Nevertheless, this standard does not revert to a purely formal approach. As O‟Connor J has emphasised, „strict scrutiny‟ does not amount to „fatal in fact‟.39 There are two interests which have been recognised as compelling under the strict scrutiny test. The first is the compelling interest of remedying the effects of past intentional discrimination.40 The second is the interest in diversity.41 Building on the well-known suggestion of Powell J in the Bakke42 case that diversity could be a compelling interest, the Court in Grutter upheld a race conscious policy by the University of Michigan on the grounds that it furthered the legitimate aim of diversity in higher education. Grutter also elaborated on the standards to be applied for an affirmative action or race conscious programme to be „narrowly tailored‟. It is not legitimate to impose inflexible quotas; instead, to fulfil the test of strict scrutiny, a programme must be flexible enough to ensure that each applicant is evaluated as an individual and not in a way that makes race or ethnicity the defining feature of the application. Thus race or ethnicity may be considered, but only as a “ „plus‟ in a particular applicant's file.‟43 At the same time, the Court rejected the argument that the Law School should have used race-neutral means to obtain the educational benefits of student body diversity. „Narrow tailoring does not require exhaustion of every conceivable race-neutral alternative or mandate that a university choose between maintaining a reputation for excellence or fulfilling a commitment to provide educational opportunities to members of all racial groups.‟44 The standard of strict scrutiny is nevertheless clearly inconsistent with a genuine adherence to the principle of substantive as against formal equality. Does this mean, however, that Courts should entirely abdicate their supervisory function? The Canadian Court in Kapp took the view that review should be primarily directed to the genuineness of the government‟s stated ameliorative purpose. As will be recalled, section 15(2) states that a programme does not breach the equality guarantee if it has 39 Adarand Constructors v Pena 515 U.S. 200 (1995) (United States Supreme Court) at 237 Parents Involved in Community Schools v. Seattle School Dist. No. 1 127 S.Ct. 2738 (2007) (US Supreme Court) at 2752. 41 Grutter v. Bollinger 539 U.S. 306; 123 S.Ct. 2325 (2003) (US Supreme Court) at 328. 42 Regents of the University of California v Bakke 438 U.S. 265 (1978) (United States Supreme Court) 314-315. 43 Grutter 539 US at 317. 1 44 Grutter 539 US at 309. 40 „as its object the amelioration of conditions of disadvantaged individuals or groups‟. The Court rejected the argument that judicial review should focus on the outcome of a measure, since results might be unpredictable or difficult to evaluate. Instead, as required by the text of the provision, the focus should be on the purpose. Nor did amelioration need to be the sole purpose, as long as it constituted a genuine purpose. So far as the fit between means and ends was concerned, the Court required no more than a correlation between the program and the purpose. „Analysing the means employed by the government can easily turn into assessing the effect of the program. As a result, to preserve an intent-based analysis, courts could be encouraged to frame the analysis as follows: Was it rational for the state to conclude that the means chosen to reach its ameliorative goal would contribute to that purpose? For the distinction to be rational, there must be a correlation between the program and the disadvantage suffered by the target group. Such a standard permits significant deference to the legislature but allows judicial review where a program nominally seeks to serve the disadvantaged but in practice serves other, non-remedial objectives.‟ 45 There is force in the Canadian courts‟ concern with the purpose of a programme rather than its effects. The Kapp case concerned the federal government‟s strategy to enhance aboriginal involvement in commercial fishing. As part of the strategy, the government had issued a communal fishing licence to three aboriginal bands permitting only fishers designated by the bands to fish for salmon in the mouth of the Fraser River for a period of 24 hours and to sell their catch. The appellants, commercial fishers, mainly non-aboriginal, were excluded from the fishery during this 24-hour period. They argued that the communal fishing licence discriminated against them on the basis of race. The trial judge found upheld their claim. In the Supreme Court of Canada , Mr Kapp argued that the program did not offer a benefit that effectively tackled the problems faced by these bands. The Court was rightly reluctant to adjudicate on the effectiveness of the programme in this sense. As it stated, „If the sincere purpose is to promote equality by ameliorating the conditions of a disadvantaged group, the government should be given some leeway to adopt innovative programs, even though some may ultimately prove to be unsuccessful. The 45 Kapp para 49 government may learn from such failures and revise equality enhancing programs to make them more effective.‟46 Nevertheless, it was alive to the fact that courts have regarded a programme as ameliorative in surprising circumstances. Particularly worrying are measures which place restrictions on individuals which the Government claims to be in their best interests.47 Thus the Court drew a careful line between deference to the judgement of the executive, and appropriate judicial supervision. „The meaning of “amelioration” deserves careful attention in evaluating programs under s. 15(2). We would suggest that laws designed to restrict or punish behaviour would not qualify for s. 15(2) protection. Nor, as already discussed, should the focus be on the effect of the law. This said, the fact that a law has no plausible or predictable ameliorative effect may render suspect the state‟s ameliorative purpose. Governments, as discussed above, are not permitted to protect discriminatory programs on colourable pretexts.‟48 The South African Constitutional Court, while rejecting the strictness of „narrow tailoring‟, has nevertheless retained for itself a role which goes beyond checking only for genuine ameliorative purpose. Instead, it requires a measure to meet the standard of reasonableness. Although the constitutional text also talks of measures „designed to protect and advance‟, in van Heerden, Moseneke J stated that the remedial measures must be „reasonably capable of attaining the desired outcome‟, namely to protect or advance individuals or categories of persons who have been disadvantaged by unfair discrimination. This excludes measures which are arbitrary, capricious or display naked preference, or are not reasonably likely to achieve the end of advancing or benefiting the interests of those who have been disadvantaged by unfair discrimination. 