SAFLII [Home] [Databases] [WorldLII] [Search] [Feedback]

South Africa: Constitutional Court

You are here:  SAFLII >> Databases >> South Africa: Constitutional Court >> 2004 >> [2004] ZACC 3

[Database Search] [Name Search] [Recent Decisions] [Noteup] [Help]


Minister of Finance and Other v F J van Heerden (CCT 63/03) [2004] ZACC 3 (29 July 2004)

PDF of original document.PDF of original document

.RTF of original document

Media summary in .PDF format Links to media summary
Media summary in .DOC format




CONSTITUTIONAL COURT OF SOUTH AFRICA



Case CCT 63/03

MINISTER OF FINANCE First Applicant

THE POLITICAL OFFICE BEARERS PENSION FUND Second Applicant

versus

FREDERIK JACOBUS VAN HEERDEN Respondent


Heard on : 24 February 2004

Decided on : 29 July 2004


JUDGMENT




MOSENEKE J:


Introduction

[1]This case raises important constitutional issues of equality, restitutionary measures and unfair discrimination. These issues arise within the context of a challenge to the constitutionality of certain rules of the Political Office-Bearers Pension Fund (the Fund) that provided for differentiated employer contributions in respect of members of Parliament and other political office-bearers between 1994 and 1999.

[2]The constitutional attack is mounted on two grounds. The first is that the relevant rules of the Fund offend the equality provisions of the Constitution because they are unfairly discriminatory. The second ground is that, in any event, the Fund as a whole is a nullity because it was not validly established under section 190A of the interim Constitution[1] or section 219 of the Constitution.[2] The equality challenge is contested on the basis that the differentiation in the rules of the Fund is not unfairly discriminatory because it constitutes a “tightly circumscribed affirmative action measure” permissible under the equality provisions of our Constitution.

[3]The claimant is Mr Frederik Jacobus van Heerden (respondent). He served as a National Party member of the old Parliament from 1987 to 1994. With the advent of the new democratic Parliament in 1994, he was returned to office for the same political party as member of the National Assembly until April 1999. Like many parliamentarians whose term straddled the old and new Parliaments, he is a member of the Fund and of the Closed Pension Fund (CPF).[3] He purports to act also on behalf of 145 other similarly placed members of the Fund. Thring J, sitting in the Cape High Court (High Court), upheld the claim and declared the provisions of rule 4.2.1 of the Fund to be unconstitutional and invalid in Van Heerden v The Speaker of Parliament and Others (the High Court judgment).[4] The Minister of Finance, the first applicant, and the Fund, the second applicant, are aggrieved by this decision and seek leave of this Court to appeal against it.


Factual background

[4]From 1983 to 1994, the pension benefits of members of the tricameral Parliament and of other political office-bearers were regulated by statute.[5] In 1993, at the Kempton Park constitutional negotiations,[6] the ruling party of the time raised concern regarding the security of existing pensions of political office-bearers. There had been speculation by parliamentarians and other political office-bearers of the time that the new political regime may not continue to pay their pension benefits. The negotiating parties agreed that a pension fund exclusive to members of the old Parliament and other political office-bearers of the time would be established and fully funded to pay defined benefits to its members, whether they were re-elected or not as members of the first democratic Parliament of 1994.

[5]Pursuant to this agreement, legislation established the CPF.[7] It came into operation on 5 January 1994. As its name intimates, the CPF had several exclusionary features. Only members of Parliament and political office-bearers who held office before 1994 could become its members.[8] No new members could be admitted. It follows that persons who were elected to Parliament for the first time in the 1994 general elections were excluded. A further significant feature is that members of the old Parliament who on the 26 April 1994 had not served for a period of seven and a half years were entitled only to a gratuity.[9]

[6]Yet another distinguishing feature of the CPF is its financing provisions. The pension liability of the CPF to its beneficiaries was to be fully financed by public funds and not based on employer or employee contributions.[10] As a result, after January 1994, its members were not required to make any contributions to the CPF irrespective of whether they were returned to office or not in the 1994 general elections.

[7]Another relevant sequel to the negotiations at Kempton Park was the establishment of a Special Pension Fund to provide for people who had undergone sacrifices in order to bring about the new democratic order.[11]

[8]As the new democratic Parliament of April 1994 convened, it and its members had no pension arrangements. A new pension fund for the new Parliament had to be brought into being. This was in fact a constitutional obligation under section 190A of the interim Constitution.[12] Clearly, this constitutional obligation could not be achieved at the outset. The setting up of a new pension fund was a venture that would take time. As an interim measure, all concerned agreed that from 27 April 1994, the National Assembly and each of its members would contribute 12,5% and 7,5% of a member’s pensionable annual income respectively towards the pension fund to be formed.[13] Pending the creation of the envisaged pension fund, employer and member contributions were paid to the Public Investment Commission, subject to the accrued aggregate capital and interest thereon being refundable to the pension fund to be formed.

[9]For reasons not immediately apparent, four years elapsed before Parliament turned its attention to its own new pension scheme. In June 1998, Parliament supported recommendations on the formation of the new pension fund, with four political parties in Parliament dissenting.[14] On 3 August 1998 a parliamentary committee[15] tabled before the National Assembly a further report on the nature, benefits and management of the new pension fund.[16] The report included a proposal that pension contributions by employers for the period 27 April 1994 to 30 April 1999 should be paid retrospectively on a differentiated basis to new and continuing political office-bearers. On 13 August 1998, the report was debated and adopted by the National Assembly with only one party dissenting.[17] Towards the end of 1998, only a few months before the end of the first term of the new Parliament, the Fund was established but took effect retrospectively from 27 April 1994. The rules of the new fund were finalised and registered in terms of section 4(4) of the Pension Funds Act.[18] Predictably, the main object of the Fund was to provide for retirement, death and other benefits for serving and retired parliamentarians.[19]


The rules of the Fund

[10]The rules of the Fund create three categories of members. Rule 2 spells out the categories:

“‘Category A Member’ shall mean a Member who has been notified to the Trustees by the Employer as a Member who has not reached age 49 years and who is not a member of the Closed Pension Fund.
‘Category B Member’ shall mean a Member who has been notified to the Trustees by the Employer as a Member who has reached age 49 years and who is not a member of the Closed Pension Fund.
‘Category C Member’ shall mean a Member who is a member of the Closed Pension Fund.”


The rules require that each member must make a contribution to the Fund towards retirement benefits at a monthly and uniform rate of one-twelfth of 7,5% of his or her annual pensionable salary. However, the contributions payable by the various employers[20] within the Fund are calculated according to a differentiated scale. Rule 4.2.1 prescribes the variance in this manner:


“The Employer shall make contributions towards the retirement benefit of each Member in its Service at the rate of:
(a)in the case of a Category A Member, one twelfth of 17 per cent of his Pensionable Salary;
(b)in the case of a Category B Member

(i) for the period of 27 April 1994 to 30 April 1999, one twelfth of 20 per cent of his Pensionable Salary;

. . . .

(c)in the case of a Category C Member

(i) for the period of 27 April 1994 to 30 April 1999, one twelfth of 10 per cent of his Pensionable Salary”.

From 1 May 1999, the differentiation between the three categories fell away, and the contribution of employers became standardised for all members at a rate of one-twelfth of 17 percent of their annual pensionable salaries.

[11]The nub of the respondent’s unfair discrimination complaint is that over the designated five years the differentiated employer contributions scheme improperly disfavours him and other category C members who are in receipt of pensions from the CPF, in comparison with new parliamentarians who are either below or above 49 years of age and do not receive pension benefits from the CPF.


The High Court

[12]The High Court found that the challenged provisions are not mere “differentiation” but rather “discriminatory in nature” because for five years lower employer contributions were paid for the less favoured class of members of the Fund to which they all belonged and contributed equally, with the result that the less favoured class of members will receive substantially smaller pensions than will members of the favoured classes. It also found the differentiation to be “prima facie unfair” because first, it is arbitrary as no reason is advanced for it and secondly, it is based on intersecting grounds of race and political affiliation — the latter a matter of conscience and belief — all being prohibited grounds listed in section 9(3) of the Constitution.[21]

[13]The High Court reasoned that a person who relies on section 9(2) to justify discriminatory measures bears the “onus” of establishing on a balance of probabilities that the measures have been taken to promote the achievement of equality and that “generally speaking it cannot be an easy onus to discharge”. The discrimination, it held, has to be “convincingly justified” to discharge the presumption of unfairness under section 9(5).

[14]The High Court found that the Minister and the Fund had failed to discharge the “onus” that the impugned measures are justified under section 9(2) because the measures relied upon do not bear a rational connection to the end they purport to achieve. It held that there is no “causal nexus” between means and ends because it has not been shown that in order to benefit members of the favoured categories less must be given to the disfavoured category.

[15]The High Court took the view that even if the measures were assumed to be directed at promoting the achievement of equality they were unlikely to do so because on “various calculations . . . the alleged inequality between categories A and B, on the one hand and category C, on the other, subsisted.”[22] It also held that even accepting that the disfavoured members (category C) are now better off than the other categories (A and B) that would not cure the defect in the ameliorative measures. It concluded that the measures were not only “haphazard, random and overhasty” but also “arbitrary”. The High Court held the differentiated employer contributions to be unconstitutional and declared them invalid. However, no order was made regulating the consequences of the declaration of invalidity on the Fund and its members.

[16]The High Court dismissed the assertion by the Minister and the Fund that certain interested parties to the proceedings had not been joined and condoned the delay on the part of the claimant in instituting the claim. Since the claimant had succeeded on other grounds, the High Court found it unnecessary to decide the merits of the claim that the Fund had been improperly established in breach of the President’s constitutional obligations under section 190A of the interim Constitution.


Equality submissions

[17]Before us, the gravamen of the applicants’ complaint is that the High Court misconceived the true nature of the equality protection recognised by our Constitution, by resorting to a formal rather than a substantive notion of equality. They argued that the purpose of the differentiated scheme of employer benefits was to advance equality by identifying three separate indicators of need for increased pension benefits over a finite period. In that way the scheme rationally pursues a legitimate governmental purpose of distributing pensions on an equitable basis.