49 Importantly, he emphasised that the Constitution did not „postulate a standard of necessity between the legislative choice and the governmental objective. The text requires only that the means should be designed to protect or advance. It is sufficient if the measure carries a reasonable likelihood of meeting the end.‟50 46 Kapp para 47 Kapp para 53 48 Kapp para 54 49 Minister of Justice v Van Heerden 2004 (6) SA 121 (CC); 2004 (11) BCLR 1125 (South African Constitutional Court) para 38 - 40 50 Van Heerden para 41 47 This can be contrasted with the standard of scrutiny applied by the European Court of Justice, which continues to regard affirmative action as a derogation from the principle of equality, and therefore to be construed strictly. According to the ECJ: „In determining the scope of any derogation from an individual right such as the equal treatment of men and women laid down by the Directive, due regard must be had to the principle of proportionality, which requires that derogations must remain within the limits of what is appropriate and necessary in order to achieve the aim in view and that the principle of equal treatment be reconciled as far as possible with the requirements of the aim thus pursued.‟51 Ultimately, the role of the court should be to further substantive equality, and therefore to support government measures which aim to achieve substantive equality. There is no equivalence between classification which perpetuates disadvantage or causes detriment, and measures which use status in order to achieve substantive equality. However, judicial support for affirmative action which furthers substantive equality should not be confused with deference to governmental decisions. Courts should still require governments to demonstrate that the aim of the measure is indeed to further substantive equality and to justify their choice of means. Justification need not amount to proof that there were no suitable alternatives, nor should the State have to demonstrate that the programme is effective. This preserves the space for innovation, as emphasised by the Canadian Court. But justification does require that the State demonstrate that its measures are not based on assumptions, generalisations or stereotypes. In both Hugo and Gosselin, the South African and Canadian Courts were prepared simply to assume facts in favour of the State. Thus in Hugo, although the President had relied exclusively on his desire to protect the interests of young children, the Court was particularly influenced by what it regarded as the impossibility of extending the remission to fathers of young children. „Male prisoners outnumber female prisoners almost fiftyfold. A release of all fathers would have meant that a very large number of men prisoners would have gained their release. As many fathers play only a secondary role in child rearing, the 51 Briheche 24, C-476/99 Lommers v Minister Van Landbouw, Natuurbeheer En Visserij [2004] 2 C.M.L.R. 49 (ECJ) at [39], E-1/02 EFTA Surveillance Authority v Norway [2003] 1 C.M.L.R. 23 (EFTA Court) at [37] –[43] release of male prisoners would not have contributed as significantly to the achievement of the President‟s purpose as the release of mothers. In addition, the release of a large number of male prisoners in the current circumstances where crime has reached alarming levels would almost certainly have led to considerable public outcry. In the circumstances it must be accepted that it would have been very difficult, if not impossible, for the President to have released fathers on the same basis as mothers. Were he obliged to release fathers on the same terms as mothers, the result may have been that no parents would have been released at all.‟ Yet as Kriegler J stated, the President had not sought to justify his decision by reference to the numbers of male prisoners who would have gained their release and no evidence was led to that effect. Nor was there evidence to the effect that such numbers would have produced public disquiet, nor did the Court know anything about the administrative burden of weighing the family circumstances of individual prisoners. The Court simply assumed these issues by way of justification for the differential treatment. This technique, namely a reliance on judicial notice or assumptions of unproved facts, is a familiar one for a Court to resort to when it is unwilling to deflect what it regards as a as a legitimate government policy to further substantive equality. In the Gosselin case in Canada, for example, the Court accepted without substantiating evidence the State‟s assertion that it was easier for young people to find jobs than older people. Judicial deference in this context operates as an obstacle to transformation because it protects the State from having to explain its decision to exclude a particular group from a particular benefit. Respect for the democratic process requires greater, not less, attention to the duty to account and explain. Providing an explanation for a decision enhances the accountability of decision-makers, exposing them to public scrutiny and debate. Requiring the state to provide a reasoned explanation does not require the judges to substitute their own opinions for those of policy makers on resource allocation questions. The judicial role in applying substantive equality remains a supervisory one, to guard against stereotypical assumptions and unwarranted generalisations which can cause or perpetuate disadvantage. This role is not easy to define. However, by protecting the state from having to explain an allocation in the first place, courts abdicate their rightful role in developing an appropriate judicial standard of scrutiny52. (iii) Achieving equality or reinforcing stereotypes: pregnancy and parenthood The dangers of avoiding the formulation of an appropriate level of scrutiny are highlighted in relation to pregnancy and maternity, where a deep sensitivity to the structural nature of gender discrimination is necessary to check whether a measure is likely to bring about change or simply reinforce stereotypes. Whereas special measures for women are necessary for child-bearing, the question is more complex in relation to child-care and family work. Genuinely transformative change can only occur when both parents are equally responsible for child-care53. Special measures for women, however well-intentioned, run the risk of reinforcing their primary role as child-carers, and therefore perpetuating their disadvantage. This does not mean that there should be no special provision for parents. That would be to revert to formal equality, which contemplates levelling-down as an appropriate response. Instead, such measures will only achieve real change if they refer to both parents. Yet, in a string of cases in various jurisdictions, courts have upheld measures which excluded men from rights related to parenting on the ground that they achieved equality for women by recognising their position as mothers. We have already seen this in the Hugo case, where mothers of minor children pardoned, but not fathers. As Kriegler J stated in his dissent, the stereotype that fathers were not responsible for children and that mothers were had a more invidious effect than the benefit to the few hundred women released under the scheme: „The benefits in this case are to a small group of women - the 440 released from prison and the detriment is to all South African women who must continue to labour under the social view that their place is in the home. In addition, men must continue to accept that they can have only a secondary/surrogate role in the care of their children. …There is also more diffuse disadvantage when society imposes roles on men and 52 See further S Fredman ''Providing Equality: Substantive Equality and the Positive Duty to Provide'' (2005) 21 SAJHR 163 53 See further S Fredman Women and the Law women, not by virtue of their individual characteristics, qualities or choices, but on the basis of predetermined, albeit time-honoured, gender scripts..‟54 A similar result can be seen in the European Union case-law. The European Court of Justice has generally been strict in its scrutiny of affirmative action measures, regarding them as a derogation from the equality principle. But in a recent case, it held that to provide child-care facilities for working mothers and not working fathers was a measure which fell within the exception in EU law for measures which, „although discriminatory in appearance, are in fact intended to eliminate or reduce actual instances of inequality which may exist in the reality of social life‟.55 The case concerned a programme according to which a work-place nursery provided places for children of female but not male employees. The measure was defended by the employer, the Dutch Ministry of Agriculture, on the basis that the distinction drawn on grounds of sex reflected its determination to tackle the under-representation of women in the Ministry. According to Dutch law, the programme was legitimate because it was a “ distinction intended to place women in a privileged position in order to eliminate or reduce de facto inequalities” and was reasonable in relation to that aim.56 The European Court of Justice endorsed this approach. The Court pointed out that there was a significant under-representation of women in the Ministry of Agriculture and particularly in the higher grades, and that there was similarly a proven insufficiency of suitable and affordable nursery facilities, which, it regarded as likely to induce women employees to give up their jobs. The Court was not oblivious to the danger that „a measure such as that at issue …, whose purported aim is to abolish a de facto inequality, might nevertheless also help to perpetuate a traditional division of roles between men and women.‟57 However, it was not prepared to hold that the measure was disproportionate for failing to include working fathers, particularly in light of the very limited number of nursery places available, and the fact the employer could grant requests from male officials in cases of emergency, to be determined by the employer. The Court did stress that a measure 54 Hugo para 83 C-476/99 Lommers v Minister Van Landbouw, Natuurbeheer En Visserij [2004] 2 C.M.L.R. 49 (ECJ) para 32; see Equal Treatment Directive, Article 2(4). 56 Art.5 of the Netherlands Law on Equal Treatment of Men and Women. 57 C-476/99 Lommers v Minister Van Landbouw, Natuurbeheer En Visserij [2004] 2 C.M.L.R. 49 (ECJ) para 41 55 which excluded male officials who take care of their children by themselves from access to a nursery scheme subsidised by their employer would go beyond the permissible derogation. But it seemed reassured on this score by the indication given by the Ministry of Agriculture that male officials who bring up their children by themselves should, on that basis, have access to the nursery scheme at issue. Valuable guidance as to how to distinguish between discrimination and affirmative action in this context is given by the CEDAW Committee. In its Recommendation 25 on special temporary measures, it states that „women‟s biologically determined permanent needs and experiences should be distinguished from other needs that may be the result of past and present discrimination against women by individual actors, the dominant gender ideology, or by manifestations of such discrimination in social and cultural structures and institutions.‟ Moreover, „as steps are being taken to eliminate discrimination against women, women‟s needs may change or disappear, or become the needs of both women and men. Thus, continuous monitoring of laws, programmes and practices directed at the achievement of women‟s de facto or substantive equality is needed so as to avoid a perpetuation of non-identical treatment that may no longer be warranted.‟58 (iv) Demarcating beneficiaries The second challenge for courts in their supervisory role concerns the question of whether the class of beneficiaries is appropriately delineated, or whether it is over- or under-inclusive. A key strength of substantive equality is its move away from the individualism of formal equality, which requires proof that an individual has suffered discrimination at the hands of an identified perpetrator before an individual remedy can be granted. Affirmative action, understood as an element of substantive equality, does not require proof of individual detriment in this sense. Therefore the class of beneficiaries need not consist only of proved victims. Nevertheless, there needs to be some correlation between the beneficiaries and the previous disadvantage. This raises 58 General recommendation No. 25, on article 4, paragraph 1, of the Convention on the Elimination of All Forms of Discrimination against Women, on temporary special measures, para 11. the important question of the relationship between socio-economic disadvantage and status groups. If the aim of the measure is to address socio-economic disadvantage, then a group demarcated by status, such