[18]The applicants urged us to have regard to the actual impact of the differentiated employer contribution scheme. Its effect on the respondent and members of his class is that they remain considerably privileged and better off in respect of their pension benefits than members of the favoured categories A and B. Moreover, it was submitted, the scheme is neither unfair nor invasive of the dignity of anyone. The complaint of the respondent and his class is not one that says the scheme invades their dignity but rather one propelled by financial benefit out of public funds and a desire to maintain historical privilege.

[19]In this Court, the argument advanced by the respondent had three components. He argued that ameliorative measures under section 9(2) of the Constitution, if based on any of the anti-discrimination grounds listed in section 9(3), constitute, in his words, “positive discrimination” and must be presumed unfair. The party implementing the measures must show them to be fair. The differentiation here, he argues, is informed by race because the scheme has a disproportionate impact on 143 white, coloured and Indian members of Parliament as against two black members.[23] He urged us to take the view that the applicants have failed to rebut the resultant presumption of unfairness of the discriminatory measures.

[20]A further contention of the respondent is that the scheme is unfair because the state does not allege that in order to benefit the favoured group it was essential that the disfavoured group should receive lower employee benefits. In his view, limited resources do not necessitate the scheme, as the state cannot credibly claim that it cannot afford to pay pension contributions for all members at the same level. After all from 1999 it did. In emphasising the point that the state is not out of pocket, the respondent draws attention to an announcement by the Minister on 12 November 2003, a date after the judgment of the High Court, that the national treasury plans to put aside as a budget item R400 million for additional service benefits for members of Parliament and of provincial legislatures.

[21]The respondent concedes the correctness of the comparative actuarial calculations, presented by the Minister and the Fund, which indicate that members of Parliament who are also members of the CPF are better off than those who are not despite the increased employer contribution. It is contended, however, that this is not so in all cases. The respondent points to 15 category C members who are saddled with membership of the CPF without the benefit of generous pensions.[24] He regards these cases as jammergevalle.[25] He argues that in testing the constitutional invalidity of the challenged scheme, an objective approach would require that the position of all members affected by the challenged measure should be considered. As a result, he submitted, the adverse impact of the scheme on the jammergevalle is sufficient to render the employer contributions scheme as a whole unfairly discriminatory.


Equality and unfair discrimination

[22]The achievement of equality goes to the bedrock of our constitutional architecture.[26] The Constitution commands us to strive for a society built on the democratic values of human dignity, the achievement of equality, the advancement of human rights and freedom.[27] Thus the achievement of equality is not only a guaranteed and justiciable right in our Bill of Rights but also a core and foundational value; a standard which must inform all law and against which all law must be tested for constitutional consonance.[28]

[23]For good reason, the achievement of equality preoccupies our constitutional thinking. When our Constitution took root a decade ago our society was deeply divided, vastly unequal and uncaring of human worth. Many of these stark social and economic disparities will persist for long to come. In effect the commitment of the Preamble is to restore and protect the equal worth of everyone; to heal the divisions of the past and to establish a caring and socially just society. In explicit terms, the Constitution commits our society to “improve the quality of life of all citizens and free the potential of each person”.[29]

[24]Our supreme law says more about equality than do comparable constitutions. Like other constitutions, it confers the right to equal protection and benefit of the law and the right to non-discrimination. But it also imposes a positive duty on all organs of state to protect and promote the achievement of equality[30] — a duty which binds the judiciary too.[31]

[25]Of course, democratic values and fundamental human rights espoused by our Constitution are foundational. But just as crucial is the commitment to strive for a society based on social justice.[32] In this way, our Constitution heralds not only equal protection of the law and non-discrimination but also the start of a credible and abiding process of reparation for past exclusion, dispossession, and indignity within the discipline of our constitutional framework.
[26]The jurisprudence of this Court makes plain that the proper reach of the equality right must be determined by reference to our history and the underlying values of the Constitution.[33] As we have seen a major constitutional object is the creation of a non-racial and non-sexist egalitarian society underpinned by human dignity, the rule of law, a democratic ethos and human rights.[34] From there emerges a conception of equality that goes beyond mere formal equality and mere non-discrimination which requires identical treatment, whatever the starting point or impact.[35] Of this Ngcobo J, concurring with a unanimous Court, in Bato Star Fishing (Pty) Ltd v The Minister of Environmental Affairs and Tourism and Others[36] observed that:

“In this fundamental way, our Constitution differs from other constitutions which assume that all are equal and in so doing simply entrench existing inequalities. Our Constitution recognises that decades of systematic racial discrimination entrenched by the apartheid legal order cannot be eliminated without positive action being taken to achieve that result. We are required to do more than that. The effects of discrimination may continue indefinitely unless there is a commitment to end it.”[37]

[27]This substantive notion of equality recognises that besides uneven race, class and gender attributes of our society, there are other levels and forms of social differentiation and systematic under-privilege, which still persist. The Constitution enjoins us to dismantle them and to prevent the creation of new patterns of disadvantage.[38] It is therefore incumbent on courts to scrutinise in each equality claim the situation of the complainants in society; their history and vulnerability; the history, nature and purpose of the discriminatory practice and whether it ameliorates or adds to group disadvantage in real life context,[39] in order to determine its fairness or otherwise in the light of the values of our Constitution. In the assessment of fairness or otherwise a flexible but “situation-sensitive”[40] approach is indispensable because of shifting patterns of hurtful discrimination and stereotypical response in our evolving democratic society. The unfair discrimination enquiry requires several stages. These are set out by this Court in Harksen v Lane NO and Others.[41]


Restitutionary measures

[28]A comprehensive understanding of the Constitution’s conception of equality requires a harmonious reading of the provisions of section 9. Section 9(1) proclaims that everyone is equal before the law and has the right to equal protection and benefit of the law. On the other hand, section 9(3) proscribes unfair discrimination by the state against anyone on any ground including those specified. Section 9(5) renders discrimination on one or more of the listed grounds unfair unless its fairness is established. However, section 9(2) provides for the achievement of full and equal enjoyment of all rights and freedoms and authorises legislative and other measures designed to protect or advance persons or categories of persons disadvantaged by unfair discrimination. Restitutionary measures, sometimes referred to as “affirmative action”, may be taken to promote the achievement of equality. The measures must be “designed” to protect or advance persons disadvantaged by unfair discrimination in order to advance the achievement of equality.

[29]Section 9(1) provides: “Everyone is equal before the law and has the right to equal protection and benefit of the law.” Of course, the phrase “equal protection of the laws” also appears in the 14th Amendment of the US Constitution. The American jurisprudence has, generally speaking, rendered a particularly limited and formal account of the reach of the equal protection right.[42] The US anti-discrimination approach regards affirmative action measures as a suspect category which must pass strict judicial scrutiny. The test requires that it be demonstrated that differentiation on the grounds of race is a necessary means to the promotion of a compelling or overriding state interest. A rational relationship between the differentiation and a state interest would be inadequate.[43] Our equality jurisprudence differs substantively from the US approach to equality. Our respective histories, social context and constitutional design differ markedly. Even so, the terminology of “affirmative action” has found its way into general use and into a number of our statutes directed at prohibiting unfair discrimination and promoting equality, such as the Employment Equity Act 55 of 1998 and the Promotion of Equality and the Prevention of Unfair Discrimination Act 4 of 2000. But in our context, this terminology may create more conceptual and other difficulties than it resolves. We must therefore exercise great caution not to import, through this route, inapt foreign equality jurisprudence which may inflict on our nascent equality jurisprudence American notions of “suspect categories of state action” and of “strict scrutiny”. The Afrikaans equivalent “regstellende aksie” is perhaps juridically more consonant with the remedial or restitutionary component of our equality jurisprudence.

[30]Thus, our constitutional understanding of equality includes what Ackermann J in National Coalition for Gay and Lesbian Equality and Another v Minister of Justice and Another[44] calls “remedial or restitutionary equality”.[45] 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”[46] 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”.[47] A disjunctive or oppositional reading of the two subsections would frustrate the foundational equality objective of the Constitution and its broader social justice imperatives.

[31]Equality before the law protection in section 9(1) and measures to promote equality in section 9(2) are both necessary and mutually reinforcing but may sometimes serve distinguishable purposes, which I need not discuss now. However, 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. Absent a positive commitment progressively to eradicate socially constructed barriers to equality and to root out systematic or institutionalised under-privilege, the constitutional promise of equality before the law and its equal protection and benefit must, in the context of our country, ring hollow.

[32]The High Court favoured the approach that in effect, the measures under attack were not mere differentiation but discriminatory and that they must be convincingly justified because they are premised on grounds listed in section 9(3) and therefore attract an onus “that cannot be easy to discharge”. In Public Servants Association of South Africa and Others v Minister of Justice and Others,[48] Swart J, in dealing with the “affirmative action” claim of the government in that case, adopted an equivalent route in the interpretation and application of section 8(3)(a) of the interim Constitution. I am unable to agree with that approach. Legislative and other measures that properly fall within the requirements of section 9(2) are not presumptively unfair. Remedial measures are not a derogation from, but a substantive and composite part of, the equality protection envisaged by the provisions of section 9 and of the Constitution as a whole. Their primary object is to promote the achievement of equality. To that end, differentiation aimed at protecting or advancing persons disadvantaged by unfair discrimination is warranted provided the measures are shown to conform to the internal test set by section 9(2).


Onus of proof and section 9(2)

[33]It seems to me plain that if restitutionary measures, even based on any of the grounds of discrimination listed in section 9(3), pass muster under section 9(2), they cannot be presumed to be unfairly discriminatory.[49] To hold otherwise would mean that the scheme of section 9 is internally inconsistent or that the provisions of section 9(2) are a mere interpretative aid or even surplusage.[50] I cannot accept that our Constitution at once authorises measures aimed at redress of past inequality and disadvantage but also labels them as presumptively unfair. Such an approach, at the outset, tags section 9(2) measures as a suspect category that may be permissible only if shown not to discriminate unfairly. Secondly, such presumptive unfairness would unduly require the judiciary to second guess the legislature and the executive concerning the appropriate measures to overcome the effect of unfair discrimination.