as race or gender, might be over-inclusive in including wealthier blacks or women, and under-inclusive, in excluding poor white men. If the aim is to address mis-recognition or status inequalities, then the demarcation might be appropriate. It is here that the controlling function of proportionality comes in: are the means appropriately correlated with the ends? But it is here too that the problem is overlaid by the need to take account of relative institutional competence. Should judges be leaving this decision to law-makers and the executive, or should they be exercising some supervision? Both the Canadian and South African courts have addressed this question. In Kapp, the question arose as to whether the three Indian bands were an „identifiable disadvantaged group‟ for the purposes of section 15(2), which it will be recalled, permits ameliorative measures whose object is „the amelioration of conditions of disadvantaged individuals or groups including those that are disadvantaged because of race, national or ethnic origin, colour, religion, sex, age or mental or physical disability‟. It is notable that this formulation focuses on status as a means of identifying the group, but does not distinguish openly between attendant socioeconomic disadvantage and status disadvantage. The Court in Kapp regarded them as intertwined. Thus, having identified the distinction as being based on race, it went on to establish whether, as required by s15(2), the program targeted a „disadvantaged group identified by‟ race. For this, it referred to both status disadvantage - “the legacy of stereotyping and prejudice against Aboriginal peoples” - and socio-economic disadvantage: “the evidence shows in this case that the bands granted the benefit were in fact disadvantaged in terms of income, education and a host of other measures.” However, it did not require the classification to be solely concerned with such disadvantage. Thus it held that not all members need be disadvantaged. „The fact that some individual members of the bands may not experience personal disadvantage does not negate the group disadvantage suffered by band members.‟59 59 Kapp para 59 In Van Heerden, a similar question was faced. Although an overwhelming majority of the new members of Parliament were excluded from parliamentary participation by past apartheid laws, not all new parliamentarians of 1994 belonged to the class of persons prejudiced by past disadvantage and unfair exclusion. Moseneke J recognised that it would often be „difficult, impractical or undesirable to devise a legislative scheme with “pure” differentiation demarcating precisely the affected classes. Within each class, favoured or otherwise, there may indeed be exceptional or “hard cases” or windfall beneficiaries. That however is not sufficient to undermine the legal efficacy of the scheme. The distinction must be measured against the majority and not the exceptional and difficult minority of people to which it applies…‟60 He thus held that the measure should be judged by „[whether an overwhelming majority of members of the favoured class are persons designated as disadvantaged by unfair exclusion.‟ 61 Like the Canadian Court, he regarded the validity of the remedial measures as unaffected by the existence of a tiny minority of members of Parliament who were not unfairly discriminated against, but who benefited from the measure. This can be contrasted with the decision of Mogkoro J, who held that a more exact fit was necessary. If the class of beneficiaries could not be precisely delineated, then the measure should have to reach the high standards of fairness required in relation to an ordinary discrimination claim. She regarded the measure as too loosely related to a proscribed group to fall within section 9(2), which relieves the state of the burden of proving unfairness. Instead, it needed to pass muster under section 9(3), which, on the facts, it did. The difference between Moseneke and Mokgoro JJ reflects an underlying principle about the role of judicial scrutiny. Requiring a higher level of scrutiny does not in itself mean that affirmative action programmes will be struck down as a breach of equality. Judges could exercise supervisory control without reverting to a formal equality approach. But this means that more attention needs to be given to the criteria of review which would ensure that an affirmative action measure is consistent with substantive equality. Most important in the context of the delineation of the beneficiary group is the relationship between redistribution and recognition, or socioeconomic disadvantage and status inequality. This relationship need not achieve much 60 61 Van Heerden para 39 Van Heerden para 40 attention when there is a substantial overlap, as in the case of the First Nation tribes in the Kapp case. However, as affirmative action measures become more effective, and some members of the status group begin to prosper, questions arise as to the legitimacy of an over-inclusive group. One response is to overlay status with socio-economic disadvantage in demarcating the group of beneficiaries. To qualify for the benefit, the individual must show socioeconomic disadvantage as well as membership of the status group. This can be seen in both India and the US. The Indian Constitution permits special provision to be made for two categories of disadvantaged groups. The first are known as „Scheduled Castes and Scheduled Tribes‟ and are specified by the President.62 The second are referred to as „socially and educationally backward classes of citizens,‟63 or other „backward classes‟.64 The latter are specified in a list drawn up by the National Commission for Backward Classes.65 The equality guarantee in the Indian Constitution provides that special provision may be made for the advancement of any these categories.66 Particularly controversial has been the question of whether the reservations should be available to the „creamy layer‟, or those members of the certified groups which are in fact no longer socially or educationally disadvantaged. In two of the foremost cases on reservations, the Indian Supreme Court has made it clear that exclusion of the creamy layer is not an addition principle, but one that goes to the very purpose of the reservation. As Justice Jeevan Reddy stated in Indra Sawhney's case, „In our opinion, it is not a question of permissibility or desirability of such test but one of proper and more appropriate identification of a class as a backward class.‟67 Similarly, the Chief Justice in Ashoka Kumar stated: „To fulfil the conditions and to find out truly what is socially and educationally backward class, the exclusion of "creamy layer" is essential.‟68 In other words, the definition of the beneficiary class must correspond with the purpose of the provision, namely to advance those who are disadvantaged. However, the requirement that a group display 62 Constitution of India http://lawmin.nic.in/coi/coiason29july08.pdf Articles341(1), 342(2). See Constitution of India s15(4) (http://lawmin.nic.in/coi/coiason29july08.pdf) 64 See Constitution of India Article 16(4) 65 See Constitution of India Article 340(1); the list can be found at http://ncbc.nic.in/backwardclasses/index.html 66 Article 15(4) 67 Indra Sawhney p.724. 68 Ashoka Kumar para 149. 63 both misrecognition and socio-economic disadvantage is only imperfectly executed. On the one hand, there is no creamy layer exclusion for the Scheduled Tribes or Scheduled Castes. On the other hand, there is no provision for reservation in favour of disadvantaged Muslims. A similar attempt to create a closer fit between status and socio-economic disadvantage is found in US affirmative action programmes, particularly in respect of set-aside programmes authorizing preferential treatment in the award of government contracts. In order to comply with the strict scrutiny test, Federal legislation mandating set asides for status groups now also includes a requirement of evidence of socio-economic disadvantage. This can be seen in the major vehicle for present-day set-asides, section 8(a) of the Small Business Act (SBA). 69 To participate in what has become known as a section 8(a) program, a business must be 51% owned by individuals who qualify as “socially and economically disadvantaged.” 70 The Act defines “socially disadvantaged individuals” as “those who have been subjected to racial or ethnic prejudice or cultural bias because of their identity as a member of a group without regard to their individual qualities.”71 It defines “economically disadvantaged individuals” as “those socially disadvantaged individuals whose ability to compete in the free enterprise system has been impaired due to diminished capital and credit opportunities as compared to others in the same business area who are not socially disadvantaged.”72 There is a rebuttable presumption that members of certain groups are socially disadvantaged, including: Black Americans; Hispanic Americans; Native Americans; Asian Pacific Americans; subcontinent Asian Americans and members of other groups designated from time to time by SBA.73 The presumption of social disadvantage may be overcome with „credible evidence to the contrary‟, which can be submitted by third parties74. An individual who is not a member of a listed group can also be included if she or he can „establish individual social disadvantage by a preponderance of the evidence.75‟ The presumption does not apply 69 15 U.S.C. § 631 et seq. As defined in 13 C. F. R. § 124.105 71 15 U.S.C. § 637(a)(5) 72 15 U.S.C. § 637(a)(6)(A). 73 13 C.F.R § 124.103 (b) 74 13 C.F.R § 124.103 (b)(3) 75 13 C.F.R § 124.103(c) 70 to economic disadvantage: each participant to must prove such disadvantage.”76 The result is that businesses which, regardless of race, are not in fact socially or economically disadvantaged are excluded, a similar effect to that achieved by the Indian „creamy layer‟ provisions above. Amendments adopted in 1994 specifically included “small business concerns owned and controlled by women” in addition to “socially and economically disadvantaged individuals.” Also of importance has been the Transportation Equity Act for the 21st Century (known as TEA-21),77 which authorizes the use of race- and sex-based preferences in federally funded transportation contracts. As in the SBA regulations, the TEA-21 regulations presume that Black Americans, Hispanic Americans, Native Americans, Asian-Pacific Americans, Subcontinent Asian Americans, and women are socially and economically disadvantaged,78 a presumption which is rebutted where the individual has a personal net worth of more than $750,000 or a preponderance of the evidence demonstrates that the individual is not in fact socially and economically disadvantaged.79 Firms owned and controlled by someone who is not presumed to be disadvantaged (i.e., a white male) can qualify for DBE status if the individual can demonstrate that he is in fact socially and economically disadvantaged.80 This raises deep questions as to the function of affirmative action and the relationship between status and socio-economic disadvantage. By requiring individuals to manifest both socio-economic disadvantage and membership of an out-group, it is assumed that the function of affirmative action is only in relation to the socio-economic dimension of status wrongs. Pure socio-economic disadvantage without status wrongs and pure status wrong without socio-economic disadvantage both fall outside the affirmative action purview. It could be argued that affirmative action is an inappropriate remedy in cases in which status wrongs need to be addressed but socio-economic disadvantage is no longer an issue. On this view, status-only wrongs are better addressed through other measures, such as prohibitions on discrimination and 76 13 C. F. R. § 124.104 Transportation Equity Act for the 21st Century (TEA-21), extended through FY2009 by P.L. 109-59, signed into law during the 109th Congress. 78 .§ 26.67(a). 79 .§ 26.67(b). 80 .§ 26.67(d). 77 harassment. Indeed, to continue to use affirmative action for purely status wrongs could, on this view, simply reinforce stereotypes. However, this unnecessarily narrows the role of affirmative action. Its aim should not be regarded as merely redistributive. Affirmative action also functions in three further ways.81 Firstly, hidden and structural barriers to advancement can be more easily overcome by affirmative action than by individual claims for indirect discrimination. Despite apparently objective eligibility standards and ostensible equal opportunity policies, there remain many hidden obstacles to the advancement of women, historically disadvantaged South Africans, minorities and others. On the assumption that, in the absence of barriers, there would be a random spread of men and women, blacks and whites, and members of different ethnic groups across the labour force and government, the very fact that a group is seriously under-represented in a sphere or activity is evidence of the subtle operation of often invisible barriers. Yet could this not be dealt with by the familiar principle of indirect discrimination? Indirect discrimination expressly aims to remove apparently neutral barriers which in fact function to exclude more women than men or more blacks