[34]Following the reasoning in Public Servants Association,[51] the High Court made much of the presumption of unfairness against the differentiated pension scheme and the burdensome onus it attracts. I have concluded that legislative and other measures, which properly fall within the provision of section 9(2), do not attract any such burden.

[35]It follows that the High Court is clearly mistaken in approaching this matter on the limited basis that it need not decide whether and the extent to which members of Parliament who were members of the CPF were better off than those who were not,[52] since the applicants had omitted to make certain averments, which the court regarded as essential to discharge the section 9(5) onus.


Requirements of section 9(2)

[36]The pivotal enquiry in this matter is not whether the Minister and the Fund discharged the presumption of unfairness under section 9(5), but whether the measure in issue passes muster under section 9(2). If a measure properly falls within the ambit of section 9(2) it does not constitute unfair discrimination. However, if the measure does not fall within section 9(2), and it constitutes discrimination on a prohibited ground, it will be necessary to resort to the Harksen test in order to ascertain whether the measures offend the anti-discrimination prohibition in section 9(3).

[37]When a measure is challenged as violating the equality provision, its defender may meet the claim by showing that the measure is contemplated by section 9(2) in that it promotes the achievement of equality and is designed to protect and advance persons disadvantaged by unfair discrimination. It seems to me that to determine whether a measure falls within section 9(2) the enquiry is threefold. The first yardstick relates to whether the measure targets persons or categories of persons who have been disadvantaged by unfair discrimination; the second is whether the measure is designed to protect or advance such persons or categories of persons; and the third requirement is whether the measure promotes the achievement of equality.

[38]The first question is whether the programme of redress is designed to protect and advance a disadvantaged class. The measures of redress chosen must favour a group or category designated in section 9(2). The beneficiaries must be shown to be disadvantaged by unfair discrimination. In the present matter, the Minister and the Fund submitted that the differentiated contribution scheme was set up to promote the attainment of equality between members of the CPF and new members who were in the past excluded on account of race and or political affiliation. This objective they would advance by identifying three separate indicators of need for increased pensions for new parliamentarians. On the facts, however, it is clear that not all new parliamentarians of 1994 belong to the class of persons prejudiced by past disadvantage and unfair exclusion. An overwhelming majority of the new members of Parliament were excluded from parliamentary participation by past apartheid laws on account of race, political affiliation or belief. [53]

[39]The starting point of equality analysis is almost always a comparison between affected classes. However, often it is 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. In this regard I am in respectful agreement, with the following observation of Gonthier J, in Thibaudeau v Canada: [54]

“The fact that it may create a disadvantage in certain exceptional cases while benefiting a legitimate group as a whole does not justify the conclusion that it is prejudicial.”[55]

[40]In the context of a section 9(2) measure, the legal efficacy of the remedial scheme should be judged by whether an overwhelming majority of members of the favoured class are persons designated as disadvantaged by unfair exclusion. It is clear that the existence of exceptional cases or of the tiny minority of members of Parliament who were not unfairly discriminated against under the apartheid regime, but who benefited from the differential pension contribution scheme, does not affect the validity of the remedial measures concerned.

[41]The second question is whether the measure is “designed to protect or advance” those disadvantaged by unfair discrimination. In essence, the remedial measures are directed at an envisaged future outcome. The future is hard to predict. However, they must be reasonably capable of attaining the desired outcome. If the remedial measures are arbitrary, capricious or display naked preference they could hardly be said to be designed to achieve the constitutionally authorised end.[56] Moreover, if it is clear that they are not reasonably likely to achieve the end of advancing or benefiting the interests of those who have been disadvantaged by unfair discrimination, they would not constitute measures contemplated by section 9(2).

[42]In Public Servants Association,[57] Swart J, in interpreting section 8(3)(a) of the interim Constitution, held that:

“The measures must be designed to achieve something. This denotes . . . a causal connection between the designed measures and the objectives.”[58]


In the present matter Thring J followed this approach and held that no such causal nexus is present because the sponsor of the differentiated employer contribution scheme does not say that less had to be paid for the disfavoured category in order to give more to the favoured group. I cannot support this approach. Section 9(2) of the Constitution does 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. To require a sponsor of a remedial measure to establish a precise prediction of a future outcome is to set a standard not required by section 9(2). Such a test would render the remedial measure stillborn, and defeat the objective of section 9(2).

[43]It is untenable to require, as Thring J did, that a sponsor of remedial measures must show a necessity to disfavour one class in order to uplift another. The provisions of section 9(2) do not prescribe such a necessity test because remedial measures must be constructed to protect or advance a disadvantaged group. They are not predicated on a necessity or purpose to prejudice or penalise others, and so require supporters of the measure to establish that there is no less onerous way in which the remedial objective may be achieved. The prejudice that may arise is incidental to but certainly not the target of remedial legislative choice. On the facts of this case, the members of the disfavoured class, barring a few, are beneficiaries of a generous publicly funded pension scheme which pre-dates the differential measure. The favoured categories are, in the main, not. The disfavoured category was and, as the High Court observed, remains better situated than its new parliamentary counterparts as far as state-funded pension benefits go.

[44]The third and last requirement is that the measure “promotes the achievement of equality”. Determining whether a measure will in the long run promote the achievement of equality requires an appreciation of the effect of the measure in the context of our broader society. It must be accepted that the achievement of this goal may often come at a price for those who were previously advantaged. Action needs to be taken to advance the position of those who have suffered unfair discrimination in the past. As Ngcobo J observed in Bato Star:

“The measures that bring about transformation will inevitably affect some members of the society adversely, particularly those coming from the previously advantaged communities.”[59]

However, it is also clear that the long-term goal of our society is a non-racial, non-sexist society in which each person will be recognised and treated as a human being of equal worth and dignity. Central to this vision is the recognition that ours is a diverse society, comprised of people of different races, different language groups, different religions and both sexes. This diversity, and our equality as citizens within it, is something our Constitution celebrates and protects. In assessing therefore whether a measure will in the long-term promote equality, we must bear in mind this constitutional vision. In particular, a measure should not constitute an abuse of power or impose such substantial and undue harm on those excluded from its benefits that our long-term constitutional goal would be threatened.

Discussion

[45]At the threshold, the challenged pension contribution scheme differentiates among its members. The differentiation is based on several indicators. However, the discontent of the respondent is confined to the distinction made between state pension contributions in respect of pre- and post-1994 parliamentarians. In my view, we are obliged to look at the scheme as a whole. We must bear in mind its history of transition from the old to the new 1994 Parliament; the duration, nature and purpose of the scheme; the position of the complainant and the impact of the disfavour on the respondent and his class.

[46]The scheme has a finite lifespan of five years. It is a transitional, limited and temporary tool to allocate public resources. Its effect is retrospective. Nothing significant turns on that. The scheme was set up late in the life of the first democratic Parliament. Properly so, the pension benefits of all concerned were best protected by a retrospective date of commencement. Otherwise all members would have found themselves without pension benefits, although they had served Parliament for nearly four and a half years since April 1994.

[47]The scheme creates several classes of members in regard to employer pension contributions. The first class (category A) focuses on parliamentarians younger than 49 years of age, who are not members of the CPF. They receive employer contributions of 17% of their annual pensionable salary. Their colleagues older than 49 years (category B) who are not members of the CPF get a higher contribution of 20%. The third class (category C) receives pension benefits from the CPF and is allocated employer contributions of 10%. Lastly, the class of those over 49 years (category B) who left office in 1999 continue to receive a 5% employer contribution for five years after they left office. Those of the same class who remained in office receive no comparable benefit.

[48]It is clear to me that the differentiated scale of employer contributions was one decided and applied to ameliorate past disadvantage related to the pension benefits need of new political office-bearers, premised on three indicators. First, members who did not have access to the generous benefits of the CPF, as a class, had a greater need for pension benefits than the class of members who were already in receipt of these benefits. The inequality of pensions between the overwhelming majority of new parliamentarians and the vast majority old parliamentarians arises from past unjustified legislative and other exclusions of the former. That, in a large measure, explains the line drawn between new and old parliamentarians. Although the class of the new parliamentarians of 1994 is drawn predominantly from disadvantaged backgrounds, it is racially and gender diverse and drawn from different political parties.

[49]Within the class of new parliamentarians a sharper indicator of need is utilised. Members over the age of 49 years (category B) being, as a class, closer to retirement, had a greater need for increased pension benefits than members under that age (category A). The older class was accordingly given 20% employer contributions while the younger class received 17% contributions.

[50]Thirdly, within the class of category B, members who left office in 1999, as a class, had a greater need for increased pension benefits than those who remained in office because the latter would continue to accumulate benefits under the Fund. The class that left office in 1999 accordingly continued to receive a 5% employer contribution for five years after they left office. The class that remained in office received no comparable benefit because the latter would continue to accumulate benefits under the Fund.

[51]Within each class of members of the Fund, individual variations are to be expected. Conceivably some new members of Parliament who do not receive pensions from the CPF may have accumulated pension benefits before joining Parliament. Conversely, some old parliamentarians may not receive pensions as generous as most members of the CPF. Comparable individual variations may be found amongst younger members who leave office early or older members who are elected to office several times. In my view, none of these possible exceptions to the three membership categories diminishes the efficacy of the indicators as general guides to the payment of the relative increased pension benefits.

[52]I am satisfied that the evidence demonstrates a clear connection between the membership differentiation the scheme makes and the relative need of each class for increased pension benefits. The scheme was designed to distribute pension benefits on an equitable basis with the purpose of diminishing the inequality between privileged and disadvantaged parliamentarians. In that sense the scheme promotes the achievement of equality. It reflects a clear and rational consideration of the need of the members of the Fund and serves the purpose of advancing persons disadvantaged by unfair discrimination.

[53]The high watermark of the respondent’s complaint is that the impact of this differentiation on him and others in his position is that he will earn from the Fund less pension than otherwise. That is so.[60] The argument the respondent did not advance is that, as a class, new parliamentarians who are members of the Fund earn an average annual pension higher than that earned by him and his class of parliamentarians who are also members of the CPF. He cannot credibly advance that assertion. The actuarial evidence tendered by the Minister and the Fund[61] demonstrates that the applicant and his class remain a privileged class of public pension beneficiaries notwithstanding the challenged remedial measures. Their pensions are indeed generous and several times more generous than they would, on their pensionable annual salaries, have been entitled to under comparable public sector pension funds. Moreover, they are considerably more generous than pensions payable out of the Special Pension Fund to people who had undergone sacrifices in order to bring about the new democratic order.[62]

[54]The respondent does not claim that he and his class of parliamentarians are in any sense vulnerable or marginalized or that in the past they were unfairly excluded or discriminated against. Nor do I think that he and his class were. He does not complain that the scheme is invasive of his dignity or of any of the members of the CPF. There is no evidence to suggest any indignity. His claim appears to be propelled by a desire to earn more in circumstances where his pensions benefit is well ahead of that of his newer colleagues in parliament, despite the remedial measures challenged.


Jammergevalle

[55]Jammergevalle is an appellation that both counsel used to describe the class of some 15 members[63] of the Fund who were also members of the CPF, but did not receive the generous payments that accrued to the overwhelming majority of parliamentarians who are members of the CPF. Their terminal benefits were calculated in accordance with a formula under section 11 of the Members of Parliament and Political Office-Bearers Pension Scheme Act.[64] Ordinary members of the old Parliament who had rendered less than seven and a half years service at April 1994 were entitled to no more than a gratuity. Other office-bearers with less than five years of service at April 1994 also fall into the category. It is thus clear that within category C not all members receive generous benefits from the CPF. Their relatively limited terms of office before the advent of the new Parliament earned them only a lump gratuity payment in the CPF. Nonetheless, under the differentiated scheme of the Fund, they fall within the disfavoured category C membership.

[56]The question is whether the adverse impact of the employer contribution scheme on jammergevalle is such as to render it unfairly discriminatory. One must, however, keep in mind that they are a notional sub-class comprising approximately 10% of the total class of 146 category C members of the Fund, even on their argument. In many respects they do not, in terms of state funded pension benefits, share the financial attributes of 90% of the respondent and class he seeks to represent. Put differently, jammergevalle are unrepresentative of the class complaining of unfair discrimination and are therefore not an appropriate comparator. The comparison to be made must be with the overwhelming majority of the class asserting the equality claim. I am satisfied that the circumstances of this sub-class of category C members do not invalidate the legal efficacy of the scheme of the Fund.


Conclusion

[57]I have come to the conclusion that it is in the interests of justice to grant this application for leave to appeal from the decision of the High Court. The order of the High Court declaring rule 4.2.1 of the Fund unconstitutional and invalid cannot be supported. The appeal has merit and must be upheld.


Section 190A of the interim Constitution

[58]In his fourth set of affidavits before the High Court, the respondent raised a new cause of action that the Fund as a whole is invalid as it was not properly established under section 190A or its employer contributions are not set at a rate determined by the President.

[59]The High Court declined to decide this cause of action because it had disposed of the matter on the basis of unfair discrimination. In its further judgment on the application for leave to appeal and for a certificate in terms of the old Rule 18 of this Court,[65] the High Court took the view that the section 190A contention did not form part of its reasons for judgment and thus cannot be the subject of any appeal. I respectfully agree that “an argument in support of an appeal on this ground would be virtually impossible to formulate in logic.” However, in this Court the respondent persisted in this argument.

[60]The crux of the respondent’s contention is that the Fund has no legal effect because it was not established in terms of section 190A or, if it was, the levels of employer contributions to the Fund were set by cabinet resolution and not by the President as required by section 190A(5). In response, the Minister and the Fund disavowed any reliance on section 190A for the establishment of the Fund. They argue that mere reference to section 190A in the affidavit of one of their deponents[66] does not convey that the Fund was created under that constitutional provision.

[61]It appears to me plain that the Fund could not be established under the provisions of section 190A. The Fund came into force on 23 September 1998.[67] Section 190A was repealed on 4 February 1997, the day the final Constitution took effect.[68] Accordingly, the question whether the level of employer contributions of the impugned rules of the Fund was set by the cabinet rather than the President in accordance with the requirements of section 190A(5) does not arise. The legal power to set up a pension fund could not possibly arise from a repealed and therefore lifeless constitutional provision, irrespective of the mistaken views or preferences of those concerned. The impugned pension scheme could not be set up pursuant to the repealed provisions of section 190A of the interim Constitution.


Section 219 of the Constitution and section 8 of the Remuneration Act

[62]Before us the respondent advanced a new reason why the Fund as a whole should be invalidated. He argues that levels of challenged employer contributions were determined by the cabinet and not in compliance with section 219[69] of the Constitution and section 8[70] of the Remuneration of Public Office-Bearers Act (the Remuneration Act).[71]
[63]It is so that the Constitution does not contain a direct equivalent of section 190A. Unlike section 190A of the interim Constitution, section 219 of the Constitution enjoins Parliament to create a legislative framework for determining salaries, allowances and benefits of members of the National Assembly and other persons holding public office. Such legislation is the Remuneration Act. Section 8(2) of the Remuneration Act requires the Minister of Finance to determine the amount of the contributions to be made to the pension fund by the national government, after taking into consideration the recommendations of the Independent Commission for the Remuneration of Public Office-Bearers (Remuneration Commission), established in terms of the Independent Commission for the Remuneration of Public Office-Bearers Act (Commission Act),[72] which legislation came into operation on 29 June 1998.

[64]The respondent raised the contention on section 219 for the first time before us on appeal. None of the facts relevant to the issue are traversed in the affidavits lodged in the High Court. Neither the Minister nor the Fund had a proper opportunity to deal with the matter in written argument or during the hearing before us. The matter is not an issue in the appeal. There is no application for direct access. The matter is not properly before us and must be dismissed without deciding its merits.

Unreasonable delay and non-joinder

[65]The applicants urged us to uphold the appeal on the ground that the High Court ought to have disallowed the respondent’s claim because there was unreasonable delay in bringing it before court. They also raised issues of non-joinder, contending that provincial legislatures and the President should have been joined as parties. In the light of the decision I have come to on the merits of the case, it is unnecessary to decide these issues.


Costs

[66]The applicants sought an order for costs against the respondent. The respondent has raised issues of broad public concern and constitutional importance. In circumstances such as these, this Court seldom makes a cost order.[73] I am, therefore, not minded to grant an adverse cost order against the respondent.


Order
The following order is made:

(a)The application for leave to appeal against the judgment and order of the High Court made on 12 June 2003 is granted.
(b)The appeal is upheld.
(c)The order of the High Court of 12 June 2003 declaring the provisions of rule 4.2.1 of the Political Office-Bearers Pension Fund to be unconstitutional and invalid including its order as to costs is set aside.
(d)No order as to costs of the present application is made.



Chaskalson CJ, Langa DCJ, Madala J, O’Regan J, Sachs J, Van der Westhuizen J and Yacoob J concur in the judgment of Moseneke J.


MOKGORO J:


Introduction

[67]I have read the judgment prepared by my colleague Moseneke J. I agree with the order that he proposes as well as his findings in relation to section 190A of the interim Constitution and section 219 of the Constitution. I also agree with his conclusion that the impugned measure does not violate section 9 of the Constitution,[73]

but am unable to agree with the route taken to arrive at this conclusion. Whereas Moseneke J concludes that section 9(2) of the Constitution applies to this case, I am of the view that the facts of this case are to be decided in terms of section 9(3) of the Constitution.

[68]The facts of this case have been clearly set out in the main judgment.[2] The High Court upheld the claim of Mr Van Heerden (the respondent) that the differentiation between the amounts of contributions of members of the Political Office-Bearers Fund (POBF) amounted to unfair discrimination on the basis of race and political affiliation. As the main judgment makes clear, the High Court held that the state bears an onus to show that the impugned measures fall under section 9(2) of the Constitution. The High Court held that the applicant (the Minister) had failed to discharge this onus. One of the main reasons for this finding was that the Minister had failed to show that it was necessary to require that the various employers identified under the Fund’s rules contribute less to the pensions of category C members in order for category A and B members to receive greater contributions. In other words, the High Court held that there was no causal connection between the benefit to the new members of Parliament and the disadvantage to the old members. Moseneke J has dealt sufficiently with the issue that the High Court’s approach to section 9(2) and the nature of the burden to be discharged were incorrect. I agree with the analysis and conclusions of Moseneke J in this regard.


Arguments in this Court

[69]The Minister argues that the High Court adopted an approach to the interpretation of section 9 which is based on the notion of formal equality. In contradistinction to the notion of formal equality it was argued that the Constitution embraces substantive equality which permits remedial measures to be enacted to address past unfair discrimination. According to the Minister, the differentiation that occurs under the rules of the POBF constitutes a positive measure to create equity amongst parliamentarians in respect of the pensions that they will ultimately draw. This measure, it was argued, is of the type envisaged by section 9(2) of the Constitution because it aims to advance persons previously disadvantaged by unfair discrimination and in so doing promotes the achievement of equality.

[70]Before this Court, the respondent persists with his argument that the measure unfairly discriminates on the basis of race and political affiliation. The respondent argues that any measure, which is restitutionary in nature but discriminates against persons on one of the listed grounds, attracts the presumption of unfairness in terms of section 9(5). The state must show the measures to be fair in order successfully to resist his equality challenge. The respondent further argues that the state has failed to show that it cannot afford to pay all parliamentarians equally and, as such, has failed to show the necessity of discriminating in the way that the POBF does. As a consequence, so the argument goes, the measure is unfair and unconstitutional.


Equality and unfair discrimination

[71]The role of the right to equality in our new dispensation cannot be overstated. Apartheid was not merely a system that entrenched political power and socio-economic privilege in the hands of a minority nor did it only deprive the majority of the right to self actualisation and to control their own destinies. It targeted them for oppression and suppression. Not only did apartheid degrade its victims, it also systematically dehumanised them, striking at the core of their human dignity. The disparate impact of the system is today still deeply entrenched.