than whites unless they can be justified.82 However, it has proved to be too clumsy a tool to achieve its aims. Applicants seeking to prove indirect discrimination have initiate court proceedings and show disparate impact, often on the basis of complex statistics. Finally, even if she can surmount all these barriers, she may find that an employer successfully shows that the criteria, despite being exclusionary, are justifiable by reference to the needs of the business. Affirmative action resolves many of these difficulties. Instead of relying on litigation by individual victims, the employer takes the initiative. Nor is it necessary to prove that an exclusionary rule has had a disproportionate impact. Instead, it is sufficient to demonstrate a clear pattern of under-representation of women in particular grades or occupations. The complex questions above are unnecessary. Moreover, discriminatory selection criteria are unequivocally removed: by creating a presumption in favour of women in conditions of equal merit, it makes it impossible 81 The remainder of this section is taken from S Fredman Discrimination Law (Clarendon Law Series Oxford University Press Oxford 2002), chap 5. 82 Sex Discrimination Act 1975, s. 1(1)(b)(i). for such criteria to be reintroduced surreptitiously through subjective decisionmaking. Phrased in this way, affirmative action can be legitimated as an effective means of overcoming hidden barriers. At the same time, this formulation reveals its very limited impact. Most importantly, while preference policies may change the gender or racial composition of some higher paid occupations, they do not challenge the underlying structural and institutional forces leading to the discrimination. As Young argues83, because affirmative action diagnoses the problem as one of maldistribution of privileged positions, its objective is limited to the redistribution of such positions among under-represented groups. However, this narrow distributive definition of racial and gender justice leaves out the equally important issues of institutional organisation and decision-making power. The under-representation of women and historically disadvantaged South Africans in higher positions in the employment ladder, both public and private, is only partially solved by inserting some women or black South Africans into those positions. While some women „make it to the top‟, the vast majority will remain in poorly paid, low status jobs. It is not surprising that, in practice, reverse discrimination is often found to do no more than favour middle class women or blacks who are already relatively privileged in society.84 For fundamental change to occur, the structural and institutional causes of exclusion need to be changed, including the division of labour in the home, the interaction between work in the family and work in the paid labour force, education and others. A second, more dynamic way of justifying the use of affirmative action policies is to argue that the very presence of women or blacks in higher status positions will lead to structural changes. This is to argue, as Young does, that the particular life experience of the decision-maker is reflected in his or her view. Since gender and race remain such strong determinants of a person‟s life experience, the overwhelming 83 I. Young, Justice and the Politics of Difference (Princeton University Press, 1990), p. 193. 84 W. F. Menski „The Indian Experience and its Lessons for Britain‟ in B. Hepple and E. Szyszcak (eds.), Discrimination and the Limits of the Law (Mansell, 1992), 300 at p. 330. predominance of one gender or race in decision-making fora make it unlikely that the experience and perspectives of the excluded group will be articulated85. Indeed, a recent study in Britain demonstrated that the biggest barrier to advancement for ethnic minorities, women and disabled people within the senior Civil Service is believed to be a deeply embedded culture which acts to exclude those who are different from traditional Civil Service employees, who are generally middle class middle-aged white men86. On this view, it is possible to characterise the presence of women and historically disadvantaged groups as functioning to open up new perspectives on decision-making, to cast light on assumptions that the dominant group perceives as universal, and to enhance the store of „social knowledge.‟ This approach makes sense of the notion that women or minorities may have distinct perspectives, which the very process of exclusion negates, and therefore which need to be guaranteed a place in deliberative decision-making. In addition, it demonstrates the necessity for a critical mass both to reflect a diversity within the social group in question and to make the common interests more audible. At the same time, it is important to move beyond essentialising a status group, or regarding all members as sharing the same interests or perspectives. Young argues that groups are better understood, not as fixed categories with impermeable boundaries, but as a set of relationships between different people. Such a relational understanding moves beyond the notion that a group consists of members who all share the same fixed attributes and have nothing in common with members of other groups. Instead, a group is characterised as a social process of interaction in which some people have an affinity with each other. Assertion of affinity with a group may change with social context and with life changes; and members may have interests which differ from other members of the group but are similar to members of other groups87. A third and related justification for reverse discrimination is that it provides diversity in an educational institution or workplace; and that it facilitates the provision of role 85 Phillips, above n. 97, p. 52. 86 (1999) 87 EOR 4. 87 Young, above n. 93, pp. 171-2 models. In Bakke88, Powell J justified affirmative action in university admissions thus: „An otherwise qualified medical student with a particular background - whether it be ethnic, geographic, culturally advantaged or disadvantage - may bring to a professional school of medicine, experiences, outlook, and ideas that enrich the training of its student body and better equip its graduates to render with understanding their vital service to humanity.