[72]It was with this in mind that the interim Constitution recognised in its preamble the need to create a society “in which there is equality between men and women and people of all races so that all citizens shall be able to enjoy and exercise their fundamental rights and freedoms”. The Constitution now makes clear the fundamental importance of equality in our constitutional framework by establishing that one of the fundamental values upon which our society is founded is the “achievement of equality”.[3] As this Court held in Prinsloo v Van der Linde and Another:

“Our country has diverse communities with different historical experiences and living conditions. Until recently, very many areas of public and private life were invaded by systematic legal separateness coupled with legally enforced advantage and disadvantage. The impact of structured and vast inequality is still with us despite the arrival of the new constitutional order. It is the majority, and not the minority, which has suffered from this legal separateness and disadvantage.”[4]

[73]It is no mistake that our Constitution uses the phrase “achievement of equality”. The tremendous indignity and political oppression that characterised the years of apartheid was coupled with the systemic entrenchment of economic disadvantage for millions of South Africans. The vast majority of this country’s wealth remained then and remains still, as a consequence of the entrenched disadvantage, in the hands of a minority. Sprawling and over-crowded informal townships inhabited by poor and jobless people without property to call their own and without many of the basic amenities necessary for a dignified human existence sit beside most affluent neighbourhoods with people who have access to the best schools, the best jobs and the best opportunities. The use of the phrase “achievement of equality” therefore recognises that the creation of democracy and equal treatment before the law are not enough to foster substantive equality. Unless the disparity that currently exists is consciously and systematically obliterated, it can easily be overlooked and will as a result continue to define our society for a long time to come.

[74]In Brink v Kitshoff NO[5] this Court remarked that

“[a]s in other national constitutions, section 8 is the product of our own particular history. Perhaps more than any of the other provisions in chap 3, its interpretation must be based on the specific language of section 8, as well as our own constitutional context. Our history is of particular relevance to the concept of equality. The policy of apartheid, in law and in fact, systematically discriminated against black people in all aspects of social life. Black people were prevented from becoming owners of property or even residing in areas classified as ‘white’, which constituted nearly 90% of the landmass of South Africa; senior jobs and access to established schools and universities were denied to them; civic amenities, including transport systems, public parks, libraries and many shops were also closed to black people. Instead, separate and inferior facilities were provided. The deep scars of this appalling programme are still visible in our society. It is in the light of that history and the enduring legacy that it bequeathed that the equality clause needs to be interpreted.”[6]


Although these remarks refer to section 8 of the interim Constitution they are equally apposite today.

Restitutionary measures

[75]Moseneke J is indeed correct when he points out that the provisions of section 9 must be understood against the backdrop of the circumstances highlighted above and the need to foster substantive equality. Section 9(2) in particular was enacted with the idea that true equality can never be said to exist until the patterns of disparity which were created in the past have been eradicated. The measures it envisages therefore form an integral part of our overall conception of equality. When in 1994 democracy was established in South Africa the right to equality for all South Africans was constitutionally protected. However, section 9(2) acknowledges that our notion of substantive equality requires measures to be enacted to make up for that part of the past which cannot simply be corrected by removing the legal bars to equality of treatment.
[76]For this reason, I join Moseneke J in his criticism of the High Court’s approach to section 9(2). To require the state to demonstrate that it is necessary to give less to one group in order to advance another would be to undermine the scheme of section 9(2). The reason for the enactment of section 9(2) is to authorise restitutionary measures for the advancement of those previously disadvantaged by unfair discrimination. Whenever a group is given certain advantages it must follow that it receives more than others in the context of the particular measure which is being enacted. But the measure will not necessarily be enacted with the aim of taking from one group to give to another. The logical consequence of the respondent’s submissions is that practically no measure may be enacted of a restitutionary nature because each time the state attempts to do so it will, more often than not, fail to prove the necessity of giving more to one person or group than another. The approach of the High Court presupposes that it will only be permissible to favour a particular group if there are insufficient resources to give equally to everyone. The Minister is correct when he argues that the approach of the High Court is premised on a notion of formal equality which is at odds with the vision of substantive equality in our Constitution. It would be contrary to the spirit of section 9(2) and inimical to its purpose to require the state to show that it has insufficient resources to give advantages equally, every time that it attempts to enact a restitutionary measure which advances those previously disadvantaged.

[77]I further agree with the judgment of Moseneke J in its approach to the interaction between section 9(2) and section 9(3). The whole structure of our equality clause and the important aim of substantive equality would be undermined by an approach which requires the state to show that measures which aim at advancing the substantive notion of equality and fostering a society which no longer resembles that of the South Africa of old are fair. It is an invariable consequence of enacting measures that advance certain groups that other groups will be disadvantaged in that regard, albeit that this would not be the intention of such measures. More often than not, such disadvantage will be on the basis of one of the listed grounds in section 9(3). The logical consequence of the approach advanced by the respondent is that practically all restitutionary measures would attract a presumption of unfairness. This cannot be what section 9(2) envisages. An interpretation of the Constitution which renders certain provisions redundant should be avoided.

[78]I wish to make one further observation about the difference between section 9(2) and section 9(3). Section 9(2) is forward looking and measures enacted in terms of it ought to be assessed from the perspective of the goal intended to be advanced. The measures must promote the achievement of equality by advancing those previously disadvantaged in the manner envisaged. This is not to say that the interests of those not advanced by the measure must necessarily be disregarded. However, the main focus in section 9(2) is on the group advanced and the mechanism used to advance it.

[79]Our equality jurisprudence in terms of section 9(3) is, however, different. When assessing a measure under section 9(3), the focus is on the group or person discriminated against. Here, the impact on the complainant and his or her position in society is of utmost importance. The aim of the challenged measure and whether it advances a legitimate government purpose will of course be important. However, the main focus is on the complainant and the impact of the measure on him or her.

[80]This distinction is in my view important. It would frustrate the goal of section 9(2) if measures enacted in terms of it paid undue attention to those disadvantaged by the measure when that disadvantage is merely an invariable result and not the aim of the measure. The goal of transformation would be impeded if individual complainants who are aggrieved by restitutionary measures could argue that the measures unfairly discriminated against them because of their undue impact on them. As Ngcobo J said in Bato Star Fishing (Pty) Ltd v Minister of Environmental Affairs and Others:[7]

“There are profound difficulties that will be confronted in giving effect to the constitutional commitment of achieving equality. We must not underestimate them. The measures that bring about transformation will inevitably affect some members of the society adversely, particularly those coming from the previously advantaged communities. It may well be that other considerations may have to yield in favour of achieving the goal we fashioned for ourselves in the Constitution.”[8]

It is for this reason that the equality jurisprudence developed by this Court in the context of section 9(3) is unsuited to analysis under section 9(2). The test as established by cases such as Harksen v Lane NO and Others[9] and President of the Republic of South Africa and Another v Hugo[10] would focus unduly on the position of the complainant to be appropriate to a section 9(2) analysis.

[81]Because of this distinction it is important that a measure purportedly enacted under section 9(2) fits properly within it. If measures are incorrectly defended under it, insufficient weight will be given to the position of the complainant. Conversely, if the equality jurisprudence under section 9(3) is built into the test for section 9(2), the process of transformation, as envisaged by the Constitution, will be unduly hampered.


The correct approach to section 9(2)

[82]Given my view that section 9(2) measures ought not to be tested against section 9(3), it is clearly necessary to ascertain what requirements a section 9(2) measure must meet. I endorse the three aspects of the review standard identified by Moseneke J in the main judgment, namely that “[t]he first yardstick relates to whether the measure targets persons or categories of persons who have been disadvantaged by unfair discrimination; the second is whether the measure is designed to protect or advance such persons or categories of persons; and the third requirement is whether the measure promotes the achievement of equality.”[11]

[83]I further support the approach of the main judgment to the question of the purpose of the measure[12] and also the connection between the means employed and the end sought to be achieved.[13] I cannot, however, support the approach taken by Moseneke J to what he describes as the “first yardstick” – that the measure must be aimed at advancing persons or categories of persons previously disadvantaged by unfair discrimination.

[84]In a case such as the present, an applicant will approach the courts and claim that a particular measure unfairly discriminates against him or her. If, as a defence, the state successfully demonstrates that a measure falls within the ambit of section 9(2), the state in my view is relieved of the burden to show that the measure is fair, which it might otherwise have borne. Because a restitutionary measure which discriminates at all will almost certainly discriminate on the basis of one of the listed grounds, if it were not for section 9(2), a restitutionary measure would invariably attract a presumption of unfairness. Section 9(2) therefore relieves the state of having to show that the discrimination in question is fair. It is important that this should be so, for the reasons regarding transformation mentioned above. It would be inimical to the pursuit of substantive equality if the state was required to show that each restitutionary measure that it enacted was fair, as would be required by section 9(3).

[85]Another aspect of section 9(2) is that it allows a person or categories of people to be advanced. This is important because of the nature of the unfair discrimination that was perpetrated by apartheid. The approach of apartheid was to categorise people and attach consequences to those categories. No relevance was attached to the circumstances of individuals. Advantages or disadvantages were metered out according to one’s membership of a group. Recognising this, section 9(2) allows for measures to be enacted which target whole categories of persons. Therefore a person or groups of persons are advanced on the basis of membership of a group. The importance of this is that it is unnecessary for the state to show that each individual member of a group that was targeted by past unfair discrimination was in fact individually unfairly discriminated against when enacting a measure under section 9(2). It is sufficient for a person to be a member of a group previously targeted by the apartheid state for unfair discrimination in order to benefit from a provision enacted in terms of section 9(2).

[86]On this understanding of section 9(2), it is clear that various consequences attach to the state when invoking it. The state need not show that any discrimination caused by the measure is fair, or that each individual member of the advanced group actually suffered past disadvantages as long as an individual was part of a group targeted. Because section 9(2) relieves the state of these burdens, it is my view that care should be taken to ensure that measures enacted under it actually do fall within the ambit intended by the section. If the aim of the section is to advance persons or groups previously disadvantaged by unfair discrimination, the section should be used for that purpose alone. To do otherwise would be to allow the section to be used to enact measures which should not be tested under section 9(2) because they benefit persons who do not belong to groups previously disadvantaged by unfair discrimination.

[87]Section 9(2) is a unique constitutional provision which has been enacted to respond decisively to the particular history of inequality and the impact of that history on our society. It makes clear that restitutionary measures are part of the scheme for the realisation of substantive equality. A measure which is part of the framework for the advancement of equality cannot ever be said to discriminate unfairly. That being the case, once a measure can properly be said to satisfy the internal test in section 9(2) and fall within the ambit of the section, the scrutiny that other measures are subjected to in terms of section 9(3) does not apply. Once the state successfully demonstrates that a measure falls within section 9(2), that measure is constitutionally compliant without any further justification. That being the case, section 9(2) must be used only in appropriate cases and with great circumspection. The vision of substantive equality and the need for transformation cannot be underestimated. For that reason section 9(2), as an instrument for transformation and the creation of a truly equal society, is powerful and unapologetic. It would therefore be improper and unfortunate for section 9(2) to be used in circumstances for which it was not intended. If used in circumstances where a measure does not in fact advance those previously targeted for disadvantage, the effect will be to render constitutionally compliant a measure which has the potential to discriminate unfairly. This cannot be what section 9(2) envisages.
[88]The main judgment is compelling in its argument that a measure will fall under section 9(2) if the “overwhelming majority of members of the favoured class are persons designated as disadvantaged by unfair exclusion”.[14] By this reasoning, a measure will still fall under section 9(2) even if some of those who benefit from it were not members of groups targeted for unfair discrimination in the past. Moseneke J might be correct when he says that “the existence of exceptional cases or of the tiny minority of members of Parliament who were not unfairly discriminated against under the apartheid regime, but who benefited from the differential pension contribution scheme, does not affect the validity of the remedial measures concerned.”[15] However, it is not necessary for me to decide the correctness of this test for determining a category or group of persons under section 9(2) because it is my view that this case does not concern exceptional cases or tiny minorities.

[89]As already indicated, it is not necessarily the case, and I leave this question open, that every person in whose favour a restitutionary measure has been enacted must be shown to be a member of a group which has previously been disadvantaged by unfair discrimination. However, in the light of my remarks about the importance of section 9(2) and the burden that it removes from the state, it is my view that a measure must be more carefully crafted in relation to the group targeted for advancement than the present one, to fall under section 9(2).

[90]In this case, there are two possible grounds of previous disadvantage which this measure might have been enacted to redress: race and political affiliation. The notion of discrimination on the basis of political affiliation is complex. It might seem on the surface as if the majority of Members of Parliament (MPs) who joined for the first time in 1994 were previously unable to be members, on the basis of their political affiliation or belief. However, it is my view that this is not a conclusion which is supported by the evidence. Nor, in my view, is it capable of being supported by such evidence. The majority of parliamentarians who joined the legislatures in 1994 were black. Black people were prohibited by law on account of their race from standing for national election in South Africa and were only permitted to do so in the so-called independent homelands. They therefore did not have any choice whatsoever about standing for national elections.

[91]Unlike some of the white MPs who stood for election for the first time in 1994, who might claim that even though as whites they were entitled to vote under the old dispensation their disgust with the system prevented them from participating, black persons, even if they had no objection to running for Parliament, by law had no choice. It is artificial to say, therefore, that such black persons were previously disadvantaged on the basis of political affiliation. If black people had not been prohibited from voting or standing for national office prior to 1994, it might be that some of those MPs elected for the first time in 1994 might have run for office previously, as some did in the homelands. Even if this is unlikely it is impossible to speculate about decisions people might have taken when in reality they had no choice. The prohibition on standing for office based on race was so integral to the system of apartheid that it is hard to speculate what might have been, had black people been allowed to vote at national level.

[92]The political affiliation argument is based on the premise that a person was excluded on the basis of his or her belief. It was introduced in this case to explain why white people could justifiably benefit from the restitutionary measure in the POBF. The idea is that the white beneficiaries, who ran for office for former liberation movements in 1994, could not participate in Parliament prior to 1994 not because of a legal bar but because of an ideological distaste for the system. Unlike their black counterparts, they were not discriminated against under the law. Their non-participation in the parliamentary process arose from a choice that they made – a choice which must be acknowledged for its significance in the fight for democracy and the very equality envisaged by section 9(2). The significance of this choice should not be underemphasised. Be that as it may, since it has not been ascertained or shown whether this is the case for the vast majority of MPs elected for the first time in 1994, it is my view that the POBF cannot be seen as a restitutionary measure aimed at redressing previous discrimination on the basis of political affiliation.

[93]I turn to the question of race as a basis for advancement. On the evidence before this Court, it seems that there were 251 members elected to the National Legislature for the first time in 1994. There were therefore 251 people who benefited from the higher pension rate provided for in the POBF. Of these, 53 were white. This means that 79 per cent of the beneficiaries of the higher rate were black. However, within this group were also black people who were not excluded from membership of Parliament during apartheid. Because of the system’s approach to race classification, they were permitted to be part of the tri-cameral Parliament. Therefore, although it seems that 79 per cent of the new members were previously excluded on the basis of their race, the figure may be significantly less. In my view, and I limit my remarks to the facts of this case, the evidence does not show that the advanced group are in the overwhelming majority designated in terms of race. It has also not been shown that this is the case in terms of political affiliation. In my view, unless a measure is shown to stand the internal test in section 9(2), it does not qualify as a section 9(2) measure.


The scheme of the equality clause

[94]In the present matter, the respondent approached the High Court claiming that the differentiation in terms of the POBF unfairly discriminated against him on the basis of race and political affiliation. In response, the Minister did not contend that the measure constitutes fair discrimination or mere differentiation. Rather, the Minister raised section 9(2) as a defence and argued that the measure was restitutionary in nature. This raises the question whether the Minister must stand or fall by his submissions. It also raises the question whether the fact that the Minister relied on section 9(2) precludes a court from finding that the measure, although not restitutionary in the terms of section 9(2), is nevertheless fair having subjected it to equality analysis in terms of section 9(3).

[95]The main judgment has made it clear that section 9(2) is part of a unified view of the right to equality in section 9. I support that view. A measure enacted in terms of section 9(2) is not an exception to our notion of equality; it is an integral part of it. From this must follow that section 9 must be viewed as a whole and any matter which engages the issue of equality engages the whole section. This is not to say that all five subsections will always be relevant to every enquiry. Certain forms of discrimination might be so irrational that they do not even survive challenge in any of the terms of section 9. Other forms might attract a presumption of unfairness which can be rebutted. A measure might fall squarely under section 9(2) in which case it will not attract a presumption of unfairness and will not need to be tested in terms of section 9(3). What is important is to avoid a notion of equality which divides section 9 into artificial parts.

[96]In the present matter, the Minister has relied on facts in support of his contention that the measure falls under section 9(2). These facts in my view also support a finding that the discrimination in this case is fair. As I have found above, the measure does not meet the requirements of section 9(2). However, as I make clear below, it is my view that the measure does not constitute unfair discrimination. Many of the factors that Moseneke J advances in his judgment in support of his contention that the differentiation in the POBF is an acceptable restitutionary measure are, in my view, relevant to the fairness of the measure.[16] Given that section 9 must be viewed as a whole and given that the Minister relies on facts which demonstrate that the measure is fair, it would not be logical to hold that the appeal on the basis of section 9(2) must either be upheld or dismissed altogether.


Section 9(3)

[97]Nothing in section 9(2) limits the circumstances in which the state can enact measures to advance a purpose other than to remedy disadvantage caused by past unfair discrimination. Such measures will then need to be tested in terms of section 9(3). It is important to observe that a measure might resemble a restitutionary measure because it is aimed at creating equity between groups of persons but falls short of protection in terms of section 9(2). This would be the case when any of the three requirements identified by Moseneke J are not fulfilled. In view of the approach I take of the group targeted for disadvantage in the past, the inclusion of those not so targeted affects the group in a way that disqualifies it for advancement under a section 9(2) remedial measure. Such a measure may, generally on the basis of justification in terms of section 9(3) and particularly in view of the objective of the measure, pass muster. The evidence for advancement of the group or for justification may be the same or it may be different, depending on the circumstances of each case. It would be untenable to strike the measure down only because it does not fall under section 9(2) when it could be decided under section 9(3). Doing so would frustrate any programme designed for the achievement of equity.

[98]The measure created by rule 4.2.1 is most certainly aimed at the achievement of equity. However it falls short of all the requirements of section 9(2) in that it fails to target a group previously disadvantaged by unfair discrimination. In my view it must thus be tested against section 9(3) of the Constitution.

[99]In Harksen v Lane NO[17] the Court considered the following to be relevant to whether discrimination is unfair:
“(a) [T]he position of the complainants in society and whether they have suffered in the past from patterns of disadvantage, whether the discrimination in the case under consideration is on a specified ground or not;
(b) the nature of the provision or power and the purpose sought to be achieved by it. If its purpose is manifestly not directed, in the first instance, at impairing the complainants in the manner indicated above, but is aimed at achieving a worthy and important societal goal, such as, for example, the furthering of equality for all, this purpose may, depending on the facts of the particular case, have a significant bearing on the question whether complainants have in fact suffered the impairment in question. In Hugo, for example, the purpose of the Presidential Act was to benefit three groups of prisoners, namely, disabled prisoners, young people and mothers of young children, as an act of mercy. The fact that all these groups were regarded as being particularly vulnerable in our society, and that in the case of the disabled and the young mothers, they belonged to groups who had been victims of discrimination in the past, weighed with the Court in concluding that the discrimination was not unfair;
(c) with due regard to (a) and (b) above, and any other relevant factors, the extent to which the discrimination has affected the rights or interests of complainants and whether it has led to an impairment of their fundamental human dignity or constitutes an impairment of a comparably serious nature.”