‟89 In other words, where a group has been excluded from a particular setting, be it a workforce or an educational institution, the likelihood is that the perspectives and experiences of members of the excluded group, particularly those relating to its exclusion, will be undervalued, misunderstood or ignored by the dominant group, making it impossible for the excluded group to change its disadvantaged position. While diversity operates to change the perspectives of the dominant group, the provision of role models operates on the self-perception of excluded groups, piercing stereotypes and giving them the self confidence to move into non-traditional positions. This too suggests the importance of critical mass. This is well illustrated by Canadian Supreme Court in Action Travail des Femmes v Canadian National Railway Co .90As Chief Justice Dickson put it, the aim of an employment equity programme (in this case setting a quota of one woman in four new hirings until a goal of 13 per cent women in certain blue collar occupations was reached) is not to compensate past victims; but "an attempt to ensure that future applicants and workers from the affected group will not face the same insidious barriers that blocked their forebears".91 He identified at least two ways in which such a programme is likely to be more effective than one which simply relies on equal opportunities or the proscription of intentional prejudice. Firstly, the insistence that women be placed in non-traditional jobs allows them to prove that they really can do the job, thereby dispelling stereotypes about women's abilities. This was particularly evident in the case at hand, in which the quotas ordered by the tribunal concerned traditionally male 88 Regents of the University of California v Bakke 438 U.S. 265 (1978) (United States Supreme Court) ibid., at 314, 2760. Upheld in Grutter v. Bollinger 539 U.S. 306; 123 S.Ct. 2325 (2003) (US Supreme Court) 89 90 Action Travail des Femmes v Canadian National Railway Co [1987] 1 SCR 1114, 40 DLR (4th) 193. 91 ibid. at 213. jobs such as "brakeman" or signaller at Canadian National Railways. Secondly, an employment equity programme helps to create a "critical mass" of women in the work-place. Once a significant number of women are represented in a particular type of work, "there is a significant chance for the continuing self-correction of the system."92 The critical mass overcomes the problem of tokenism, which would leave a few women isolated and vulnerable to sexual harassment or accusations of being imposters. It would also generate further employment of women, partly by means of the informal recruitment network and partly by reducing the stigma and anxiety associated with strange and unconventional work. Finally, a critical mass of women forces management to give women's concerns their due weight and compels personnel offices to take female applications seriously. As the Chief Justice concluded: "It is readily apparent that, in attempting to combat systemic discrimination, it is essential to look to the past patterns of discrimination and to destroy those patterns in order to prevent the same type of discrimination in the future."93 At the same time, it is important to stress the limitations of affirmative action as a strategy. The introduction of new perspectives, while an important goal, can only have a limited impact: entrenched structures are often resilient and indeed have powerful conformist pressures. Women, blacks or minorities may find themselves forced to hide their views and ignore their own needs and interests in order to ensure that their continued participation is viable. Even if they do articulate their perspectives, the process of recognition and affirmation is halting and erratic. Thus affirmative action needs to be only one part of a broad based and radical strategy, which does more than redistribute privileged positions but refashions the institutions which continue to perpetuate exclusion. 92 ibid. at 214. 93 ibid. at 215. 4. Should poverty be a ground for discrimination? It has been argued thus far that a major strength of substantive equality is its ability to achieve a synthesis between status inequality and socio-economic disadvantage. Poverty, however, is generally associated only with socio-economic disadvantage. It is often thought not to correspond to any status group, and similarly not to give rise to recognition inequality. For this reason, only measures dealing with socio-economic disadvantage are thought to be appropriate, and poverty is not regarded as being a separate ground within discrimination law. These assumptions are vividly demonstrated by the United States Supreme Court, which has refused to shape the Constitutional equality guarantee in the US Constitution to cover groups defined only by their poverty. In a case in which the claimants argued that a school system funded by local taxes discriminated against those who lived in poor areas, the Court drew a bright line between claims based on poverty and those traditionally identified as „suspect‟ under the Equal Protection Clause. The difference between status and poverty is highlighted in the judgement of Powell J: „The system of alleged discrimination and the class it defines have none of the traditional indicia of suspectness: the class is not saddled with such disabilities, or subjected to such a history of purposeful unequal treatment, or relegated to such a position of political powerlessness as to command extraordinary protection from the majoritarian political process.‟94 Questions of this sort, it held, were inappropriate for strict scrutiny since they involved „the most delicate and difficult questions of local taxation, fiscal planning, educational policy, and federalism, considerations counselling a more restrained form of review‟.95 The Court found it particularly difficult to accept that there could be justiciable criteria for what amounted to poverty: „The Texas system of school financing might be regarded as discriminating (1) against „poor‟ persons whose incomes fall below some identifiable level of poverty or who might be characterized as functionally „indigent,‟ or (2) against those who are relatively poorer than others, 94 San Antonio Independent School District v. Rodriguez 411 U.S. 959, 93 S.Ct. 1919. (U.S.Supreme Court)1294. 95 Ibid 1302. or (3) against all those who, irrespective of their personal incomes, happen to reside in relatively poorer school districts. Yet appellees have not defined the term „poor‟ with reference to any absolute or functional level of impecunity.‟96 Even if it had been able to define poverty, the Court was unwilling to view redistribution in egalitarian terms. Instead, it was „“a sufficient answer to appellees” argument that, at least where wealth is involved, the Equal Protection Clause does not require absolute equality or precisely equal advantages‟.97 However, this approach entirely ignores the fact that poverty is not only about socioeconomic disadvantage. It is also about recognition inequality. Welfare claimants are frequently stigmatised as scroungers, lazy, or irresponsible, and authorities often devote disproportionate resources to „welfare fraud‟, as compared to fraud committed by powerful politicians or entrepreneurs. Bruce Porter, describing Canadian welfare fraud campaigns, describes the consequences vividly in terms of recognition inequality: „It is difficult to appreciate just how profound is the social exclusion that results from this kind of government-endorsed promotion of discrimination and scapegoating. It transforms social assistance from an entitlement of citizenship linked with the right to security and dignity, into a source of shame, guilt, and insecurity.‟98 Should legislation therefore prohibit discrimination on grounds of poverty? To answer this question, it is worth revisiting the four dimensions of equality set out in the first part of this paper. These were to, first, to break the cycle of disadvantage associated with status or out-groups; secondly, to promote respect for the equal dignity and worth of all, thereby redressing stigma, stereotyping, humiliation, and violence; thirdly to accommodate and positively affirm different identities; and finally to redress under-representation in decision-making and social exclusions. It is clear that anti-discrimination law cannot address the primarily economic manifestations of poverty: the welfare state or socio-economic rights generally aim to redress socioeconomic disadvantage. Nevertheless, it remains paradoxical that disadvantaged or socially excluded individuals who are not members of historically stigmatised status groups gain no protection, and might even be called upon to make way for members 96 Ibid 1289. Ibid 1291. 98 Porter „Claiming Adjudicative Space: Social Rights, Equality and Citizenship‟. 97 of status groups in respect of affirmative action programmes. Instead, including poverty or social exclusion as a ground of discrimination would add a „recognition‟ dimension to this primary emphasis on redressing economic disadvantage within the welfare state. In this context, the role of anti-discrimination law would be to prohibit stigma and hostility, to affirm individual dignity and worth, and to facilitate social inclusion and participation in decision-making. South Africa has tentatively included „socio-economic status‟ as a potential ground of unfair discrimination, although at present, this ground is only a „directive principle‟ and awaits further activation.99 It is also possible to argue that poverty, or socioeconomic status, is an unenumerated ground under s.9 of the Constitution. This leaves open the question of how to define the status-ground. The South African Promotion of Equality and Prevention of Unfair Discrimination Act 4 of 2000 defines socioeconomic status to include the „social or economic condition or perceived condition of a person who is disadvantaged by poverty, low employment status, or lack of or lowlevel educational qualifications‟.100 A complementary suggestion could be to focus on social exclusion, a concept which has come to replace poverty as the defining criterion for state intervention. The focus would then be on those who are systematically excluded from participation in society. This means systematic exclusion from participation in economically or socially valuable activities, from political engagement and from social interaction as well as lacking the capacity to purchase goods and services.101 Social exclusion in this form arises from a variety of causes, which are partly material, but also relate to other issues, such as living in a deprived area, suffering partnership breakdown, being a elderly or disabled or a member of a historically disadvantaged group. Social exclusion is not just a temporary phase of poverty; it is systemic, often passed from generation to generation and can be self-perpetuating. The result of including socio-economic status defined as social exclusion would be that less favourable treatment on grounds of indicators of social exclusion, such as living in an informal settlements. Similarly, if a training scheme operated in such a way as to disproportionately exclude those who are socially excluded, for example, by being located a distance away from a council 99 Section 34(1). Section 1(xxvi). 101 T Burchardt, J Le Grand & D Piachaud „Degrees of Exclusion‟ in J Hills, J Le Grand & D Piachaud (eds) Understanding Social Exclusion (2001) 30–32. 100 estate, or requiring participants to have had relevant job experience, a claim of indirect discrimination could be pursued unless it was demonstrated that the requirement was necessary. However, it is unlikely that incorporating a status ground defined according to social exclusion would go far enough to address the dichotomy between the two spheres. It simply tacks an expressly redistributive claim onto a framework geared for recognition claims, which in turn is based on individualised, retrospective and monetary remedies. Status-based claims as presently constituted, as argued above, cannot command the resources necessary genuinely to address social exclusion. It is only as a way of challenging stigmatic sequels of social exclusion that it might have some marginal effect. Conclusion As the concept of substantive equality receives growing acknowledgement, both internationally and in domestic jurisdictions, more attention needs to be given to its specific meaning and impact. This paper has attempted to frame some of the complex questions which substantive equality raises. These primarily concern the relationship between status and socio-economic disadvantage in relation to three major issues: dignity, affirmative action and poverty. Cutting across these issues concerns the question of the appropriate role of the judiciary. The key to the their resolution lies in achieving a synthesis between both dimensions, rather than regarding one as trumping the other. But to achieve this, requires a greater sensitivity to structural inequality is required. At the same time, more attention is needed to the standard of judicial review. Courts should not cut across government measures to further substantive equality; indeed, courts should support such measures. However, courts should not regard this as a reason for deference. Respect for the democratic process requires greater, not less, attention to the duty to account and explain. The judicial role in applying substantive equality remains a supervisory one, to guard against stereotypical assumptions and unwarranted generalisations which can cause or perpetuate disadvantage. View publication stats