Various factors are therefore relevant to an analysis of unfair discrimination. Of importance is the position of the complainants in society and whether they have suffered in the past from patterns of disadvantage. So too, whether the discrimination in the case under consideration is on a specified ground. The nature of the provision or power and the purpose sought to be achieved by it is also important. The question to be asked is whether the provision is aimed at an important societal goal. Unlike under section 9(2), other factors to emphasise include the extent to which the discrimination has affected the rights or interests of the complainants and whether the discrimination is of a serious nature and impairs the fundamental dignity of the complainants.

[100]It is my view that there is clearly discrimination on the facts of this case, but that such discrimination is not on a listed ground. The discrimination is between those members who served in Parliament prior to 1994 and those who did not. However, it is possible to assume in favour of the respondent that the discrimination in question is based on one of the listed grounds, either race or political affiliation.

[101]Assuming in favour of the respondent that the discrimination is based on race or political affiliation attracts the presumption that the measure unfairly discriminates. Even so, I am of the view that the measure is fair. The main judgment points out that the actuarial evidence before this Court shows that the respondent and the majority of his group “remain a privileged class of public pension beneficiaries notwithstanding the challenged remedial measures.”[18] This suggests that the consequences of the measure do not impact unduly on the interests of the respondent. In fact, the respondent concedes that the majority of members of Parliament who are members of the Closed Pension Fund (CPF) are better off than those who benefit from the increased contributions in terms of the POBF. He argues, however, that there are sufficient exceptions – referring to the so-called ‘jammergevalle’[19] – to conclude that the measure is unfair. I return to the case of the ‘jammergevalle’ below. In so far as the respondent has correctly conceded that the majority of members of the CPF are in fact better off than their colleagues who joined Parliament for the first time in 1994 as a result of their membership of the CPF, it cannot be said that they are victims of unfair discrimination. It cannot be said that a measure which creates no disadvantages unfairly discriminates unless it attracts one of the other characteristics which this Court has held in previous equality cases to constitute a violation of section 9, such as a negative impact on the complainant’s dignity. Moseneke J correctly points out that the measures do not impact negatively on the dignity of the complainants.[20] The scheme does not have an impact on their dignity, because it does not negatively impact on the complainants’ sense of self worth. Furthermore, the respondent conceded in argument that the only loss suffered was pecuniary in nature. His motivation for contesting the measure was indeed to earn more.

[102]Another factor of importance is whether the measure advances an important societal goal or whether it is aimed at impairing the complainant. It is clear that the current measure advances an important societal goal. It is aimed at creating equity between new MPs and those members of the current Parliament who, because of the fact that they were also members of the tri-cameral Parliament, are members of the CPF. It cannot be contested that a person who was not unfairly excluded in the past could have chosen to run for a right-wing party for the first time in 1994 and still benefit from the POBF. The scheme was instituted to benefit all newcomers, rather than those excluded on the basis of their race or political affiliation, because it was seen as a government concern that new MPs did not have a substantial pension to fall back on in their retirement. That is a legitimate objective in terms of section 9(3). The scheme in no way targets the complainants, nor does it seek to impair them.


Jammergevalle

[103]While acknowledging that the majority of parliamentarians who receive lower contributions in terms of the POBF are actually still better off because of their membership of the CPF, the respondent argues that the existence of the ‘jammergevalle’ – those members of the POBF who are worse off even though they are members of the CPF because of when they joined Parliament – is sufficient to render the scheme unfair.

[104]There is a dispute between the parties as to the number of ‘jammergevalle’. According to the respondent, 15 people fall into this category. According to the Minister, however, only 13 people are in fact worse off – a difference which is rather insignificant. Regardless of whether 15 or 13 people are affected, I am of the view that the measure is fair. As the main judgment holds, in any legislative scheme which differentiates between classes, there will be hard cases. These hard cases should not

prevent a court from concluding that a measure is not unfair and is therefore constitutionally compliant.

[105]I have cautioned that, in the context of section 9(2), great care must be taken to define the group because of the nature of the subsection and the advantage of not having to justify the measure on the part of the author of the remedial measure in invoking it. It is my view that the facts of this case are such that the measure is not one envisaged by section 9(2). The basis for this conclusion is that a significant number of the beneficiaries are not members of a category previously disadvantaged by unfair discrimination. There is a significant difference between a finding that a measure must be tightly crafted to fall under section 9(2) because of its specific requirements and the consequences which attach to that section and a finding that the existence of exceptional circumstances does not render an otherwise fair measure unfair. My conclusion in this regard is not at odds with my conclusion that the ‘jammergevalle’ do not constitute an obstacle to finding the present measure to be fair. The conclusion in respect of section 9(2) is based on the narrow purpose for which it was designed and its special place in our equality jurisprudence in view of the history of inequality in our society. The conclusion in respect of the ‘jammergevalle’ is based on an acknowledgment that, under our section 9(3) jurisprudence, to allow hard cases to undermine otherwise constitutionally compliant schemes would place a burden on government that would unduly impede its ability to transform our society.


Conclusion

[106]A consideration of the factors mentioned above leads me to the conclusion that the measure in this case does not unfairly discriminate against the respondent. I would therefore agree with the order proposed by Moseneke J and also uphold the appeal.




Sachs J and Skweyiya J concur in the judgment of Mokgoro J.



NGCOBO J:


Introduction

[107]At the centre of this application for leave to appeal are the provisions of rule 4.2.1 read with the relevant definitions of the rules of the Political Office-Bearers Pension Fund (the Fund). The impugned rules provide for differentiated employer contributions in respect of members of Parliament. They treat members of Parliament who came to Parliament for the first time in 1994 (new members) more favourably than those who were members of Parliament prior to 1994 (old members). The respondent attacked these rules on the grounds that they discriminate unfairly against old members. The applicant resisted this attack on the grounds that the rules constituted a “limited affirmative action measure” in favour of new members of Parliament under section 9(2) of the Constitution. The High Court found that the impugned rules did not fall under section 9(2) and concluded that the impugned rules violate the equality clause of the Constitution. The main judgment finds that the rules fall under section 9(2) and are therefore within the constitutional bounds.

[108]The main judgment holds that for a measure to come under section 9(2) it must meet three requirements, namely, it must: (a) target persons or categories of persons who have been disadvantaged by unfair discrimination; (b) be designed to protect or advance such persons; and (c) promote the achievement of equality. With this, I agree. I doubt whether section 9(2) applies to the facts of this case. In particular, I doubt whether on the facts of this case the requirement that the measure must target persons or categories of persons who have been disadvantaged by discrimination has been met. The beneficiaries of the measure included persons who were not disadvantaged by past discrimination. This issue was not fully argued in this Court. However, in the view I take of the central question whether the impugned rules discriminate unfairly against the respondent, I consider it unnecessary to reach any firm conclusion in this regard.

[109]The fact that a remedial measure under constitutional challenge does not come under section 9(2) of the Constitution does not necessarily mean that it violates the equality clause. In National Coalition for Gay and Lesbian Equality and Another v Minister of Justice[74] this Court held that the principles underlying remedial equality do not operate only in the context of section 9(2). It follows therefore that the constitutional validity of the impugned rules must still be determined in light of the equality guarantee. The respondent contended that the impugned rules unfairly discriminate against him and those similarly situated and are therefore irrational. This contention must be considered in the light of the equality guarantee.


Equality Analysis

[110]The relevant provision of the Constitution in section 9 provides:
“(1) Everyone is equal before the law and has the right to equal protection and benefit of the law.
(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.
(3) The state may not unfairly discriminate directly or indirectly against anyone on one or more grounds, including race, gender, sex, pregnancy, marital status, ethnic or social origin, colour, sexual orientation, age, disability, religion, conscience, belief, culture, language and birth.
(4) No person may unfairly discriminate directly or indirectly against anyone on one or more grounds in terms of subsection (3). National legislation must be enacted to prevent or prohibit unfair discrimination.
(5) Discrimination on one or more of the grounds listed in subsection (3) is unfair unless it is established that the discrimination is fair.”
[111]The proper approach to the question whether the impugned rules violate the equality clause involves three basic enquiries: first, whether the impugned rules make a differentiation that bears a rational connection to a legitimate government purpose; and if so, second, whether the differentiation amounts to unfair discrimination; and if so, third, whether the impugned rules can be justified under the limitations provision. If the differentiation bears no such rational connection, there is a violation of section 9(1) and the second enquiry does not arise. Similarly, if the differentiation does not amount to unfair discrimination, the third enquiry does not arise.[75] It is to the first enquiry that I now turn.


Rationality of Differentiation

[112]It is common cause that the impugned rules differentiate between old and new members of Parliament in relation to parliamentary pension benefits. The need for differentiation arose because old members of Parliament were members of the Closed Pension Fund (CPF) and thus entitled to pension benefits from that fund. New members of Parliament were excluded from the CPF and thus were not entitled to any benefits under that fund. When the new fund was created after April 1994, old members of Parliament became entitled to benefits under the new fund. This resulted in the old members of Parliament being entitled to parliamentary benefits from two pension funds. The differentiation in contributions to be made in respect of different categories of members was designed to bring about equity in the spread of parliamentary pension benefits amongst old and new political office-bearers.

[113]The legitimacy of this purpose cannot be gainsaid. There was inequality in the entitlement to pension benefits in that old members of Parliament were entitled to benefits from a parliamentary pension fund from which new members were excluded. Nor can there be any doubt as to the existence of a rational connection between the differentiation created by the impugned rules and the legitimate governmental purpose. It follows therefore that the contention by the respondent that the differentiation was irrational must fail. The question which falls to be determined is whether this differentiation amounted to unfair discrimination.


Discrimination

[114]It was contended on behalf of the respondent that race was an important factor in the differentiation. There can be no question that the differentiation had a disproportionate impact on persons who were previously classified as white, coloured and Indian. These racial groups were the only racial groups that were eligible to be members of the tri-cameral parliament. It is also clear from the papers that one of the primary considerations in adopting the impugned rules was the fact that an overwhelming majority of those who were excluded from the CPF were excluded from the tri-cameral parliament because of race and political affiliation.

[115]In all the circumstances we are concerned here with a differentiation on a listed ground. But the rules are facially neutral as far as race and political affiliation is concerned. This finding raises a rebuttable presumption that the impugned rules indirectly discriminate unfairly against the respondent. The ultimate question, however, is whether in fact the impugned rules indirectly discriminate unfairly as contended by the respondent.


Do the impugned rules discriminate unfairly?

[116]At the heart of our equality guarantee is the prohibition of unfair discrimination and remedying the effects of past unfair discrimination. Human dignity is harmed by unfair treatment that is premised upon personal traits or circumstances that do not relate to the needs, capacities and merits of different individuals. Often such discrimination is premised on the assumption that the disfavoured group is not worthy of dignity. At times, as our history amply demonstrates, such discrimination proceeds on the assumption that the disfavoured group is inferior to other groups.[76] And this is an assault on the human dignity of the disfavoured group. Equality as enshrined in our Constitution does not tolerate distinctions that treat other 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”.[77]

[117]In President of the Republic of South Africa and Another v Hugo, this Court outlined the purpose of the equality clause, in particular, the prohibition of unfair discrimination and said:
“The prohibition on unfair discrimination in the interim Constitution seeks not only to avoid discrimination against people who are members of disadvantaged groups. It seeks more than that. At the heart of unfair discrimination lies recognition that the purpose of our new constitutional and democratic order is the establishment of a society in which all human beings will be accorded equal dignity and respect regardless of their membership of particular groups. The achievement of such a society in the context of our deeply inegalitarian past will not be easy, but that that is the goal of the Constitution should not be forgotten or overlooked.”[78]

[118]However, it is not every distinction or differentiation in treatment which falls foul of the equality guarantee. Legislatures, to govern effectively, may treat different individuals and groups in different ways. In Prinsloo v Van der Linde and Another,[79] this Court accepted that in order to govern a modern country efficiently and to harmonise the interests of all its people for common good, it may be necessary for government to make distinctions. Such distinctions will “very rarely” constitute unfair discrimination to such regulation, without the addition of a further element.[80]

[119]Our concept of equality therefore recognises that at times it may be necessary to treat people differently for example when it is necessary to recognise the different social or economic situations in which individuals are situated. This is a recognition of the fact that treating unequals as if they are equals may produce inequality. Our concept of unfair discrimination therefore recognises that:

...“[A]lthough a society which affords each human being equal treatment on the basis of equal worth and freedom is our goal, we cannot achieve that goal by insisting upon identical treatment in all circumstances before that goal is achieved. Each case, therefore, will require a careful and a thorough understanding of the impact of discriminatory action upon the particular people concerned to determine whether its overall impact is one which furthers the constitutional goal of equality or not. A classification which is unfair in one context may not necessarily be unfair in a different context.”[81]

[120]As noted previously,[82] it is also important to note that the principles of remedial equality do not only operate in the context of section 9(2) of the Constitution. This Court has recognised that they are relevant in deciding whether the discriminatory provisions have impacted unfairly on complainants.[83] Thus in Harksen when dealing with the purpose of the provision or power as a factor to be considered in deciding whether discrimination has impacted unfairly on the complainant, this Court held that:

“If its purpose is manifestly not directed, at the first instance, at impairing the complainants in the manner indicated above, but is aimed at achieving a worthy and important societal goal, such as, for example, the furthering of equality for all, this purpose may, depending on the facts of the particular case, have a significant bearing on the question whether complainants have in fact suffered the impairment in question.”[84]
[121]The question which falls to be determined therefore is the impact of discrimination on the respondent and those similarly situated. And in determining this question relevant considerations include the position of the respondent and those similarly situated in society, the purpose sought to be achieved by the discrimination, the extent to which the rights or the interests of the respondent have been affected and whether the discrimination has impaired the human dignity of the respondent.[85] It is to that enquiry that I now turn.


The position of the members of the affected group in society

[122]The majority of the group affected is not one which has suffered discrimination in the past.[86] Members of this group are all politicians, and have some political power. This group cannot, in my view, be said to be a vulnerable group. That in itself does not render the discrimination fair.


The nature and the purpose of the power exercised by Parliament

[123]In adopting the impugned rules, Parliament was fulfilling its constitutional obligation to create a pension fund for political office-bearers. Under the interim Constitution this obligation was imposed by section 190A.[87] Under the Constitution that obligation is imposed by section 219.[88] The purpose of the impugned rules is, broadly speaking, to give effect to this constitutional obligation.

[124]The purpose behind the impugned rules is given as follows by Mr Maritz, the Chief Director in the Directorate of Pensions Administration of the Department of Finance and Deputy-Chairperson of the Political Office-Bearers Pension Fund, the Third Respondent herein:
“15. The pension arrangements which applied in respect of political office bearers after the commencement of the 1983 tri-cameral Parliament were contained in the Members of Parliament and Political Office Bearers Pension Scheme Act 112 of 1984 (“the 1984 Act”). The pension scheme established in terms of the 1984 Act was a so called “pay as you go” scheme. This meant that no special fund was established for the payment of contributions. Rather, in terms of the 1984 Act, ordinary members of Parliament were required to pay 7% (seven percent) of their pensionable salary to the State Revenue Fund. When a member of Parliament retired, he or she became entitled to pension benefits in terms of the Act, and these benefits were paid out of the State Revenue Fund. No specific pension fund was established for purposes of payment of pensions in terms of the 1984 Act.
16. During the negotiations held in Kempton Park in the early 1990’s, for the establishment of a democratic government in South Africa and the determination of a democratic constitution, the question of the pensions of members of the previous Parliament was raised. It was agreed that a closed pension fund would be established, and the actuarial interest of every member and existing pensioner of the pension scheme established under the 1984 Act would be determined and paid into that fund by the State. Consequent upon this agreement, the Closed Pension Fund Act 197 of 1993 was passed, (“the Closed Pension Fund Act”) in terms of which the Closed Pension Fund was established. The total actuarial liability of that fund was about R773 700 000, 00 (SEVEN HUNDRED AND SEVENTY MILLION RAND) as at 1 February 1994. This liability which was funded to an amount of some R440 000 000, 00 (FOUR HUNDRED AND FORTY MILLION RAND) by the issuing of government stock, and the remaining obligation of some R333 700 000, 00 (THREE HUNDRED AND THIRTY THREE MILLION SEVEN HUNDRED THOUSAND RAND) by way of monies voted by Parliament under the budget vote of the Department of Finance. The latter amount was payed over several years together with interest of some R220 000 000, 00 (TWO HUNDRED AND TWENTY MILLION RANDS).
17. The Closed Pension Fund Act came into operation on 5 January 1994. As its name suggested, the membership of the Closed Pension Fund was closed from the inception of the Fund – it was limited to persons who, in their capacity as political office bearers of the South African state prior to the interim Constitution, received pensions, or were entitled to pension benefits in that capacity. Persons who were not already members of the Closed Pension Fund, and who were elected to Parliament or the provincial legislatures in the first democratic elections in April 1994, did not thereby become members of the Closed Pension Fund.
18. In addition to being a closed fund, the Closed Pension Fund was an extremely generous fund. I attach marked “PM1” an affidavit of ERICH POTGIETER, the actuary of the Third Respondent. In annexure PM1 POTGIETER analyses the benefits provided by the Closed Pension Fund and shows that they were more than 2.5 times as generous as those provided by the Third Respondent to Category B members (the most privileged class of members of the Third Respondent), and just under 4 times as generous as the benefits provided by the typical defined benefit pension funds operating in the private sector.
19. Because the Closed Pension Fund was closed, after the first democratic elections in April 1994, it became necessary for a new pension dispensation to be established for members of Parliament and other political office bearers. The creation of a new pension fund for political office bearers (“the new fund”) was, in fact, a constitutional obligation imposed by section 190A of the interim Constitution.”

[125]And the rationale for the differentiation is given as follows by Mr Maritz:
“27 The rationale for the differentiation is the following:
27.1 With more pressing calls on the public purse and the expansion of Parliament and the creation of the provincial legislatures after the 1994 elections it was not affordable to create a pension scheme providing political office bearers with benefits as generous as those provided under the 1984 Act and the Closed Pension Fund.
27.2 The limited resources available for the pensions of political office bearers had to be spread in an equitable fashion.
27.3 Members of the Closed Pension Fund were already in receipt of generous pension benefits which were far in excess of those available to new political office bearers.
27.4 The overwhelming majority of new political office bearers had been excluded from access to political office under the tri-cameral regime (and thereby from access to the generous benefits of the Closed Pension Fund) by virtue of either their race or their political affiliation or both their race and their political affiliation.
27.5 In this context, the differentiated scheme of employer contributions under the Rules of the Third Respondent was designed to benefit new political office bearers whose exclusion from the benefits of the Closed Pension Fund by virtue of historical circumstances left them with a need for more generous pension benefits than their colleagues who had access to Closed Pension Fund benefits.
27.6 Within the class of new political office bearers, the differentiated scheme also conferred additional benefits on office bearers who were over the age of 50 and whose advanced age accordingly increased their immediate need for more generous pension benefits.
27.7 Consistent with its origins in a particular transitional historical context, the differentiation effected by the scheme was a limited affirmative action measure which operated only for the first five years of the democratic era. From 1 May 1999 there was to be a uniform employer contribution of 17% in respect of all members of the Third Respondent.”

[126]From what Mr Maritz says, it is clear that as at April 1994 members of Parliament and other political office-bearers who held office prior to 1994 enjoyed extremely generous pension benefits under the CPF. The CPF was fully funded by public funds. This fund was especially reserved for the benefit of this group. Persons elected to Parliament for the first time in 1994 were excluded from this fund. It was, as its name suggests, a closed fund. But for that exclusion, new members would have been entitled to join the same fund and benefit from its generous provisions. After the first democratic elections it became necessary to establish a new pension scheme for members of Parliament and other political office-bearers. Parliament was under a constitutional duty to do so.

[127]But the reality was that old members of Parliament already had a generous, publicly funded pension scheme. This had to be kept in mind when creating a new pension scheme. Old members of Parliament could not be excluded from the new pension f